Advances in

Carbohydrate Chemistry and Biochemistry

Volume 35
This Page Intentionally Left Blank
 Advances in
Carbohydrate Chemistry
and
Biochemistry

Editors
R. STUART TIPSON
DEREK HORTON

Board of Advisors
LAURENSANDERSON BENCT LINDBERC
STEPHENJ. ANGYAL HANS PAULSEN
GUY G. S. DUTTON NATHANSHARON
ALLAN B. FOSTER MAURICESTACEY
DEXTER FRENCH ROY L. WHISTLER

Volume 35
1978

ACADEMIC PRESS New York San Francisco London

A Subsidiary of Harcourt Brace Jovanovich, Publishers
COPYRIGHT @ 1978, BY ACADEMIC PRESS, INC.
ALL RIGHTS RESERVED.
NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR
TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONlC
OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY
INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT
PERMISSlON IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by

ACADEMIC PRESS, INC. (LONDON)LTD.
24/28 Oval Road, London N W I 7DX

LIBRARY
OF CONGRESS CATALOG CARD NUMBER:
45-11351

ISBN 0-12-007235-1

PRINTED IN THE UNITED STATES OF AMERICA

 CONTENTS

LIST OF CONTRIBUTORS. . . . . . . . . . . . . . ............ vii

PREFACE . . . . . . . . . . . . . . . . . . . . . ............ ix

Edmund Langley Hirst (1898-1975)

MAURICESTACEY AND DAVIDJ . MANNERS
Text ................................... 1

Carbohydrate Boronates
ROBERTJ . FERRIER

I . Introduction ............................ 31
.
I1 Synthesis of Boronates ....................... 37
I11. Structures of Carbohydrate Boronates .............. .. 41
.
IV Boronates in Chemical Reactions . . . . . . . . . . . . . . . . . . 48
.
V Separations of Carbohydrates by Use of Their Boronates . . . . . . . 57
.
VI Mass Spectrometry of Boronates . . . . . . . . . . . . . . . . . . . 65
VII . Nuclear Magnetic Resonance Spectroscopy ofBoronates . . . . . . . . 70
VIII . Borinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
1X.Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
X.Addendum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Biosynthesis of Sugar Components of Antibiotic Substances

HANS GRISEBACH

I . Introduction ..................... ....... 81

I1. Branched-chain Sugars ................ ....... 82
I11. Aminocyclitol Antibiotics ............... ....... 102
.
IV Amino Sugars Not Occurring in Aminocyclitol Antibiotics . . . . . . . 122
.
V Nucleoside Antibiotics ................ ....... 122

The Lectins: Carbohydrate-binding Proteins of Plants and Animals

IRWIN J . GOLDSTEINAND COLLEENE . HAYES

I . Introduction ............................ 128

I1. D-Mannose(D-Glucose)-binding Lectins . . . . . . . . . . . . . . . . 150
I11. 2-Acetamido-2-deoxy-~-glucose-binding Lectins . . . . . . . . . . . . 206
IV . 2-Acetamido-2-deoxy-Dgalactose-binding Lectins . . . . . . . . . . . 226
V . D-Galactose-binding Lectins . . . . . . . . . . . . . . . . . . . . . 254
VI . L-Fucose-binding Lectins ...................... 277
VII . Other Lectins . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
VIII . Cell-surface. Lectin-reactive Glycoproteins . . . . . . . . . . . . . . 317
IX.Tables ............................... 334

V
vi CONTENTS

Biochemistry of Plant Galactomannans

PRAKASH M . DEY

.
I Introduction ...... ...................... 341
.
I1 Biosynthesis ...... ...................... 352
I11. Biochemical Degradation ...................... 356
.
IV Function . . . . . . . . ...................... 375

Bibliography of Crystal Structures

of Polysaccharides. 1975
PUDUPADI R. SUNDARARAJAN AND ROBERT H . MARCHESSAULT

.
I Introduction ................ ............ 377
I1. Amylose and Other a-D-Glycans . . . . . . . . . . . . . . . . . . . 378
.
I11 Cellulose and Other p-D-Glycans ..... . . . . . . . . . . . . . 379
IV. Glycosaminoglycans (Amino Polysaccharides) . . . . . . . . . . . . . 381
AUTHOR INDEX FOR VOLUME 35 . . . . . . . . . . . . . . . . . . . . . . 387
SUBJECTINDEXFOR VOLUME35 . . . . . . . . . . . . . . . . . . . . . . 415
CUMULATIVE AUTHOR INDEX FOR VOLUMES 31-35 . . . . . . . . . . . . . 429
CUMULATIVE SUBJECT INDEX FOR VOLUMES 31-35 . . . . . . . . . . . . . 431
ERRATUM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
 LIST OF CONTRIBUTORS
Numbers in parentheses indicate the pages on which the authors’ contributions begin.

PRAKASH M. DEY,Department of Biochemistry, Royal Holloway College, University

of London, Egham Hill, Egham, Surrey TW20 OEX, England (341)
ROBERT J. FERRIER,Department of Chemistry, Victoria University of Wellington,
Private Bag, Wellington, New Zealand (31)
IRWIN J. GOLDSTEIN,Department of Biological Chemistry, The university of Mich-
igan, Ann Arbor, Michigan 48109 (127)
HANS GRISEBACH,Biologisches Znstitut 11, Biochemie der Pjlanzen, Universitat Frei-
burg i. Br., D 7800 Freiburg i m Breisgau, Germany (81)
COLLEEN E. HAYES, Immunobiology Research Center, University of Wisconsin,
Madison, Wisconsin 53706 (127)
DAVIDJ. MANNERS,Department of Brewing and Biological Sciences, Heriot-Watt
University, Chambers Street, Edinburgh EHl I H X , Scotland (1)
ROBERTH . MARCHESSAULT,Department of Chemistry, University of Montreal, P. 0.
Box 6210, Succursale A , Montreal, Quebec H3C 3V1, Canada (377)
MAURICE STACEY,Department of Chemistry, The University of Birmingham, P. 0.
Box 363, Birmingham B15 ZTT,EngZand (1)
PUDUPADI R. SUNDARARAJAN, Xerox Research Centre of Canada, 2480 Dunwin Drive,
Mississauga, Ontario L5L 119, Canada (377)

vii
This Page Intentionally Left Blank
 PREFACE

In 1961, Ferrier (Wellington, N. Z.) began a study of the condensa-

tion of phenylboronic acid with the diol groupings of various glyco-
sides, and since then, as a result of research by him and many other
workers, the subject has developed rapidly and has afforded informa-
tion that is of great potential value to synthetic carbohydrate chemists;
Ferrier now provides us with a fascinating account of the progress
made to date in the study of carbohydrate boronates. This unifying
picture of the use of these cyclic, protected sugar derivatives should
afford considerable help to the chemist searching for novel approaches
in synthesis by use of these versatile protecting groups.
In an article that focuses particularly on some of the more-recent
developments, Grisebach (Freiburg im Breisgau) discusses the bio-
synthesis of sugar components of antibiotic substances, a field that has
shown major advances since the time of the article by Dutcher in
Volume 18 of Advances. The chapter constitutes an integrating com-
plement to the articles on aminocyclitol antibiotics by the Umezawas
(Volume 30).
Goldstein (Ann Arbor, Michigan) and Hayes (Madison) contribute
a monumental chapter on lectins, the specific carbohydrate-binding
proteins present in both plant and animal species. The remarkable
specificity of certain carbohydrate-protein interactions has far-
reaching implications in biochemistry, the full significance of which is
only just beginning to be properly understood. As this subject has not
previously been treated in depth in Aduances, these authors have
written a comprehensive history of the subject, starting with Stillmark’s
discovery of plant agglutinins in 1888, and proceeding to 1977; this
article brings together in one place an enormous amount of informa-
tion scattered throughout the literature, and should constitute the
definitive treatment of lectins for many years to come.
Another biochemical topic, the biochemistry of plant galacto-
mannans, is discussed by Dey (Egham, Surrey); the article rounds out
aspects of the field that are complementary to those treated by Gorin
and Spencer in Volume 23 (on fungal polysaccharides) and Dea and
Morrison in Volume 31 (on the chemistry and interactions of seed
galactomannans).
In a continuation of the series of bibliographic articles on the struc-
tures of polysaccharides as established by X-ray crystallographic
methods, Sundararajan (Mississauga, Ontario) and Marchessault

ix
X PREFACE

(Montreal) present the information recorded in the literature during

1975, thus updating their article in Volume 33. Since the latter was
written, SI units have become generally adopted; the time-honored
Angstrom unit (A) has now fallen into disuse, and so it will no longer
be employed in this Series.
The death of our friend and mentor Sir Edmund Hirst was briefly
noted in the Preface to Volume 32. A full account of his career and
wide-ranging achievements is given here by Stacey (Birmingham)
and Manners (Edinburgh).
The Subject Index was compiled by Dr. L. T. Capell.
Kensington, Maryland R. STUART TIPSON
Columbus, Ohio DEREK HORTON
February, 1978
 Advances in
Carbohydrate Chemistry and Biochemistry

Volume 35
z<*-%K.
1898 - 1975
 EDMUND LANGLEY HIRST*

1898- 1975

Edmund Langley Hirst was the elder son of the Reverend Sim Hirst,
a Baptist minister, and Elizabeth Hirst (nke Langley). He was born in
Preston, Lancashire, on 21st July, 1898, and he had a somewhat unset-
tled childhood, due to the many changes in his father’s ministry. His
father’s family had been long established in the town of Clayton, near
Bradford, Yorkshire, where his grandfather and several uncles worked
in the woollen mills, and where other uncles were shopkeepers. All the
family were stout Nonconformists and loyal to the local Baptist church.
The Reverend Sim was born in Clayton in 1856, the son of John Hirst, a
weaver, and Martha Hirst. He graduated B.A. from the University of
Durham, was trained for the ministry at Rawdon College near Leeds,
Yorkshire, and later gained the Bachelor of Divinity degree of St.
Andrews University. He was for various periods Minister at churches
in Stoke-on-Trent, St. Andrews (for three different periods), Durham,
Preston, Burnley, and Ipswich; he died suddenly in 1923. Hirst’s
mother (Elizabeth), to whom Edmund was greatly attached, was born
in 1869, the daughter of Joseph and Mary Langley of Liverpool, where
Joseph was a successful flour merchant and baker, and she was edu-
cated at private schools. She was of mixed Welsh and North Country
stock, and her family had farmed land in the English Lake District for
generations. They had married in 1897 and she, living to the age of 86,
survived her husband by 32 years. The marriage was particularly
happy, for, although she belonged to the Church ofEngland, she was an
understanding and loyal Baptist Minister’s wife. She made sure that
Edmund and his younger brother Sim had a sound schooling, despite
the many changes of homes.
Hirst first attended a kindergarten school in Burnley, and then had
private tutoring in other towns, and a governess in Ipswich where, at
the age of 8, he attended Ipswich Municipal Secondary School. The
following year, he suffered a sharp set-back due to a severe mastoid
illness. In 1910, he was more settled in St. Andrews, where after a
considerable personal effort, he succeeded in gaining entrance to the
* The authors are indebted to Lady Hirst for her advice during the preparation of this
article, and to Walter Bird for the photograph.

1
2 MAURICE STACEY A N D DAVID J. MANNERS

famous Madras College in that ancient city. He received a sound

grounding in Greek and mathematics, and finished school at the age of
16, as Dux (Head Boy) of the School, winning entrance to St. Andrews
University with a i40-a-year bursary and a Carnegie Scholarship.
In his first year, he studied mainly mathematics, Greek, and Latin, at
which he did so well that he almost changed over to take a degree in
Classics. It was the advent ofworld War I that turned him to Chemistry.
In the summer vacation of 1915,he volunteered to assist the war effort
by joining a group in the Chemistry Department led by Professor (later
Principal Sir James) Irvine and Dr. (later Sir Norman) Haworth, who
were making such chemicals as local anesthetics, previously obtained
from Germany. He was able to continue this work during the following
summer, when he gained excellent experience in preparative chemis-
try and had his first taste of carbohydrate chemistry by the preparation
of D-galactose and galactitol (dulcitol) from lactose.
In 1917, he was called up for military service, but was immediately
seconded back to the University for urgent work on the large-scale
preparation of mustard gas. In 1918, he volunteered for the Special
Branch of the Royal Engineers, and was despatched to Northern
France. The following February, he was demobilized and was able to
return to complete his studies. His performance in examinations was
brilliant, for he gained his Bachelor’s degree with distinction in
chemistry, mathematics, and natural philosophy, and was awarded
First Class Honours in mathematics and natural philosophy in his
Master’s degree. He gained many medals during his degree courses.
Following discussions with Haworth, he decided to join him in order
to follow a research career in carbohydrate chemistry. He was awarded
a Carnegie Teaching Fellowship to study the structure of cellobiose.
He was immediately successful and gained his Ph.D. in 1921, after only
two years of research. When Haworth moved to a Chair at Armstrong
College, Newcastle (part of the University of Durham), Hirst continued
to work on cellulose in collaboration with Sir James Irvine, and he also
began independently his important studies on the ring structures of
sugars, particularly xylose. At this time, Irvine became Principal of St.
Andrews University, and Robert Robinson was appointed Professor of
Chemistry. Robinson’s stay, however, was brief, for he moved to the
Manchester Chair in 1922, and in the following year, Robinson invited
Hirst to join his staff. From 1924 to 1926, Hirst was a lecturer at Arm-
strong College, Newcastle, first with Haworth and then with H. V. A.
Briscoe. Haworth moved to Birmingham in 1925, and was joined by
Hirst in 1927, when their famous partnership really began.
There was already in Birmingham a strong staff, a number of whom,
including H. D. K. Drew, C. E. Wood, C. R. Porter, S. R. Carter, and E.
 OBITUARY-EDMUND LANGLEY HIRST 3

G. Cox, were brought into the carbohydrate team to join the research
students, including Stanley Peat, whom Haworth had brought from
Newcastle. Hirst was appointed Lecturer and Assistant Director of Re-
search, and also undertook a heavy teaching load. He was, however,
spared administrative and University committee work. His years at
Birmingham were intensive and highly productive of important results
in the carbohydrate field, including the vitamin C work, as will be
described later. He was awarded his Birmingham D.Sc. in 1929, and
was made Reader in the Chemistry of Natural Products in 1934,the year
he was elected a Fellow ofthe Royal Society. In 1936, he was appointed
to the Alfred Capper Pass Chair of Organic Chemistry in the University
of Bristol, in succession to Professor F. Francis. He took with him to
Bristol Dr. (later Professor) J. K. N. Jones and Dr. G. T. Young, who
assisted in the direction of his research groups. With Bristol graduates,
they quickly built up a vigorous research team working on numerous
carbohydrate problems, which included starches, pectic substances,
and certain plant gums. This work was suspended early in 1939, for the
Bristol Laboratories, of which Professor W. E. Garner was the Head,
were turned over for urgent Government work on explosives. Garner
and Hirst and their staff supervised a group of workers from the Re-
search Branch of Woolwich Arsenal. Hirst, already heavily involved in
Civil Defence work as Senior Gas Adviser to the South West Region,
engaged in the work with his usual vigor. He regularly visited various
ordnance factories concerned with manufacturing and filling. He was a
member of several committees of the Scientific Advisory Council of the
Ministry of Supply and of other Committees on explosives. J. K. N.
Jones was the senior assistant for this work, which now occupied 90%of
Hirst’s time. Important help, especially for undergraduate teaching,
came from Professor G. M. Bennett and his staff, who were evacuated
from King’s College, London, to Bristol.
In 1944, Hirst was appointed to succeed Alexander Todd (now Lord
Todd, P.R.S.) in the Sir Samuel Hall Chair of Organic Chemistry at
Manchester University. He went there in January, 1945, taking with
him J. K. N. Jones, as Senior Lecturer, and some members ofthe Bristol
team, to complete the explosives researches. He immediately became
involved with Professor Michael Polanyi in much reorganization and in
planning new laboratories to cope with the large intake of ex-Service
students. H e was also in great demand for public service, especially as
Chairman of the Research Section of a Working Party reporting on the
state ofthe cotton industry of Lancashire. He also gave great help to the
Shirley Institute for Cotton Research as a member of their research
committees.
In 1947, he was delighted to accept an invitation to return to his
4 MAURICE STACEY AND DAVID J. MANNERS

beloved Scotland. This came from Professor James Kendall of the

University of Edinburgh, who had created the Forbes Chair of Organic
Chemistry, which Hirst moved to occupy in that year. He remained in
Edinburgh for the rest of his life, making during the 21 years of his
Professorship a great impact on that University, on Scottish higher
education, and on the applications of science to Scottish industry and
agriculture.
On the personal side, Hirst was always a charming companion, and a
valued friend to many people, particularly his former students. He was
a good listener and always had sound advice to give. His early domestic
life in Birmingham had not been easy. In 1925, he married Beda
Winifred Ramsay, daughter of Frank and Mrs. Ramsay of Glasgow, who
were friends of the family. Early in their married life, his wife de-
veloped a mental illness which necessitated continuous hospital care
until her death. The marriage was dissolved in 1948, and, in 1949, he
married Kathleen (Kay) Jennie Harrison. Her father was a well known
Birmingham headmaster, and her two brothers were chemistry
graduates of Birmingham University. Kay was one of His Majesty’s
Inspectors of Schools, and later became well known for her charitable
interests. This marriage was ideally happy, and Kay gave Edmund the
fullest support in all his endeavors. She took part, as a Committee
member, in the many activities of the University Ladies Club and the
Women Students’ Union. She was involved in work for crippled chil-
dren, the Queensberry House Hospital for old people, the Edinburgh
College of Domestic Science (now Queen Margaret College), and the
Moray House College of Education, and was on the Council, and
Chairman for two years, of an Edinburgh Training School for nurses.
Edmund was immensely proud of Kay, who understood him so well,
and together they made a fine team.
Hirst’s first personal research venture into the structural carbohy-
drate field showed remarkable insight. He selected xylose for investi-
gation, and discovered that mild oxidation of the methylated sugar with
bromine-water formed the methylated lactone which, on stronger oxi-
dation with nitric acid, gave a number of methylated dibasic acids, the
identification of which could be used to decide the size of the ring in the
parent sugar. With A. Carruthers in 1922, and C. B. Purves in 1923, he
showed that, on oxidation with bromine-water, crystalline tri-0-
methyl-D-XylOSe gave 2,3,4-tri~-methy~-D-xy~ono~actone which, on
further oxidation with nitric acid, yielded “i-xylo-trimethoxyglutaric
acid” (tri-0-methylxylaric acid). This proved that the original xylose
possessed a six-membered ring.
Before going to Birmingham he made, with a succession of research
 OBITUARY-EDMUND LANGLEY HIRST 5

students, important studies on methylation, on the structure of cel-

lulose, on other monosaccharides, and on raffinose. The methylation-
oxidation techniques were applied systematically and with great vigor
and success in Birmingham during the ten years of the Haworth-Hirst
partnership. The ring-structures of most of the sugars known at that
time, mono-, di-, tri-, oligo-, and poly-saccharides, including the
difficult problem presented by sucrose, were elucidated. A number of
innovations speeded the work considerably, one of the most important
being the introduction into Britain by Drew and Haworth, in 1925, of
the Kuhlmann balance and Pregl’s methods of microanalysis. Another
was the appreciation that physicochemical methods could be used to
solve structural problems; such tools as U.V. absorption spectra, optical
rotatory dispersion, and X-ray crystallographic analysis, introduced by
E. G. Cox (now Sir Gordon), were pioneered. The research workers
studying the relationship between carbohydrate structure and optical
rotatory power included R. S. Tipson. The Haworth-Hirst partnership,
aided by a succession of able and enthusiastic research students, in-
cluding many from overseas, resulted in over one hundred publications
and the establishment of the Haworth system of furanose and pyranose
nomenclature and of perspective and conformational depictions of
sugars.
Both Haworth and Hirst believed in team-work, and this was seen to
advantage in the research that settled the notable controversy with C. S.
Hudson over the structure of D-mannose. This was concluded at a
famous meeting with Haworth, Hudson, and Purves in 1930, which was
fully described by Hirst in The Hudson Memorial Lecture [J. Chem.
SOC., 4042-4058 (1954)l. Hirst always had great admiration for Hud-
son’s researches on the optical properties of carbohydrates, and con-
ducted extensive personal experimental work (with C. E. Wood) on
optical rotatory dispersion.
Numerous other fascinating projects were in progress at this time,
notably studies on novel forms of stereochemistry, the beginnings of
conformational analysis, crystalline sugar carbonates, furanose forms of
sugars, and structural studies on polysaccharides, including glycogens
and numerous mold and bacterial polysaccharides. A major step for-
ward in polysaccharide chemistry was the demonstration of cellobiose
residues in cellulose. From these investigations and the numerous
methylation studies, a valuable collection of reference compounds was
synthesized and generously supplied to workers all over the world.
All of this work was set aside in 1932, when Haworth formed most of
the carbohydrate workers into a team (or “syndicate”) to investigate the
structure of Szent-Gyorgyi’s “hexuronic acid,” which proved to b e vita-
6 MAURICE STACEY A N D DAVID J. MANNERS

min C, and was later named ascorbic acid. These were exciting times,
and the rapid and successful outcome of the work was due in large
measure to Hirst’s remarkable intuition and his real inspiration to the
team. Consolidation of the ascorbic acid work was achieved in close
collaboration with F. Smith and J. K. N. Jones who, during the next few
years, produced a fine series of papers on analogs of ascorbic acid. A
number of pioneer studies pointed the direction to future work in
Bristol, Manchester, and Edinburgh. Among these were researches on
the carbohydrates of grasses and barley leaves, and on hexuronic acids,
particularly the “aldobionic acid” from gum arabic on which F. Smith
was working with such vigor. Inspiration came from a visit in 1936 to
England by Michael Heidelberger, who impressed on Hirst the close
relationship between gums and immunopolysaccharides.
The move to Bristol, and thence to Manchester, as already indicated,
brought Hirst additional duties away from chemistry, and, without the
devoted help given by J. K. N. Jones at both Manchester and Bristol,
and by G. T. Young at Bristol and T. G. Halsall at Manchester, his
contributions to carbohydrate chemistry would have been lessened. It
is a remarkable fact that the Bristol and Manchester periods resulted in
over 70 papers in which the constitutions of a whole range of plant
gums, pectic substances, and mucilages were determined. Amongst
these were damson gum, slippery-elm mucilage, and various arabinans
and galactomannans.
During this period, chromatographic methods for the separation of
sugars were being developed, and, with J. K. N. Jones and also Leslie
(now Professor) Hough, Hirst played a full part in this work. Methods
for the separation and identification of sugars, especially methylated
sugars, by paper chromatography were devised, and in a series of three
papers, they described standardized methods for their quantitative
determination. Their methods were reported in a Discussion of the
Faraday Society in 1949, and their application to the analysis of Ster-
culia setigera gum was published in the same year. These methods, and
subsequent improvements, were later widely used in the Edinburgh
laboratories.
When Hirst returned to Scotland in 1947, he knew that the Edin-
burgh laboratories already contained a very active research school in
carbohydrate chemistry that had been built up by E. G. V. Percival,
whose work was mainly concerned with the structure of plant muci-
lages and seaweed polysaccharides. These interests were complemen-
tary to Hirst’s own recent studies in Manchester on the structure of
polysaccharides, particularly the plant gums, and the next few years
saw a renewal of the successful collaboration of the Birmingham period
1930-1933, until the untimely death of Percival in 1951.
 OBITUARY-E DMU N D LANGLEY HI RST 7

Several papers resulted from this collaboration, some of which repre-

sented the first stages ofcontinuing investigations later completed with
other members of the staff, particularly G. 0. Aspinall and Elizabeth
Percival. With S. K. Chanda, structural analyses of the xylan and an
arabinose-rich hemicellulose from esparto grass, and of a xylan from
pear cell-walls, were carried out. Preliminary studies on the overall
molecular structure of the two mannans from ivory nuts, the reserve
p-D-glucan from Laminaria cloustoni, the alginic acid from Laminaria
digitata, and the seed mucilage from Plantago arenaria were reported.
During this period, two laboratories in Britain were particularly in-
terested in polymers of D-fructose. D. J. Bell in Cambridge had made
progress with various samples of inulin and plant levans [see Adv.
Carbohydr. Chem. Biochem., 30,l-8 (1974)],whilst, in Edinburgh, his
personal friends Hirst and Percival reported similar results with inulin
from dahlia tubers, and with a fructan from Dactylis glomerata. Both
laboratories confirmed the presence of chains of (2+1)- or (2+6)-linked
p-D-fmctofuranosyl residues, which were terminated at the potential
reducing end by a sucrose residue. This latter structural feature had not
been revealed in earlier investigations, and had an important bearing
on the biosynthesis of the D-fructans from sucrose by transfructosyl-
ation reactions.
During the early 1950s, Hirst appointed several new members of staff
having interests in certain specialized areas of carbohydrate chemistry.
These included D. M. W. Anderson, G . 0. Aspinall, C. T. Greenwood,
D. J. Manners, Elizabeth Percival, D. A. Rees, and J. C. P. Schwarz
who, with their respective research students, and visiting post-doctoral
fellows from many parts of the world, together helped to constitute a
large and successful research school. Collectively, substantial progress
was made, not only in most branches of polysaccharide chemistry but
also with many topics in monosaccharide chemistry. The carbohydrate
staff held regular meetings to discuss the progress oftheir work, and the
significance of papers in the current literature, especially those from
the laboratories of Hirst’s former colleagues, Professors Maurice Stacey
and Stanley Peat at Birmingham and Bangor, respectively. There were
times when the monthly issues of the Journal of the Chemical Society
were renamed the “Book(s) of Revelation.” For many of us, these were
days full of incident and interest, before the advent of the current
spectrometric techniques, when research workers still prepared crys-
talline derivatives of monosaccharides, and when the successful iden-
tification of these derivatives by the measurement of appropriate phys-
ical constants gave immense satisfaction to research student and super-
visor alike. We recall the interest and excitement generated by the
isolation of 6-0-methyl-~-galactosefrom a native polysaccharide, the
8 MAURICE STACEY AND D,4VID J. MANNERS

synthesis by Peter Schwarz OfD-xylOSe containing sulfur in the ring, the

confirmation by Elizabeth Percival of the presence of the hitherto rare
L-guluronic acid as a constituent of alginic acid, and, later, the first
applications by Dai Rees of conformational analysis to polysaccharide
structure and functions. During those days, there was a constant flow of
visitors from other carbohydrate laboratories, particularly external
examiners for the numerous Ph.D students.
The other noticeable product of this era was a steady stream of
publications, mainly to theJourna1 of the Chemical Society, but also to
the Biochemical Journal, to Chemistry and Industry, and later, to Car-
bohydrate Research. The atmosphere of this period is conveyed in
several review lectures given by Hirst; for example, the fourteenth
Pedler lecture to the Chemical Society (1955) on “Some Problems in
the Chemistry of the Hemicelluloses,” the Presidential Addresses to
the Chemical Society (1957 and 1958) on “Some Aspects of the Chemis-
try of the Fructosans” and “Polysaccharides of the Marine Algae,”
respectively, on “Plant Gums,” at the IVth International Congress of
Biochemistry, Vienna (1958), and the Bakerian Lecture to the Royal
Society (1959) on “Molecular Structure in the Polysaccharide Group.”
These lectures were delivered with a quiet authority, and the pub-
lished manuscripts show meticulous attention to detail.
In the structural analysis of polysaccharides, the importance of
methylation, introduced by Purdie and Irvine, and developed by
Haworth, is widely accepted, and was naturally fully utilized by Hirst
and his many collaborators. Perhaps less widely known is his develop-
ment of methods based on periodate oxidation for the analysis of a wide
range of polysaccharides. With F. Brown, T. G. Halsall, and J. K. N.
Jones, the first end-group assays of starches and glycogens were carried
out. The experimental conditions for the analysis of starches were later
refined, with Anderson and Greenwood. In these days of micro-scale
methylation analyses using g.1.c.-m.s. techniques, the real advance
represented by the periodate oxidation methods has tended to be over-
looked. The latter methods lessened the amount of material required,
from the gram scale to the decigram or even milligram scale, and the
overall time of analysis from weeks to days. Periodate oxidation was
also used to examine the nature of the inter-chain linkages in starches
and related polysaccharides. The absence of periodate-resistant resi-
dues was evidence for the presence of glucosyl residues triply linked at
C-1, C-4, and C-6, whilst the presence of some periodate-resistant
residues could, assuming that the oxidation was complete, indicate the
presence of other types of linkage. The fine structure of the (1-*3)-p-~-
glucan from Laminaria cloustoni was examined by other periodate-
 OBITUARY-EDMUND LANGLEY HIRST 9

oxidation methods, involving estimation of the production of formalde-

hyde from the borohydride-reduced glucan, and hence, giving the
degree of polymerization, and b y the application of the Barry and Smith
methods of degradation. These latter provided evidence for the pres-
ence of ( 1 ~ 6 ) - ~ - g l u c o s i dinter-chain
ic linkages rather than (1+6)-
inter-residue linkages, as had been suggested by other workers.
The complex nature of the molecular architecture of the plant gums
and mucilages had long been a major challenge to Hirst. Throughout
the Edinburgh period, first with Aspinall, and later with Anderson, he
published more papers on this than any other group of polysaccharides,
the last appearingin 1968. Systematic studies on various exudate gums,
involving determination of the composition, partial acid hydrolysis,
methylation, and periodate oxidation studies (including multiple
Smith degradations) led to the recognition of distinct groups of gums.
The largest group contain a core of D-galactopyranosylresidues united
by (1+3)- and (l+6)-linkages, to which are attached various propor-
tions of D-glucuronic acid, 4-O-methy~-D-glucuronicacid, L-arabino-
furanose, and L-rhamnopyranosyl residues. The central, galactan
core showed multiple branching. Examples include gum arabic
(Acacia senegal gum), and the gums fromA. pycnantha andA. arabica;
Dr. (now Professor) A. S. Perlin was amongst those involved in the early
studies of this group, as were two visitors from South Africa, Dr. A. J.
Charlson and Dr. (now Professor) A. M. Stephen.
A second group of gums contains a main backbone of alternating,
4-0-substituted D-glucosyluronic acid and 2-0-substituted D-mannOSyl
residues, to which side chains containing various D-galaCtOSe,
L-arabinopyranose, and D-glucuronic acid residues are attached. Gum
ghatti (Indian gum), examined with Dr. (now Professor) A. Wickstrdm
and (Mrs.) B. J. Auret, is a good example of this type.
Gums of the third group contain interior residues of D-galacturonic
acid and L-rhamnose, to which various side chains are attached, and
include Khaya grandifolia gum and certain Sterculia gums, which
show characteristic differences in fine structure from the Khaya gums.
In many cases, the overall complexity ofthe gums was so great that only
partial structures could be proposed. Full details of much of this work
have been given by Aspinall [Adv. Carbohydr. Chem. Biochem., 24,
333-379 (1969)l.
A second major interest during the Edinburgh period was the nature
of the reserve and structural polysaccharides of the marine algae. This
work paralleled the establishment ofthe Institute of Seaweed Research
at Musselburgh, near Edinburgh. Hirst was a member of the Board of
Governors, and Chairman of the Research Committee. A full exchange
10 MAURICE STACEY AND DAVID J. MANNERS

of information, samples, and materials operated to the mutual benefit of

all concerned. Hirst’s first paper on alginic acid was published in 1939,
and this, and later work in 1952, had established the presence of
(1+4)-linked D-mannuronic acid residues. However, Fischer and Dor-
fell obtained evidence for the presence of L-guluronic acid residues,
and this was confirmed by the isolation of L-threaric acid from a hy-
drolyzate of alginic acid after oxidation with periodate and then with
bromine. Subsequent examination of reduced alginic acid gave a par-
tial, acid hydrolyzate containing L-gulose, 4-O-~-D-mannosy~-L-gu~ose,
and 4-0-p-D-mannosyl-D-mannose, thus showing the existence of both
kinds of uronic acid residues in the same chain.
Chemical and enzymic studies were carried out on the reserve p-D-
glucans from various species oflaminaria, and from a mixed culture of
diatoms. Although they all contained chains of (l+3)-linked resi-
dues, some of the algal samples contained 2-3% of mannitol residues
which terminated a proportion of the chains; certain samples also
showed a low degree of branching involving C-6. Other seaweed
polysaccharides that were examined included those from Cladophora
rupestris and Chaetomorpha spp.; these proved to be complex sul-
fated heteropolysaccharides containing residues of D-galactOSe,
L-arabinose, and D-XylOSe. Floridean starch from the red seaweed
Dilsea edulis was also characterized; it was found to resemble the
amylopectin component of starch in some, but not all, respects.
Hirst had a wide interest in the plant kingdom, as an examination of
his personal library has shown, and his researches included studies of
polysaccharides from a large number of different plants. With Aspinall,
the hemicellulose A of beechwood and the egalactan from larch were
investigated. With Rees, he surveyed the polysaccharide components
of mustard seeds, with special reference to their embryos and their role
in germination. The polysaccharide components of certain mosses
were examined. With N. B. Chanda, the constitutions of the a-D-glucan
and P-D-gluCan from Iceland moss (Cetraria islandica) were estab-
lished; both are linear polysaccharides containing various proportions
of both (1+3)- and (1*4)-linkages. By contrast, reindeer moss
(Cladonia alpestris) synthesizes a highly branched galactoglucoman-
nan.
The constitution of starch had been a major research interest since
the Birmingham days, and Hirst continued this work in Edinburgh. In
collaboration with the Brewing Industry Research Foundation, for

(1) F. G . Fischer and H. Dbrfel, Hoppe-Seyler’s Z. Physiol. Clzem., 302, 186-203

(1955).
 OBITUARY-EDMUND LANGLEY HIRST 11

whom he acted as a consultant for many years, h e obtained the first

evidence of the structural differences between the starches from
malted barley and that from the ungerminated cereal, which had previ-
ously been examined by MacWilliam and E. G. V. Percival. As an
organic chemist, Hirst tended to be somewhat suspicious of enzymic
methods of analysis, and, during this period, when new starch-
degrading enzymes were being “discovered,” and later, when the
existence of some was refuted, these reservations were understand-
able. How many now remember the first published reports of isophos-
phorylase, Z-enzyme, R-enzyme, D-enzyme, and T-enzyme, and the
years of endeavor that followed in various laboratories, before their
specificity and properties were finally clarified? However, long before
Hirst retired, he came to accept the view that, when used with appro-
priate caution, enzymic methods really are valuable tools!
For many years, Hirst was a member of the Board of Governors of the
Rowett Research Institute, Aberdeen, where he took a keen interest in
the work of the Institute as a whole, and particularly that of Dr. A. E.
Oxford and the microbiology department. This led to collaboration on
the nature of the reserve carbohydrate synthesized by Cycloposthium
and by the holotrich ciliates present in sheep’s rumen. Both protozoal
polysaccharides were shown to be amylopectin in type. The work was
continued with other protozoa, in collaboration with J. F. Ryley, and
the presence of starch or amylopectin-type polysaccharides was estab-
lished in Chilomonas paramecium, Haemotococcus pluuialis, and Tet-
raselmis carteriiformis.
Finally, Hirst’s later contributions to monosaccharide chemistry
must not be overlooked. With E. G. V. and Elizabeth Percival, four
different trimethyl, 3,4- and 4,5-dimethyl, and the 4-methyl ethers
of D-fructose were synthesized from isopropylidene derivatives
having well established structures. Certain methyl ethers of
D-mannuronic acid and D-ghcuronic acid were also prepared, and their
periodate oxidation was compared with that of related ethers of
D-galacturonic acid. With Elizabeth Percival, he also contributed arti-
cles on the methyl ethers of mono- and di-saccharides, and on
glycofuranosides from cyclic carbonates, to Methods in Carbohydrate
Chemistry, Vol. 2 (1963).
Although this obituary is primarily concerned with Hirst’s achieve-
ments in carbohydrate chemistry, it must b e emphasized that the Edin-
burgh laboratories contained research groups active in other branches
of organic chemistry. The largest of these dealt with polycyclic, aro-
matic hydrocarbons, and was directed by Dr. (later Professor) Neil
Campbell, who, throughout the period, was an effective lieutenant in
12 MAURICE STACEY AND DAVID J. MANNERS

the Department, a loyal colleague, and a close friend. Many chemists

who were not involved with carbohydrates have obtained senior posi-
tions in other universities, or in the chemical industry, and Hirst did
not fail to give them his advice, help, and encouragement, as appro-
priate.
Hirst had an extensive knowledge of organic chemistry as a whole, as
visiting lecturers on non-carbohydrate subjects sometimes realized
from the shrewd questioning at the end of a lecture. He was much in
demand as an external examiner in other universities. The department
of Professor Wesley Cocker at Trinity College, Dublin, was a special
favorite of his, and he enjoyed returning to Birmingham, Manchester,
and St. Andrews. Hirst was always concerned with both the theoretical
and experimental aspects of organic chemistry; he had little time for
students who could recite the detailed mechanisms of the nitration of
aromatic compounds, but did not know the products of the nitration of
phenol!
When Hirst arrived in Edinburgh, Professor James Kendall, F. R. S.,
was the Head of the Department. On Kendall’s retirement in 1959,
Hirst took over this major responsibility, and helped to plan and co-or-
dinate the steady growth of the Department, in terms of numbers of staff
and students, and in respect of accommodation, including the adapta-
tion of parts of the original building.
Within the University of Edinburgh, Hirst played an extremely full
part in its committee work. From 1959 to 1962, h e served as Dean of the
Faculty of Science, and, in 1963, was elected to the University Court-
the most senior body, with financial and legal responsibilities for the
administration ofthe University. He continued to be a member ofcourt
from 1965, when he became one of the Curators of Patronage (a com-
bined committee of the University and the City of Edinburgh having
particular responsibilities towards appointments to certain chairs),
until his retirement.
Within the City of Edinburgh, he played important roles with other
academic institutions and with the Royal Society of Edinburgh. H e was
elected a Fellow ofthe last in 1948, and held a number ofoffices, before
serving as President during 1959-1964. In 1965, he received the Gun-
ning Victoria Jubilee Prize of the Society, for the period 1960-1964. He
was a member ofthe Board of Governors ofthe East of Scotland College
of Agriculture, and, until 1968, was convenor of the Appointments
Committee. He represented Edinburgh University on the Governing
Body of the Heriot-Watt College, and, following its elevation to Uni-
versity status in 1966, he became a member of its University Court,
where his experience and judgment were invaluable. He became a
 OBITUARY-EDMUND LANGLEY HIRST 13

member of the Watt Club (a graduate club of the Heriot-Watt), and

served as President in 1967. Hirst received the Honorary Fellowship of
the Heriot-Watt College in 1964, and the D.Sc. degree (honoris causa)
in 1968; both of these awards were greatly appreciated by him.
In Scotland as a whole, Hirst made many important contributions to
the academic and scientific worlds. He served as Chairman of the
Academic Advisory Committee that was involved in the transformation
of the Royal Technical College, Glasgow, into the University of
Strathclyde. This demanding work, and a subsequent appointment as a
member of Court from 1965 to 1970 gave him special satisfaction and
pleasure. He was a member of the Council of Management of the
Macaulay Institute for Soil Research, Aberdeen, where his specialist
knowledge was particularly welcomed by Dr. J. S. D. Bacon and his
coworkers. Until 1968, h e was a member of the Governing Body, and
chairman of the Research Committee, of the Hill Farming Research
Organisation, Edinburgh; his continued involvement with the Rowett
Research Institute and the Institute of Seaweed Research has already
been noted. In 1964, h e was elected an Honorary Fellow of the Royal
Scottish Society of Arts.
However, Hirst’s contributions were by no means confined to Scot-
land, and, in fact, his major service to Britain and its scientific commu-
nity resulted from his regular visits to London. This service was recog-
nized by the award of a Coronation Medal in 1953, and h e was made a
C.B.E. in 1957 and a Knight Bachelor in 1964.
Hirst was elected a Fellow of the Chemical Society in 1922; he
served on various committees, including the Council (from 1939 to
1942), and was elected President for 1956-1958. H e gave the Tilden
Lecture in 1940, the Hugo Muller Lecture in 1948, and the C. S. Hud-
son Memorial Lecture in 1954, and was awarded the Longstaff Medal
of the Society in 1957. He was a respected member cf the British Car-
bohydrate Nomenclature Committee. H e was elected Fellow of the
Royal Institute of Chemistry in 1934, and served on several commit-
tees, including the Council (from 1942 to 1944), becoming Vice- Presi-
dent for 1967-1969. From 1934, h e was an active member of the local
sections of the Society of Chemical Industry in Bristol, Manchester,
and Edinburgh, respectively, serving on the Council as Chairman of
the last in 1955-1956. H e was also a long-standing member of the
Faraday Society, the Biochemical Society, and the Society for Experi-
mental Biology.
Hirst was elected a Fellow ofthe Royal Society in 1934. H e served on
the Council from 1945to 1947 and again from 1962 to 1964, and was, for
some years, a member of the Chemistry Sectional Committee and the
14 MAURICE STACEY AND DAVID J. MANNERS

Government Grant Board. H e received the Davy Medal of the Royal

Society in 1948. For many years, he was a member of the select Royal
Society Dining Club. Following the end of the 1939-1945 war, he
continued his work for the Scientific Advisory Council of the Ministry
of Supply, especially with the Chemical Engineering Committee. For a
period, he was a member of the Plants and Soils Committee of the
Agricultural Research Council, and was a member of the Blackman
Committee which reported on research and related matters for the
natural-rubber industry. His work for the Department of Scientific and
Industrial Research (D.S.I.R., later the Science Research Council) was
of special importance; here, he served on the Studentships Committee
and as Chairman of the Chemistry Sub-Committee, and, during 1950-
1955, he served as Chairman of the Chemistry Research Board. For
several years, he acted as D.S.I.R. visitor to the Jute Industries Re-
search Association, Dundee, and he was for a time a member of the
Forest Products Research Board.
From his early days as a research student, Hirst was actively in-
terested in the British Association for the Advancement of Science. In
1922, he read a paper on esparto cellulose, under the Presidency of
Professor J. C. Irvine. He regularly attended the annual meeting, being
President of Section B in Birmingham in 1950, and was elected a
member of the Council for 1964-1968.
Hirst’s scientific achievements, and services to the academic world
were recognized by the award of honorary degrees by the Universities
of St. Andrews, Aberdeen, Birmingham, Strathclyde, and Trinity Col-
lege, Dublin, and as already mentioned, the Heriot-Watt University.
Hirst’s scientific contributions extended far beyond the United
Kingdom. Many will know ofhis membership ofthe Editorial Advisory
Boards ofAdvances in Carbohydrate Chemistry, of Carbohydrate Re-
search (which published a special issue in July, 1968, to honor his 70th
birthday”, and-a special interest-of the encyclopedic Rodd’s
Chemistry of Carbon Compounds. Formany years, he was a member of
the American and Swiss Chemical Societies; from 1959, he was an
Honorary Member ofthe Polish Chemical Society; and, in 1967, h e was
elected an Honorary Member of the Royal Irish Academy.
Hirst enjoyed travelling, and visited many countries either to attend
an international conference or as a visiting lecturer. His most important
conference was the International Union meeting at Likge in 1930, at
which the historic discussions with Haworth, Hudson, and Purves
(referred to earlier) took place. He visited the U.S.A. in 1951, to attend

(2) For a Memorial Issue, dedicated to Sir Edmund, see Carbohydr. Res., 57 (1977).
 OBITUARY-EDMUND LANGLEY HIRST 15

an American Chemical Society meeting in New York and to participate

in a Starch Round Table, and again, in 1966, to visit several institutions.
In 1961, he visited Canada, to attend the International Union meeting
in Montreal, and visit his old colleague and friend, J. K. N. Jones at
Kingston. His own biographical notes record visits to conferences in
several cities in Europe, including Stockholm (1953), Hamburg
(1956),Galway (1958),Vienna (1958),and Warsaw (1959),and he was a
visiting lecturer at various universities in Norway and West Germany,
in 1955 and 1956, respectively. His lectures and personal discussions
with other carbohydrate chemists are long remembered.
The foregoing account describes many aspects of Hirst’s scientific
career, but is by no means complete. His advice was widely sought by
University Vice-Chancellors, by other professors of chemistry, and by
the senior officials of Government bodies and Scientific Societies.
Throughout, his work was characterized by thorough preparation, and
when verbal comment was required, this was succint. He was a quiet
but effective chairman of meetings, but could be very firm when the
occasion demanded it. On his retirement, it was typical of Hirst that he
retained a room in the Chemistry Department, and he used this regu-
larly during the following years. He had a substantial correspondence
with former students and colleagues, who still continued to seek his
help and advice, and, as always, this was freely given.
Many of his former colleagues now hold high positions in the
academic and industrial worlds, and all readily acknowledge the great
debt that they owe him. Despite his many heavy commitments, he was
never too busy to discuss scientific and academic matters with his
colleagues, and he showed a kindly interest in their well-being and that
of their families. Many of the children of the staff received unexpected
Christmas presents from Edmund and Kay, to their mutual pleasure.
Although Hirst tended to be somewhat reserved on public occasions,
he was good company in smaller groups. He developed a special rap-
port with Neil Campbell and the late Dr. Mowbray Ritchie, the Reader
in Physical Chemistry, which was characterized by a sense of gentle
good humor. If a junior colleague should ever show signs of brashness,
this was soon halted by effective leg-pulling.
The numerous commitments in England and Scotland, especially
during the years of the Second World War, involved Hirst in extensive
travel by railway. Whilst this was tedious to others, he enjoyed it, and
he possessed a detailed knowledge ofthe railways and their timetables.
Once, while waiting in a train in a station in Glasgow to return home, a
large express train drew into the neighboring platform, whereupon
Hirst glanced at his watch and startled his companions with the remark
16 MAURICE STACEY AND DAVID J. MANNERS

“That will be the 1.00 p.m. from Euston.” And it was! His return to the
Department from a visit usually brought forth some amusing anecdote
about the meeting or his companions, or some detail of the railway
journey, or of the countryside that he enjoyed so much. One particular
railway journey paid handsome dividends. Conversation with a fellow
traveller eventually led to the award of a $23,650 grant to the Depart-
ment from the U.S. Department ofAgriculture-which, nearly 20 years
ago, was a not inconsiderable sum.
Away from the Department, Edmund Hirst enjoyed many hobbies.
He was a keen and knowledgeable photographer, both with black-
and-white and color films; with Kay’s help, he kept a vegetable and fruit
garden, although he preferred growing flowers rather than vegetables.
In his early days, he and a companion, usually E. G. Cox, climbed in the
Welsh hills, and toured-often on foot, but once on a tandem
bicycle-a great part of rural England; he was a member of the Cyclists
Touring Club until his death. He and Kay regularly attended, for many
years, the Scottish National Orchestral Concerts in the Usher Hall,
Edinburgh. His taste was exclusively classical and, preferably, pre-
Bartok, although he derived great pleasure from the operas of Gilbert
and Sullivan, and the music of Johann Strauss the elder and his sons. He
read widely, but his taste in novels was classical-Scott, Dickens, Jane
Austen, the Bronte sisters, and Trollope. Hirst adored dogs, and, during
the last few weeks of his life, he sat for hours in his arm-chair reading
“Pickwick Papers,” with Jennie, a little Cairn terrier, curled up on his
lap. In 1973, Hirst developed Hodgkin’s disease, and his health gradu-
ally deteriorated until his death on 29th October, 1975.
From 1949 on, Hirst was supported in all his work by Kay. Visitors,
especially from abroad, always received a warm welcome at their
home, and the special kindness which they showed to individual guests
will not be forgotten.
In presenting Hirst for an honorary degree at Strathclyde University,
the orator included this sentence: “Quiet and unassuming, but decisive
in thought and action, his wise guidance and judgments have been of
the greatest value to us.” We offer these words now, as atribute to a man
who achieved so much for carbohydrate chemistry, and who is sadly
missed by so many of us.

MAURICE STACEY
DAVIDJ. MANNERS
 OBITUARY-EDMUND LANGLEY HIRST 17

APPENDIX
The following is a chronological list of the scientific publications of
Sir Edmund Hirst and his colleagues.
“The Constitution ofthe Disaccharides. Part V. Cellobiose (Cellose),” W. N. Haworth
and E. L. Hirst,J. Chem. Soc., 119, 193 (1921).
“The Labile Nature of the Halogen Atom in Organic Compounds. Part V. The Action of
Hydrazine on the Halogen Derivatives of Some Esters and Substituted cycloHexanes,”
E. L. Hirst and A. K. Macbeth,J. Chem. Soc., 121,2169 (1922).
“Methylation of Xylose,” A. Carruthers and E. L. Hirst,J. Chem. Soc., 121, 2299
( 1922).
“The Labile Nature of the Halogen Atom in Organic Compounds. Part VI. The Action
of Titanous Chloride and of Ammonia on Representative Halogen Compounds,” I. A.
Black, E. L. Hirst, and A. K. Macbeth,J. Chem. Soc., 121,2527 (1922).
“The Constitution of Polysaccharides. Part VI. The Molecular Structure of Cotton
Cellulose,” J. C. Irvine and E. L. Hirst,J. Chem. Soc., 123, 518 (1923).
“The Structure of the Normal Monosaccharides. Part I. Xylose,” E. L. Hirst and C. B.
Purves,]. Chem. Soc., 123, 1352 (1923).
“The Constitution ofRaffinose,” W. N. Haworth, E. L. Hirst, and D. A. RuellJ. Chern.
Soc., 123,3125 (1923).
“The Action of Highly Concentrated Hydrochloric Acid on Cellulose and on Some
Derivatives of Glucose and of Xylose,” E. L. Hirst and D. R. Momson,]. Chem. Soc.,
123,3226 (1923).
“Constitutional Studies in the Monocarboxylic Acids Derived from Sugars. Part 111.
The Isomeric Tetramethyl Galactonolactones and Trimethyl Arabonolactones,” J. Pryde,
E. L. Hirst, and R. W. Humphreys,J. Chem. Soc., 127,348 (1925).
“The Constitution ofthe Normal Monosaccharides. Part 11.Arabinose,” E. L. Hirst and
G. J. Robertson, J. Cheni. Soc., 127, 358 (1925).
“The Structure of the Normal Monosaccharides. Part 111. Rhamnose,” E. L. Hirst and
A. K. Macbeth,J. Chem. SOC., 22 (1926).
“The Structure of the Normal Monosaccharides. Part IV. Glucose,” E. L. Hirst,
J. Chem. Soc., 350 (1926).
“The Structure of Fructose, y-Fructose and Sucrose,” W. N. Haworth and E. L. Hirst,
J. Chern. Soc., 1858 (1926).
“The Structure of Normal Fructose: Crystalline Tetramethyl p-Methyl-fructoside and
Crystalline Tetramethyl Fructose (1:3:4:5),” W. N. Haworth, E. L. Hirst, and A. Learner,
J. Chem. Soc., 1040 (1927).
“The Constitution of the Disaccharides. Part XIII. The y-Fructose Residue in Su-
crose,” W. N. Haworth, E. L. Hirst, and V. S. Nicholson,J. Chem. Soc., 1513 (1927).
“The Constitution of the Disaccharides. Part XV. Sucrose,” J. Avery, W. N. Haworth,
and E. L. Hirst,J. Chem. Soc., 2308 (1927).
“The Ring Structure in Normal Galactose. Oxidation of Tetramethyl
6-Galactonolactone,” W. N. Haworth, E. L. Hirst, and D. I. Jones,J. Chem. Soc., 2428
(1927).
“1:3:4:6-TetramethyI (y-) Fructose and 2:3:5-Trimethyl (y-) Arabinose. Oxidation ofd-
and Z-Trimethyl y-Arabonolactone,” W. N. Haworth, E. L. Hirst, and A. LearnerJ. Chem.
Soc., 2432 (1927).
“The Structure of the Normal and y-Forms of Tetramethyl Glucose. Oxidation of
Tetramethyl S and y-Gluconolactones,” W. N. Haworth, E. L. Hirst, and E. J. Miller,
I. Chem. Soc., 2436 (1927).
18 MAURICE STACEY AND DAVID J. MANNERS

“Derivatives of Orcinol. Part I,” E. L. Hirst,]. Chem. Soc., 2490 (1927).

“Ring Structure and Optical Relationships in the Mannose-Rhamnose-Lyxose Series
of Sugars. Isolation of a New Form of Lyxose,” W. N. Haworth and E. L. Hirst,]. Chem
Soc., 1221 (1928).
“Polysaccharides. Part 11. The Acetylation and Methylation of Starch,” W. N. Haworth,
E. L. Hirst, and J. I. Webb,]. Chem. Soc., 2681 (1928).
“Organic Chemistry. Part I . (Aliphatic Division),” W. N. Haworth and E. L. Hirst,
Annu. R e p . Prog. Cheni., Chem. S O C . London, 25,67 (1928).
“Polysaccharides. Part IV. The Constitution ofxylan,” H. A. Hampton, W. N. Haworth,
and E. L. Hirst,]. Chem. Soc., 1739 (1929).
“The Structure of Normal Monosaccharides, Part VI. 2:3:4-Trimethyl Rhamnonolac-
tone,” J. Avery and E. L. Hirst,]. Chem. Soc., 2466 (1929).
“The Development ofa Novel Form ofStereoisomerism in the Sugar Series. Part I. The
Third Variety of Triacetyl Methylrhamnoside,” W. N. Haworth, E. L. Hirst, and E. J.
Miller,]. Chem. SOC., 2469 (1929).
“Polysaccharides. Part V. Glycogen,” W. N. Haworth, E. L. Hirst, and J. I. Webb,
J. Chem. Soc. 2479 (1929).
“Organic Chemistry. Part I. (Aliphatic Division),” W. N. Haworth and E. L. Hirst,
Annu. R e p . Prog. Chem., Chem. Soc. London, 26,74 (1929).
“Crystalline cu-Methylmannofuranoside (y-Methylmannoside). Part 11,” W. N.
Haworth, E. L. Hirst, and J . I. Webb,]. Chem. Soc., 651 (1930).
“Derivatives of Lyxofuranose,” H. G . Bott, E. L. Hirst, and J. A. B. Smith,]. Chem.
Soc., 658 (1930).
“The Development of a Novel Form of Isomerism in the Sugar Series. Part 11. The
Third Variety of Tetra-acetyl Methylmannoside,” H. G. Bott, W. N. Haworth, and E. L.
Hirst, J . Chem. SOC., 1395 (1930).
“The Conversion of 1:2:3:4-Tetra-acetyl P-d-Glucose into 2:3:4:6-Tetra-acetyl
P-Methylglucoside,” W. N. Haworth, E. L. Hirst, and (Miss) E. G. Teece,]. Chem. Soc.,
1405 (1930).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part I. General
Introduction,” W. N. Haworth and E. L. Hirst,]. Chem. Soc., 2615 (1930).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part 11.
4-Glucosido-cu-niannose and its Derivatives,” W. N. Haworth, E. L. Hirst, H. R. L.
Streight, H. A. Thomas, and J. I. Webb,]. Chem. Soc., 2636 (1930).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part 111.
4-Galactosido-c~-mannoseandits Derivatives,” W. N. Haworth, E. L. Hirst, (Miss) M. M.
T. Plant, and R. J. W. Reynolds,]. Cheni. S O C . , 2644 (1930).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part IV. Deriva-
tives ofa- and P-Methylmannopyranoside,” H. G. Bott, W. N. Haworth, E. L. Hirst, and R.
S. Tipson, J . Chem. Soc., 2653 (1930).
“The Structure ofcarbohydrates and their Optical Rotatory Power. Part V. The Optical
Rotatory Powers of Methylated Lactones Derived from the Simple Sugars,” W. N.
Haworth, E. L. Hirst, and J. A. B. Smith,]. Chem. Soc., 2659 (1930).
“The Existence of the Cellobiose Residue in Cellulose,” W. N. Haworth, E. L. Hirst,
and H. A. Thomas, Nature, 126, 438 (1930).
“The Compound Uronic Acids. Structure ofthe Aldobionic Acid from Gum Arabic,” S .
W. Challinor, W. N. Haworth, and E. L. Hirst,]. Chem Soc., 258 (1931).
“Polysaccharides. PartVI. Trimethyl Cellulose,” W. N. Haworth, E. L. Hirst, and H. A.
Thomas,]. Chem. SOC., 821 (1931).
“Polysaccharides. Part VII. Isolation of Octamethyl Cellobiose, Hendecaniethyl Cel-
 OBITUARY-EDMUND LANGLEY HIRST 19

lotriose, and a Methylated Cellodextrin (Cellotetrose?) as Crystalline Products of the

Acetolysis of Cellulose Derivatives,” W. N. Haworth, E. L. Hirst, and H. A. Thomas,
J . Chem. Soc., 824 (1931).
“The Structure of Glucal,” E. L. Hirst and C. S. Woolvin,J. Chem. Soc., 1131 (1931).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part VI.
4-Glucosido-mannose and its Methylated Derivatives,” W. N. Haworth, E. L. Hirst, and
H. R. L. Streight,J. Chem. Soc., 1349 (1931).
“The Structure of Carbohydrates and their Optical Rotatory Power. Part VII.
4-Galactosido-niannose and its Methylated Derivatives,” W. N. Haworth, E. L. Hirst, and
(Miss) M. M. T. Plant,]. Chem. SOC., 1354 (1931).
“Conversion of 2:3:4-Triacetyl a-Methylglucoside into 3:4:6-Triacetyl 2-Methyl
a-Methylglucoside,” W. N. Haworth, E. L. Hirst, and (Miss) E. G. Teece,J. Chem Soc.,
2858 (1931).
“The Third Variety ofTriacety1 Methyl-rhamnoside,” W. N. Haworth, E. L. Hirst, and
H. Samuels,J. Chem. Soc., 2861 (1931).
“Walden Inversion in the a-Glucoheptose Series. The Preparation ofNew Derivatives
and the Determination of the Structure of Methyl-a-glucoheptoside,” W. N. Haworth, E.
L. Hirst, and M. Stacey,J. Chem. Soc., 2864 (1931).
“The Rble of Tobacco-smoking in the Production of Cancer,” E. A. Cooper, H. G. Bott,
H. D. Cheeseworth, R. S. Tipson, F. W. Lamb, E. Sanders, and E. L. Hirst,
J . Hyg., 32,293 (1932).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part I,” T. L. Harris, E. L.
Hirst, and C . E. Wood,]. Chem. Soc., 2108 (1932).
“Polysaccharides. Part XII. Acetolysis Products of Cellulose,” W. N. Haworth, E. L.
Hirst, and 0. Ant-Wuorinen,J. Chenz. Soc., 2368 (1932).
“Polysaccharides. Part XIV. The Molecular Struchne of Amylose and of Aniylopec-
tin,” E. L. Hirst, (Miss) M. M. T. Plant, and (in part) (Miss) M. D. Wilkinson,J. Chem.
Soc., 2375 (1932).
“Polysaccharides. Part XV. The Molecular Structure of Inulin, ” W. N. Haworth, E. L.
Hirst, and E. G. V. Percival, J . Chem. Soc., 2384 (1932).
“Methylation of Monocarboxylic Acids Derived from Aldoses. Structure of Penta-
methyl a-Gluco-heptono-y-lactone,” W. N. Haworth, E. L. Hirst, and M. Stacey, J .
Chem. Soc., 2481 (1932).
‘‘Formation of Furfural from Methylated Pentoses,” H. G. Bott and E. L. Hirst,
J . Chem. Soc., 2621 (1932).
“Hexuronic Acid as the Antiscorbutic Factor,” E. L. Hirst and R. J. W. Reynolds,
Nature, 129,576 (1932).
“The Absorption Spectrum of Hexuronic Acid,” R. W. Herbert and E. L. Hirst,
Nature, 130,205 (1932).
“Constitution of Vitamin C,” E. G. Cox and E. L. Hirst, Nature, 131,402 (1933).
“Constitution of Ascorbic Acid,” E. L. Hirst, E. G. V. Percival, and F. Smith, Nature,
131,617 (1933).
“The Structure of Ascorbic Acid,” E. L. Hirst, J . Soc. Chem. Znd. London, 52, 221
(1933).
“Synthesis of Ascorbic Acid,” W. N. Haworth and E. L. Hirst, J . SOC. Chem. Znd.
London, 52, 645 (1933).
“Ascorbic Acid as the Antiscorbutic Factor,” E. L. Hirst and S. S. Zilva, Biochem.J.,
27, 1271 (1933).
“The Constitution ofAscorbic Acid,” R. W. Herbert, E. L. Hirst, E. G. V. Percival, R. J.
W. Reynolds, and F. Smith,]. Chem. Soc., 1270 (1933).
20 MAURICE STACEY AND DAVID J. MANNERS

“Synthesis of d- and l-Ascorbic Acid and of Analogous Substances,” R. G. Ault, D. K.

Baird, H. G. Carrington, W. N. Haworth, R. W. Herbert, E. L. Hirst, E. G. V. Percival, F.
Smith, and M. Stacey,]. Chem. SOC.,1419 (1933).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part 11. Ascorbic Acid,” R.
W. Herbert, E. L. Hirst, and C. E. Wood,]. Chem. Soc., 1564 (1933).
“The Molecular Structure of Polysaccharides,” W. N. Haworth and E. L. Hirst, Truns.
Furuduy SOC.,29, 14 (1933).
“Sucrose and Other Disaccharides. Sir James Irvine’s ‘Correction’,” H. C. Carrington,
W. N. Haworth, and E. L. Hirst,].Am. Chem. SOC.,55, 1084 (1933).
“Ascorbic Acid and Synthetic Analogues,” D. K. Baird, W. N. Haworth, R. W. Herbert,
E. L. Hirst, F. Smith, and M. Stacey,]. Chem. Soc., 62 (1934).
“Isolation of a Crystalline Dimethyl Anhydromethyl-hexoside. Characterisation of
3:4:6-Trimethyl Glucose,” W. N. Haworth, E. L. Hirst, and L. Panizzon,]. Chem. Soc.,
154 (1934).
“Malta1 and 4-a-Glucosidomannose,” W. N. Haworth, E. L. Hirst, and R. J. W.
Reynolds, J. Chem. SOC.,302 (1934).
“Polysaccharides. Part XVII. The Constitution and Chain Length of Levan,” S. W.
Challinor, W. N. Haworth, and E. L. Hirst,]. Chem. Soc., 676 (1934).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part 111. Tetra-methyl
a-Methylglucopyranoside and Tetramethyl a-Methylmannopyranoside,” R. W. Herbert,
E. L. Hirst, and C. E. Wood,]. Chem. Soc., 1151 (1934).
“Physiological Activity of Synthetic Ascorbic Acid,” W. N. Haworth, E. L. Hirst, and S.
S. Zilva,]. Chem. SOC.,1155 (1934).
“Synthesis of Ascorbic Acid and its Analogues: The Addition of Hydrogen Cyanide to
Osones,” W. N. Haworth, E. L. Hirst, J. K. N. Jones, and F. Smith,J. Chem. SOC.,1192
( 1934).
“Methyl Ethers ofAscorbic Acid,” W. N. Haworth, E. L. Hirst, and F. Smith,]. Chem.
SOC., 1556 (1934).
“The Carbohydrates of Grass. Isolation of a Polysaccharide of the Levan Type,” S. W.
Challinor, W. N. Haworth, and E. L. Hirst,]. Chem. SOC., 1560 (1934).
“Preparation ofArabinose from Gum Acacia (Gum Kordofan),” H. C. Carrington, W. N.
Haworth, and E. L. Hirst,]. Chem. SOC., 1653 (1934).
“The Constitution of Ascorbic Acid. Action of Sodium Hypochlorite on a-Methoxy-
acid Amides,” R. G. Ault, W. N. Haworth, and E. L. Hirst,J. Chem. Soc., 1722 (1934).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part IV. Tetramethyl-y-man-
nonolactone,” T. L. Harris, E. L. Hirst, and C. E. Wood, J. Chem. Sac., 1825 (1934).
“Polysaccharides. Part XVIII. The Constitution ofXylan,” W. N. Haworth, E. L. Hirst,
and E. Oliver,]. Chem. Soc., 1917 (1934).
“The Primary Product of the Synthesis of Ascorbic Acid and its Analogues,” W. N.
Haworth and E. L. Hirst, Helv. Chim. Actu, 17, 520 (1934).
“Organic Chemistry, Part I (Aliphatic Division),” H. D. K. Drew, E. L. Hirst, R. S.
Morrell, and S. Peat, Annu. Rep. Prog. Chem., Chem. Soc. London, 31, 143 (1934).
“Polysaccharides. Part XIX. The Molecular Structure of Waxy Maize Starch,” W. N.
Haworth, E. L. Hirst, and M. D. Woolgar,]. Chem. Soc., 177 (1935).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part V. Tetramethyl
y-Gluconolactone,” R. W. Herbert, E. L. Hirst, H. Samuels, and C. E. Wood,]. Chem.
SOC.,295 (1935).
“Preparation of d-Mannuronic Acid and its Derivatives,” R. G. Ault, W. N. Haworth,
and E. L. Hint,]. Chem. SOC., 517 (1935).
“Acetone Derivatives of Methylglucosides,” R. G. Ault, W. N. Haworth, and E. L.
Hirst,]. Chem. Soc., 1012 (1935).
 OBITUARY-EDMUND LANGLEY HIRST 21

“Polysaccharides. Part XX. The Molecular Size of Amylose and the Relationship
Between Amylose and Starch,” D. K. Baird, W. N. Haworth, and E. L. Hirst,]. Chem. Soc.,
1201 (1935).
“Polysaccharides. Part XXI. The Constitution and Chain Length of Some Starch Dex-
trins,” W. N. Haworth, E. L. Hirst, and M. M. T. Plant, J . Chem. Soc., 1214 (1935).
“Polysaccharides. Part XXII. Constitution and Molecular Structure of
a-Amylodextrin,” W. N. Haworth, E. L. Hirst, and A. C. Waine, J . Chem. Soc., 1299
(1935).
“Absorption Spectra in Relation to the Constitution of Derivatives of Isatin and Carbo-
styril,” R. G. Ault, E. L. Hirst, and R. A. Morton,]. Chem. Soc., 1653 (1935).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part VI. The Amide Rotation
Rule,” T. L. Harris, E. L. Hirst, and C. E. Wood,]. Chem. Soc., 1658 (1935).
“Absorption Spectra of the Metabolic Acids of Penicillium charlesii and Their Rela-
tionship to the Absorption Spectrum of Ascorbic Acid,” R. W. Herbert and E. L. Hirst,
Biochem. ]., 29,1881 (1935).
“Organic Chemistry. Carbohydrates,” E. L. Hirst and S. Peat, Annu. Rep. Prog.
Chein., Chem. Soc. London, 32, 272 (1935).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part VII. The Glucal Series,”
T. L. Harris, R. W. Herbert, E. L. Hirst, C. E. Wood and H. Woodward,]. Chem. Soc.,
1403 (1936).
“Organic Chemistry. Carbohydrates,” E. L. Hirst and S. Peat, Annu. Rep. Prog.
C l z e n . , Chem. Soc. London, 33, 245 (1936).
“The Chemistry of the Carbohydrates and the Glucosides,” W. N. Haworth and E. L.
Hirst, Annu. Reu. Biochem., 5, 81 (1936).
“Gluco-ascorbic Acid,” W. N. Haworth, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc.,
549 (1937).
“Polysaccharides. Part XXIII. Determination of the Chain Length of Glycogen,” W. N.
Haworth, E. L. Hirst, and F. A. Isherwood,]. Cheni. Soc., 577 (1937).
“The Ring Structure ofXylal,” W. N. Haworth, E. L. Hirst, and C. S. Woolvin,]. Chem.
Soc., 780 (1937).
“Degradation of Methylated Inulin to Hexamethyl Difructosan,” W. N. Haworth, E. L.
Hirst, and F. A. Isherwood,]. Chem. Soc., 782 (1937).
“Polysaccharides. Part XXIV. Yeast Mannan,” W. N. Haworth, E. L. Hirst, and F. A.
Isherwood,]. Chem. Soc., 784 (1937).
“Polysaccharides. Part XXV. a-Amylodextrin,” W. N. Haworth, E. L. Hirst, H. Kitchen,
and S. Peat,]. Chem. Soc., 791 (1937).
“Acetone Derivatives of Gluconic Acid,” W. N. Haworth, E. L. Hirst, and K. A.
Chamberlain, J . Chem. Soc., 795 (1937).
“Isomerisation of 2:3-Dimethyl Ascorbic Acid,” W. N. Haworth, E. L. Hirst, F. Smith,
and W. J. Wilson,]. Chem. Soc., 829 (1937).
“Optical Rotatory Dispersion in the Carbohydrate Group. Part VIII. Tetramethyl
8-Gluconolactone and Tetramethyl 6-Galactonolactone,” T. L. Harris, E. L. Hirst, and C.
E. Wood, J . Chem. Soc., 848 (1937).
“Polysaccharides. Part XXVI. Xylan,” R. A. S. Bywater, W. N. Haworth, E. L. Hirst, and
S. Peat,”]. Cheni. Soc., 1983 (1937).
“A Water-soluble Polysaccharide from Barley Leaves,” W. N. Haworth, E. L. Hirst, and
R. R. Lyne, Biochem. ]., 31, 786 (1937).
“The Chemistry of the Carbohydrates and the Glycosides,” W. N. Haworth and E. L.
Hirst, Annu. Reo. Biochem., 6, 99 (1937).
“Pectic Substances. Part I. The Araban and Pectic Acid ofthe Peanut,” E. L. Hirst and
J. K. N. Jones,]. Chem. Soc., 496 (1938).
22 MAURICE STACEY AND DAVID J. MANNERS

“Analogues of Ascorbic Acid Containing Six-membered Rings,” W. N. Haworth, E . L.

Hirst, and J. K. N. Jones,]. Chem. SOC., 710 (1938).
“The Constitution of Damson Gum. Part I. Composition of Damson Gum and Structure
of an Aldobionic Acid (Glycuronosido-2-mannose) Derived from it,” E. L. Hirst and J. K.
N. Jones,]. Chem. SOC., 1174 (1938).
“Polysaccharides. Part XXVIII. The ‘End-group’ Method as Applied to Starch. An
Improved Method for the Estimation of Tetramethyl Glucose in Admixture with
Trimethyl Glucose,” E. L. Hirst and G. T. Young,]. Chem. Soc., 1247 (1938).
“The Ring Structure of Methylgalactofuranoside,” W. N. Haworth, E. L. Hirst, D. I.
Jones, and H. Woodward,]. Chem. Soc., 1575 (1938).
“Methylation of a-Methylglucoside by Thallous Hydroxide and Methyl Iodide,” C. C.
Barker, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 1695 (1938).
“Estimation of Uronic Anhydride Residues in Polysaccharides,” E. L. Hirst and G. T.
Young, Nature, 142, 912 (1938).
“The Chemistry of Ascorbic Acid (Vitamin C) and its Analogues,” W. N. Haworth and
E. L. Hirst, Ergeb. Vitam.Hormonforsch., 160, 191 (1939).
“The Structure and Synthesis of Vitamin C (Ascorbic Acid) and its Analogues,” E. L.
Hirst, Fortschr. Chem. Org. Naturst., 2, 132 (1939).
“Structure of Alginic Acid,” E. L. Hirst, J. K. N.Jones, and W. 0. Jones, Nature, 143,
857 (1939).
“Methyl Ethers ofArabo-ascorbic Acid and their Isomerism,” E. G. E. Hawkins, E. L.
Hirst, and J. K. N. Jones,]. Chem. SOC., 246 (1939).
“Pectic Substances. Part 11. Isolation ofan Araban from the Carbohydrate Constituents
of the Peanut,” E. L. Hirst and J. K. N. Jones,J. Chem. Soc., 452 (1939).
“Pectic Substances. Part 111.Composition ofApple Pectin and the Molecular Structure
of the Araban Component of Apple Pectin,” E. L. Hirst and J. K. N. Jones,]. Chem.
Soc., 454 (1939).
“Polysaccharides. Part XXXI. Constitution of Wheat Starch and Horse-chestnut
Starch,” E. L. Hirst and G . T. Young,]. Chem. Soc., 951 (1939).
“Constitution ofthe Mucilage from the Bark of Ulmusfuloa (Slippery Elm Mucilage).
Part I. The Aldobionic Acid Obtained by Hydrolysis of the Mucilage,” R. E. Gill, E. L.
Hirst, and J. K. N. Jones,]. Chem. Soc., 1469 (1939).
“Polysaccharides. Part XXXII. The Molecular Constitution of Rice Starch,” E. L. Hirst
and G. T. Young,]. Chem. Soc., 1471 (1939).
“The Constitution of Damson Gum. Part 11. Hydrolysis Products from Methylated
Degraded (Arabinose-free) Damson Gum,” E. L. Hirst and J. K. N. Jones,]. Chem. Soc.,
1482 (1939).
“Pectic Substances. Part IV. Citrus Araban,” G. H. Beaven, E. L. Hirst, and J. K. N.
Jones,]. Chem. SOC., 1865 (1939).
“2:3:4-Trimethyl Mannose,” W. N. Haworth, E. L. Hirst, F. A. Isherwood, and J . K. N.
Jones, J . Chem. SOC., 1878 (1939).
“The Structure of Alginic Acid. Part I,” E. L. Hirst, J. K. N. Jones, and (Miss) W. 0.
Jones,]. Chem. Soc., 1880 (1939).
“Polysaccharides. Part XXXIII. The Methylation of Cellulose in Air and in Nitrogen,”
W. N. Haworth, E. L. Hirst, L. N. Owen, S. Peat, and (in part) F. J. AverillJ. Chem. Soc.,
1885 (1939).
“Polysaccharides. Part XXXV. Hydrocellulose,” H. C. Carrington, W. N. Haworth,
E. L. Hirst, and M. Stacey,]. Chem. Soc., 1901 (1939).
“Polysaccharides. Part MXVIII. The Constitution of Glycogen from Fish Liver and
Fish Muscle,” W. N. Haworth, E. L. Hirst, and F. Smith,]. Chem. Soc., 1914 (1939).
“The Nature of the Bonds in Starch,” C. E. H. Bawn, E. L. Hirst, and G. T. Young,
Trum. Faraday Soc., 36, 880 (1940).
 OBITUARY-EDMUND LANGLEY HIRST 23

“The €-Galactan of Larch Wood,” E. L. Hirst, J. K. N. Jones, and W. G. Campbell,

Nature, 147,25 (1941).
“Linkage Between the Repeating Units in the Starch Molecule,” C. C. Barker, E. L.
Hirst, and G. T. Young, Nature, 147, 296 (1941).
“Recent Progress in the Chemistry of the Pectic Materials and Plant Gums” (Tilden
Lecture), E. L. Hirst,]. Chem. Soc., 70 (1942).
“Nitrogenous Substances Synthesized by Moulds,” A. H. Campbell, M . E. Foss, E. L.
Hirst, and J. K. N. Jones, Nature, 155, 141 (1945).
“Application of New Methods of End-group Determination to Structural Problems in
the Polysaccharides,” F. Brown, Sonia Dunstan, T. G. Halsall, E. L. Hirst, and J. K. N.
Jones, Nature, 156, 785 (1945).
“The Constitution of Damson Gum. Part 111. Hydrolysis Products from Methylated
Damson Gum,” E. L. Hirst and J. K. N. Jones,]. Chem. Soc., 506 (1946).
“Methylation of P-Methylglucopyranoside and cup-Methylxylopyranosides by
Thallous Hydroxide and Methyl Iodide,” C. C. Barker, E. L. Hirst, and J. K. N. Jones,
J. Chem. Soc., 783 (1946).
“Constitution of the Mucilage from the Bark of Ulmusfuloa (Slippery Elm Mucilage).
Part 11. The Sugars Formed in the Hydrolysis of the Methylated Mucilage,” R. E. Gill,
E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 1025 (1946).
“The Chemistry of Pectic Materials,” E. L. Hirst and J. K. N. Jones, Adv. Carbohydr.
Chem., 2,235 (1946).
“Structure of Starch and Cellulose,” T. G. Halsall, E. L. Hirst, and J. K. N. Jones,
Nature, 159, 97 (1947).
“Quantitative Estimation of Mixtures of Sugars by the Paper Chromatogram Method,”
A. E. Flood, E. L. Hirst, and J. K. N. Jones, Nature, 160, 86 (1947).
“Structure of Starch: Mode of Attachment of the Side-chains in Amylopectin,” T. G .
Halsall, E. L. Hirst, J. K. N. Jones, and A. Roudier, Nature, 160, 899 (1947).
“The Chemistry of Some Plant Gums and Mucilages,” E. L. Hirst and J. K. N. Jones,
J. Soc. Dyers Colour., 63,249 (1947).
“The Reaction of 1-Nitropropane with Formaldehyde and Ammonia,” E. L. Hirst, J. K.
N. Jones, (Mrs) A. Minahan, F. W. Ochynski, A. T. Thomas, and T. Urbanski,]. Chem.
Soc., 924 (1947).
“The Quantitative Determination of Galactose, Mannose, Arabinose and Rhamnose,”
E. L. Hirst, J. K. N. Jones, and E. A. Woods,]. Chem. Soc., 1048 (1947).
“The Synthesis of 3-Methyl and 3:5-Dimethyl I-Arabinose,” E. L. Hirst, J. K. N. Jones,
and (Miss) E. Williams,]. Chem. Soc., 1062 (1947).
“The Constitution of Egg-plum Gum. Part I,” E. L. Hirst and J. K. N. Jones,]. Chem.
Soc., 1064 (1947).
“Pectic Substances. Part VI. The Structure ofthe Araban fromArachis hypogea,” E. L.
Hirst and J. K. N. Jones,J. Chem. Soc., 1221 (1947).
“Pectic Substances. Part VII. The Constitution ofthe Galactan from Lupinus albus,” E.
L. Hirst, J. K. N. Jones, and (Mrs) W. 0. Walder,]. Chem. Soc., 1225 (1947).
“The Structure of Glycogen. Ratio of Non-terminal to Terminal Glucose Residues,” T.
G. Halsall, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc.,1399 (1947).
“Oxidation of Carbohydrates by the Periodate Ion,” T. G. Halsall, E. L. Hirst, and J. K.
N. Jones,]. Chem. Soc., 1427 (1947).
“The Galactomannan of the Lucerne Seed,” E. L. Hirst, J. K. N. Jones, and (Mrs)W. 0.
Walder,]. Chem. Soc., 1443 (1947).
“Separation and Identification of Methylated Sugars on the Paper Chromatogram,” F.
Brown, E. L. Hirst, L. Hough, J. K. N. Jones, and H. Wadman, Nature, 161,720 (1948).
“The Structure of Starch. The Ratio of Non-terminal to Terminal Groups,” F. Brown, T.
G. Halsall, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 27 (1948).
24 MAURICE STACEY AND DAVID J. MANNERS

“The Structure of Egg-plum Gum. Part 11.The Hydrolysis Products Obtained from the
Methylated Degraded Gum,” E. L. Hirst and J. K. N. Jones,]. Chem. Soc., 120 (1948).
“The €-Galactan ofLarch Wood(Larixdeciduu),” W. G. Campbell, E. L. Hirst, and J. K.
N. Jones,]. Chem. Soc., 774 (1948).
“The Galactomannan of Carob-seed Gum (Gum Gatto),” E. L. Hirst and J. K. Jones,
J . Chem. Soc., 1278 (1948).
“The Structure of Almond-tree Gum. Part I. The Constitution of the Aldobionic Acid
Derived from the Gum,” F. Brown, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 1677
(1948).
“Quantitative Analysis of Mixtures of Sugars by the Method of Partition Chromatog-
raphy. Part I. Standardisation of Procedure,” A. E. Flood, E. L. Hirst, and J. K. N. Jones,
J . Chem. Soc., 1679 (1948).
“Structure of Acorn Starch,” E. L. Hirst, J. K. N. Jones, and A. J. Roudier,]. Chem.
Soc., 1779 (1948).
“Pectic Substances. Part VIII. The Araban Component of Sugar-beet Pectin,” E. L.
Hirst and J. K. N. Jones,]. Chem. Soc., 2311 (1948).
“The Amylose Content of the Starch Present in the Growing Potato Tuber,” T. G.
Halsall, E. L. Hirst, J. K. N. Jones, and F. W. Sansome, Biochem. ]., 43, 70 (1948).
“La Structure chimique de l’Amidon: Mode de Liaison. Des Chaines Laterales de
l’hylopectine,” T. G . Halsall, E. L. Hirst, J. K. N. Jones, and A. Roudier, Mkm. Sera
Chim. Etat, 34 (1948).
“Composition of the Gum of Sterculia setigera: Occurrence of D-Tagatose in Nature,”
E. L. Hirst, L. Hough, and J. K. N. Jones, Nature, 163, 177 (1949).
“Chromatographic Analysis. The Application of Partition Chromatography to the Sep-
aration of the Sugars and Their Derivatives,” E. L. Hirst and J. K. N. Jones, Discuss.
Furaday Soc., No. 7, 268 (1949).
“Pear Cell-wall Cellulose,” E. L. Hirst, F. A. Ishenvood, M. A. Jermyn, and J. K. N.
Jones,]. Chem. Soc., S182 (1949).
“The Occurrence and Significance ofthe Pentose Sugars in Nature, and their Relation-
ship to the Hexoses,” E. L. Hirst,]. Chem. Soc., 552 (1949).
“Quantitative Analysis of Mixtures of Sugars by the Method of Partition Chromatog-
raphy. Part 11. The Separation and Determination of Methylated Aldoses,” E. L. Hirst,
L. Hough, and J. K. N. Jones,]. Chem. Soc., 928 (1949).
“The Polysaccharides of the Florideae. Floridean Starch,” V. C. Barry, T. G. Halsall,
E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 1468 (1949).
“Quantitative Analysis of Mixtures of Sugars by the Method of Partition Chromatog-
raphy. Part 111.Determination ofthe Sugars by Oxidation with Sodium Periodate,” E. L.
Hirst and J. K. N. Jones,]. Chem. Soc., 1659 (1949).
“The Constitution of Egg-plum Gum. Part 111. The Hydrolysis Products Obtained
from the Methylated Gum,” F. Brown, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc.,
1757 (1949).
“Cholla Gum,” F. Brown, E. L. Hirst, and J. K. N. Jones,]. Chem. Soc., 1761 (1949).
“The Structure of Sterculia setigera Gum. Part I. An Investigation by the Method of
Paper Partition Chromatography of the Products of Hydrolysis of the Gum,” E. L. Hirst,
L. Hough, and J. K. N. Jones,]. Chem. Soc., 3145 (1949).
“The Action of PAmylase on Amylopectin and on Glycogen,” T. G. Halsall, E. L.
Hirst, L. Hough, and J. K. N. Jones,]. Chem. Soc., 3200 (1949).
“The Industrial Utilisation of Agricultural Products and of Seaweed,” E. L. Hirst,
Proc. Colloq. R. Inst. Chem., Third Session, Dublin (1949).
“Chemical Constitution of Slippery Elm Mucilage: Isolation of 3-Methyl D-Galactose
 OBITUARY-EDMUND LANGLEY HIRST 25

from the Hydrolysis Products,” E. L. Hirst, L. Hough, and J. K. N. Jones, Nature, 165,34
(1950).
“Sir Norman Haworth, F.R.S.: Obituary Notice,” E. L. Hirst, Nature, 165,587 (1950).
“Modern Development in Carbohydrate Chemistry,” E. L. Hirst, Adu. Sci., No. 26,
September (1950).
“On the Structure of Knudsen’s Base and of Related Compounds. Part I,” M. E. Foss,
E. L. Hirst, J. K. N. Jones, H. D. Spr,ingall, A. T. Thomas, and T. Urbanski,]. Chem. SOC.,
624 (1950).
“The Constitution ofXylan from Esparto Grass (Stipatenacissima, L.),” S . K. Chanda,
E. L. Hirst, J. K. N. Jones, and E. G. V. Percival,]. Chem. Soc., 1289 (1950).
“Studies in Fructosans. Part I. Inulin from Dahlia Tubers,” E. L. Hirst, D. I.
McGilvray, and E. G. V. Perciva1,J. Chem. Soc., 1297 (1950).
“On the Structure of Knudsen’s Base and of Related ComDounds. Part 11,” M. E. Foss.
E. L. Hirst, J K. N. Jones, H. D. Springall, A. T. Thomas, an2 T. Urbanski,j. Chem. Soc.;
1691 (1950).
. ,
“Grapefruitand Lemon Gums. Part I. The Ratio of Sugars Present in the Gums and the
Structure of the Aldobionic Acid (4-~-Glucuronosido-~-galactose) Isolated by Graded
Hydrolysis of the Polysaccharide,” J. J. Connell, (Miss) Ruth M. Hainsworth, E. L.
Hirst, and J. K. N. Jones,]. Chem. Soc., 1969 (1950).
“The Constitution of Laminarin. Part I. An Investigation on Laminarin Isolated from
Laminaria cloustoni,” J. J. Connell, E. L. Hirst, and E. G. V. Percival,]. Chem. Soc.,
3494 (1950).
“Constitution of the Mucilage from the Bark of Ulmusfulua (Slippery Elm Mucilage).
Part 111. The Isolation of 3-Monomethyl D-Galactose from the Products of Hydrolysis,”
E. L. Hirst, L. Hough, and J. K. N. Jones,J. Chem. Soc., 323 (1951).
“The Constitution of a Pear Cell-wall Xylan,” S. K. Chanda, E. L. Hirst, and E. G. V.
Percival,]. Chem. Soc., 1240 (1951).
“Walter Norman Haworth, 1883-1950. Obituary Notice,” E. L. Hirst,]. Chem. Soc.,
2790 (1951).
“Wood Starches. Part I,” W. G. Campbell, J. L. Frahn, E. L. Hirst, D. F. Packman, and
(the late) E. G. V. Percival,]. Chem. Soc., 3489 (1951).
“The Chemistry ofplant Gums and Mucilages,” E. L. Hirst,En&aoour, 10,106 (1951).
“The Chemistry of the Plant Gums,” E. L. Hirst and J. K. N. Jones, Research, 4,411
(1951).
“Walter Norman Haworth, 1883-1950. Obituary Notice,” E. L. Hirst, R. Soc. London,
Obituary Notices, 7, (Nov) 373 (1951).
“Obituary Notice. Dr. E. G. V. Percival,” E. L. Hirst, Nature, 168, 855 (1951).
“Walter Norman Haworth, 1883-1950,” E. L. Hirst, Adu. Carbohydr. Chem., 6, 1
(1951).
“Obituary Notice. Edmund George Vincent Percival,” E. L. Hirst,]. Chem. Soc., 1557
(1952).
“The Structure of Alginic Acid. Part 11,” S. K. Chanda, E. L. Hirst, (the late) E. G. V.
Percival, and A. G. Ross,]. Chem. Soc., 1833 (1952).
“Obituary Notice. W. G. Campbell,” E. L. Hirst, Nature, 169,524 (1952).
“Sir James Irvine, K.B.E., F.R.S., 1877-1952,” E. L. Hirst,]. Am. Chem. SOC., 75,253
(1953).
“Studies on Fructosans. Part IV. A Fructosan from Dactylis glomerata,” G. 0.
Aspinall, E. L. Hirst, (the late) E. G. V. Percival, and R. G. J. Telfer,]. Chem. Soc., 337
(1953).
“The Hemicelluloses of Esparto Grass (Stipa tenacissima L.).The Arabinose-rich
26 MAURICE STACEY AND DAVID J. MANNERS

Fraction,” G. 0. Aspinall, E. L. Hirst, R. W. Moody, and (the late) E. G. V. Percival,

J. Chem. SOC., 1631 (1953).
“Protozoal Polysaccharides. Structure of the Polysaccharides Produced by the
Holotrich Ciliates Present in Sheep’s Rumen,” G. Forsyth and E. L. Hirst, J. Chem.
SOC., 2132 (1953).
“Protozoal Polysaccharides. Structure of a Polysaccharide Produced by Cyclo-
posthium,” G . Forsyth, E. L. Hirst, and A. E. Oxford,J. Chem. SOC., 2030 (1953).
“The Structure of Karaya Gum (Cochlospermum gossypium),” E. L. Hirst and Sonia
Dunstan,]. Chem. SOC., 2332 (1953).
“Syntheses of Methyl Ethers o f Fructose,” E. L. Hirst, W. E. A. Mitchell, Elizabeth E.
Percival, and (the late) E. G. V. Perciva1.J. Chem. SOC., 3170 (1953).
“The Mannans o f Ivory Nut (Phytelephas macrocarpa). Part I. The Methylation o f
Mannan A and Mannan B,” G. 0.Aspinall, E. L. Hirst, (the late) E. G. V. Percival, and I.
R. Williamson,J. Chem. SOC., 3184 (1953).
“Schools of Chemistry in Great Britain and Ireland-VII. The University of
Edinburgh,” E. L. Hirst and M. RitchieJ. R . Znst. Chem., 77 (November), 505 (1953).
“SirJames ColquhounIrvine, K.B.E., D.C.L.,LL.D., F.R.S.: ObituaryNotice,”R. SOC.
Edinburgh, Year Book, 1951-1952, 22 (1953).
“James Colquhoun Irvine, 1877-1952,” E. L. Hirst, Adv. Carbohydr. Chem. 8, xi
(1953).
“Studies on Seed Mucilages. Part VI. The Seed Mucilage ofPlantago arenaria,” E. L.
Hirst, (the late) E. G. V. Percival, and Clare B. Wylam,J. Chem. SOC., 189 (1954).
“Multiple Branching in Amylopectin,” E. L. Hirst and D. J. Manners, Chem. Ind.,
(London), 224 (1954).
“Hemicellulose A of Beechwood(Fagus sylvatica),” G. 0.Aspinall, E. L. Hirst, and R.
S . Mahorned,]. Chem. Soc., 1734 (1954).
“The Gum of Acacia pycnantha,” E. L. Hirst and A. S. Perlin, J. Chem. SOC., 2622
(1954).
“The Hudson Memorial Lecture. Claude Silbert Hudson,” E. L. Hint,]. Chem. SOC.,
4042 (1954).
“Physicochemical Studies on Starches. Part 11. The Oxidation o f Starches by
Potassium Metaperiodate,” D. M. W. Anderson, C. T. Greenwood, and E. L. Hirst,
J. Chem. SOC., 225 (1955).
“The Alkali-soluble Polysaccharides o f the Lichen Cladonia ulpestris (Reindeer
Moss),” G . 0. Aspinall, E. L. Hirst, and (Mrs.) Margaret Warburton,J. Chem. SOC., 651
(1955).
“Gum Ghatti (Indian Gum). The Composition o f the Gum and the Structure of Two
Aldobiouronic Acids Derived from It,” G. 0. Aspinall, E. L. Hirst, and A. Wickstrdm,
J. Chem. SOC., 1160 (1955).
“The Synthesis of Methyl Ethers of Mannuronic and Glucuronic Acid, and their
Reaction with Periodate,” R. A. Edington, E. L. Hirst, and Elizabeth E. Percival,
J . Chem. SOC., 2281 (1955).
“Edmund George Vincent Percival, 1907-1951,” E. L. Hirst and A. G. Ross, Ado.
Carbohydr. Chem. 10, xiii (1955).
“Some Problems in the Chemistry of the Hemicelluloses,” E. L. Hint,]. Chem. SOC.,
2974 (1955).
“The Constitution o f a Modified Starch from Malted Barley,” G. 0. Aspinall, E. L.
Hirst, and W. McArthur,J. Chem. SOC., 3075 (1955).
“The Analysis o f Plant Gums and Mucilages,” E. L. Hirst and J. K. N. Jones, in
“Modern Methods of Plant Analysis,” K. Paech and M. Y. Tracey, eds. Springer-
Verlag, Berlin, 1955, Vol. 11, p. 275.
 OBITUARY-EDMUND LANGLEY HIRST 27

“Plant Gums of the Genus Khaya. The Structure of Khaya grandifolia Gum,” G. 0 .
Aspinall, E. L. Hirst, and N. K. Matheson,]. Chem. SOC., 989 (1956).
“a-l:4-Glucosans. Part IV. A Re-examination of the Molecular Structure of Floridean
Starch,” I. D. Fleming, E. L. Hirst, and D. J. Manners,]. Chem. SOC.,2831 (1956).
“Carbohydrates,” E. L. Hirst, in “Perspectives in Organic Chemistry,” A. Todd, ed.,
Interscience, New York, 1956,p. 214.
“Some Aspects of the Chemistry of the Fructosaus,” E. L. Hirst, Proc. Chem. Soc.
London, 193 (July, 1957).
“A Comparison of Isolichenin and Lichenin from Iceland Moss (Cetraria islandica),”
N. B. Chanda, E. L. Hirst, and D. J. Manners,]. Chem. Soc., 1951 (1957).
“Periodate Oxidation of Laminarin,” F. B. Anderson, E. L. Hirst, and D. J. Manners,
Chem. Ind. (London), 1178 (1957).
“Changes in Organic Acid Content of Perennial Rye-grass During Conservation,” E.
L. Hirst and S . Ramstad, J. Sci. Food Agric., 8, 727 (1957).
“Introductory Remarks on the Chemical Reactivity of Sucrose,” E. L. Hirst, C . R .
Assemblie Comm. Intern. Tech. Sucrerie, lo“,Londres, 12 (1957).
“Gum Ghatti (Indian Gum). Part 11. The Hydrolysis Products Obtained from the
Methylated Degraded Gum and the Methylated Gum,” G. 0. Aspinall, (Mrs) Barbara J.
Auret, and E. L. Hirst,]. Chem. Soc., 221 (1958).
“The Constitution of Larch +Galactan,” G. 0.Aspinall, E. L. Hirst, and Else Ramstad,
J. Chem. Soc., 593 (1958).
“The Structure of Brachychiton diversifolium Gum (Sterculia caudata),” E. L. Hirst,
Elizabeth Percival, and R. S. Williams,]. Chem. SOC., 1942 (1958).
“The Constitution of Laminarin. Part 111.The Fine Structure of Insoluble Laminarin,”
F. B. Anderson, E. L. Hirst, D. J. Manners, and A. G. Ross,]. Chem. SOC., 3233 (1958).
“Gum Ghatti (Indian Gum). Part 111. Neutral Oligosaccharides Formed on Partial Acid
Hydrolysis of the Gum,” G. 0.Aspinall, (Mrs) Barbara J. Auret, and E. L. Hint,]. Chem.
SOC., 4408 (1958).
“The Gums and Mucilages of Plants,” E. L. Hirst and J. K. N. Jones, Handb.
Pjlanzenphysiol., 6,500 (1958).
“Structural Chemistry of Plant Polysaccharides with Reference to their Colloidal
Character,” E. L. Hirst, Verhandlungsber. Kolloid-Ges., 18, 104 (1958).
“Barry Degradation of Laminarin,” E. L. Hirst, J. J. O’Donnel1,and Elizabeth Percival,
Chem. Ind. (London), 834 (1958).
“The Presence of L-Guluronic Acid Residues in Alginic Acid,” D. W. Drummond, E.
L. Hirst, and Elizabeth Percival, Chem. Ind. (London), 1088 (1958).
“Chemical Structure in the Hemicellulose Group,” E. L. Hirst, Fundamentals Paper-
making Fibers, Trans. Symp. Cambridge, Engl. 1957, 93 (1958).
“Polysaccharides ofthe Marine Algae” (Presidential Address), E. L. Hirst, Proc. Chem.
Soc. London, 177 (July, 1958).
“Seaweed Mucilages,” E. L. Hirst,Abstr. Int. Seaweed Symp., 3rd, Galway, 52 (1958).
‘‘Analytical Studies on the Carbohydrates of Grasses and Clovers. IX. Changes in
Carbohydrate Composition during the Growth oflucerne,” E. L. Hirst, D. J. Mackenzie,
and Clare B. Wylam,]. Sci. Food Agric., 10, 19 (1959).
“The Structure ofAcacia pycnantha Gum,” G. 0.Aspinall, E. L. Hirst, and A. Nichol-
son,J. Chem. Soc., 1697 (1959).
“Molecular Structure in the Polysaccharide Group,” E. L Hirst, Proc. R. SOC. London,
Ser. A , 252,287 (1959).
“Studies on Uronic Acid Materials. 11. The Variation in Composition of Gum Nodules
from Combreturn leonense,” D. M. W. Anderson, E. L. Hirst, and N. J. King, Talanta, 3,
118 (1959).
28 MAURICE STACEY AND DAVID J. MANNERS

“Plant Gums,” E. L. Hirst. Proc. Int. Congr. Biochem., 4th, Vienna, 1958,1,31(1959).
“Studies on the Metabolism of the Protozoa. Part VIII. The Molecular Structure of a
Starch-type Polysaccharide from Chilomonas paramecium,” A. R. Archibald, E. L. Hirst,
D. J. Manners, and J. F. Ryley,]. Chem. Soc., 556 (1960).
“Studies on the Metabolism of the Chrysophyceae. Comparative Structural Investiga-
tions on Leucosin (Chrysolaminarin) Separated from Diatoms and Laminarin from the
Brown Algae,” Anne Beattie, E. L. Hirst, and Elizabeth Percival, Biochem. J., 79,
531 (1961).
“The Structure of Polysaccharides,” E. L. Hirst, Biochem. SOC.Symp., No. 21, 45
( 1962).
“The Constitution of Alginic Acid,” D. W. Drummond, E. L. Hirst, and Elizabeth
Percival,]. Chem. SOC.,1208 (1962).
“The Position ofMannit01 in Laminarin,” W. D. Annan, E. L. Hirst, and D. J. Manners,
Chem. Ind. (London), 984 (1962).
“The Acidic Sugar Components ofCochlospemnum gossypium Gum,” G. 0.Aspinall,
E. L. Hirst, and (Miss) Margaret J. Johnston, J. Chem. SOC.,2785 (1962).
“The Chemical Structure of the Hemicelluloses,” E. L. Hirst, Pure Appl. Chem., 5,53
(1962).
“Structural Studies ofAlginic Acid,” E. L. Hirst, E. Percival, and J. K. Wold,Chem. Ind.
(London), 257 (1963).
“The Location of L-Rhamnopyranose Residues in Gum Arabic,” G. 0. Aspinall, A. J.
Charlson, E. L. Hirst, and R. Young,]. Chem. Soc., 1696 (1963).
“Methyl Ethers of Mono- and Disaccharides,” E. L. Hirst and Elizabeth Percival,
Methods Carbohydr. Chem., 2, 145 (1963).
“Glycofuranosides from Cyclic Carbonates,” E. L. Hirst and Elizabeth Percival,
Methods Carbohydr. Chem., 2,349 (1963).
“Professor John Read, F.R.S.,” E. L. Hirst, Nature, 198,336 (1963).
“Professor John Read, 1884-1963 (Obituary Notice),” E. L. Hirst. Proc. Chem. SOC.
London, 353 (1963).
“John Read, 1884-1963,” E. L. Hirst, Biogr. Mem. Fellows R. SOC.,9,237 (1963).
“The Structure of Alginic Acid. Part IV. Partial Hydrolysis of the Reduced Polysac-
charide,” E. L. Hirst, Elizabeth Percival, and J. K. Wold,J. Chem. Soc., 1493 (1964).
“The Constitution of Laminarin. Part IV. The Minor Component Sugars,” W. D.
Annan, Sir Edmund Hirst, and D. J. Manners,J. Chem. Soc., 220 (1965).
“The Constitution of Laminarin. Part V. The Location of 1,6-Glucosidic Linkages,” W.
D. Annan, Sir Edmund Hirst, and D. J. Manners,]. Chem. SOC., 885 (1965).
“The Structure of Alginic Acid. Part V. Isolation and Unambiguous Characterization of
Some Hydrolysis Products of the Methylated Polysaccharide,” Sir Edmund Hirst and D.
A. Rees,]. Chem. SOC.,1182 (1965).
“The Water-soluble Polysaccharides of Cladophora rupestris and of Chaetomorpha
spp., Part 11. The Site of Ester Sulphate Croups and the Linkage between the Galac-
tose Residues,” Sir Edmund Hirst, W. Mackie, and Elizabeth Percival,]. Chem. Soc.,
2958 (1965).
“Seed Polysaccharides and their Role in Germination. A Survey of the Polysaccharide
Components ofMustard Seeds with Special Reference to the Embryos,”E. L. Hirst, D. A.
Rees, and N. G. Richardson, Biochem. I. , (1965).
95,453
“The Role of Sugars as Energy Reserves in Nature,” Sir Edmund Hirst,]. R. Soc. Arts,
64,290 (1966).
“Studies on Uronic Acid Materials. Part XVII. Some Structural Features of Acacia
senegal Gums (Gum Arabic),” D. M. W. Anderson, Sir Edmund Hirst, and J. F. Stoddart,
/. Chem. Soc., 1959 (1966).
 OBITUARY-EDMUND LANGLEY HIRST 29

“Review of the Chemistry of Water-soluble Gums,” Sir Edmund Hirst, in Monograph

No. 24 (“The Chemistry and Rheology of Water-soluble Gums and Colloids”), The So-
ciety of Chemical Industry, London, 1966, p. 3.
“Studies on Uronic Acid Materials. Part XVIII. Light-scattering Studies on Some
Molecular Weight Fractions fromAcacia senegal Gum,” D. M. W. Anderson, Sir Edmund
Hirst, S. Rahman, and G. Stainsby, Carbohydr. Res., 3,308 (1967).
“Studies on Uronic Acid Materials. Part XXI. Some Structural Features of Acacia
arabica Gum,” D. M. W. Anderson, Sir Edmund Hirst, and J. F. StoddartJ. Chenz. Soc.,
1476 (1967).
“The Ongins of Chemistry,” by Robert P. Multhauf (book review), J . R . Soc. Arts,
116,75 (1967).
“Studies on Uronic Acid Materials. -11. Some Structural Features of the Gum
Exudate from Acacia seyal Del.,” D. M. W. Anderson, I. C. M. Dea, and Sir Edmund
Hirst, Carbohydr. Res., 8, 460 (1968).
“Characterisation of Polysaccharide Structures by Glycoside Stabilisation with
Toluene-p-sulphonates: Model Experiments with Dextran,” D. A. Rees, N. G.
Richardson, N. J. Wight, and Sir Edmund Hirst, Carbohydr. Res., 9, 451 (1969).
“Stanley Peat,” Sir Edmund Hirst, Biogr. Mem. Fellows H. Soc., 16,441 (1970).
“a-(1+4)-~Glucans. Part XXI. The Molecular Structure of Starch-type Polysac-
charides from Haemutococcus pluvialis and T e t r a s e h i s carteri$omis,” Sir Edmund
Hirst, D. J. Manners, and I. R. Pennie, Carbohydr. Res., 22, 5 (1972).
“Paul Karrer-Obituary Notice,” R. Soc. Edinburgh, Year Book, 1971-72,50 (1972).
“R. C. Menzies-Obituary Notice,” R. Soc. Edinburgh, Year Book, 1973, 42 (1973).
This Page Intentionally Left Blank
 CARBOHYDRATE BORONATES

BY ROBERTJ . FEFWER
Department of Chemistry. Victoria University of Wellington.Wellington.New Zealand

I . Introduction .......................................................... 31
I1. Synthesis of Boronates ................................................ 37
1. Direct Condensation of Carbohydrates with Boronic Acids . . . . . . . . . . . . 37
2 . From Borinates .................................................... 39
I11. Structures of Carbohydrate Boronates .................................. 41
1. Alditol Boronates .................................................. 42
2 . Sugar Boronates .................................................... 43
3 . Boronates of Glycosides, Nucleosides. and Related Compounds . . . . . . . 45
IV . Boronates in Chemical Reactions ...................................... 48
1. Boronates in Aqueous Systems ...................................... 48
2 . Removal of Boronate Groups ........................................ 52
3. Stability of Boronates during Chemical Reactions .................... 53
.
V Separations of Carbohydrates by Use of Their Boronates ................ 57
1. Use of Isolated Boronates ........................................... 57
2 . Use in Paper Chromatography ...................................... 58
.
3 Use in Electrophoresis ............................................. 62
4 . Use in Column Chromatography .................................... 63
5 . Use in Gas-Liquid Chromatography ................................. 65
VI . Mass Spectrometry of Boronates ....................................... 65
VII . Nuclear Magnetic Resonance Spectroscopy of Boronates ................ 70
VIII . Borinates ............................................................ 70
IX . Tables ............................................................... 71
X . Addendum ........................................................... 80

I . INTRODUCTION
Condensation between D-ghCOSe and boric acid in acetone. in the
presence of a strong-acid catalyst. gives the discrete. cyclic ester 1.2-0 .
isopropylidene-a-D-glucofuranose3.5.boratel (1). which may be used
in the preparation of 6-substituted derivatives of D.glucose . Such
instances of the isolation and syntheti.c application of specific
carbohydrate borates are. however. few. because of the complexity of
.
boric acid as an esterifying agent As well as affording the 3.5.cyclic
ester 1. the aforementioned reaction could have led to ( a ) a dimeric

(1) L . Vargha. Ber., 66. 704-707 (1933).

31
32 ROBERT J. FERRIER

product containing boron-oxygen-boron linkages: (b) various dimers

linked by one boric acid unit, or (c) an intramolecular triester. With a
view to exploiting the clear potential of boric acid derivatives in
preparative, carbohydrate chemistry, but with reagents not subject to
such diverse means of reaction, the present autho9 began, in 1961, a
study of the condensation undergone between phenylboronic acid
[PhB(OH),I and the diol systems of various glycosides. At that time, it
had been found that free sugars interact with this acid in aqueous
soluti0n,4-~and several boronates of free sugars’ and alditols8.9 had
been isolated, but, despite initial efforts,s no discrete glycoside
derivatives had been reported, no successful structural work had been
carried out, and the esters had not found application in carbohydrate
chemistry. In the ensuing fifteen years, a wide range of specific
boronates, for example, 1,2-O-isopropylidene-a-D-g~ucofuranose 3,5-
phenylboronate (2), have been prepared, and many workers have

1 R=OH
2 R=Ph

shown how useful they can be as suitable crystalline derivatives, as

intermediates in synthesis, in a range of separatory techniques, and as
reaction catalysts. The present article surveys the progress made
during this time. The role of boronic acids in carbohydrate biochem-
istry is not dealt with, although, in plants, they have been found to
influence the storage of polysaccharides and aspects of their growth5

(2)W.Gerrard, M.F. Lappert, and B. A. Mountfield,J. Chem. SOC., 1529-1535 (1959);N.

N. Greenwood, in “Comprehensive Inorganic Chemistry,” J. C. Bailar,H. J.
Eniekus, R. Nyholm, and A. F. Trotman-Dickenson,eds., Pergamon Press, Oxford,
1973,Vol. 1, p. 895.
(3)R. J. Ferrier,]. Chem. SOC., 2325-2330 (1961).
(4)K. Torssell, Ark. Kemi, 10, 541-547 (1957);Chem. Abstr., 52, 14,555(1958).
(5)K. Torssell, J. H. McClendon, and G. F. Somers,Acta Chem. Scand., 12,1373-1385
(1958).
(6)J. P. Lorand and J. 0. Edwards,]. Org. Chem., 24,769-774 (1959).
(7)M.L.Wolfrom and J. Solms,J. Org. Chem., 21,815-816 (1956).
(8)H.G. Kuivila, A. H. Keough, and E. J. Soboczenski, J. Org. Chem., 19, 780-783
(1954).
(9)J. M.Sugihara and C. M. Bowman,]. Am. Chem. SOC., SO, 2443-2446 (1958).
 CARBOHYDRATE BORONATES 33

Boron, having the electronic configuration ls22s22p,has 3 valence

electrons, and forms planar, tricovalent derivatives that are electron
deficient, and which, as Lewis acids, accept two electrons from bases to
complete the boron outer-shell octet and give tetrahedral adducts.
Boric acid exemplifies this behavior by ionizing, in aqueous solution,
not by direct deprotonation, but by hydration and subsequent
ionization, to give the symmetrical borate anion:
-
B(OH)B+ H 2 0 B(OH)4+ H+.

Wide attention has been given to the interaction of carbohydrates with

these ions,l’Jdespite the complexity of the systems produced; reaction
of a conformationally unrestricted, contiguous trio1 could, for instance,
result in the formation of many species exemplified by the following.

B B O H PfoH ’ ‘OH 0
‘

With boronic acids, formation of dimeric species is essentially

precluded, but cyclic products having different ring-sizes and
electronic configuration at boron can b e formed, and it is here
suggested that isomerization of certain cyclic boronates may conse-
quently occur much more readily than has been recognized. In the case
of the general triols (3, the five-membered boronates (4) could
conceivably rearrange to the six-membered structures (5) by way of
accessible ionic intermediates (6),and thus 4 and 5 may be thought of as

(10) J. Boeseken,Adu. Carbohydr. Chem., 4,189-210 (1949);A. B. Foster, ibid.,12,81-

115 (1957); T. E. Acree,Adu. Chem. Ser., 117, 208-219 (1973); J. Dale,./. Chem.
SOC., 922-930 (1961); E. W. Malcolm, J. W. Green, and H. A. Swenson, ihid.,
4669-4676 (1964).
34 ROBERT J. FERRIER

4 5 6

tautomers, as well as isomers, bearing the same relationship to each

other as do furanose and pyranose forms of free sugars, and thereby
presenting classically difficult problems of structural analysis. In the
chemistry of alditol boronates, this issue is potentially of wide
significance (see, for example, the inconsistencies reported in the
chemistry of glycerol phenylboronate; Section III,l), but fortunately,
boronates of cyclic carbohydrates are not susceptible (except,
conceivably, in such compounds as ribopyranoside esters), and most of
the work described in this Chapter relates to discrete compounds of
tricovalent boron-all of which can, however, form tetrahedral
complexes with Lewis bases.
The relevant bond-lengths and bond-angles associated with trigonal
and tetrahedral boron in boronates are as shown. Consequently, both

can be accommodated in strainless, six-membered rings formed from

l,Sdiols, whereas five-membered, cyclic boronates are free from angle
strain only when the boron is tetrahedral. Thus, esters formed from
vicinal diols are more strained,ll and exhibit a stronger tendency to
react with bases12 (which frequently leads to relatively easy hydrol-
y s i P ) , than their six-membered counterparts. These characteristics
(11) A. Finch, P. J. Gardner, P. M. McNamara, and G. R. Wellum,J. Chem. Soc., A, 3339-
3345 (1970).
(12) A. Finch and J. C. Lockhart,J. Chem. SOC., 3723-3726 (1962).
(13) R. A. Bowie and 0. C. Musgrave,J. Chem. SOC., 3945-3949 (1963).
 CARBOHYDRATE BORONATES 35

therefore parallel, as expected, those of five- and six-membered, cyclic

borates.14 Apart from this generalization, however, the question of
stabilities of carbohydrate boronates is complex, because, not only are
they dependent upon ring size but also upon the carbon-bonded
radicals (ring strain in alkylboronates being substantially higher than
in aryl analogs1'), and, notably, upon the presence and type of
substituent groups on the ester rings and elsewhere in the molecules.
Groups that inhibit coordination of water with boron stabilize the
esters towards hydrolysis,l3 as, especially, do groups which can
themselves act as fourth ligands. As, in carbohydrate boronic esters,
oxygen atoms are frequently available for this function, this specific
means of stabilization is of great importance; for example, in
determining the degree of complexing of polyhydroxy compounds in
aqueous systems and, consequently, their mobility as boronates on
chromatograms or electrophoretograms (see Section V). Specific
solvation, such as occurs15 in the complex 7, is also likely to influence
ester stability to some extent.

Cyclic boronic esters can be formed readily from simple, acyclic 1,2-,
1,3-, and 1,4-diolsYand five-, six-, and, to a lesser extent, seven-
membered esters are, consequently, frequently encountered. Initial
attempts9 to obtain an eight-membered, cyclic product from 1,5-
pentanediol were not successful, presumably because linear diesters
or macrocyclic products preponderated, but it has been shown by mass-
spectrometric methods16 that some of the cyclic phenylboronate can be
formed and, furthermore, 1,6-hexanediol affords some nine-mem-
bered, cyclic phenylboronate. In model, cyclic systems, cis-lY2-diols
on cyclopentane and cyclohexane rings readily give bicyclic boronates,
whereas trans-related analogs react with two molar equivalents of
boronic acids, to give seven-membered, cyclic diesters (for example, 8
(14) A. J . Hubert, B. Hartigay, and J. Dale,J. Chem. SOC., 931-936 (1961).
(15)J. D. Morrison and R. L. Letsinger,J. Org. Chem., 29, 3405-3407 (1964).
(16)E. J. Bourne, I. R. McKinley, and H. Weigel, Carbohydr. Res., 35,141-149 (1974)
36 ROBERT J. FERRIER

and 9) of diboronic acids.9 In the case of 1,3- and 1,4-cis-cyclo-

hexanediols, forcing conditions permit the synthesis of cyclic
boronate s.l 7

8 n=l
9 n=2

Esters may be prepared from a range of boronic acids that vary

widely in their properties, including their aciditiesls (boric acid K, 6.53
x 10-10, phenylboronic acid, 13.7 x and butylboronic acid, 0.18
x 10-10), but most of the known carbohydrate boronates are phenyl
esters, although substituted-aryl compounds have received some slight
attention. Butyl, ethyl, and methyl analogs, although seldom crystal-
line compounds, unlike the majority of carbohydrate phenylboronates,
have been utilized mainly because of their enhanced volatility and
their consequent value in gas-liquid chromatography and mass
spectrometry.
Naming of carbohydrate boronates may be based on three principles:
( a ) ,the central use of the ester heterocyclic rings: 1,3,2-dioxaborolane
(five-membered), lY3,2-dioxaborinane(six-membered), and 1,3,5,2,4-
trioxadiborepane (seven-membered, for example, 8 and 9); (b),the use
of radical prefixes: “borylene” (Chem. Abstr.) or “boranediyl”
(Z.U.P.A.C.) for )BH; or (c), ester terminology. Thus, for example,
according to these respective procedures, the glycerol derivative 10

may be named ( a ) 5-hydroxy-2-phenyl-l,3,2-dioxaborinane, ( b ) 1,3-0-

(phenylborylene)glycerol,or 1,3-0-(phenylboranediyl)glycerol,or ( c )
glycerol lY3-phenylboronate.In this Chapter, the third method will be
adopted, and, in accordance with current, Chemical Abstracts usage,
(17) W. V. Dahlhoff and R. Koster,Justus Liebigs Ann. Chem., 1625-1636 (1975).
(18) G. E. K. Branch, D. L. Yabroff, and B. Bettman,]. Am. Chem. SOC., 56,1850-1857
(1934).
 CARBOHYDRATE BORONATES 37

alkyl- and aryl-boronate (rather than alkane- and arene-boronate)

terminology will b e used.
A brief summary is included of carbohdyrate borinates-acyclic
esters derived from borinic acids (R,BOH)-which have received less
attention than the related boronates.
Earlier, short reviews have dealt with m o n o ~ a c c h a r i d eand
~~
nucleoside phenylboronates,20 and the application of phenylboronic
acid in carbohydrate chemistry.21

11. SYNTHESIS OF BORONATES

1. Direct Condensation of Carbohydrates with Boronic Acids
Boronic acids readily undergo self-condensation to give cyclic
anhydrides, and either these trisubstituted boroxins (ll),or the acids
themselves, react spontaneously, but not necessarily completely, in
solution, with suitable diols, to give cyclic boronates 12 (see Scheme l),

\
/ 12

11
Scheme 1

for the isolation of which the water and solvent have simply to be
removed. Frequently, a suitable procedure involves treatment of the
carbohydrate with the required number of molar equivalents of the
acid or anhydride in boiling benzene, and azeotropic removal of the
water formed (which can conveniently be collected in a Dean-Stark
distillation head, and, provided that sufficient quantities are evolved,
measured, in order to monitor the progress of the reaction). Methyl p-D-
xylopyranoside (16.8 g) treated in this way with triphenylboroxin (10.6
g, 0.33 mol. equiv.) liberates water (1.8ml) during conversion into its
(19) A. M. Yurkevich and S. G. Verenikina, Vitam. Vitam.Prep., 247-256 (1973);Chem.
Abstr., 80, 48,2474 (1974).
(20) I. I. Kolodkina and A. M. Yurkevich, Vitam. Vitam. Prep., 257-268 (1973);Chem.
Abstr., 79, 146,774h (1973).
(21) R. J . Ferrier, Methods Carbohydr. Chem., 6,419-426 (1972).
38 ROBERT J. FERRIER

2,4-phenylboronate7 which may be isolated in excellent yield by

removal of most of the solvent and addition of dry, light petroleum.22
Some carbohydrate derivatives are too insoluble in benzene to be so
esterified, and, for them, a suitable, alternative solvent is 174-dioxane,
the water azeotrope of which may, by careful, fractional distillation be
removed before the solvent, and the water evolved in the reaction
determined, if desired, by the Karl Fischer method. Whereas methyl
a-D-glucopyranoside can be esterified by the benzene procedure, the
analogous a-D-mannoside ester must be prepared by use of 1,Cdioxane
or by some other variation.3
A further method applicable to benzene-insoluble carbohydrates
involves the addition of the compounds, in water, to triphenylboroxins
(preferably in methanoP), and, under these conditions, some boronic
esters have been found to precipitate. Otherwise, the reactants may be
fused under vacuum7 (a procedure not in current use, and not
recommended) or treated in such solvents as acet0ne,~*23,24 or 2-
methoxyethanol.25~28Particularly for the formation of nucleoside
boronates, pyridine has often been used as the solvent, and the
reactants have been heated at the boiling point, or in sealed tubes27-30;
on occasion, N7N-dimethylformarnide2s*31has been employed. A
feature of the use of pyridine is the occasional isolation-not
surprisingly-of the products as pyridine complexes.32
A slight modification of the reaction in acetone incorporates use of
sulfuric acid as a catalyst and provides, from the corresponding, free
sugars, moderately efficient, one-step procedures for obtaining 1,2-0-
isopropylidene-a-D-xylofuranose3,5-phenylboronate and the D-glU-

(22) R. J. Ferrier, D. Prasad, A. Rudowski, and I. Sangster,]. Chern. SOC.,3330-3334

(1964).
(23) E. J. Bourne, E. M. Lees, and H. Weigel,]. Chem. Soc., 3798-3802 (1965).
(24) A. M. Yurkevich, S. G . Verenikina, E. G . Chauser, and N. A. Preobrazhenskii, Zh.
Obshch. Khim., 36,1746-1749 (1966).
(25) S. G. Verenikina, A. M. Yurkevich, and N. A. Preobrazhenskii, Zh. Obshch. Khim.,
37,2181-2184 (1967).
(26) A. S. Guseva, S. G . Verenikina, S. F. Dymova, and A. M. Yurkevich, Zh. Obshch.
Khim., 44,2327-2331 (1974).
(27) A. M. Yurkevich, L. S. Varshavskaya, I. I. Kolodkina, and N. A. Preobrazhenskii,Zh.
Obshch. Khim., 37,2002-2006 (1967).
(28) A. M. Yurkevich, I. I. Kolodkina, L. S. Varshavskaya, V. I. Borodulina-Shvetz, I. P.
Rudakova, and N. A. Preobrazhenskii, Tetrahedron, 25,477-484 (1969).
(29) J. J. Dolhun and J. L. Wiebers,J. Am. Chem. SOC.,91, 7755-7756 (1969).
(30) S. A. Ermishkina and A. M. Yurkevich, Zh. Obshch. Khim., 40,652-655 (1970).
(31) A. M. Yurkevich, V. I. Borodulina-Shvetz, I. I. Kolodkina, and N. A. Preobrazhen-
skii, Zh. Obshch. Khim., 37,21762180 (1967).
(32) P. J. Wood and I. R. Siddiqui, Carbohydr. Res., 36,247-256 (1974).
 CARBOHYDRATE BORONATES 39

cose analog, 1,2-0-isopropylidene-~-~-arabinopyranose 3,4-phenyl-

boronate and the a-D-galactose analog, and 2,3-0-isopropylidene-D-
mannohranose 5,6-~henylboronate.~~
Koster and Dahlhoff introducedsa a variation by using, as
the esterifying agent, the ethylboronic anhydride derivative
(Me,CCO,BEt),O.

2. From Borinates
1,2-Ethanediol undergoes reaction at 380” with trimethylborane, to
give 2-methyl-1,3,2-dioxaborolane34(14), presumably by way of the
intermediate dimethylborinate (13) (see Scheme 2), and Russian
workers unexpectedly encountered an analogous reaction during

CH,OH
I
CH,OH
+ Me,B - H,COBMe,
I
CH,OH
-Hzlo\BMe
H, 0’

13 14
Scheme 2

heating of ribonucleosides in pyridine with equimolar amounts of

isobutyl diphenylborinate, as, instead of the borinic esters anticipated
(presumably 5’), they obtained almost quantitative yields of the 2’,3’-
phenylboronate~.35-~* The procedure may be used for preparative
purposes, and has the advantage of not producing water as the
byproduct.
German workers independently found that 1,2- 1,3-, and some 1,4-
diols can be converted into bis(diethy1borinates) which, on heating to
-200°, lose triethylboron to give cyclic ethylboronates. Otherwise,
these boronates can be prepared directly from the diols by heating with
triethylboron in the presence of diethylboryl pivalate, or by dismuta-
tion from equimolar proportions of diols and their bis(diethylbori-

(33) B. E. Stacey and B. Tierney, Carbohydr. Res., 49, 129-140 (1976).

(33a) R. Koster and W. V. Dahlhoff, Justus Liebigs Ann. Chem., 1925-1936 (1976).
(34) D. W. Wester, F. Longcor, and L. Barton, Synth. Inorg. Met.-Org. Chem.,3,115-123
(1973).
(35) A. M. Yurkevich, L. S. Varshavskaya, and I . I. Kolodkina, Zh. Obshch. Khim., 38,
21 15 (1968).
(36) I . I. Kolodkina, A. S. Guseva, E. A. Ivanova, L. S. Varshavskaya, and A. M.
Yurkevich, 2%. Obshch. Khim., 40,2489-2493 (1970).
(37) V. N. Rekunova, I. P. Rudakova, and A. M. Yurkevich,Zh. Obshch. Khim., 44,1182-
1187 (1974).
(38) V. N. Rekunova, I. P. Rudakova, and A. M. Yurkevich, Tetrahedron Lett., 281 1-2814
(1973).
40 ROBERT J. FERRIER

nates), and these procedures have been used to obtain ethylboronates

of simple diols,17 triols and tetrols,39 alditols,40-42 and the cis-isomers
of 1,2-, 1,3-, and 1,4-cyclohexanediol.~~ Pyrolysis of acyclic 1,4- and 1,5-
diols as their diethylborinates did not give seven- and eight-
membered, cyclic products but, instead,l7 the macrocycles 15 and 16.

Et
15 n = 4
l6n=5

In the case of the alditols, the per(diethylboriny1) derivative may be

selectively converted into mixed borinate-boronates, the D-mannitol
derivative giving D-mannitOl 1,2,5,6-tetrakis(diethylborinate) 3,4-
ethylboronate (17), which can be isolated prior to its conversion into
the 1,2:3,4:5,6-triboronate,4'and the galactitol ester giving galactitol
1,6-bis(diethylborinate) 2,3:4,5-bis(ethylboronate)(18) prior to the
formation of galactitol 176:2,3:4,5-tris(ethylboronate) (19) having an

%
BEt,

%
Et,BOCH, ,OBEt, Et,BOCH, ,OBEt,

17 18

Ef
19

(39) W. V. Dahlhoff and R. Koster, Justus Liebigs Ann. Chem., 1914-1925 (1975).
(40) W. V. Dahlhoff and R. Koster,Justus Liebigs Ann. Chem., 1926-1933 (1975).
(41) W. V. Dahlhoff, W. Schussler, and R. Koster,Justus Liebigs Ann. Chem., 387-394
(1976).
(42) W. V. Dahlhoff and R. KosterJ. Org. Chem., 41, 2316-2320 (1976).
 CARBOHYDRATE BORONATES 41

unusual, nine-membered ring.42 Selective deborination provides a

means of obtaining, from compounds 17 and 18, D-mannitol 3,4-
e t h y l b ~ r o n a t and
e ~ ~ galactitol 2,3:4,5-bi~(ethylboronate).~*
Koster and
Dahlhoff have reviewed their work with carbohydrate ethylboron
esters.qa
From this work, it follows that boronates of polyhydric alcohols can
be prepared directly by heating with trialkyl- or triaryl-boranes.
Galactitol heated with triphenylborane in refluxing toluene gives the
same tris(phenylboronate), in 93%yield22 as can be prepared by use of
phenylboronic acid.9
It seems probable that a fundamental distinction may exist between
the two methods available for the synthesis of carbohydrate boronates,
in that, under dehydration conditions, esterifications may be revers-
ible, whereas boronate formation from borinates is not. In the first,
therefore, products may be thermodynamically controlled, whereas
kinetic control will operate in the second. It is noteworthy, however,
that products obtained from several glycopyranosides (a) by use of
an ethylboronate derivative, and (b) by way of the per(diethy1bor-
inates) had the same

111. STRUCTURES OF CARBOHYDRATE BORONATES

Chemical methods for determining boronate structures usually
involve the substitution of unesterified hydroxyl groups, and the
characterization of the products, either by examination of the
compounds obtained after removal of the boronate groups, or by
independent synthesis of the fully substituted compounds by
boronation of known derivatives. Occasionally, the positions of
sulfonic ester groups in fully substituted boronates have been
determined by means of nucleophilic displacement-reactions. Al-
though these methods are often well suited to derivatives containing
one boronate group, they cannot always be applied unambiguously if
more than one boronate group is present. The most useful physical
methods of structural analysis, apart from X-ray diffraction, are nuclear
magnetic resonance (n.m.r.) spectroscopy and mass spectrometry.
Newer applications of 13C- and 'lB-n.m.r. spectroscopy (see Section
VII) indicate their value in the determination of the ring size of
boronates, and mass spectrometry (see Section VI) is of importance
largely for the same reason. Proton n.m.r. (p.m.r.) spectroscopy is of

(42a) R . Koster and W. V. Dahlhoff,Am. Chem. SOC. Symp. Ser., 39, 1-21 (1977).
42 R O B E R T J. FERRIER

great, general significance, and on occasion, infrared spectroscopy has

been a useful structural tool, as it can offer means of determining the
nature of the intramolecular hydrogen-bonding in which unsubstituted
hydroxyl groups may be engaged, and this may permit structural
assignment.

1. Alditol Boronates
For some properties of alditol boronates, see Table V.
Glycerol phenylboronate (m.p. 75.5-76.5') is obtainable in high
yield,l3 and was assigned the 1,2-cyclic structure on examination of the
glycerol N-phenylcarbamate formed from it by treatment with phenyl
isocyanate, followed by removal of the boronate ring.23 Consistent
with this conclusion are the characteristics of hydrolysis of the parent
ester13 and the fact that phenylboronic acid complexes more strongly
with 1,2- than with 1,3-diols.6 However, a re-inve~tigation~~ of the
glycerol phenylcarbamate revealed that it reduces 0.84 molar
equivalent of periodate, instead of the 1 molar equivalent expected,
which was taken to indicate that it was contaminated with some of the
unoxidized, 2-substituted isomer. A new procedure for structural
analysis was then developed to check this conclusion. Glycerol
phenylboronate was methylated with diazomethane in dichlorometh-
ane containing boron trifluoride etherate, the boronate ring was
removed, and the resultant diols were acetylated, to give two mono-0-
methylglycerol diacetates which were characterized, after g.1.c.
separation, by mass spectrometry. Surprisingly, this method revealed
that the initial phenylboronate was a mixture comprising the six-
membered ester as the main product.
In the present author's opinion, the control experiment conducted on
the critical methylation step within the foregoing procedure may not
have been sufficiently rigorous, on the grounds that methylation of a-D-
glucofuranose 1,2:3,5-bis(phenylboronate)with diazomethane-boron
trifluoride may not establish that all boronates react without change
under these conditions. As was indicated in the Introduction, chemical
methods may be seriously deficient for characterizing such com-
pounds, and techniques that do not disturb delicate chemical states
may have to be used in just such cases. However, when applied to
glycerol phenylboronate, "B-n.m.r.-spectral analysis did not unam-
biguously resolve the issue either, because, at 20', the boron chemical-
shift indicated the presence of a six-membered ring, whereas, at 80', it
moved to a position indicative of a five-membered ring.26
Chemical procedures have to be used with the greatest care, and

(43) I. R. McKinley and H. Weigel, Carbohydr. Res., 31, 17-26 (1973).

 CARBOHYDRATE BORONATES 43

McKinley and Weigel's results43 with other triols should also be

considered with this in mind. By applying the diazomethane
methylation procedure, they also investigated the phenylboronates
obtained from 3-deoxy-~~-glycero-tetritol, 4-deoxy-~-erythritol,and
1,5-dideoxy-~-arabinitol,-ribitol, and -xylitol, and found that all except
that derived from the last trio1 were mixtures. Their conclusion was
that, where six-membered rings that do not have axial substituents can
be formed, their production is favored (glycerol, 3-deoxy-~~-glycero-
tetritol, 4-deoxy-~-erythritol,and 175-dideoxyribitol);in other cases,
substantial proportions of five-membered ring esters are produced.
However, other points that appear to increase suspicion of the validity
of the methylation procedure are: (i) the inconsistency between the
periodate and the methylation results, as applied to glycerol
phenylboronate, ( i i ) the finding of the formation, from some triols, of
some thermodynamically unfavored, five-membered, cyclic products,
and ( i i i ) the conclusion that the phenylboronate obtained from 4-
deoxy-L-erythritol comprises all three possible isomers, despite its
being a sharp-melting compound.
Glycerol ethylboronate produced by the borinate method39 has the
five-membered, cyclic structure, consistent with its being a kinetically
controlled product, whereas that obtained from 3-deoxy-D~-glycero-
tetritol contains a six-membered ring.
From these findings with triols, it follows that, apart from the
expectation that formation of five- and six-membered rings would be
favored (see, however, the exceptional compound 19), no general
conclusions can be drawn regarding the structures of boronates
derived from more-complex polyhydric alcohols. In Table V, alditol
boronates are listed with structures when these can be concluded
either from the method of synthesis, from physical studies, or by
deduction (as with the 1,2:5,6-diesters formed from 3,4-di-O-substi-
tuted mannitols).

2. Sugar Boronates
For some properties of boronates of sugars, see Table 111.
Although boronates of sugars were amongst the first such carbohy-
drate esters to be prepared,' little structural work was performed until
the advent of n.m.r.-spectroscopic and mass-spectrometric methods.
The high-yielding condensations undergone by D-xylose and L-
arabinose with phenyl- and butyl-boronic acid have thus been shown
to give a-D-xylose 1,2:3,5-bis(phenyl-and butyl-boronate) (20)and p-L-
arabinose 1,2:3,4-bis(phenyl- and butyl-boronate)44 (21).

(44) P. J. Wood and I. R. Siddiqui, Curbohydr. Res., 33,97-104 (1974).

44 ROBERT J. FERRIER

20 R = Bu or Ph 21

In 1956, a ribose diphenylboronate (map.140-142") was prepared in

modest yield by a fusion method,' but it was later shown26 that an
almost quantitative yield of a mono-ester was obtainable by conducting
the condensation in hot 2-methoxyethanol, and this was a-D-ribose 2,4-
phenylboronate (22), as indicated by p.m.r. and "B-n.m.r.
and conversion by selective substitution into the S-p-toluenesulf-
~ n a t eIt. ~is ~here suggested that the 2,4-ester (22) might be very sus-

22

ceptible to structural change, as it could give the 1,2-, 1,3-, 2,3-, and
3,4-isomers by a set of tautomeric rearrangements by way of the acces-
sible, boronate anions (23 and 24).

23 24

Condensation of D-ribose with two molar equivalents of phenyl-

boronic acid in 2-methoxyethanol gave a diester to which was assigned
the 1,5:2,3-P-furanose structure (25) on the basis of as-yet-unreported
n.m.r. data.26

(45) M. G. Edelev, T. M. Filippova, V. N. Robos, I. K. Shmyrev, A. S. Guseva, S. G.

Verenikina, and A. M. Yurkevich, Z h . Obshch. Khim., 44,2321-2327 (1974).
(46) A. S. Guseva, I. P. Rudakova, S. G. Verenikina, and A. M. Yurkevich, Zh. Obshch.
Khim., 44, 1187-1193 (1974).
 CARBOHYDRATE BORONATES 45

D-Glucose gives a crystalline 172:3,5-bis(phenylboronate)(26)

Ph

phAFGy
&H 0

b-hPh
B
Ph
25 26

( ~ . m . r . ~and
* * 13C-n.m.r.45
~~ evidence) which can b e used to prepare 6-
substituted esters32.4' and ethers** of the sugar. Likewise, 1,2-0-
isopropylidene-a-D-glucose gives a crystalline 3,5-phenylboronate724.
32*33 and, with two molar equivalents of the boronating agent, a further
crystalline derivative, characterized as the 5,6-dib0ronate.~~ When a
substituent is present at C-3, as in 3-deoxy-3-fluoro-1,2-O-isopropyl-
idene-a-~-glucofuranose,4~ direct boronation takes place, as expected,
at the 5,6-diol.
D-Fructose reacts in the P-pyranose form, to give the 2,3:4,5-
bis(phenylboronate)32 (27), and 6-deoxy-a-~-galactoseaffordP the
stereochemically related 1,2:3,4-die ster (28).

Phb-0

27 R' = H, R2 = CH,OH
28 R' = Me. Rz = H

Little structural work has been performed on derivatives of other

hexoses, or higher sugars.

3. Boronates of Glycosides, Nucleosides, and Related Compounds

For some properties of boronates of glycosides, nucleosides, and
related compounds, see Tables IV, VI, and VII.
Glycosides and related compounds tend to give boronates easier to

(47) L. G. Mogel and A. M. Yurkevich, Zh. Obshch. Khim., 39, 1882-1886 (1969).
(48) E. J. Bourne, I. R. McKinley, and H. Weigel, Carbohydr. Res., 25,516-517 (1972).
(49) A. B. Foster, R. Hems, and J. M . Webber, Carbohydr. Res., 5,292-301 (1967).
46 ROBERT J. FERRIER

characterize than corresponding derivatives of alditols and sugars,

because of the specific diol systems they present to the boronating
reagents. Most interestingly, cis-173-related diols on pyranoid rings
condense to give cyclic boronates [for example, methyl a-D-xylo-
pyranoside 2,4-phenylboronate (29) and N-(p-bromopheny1)-a-D-ribo-
pyronosylamine 2,4-phenylboronate (30)] which, in the ~ y l o s e ~ ~ , ~ ~ ~

Ph Ph
29 30

and ribose51 series, afford useful means for obtaining 3-substituted

derivative~2~*~~.52 and glycosid-3-uloses.53 The specific complexing
that occurs between boronic acids and contiguous cis,cis-triols (as in
ribopyranosides; see Section V) is postulated as resulting from
stabilization of the cyclic esters formed from 1,3-related7axial hydroxyl
groups by coordination from the oxygen atom of the central hydroxyl
groups. However, although the crystal structure of N-(p-bromophen-
y1)-a-D-ribopyranosylamine 2,4-phenylboronate (30) indicates that the
pyranoid ring is in the expected IC4conformation with the ester oxygen
bonds axial, it also reveals that the boron is trigonal and that 0 - 3 is too
far from it (305 pm; 3.05 A)* for coordination. The propensity for this
oxygen atom to co-ordinate is, however, satisfied by hydrogen bonding
with the N-bonded proton of an adjacent molecule.54
The a r a b i n o p y r a n ~ s i d e sand
~ ~ ~lyxopyranosides are esterified, as
expected, at the 3,4- and 2,3cis-diols7 respectively5’; in the ribo-
furanosyl series, esterification occurs at 0-2’,0-3’, and a range of
nucleoside phenylboronates is known, mainly through the work of
Yurkevich and his associates (see Table VII).
In the hexopyranoside series, 4,6-phenylboronates [for example,
methyl a-D-glucopyranoside 4,6-phenylboronate (31)l are obtain-

(50) R. J. Ferrier, D. Prasad,and A. Rudowski, Chern. lnd. (London), 1260-1261 (1964).

(51) R. J. Ferrier and D. Prasad,J. Chem. Soc., 7425-7428 (1965).
(52) R. J. Ferrier and D. Prasad,J. Chern. Soc., 7429-7432 (1965).
(53) B. Lindberg and K. N. Slessor, Curbohydr. Res., 1,492-493 (1966); Actu Chern.
S c u d . , 21, 910-914 (1967).
(54) H. Shimanouchi, N. Saito, and Y. Sasada, Bull. Chern. SOC.Jpn., 42, 1239-1247
(1969).

*Calculated by Dr. J. H. Johnston of this Department.

 CARBOHYDRATE BORONATES 47

able,3.33a,55
and residual diols may react further, to lead to fully sub-
stituted compounds [for example, methyl a-D-glucopyranoside
4,6-phenylboronate 2,3-(diphenyl~yclodiboronate)~*~~ (32)l. Chemical

PCHa p""'
PhBQ OHOMe
Ph\&PhB O, Me
b-B,
31 Ph
32

analyses55 indicated that the galactopyranosides likewise give 4,6-

esters in preference to the five-membered-ring derivatives that might
have been formed from the cis-related 3,4-diols. However, the
situation appears to be delicately balanced in compounds containing
such 3,4,6-triol groupings, as some undergo reaction to give 4,6-esters
(33), whereas others (presumably different in flexibility and time-
averaged conformation) afford mainly the five-membered derivatives
(34). Later work, based on mass spectrometry,56 suggested that the
esterifications of the methyl galactopyranosides are not so unambigu-
ous as suggested by the chemical methods,55and revealed that 3,4- as
well as 4,6-esters are produced. Here, again, it is possible that facile
isomerization of the initial products might have occurred during their
isolation or subsequent analysis, and, for this reason, the available
information should be considered with reservation.

"0
Phl3/OcH,

h.1 R'
R' R2 R4Rs R' R2 RS R4
OH H OMe H H H OMe H
OH H H OMe
H H H OMe
O H H H H
H H H H
33 34

(55) R. J. Ferrier, A. J . Hannaford, W. G. Overend, and B. C. Smith, Carbohydr. Res., 1,

38-43 (1965).
(56) V. N. Reinhold, F. Wirtz-Peitz, and K. Biemann, Carbohydr. Res., 37, 203-221
(1974).
48 ROBERT J. FERRIER

Methyl a-D-mannopyranoside appears to give mixed monoesters

(presumably 2,3- and 4,6-), but reacts readily to afford the 2,3:4,6-
d i e ~ t e P . ~ (35)
~ * ~with
~ * two
~ ' molar equivalents of acid, whereas, in the
6-deoxyhexoside series, reaction occurs at vicinal, cis-diol sites (2,3 for
methyl 6-deoxy-a-~-mannopyranoside,58 and 3,4 for methyl 6-deoxy-a-
~-galactopyranoside),5Bbut, again, a compound having a cis-related,
diol grouping at C-2 and C-4 gives a six-membered, cyclic ester. For
methyl 6-deoxy-/3-~-allopyranoside,this is somewhat surprising, in
view of the diaxial relationship between the groups at C-1 and C-5 in
the product58,60 (36).Methyl 6-deoxy-a-glucopyranosidelikewise

/b
gives59 the 2,4-ester (37).

ph\ dB,o
Me
@ Q
o

Me

-B'
35 B/O
Ph Ph

36 37
With the 1,6-anhydrohexopyranoses,condensation occurs at vicinal
cis-diols or, in the case of 1,6-anhydro-/3-~-glucopyranose,
at the diaxial
2,4-~ites,~ldespite the observation that this anhydride does not readily
complex with the acid (see Section V,2).

Iv. BORONATES
IN CHEMICAL REACTIONS

1. Boronates in Aqueous Systems

a. Interaction Between Carbohydrates and Boronic Acids in Aqueous
Media.-Initial1 y, Torssel14 detected interaction between phenyl-
boronic acid and D-fructose, and, by potentiometric titration with
sodium hydroxide solution, determined a formation constant for a 1:1
complex:
0 OH
PhB(OH), + D-fructose * (D-fructose)/ \ I

(57) D. S. Robinson, J. Eagles, and R. Self, Carbohydr. Res., 26,204-207 (1973).

(58)J. S.Brimacombe, F. Hunedy, and A. Husain, Carbohydr. Res., 10,141-151 (1969).
(59)J. S.Brimacombe, A. Husain, F. Hunedy, and M. Stacey,Ado. Chem. Ser., 74,56-69
(1968).
(60) J. S. Brimacombe and D. Portsmouth,]. Chem. SOC., C , 499-501 (1966).
(61) F. Shafizadeh, G.D. McGinnis, and P. S. Chin, Carbohydr. Res., 18,357-361(1971).
 CARBOHYDRATE BORONATES 49

With colleagues,5 he then proceeded to examine this complexing for a

range of substituted phenylboronic acids (as part of a study of their
influence in certain aspects of plant biochemistry), and found that, as
expected, electron-withdrawing groups on the aromatic ring increased
the stability of the complexes formed.
By examining pH depressions, other workers6 determined formation
constants for the phenylboronate complexes formed with various diols
and polyhydroxy compounds, and found that, of several sugars
examined, D-frUCtOSe formed much the most stable complex. Fourteen
years later, S. A. Barker and his colleagues62 conducted a detailed,
polarimetric analysis of the complexing undergone between D-
glucose, D-mannose, and D-fructose (separately) and phenylboronic
acid and its m-nitro and p-methoxy derivatives, respectively, and
showed that the favored complexing with the ketose is pH-dependent.
Their observations indicated that, with phenylboronic acid, D-glucose
is uncomplexed up to pH 6, and fully complexed beyond pH 9, whereas
D-fructose begins to form a complex near pH 5 and, in the experiment
reported, is -30% complexed before D-glucose begins to react. The
ketose is also fully complexed at, and above, pH 9. Similar effects were
observed for the substituted acids, the pH ranges within which partial
complexing occurs going to lower values with the m-nitrated acid and
to higher values with the p-methoxy compound.
It was then demonstrated63 that these arylboronic acids displace, in
favor of the ketose, the D-glucose-D-mannose-D-fructose pseudo-
equilibrium that is established in alkaline solution, so that the usual,
single-step, conversion efficiency of D-glUCOSe to D-fructose of -30%
may be increased to as high as 81%. A detailed investigation of the
effects of pH, temperature, and concentration on this phenomenon was
undertaken with a view to optimizing a commercial preparation of D-
fructose. This displacement may also be effected with polymeric
arylboronic acids which, as expected, also show differential binding of
the ketose.64
Yurkevich and coworkers,65 using the pH-depression procedure,
reported complexing constants of nucleosides and nucleotides, and
showed that the depressions are themselves pH-dependent, adenosine
exhibiting maximal effects near pH 7.8, 7.3, and 8.2 on mixing with

S. A. Barker, A. K. Chopra, B. W. Hatt, and P. J. Somers,Carbohydr. Res., 26,33-40

(1973).
S. A. Barker, B. W. Hatt, and P. J. Somers, Carbohydr. Res., 26,41-53 (1973).
S. A. Barker, B. W. Hatt, P. J. Somers, and R. R. Woodbury, Carbohydr.Res., 26,55-
64 (1973).
E. A.Ivanova, I. I. Kolodkina, and A. M. Yurkevich,Zh. Obshch. Khim., 41,455-459
(1971).
50 ROBERT J. FERRIER

phenylboronic acid and its p-nitro and p-methyl derivatives, respec-

tively. Here, it may safely be assumed that the 2’,3’-diol is involved
in complex-formation; in all of the other work referred to in this
Section, no specific evidence is available regarding the structures of
the complexes.

b. Hydrolysis of Boronates-There is very little good information

available on the stability to hydrolysis of an adequate range of
carbohydrate boronates. Some evidence indicates that at least certain
esters are stable in aqueous systems: phenylboronates of sugars and
alditols may be isolated by crystallization from aqueous m e t h a n ~ l ~ , ~ ~ ;
p-tolylboronates of various diols and of 1,2-0-(trichloroethylidene)-a-
D-glucofuranose can be precipitated by acidification of aqueous
alkaline solutions66; and 1,3-O-benzylidene-~-arabinitol gives a
phenylboronate that is apparently unchanged on heating in aqueous
sodium hydr0xide.6~It has been suggested9 that the first of these points
does not establish the stability, but rather the insolubility of the esters
in water, and the second point conceivably reflects the insolubility of
trigonal esters relative to their tetrahedral, anionic analogs. In the case
of the L-arabinitol derivative, after the hydrolysis, the solution was
extracted continuously with chloroform for several hours, and it is here
suggested that, during this treatment, phenylboronic acid and the
benzylidene acetal may have been separately taken into the chloro-
form, to recondense during the subsequent removal ofthis solvent. The
finding that phenylboronic acid could be removed at room temperature
by chromatographic separation on an anionic resin or on neutral
alumina adds support to the conjecture that hydrolysis had occurred in
the alkaline medium. It is, therefore, suggested that none of this
evidence establishes the stability of any ofthe esters in aqueous media.
All other evidence indicates that carbohydrate boronates readily
undergo hydrolysis on addition of water to their solutions in organic
solvents. Thus, early in the history of the compounds, it was found that
very mild hydrolysis of L-arabinose bis(phenylboronate)7 and D-
mannitol tri~(pheny1boronate)~ gave the respective starting-materials,
which could readily be separated from each other. In all cases, when
water has been added to solutions of boronates in dry solvents, optical
rotational changes in magnitude occur in directions consistent with
their being caused by hydrolysis. In this way, Hannaf0rd5~468

(66)Brit. Pat. 885,766(1960);Chem. Abstr., 56, 15,546f(1962).

(67)A.B.Foster, A. H. Haines, T. D. Inch, M. H. Randall, and J. M. Webber, Carbohydr.
Res., 1, 145-155 (1965).
(68)A. J. Hannaford, Ph.D. Thesis, University of London, 1964.
 CARBOHYDRATE BORONATES 51

calculated the equilibrium constants for the reaction

0
/ \
Carbohydrate BPh + H 2 0 * carbohydrate + PhB(OH),
\ /
0

to be 0.2 k O . 1 for methyl a- and P-D-glucopyranoside and their 2,3-di-

O-methyl derivatives and for D-glucal. This means that -4% of water
caused complete hydrolysis of the esters dissolved (2%)in 1,cdioxane.
Similar, semiquantitative work on methyl xyloside phenylboronates
revealed much more variable susceptibilities, the percentages of water
needed to cause complete hydrolysis ofthe esters (29,38,39,and 40) of

OMe

JpQ(
0
OH
\/O
Ph 39 R’ = H. R* = OMe
40 R1 = OMe, RZ = H
38

the a- and p-pyranoside and a- and p-furanoside (1%, in 174-dioxane)

being 1,3,9, and 30, r e s p e c t i ~ e l yAs
. ~ expected,
~ therefore, the highly
strained, pyranoside derivatives are highly susceptible to hydrolysis,
whereas the furanoside analogs, particularly the p anomers, are
appreciably more stable. Although the trans-relationships of the
groups at C-1 and C-2 of the p-furanoside (40) make it thermody-
namically the more stable, it does not seem likely that its marked
hydrolytic stability can be attributed to this factor alone (see later).
Other compounds examined by this type of procedure are the 2’,3’-
phenylboronates of nu~leosides,~7~28~3~ and the butyl- and phenyl-
boronates of D-xylose and L-arabinose,” and these, also, are hydrolyt-
ically unstable.
It is here concluded that carbohydrate boronic esters should
normally be treated as being readily susceptible to hydrolysis and,
presumably, alcoholysis, and that their occasional isolation in
crystalline form from aqueous solvents**23.70or ethanol67 should be

(69) R. J. Ferrier, D. Prasad, and A. Rudowski,J. Chem. SOC.,858-863 (1965).

(70)R. J. Ferrier and L. R. Hatton, Carbohydr. Res., 5, 132-139 (1967).
52 ROBERT J. FERRIER

viewed as evidence of their insolubility and not of their stability in

these solvents. Although it might have been expected that such
solvents as pyridine could stabilize the esters by co-ordination with the
boron atoms, this does not appear to be so44; however, intramolecular
co-ordination by suitably oriented oxygen atoms does stabilize them
considerably, and it is probably this factor that is responsible for the
observed character (see earlier) of methyl /3-D-xylofuranoside 3,5-
phenylboronate (40), in which 0 - 1 can specifically bond to boron.
Although no formal studies of hydrolyses of carbohydrate derivatives
stabilized in this way appear to have been reported, their stability as
revealed by their behavior in chromatography and electrophoresis is
well recognized; this is discussed in Section V,2 and 3.
Products of the partial hydrolysis of complex esters have not often
been isolated, but the very labile, seven-membered ring of the
diphenyldiboronate 32 can be selectively removed, to give methyl a-
D-ghcopyranoside 4,6-phenylboronate (31)in good yield.3 In their
work with alditol derivatives containing ethylboronate and diethyl-
borinate groups, Dahlhoff and K o ~ t e r ~ Ofound
- ~ ~ that the latter may be
selectively removed by treatment with either methanol or 2,4-
pentanedione and, in the case of xylitol 2-diethylborinate 1,3:4,5-
bi~(ethylboronate),4~ they were able to remove the six-membered ring
selectively. This is against expectations (see Section I), and may result
from intramolecular stabilization of the smaller ring by co-ordination
from 0-2.

2. Removal of Boronate Groups

When boronic esters are utilized as protecting groups in the
synthesis of specifically substituted, or otherwise modified, carbohy-
drate derivatives, the cleavage of the esters must be followed by
removal of the by-products liberated. With esters of strongly
hydrophilic, carbohydrate compounds, the boronic acids can be
specifically extracted from aqueous into organic solvents, and, in this
way, alditolss and free sugars' have been recovered from their
phenylboronates, but a more widely applicable procedure involves
exchanging the boronic acids from the carbohydrate to 1,e-ethane-
dio140,41*71
or, more usually, 1,3-propanediol,22*58~72with which they
condense to give volatile, cyclic derivatives. Addition of 1,3-pro-

(71)A. A. Amagaeva, A. M. Yurkevich, I. P. Rudakova, L. V. Khristenko, I. M.

Kustanovich, and N. A. Preobrazhenskii, Khim. Prfr. Soedfn., 4, 304-307 (1968);
Chem. Abstr., 70, 115,471s(1969).
(72)L.G.Mogel and A. M. Yurkevich, Zh. Obshch. Khim., 40,708(1970).
 CARBOHYDRATE BORONATES 53

panediol to a solution of a carbohydrate boronate in acetone, and

removal of the volatile products, usually offers a very efficient method
of deboronation.
Column separations of the products of hydrolysis of carbohydrate
boronates by use of anionic resins provides an alternative, efficient
means of deb0ronation,~~,67,69 and other similar separations have used
columns of cellulose,48 alumina,67 and, for carbonyl-containing
compounds, anion-exchange resins in the hydrogensulfite In
special cases, e l e c t r o p h o r e ~ i sand
~ ~ direct crystallizationz3 have been
employed.
An interesting, alternative procedure, applicable at least to the
removal of phenylboronic acid, involves its conversion, by treatment
with bromine-water, into bromobenzene and boric acid, and the
removal of these by distillation-the latter as its trimethyl ester.43

3. Stability of Boronates during Chemical Reactions

a. Esterificatioa-Successful acetylation of unsubstituted hydroxyl
groups in carbohydrate boronates has been reported on several
occasions. Acetyl chloride in pyridine has usually been used as the
acetylating agent, and the products have frequently been distilled prior
to crystallization. Acetates of boronates of glycosides,3~22*51*~5~~9
a l d i t o l ~and
, ~ ~ nucleosides3" have been reported.
Benzoylation is, likewise, usually a satisfactory process, but the
products are insufficiently volatile for distillation. Fully substituted
aldito1,41,67 and nucleoside36 esters have
been prepared, and, sometimes, the boronate groups have been
removed without isolation of the (benzoylated) first products.
An instance of the use of an insoluble poly(styry1boronic acid) in the
preparation of methyl 2,3-di-0-benzoy~-a-D-ghcopyranoside and
-galactopyranoside has been reported.72aUsually, benzoyl chloride is
the reagent used, but, with this, methyl a-D-xylopyranoside 2,4-
phenylboronate gave only 37% of the 3-benzoate7 whereas the yield
was doubled by use of benzoic anhydride.22
p-Toluenesulfonylation of methyl a-D-glucopyranoside 4,6-phenyl-
boronate did not give an isolable product, although the bis-p-
toluenesulfonate could be prepared by boronation of the appropriate D-
glucoside derivatives; othersZ3 have reported failure to esterify
galactitol bis( phenylboronate). Despite these findings, there is little
reason to doubt that p-toluenesulfonylation of boronates can be
satisfactorily achieved. Thus, direct products have been obtained from

(72a) E. Seymour and J. M. J. Frechet, Tetrahedron Lett., 1149-1152 (1976).

54 ROBERT J . FERRIER

alditol67 and sugar47 boronates having free primary-hydroxyl groups,

and secondary p-toluenesulfonates from 1,6-anhydroaldohexose bor-
onates61 (although the D-gdo compound was slow to react); D-ribose
2,4-phenylboronate gave,"6 selectively, the 3-p-toluenesulfonate at
-10". Several reports have appeared on the preparation of 5'-p-
toluenesulfonates of nucleoside 2',3'-phenylb0ronates,37,7~*73*74 but
these still remain poorly characterized derivatives.
The preparation of N-phenylcarbamates from incompletely substi-
tuted, carbohydrate boronates is, perhaps, the least destructive of
esterifying reactions, requiring only that the compounds be heated
with phenyl isocyanate in dry benzene or toluene. Several reports have
appeared on successful application to sugar,47and
aldito123 derivatives.
By using the 2',3'-phenylboronate protecting-group, Yurkevich
and his colleagues prepared 5'-phosphates of adeno~ine~28.75-77
uridine, 27,76 cytidineYz7
and g ~ a n o s i n e Usually,
.~~ 2-rnorpholino-1,l-
diphenylpyrophosphorochloridate has been used as the phosphoryla-
ting reagent, but the morpholinophosphorodichloridate and 2-cyano-
ethyl phosphate-N,N'-dicyclohexylcarbodiimide procedures have
also been employed.28 Nucleoside 2',3'-phenylboronates may also be
used in the synthesis of dinucleoside phosphates.30
Chloroacetylation can be effected by use of chloroacetic anhydride
in pyridine, and, in this way, the 6- and 3-esters of l,%O-isopropyli-
dene-a-D-glucofuranose were prepared, in good yield, by using the 3,5-
phenylboronate and 5,6-(diphenylcyclodiboronate), respectively, as
starting materials, and, from the chloroacetates, trimethylammonio-
acetyl salts were prepared.24In further work, connected with pangamic
acid, 6-O-(N,N-dimethylglycyl)-~-glucose was synthesized by use of
N,N-dimethylglycine and N,N'-dicyclohexylcarbodiimide in aceto-
nitrile-pyridine,25 and pangamolactone [6-0-(N,N-dimethylglycyl)-~-
glucono-1,4-lactone] by a similar procedure from D-glucono-1,4-
lactone 3,5-phenylboronate.78
(73) A. M. Yurkevich, A. A. Amagaeva, I. P. Rudakova, and N. A. Preobrazhenskii, Zh.
Obshch. Khim., 39,434-440 (1969).
(74) I. P. Rudakova,T. A. Pospelova, and A. M. Yurkevich,Zh. Obshch. Khim., 40,2493-
2499 (1970).
(75) A. M. Yurkevich, I. I. Kolodkina, and N. A. Preobrazhenskii, Dokl. Akad. Nauk
S S S R , 164,828-830 (1965);Chem. Abstr., 64,3661g (1966).
(76) I. I. Kolodkina, L. S. Varshavskaya,A. M. Yurkevich, and N. A. Preobrazhenskii,Zh.
Obshch. Khim., 37,1996-2002 (1967).
(77) A. M. Yurkevich, I. I. Kolodkina, G. S. Evdokimova, E. T. Bazhanova, and N. A.
Preobrazhenskii, Khim. Org. Soedin. Fosfora, Akad. Nauk SSSR, Otd. Obshch.
Tekh. Khim., 215-220 (1967);Chem. Abstr., 69, 10,667m (1968).
(78) K. Murase and M. Murakami, Yamanouchi Seiyaku Kenkyu Hokoku, 2, 62-65
(1974); Chem. Abstr., 83, 179,446p (1975).
 CARBOHYDRATE BORONATES 55

Methacrylic anhydride in pyridine, applied to 1,6-anhydro-~-

glucose 2,4-phenylboronate, gave the 3-methacrylate7from which an
addition polymer was prepared.79

b. Etherification.-Successful methylation of hydroxyl groups in

partially substituted carbohydrate boronates has not often been
reported. The first attempt, in which methyl iodide, silver oxide, and a
drying agent were applied to methyl a-D-glucopyranoside 4,6-
phenylboronate, gave,3 after distillation of the product, a poor yield of
the diether, chromatographic evidence being obtained that some
triether (but no tetraether) was formed during the reaction. Applied .to
methyl P-D-xylopyranoside 3,5-phenylboronate in N,N-dimethyl-
formamide, the same reagent gavez2only 18% of the 2-ether7 which
could be purified by sublimation. Other workers have reported no
success on attempted methylation of galactitol bi~(pheny1boronate)~~
and methyl 6-deoxy-/3-~-allopyranoside2,4-phenylboronate.60
The first, satisfactory methylation appears to have been effected by
Bourne and his colleague^,^* who prepared 6-0-methyl-D-glucose in
high yield from the 1,2:3,5-bis(phenylboronate),using diazomethane
and boron trifluoride etherate in dichloromethane for methylation, and
a final separation on a column of cellulose powder. It is this procedure
which the same group used for structural analysis of trio1 phenylboron-
ate^^^ (see Section 111,l); others60have attempted to use it, but with-
out success.
The phenylboronate group has been shown to be stable under
Koenigs-Knorr glycosylation conditions, and, from benzyl p-D-xylopy-
ranoside 2,4-phenylboronate73-0-~-D-g~ucopyranosy~-D-xy~ose and 3-
0 - a - and p-D-xylopyranosyl-D-xylose have been synthesized52by using
nitromethane as the solvent. The anomer of the initial boronate was
less reactive under these conditions, conceivably because its 3-
hydroxyl group cannot become hydrogen-bonded intramolecularly
and thereby gain enhanced nucleophilicity.
Tritylation of the nucleoside 2’,3’-phenylboronate has been used to
obtain 5’-O-trityladenosine,’T and trimethylsilyl ethers of glycoside
phenyl-, methyl-, and butyl-boronates have been produced by use
of chlorotrimethylsilane and trifluorobis(trimethy1silyl)acetamide in
pyridine for mass-spectrometric studies56 (see Section VI).

c. Nucleophilic Displacement Reactions-Several reports have

appeared on the introduction of nucleophilic groups into carbohydrate
(79) S . P. Valueva, E. P. Cherneva, V. A. Kargin, and N. M. Merlis, Vysokomol. Soedin.
Ser., 13, 468-470 (1971); Chern. Abstr., 75, 152,153~(1971).
56 ROBERT J. FERRIER

derivatives carrying boronic ester groups. In the most straightforward,

D-glucose 1,2:3,5-bis(phenylboronate)was separately treated with
triphenylphosphine in bromoform and carbon tetrachloride, and the
initial products were deboronated, to give 6-bromo- and 6-chloro-6-
deoxy-D-glucose in 81 and 79% yield, r e ~ p e c t i v e l y .Although
~~
displacements of the sulfonyloxy groups of 1,3-O-benzylidene-5-OO-p-
tolylsulfonyl-L-arabinitol 2,Cphenylboronate and 6-deoxy-3,4-0-iso-
propylidene-5-0-p-tolylsulfonyl-~-mannitol 1,2-phenylboronate did
not occur so readily when the esters were heated in refluxing N , N -
dimethylformamide with sodium benzoate and sodium azide, respec-
tively, direct substitution did occur, to give products whose structures
were used to establish the sites of bor0nation.~7
Yurkevich and coworkers46 used 3-0-p-tolylsulfonyl-D-ribose 2,4-
phenylboronate to prepare a cobalt-containing carbohydrate com-
pound which, without reported evidence, they claimed has a C-3-
cobalt bond and the D-XY~O configuration. Similarly,74 they prepared,
from the 5’-p-toluenesulfonates, nucleoside derivatives having C-5’-
cobalt bonding, and, in related work, they obtained nucleoside
derivatives bonded directly to cobalt at C-5’ by use of an intermediate
which, on the basis of carbon and hydrogen analytical data, they
proposed had structure 41 (R = H). Synthesis of this compound was
a~hieved71.8~ by treatment of 5’-O-p-tolylsulfonyladenosine2’,3’-
phenylboronate with lithium bromide in acetic anhydride at 100’-
conditions that would be expected81to yield the acetylated analog (41,
R = Ac). As such a compound would have carbon and hydrogen
compositions similar to those of the monoboronate (41, R = H), the two

NHAc
I

-b b
PhBOR, Ac
41

(80) I. P. Rudakova, V. I. Sheichenko, T. A. Pospelova, and A. M. Yurkevich,Zh.Obshch.

Khim., 37, 1748-1753 (1967).
(81) B. C. Maiti, 0.C. Musgrave, and D. Skoyles,]. Chem. SOC. Chem. Commun., 244-
245 (1976).
 CARBOHYDRATE BORONATES 57

would not be readily distinguishable on this basis. The same 5’-bromo-

5‘-deoxy intermediate was also used to obtain 5’-thioadenosine
derivatives .73

d. Oxidation Reactions.-The phenylboronate group has been found

to be stable during acetic anhydride-dimethyl sulfoxide oxidation of
hydroxyl groups, and this has made possible53the synthesis of methyl
a-D-erythro-pentopyranosid-3-ulose in 50-60% yield from methyl Q-D-
xylopyranoside 2,4-phenylboronate (29); the P-glycosidulose was
obtained similarly from the p-ester 38 in 80% yield. Oxidations were
effected at 40°, and the stability of the esters during the reactions was
indicated by polarimetry, as the first product obtained from the
boronate 29 is levorotatory, whereas the derived, deboronated ketone
is dextrorotatory. Furthermore, the oxidations were highly specific,
indicating that the readily oxidized, ketonic products remained
protected during the reactions. The oxidized boronates were not,
however, isolated, the de-esterified products being obtained by
column chromatography. Other worker@ reported the unreactivity of
methyl 6-deoxy-P-D-allopyranoside2,4-phenylboronate under these
conditions.
Oxidation of phenylboronates with the periodate ion has been used
in structural analysis,3*22*23,56as, in aqueous 1,4-dioxane7 they are
hydrolyzed to release diols that are susceptible to oxidation only when
they are vicinally related. Thus, the esters of mOnO-O-aCetyl-D-glUCal
and -D-galactal were found to reduce 0.09 and 1.07 molar equivalents
of periodate, respectively, and were, therefore, assigned55 the
structures 42 and 43. Whereas, in our lab0ratory,3*2~*~5 we have found it
unnecessary in these analyses to correct for oxidation of reaction
components other than the carbohydrates, other workers23 have
reported the use of substantial corrections.
$!H,OAc

42 43

v. SEPARATIONS OF CARBOHYDRATES BY USE OF THEIRBORONATES

1. Use of Isolated Boronates
As only the cis isomers of 1,2-, 1,3-, and 1,4-cyclohexanediols can
form simple, cyclic boronates, whereas the trans compounds give
58 ROBERT J. FERRIER

polymeric diesters, the epimeric diols may be separated by distillation

of their boronates, and this procedure has been applied by using
butylboronates prepared directly with butylboronic acid,82 and
ethylboronates obtained by thermal cyclization of bis(diethy1bori-
nates).17
The anomeric methyl 3,5-O-isopropylidene-~-xylofuranosides are
separable by distillation,83 and, therefore, the corresponding 3,5-
cyclic boronates (for example, 39 and 40) could also afford means of
separating the methyl xylofuranosides by this method. This procedure
has, apparently, not yet been attempted, but the intramolecular
hydrogen-bonding responsible for making a anomers in this series
relatively volatile makes them, also, very much more soluble in
nonpolar solvents, and this characteristic has been usefully applied to
the isolation of the anomeric methyl D - x y l o f u r a n o s i d e ~Their
. ~ ~ ~ ~3,5-
~
phenylboronates prepared from a glycoside mixture rich in furanosides
can be fractionally crystallized from light petroleum [solubilities
(g/lOO ml): a, 17.3; /3, 0.641, thus providing a means of obtaining the
unsubstituted a- and P-glycosides in 21 and 29% yield, respectively.
With the methyl D-xylopyranosides, it is the ester (38) of the P
anomer that has the intramolecular, strongly hydrogen-bonded
hydroxyl group, making it the more soluble anomer [solubilities50
(g/lOO ml in light petroleum): a,0.07; P, 531. By virtue of this great
difference, methyl a-D-xylopyranoside 2,4-phenylboronate can readily
be separated from its isomers, to provide a convenient means of
isolating this glycoside, which is otherwise difficult to obtain.69Methyl
a-and /3-D-~ylopyranoside-5-~~0 have both been prepared by applica-
tion of this procedure,83aand the stereoisomers of cis-3,4-thiolanediol
l-oxide have been separated as their p h e n y l b ~ r o n a t e s . ~ ~ ~

2. Use in Paper Chromatography

It has been shown by two groups of w0rkers84-8~that phenylboronic
acid incorporated into paper-chromatographic solvents specifically
(82) H. C. Brown and G . Zweifel,]. Org. Chem., 27,4708-4709 (1962).
(83) B. R. Baker, R. E. Schaub, J. P. Joseph, and J. H. Williams,]. Am. Chem. Soc., 76,
4044-4045 (1954).
(83a) W. D. Hitz, D. C. Wright, P. A. Seib, M. K. Hoffman, and R. M. Caprioli, Carbo-
hydr. Res., 46, 195-200 (1976).
(83b) J. E. McCormick and R. S. McElhinney,]. Chem. Soc. Perkin Trans. 1 , 2533-
2540 (1976).
(84) H. M. Wall, M.Sc. Thesis, University of London, 1964.
(85) R. J. Ferrier, W. G. Overend, G. A. Rafferty, H. M. Wall, and N. R. Williams, Proc.
Chem. Soc., 133 (1963).
(86) E. J. Bourne, E. M. Lees, and H. Weigel,]. Chrornatogr., 11,253-257 (1963).
 CARBOHYDRATE BORONATES 59

TABLEI
Enhancement Factors for Paper-chromatographic Mobility of
Free Sugars by Phenylboronic Acid

Enhancement Enhancement
sugar factora Sugar factor"

GIycerose 1.05: 1.0' G1ycerone 1.oc

Erythrose 2.7b glycero-Tetrulose 1.0'
Threose 1.7,b 1.0' erythro- Pentulose 1.8=
Ribose 2.0,b 2.0e threo-Pentulose 2.3"
Arabinose 0.9,b 0.7' Fructose 1.1: 0.9'
Xylose 1.0: 0.7" Sorbose 1.6b
Lyxose 1.0: 0.8' gluco- Heptulose 2.2'
Allose 1.4" 2-Deoxy-erythro-pentose 0Bc
Altrose 0.95" 2-Deoxy-ribo- hexose 0.8'
Glucose 1.0,* 0.75" 2-Deoxy-arabino-hexose 0.8e
Mannose l . l , b 0.7' 2-Deoxy-lyxo-hexose 1.3c
Gulose 2.l,* 2.1' 3-Deoxy-ribo-hexose 0.9'
Idose 1.8: 2.0' 3-Deox y-xylo-hexose 1,4c
Galactose 1.3: 1.0" 4-Deoxy-xylo-hexose 0.8''
Talose 2.7' 6-Deoxytalose 2.0e

"Ratios of RF values in solvents (ii) and (i).The enhancement factors determined in

the butanol-containing solvents are frequently somewhat less than unity, presumably
because the solvents differed slightly in their ethanol content (see footnote c). bDescend-
ing chromatograms with solvents (i) 9:2:2 ethyl acetate-acetic acid-water, and (ii) the
same, but with phenylboronic acid (0.55%) added.86'Descending chromatograms with
solvents (i) butanol-ethanol-water (4:1:5, upper phase), and (ii) the same, but with
phenylboronic acid (5%) added. This caused slight phase-separation, and so ethanol
(2.5%) was then also added.84

enhances the mobilities of compounds that possess trio1 systems from

which particularly stable, boronic esters are formed (see Tables I and
11). From work with six-membered-ring compounds, it was concluded
that this stabilization arises from contiguous cis&-triols on such rings,
which use 1,3-related, axial hydroxyl groups to give six-membered,
cyclic esters (44) stabilized to hydrolysis by the intervening, equatorial

44
60 ROBERT J. FERRIER

TABLEI1
Enhancement Factorsa for Paper-chromatographic Mobility of
Alditols, Inositols, and Anhydro Compounds by Phenylboronic Acid

Enhancement Inositols and Enhancement

Alditol factop anhydro compounds factop

Glycerol 1.1; 0.85' d o - Inositol 2.7b

Erythritol 1.4; 1.6' epi-Inositol 4.0b
Ribitol 3.4," 2.85' chiro-Inositol 0.7b
Arabinitol 3.6,b3.3c muco-Inositol 1.06
Xylitol 3.2,b4.2' myo-Inositol 1.@
Allitol 2.5' scyZZo- Inositol LO*
Alhitol 3.2b l,6-Anhydro-P- 1.0; 0.9'
altropyranose
Glucitol 5.6,b7.8' 1,6-Anhydro-p- 1 . 0 , b 0.9"
glucopyranose
Mannitol 5.4," 4.6c 1,6-Anhydro-P- 1.2: 2.w
mannopyranose
Galactitol 6.7," 5.7' 1,6-Anhydro-P- 1.0; 1.w
gulopyranose
2-Deoxy-erythro- 1.46 1,6-Anhydro-P- l.lb
pentitol galactopyranose
1-Deoxygalactitol 2.26 Methyl 3,B-anhydro-a- 1.2"
glucopyranoside
1,6-Dideoxygalactitol 1.5b Methyl 3,6-anhydro-P- 1.2'
glucopyranoside

aFootnotes as for Table I.

hydroxyl group. Of the pentoses, ribose alone has such a triol grouping
(P-D-lyxopyranose is excluded, because of its instability in the lC4
conformation) and is specifically affected (see Table I and Fig. 1).In
the aldohexose series, allose, gulose, and talose show enhancements,
presumably because, in the pyranose modifications, they contain the
required cis,cis-trio1 groupings, but mannose does not, and is thus
analogous to lyxose. Idose does not possess a cis,cis-trio1 grouping but
shows marked boronate enhancement, suggesting that either its
boronate is formed at 0-2,O-4 of a pyranose form, and the ester is
stabilized by the ring-oxygen atom, or else a furanoid form presents a
suitable triol grouping to the reagent.
Other than glycerol, the alditols (see Table 11) all form strong
complexes, so the technique is usually suitable for separations of free
sugars from their reduction products.86 Removal of oxygen atoms from
these derivatives, as expected, lessens their association with the acid.
 CARBOHYDRATE BORONATES 61

FIG. 1.-Paper Chromatograms of the Aldopentoses (1, Arabinose; 2, Lyxose; 3,

upper phase); B, the
Ribose, and 4, Xylose). [Solvent: A, butanol-ethanol-water (4:1:5,
same, but containing 2% of phenylboronic acid.]

Evidence of complexing was obtained with the anomeric methyl 3,6-

anhydro-D-glucopyranosides(see Table II), which suggests that the
oxygen atoms on the pyranoid ring can play a role in ester stabilization
(45); but, with 1,6-anhydro-/3-~-glucopyranose (46), no complex was

45
62 ROBERT J. FERRIER

formed, because of the splayed relationship of its carbon-oxygen

bonds at C-2 and C-4. The 0-2-0-4 distances in these compounds87 of
276 and 330 pm (2.76 and 3.30 A) are in agreement with these
observations, but it should be noted that the 1,6-anhydride can be
converted into an unstable 2,4-phenylb0ronate.~l
The complexes formed by the methyl 3,6-anhydroglucosides and 1,6-
anhydro-P-mannopyranose (see Table 11) establish that cyclic systems
other than that depicted in formula 44 may give rise to mobility
enhancements, but, nevertheless, this chromatographic technique can
be used in certain cases to detect cis&-triols and, thereby, in
configurational analysis, The configurations of 2-C-methyl-~-arabinose
and -L-ribose have thus been ascertained by determination of the
mobilities of the free sugars and their methyl P-pyranosides. The
epimer and the glycoside that showed enhanced mobilities in solvents
containing phenylboronic acid were assigned the ribo structure.88

3. Use in Electrophoresis
Sulfonylated phenylboronic acid (mainly the ortho isomer) shows
selective, complexing abilities with carbohydrates, and may thus be
used in electrophoresis to permit separations of alkali-sensitive
compounds at neutral pH values .89 Complexing occurs with all alditols
other than glycerol, that is, they migrate on the electrophoretograms,
and the relative degrees of complexing are related to the enhancement
values found during chromatography with phenylboronic acid (see
Tables I and 11). (For example, relative mobilities are: ribitol, 0.3;
arabinitol, 0.6; xylitol, 0.9; glucitol, 1.3; mannitol, 1.0; and galactitol,
1.0). Observations with free sugars indicate, however, that the two
procedures are basically different, because all aldopentoses, for
example, migrate on the electrophoretograms (mobilities relative to
glucose as unity: ribose, 4.7; arabinose, 2.4; xylose, 1.8; and lyxose,
2.3), whereas only ribose shows chromatographic enhancement.
Likewise, fructose showed no complexing during chromatography, but
underwent an extremely strong interaction with sufonylated phenyl-
boronic acid. With the information available at present, the two
methods cannot be accurately compared; both, however, appear to be
selective for cis&-trio1 groupings on six-membered rings, and the

(87) Borje Lindberg, Bengt Lindberg, and S. Svensson,Actu Chem. Scand., 27,373-374
(1973).
(88) R. J. Fernier, W. G. Overend, G.A. Rafferty, H. M. Wal1,andN. R. Williams,J. Chem.
SOC.,C, 1091-1095 (1968).
(89) P. J. Garegg and B. Lindberg, Actu Chem. Scund., 15, 1913-1922 (1961).
 CARBOHYDRATE BORONATES 63

electrophoresis results also reveal interaction with cis-1,2-diol group-

ings in furanoid systems.

4. Use in Column Chromatography

From the known, differential complexing between boronic acids and
polyhydroxy compounds, it follows that carbohydrate mixtures may be
separated b y column-chromatographic methods that exploit the
differences. Nucleoside and nucleotide boronates have been separated
on columns of anion-exchange resinsFOand sugars and alditols have
been shown to be differentially retained on such resins in the
sulfonated phenylboronic acid form,64but perhaps the best uses of
column chromatography in this connection have incorporated the
resolving powers of insoluble polymers to which boronic acid groups
have been covalently bonded. Such insoluble forms of boronates have
been synthesized either by substitution of polysaccharide derivatives,
or by polymerization of suitable arylboronic acids.
In the first approach, O-(carboxymethy1)cellulose (47) has been
converted into the acyl azide (48) and thence, by treatment with m-
aminophenylboronic acid, into the borylated form 49, to give a polymer

47 48 49

on columns of which the alditols had the relative retention-times:

erythritol, 51; ribitol, 56; arabinitol, 76; xylitol, 93; glucitol, 182;
mannitol, 109; and galactitol, 111(eluted with acetate buffer, pH 7.5).91
These values indicate that the complexing that results in enhance-
ments of paper-chromatographic mobilities of alditols when phenyl-
boronic acid is added to solvents (see Table 11) operates in a closely
similar fashion, to determine their retention by the borylated cellulose.
In the same work, it was showng1that nucleosides could be separated
on this stationary phase, and that they have retention volumes that
depend on the pH and ionic strengths of the buffers used, and on the
heterocyclic bases, but, especially, on the presence ofthe 2’,3’-diols. At
high pH-values, binding of the purine ribonucleosides was so strong

(90) I. I. Kolodkina, E. A. Ivanova, and A. M. Yurkevich, Khim. Prir. Soedin., 6,612-616

(1970); Chem. Abstr., 74, 112,363e (1971).
(91) H. L. Weith, J. L. Wiebers, and P. T. Gilham, Biochemistry, 9,4396-4401 (1970).
64 ROBERT J. FERRIER

that they could not be eluted. Yurkevich and coworkers prepared

similar column-materials (50)by treatment of 0-(2-diethylaminoethyl)-

so

Sephadex and O-(2-diethylaminoethyl)cellulose with a-bromotolyl-

boronic acid, and studied the separation of free sugarse2~93and
nucleosides and nucleotidess3~94on them at various pH values and
with different eluants. Compounds containing cis-172-diol groupings
were, again, the most strongly bound.
Solms and Deuelg5initially prepared a wholly synthetic, borylated
polymer by using m-phenylenediamine, p-aminophenylboron dichlor-
ide, and formaldehyde, and they investigated carbohydrate separations
on it, but addition polymers have usually been favored. Thus,
ribonucleosides and deoxyribonucleosides have been efficiently
separated on a column of a mixed copolymer of the methacroyl
derivativesee 51 and 52, and the method has been extended to

oligonucleotides, including free transfer-ribonucleic acids (t-RNA).

The strong complexing that again occurs with compounds containing
2’,3’cis-diol groupings makes it possible to retain ribonucleoside 5’-
phosphates, 3‘-ribonucleoside-terminatedoligo(deoxynucleotides),
and t-RNA’s on the columns, while deoxynucleotides and their
oligomers, ribonucleoside 2’-or 3’-phosphates and aminoacyl t-RNA’s

(92) E. A. Ivanova, S. I. Panchenko, I. I. Kolodkina, and A. M. Yurkevich, Zh. Obshch.

Khim., 45,208-212 (1975).
(93) A. M. Yurkevich, I. I. Kolodkina, E. A. Ivanova, and E. I. Pichuzhkina, Carbohydr.
Res., 43, 215-224 (1975).
(94)E. A. Ivanova, I. I. Kolodkina, and A. M. Yurkevich,Zh.Obshch. Khim., 44,430-434
( 1974).
(95) J. Solms and H. Deuel, Chimia, 11,311 (1957).
(96) H. Schott, Angew. Chem. Int. Ed. Engl., 11,824-825 (1972).
 CARBOHYDRATE BORONATES 65

are eluted, and this represents97 a significant development in

techniques available for the purification of t-RNA’s.
Similar polymers have been used by S . A. Barker a n d his
colleague^.^^ Radical-induced copolymerization of iminodiethyl (4-
vinylphenyl)boronate, divinylbenzene, and ethylvinylbenzene gave a
polymer on which free sugars can be differentially eluted with distilled
water. It is noteworthy that ribose was specifically bound by the
polymer (compare, Table I). Separations-particularly those of D-
glucose and D-fructose-were shown to be temperature- and pH-
dependent.

5. Use in Gas-Liquid Chromatography

Separations were first achieved with butylboronates of sugars and
alditols, many of which were shown to be separable at 200”. In the
earliest report,f’8 it was recognized that products obtained by direct
esterification of some sugars in pyridine were chromatographically
homogeneous, whereas others were not-glucose giving, for example,
three resolvable products. A concurrent study,99 however, showed that
glucose gave a single ester if the butylboronation was allowed to pro-
ceed to equilibrium, and, similarly, lyxose, fucose, arabinose, xylose,
fructose, galactose, and mannose gave essentially homogeneous prod-
ucts (eluted in that order); ribose, rhamnose, erythritol, arabinitol,
xylitol, and glucitol did not give single esters. This workggdiffered
from that reported in the first paper98 by trimethylsilylation of the
unsubstituted hydroxyl groups of the esters prior to g.1.c. examination.
A fuller report,loOwhich included mass-spectral results on the main
products obtained from D-glUCOSe, D-mannose, D-galactose, and
D-fructose showed them to be bis(buty1boronate) trimethylsilyl ethers.
Similar substitutions, involving methylboronation followed by
trimethylsilylation, have been performed in a g.1.c.-m.s. investiga-
tionS8 of the structures of esters of several glycosides and sugars (see
Section VI).

VI. MASSSPECTROMETRY
OF BORONATES

Although, in the mass spectrometer, several fragmentation-

pathways are followed by carbohydrate boronates, those which result
(97) H. Schott, E . Rudloff, P. Schmidt, R. Roychoudhury, and H. Kossel, Biochemistry,
12,932-938 (1973).
(98) F. Eisenberg, Carbohydr. Res., 19, 135-138 (1971).
(99) P. J. Wood and I. R. Siddiqui, Carbohydr. Res., 19,283-286 (1971).
(100) P. J. Wood, I. R. Siddiqui, and J. Weisz, Carbohydr. Res., 42, 1-13 (1975).
66 ROBERT J. FERRIER

in the formation ofthe cyclic, boron-containing ions shown in Scheme 3

‘p):’
Jb
‘PR
’
0
‘
R = Me
84
Ion masses
Bu
126
Ph
146

R’ i t

k
3R (R‘=H) 98 140 160

Scheme 3

are of the greatest value in structural analysis. The formation, struc-

tures, and masses of such ions formed from five- and six-membered
methyl-, butyl-, and phenyl-boronates are shown, together with those
of the ions derived from analogous a-diol cyclodiboronates. In the case
of the six-membered species, loss of hydrogen atoms can readily occur,
to give intense, conjugated ions of one mass unit less than those given
in Scheme 3. Boronates of vicinal diols are, therefore, recognizable by
the ions derived by direct excision of the two carbon atoms involved,
together with their ester substituents, whereas boronates of 173-related
diols give ions having a neighboring group (R’) attached. In terminal
systems of this type (derived, for example, from 4,6-diols of aldo-
pyranosides), ions having R’ = H are produced, and are, therefore,
readily recognized by having the mle values given in Scheme 3. The
bis(pheny1boronates) of the tetritols and pentitols produced by the
dehydration procedure all gave ions with mle 159 and 160, and were
therefore assigned 173:2,4-bicylostructures; for the D-arabinitol deriva-
tive, however, the 2,4:3,5-isomeric structure cannot be excluded.101On
the other hand, the bis(boronate) obtained from xylitol per(diethy1-
borinate) was shown in this way to have the 1,2:3,5-structure.40 With
the bis(pheny1boronates) of D-glucitol, galactitol, and D-mannitol, the
first two were each determined to contain two-fused, six-membered

(101) I. R. McKinley and H. Weige1,J. Chem. Soc. Chem. Commun., 1051-1052 (1972).
 CARBOHYDRATE BORONATES 67

ring-systems (and a five-membered ring), whereas the third gave no

ions indicative of a similar structure.101
Applied on a micro scale, and without isolation of the esters, to
products derived from mixtures of phenylboronic acid and methyl
a-D-glucopyranoside, methyl a-D-galactopyranoside, and methyl a-D-
mannopyranoside, the procedure indicated that the D-ghcoside and
D-galactoside derivatives contained 4,6-ester rings (mle 160) and
2,3-(diphenylcyclodiboronate)systems (mle 250), whereas the D-
mannoside gave the 2,3:4,6-bis(phenylboronate)structure (mle 160;
mle 146; no rnle 250).57 Molecular-ion masses confirmed these conclu-
sions, which are in agreement with the results ofchemical analysis (see
Section 111). Analogous results were obtained from the eight methyl
amino-4,6-O-benzylidenedeoxy-a-~-hexopyranosides~~~; when the
hydroxyl and amino groups at C-2 (C-3) and C-3 (C-2) were cis-related,
2-phenyl-1,3,2-oxazaborolane(53) rings were formed, whereas 2,4-
diphenyl-1,3,5,2,4-dioxazadiborepaneproducts (54) were obtained
from the trans-related glycoside derivatives (compare Section I),so the
method provides a means for determining the relative configurations of
a-diols and vicinal amino-alcohol groups on cyclic structures.

53 54

For free sugars, the procedure has been applied with isolated es-
t e r ~ ? and
~ * with
~ ~ samples prepared on a micro-scale and purified by
g.l.c.563esL-Arabinose gave a bis(buty1boronate) and a bis(pheny1boro-
nate) displaying ions mle 126 and 146, respectively, indicative of their
containing five-membered ester groups.44 No ions of mle 140 or 160
were detected (indicating the absence of six-membered rings), and
thus, 1,2:3,4-structures were assigned. On the other hand, D-XylOSe
bis(buty1boronate) and bis(pheny1boronate) gave ions having rnle 126
and 146 respectively, and also others having mle 140 (139) and 160
(159), respectively, indicating their 1,2:3,5-furanoid structures. Care
has to be taken, however, with analytical work of this type, as is illus-
trated by the spectra given by the bis(methy1boronates) of these two
sugars?e which are distinguishable mainly on the basis of ion inten-

(102) I. R. McKinley, H. Weigel, C. B. Barlow, and R. D. Guthrie, Carbohydr. Res., 32,

187-193 (1974).
68 ROBERT J . FERRIER

sities. All of the structurally significant fragment-ions appear in both

spectra.
Biemann and coworkers56 extended the g.1.c.-mass spectrometric
analysis of boronates by studying the trimethylsilyl ethers of com-
pounds which, after methylboronation, still retain unsubstituted hy-
droxyl groups. On such treatment, D-fUCOSe gives a product showing a
molecular ion mle 212, establishing that it contains two boronic ester
and no trimethylsilyl groups, whereas the isomer L-rhamnose is con-
verted into a derivative having one cyclic ester group and two ether
substituents (mle 332). The esterification process therefore occurred at
one diol site only. In this way, the xylose and arabinose results (see the
preceding) were confirmed, and the diesters respectively formed from
D-ghcose, D-mannose, and D-galaCtOSe were shown to have the 1,2:3,5,
2,3:4,6, and 1,2:3,4 structures, respectively. Similar work was per-
formed with several methyl aldopyranosides and, with an excess of
methylboronic acid, methyl a-D-mannopyranoside gave the 2,3:4,6-
diester, as expected, and, with limiting amounts of the reagent, a mix-
ture of the 2,3- and 4,g-esters (compare Section 111,3).Methyl a - ~ -
galactopyranoside was shown to be esterified to give both the 3,4- and
4,6-cyclic esters, in contrast to the conclusion drawn from chemical
evidence (see Section 111,3). One feature of this analysis that merits
comment is the failure to detect evidence for trioxadiborepane ring-
structures with such compounds as methyl a-D-glucopyranoside, de-
spite the use of a 20-fold molar excess of the reagent.
A study of the 3-0-a-D-glucopyranosyl-, 3-0-P-D-glucopyranosyl-,
3-0-a-D-galactopyranosyl-, 3-0-~-D-ga~actopyranosy~-, and 3 - 0 - m ~ -
mannopyranosyl-L-glycerols as phenylboronates, prepared in sub-
milligram quantities, illustrates ( a ) the value of the method in structural
analysis, and ( b ) the diversity of the fragmentation reactions that can
occur.1o3In Scheme 4, the full degradation pattern of the derivative of
3-0-a-D-glucopyranosy~-L-glycero~ is illustrated; all except the man-
nosy1 isomer gave the same ions, but having different intensities.
Initial work with nucleoside 2’,3’-phenylboronates and their 5’-
trimethylsilyl ethers was conducted by Dolhun and Wiebers,l04 who
identified the main fragmentation-pathways as including rupture of
the C-l‘-N and C-4’-C-5’ bonds. They then showed29 that phenyl-
boronated and then trimethylsilylated dinucleotides undergo fragmen-
tation in a way that identifies their sequence, and the method has
potential for characterizing the 3’-terminal base of oligo(ribonu-
cleotides).
(103) S. G . Batrakov, E. F. Il’ina, B. V. Rozynov, V. L. Sadovskaya,and L. D. Bergel’son,
Izu. Akad. Nauk S S S R , Ser. Khim., 821-828 (1975).
(104) J. J. Dolhun and J. L. Wiebers, Org. Mass Spectrom., 3,669-681 (1970).
 CARBOHYDRATE BORONATES 69

OH

pq
159 146 353

I t O?

P h q

160 250 4 51

phB%+- Ph phQ Ph

306 410 459 ss5

1 1
Ph
Ph

141

Ph
305 409 438 334
Scheme 4

Carbohydrate phosphates have been examined as the methyl- and

butyl-boronates of their dimethyl esters.105 Procedures involved
methylation of the enzymically prepared phosphates, and conversion
into the dimethyl esters with diazomethane, followed by the usual
boronation. Fragmentation patterns were identified for the following:
a-D-glucofuranose 172:3,5-bis(butylboronate) 6-(dimethyl phosphate),
a-D-galaCtOpyranOSe 1,2:3,4-bis(butylboronate) 6-(dimethyl phos-
phate), P-D-fructopyranose 2,3:4,5-bis(butylboronate)1-(dimethyl
phosphate), and methyl D-gluconate 2,3:4,5-bis(butylboronate)
6-(dimethyl phosphate). In the case of the D-mannose &(dimethyl
phosphate), the bis(buty1boronate) obtained was assigned the 1,2:3,5-
furanoid structure, which contrasts with the 2,3:4,6-structure of boro-
nates derived from the unsubstituted ~ u g a r . ~ 6
With the boronate phosphates,l05 provided that stereochemical fac-
tors permit, the phosphate groups stabilize species formed by loss of
(105) J. Wiecko and W. R. Sherman, Org. Mass Spectrom., 10, 1007-1020 (1975).
70 ROBERT J. FERRIER

the radicals attached to the boron atoms. Similarly, in acetates of sugar

boronic ester~,105~ the acetyl groups stabilize ions produced from
neighboring boronate rings, and this behavior can be used, for exam-
ple, to distinguish between a-D-glucofuranose 1,2-butylboronate 3,s-
methylboronate and the isomeric 3,5-butylboronate 1,2-methylboro-
nate; an acetyl group on 0-6 interacts only with the boron of the
adjacent 3,5-ester.

VII. NUCLEARMAGNETIC RESONANCE SPECTROSCOPY OF

BORONATEs
Several proton-, W - , and l'B-n.m.r. studies have been reported, and
all provide means for elucidating the structures of boronates. Results
from p.m.r. work26.32,33.44.45.108 were interpreted in the conventional
way, and gave information on the molecular structure and the
conformation of the carbohydrate portions of the molecules. The 13C
shifts of the ortho-aromatic carbon atoms of phenylboronates may
provide a means of determining the ester ring-sizes,32 but, on the basis
of studies of five compounds, Yurkevich and his ~olleagues4~ pointed
out that all three aromatic-carbon resonances (the signal from C-1 is not
observed, because of relaxation caused by bonding to boron) occur at
lower fields (0.5-1 p.p.m.) in the case of five-membered boronates.
Following the work of Cragg and Lockhart,l07 they also demonstrated
that broad-line, "B-n.m.r. spectra allow distinction between five- and
six-membered, cyclic phenylboronates.26 The former give resonances
4 p.p.m. upfield of those of the latter, and similar findings have been
reported for e t h y l b o r o n a t e ~ . ~This
~*3~ observation has assisted in the
structural analysis of cyclic ethylboronates of ~ y l i t o lD-mannitol>l
,~~
and galactit01.4~

VIII. BOFUNATES
The reaction undergone by alcohols with trialkyl- and triaryl-
boranes in the presence of pivalic acid, to give borinic esters, and the
thermal cyclization of bis(dialky1borinates) to boronates, are discussed
briefly in Section 11. Many borinates have been prepared in quantita-
tive yield from mono-, di-, oligo-, and poly-sa~charides,3~~J~*
and mixed

(105a) J. Wiecko and W. R. Sherman,J. Am. Chem. Soc., 98,7631-7637 (1976).

(106) A. B. Foster, R. Hems, and L. D. Hall, Can. J. Chem., 48,3937-3945 (1970).
(107) R. H. Cragg and J. C. Lockhart,J. Inorg. Nucl. Chem., 31,2282-2284 (1969).
(108) R. Koster, K.-L. Amen, and W. V. Dahlhoff,Justus Liebigs Ann. Chem., 752-788
(1975).
 CARBOHYDRATE BORONATES 71

borinic-boronic esters have been reported.17*33a*40-42 Treatment of

all of these compounds with methanol or 2,4-pentanedione causes de-
esterification, the borinic esters of mixed compounds being selectively
removable, to provide a means, for example, of preparing D-mannitol
3,4-ethylboronate41 and galactitol 2,3:4,5-bi~(ethylboronate)~ (see
also, Ref. 33a).
At pH 10, diphenylborinic acid gives a tetrahedral anion that
complexes with various diol systems, and thus it can be used in
electrophoresis like borate.lo9 I n a more detailed study of such
complexing,llO diols were examined by 13C-n.m.r. spectroscopy,
before and after addition of sodium diphenylborinate, and complexes
were detected, and their spectra observed, for a variety of carbohydrate
derivatives. 1,2-Diol groupings in acyclic and cis-cyclic compounds,
lY3-relateddiols at C-4,C-6 of hexopyranosides, the 3,s-diols of
glucofuranoses, and 2,4-diols of the anomeric methyl 3,6-anhydro-~-
glucopyranosides were all found to react. No interaction occurred with
1,6-anhydro-/3-~-glucopyranose (compare Section V,2).

IX. TABLES
The following Tables record some physical properties of boronates
of sugars, glycosides, C- and N-glycosyl compounds including nucleo-
sides, alditols, and anhydro sugars,

(109) P.J. Garegg and K. Lindstrom, Acta Chem. Scand., 25, 1559-1566 (1971).
(110) P.A. J. Gorin and M. Mazurek, C a n . J .Chem., 51,3277-3286 (1973).
 TABLEI11
Boronates of Sugars

Melting point [QI, Rotation

sugar Boronate (“C) (degrees) solvent References

Arabinopyranose, 8-L- 1,2:3,4bis(butyl- - +19 CHCI3 44

1,2:3,4bis(phenyl- 166 +8.5 CsH6 7.44
1,2-0-isopropylidene- 3,4phenyl- 130-131 - - 33 P
3,4-O-isopropylidene-
2-Deoxy-~erythro-pentose
l,&phenyl-
3,4-phenyl-
80-82
146147
-
-25.4 CHC13
- 33
26
sm
s
1-0-p-tolylsulfonyl- 3,Pphenyl- 117-1 18 +77.4 CHCl, 46 4
Fructopyranose, 8-D- 2,3:4,5-bis(phenyl- 89-98 - 17 CsHa 32 ?
14-benzoyl- 2,3:4,5-bis(phenyl- 139-142 -26 c6b 32 m
Galactopyranose, a - ~ - P
1,2-0-isopropylidene- 3,4-phenyl- 143-145 - - 33 z!
Galactopyranose, a - ~ -
- CHCl, 32
!L
6-deoxy- 1,2:3,4-bis(butyl- +24
1,2:3,4-bis(phenyl- 108.5-109.5 +29 C,H, 7,32
Glucohranose, Q-D- 1,2:3,5bis(phenyl- 161-162 +22 c 6 b 23,25,32,48
6-0-benzoyl- 1,2:3,5-bis(phenyl- 142-144 +81 CsHs 32
1,24-isopropylidene- 3,s-phenyl- 116-1 17 + 14 CHCI, 24,32,33
5,6-(diphenylcyclodi- 4648 - - 24
3-deoxy-3-fluoro- 5,gphenyl- 115-116 -2 CHCl, 49
6-0-(N-phenylcarbamoyl)- 3,Sphenyl- 6669 - - 47
6-0-p- tolylsulfonyl- 3,Sphenyl- 93-94 +24.4 CHCl, 47
6-0-(N-phenylcarbamoy1)- 1,2:3,5-bis(phenyl- 78-80 - - 47
6-0-p- tolylsulfonyl- 1,2:3,5bis(phenyl- 120-122 +42.4 CHCl, 47
1,2-0-(trichloroethylidene)- 3,5-p-tolyl- 95 - - 66
Lyxose, D bis(pheny1- 109-110 -60.4 C& 7
Mannofuranose, a-D
2,3Oisopropylidene- 5,Gphenyl- 171-173 - - 33
Mannose, L-
6-deoxy- bis(pheny1- 107.5 +87 CiHs 7
Ribofuranose, 8-D 1,5:2,3-bis(phenyl- 124-126 +82.8 CHCl, 26
Ribopyranose, a-D- 2,4phenyl- 90-92 -44.1 CHCI, 26
3-09- tolylsulfonyl- 2,4-phenyl- 57-58 -6.9 CHCl, 46
Ribose, D bis(pheny1- 140-142 +116 c6& 7
Xylofuranose, a-D- 1,2:3,5-bis(butyl- - +34 CHCll 44
1,2:3,5-bis (phenyl- 142-142.5 -10 C6Kl 7,44
1,20isopropylidene- 3,Sphenyl- 126-127 - 14 1,Cdioxane 32,33,69
3,5-di-O-methyl- 1,Bphenyl- 54-55 -9 1P-dioxane 69
Xylose, D,
dibutyl acetal bis(pheny1- 114-115 +25 l,.l-dioxane 70
diethyl a c e d bis(pheny1- 150-151 +28 1,4dioxane 70
ethylene acetal bis(pheny1- 180-181 +1 1,4-dioxane 70
di-isobutyl acetal bis(pheny1- 130-132 +23 1,Pdioxane 70
dimethyl acetal bis(pheny1- 168-169 +30 1,Pdioxane 70
dipropyl acetal bis(pheny1- 135-136 +22 1,4-dioxane 70
 TABLEIV
Boronates of Glycosides and C-and N-Glycosyl Compounds
2

Glycoside or glycosyl Melting point [ffID Rotation

compound Boronate (“C) (degrees) solvent References

Allopyranoside, methyl p-D

6deoxy- 2,4phenyl- 145-146 -76 GH, 58,60
3-O-(N-phenylcarbamoyl)- 2,Pphenyl- 154-155 -74 1,Pdioxane 58
Arabinopyranoside, methyl p-L- 3,4-phenyl- 73-74 +117 1,4-dioxane 51
3,4-ethyl- - + 27 MezSO 33a
2 0benzoyl- 3,4-phenyl- - +184 1,4-dioxane 51
Galactopyranoside, methyl a-D 4,6-phenyl- 119-120 +147 1,4-dioxane 55
2,3-di-O-acetyl- 4,Gphenyl- 166-167 +232 I,4-dioxane 55
6-deoxy- 3,4phenyl- - - - 59
Galactopyranoside, methyl p-D 4,6-phenyl- 176-177 -28 1,4-dioxane 55
2,Sdi-O-acetyl- 4,gphenyl- 145 +75 1,Cdioxane 55
2,3-di-O- benzoyl- 4,Gphenyl- 161-162 +125 1,4-dioxane 55
Glucopyranoside, benzyl p-D 4,Cphenyl- 177- 178 -91 1,4-dioxane 3
2,3,6-tri-O-acety1-4-0-
(2,3-di-O-acetyl-p-~
glucopyranosy1)- 4‘,6’-phenyl- 199-200 -58.9 CHC& 111
Glucopyranoside, methyl a-D 4,6-p-chlorophenyl- 164-165 +59 1,4-dioxane 3
3,4-ethyl- 46-47 + 80 1,4-dioxane 33a
4,6-m-nitrophenyl- 168-169 +49 1P-dioxane 3
4,g-phenyl- 166-167 +59 1,Pdioxane 3
4,6-phenyl 2,3-
(diphenylcyclodi- 162-163 -31 1P-dioxane 3
2,3-di-O-acetyl- 4,6-phenyl- 116-117 +74 1,4-&oxane 3
2,3-di-O-benzoyl- 4,6-phenyl- 203-204 +94 1,Pdioxane 3
6-deoxy- 2,4-phenyl- - - - 59
2,3-di-O-methyl- 4,6-phenyl- 120-122 +61 1,Cdioxane 3
2,3-di-0-p- tolylsulfonyl- 4,g-phenyl- 180-181 - 15 1,4dioxane 3
Glucopyranoside, methyl p-D- 4,6-phenyl- 188-189 -82 1,4-dioxane 3,55
4,6-phenyl-2,3-
(diphenylcvclodi- 185- 186 -127 l,4-dioxane 55
2,3-di-O-acetyl- 4,g-phenyl- 123-124 -99 1,4-dioxane 55
2,3-di-O-benzoyl 4,Gphenyl- 123-124 -2.6 1,Pdioxane 55
Hex-2-enopyranoside,p-nitro-
phenyl 2,3-dideoxy-a-~-
erythro- 4,6-phenyl- 180-181 +265 1,4-dioxane 3
Hexopyranoside, methyl
2-deoxy-a-~arabino- 4,g-phenyl- 142-143 +63 1,4-dioxane 3
Hexopyranoside, methyl
2-deoxy-a-~-lyxo- 3,4phenyl- 159 +114 1,Cdioxane 55
6-0-acetyl- 3,4-phenyl- 132-133 +39 1,4-dioxane 55
Hexopyranoside, methyl
2-deoxy-P-~-lyxo- 4,g-phenyl- 188 -71 1,Pdioxane 55
3-0-acetyl- 4,g-phenyl- 131 -87 1,4-dioxane 55
Lyxopyranoside, methyl a-D 2,3-phenyl- 66-69 +36 1P-dioxane 51
4-0-acetyl- 2,3-phenyl- - +13 1,4-dioxane 51
Mannopyranoside, methyl a-D- 2,3:4,Gbis(phenyl- 116-117 - 118 1,Pdioxane 3
2,3:4,6-bis(ethyl- - - 39 CCl, 33a
Mannopyranoside, methyl a - ~ -
6-deoxy- 2,3-phenyl- - -34 1,4-dioxane 58
4-0-(N-phenylcarbamoyl)- 2,3-phenyl- 184-185 + 13 1,4-dioxane 58
Psicofuranoside, methyl p-D 3,4-phenyl- 123-124 - 136 C& 112
l-chloro-l-deoxy- 3,4-phenyl- 95-96 -111.8 - 113
Ribofuranoside, methyl p-D 2,3-phenyl- 87.5 -64.6 - 113
Benzene
2,4,6-trimethoxy-l-p--~
ribofuranosyl- 2,3-phenyl- 136-137 - - 114
Ribopyranoside, methyl p-D- 2,4-phenyl- 149-150 -113 1,4-dioxane 51
3-0-acetyl- 2,4-phenyl- 82-83 -118 1,4-dioxane 51
3-0-(N-phenylcarbamoyl)- 2,4-phenyl- 163-164.5 -94 1,4-dioxane 51
8
(Continued)
 4
TABLEIV (Confinued) 0)

Glycoside or glycosyl Melting point 1.13. Rotation

compound Borunate ("C) (degrees) solvent References

Ribopyranosylamine,N-(p-
bromopheny1)-a-D- 2,4-phenyl- - - - 54
Xylofuranoside, ethyl
1-thio-a-D 3,Sphenyl- 102-104 +27 1,4-dioxane 115
Xylofuranoside, ethyl 6-D- 3,Sphenyl- 110-111 - 146 1,Pdioxane 116
l-thio- 3,5-phenyl- 157-158 -254 1,Pdioxane 116
Xylofuranoside, methyl a - ~ - 3,Sphenyl- 83-84 +21 1,4-dioxane 50,69
2-0-(N-phenylcarbamoyl)- 3,5-phenyl- 215-216 +97 1,4-dioxane 69 5
Xylofuranoside, methyl p-D- 3,Sphenyl- 122-123 - 158 1,4dioxane 50,69
20acetyl- 3,5-phenyl- 99-100 -88 1P-dioxane 69
Xylopyranoside, benzyl a - ~ - 2,4-phenyl- 152-153 -4 1,Pdioxane 52
Xylopyranoside, benzyl p-D 2,Pphenyl- 77-78 - 144 1,4-dioxane 52
3-0- (tetra-0-acety l-p-D
glucopyranosy1)- 2,4phenyl- 151-152 -97 1,cdioxane 52
Xylopyranoside, ethyl CY-D 2,4-phenyl- 137-138 +11 1,4dioxane 116
l-thio- 2,4phenyl- 143-145 +79.5 1,4-dioxane 115
Xylopyranoside, ethyl l-thio-
P-D- 2,4-phenyl- 109-110 -233 1,4-dioxane 115
Xylopyranoside, methyl a-D 2,4-phenyl- 175- 176 + 10 1,4dioxane 22,50
3-0-acetyl- 2,4-phenyl- 119-121 + 13 1,4-dioxane 22
3-0-benzo yl- 2,Pphenyl- 138-140 + 18 1,4-dioxan e 22
Xylopyranoside, methyl p-D- 2,Pphenyl- 85-86 - 104 1,Cdioxane 22,50
2,4-ethyl- - - 113 CCl, 33a
3-0-acetyl- 2,Pphenyl- 122-123 - 127 1P-dioxane 22
3-0-benzoyl- 2,4-phenyl- 99-100 -82 1,4-dioxane 22
3-0-methyl- 2,4phenyl- 82-84 -114 1,Pdioxane 22
3-0-(N-phenylcarbamoyl)- 2,4-phenyl- 146-147 -90 1,cdioxane 22
 TABLEV
Boronates of Alditols

Melting point DI.[ Rotation

Alditol Boronate (“C) (degrees) solvent References

L-Arabinitol bis(pheny1- 114-1 16 - - 23

5-0-benzoyl- 2,4-phenyl- 141-142 +12 CHCl, 67
1,3-O-benzylidene- 2,4-phenyl- 120-121 +ll CHCl, 67
5-0-ptolylsulfonyl- 2,4-phenyl- 141-142 + 16 CHC13 67
Erythritol bis(phenyl- 88 - - 23
1,3-O-butylidene- 2,4-phenyl- 60.5-62 +28 CHCl, 117
DErythritol
l-deoxy- phenyl- 75-76 43
Galactitol 2,3:4,5-bis(ethyl- 97-98 42
1,6:2,3:4,5tris(ethyl- - 42
1,3:4,6-bis(phenyl- 125-130 23,42
tris(pheny1- 162-163 - 9,42
2,5-di43-(N-phenylcarbamoyl)- bis(pheny1- 223-224 - 23
DGlucitol tris(pheny1- 187-190 +39.6 8,9,23
2,4di-0- benzoyl- 1,3:5,&bis(phenyl- 199 - - 23
2,402-butenylidene- 1,3:5,6bis(phenyl- 129-131 +18.6 CHCI, 118
2 , 4 0 butylidene- 1,3:5,6bis(phenyl- 82-84.5 -6.1 CHC& 119
1,3:2,4-dia-ethylidene- 5,6phenyl- 88 - - 23
2,4-O-furylidene- 1,3:5,6-bis(phenyl- 177- 178 +40.5 CHCl, 120
~~~ ~

(111) H. Kuzuhara and S. Emoto, Agric. Biol. Chem., 30,122-125 (1966);Chem. Abstr., 64, 19,736h (1966). (Continued)
(112) H. Hiehnhecky and J. FarkaS, Collect. Czech. Chem. Commun., 39, 1093-1106 (1974).
(113) J. Farkas’, Collect. Czech. C h e p . Commun., 31, 1535-1543 (1966).
(114) L. Kalvoda, J. Farkag, and F. Sorm, Tetrahedron Lett., 2297-2300 (1970).
(115) R. J. Femer, L. R. Hatton, and W. G. Overend, Carbohydr. Res., 6,87-96 (1968).
(116) R. J. Femer and L. R. Hatton, Carbohydr. Res., 6,75-86 (1968).
4
-l
 4
TABLEV (Continued) W

Melting point Iff1 Rotation

Alditol Boronate (“C) (degrees) solvent References

Glycerol phenyl- 76-78 - - 13,23,26,43

DMannitol 3,4-ethyl- 128 +28.2 CCI, 41
tris(p-bromophenyl- 204-205 +35.1 - 8
tris(p-chlorophenyl- 184-185 +45.4 - 8
1,2:3,4:5,6-tris(ethyl- - +9.7 CCI, 41
1,2:3,4:5,6tris(phenyl- 134-135 +53.4 - 8,9,23,123
tris(p-tolyl- 162-164 +45.9 - 8
1,2,5,6-tetra-O-acetyl- 3,4-ethyl- - +48.6 CCl, 41 !=
1,6-di-O-benzoyl- bis(pheny1- 150 - - 23
3,4-di-O-benzoyl- 1,2:5,6bis(phenyl- 149-150 - - m
9 !=
1,3-O-butylidene- phenyl- 116118 -10.8 CHCl, 121 4
3,4-0- butylidene- 1,2:5,6bis(phenyl- 69-72 +45 CHCI, 121 ?

L-Mannitol m
6-deoxy- P
3,4-O-isopropylidene- 1,e-phenyl- 78-80 -27 CHCl, 67 22
m
5-0-ptolylsulfonyl- 1,e-phenyl- 124-126 -23 CHCl, 67 !=
Pentaerythritol bis(pheny1- 207-208 - - 8
Xylitol 1,2:3,5bis(ethyl- - - - 40
4-0-acetyl- 1,2:3,5-bis(ethyl- - - - 40
1,e-ethyl- - - - 40
4-0- benzoyl- 1,2:3,5-bis(ethyl- - - - 40
3,4,5-tri-O-acetyl- 1,e-ethyl- - - - 40

(117) T. G. Bonner, E. J. Bourne, and D. Lewis,]. Chem. Soc., 7453-7458 (1965).

(118) T. G. Bonner, E. J. Bourne, and D. Lewis, J . Chem. Soc., 3375-3381 (1963).
(119) T. G. Bonner, E. J. Bourne, S. E. Harwood, and D. Lewis,J. Chem. SOC., 121-126 (1965).
(120) T. G. Bonner, E. J. Bourne, S. E. Harwood, and D. Lewis,]. Chem. SOC., C , 2229-2233 (1966).
(121) T. G. Bonner, E. J. Bourne, D. G. Gillies, and D. Lewis, Carbohydr. Res., 9,463-470 (1969).
 TABLEVI
Boronates of Anhydro Sugar Derivatives (Including a Lactone)

Melting point [&I, Rotation

Anhydro compound Boronate (“C) (degrees) solvent References

Altropyranose, 1,6-anhydro-p-~ 3,4-phenyl- 166-167 - 129.6 - 61

2-0-p- tolylsulfonyl- 3,4-phenyl- 173- 175 -112 - 61
Galactitol, 1,5-anhydro-~- 4,Gphenyl- 141-142 +69 1,4-dioxane 55
4,6-phenyl-2,3-(diphenyl-
cyclodi- 201-203 + 177 1,4-dioxane 55
2,3-diOacetyl- 4,6-phenyl- 143-144 +177 1,Cdioxane 55
Galactopyranose, 1,6-anhydro-
P-D- 3,4-phenyl- 169- 170 -114 - 61
2-0-p- tolylsulfonyl- 3,4-phenyl- 135-136 +50 - 61
Glucitol, 1,5-anhydro-~- 4,g-phenyl- 176- 177 -80 1,4-dioxane 3
4,6-phenyl-2,3-(diphenyl-
cyclodi- 188-189 -121 1,4-dioxane 3
D-Glucono-l,4-lactone 3,5-phenyl- - - - 78
Glucopyranose, 1,6-anhydro-
p-D- 2,4-phenyl- 122-124 -70.4 - 61
Gulopyranose, 1,6-anhydro-p-~- 2,3-phenyl- 212-216 +43.7 - 61
4-0-p- tolylsulfonyl- 2,3-phenyl- 140-141 +83.9 - 61
Hex-1-enitol, 1,5-anhydr0-2-
deoxy-Darabino- (D-glucal) 4,Gphenyl- 128 -53 1,4-dioxane 55
3-0-acetyl- 4,6-phenyl- 88-89 -90 1,4-dioxane 55
Hex-1-enitol, 1,5-anhydro-2-
deoxy-Dlyxo- (D-galactal) 3,4-phenyl- 104 -87 1,4-dioxane 55
6-0-acetyl- 3,4-phenyl- 98-99 -74 1P-dioxane 55
Hexitol, 1,5-anhydro-2-deoxy-
D-lyX0- 4,g-phenyl- 114-115 +60 1,Cdioxane 55
3-0-acetyl- 4,6-phenyl- 89 + 172 1,4-dioxane 55
Mannopyranose, 1,6-anhydro-
p-D- 2,3-phenyl- 149-150 - 104.9 - 61
4-0-p- tolylsulfonyl- 2,3-phenyl- 157-158 - 104 - 61
 03
TABLEVII 0

Boronates of Nucleosides

Melting point [ffl Rotation

Nucleoside Boronate ("C) (degrees) solvent References

Adenosine 2',3'-phenyl- 223-224 28,35,36,75,77

2',3'-m-nitrophenyl- 235-237 36
2',3'-p-tolyl- 226229 36
N6-benzoyl- 2',3'-phenyl- 187- 188 122
5'-phosphate 2',3'-phenyl- 170 36
5'4-p-tolylsulfonyl- 2',3'-phenyl- - 71,80
5'4-t~ityl- 2',3'-phenyl- 78 36
Cytidine 2',3'-phenyl- 140 27,28
N-acetyl- 2',3'-phenyl- 110 36
N-benzoyl- 2',3'-phenyl- 175-177 30
5'-O-p-tolylsulfonyl- 2',3'-phenyl- - 74
Guanosine 2',3'-phenyl- 270 35,36
N- benzoyl- 2',3'-phenyl- 160,168-170 35,36
Inosine 2',3'-phenyl- 196, 178-179 28,35,36
5'-O-p-tolylsulfonyl- 2',3'-phenyl- - 37
Uridine 2',3'-phenyl- 221-222 27,28
5'-O-p-tolylsulfonyl- 2',3'-phenyl- - 73

(122) S. G. Verenikina, E. G. Chauser, and A. M. Yurkevich, Zh. Obshch. Khim., 41, 1630-1632 (1971).

X. ADDENDUM
An X-ray structural study of D-mannitol1,2:3,4:5,6-tris(phenylboronate)
has been reported.123

(123) A. Gupta, A. Kirfel, G. Will, and G. Wulff,Acta Crystallogr., Sect. B, 33,637-641 (1977).
 BIOSYNTHESIS OF SUGAR COMPONENTS OF ANTIBIOTIC
SUBSTANCES

BY HANSGRISEBACH

Biologisches Znstitut IZ der Uniuersitat Freiburg i . Br., Lehrstuhl f u r Biochemie

der Pflanzen, D 7800 Freiburg i . Br., Germany

I. Introduction . . . . . . . . . . . . . . . . . . , , . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . 81
11. Branched-chain Sugars . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
1. Methyl-branched Sugars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
2. Sugars Having a Two-Carbon Branch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3. General Conclusions on the Biosynthesis of Methyl-branched
Sugars and on Sugars Having a Two-Carbon Branch . . . . . . . . .. . .. .. . .. 97
4. Sugars Having a Formyl or Hydroxymethyl Branch:
..
L-Streptose and L-Dihydrostreptose . . . . . . . . . . . . . . . . . . . . . . .... .. . .. .. 98
. . .. . . . .
111. Aminocyclitol Antibiotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . , , 102
.. . . . .
1. Streptomycins and Bluensomycin . . . . . . . . . . . . . . . . .. .. .
.. . . . . . . . . 102
2. Gentamicins . . . . .. . . . . . . . . . . .
.. . ., .. .. . . .. . . ... . . . . .
, . .. .. . . .. .. .. .. 110
..
3. Neomycin., . . . . . . . . . . .. . . . , . . . . . . . . .. .. . . . .. . .
. .. . . .. . .. . ..
. . . .. . , 115
4. Spectinomycin . .. .. . .. .. . . . . . . . .
.. . . . . .. . . . . . . . . . .. . . . . .. . . . . . . . . .. I 118
5. Validamycin . . . . . . . .. .. . . .. . . . . .. . ..
. . . . , . , . . . .. .. . . . . . . . . . . . . . . . . . . 120
IV. Amino Sugars Not Occurring in Aminocyclitol Antibiotics . . . . , . . , , . . . . . .. . 122
. . ..
1. Desosamine and Mycaminose.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . .. 122
..
V. Nucleoside Antibiotics.. . . . . . . . . . . . . . . . . .. . . .. .. . . . .
.. . . . . . . , .. . ... . .. 122

I. INTRODUCTION
The extensive research on the chemistry of antibiotics has led to the
discovery of numerous new sugars, many of which possess unusual
structures. Carbohydrate-containing antibiotics may be classified into
two large groups: those which are entirely carbohydrate in nature, for
example, the aminocyclitol antibiotics,’ and those which contain
sugars as glycosidic components, for example, the large family of
“macrolide” antibiotics.2 Dutchel.3 has classified the glycosidic
antibiotics into five subgroups according to the nature of the aglycon.
(1) S. Umezawa, Adu. Carbohydr. Chem. Biochem., 30,111-182 (1974);MTP Znt. Rev.
Sci.: Org. Chem. Ser. Two, 7 , 149-200 (1976).
( 2 ) W. Keller-Schierlein, Fortschr. Chem. Org. Naturst., 30,314-460 (1973).
(3) J. D. Dutcher, Adu. Carbohydr. Chem., 18,259-308 (1963).

81
82 HANS GRISEBACH

Inspired by the partly exotic structures of antibiotic sugars, a number

of chemists and biochemists have tried to unravel their biosynthesis.
As in other biosynthetic studies, labelling patterns obtained with
isotopically labelled precursors dominated the earlier phases of this
work, and this approach is still being used. Only recently has it been
possible in some cases to study the reactions at the enzymic level.
The present article concentrates mainly on recent developments.
Where it is necessary for a better understanding, the biosynthesis of
related sugars that have not yet been found in antibiotics is also
considered. The author also does not hesitate to include some
speculations on the biosynthesis of sugars for which experimental
results are still lacking, because these hypotheses might inspire future
work in this field.

11. BRANCHED-CHAINSUGARS
More than a dozen branched-chain, deoxy sugars have now been
discovered as components of antibiotics. One review on their
biosynthesis4 and ~ w o on ~ , the
~ chemistry and biochemistry of
branched-chain sugars have appeared. These sugars can be divided
into two groups according to their biosynthesis.
Group 1 includes methyl-branched sugars and sugars having a two-
carbon branch. These sugars arise by transfer of a C, or C2 unit from
appropriate donors to nucleotide-bound hexosuloses.
Group 2 consists of sugars having a hydroxymethyl or formyl branch.
These sugars are formed by intramolecular rearrangement of nucleo-
tide-bound hexosuloses, with ring contraction and expulsion of one
carbon atom.

1. Methyl-branched Sugars
a. L-Mycarose, L-Cladinose, and L-Noviose.-Early tracer studies on
the biosynthesis of L-mycarose (l),L-cladinose (2), and L-noviose (3)
led to the initially surprising observation that the C-methyl branch of
these sugars originates from the S-methyl group of L-methionine. The
hexose portion of 1,2, and 3 was shown to be formed from D-glucose
without inversion or rearrangement of the hexose chain. In the case of
mycarose and noviose, it was also shown, with the aid of ~-[methyl-’~C,
methyl-2H,]methionine that the C-methylation proceeds by way of
(4) H. Grisebach, Helo. Chim. Acta, 51, 928-939 (1968).
(5) H. Grisebach and R. Schmid, Angew. Chem. Int. Ed. Engl., 11,159-173 (1972).
(6) H. Grisebach, “Biosynthetic Patterns in Microorganisms and Higher Plants,” Wiley,
New York, 1967, pp. 66-101.
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 83

transfer ofintact CD,groups. This work has been covered in the earlier

I I
H,NCO OH
II
0
1 L-Mycarose R = H 3 L-Noviose
2 L-Cladinose R = Me

The results mentioned led to the hypothesis that C-methylation

occurs with S-adenosyl-L-methionine (AdoMet, “active methionine”)
as the methyl donor and a nucleotide-bound aldosulose as the
acceptor.6 This assumption has now been confirmed with a cell-free
system from Streptomyces rimosus that produces the antibiotic
tylosin, which contains L-mycarose. When a cell-free extract from
mycelia of S. rimosus was incubated with dTDP-D-glucose, S-adeno-
syl-~-[methyl-’~C]methionine,and NADPH, a new radioactive
product was formed that contained at least two dTDP-sugars.‘ One of
these was identified as dTDP-mycarose (6, see Scheme 1).The second
product has not yet been identified, but its properties are very similar to
those of dTDP-mycarose. Kuhn-Roth oxidation of the mixture of
the dTDP-sugars gave acetic acid that contained over 90% of the
carbon-14; this proved that both sugars must bear a C-methyl branch
originating from [14C]AdoMet. The same products were obtained
when dTDP-6-deoxy-~-xylo-hexos-4-ulose(4, see Scheme 1)was used
instead of dTDP-D-glucose as the substrate. The enzyme dTDP-D-
glucose 4,gdehydratase (EC 4.2.1.46), which catalyzes the conversion
ofdTDP-D-glucose into 4, was also shown to be present in extracts from
S. rimosus.8The D configuration of 4 formed by use of the enzyme from
S. rimosus was proved in the following way.8 Reduction of 4 with
sodium borohydride, and subsequent hydrolysis of the product, gave 6-
deoxyglucose and fucose. The former could be oxidized with D-glucose
oxidase (EC 1.1.3.4); as this enzyme is specific for D-glucose, the
configuration in 4 must also be D.
Besides the methylation step, the following reactions are necessary
for the synthesis of dTDP-L-mycarose from dTDP-D-glucose: reduc-
tion at C-2, inversion of configuration at C-5, and (possibly as the last
(7) H. Pape and G. U. Brillinger, Arch. Mikrobiol., 88,25-35 (1973).
( 8 ) H. Matem, G. U. Brillinger, and H. Pape, Arch. Mikrobiol., 88, 37-48 (1973).
84 HANS GRISEBACH

step) stereospecific reduction of the carbonyl group at C-4, leading to

the L-rib0 configuration (see Scheme 1).

dTDP-0-glucose - 0 0- OH
dTDP
-
M
Ci+
I
H
e
O
J
H O
OdTDP

4 5

H,C R H,C
6
SCHEME 1.-Postulated Reaction-sequence for the Biosynthesis of dTDP-L-mycarose
(6) from dTDP-D-glucose.

This sequence of reactions therefore requires several enzymes that

have not yet been separated from each other. As the methylation step is
NADPH-dependent,’ it may be postulated that reduction at C-2 occurs
before methylation at C-3. In Scheme 1,the 3,4-enediol of the hexos-4-
ulose 5 is shown as the nucleophilic acceptor for the methyl group from
AdoMet, but the existence of such an enediol has not been proved.
dTDP-D-glucose 4,gdehydratase activity increases during the
stationary-growth phase of S . rimosus, together with production of
tylosin and the dTDP-mycarose-synthesizing system.8This is in accord
with the assumption that these enzymes are involved in the
biosynthesis of tylosin.
The 0-methyl group of L-cladinose (2) is also derived from L-
methionine. The O-methylation step does not, however, take place at
the level of the “nucleotide-sugar,” but it occurs when the substrate is
erythromycin C, which contains L-mycarose and D-desosamine as
glycosidic componentsg (see Scheme la). The 0-methylation of the L-
mycarose moiety of erythromycin C by a partially purified enzyme from
Streptomyces erythreus was described by Alpine and Corcoran.loJ1
The reaction catalyzed is shown in Scheme la.
(9) W. Hofheinz and H. Grisebach, 2. Naturforsch. Teil B , 17, 852 (1962).
(10) T. S. McAlpine and J. W. Corcoran, Fed. Proc. Fed. Am. SOC. E x p . B i d . , 30, 1168
( 197 1).
(11) J. W. Corcoran, Methods EnzymoZ., 43, 487-498 (1975).
 SUGAR COMPONENTS O F ANTIBIOTIC SUBSTANCES 85

ErythromycinA R, = OH, R, = Me
Erythromycin C R, = OH, R, = H

[ :Lo" OQ
. Hz + AdoMet - [ :Lo" O
Q
M
:; + AdoHcy

H,C
SCHEME la.-Reaction Catalyzed by S-Adenosy1methionine:ErythromycinC 0-
Methyltransferase. (D = desosamine.)

The enzyme converts erythromycin C into erythromycin A in the

presence of AdoMet. Evidence was obtained that the enzyme is
associated with the microsomal fraction. The enzyme showed a very
high degree of substrate specificity. Aside from erythromycin C, it
failed to catalyze the methylation of any other L-mycarosyl moiety
tested. Erythromycin A and S-adenosyl-L-homocysteine (AdoHcy)
were potent inhibitors of the enzyme, and it was assumed that the Ado-
M e t AdoHcy ratio could be a major regulatory factor of the final step in
the formation of erythromycin A.

b. L-Vinelose.-Cytidine 5'-(~-vinelosyl diphosphate) (CDP-6-

deoxy-3-C-methyl-2-O-methyl-~-talose)~~-~~ (7)and its 4-(O-methyl-
glycolyl) ester15(8)were isolated from Acetobacter vinelandii strain 0 .
Although L-vinelose has not yet been found in an antibiotic substance,
(12) S. Okuda, N. Suzuki, and S. Suzuki,]. Biol. Chem., 242, 958-966 (1967).
(13) J. S. Brimacombe, S. Mahmood, and A. J. Rollins,]. Chem. SOC. Perkin Trans. 1,
1292-1297 (1975).
(14) M. Funabashi, S. Yamazaki, and J. Yoshimura, Carbohydr. Res., 44,275-283 (1975).
(15) S. Okuda, N. Suzuki, and S. Suzuki,]. Biol. Chem., 243,6353-6360 (1968).
 86 HANS GRISEBACH

the work on its biosynthesis is discussed here, because it complements

the results on the enzymic formation of L-mycarose.

Hb bMe
7 CDP-L-vinelose, R = H
8 CDP- (0-rnethylglycoly1)-
L-vinelose, R = MeOCH,CO

First, the incorporation of ~-[methyl-'~C]methionineinto CDP-

vinelose was investigated with cells of a methionine-requiring mutant
ofA, vineZandii.lsThe degradation of the radioactive CDP-vinelose by
the procedure shown in Scheme 2 showed that about half of the
radioactivity was present in the 0-methyl group (isolated as chloro-
methane), and the other half was in the C-methyl group (isolated as p -
bromophenacyl acetate).

CH,OH

9
HO

HO
OCDP H + + HO CH3

oie HO olke
OH -
NaBH,
HCdH,
.I
H,CCOH
I
HCOH
HOCH
CH,
I

BC13 &,C1

HYHO CqOH
I
HF0,H HCOH
B r G C O C H , B r
10;
.I
H,CCOH
Br + o c H , - y ~ H , +
HYO.,H I
HCOH
0
I
CHO HOCH
I I
cH3 CH,
SCHEME2.-Degradation of CDP-L-vinelose Synthesized by A. uinelandii in the
Presence of ~-[Methyl-"C]rnethionine. ( 0 , Isotope from ~-[methyl-'~C]methionine.)

When a crude enzyme-preparation from A. vinelandii was incubated

with CDP-D-[U-14C]glucoseand AdoMet, and the incubation mixture
(16) Y. Eguchi, M. Takagi, F. Uda, K. Kimata, S. Okuda, N. Suzuki, and S. Suzuki,]. B i d .
Chem., 248,3341-3352 (1973).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 87

was then digested with phosphate diesterase and phosphatase, three

new radioactive sugars (Xl, X,, and X,) were detected on paper
chromatograms.ls One of the products (XI) was also formed in the
absence of AdoMet, and was identified as 6-deoxy-~-xylo-hexos-4-
ulose. With the aid of Ad~[methyl-~H]Met, it was shown that the other
two products were methylated sugars. The corresponding CDP-sugars
were extremely labile, and were degraded to the glycosyl phosphates
immediately after the mixture was applied to the paper. All of the
evidence so far obtained indicates that one of the sugars (X,) obtained
by hydrolysis of the 1-phosphate was a 6-deoxy-3-C-methyl-hexos-4-
ulose (9, see Scheme 3). The evidence for this structure was educed
mainly from the series of reactions shown in Scheme 3. Reduction of 9
with sodium borohydride gave a pair of alditols (lo), indicating the
reduction of a keto group in the original sugar. Structure 9 was also
consistent with the isotope distribution found in the degradation
products. When, for example, a sample of r3H, 3-'4c-jx, (obtained from
incubation of CDP-6-deoxy-~-xy~o-[3-~~C]hexos-4-ulose with [methyl-
14C]AdoMet)was reduced with sodium borohydride, and the product
decomposed with sodium periodate, an acetic acid having the same
ratio of ,H to 14Cas that of the original sugar was obtained. This result
proved the attachment of the rH]methyl group to C-3.

\
/- 10

[I;:]
CH,CO,H
HCHO
+
HC+O,H
H,CCO,H
+
C6,H HC0,H
I
CHOH CAO
I I
CH3 CH3
SCHEME3.-Reactions Performed in the Structural Assignment ofthe Intermediate in
the Biosynthesis of L-Vinelose.
88 HANS GRISEBACH

When the crude enzyme-preparation was separated into five

fractions by stepwise precipitation with ammonium sulfate, a partial
separation of the X2-formingactivity from the X,-forming activity was
achieved. It was also found that one of the fractions catalyzes the
conversion of CDP-X2 into CDP-X,. Although it has not yet been
possible to assign a definite structure to sugar X,, the experimental
evidence strongly suggests that it is an isomer of sugar X,.
On the basis of these results, the authorsi6 proposed the reaction
sequence shown in Scheme 4 for the biosynthesis of CDP-L-vinelose
from CDP-D-glucose. As with L-mycarose, the 6-deoxyhexos-4-ulose
derivative (11) is the substrate for the methylation step leading to the
CDP-6-deoxy-3-C-methylhexos-4-ulose (12). The corresponding
dTDP-sugar might, therefore, be one of the unidentified reaction-
products in the bias ynthesis of dTDP-L-mycarose.

SCHEME4.-Proposed Pathways for the Conversion of CDP-D-glucose into CDP-L-

vinelose. (The configuration of 12 is still speculative.)

For the conversion of CDP-D-glucose into CDP-L-vinelose, the

following reactions must be considered: ( a ) reduction at C-6, (b)
methylation at C-3, (c) inversion of configuration at C-3 and C-5, (d)
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 89

stereospecific reduction at C-4, and (e) methylation of the 2-hydroxyl

group. In Scheme 4,the epimerization at C-3 and C-5 is postulated to
occur after the methylation step. This assumption has, however, not
been proved true, as the configuration of 12 is still unknown. A 3,5-
epimerase was found to be involved in the biosynthesis of dTDP-L-
rhamnose from dTDP-~-glucose.''~'~ This epimerase converts its
substrate dTDP-6-deoxy-~-xyZo-hexos-4-uloseinto enzyme-bound
dTDP-6-deoxy-~-lyxo-hexos-4-ulose.By analogy, it could be assumed
that, in the biosynthesis of L-vinelose, epimerization of 11 at C-3 and C-
5 takes place before methylation at C-3.
Methylation at 0-2 must occur late in the sequence, because X, (9)
and X3 do not possess an 0-methyl group.

2. Sugars Having a Two-Carbon Branch

a. D-A1dgarose.-Aldgamycin E, from cultures of Streptomyces
ZavenduZae, is a macrolide antibiotic's*20 containing, besides D-
mycinose, the (1-hydroxyethy1)-branched sugar D-a1dgarose2'(13).The
stereochemistry at C-7 of aldgarose has been clarified in connection
with a total synthesis of this s ~ g a r . ~ ~ J ~

0
13

For studies on the biosynthesis of D-aldgarose, growing or resting

cells of S . lavendulae were used. ~-[methyl-'~C]Methionine, L-[ethyl-
''C]ethionine, [1-14C]acetate, D-[U-'4C]glUCOSe, [2-'4C]pyruvate, and
[3-14C]pyruvate were t e ~ t e d as~ ~potential
,~~ precursors for the 1-

(17) R. W. Gaugler and 0. Gabriel,]. Biol. Chem., 248, 6041-6047 (1973).

(18) A. Melo and L. Glaser,]. Biol. Chem., 243, 1475-1478 (1968).
(19) H. Achenbach and W. Karl, Chem. Ber., 108,759-771 (1975).
(20) H. Achenbach and W. Karl, Chem. Ber., 108,780-789 (1975).
(21) G. A. Ellestad, M. P. Kunstmann, J. E. Lancaster, L. A. Mitscher, and G. Morton,
Tetrahedron, 23,3893-3902 (1967).
(22) H. Paulsen and H. Redlich, Chem. Ber., 107,2992-3012 (1974).
(23) J. S. Brimacombe, C. W. Smith, and J. Minshall, Tetrahedron Lett., 2997-3000
(1974).
(24) R. Schmid, H. Grisebach, and W. Karl, Eur. J. Biochem., 14,243-252 (1970).
(25) R. Schmid and H. Grisebach, Z. Naturforsch. Teil B , 25, 1259-1263 (1970).
90 HANS GRISEBACH

hydroxyethyl branch of 13. The degradation reactions outlined in

Scheme 5 were used for localization of 14C-activity.Methanolysis of
aldgamycin E yielded methyl aldgaroside and methyl mycinoside,
which were separated by thin-layer and gas-liquid chromatography.
Methyl aldgaroside was treated with sodium hydroxide solution to
saponify the cyclic carbonate. Periodate oxidation of the free sugar
yielded acetaldehyde from the 1-hydroxyethyl branch (containing C-7
and C-8) and ~-(-)-3-hydroxybutanoic acid from C-3 to C-6 of
aldgarose. The latter acid could be identified by oxidation with D-( -)-3-
hydroxybutanoic acid dehydrogenase (EC 1.1.1.30).Direct periodate
oxidation of aldgamycin C (see Scheme 5 ) also yielded acetaldehyde
and the butanoic acid from aldgarose, leaving the rest of the molecule
intact. This was proved by a mass-spectrometric i n v e ~ t i g a t i o n . ~ ~

SCHEMES.-Chemical Degradation of Aldgamycin E.

 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 91

By application of these degradation reactions, the following results

were obtained with the precursors already mentioned. L-Methionine,
L-ethionine, acetate, and Dglucose did not function as precursors of
the hydroxyethyl branch of aldgarose. In all cases, the incorporation of
carbon-14 into C-7 and C-8 of aldgarose was negligible. By contrast, 44-
45% of the radioactivity of aldgarose was located in C-7 and (2-8, with
[2-14C]- and [3-14C]pyruvateas the source of radioactive carbon. The
distribution of carbon-14 in the two-carbon branch was determined by
Kuhn-Roth oxidation of the acetaldehyde (2,4-dinitrophenyl)hydra-
zone, followed by Schmidt degradation of the acetic acid. In this way, it
was found that 88% of the carbon-14 was located in C-7, with [2-
14C]pyruvateas the precursor, and 93% in C-8, with [3-‘4C]pyruvate as
the precursor, proving the specific incorporation of C-2 and C-3 of
pyruvate into the hydroxyethyl branch. D-[U-14C]Glucosewas almost
exclusively incorporated into the hexose portion of D-aldgarose, with
equal distribution of activity in the hexose chain.
No radioactivity was found in the cyclic carbonate group with the
aforementioned precursors, but 45% of carbon-14 (related to aldgamy-
cin E) from sodium hydrogen[14C]carbonatewas located in this group.
It may, therefore, be assumed that a carboxylation reaction is involved
in the formation of the cyclic carbonate.
As aldgamycin C is also a fermentation product of Strepomyces
lauendulae,21formation of the cyclic carbonate probably occurs at the
stage of aldgamycin C.
On the basis of these results, it was postulated24that 1-(hydroxy-
ethy1)thiamine pyrophosphate (HTPP, “active acetaldehyde”) is the
actual precursor for the hydroxyethyl branch of aldgarose. Reaction of
HTPP with a hexos-8ulose would lead to the intermediate 14 (see
Scheme 6). Stereospecific reduction of the carbonyl group in 14 would
then lead to 15. Strong support for the formation of a methylcarbonyl
sugar such as 14 has come from work on the biosynthesis of the
quinocycline sugars (see Section II,2,b). However, experiments with
oxythiamine and pyrithiamine, which are antivitamins of thiamine
pyrophosphate , showed no inhibition of the formation of aldgamycin
E. The mechanism of the attachmant of the two-carbon branch is,
therefore, still undecided (see also, Section 11,2,b).

b. Sugars of the Quinocycline Complex.-The anthracycline

antibiotics 16, namely, quinocyclines A and B and isoquinocyclines A
and B from Streptomyces aureofaciens (FD 11188 Pfizer) contain as
glycosidic components the (1-hydroxyethy1)-branched sugar 17 [2,6-
dideoxy-4-C-(1-hydroxyethy1)-L-xylo-hexopyranose] in the A compo-
92 HANS GRISEBACH

OH
I
H,C--C-R'

HO H,C -C
6 OR OH
14

15
SCHEME6.-Hypothesis for the Introduction of the Two-carbon branch of Aldgarose.
(R' = thiamine pyrophosphate or group of an enzyme.)

nents, and the methylcarbonyl-branched sugar 18 [2,6-dideoxy-4-C-

acetyl-~-xy20-hexopyranose]in the B components.2s
The structural identification of the sugar from the B components as a
C-acetyl-branched sugar constituted strong support for the conclusion
reached from work on the biosynthesis of D-aldgarose (see Section
II,2,a) that a C-acetyl-branched sugar is the primary product in the
attachment of the two-carbon branch (see Scheme 6).
The incorporation of [2-'4C]pyruvate and [l-'*C]acetate into sugars 17
and 18 was investigated." Oxidation of the methyl glycosides of sugar
17 with periodate yielded acetaldehyde from the l-hydroxyethyl
branch. The acetaldehyde (2,4-dinitrophenyl)hydrazonewas further
oxidized by Kuhn-Roth oxidation to acetic acid, which was degraded
by the Schmidt reaction to methylamine and carbon dioxide. Periodate
oxidation of the methyl glycosides of sugar 18 produced acetic acid
from the C-acetyl branch. The acetic acid was isolated, and purified as
l-acetamidonaphthalene.
The following results were obtained with these degradation
reactions. [2-14C]Pyruvate was specifically incorporated into the 1-
hydroxyethyl branch of 17 and into the C-acetyl branch of 18. In each
instance, -90% of the radioactivity of the methyl glycosides was

(26) U. Matern, H. Grisebach, W. Karl, and H. Achenbach, Eur. J . Biochem., 29, 1-4
(1972).
(27) U. Matern and H. Grisebach, Eur. J. Biochem., 29, 5-11 (1972).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 93

16

17 18

Structures of the Sugar Components (17 and 18) of the

Quinocycline Complex and of the Aglycon of Isoquinocyclines
(Isoquinocycline A 16, with R = 17; isoquinocycline B
16, with R = 18.)

located in the two-carbon branch. Further degradation proved that

carbon-14 was present only in C-7 of this branch. In contrast, no
incorporation of radioactivity from [l-14C]acetate into the two-carbon
branch was found.
[‘4C]Quinocycline B (obtained from [l-14C]acetate)containing sugar
18 was efficiently converted into [14C]quinocyclineA by growing cells
of Streptornyces uureofuciens. Under the same conditions, only a very
low conversion of quinocycline A into quinocycline B took place. It
may, therefore, be concluded that quinocycline B is the precursor of
quinocycline A.
Neopyrithiamine, which is a competitive inhibitor of thiamine, had
no influence on the incorporation of [2-’4C]pyruvate into the two-
carbon branch.
On the basis of these results, the following, metabolic sequence
leading to quinocycline B and quinocycline A was postulated: XDP-
2,6-dideoxy-~-erythr-hexos-4-ulose(X = nucleoside residue) reacts
with an activated, two-carbon species originating from pyruvate, to
form a “nucleotide sugar” of 18. Sugar 18 is then transferred, with
inversion of configuration at C-1, to the aglycon to form quinocycline B,
and the carbonyl group in sugar 18 is reduced, to yield quinocycline A.
Further insight into the reaction sequence leading to the two-carbon-
branched sugars was obtained with a cell-free extract from S.
uureofuciens.28 Incubation of dTDP-[U-14C]glucosewith this extract,
(28) U. Matem and H. Grisebach, 2. Naturforsch. Teil C, 29,407-413 (1974).
94 HANS GRISEBACH

and subsequent hydrolysis of the “nucleotide sugars” led to the

formation of four radioactive products (I-IV). Addition of pyruvate to
the incubation caused an increase in the amount of 111, and a decrease
in that of 11. Product I was identified as 6-deoxy-~-xyZo-hexos-4-ulose
by its absorption spectrum after addition of sodium and by
reduction with sodium borohydride to give 6-deoxyglucose and 6-
deoxygalactose.
Incubation of dTDP-6-deoxy-~-xyEo-[U-’~C]he~0~-4-~lo~e with the
cell-free extract led to the same products 11-IV as with dTDP-D-[U-
14C]glucose as substrate. Product I11 was isolated as the methyl
glycoside from an incubation performed on a preparative scale.
The n.m.r. spectrum of I11 was very similar to that of the methyl
glycoside of 18.The signal of H-3 was, however, shifted by 0.82 p.p.m.
to lower field compared with the corresponding signal of 18,and had a
large coupling-constant of 11 Hz, indicating trans-diaxial coupling.
Furthermore, the H-5 signal was shifted by 0.42p.p.m. to higher field.
These data were consistent with the assumption that, in 111, the 3-
hydroxyl group is equatorial (not axial, as in 18).This conclusion was
supported by the fact that I11 formed a borate complex (with the
hydroxyl groups at C-3 and C-4),whereas 18 did not form such a
complex.
On the assumption that I11 belongs, as does 18, to the L series,
structure 19 may be proposed for the enzymic product from dTDP-6-
deoxy-~-xy2o-hexos-4-ulose and pyruvate.
On reduction with sodium borohydride, product I1 yielded a pair of
radioactive products that were tentatively identified as 2-deoxy-~-Zyxo-
hexose and 2,6-dideoxy-~-urubino-hexose. This result, together with
the chromatographic properties of 11, and its reaction on chromato-
grams with vanillin-perchloric acid to give a blue color, are consistent
with the structure of 2,6-dideoxy-~-Zyxo-hexos-4-ulose (20)for 11. This
structure is also consistent with the concept that I1 (20) is, very
probably, the precursor of I11 (19).

19 20

(29) R.Okazaki, T. Okazaki, J. L. Strominger, and A. M. Michelson,]. Biol. Chem., 237,

3014-3026 (1962).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 95

When a mixture of dTDP-~-[3-~H]glucoseand dTDP-D-[U-

''C]glucose was incubated with the cell-free extract, the 3H/'4Cratio in
the 6-deoxy-~-xyZo-hexosuloseremained unchanged. In contrast, there
was no tritium in products 19 and 20. This result is strong evidence for
the participation of a 3,5-epimerase13,14in a reaction converting dTDP-
6-deoxy-~-xylo-hexos-4-ulose into an enzyme-bound dTDP-6-deoxy-
~-Zyxo-hexos-4-ulose(21, see Scheme 7) which is then reduced at C-2 to
yield dTD P-2,6-dideoxy-~-lyxo-hexos-4-ulose. The mechanism for the
introduction of the deoxy function at C-2 is still undecided.
Preliminary experiments showed no influence of pyridoxamine 5'-
phosphate on the conversion of dTDP-~-[U-'~C]glucoseinto 19.
(Pyridoxamine 5'-phosphate is a cofactor in the 3-deoxygenation step in
the biosynthesis of CDP-3,6-dideoxyhe~oses.~~)
On the basis of the results discussed and of studies on the time-
dependence of formation of product from dTDP-D-glucose, the
sequence of reactions shown in Scheme 7 was postulatedz8 for the
biosynthesis of the C-acetyl-branched sugar 19. It is at present
unknown why the sugar formed in the cell-free system is the C-3
epimer of sugar 18 from quinocycline B.

SCHEME7.-Reactions Proposed for the Biosynthesis of the Oxoethyl-branched

dTDP-sugar 19 from dTDP-D-glucose in a Cell-free System from S. uureofuciens.

Further experiments were undertaken to investigate the possible

participation of thiamine pyrophosphate in the carboligase reaction.z8
Addition of TPP, Mgz+,and pyruvate to the incubation mixture gave no

(30) P. Gonzalez-Porque and J. L. Strominger,J. Biol. Chem., 247,6748-6756 (1972).

96 HANS GRISEBACH

increase of product I11 (19).The cofactor for the ligase reaction could
be partially removed by gel filtration through Sephadex G-25,but no
restoration of activity by TPP and Mg2+ was obtained. Finally,
incubation of dTDP-D-glucose and l-([l-'4C]hydroxyethyl)thiamine
pyrophosphate (HETPP) gave no radioactive product 111.
Despite these negative results, a participation of HETPP cannot be
completely excluded, because HETPP cannot be regarded as a
genuine cofactor, as it must first be transformed into the carbanion by
enzyme catalysis.31
c. Pil1arose.-Pillaromycin A, an antitumor, antibiotic substance
from cultures of Streptomycesflavovireus,is composed of a tetracyclic
aglycon 22, and a branched-chain monosaccharide residue named
pillarose. The structure of a 2,6-dideoxyhexos-4-ulosehaving a
glycoloyl branch at C-2 (23)was originally proposed for p i l l a r o ~ e . ~ ~
However, crystallographic and mass-spectral evidence,= as well as
synthetic led to a revised structure of pillarose; it was
identified as 2,3,6-trideoxy-4-C-(2-hydroxyacetyl)-~-threo-aldohexose
(24).

H OH
i

HO OH 0 0

22

C=O
I
CH,OH
23 24
Structures of Pillaromycin (22, R = 24 -H on 0-1) and Pillarose (24),
and Structure (23) Previously Proposed for Pillarose

(31) J. Ullrich and A. Mannschreck, Eur. J. Biochem., 1, 110-116 (1967).

(32) M. Asai, Chem. Pharm. Bull., 18, 1713 (1970).
(33) J . 0. Pezzanite, J . Clardy, P. Y. Lau, G . Wood, D. L. Walker, and B. Fraser-Reid,
J . Am. Chem. SOC., 97,6250-6251 (1975).
(34) D. L. Walker and B. Fraser-Reid,J. Am. Chem. SOC., 97,6251-6253 (1975).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 97

Although no biosynthetic studies on pillarose have as yet been

published, the close structural relationship of pillarose to sugar 19 from
quinocyline B (see Section II,2,b) permits postulation of the following
reactions. Linkage of an activated glycolaldehyde to a 2,3,6-trideoxy-
hexos-4-ulose leads directly to 24. Glycolaldehyde may originate from
hydroxypyruvate with the enzyme hydroxypyruvate d e c a r b ~ x y l a s e ~ ~
(EC 4.1.1.40) and “active glycolaldehyde” from hydroxypyruvate or a
ketose donor with the enzyme transketolase (EC 2.2.1.1).

3. General Conclusions on the Biosynthesis of Methyl-branched

Sugars and on Sugars Having a Two-Carbon Branch
The foregoing results permit some general conclusions regarding the
biosynthesis of methyl-branched sugars and of sugars having a two-
carbon branch. In each example investigated, the acceptor for the

M * O O X D P

XDP-a-glucose L-noviose

(a) C -methylation
I (a) C- and O-methyla-
tion
(b) carbamoylation

C-2 and C-4 ‘ Q HO

O X D P ]
CHS OH
D-Evermicose (a) C-methylation
b) reduction at
C-2and C-4

L-mycarose L-olivomycose
SCHEME8.-General Scheme for the Biosynthesis of Methyl-branched Sugars. [(i) C -
Methylation by AdoMet, with inversion of configuration; (ii) C-methylation with
retention of configuration.]

(35) J. L. Hedrick and H. J. Sallach, Arch. Biochem. Biophys., 105,261-269 (1969).

98 HANS GRISEBACH

branch is a nucleotide-bound hexos-4-ulose in which the carbon atoms

adjacent to the 0x0 group are activated for attachment of the
electrophilic methyl group from S-adenosyl-L-methionine or an
activated, two-carbon fragment. In most cases, C-methylation takes
place at C-3 (as in L-mycarose), but it may also occur at C-5 (as in L-
noviose). An exception is L-garosamine (see Section III,2) having a4-C-
methyl group. All other known examples of methyl-branched sugars5
can be assumed to be derived from XDP-6-deoxy-~-xy2o-hexos-4-ulose
by the action of known enzymes, as shown in Scheme 8. Obviously, C-
methylation can occur with retention, or inversion, of configuration.
Other variations arise by deoxygenation at C-2, and by reduction of the
keto group at C-4 (leading to different stereochemistry).

4. Sugars Having a Formyl or Hydroxymethyl Branch: L-Streptose and

L-Dihydrostreptose
The work of Baddiley and coworkers and of Bruton and Horner with
labelled precursors in connection with the biosynthesis of L-streptose
(25) from streptomycin has been r e v i e w e d . 4 ~It
~ *was
~ ~ found that the
aldehyde branch of streptose originates from C-3 of glucose, and that
the hexose unit as a whole is incorporated into the streptose molecule
(see Scheme 9).

OH HO OH
25
SCHEME 9.-Positions of the Carbon-14 label in L-Streptose (25) from D-Glucose
Labelled at C-1 (A), C-3 (o), or C-6 (m).

On the basis of the observation that dTDP-D-mannose and dTDP-L-

rhamnose occur in Streptornyces g r i s e ~ s , ~and
' that cell-free prepara-
tions from this organism can convert both dTDP-D-glucose and dTDP-
D-mannose into dTDP-~-rhamnose,~* it was suggested that the
biosynthesis of streptose and that of L-rhamnose are related.
(36) J. Walker, Lloydia, 34,363-371 (1971).
(37) N. L. Blumson and J. Baddiley, Blochem. J., 81, 114-124 (1961).
(38) J. Baddiley, N. L., Blumson, A. Di Girolamo, and M. Di Girolamo, Biochim.
Biophys. Acta, 50,391-393 (1961).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 99

Further insight into the biosynthesis of streptose was gained in

experiments in which dTDP-~-[U-'~C]glucose was incubated with a
cell-free extract from a streptomycin-producing strain of S. g r i s e ~ sIn
.~~
the presence of NADPH, and after subsequent hydrolysis of the
"
nucleotide sugars," a new radioactive product was formed that was
identified as dihydrostreptose (26, see Scheme 10) by paper chro-
matography, paper electrophoresis, gas chromatography of the tri-
methylsilyl derivative, and complex-formation with phenylboronic
acid in molybdate b ~ f f e r . 3Besides
~ dihydrostreptose, rhamnose and
a 6-deoxyhexos-4-ulose were formed during the incubation.
A compound having the properties of the dTDP derivative of
dihydrostreptose was apparently very labile, because it decomposed
readily to a compound tentatively identified as dihydrostreptosyl
phosphate. The latter was also very labile, and gave free dihydrostrep-
tose upon attempted paper chromatography.
dTDP-dihydrostreptose and dTDP-L-rhamnose were also obtained
when dTDP-6-deoxy-~-xyZo-hexos-4-ulosewas used as the substrate.
Only NADPH was needed as a cofactor for the reaction. NADH had
-50-60% of the activity of NADPH. No dihydrostreptose or streptose
could be detected when NADPH was replaced by NADP+ or NAD+.
The formation of dihydrostreptose could be completely inhibited b y
the addition of 1 mM p-(chloromercuri)benzoate, whereas formation of
L-rhamnose was not inhibited by this reagent.
As Streptomyces contain high proportions of proteases, a protease
inhibitor, diphenylcarbamoyl chloride, was added to the buffer during
purification of the "dTDP-dihydrostreptose synthase." A partially
purified, enzyme preparation from S. griseus could be obtained by
removal of nucleic acids with streptomycin and fractionation with
ammonium sulfate.4o However, when this enzyme preparation was
subjected to gel filtration on a column of Sephadex G-100, enzyme
activity was completely lost. By combining certain fractions of the
column eluate, enzyme activity could be partially restored.
It was, therefore, assumed that separation into two, or more, active
protein-fractions had occurred on the Sephadex column. As the
biosynthesis of dTDP-L-rhamnose and dTDP-L-dihydrostreptose are
related,37*39and as, moreover, a dTDP-L-Zym-4-hexulose3,5-epimerase
is necessary for the formation of dTDP-~-rhamnose,"*'~ the Sephadex
G-100 fractions were assayed for the presence of the 3,5-epimerase by

(39) R. Ortmann, U. Matern, H. Grisebach, P. Stadler, V. Sinnwell, and H. Paulsen, Eur.

J . Biochem., 43,265-271 (1974).
(40) P. Wahl, U. Matern,and H. Grisebach,Biochem. Biophys. Res. Commun., 64,1041-
1045 (1975).
100 HANS GRISEBACH

determining the loss of tritium from dTDP-6-deoxy-~-[3-~H]xyZo-

he~os-4-ulose.'~ A sharp peak of epimerase activity was found; this was
cleanly separated from a second protein fraction that catalyzed the
synthesis of dTDP-dihydrostreptose from TDP-6-deoxy-~-xylo-hexos-
4-ulose in the presence of NADPH and the 3,5-epimerase. These
results, taken together with the results already discussed, proved that
the biosynthesis of dTDP-L-dihydrostreptose from dTDP-D-glucose
requires three enzymes: dTDP-D-glucose 4,6-dehydrataseYdTDP-L-
Zyxo-4-hexulose 3,5-epimeraseYand an NADPH-dependent dTDP-
"dihydrostreptose synthase" (see Scheme 10).

dTDP-o-glucose -dTDP-o-Clucose
4 , 6 - dehy dratas e

I
OH

dTDP- ~ - " r l i k y d ~ o
ntrrptos e synlkasr "

HO OH HO OH
21 26
SCHEME10.-Biosynthesis of dTDP-L-dihydrostreptose from dTDP-D-glucose.

By analogy to results obtained in studies on the biosynthesis of

dTDP-L-rhamnose,'* dTDP-6-deoxy-~-talose,'~ and GDP-~-fucose,~l it
seems very likely that dTDP-6-deoxy-~-Zyxo-hexos-4-ulose (21),
formed by the 3,5-epimerase reaction, remains enzyme-bound.
The biosynthesis of dTDP-dihydrostreptose from dTDP-D-glucose
by way of dTDP-6-deoxy-~-xyZo-hexos-4-ulose shows a close similarity
to the biosynthesis of UDP-D-apiose" from UDP-D-glucuronic acid in
higher plants5 (see Scheme 11). In both cases, the hydroxymethyl
branch originates by ring contraction of a 4-ketoseYwith expulsion of

(41)V. Ginsburg,J. B i d . Chem., 236,2389-2393(1961).

(42)R. R. Watson and N. S. Orenstein, Adv. Carbohydr. Chem. Biochem., 31,135-184
(1975).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 101

C-3. However, for apiose, the formation of the enzyme-bound glycos-4-

ulose i n t e m ~ e d i a t e ,and
~ ~ its rearrangement and reduction at the
branched carbon atom, are catalyzed by one enzyme.
In experiments in which sodium borohydride was added to the
enzyme incubation mixture of the apiose/xylose synthase, no reduction
products could be found that would arise from epimerization at C-3 of
the 4-ketose intermediate.44In the synthesis of apiose, ring contraction
therefore seems to take place with the ~-threo-pentos-4-ulose27 (see
Scheme ll),and not with the L-erythro intermediate, which would be
analogous to the ring contraction in the biosynthesis of L-dihydro-
streptose.

QOUDP

OH HA bH
UDP-~-glucuronic acid UDP-D-apiose
SCHEME11.-Biosynthesis of UDP-D-apiose from UDP-D-glucuronicAcid.

When the dihydrostreptose synthase reaction was conducted in the

presence of either (4-pr0-R-~H) NADPH or (4-pro-SPH)NADPH, label
in dihydrostreptose appeared44ain both cases. This apparent lack of
stereospecificity for the hydride transfer needs clarification.
The unexpected result that dihydrostreptose, but not streptose, was
formed under all conditions, either in the crude extract of S . griseus or

(43) C. Gebb, D. Baron, and H. Grisebach, Eur. J . Biochem., 54,493-498 (1975).

(44) U. Matern and H. Grisebach, Eur. J . Biochem., 74,303-312 (1977).
(44a) H. P. Wahl, unpublished results.
102 HANS GRISEBACH

in the partially purified synthase was understood only later, when the
fermentation products of S . griseus were investigated more closely.45
The products formed were separated by paper chromatography and
electrophoresis, and detected by bio-autography. It could be shown
that dihydrostreptomycin is a normal product of a streptomycin-
producing strain of s. griseus. At all stages of fermentation, only
dihydrostreptomycin was found inside the mycelium, whereas
dihydrostreptomycin and streptomycin were present in the medium.
From these results, it was concluded that dihydrostreptomycin is the
primary product in the biosynthesis of streptomycin, and that
dihydrostreptomycin is oxidized to streptomycin.

111. AMINOCYCLITOL
ANTIBIOTICS

The aminocyclitol antibiotics-gentamicins, kanamycins, neomy-

cins, paramomycins, spectinomycins, streptomycins, and tobramy-
cins-constitute a group of basic oligosaccharides that have a broad,
antibacterial
Instead of describing the biosynthesis of the individual sugar
components of these antibiotics, it is advantageous to review the
biosynthesis of the total molecule. A review on the biosynthesis of
aminocyclitol (“aminoglycoside”) antibiotics has appeared?’

1. Streptomycins and Bluensomycin

Streptomycin and dihydrostreptomycin are45normal fermentation-
products of S . griseus (see Section 11,4). Certain Streptomyces strains
produce mannosido-streptomycin and hydroxystreptomycin (see 28).
Bluensomycin (29) and glebomycin are apparently identical4*;29 is a
monoguanidinated analog of dihydrostreptomycin in which a carba-
moyl group replaces4s the guanidino group at C-1.
The biosynthesis of the dihydrostreptose moiety has already been
described. The biosynthesis of the streptidine and 2-deoxy-2-(methyl-
amino)-L-glucose moieties, and their assembling to give the antibiotic,
will now be discussed. Furthermore, the biosynthesis of the bluensi-
dine moiety of bluensomycin will be compared with that of streptidine.

(45)S. Maier, U. Matern, and H. Grisebach, FEBS Lett., 49,317-319 (1975).

(46)R. Reiner, “Antibiotika,” G. Thieme Verlag, Stuttgart, 1974,pp. 136-147.
(47)K. L.Rinehart, Jr., and R. M. Stroshane,J. Antibiot., 29,319-353 (1976).
(48)M. Okanishi, H. Koshiyama, T. Ohmari, M. Matsuzaki, S. Ohashi, and H.
Kawaguchi,J. Antibiot., Ser. A, 15, 7-13 (1962).
(49)C. B. Barlow and L. Anderson,J. Antibiot., 25,281-286 (1972).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 103

YH2
C=NH
I

I1
H,N -C--NH

R2H,CQ
d

H
28

Variations in the Structure of Streptomycins (28)

(Streptomycin, R' = CHO; R* = H;Rs = OH. Dihydrostreptomycin,
R' = CH,OH, Rz = H, R s = OH. Mannosidostreptomycin,
R1 = CHO, Rz = H , Rs = p-D-mannosyl. Hydroxystreptomycin,
R' = CHO, R2 = OH, Rs = OH.)

II
H,N-C-NH

Bluensomycin
29

a. Streptidine 6-Phosphate.-Extensive investigations by Walker

and coworkers have shown that the streptidine moiety of streptomycin
is synthesized from D-glucose 6-phosphate by way ofmyo-inositol; this
104 HANS GRISEBACH

work has been covered in two review^,^^*^^ and is therefore only

summarized here. The sequence of reactions that are now assumed to
occur in the formation of streptidine 6-phosphate from D-ghCOSe 6-
phosphate is shown in Scheme 12. It involves two analogous
sequences of five enzymic reactions each, operating in series: a
dehydrogenation, transamination, phosphorylation, transamidation,
and dephosphorylation. The individual enzymic reactions have been
described by Walker.51D-Glucose 6-phosphate is cyclized to ~ - m y o -
inositol l-phosphate (30),as has been observed with enzymes from
higher and lower plants, as well as from mammals.52After dephospho-
rylation, myo-inositol (31)is oxidized to 2-keto-scyllo-inositol (32),
which by transamination gives l-amino-l-deoxy-scyllo-inositol (33).
After phosphorylation of 33 by a kinase, to give l-amino-l-deoxy-scyllo-
inositol4-phosphate (34),an amidinotransferase transfers the amidine
residue from arginine to the amino group of the inosamine, to afford
N-amidino-O-phosphono-scy llo-inosamine (35).After dephosphoryla-
tion of 35 to give 36,oxidation at C-3 leads to N-amidino-3-keto-scyllo-
inosamine (37),which is transformed in a second transamination step
into N-amidinostreptamine (38).The latter is rephosphorylated, to give
39,which undergoes further amidination, to yield streptidine 6-phos-
phate (40).
For certain pairs of corresponding reactions in the two sequences, it
is known that different enzymes are involved, although some have
overlapping substrate-specificities. Additional information on some of
these enzymes was obtained in studies on the biosynthesis of
bluensidine (see Section III,l,b). Rinehart and have
pointed out that earlier analysis did not allow an unequivocal
correlation between the labelling pattern in streptidine obtained with
differently labelled D-glUCOSe and the label of the D-glucose
precursors. As they found that biosynthesis of deoxystreptamine in S .
fradiae follows53an alternative pathway (see Scheme 20 and Section
111,3), the biosynthesis of streptidine was reinvestigated with D-[6-
13C]glucoseas the The 13C-n.m.r.spectrum of streptomycin
showed resonances for each ofthe 21 carbon atoms.55The streptomycin

(50) A. L. Demain and E. Inamine, Bacteriol. Rev., 34, 1-19 (1970).

(51) J. B. Walker, Methods Enzymol., 43,429-470 (1975).
(52) F. Pittner, W. Fried, and 0.Hohann-Ostenhof, Hoppe-Seyler’s Z. Physiol. Chem.,
355,222-224 (1974), and references cited therein.
(53) K. L. Rinehart, Jr., J . M. Malik, R. F. Nystrom, R. M. Stroshane, S. G. Truitt, M.
Taniguchi, J. P. Rolls, W. J. Haak, and B. A. RuffJ. Am. Chem. Soc., 96,2263-2265
(1974).
(54) M. H. G. Munro, M. Taniguchi, K. L. Rinehart,Jr., D. Gottlieb, T. H. Stoudt, and T.
0. Rogers,]. Am. Chem. SOC., 97,4782-4783 (1975).
(55) M. H. G . Munro and K. L. Rinehart, Jr.,J. Am. Chem. Soc., in press.
 OH OH OH
30 31 32 33

NH NH NH NH
1
I II II

HO’ HO’
OH OH OH OH
36 37 36 35 34

NH NH NH

m
cn

OH ODSBA
3s Dihydrostreptomycin 41
40

SCHEMEl2.-Sequence of Reactions Leading from D-Glucose 6-Phosphate to

Skeptidine 6-Phosphate (40), or Bluensidine (41, R = H). (Abbreviations: DSBA,
tt r
dihydrostreptobiosamine;Gln, Glutamine; aKGN, a-keto-glutaramate;Om, ornithine; Bluensomycin 0
u1
Arg, arginine; Ala, alanine; 08,
phosphate and pyr, pyruvate.) 29
106 HANS GRISEBACH

isolated from the experiment with D-[6-'3C]glucose displayed three

enhanced resonances, at 13.4, 61.2, and 72.4 p.p.m. from tetramethyl-
silane, which are the respective signals for C-5' of streptose, C-6" of 2-
deoxy-2-(methylamino)-~-glucose, and C-6 of streptidine. Labelling of
C-6 of streptidine b y ~-[6-'~C]glucoseis consistent with path A (see
Scheme 13), proposed earlier on the basis of labelling studies and
enzymic work (compare Scheme 12),but not with path B, which leads
to deoxystreptamine.

OOH
CH,OH

HO = Q= Q OH
HO OH
OH
HO OH

32

NHR NHR

I
OH OH
SCHEME 13.-Biosynthetic Conversion of D-Glucose into Streptidine by way of Keto-
scyllo-inositol 32 (compare Scheme 12), with C-6 of D-Glucose Labelling C-6 of
Streptidine by Path A. (Path B is analogous to the pathway found for formation of
deoxystreptamine by S. frudiue.)

Support for the assumption that streptidine or its phosphorylated

derivative is an intermediate in the synthesis of streptomycin came
through studies with a mutant of S . griseus which is presumably
blocked in the biosynthesis of streptidine or streptidine phosphate.56
This mutant only produced antibiotics when the medium was
supplemented with streptidine dihydrogensulfate. The maximal
amount of antibiotic (530 pg/ml, calculated as streptomycin) was
produced in the presence of 1.000 p g of streptidine per ml. Two
substances having antibiotic activity were produced; they respectively
had chromatographic mobilities identical with those of streptomycin
and mannosidostreptomycin. Several other aminocyclitols and guani-
docyclitols were also tested for their ability to support production of
new antibiotics by the mutant. Only with 2-deoxystreptidine was a new
antibiotic produced; this was named streptomutin A. Presumably,
(56) K. Nagaoka and A. L. Demain,]. Antibiot., 28,627-635 (1975).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 107

streptomutin A is the streptomycin analog having a 2-deoxystreptidine

moiety, but the structure has not yet been proved.
b. B1uensidine.-Bluensidine (41,R = H;see Scheme 12) is the
monoguanidinated inositol moiety of bluensomycin (29).The bio-
synthesis of bluensidine was studied with cell-free extracts of S.
glebosus, and compared with the biosynthesis of ~treptidine.~'
The following enzymic reactions were observed. Dialyzed extracts
of S. glebosus catalyzed the conversion of rnyo-[U-'4C]inositol (31,
Scheme 12)to aminodeoxy-scyllo-[U-14C]inositol (33)when both NAD+
and an amino donor were provided. NAD+was presumably required in
order to form ket~-scyllo-[U-'~C]inositol (32),and the amino donor (for
example, L-glutamine) was required in order to convert 32 into 33.The
transamination reaction was confirmed in a separate assay. Assays for
reactions H and I (see Scheme 12) with S. glebosus were negative.
The substrate specificity of S. glebosus amidinotransferase could not
be distinguished from the substrate specificity of the amidinotrans-
ferase of streptomycin producers.
In summary, extracts of S. glebosus could catalyze reactions C, D, F,
and G, but not reactions H and I; these reactions are apparently
catalyzed by enzymes similar to those involved in the biosynthesis of
the first guanidino group of the diguanidinated inositol (streptidine)
moiety of dihydrostreptomycin. The scheme for the biosynthesis of
bluensidine predicts a labelling pattern from D-glucose 6-phosphate,
different from that observed for the biosynthesis of streptidine.
Whereas C-5 and C-6 of streptidine arise from C-1 and C-6 of D-glucose
, ~ ~ ~ ~ ~13 predicts that C-2 of bluen-
6-phosphate, r e s p e ~ t i v e l y Scheme
sidine would be derived from C-6 O f D-glucose 6-phosphate. This is the
labelling pattern that has been found in the biosynthesis of
deoxystreptaminea (see Section 111,3).
c. 2-Deoxy-2-(methylamino)-~-glucose.-Notmuch information is
available on the biosynthesis of the 2-deoxy-2-(methylamino)-~-
glucose moiety (42),which is common to all streptomycins and to
bluensomycin.

CH,OH

HO
42

(57) J. B. Walker,]. Biol. Chent., 249,2397-2404 (1974).

(58) R. M. Bruce, H. S. Ragheb, and H. Weiner, Biochim. Biophys. Acta, 158,499-500
( 1968).
108 HANS GRISEBACH

Tracer studies have s ~ o w ~that~ the ~ individual

, ~ ~ , carbon
~ ~ atoms of
D-glucose are incorporated into the corresponding carbon atoms of 42.
It must, therefore, be assumed that inversion at all four chiral centers of
D-glucose takes place during the biosynthesis. Experiments with 2-
deoxy-2-(methylamino)-~-[’~C]glucoseand 2-amino-2-deoxy-~-[l-
l4 C]glucose gave predominant incorporation into 42, although it is not
clear if the amino group stays attached to the hexose moiety during this
conversion. Experiments with 15N-labelledprecursors are, therefore,
needed, in order to clarify this question. ~-[‘~C]Glucose does not seem
to be a precursor of 42.
Epimerization at C-2 of a “sugar nucleotide” has been described.
An enzyme from Escherichia coli catalyzes the epimerization of
UDP-2-acetamido-2-deoxy-~-glucoseto UDP-2-acetamido-2-deoxy-~-
m a n n o ~ e Such
. ~ ~ an epimerization, together with the 3,5-epimerase
reaction of a 4-ketose and stereospecific reduction at C-4, could lead to
inversion at all of the chiral centers of, for instance, 2-amino-2-deoxy-~-
glucose.
L-Methionine has been shown to be the source of the N-methyl
group. The following results led Heding and coworkers to the
conclusion that the last step in the biosynthesis of streptomycin is N -
methylation of N-demethylstreptomycin. By adding the methylation
inhibitor ethionine to the culture medium of S . griseus, it was possible
to isolate N-demethylstreptomycin.60 Addition of ~-[rnethylJ~C]me-
thionine and N-demethylstreptomycin to a culture led to labelled
streptomycin, which could have been formed either by de no00
synthesis or by methylation of the N-demethylstreptomycin.61 To
decide between these two possibilities, N-demethyldihydrostrepto-
mycin was used as the potential methyl-acceptor, because it was
assumed that this compound could be not a natural substrate. It was
found that [14C]dihydrostreptomycinwas formed in the presence of the
labelled methionine. N-Demethyldihydrostreptomycin also seemed to
inhibit de novo biosynthesis of streptomycin. Also, because, in the
presence of a larger excess of N-demethylstreptomycin and N -
demethyldihydrostreptomycin, both were converted into the corre-
sponding methylated products in the presence of L-methionine, it was
concluded that radioactivity had been incorporated by N-methylation
of these substrates, and not by de no00 biosynthesis.61The interpreta-

(59)T. Kawamura, N. Ichihara, N. Eshimoto, and E. Eto, Biochem. Biophys. Res.

Commun., 66, 1506-1512 (1975).
(60) H. Heding, Acta Chem. Scand., 22, 1649-1654 (1968).
(61) H.Heding and K. Bajpai,]. Antibiot., 26,725-727 (1973).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 109

tion of these results is, however, not unambiguous. Attempts to

demonstrate methylation of N-demethyldihydrostreptomycin with S-
adenosyl-~-['~C]methionine in a cell-free extract from S. griseus have
thus far been
d. Assembling of the Streptomycin Molecule.-Walker and co-
WorkerP postulated that L-streptose and 2-deoxy-2-(methylamino)-~-
glucose are transferred from the corresponding "nucleotide sugars" to
streptidine 6-phosphate and O-a-~-streptose-(144)-streptidine 6-
phosphate, respectively, to form streptomycin 6-phosphate.
With the possibility of preparing dTDP-~-[U-'~C]dihydrostreptose~~
(see Section II,4),61a*62the enzymic transfer of the dihydrostreptose
moiety to streptidine 6-phosphate could now be tested experimen-
tally.63
Streptidine phosphate was incubated with dTDP-~-[U-'~C]dihydro-
streptose, obtained in s i t u from dTDP-~-[U-'~C]glucosewith a cell-
free extract from s. griseus. Two new, positively charged, radioactive
products were obtained which, upon hydrolysis, gave dihydrostrep-
tose as the only radioactive product. Comparison with a synthetic
sample of 0-a-L-dihydrostreptose-(1+4)-streptidine (43,see Scheme
14) proved that one of the products was identical with this
pseudodisaccharide, and the second product was identified as the
corresponding 6-phosphate (44). It was concluded that the phospho-
rylated product had been partially hydrolyzed during the incubation
by streptomycin 6-phosphate phosphatase present in the cell-free
extract. No transfer-products were formed in controls in which ( a )
denatured extract was used, ( b )streptidine phosphate was absent, or(c)
streptidine was substituted for streptidine phosphate.
From these results, it may be concluded that the pseudo-di-
saccharide 44 is the first intermediate in the assembling of the three
components of dihydrostreptomycin. The fact that no transfer-product
was observed with streptidine is in agreement with the conclusions
from a number of studies in which the importance of phosphorylated
intermediates in the biosynthesis of streptomycin was estab-
lished.36,64,65

(61a) S. Maier and H. Grisebach, unpublished results.

(62) M. S. Walker and J. B. Walker,J. B i d . Chem., 246, 7034-7040 (1971).
(63) B. Kniep and H. Grisebach, FEBS Lett., 65,44-46 (1976).
(64) J. B. Walker and M. Skorvaga,]. BioZ. Chem., 248,2441-2446 (1973).
(65) 0.Nimi, H. Kiyohara, T. Mizoguchi, Y. Ohata, and R. Nomi,Agric. B i d . Chem., 34,
1150-1156 (1970).
110 HANS GRISEBACH

NH
II

-t 0 + dTDP
I
I
, I

HO OH OR

HO OH
43 R = H
44 R =-POSH,
SCHEME14.-Formation of 0-a-L-Dihydrostreptose-(1+4)-streptidine 6-Phosphate
(44) from dTDP-L-dihydrostreptose and Streptidine 6-Phosphate with a Cell-free Extract
from S . griseus.

2. Gentamicins
The gentamicin complex, produced by Micromonospora sp.,
consists66of the three major components, gentamicins C,, C2,and C,a
and a larger number of minor component^.^' The major C components
differ only in their degree of methylation, and are closely related to the
kanamycins. Minor components closely related to the C components
are CZb,a 6’-(aminomethy1)gentamicinCla, and C2a,a 6’-methyl epimer
of gentamicin C,. As shown, gentamicins Cla, C2,and C, respectively
possess one C-methyl and one N-methyl group, two C-methyl groups
and one N-methyl group, and two C-methyl and two N-methyl groups.
When ~-[methyl-’~C]methioninewas added to a growing culture of
Micromonospora purpurea, a high incorporation of radioactivity into
gentamicins was observed.68 As expected, the radioactivity incor-
porated into gentamicins increased with the number of C-methyl and
N-methyl groups in the molecule (Cia 3, C, 18, and C, 60% of the total
incorporation).
(66) D. J. Cooper, M. D. Yudis, R. D. Guthrie, and A. M. Prior,J. Chem. SOC., C. 960-
963 (1971).
(67) P. J. L. Daniels, in “Drug Action and Drug Resistance on Bacteria,” S. Mitsuhashi,
ed., Tokyo Univ. Press, Tokyo, 1975, pp. 77-111.
(68) B. K. Lee, R. T. Testa, G. H. Wagman, C. M. Lin, L. McDaniel, and C. Schaffner,
J . Antibiot., 26,728-731 (1973).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 111

OH
Structures of Some Components of the Gentamicin Complex
R' R2 R3
Gentamicin C, -NHMe -Me H
Gentamicin C, -NH, -Me H
Gentamicin C,, H H -m,
Gentamicin C,, H H -NHMe
Gentamicin C,, H -Me -NH,

Later studies, with ~-[methyl-'~C]methionine,in which the genta-

micin C components were separated and then hydrolyzed to the
individual sugars supported the earlier results.60
Further information on the origin of the methyl groups was
obtained'O from experiments with ~-[methyl-'~C]methionineand L-
[methyl-2H,]methionine, and analysis of the n.m.r. and mass spectra of
gentamicins C,, C2,and Cia. The 13C-n.m.r.spectrum of, for instance,
gentamicin C, showed enrichment of the two C-methyl and two N -
methyl groups, proving the origin of all four methyl groups by
transmethylation from L-methionine, probably by way of S-adenosyl-L-
methionine (see Section II,l,a,b).
High-resolution, mass spectroscopy using the [M + 13' peak gave the
following incorporation of deuterium in gentamicin C, : 1CD, 15.5%;2
CD, 3.7%; 3 CD, -1%; 4 CD, <0.5%. No peaks corresponding to the
presence of CHDz or CHDz + CD3 groups in gentamicin could be
detected. These results showed that methyl groups from methionine
are incorporated substantially intact into all four positions of
gentamicin C,.

(69) B. K. Lee, R. G. Condos, A. Murawski, and G . H. Wagman,J.Antibiot., 28,163-166

(1975).
(70) P. J. L. Daniels, Bloomfield, New Jersey, personal communication.
112 HANS GRISEBACH

The sequence of reactions shown in Scheme 15 may be postulated

for the biosynthesis of L-garosamine (compare Section 1,3). A
nucleotide-bound pentos-4-ulose (probably enzyme-bound) that could
arise from the “nucleoside diphosphate (NucDP) D-glucuronic acid”
(compare Scheme 11) is methylated at C-4. Transamination and N -
methylation then gives NucDP-L-garosamine.

Nucleoside 5 ‘-(D- glucosyl-

uronic acid diphosphate)

OH

OH
NucDP- L-garosamine
SCHEME 15.-Proposed Biosynthetic Sequence for L-Garosamine. [( 1) Transamina-
tion; (2) N-methylation.]

Studies on the biosynthetic relationships of the various gentamicins

were conducted with mutants blocked in regard to formation of
gentamicin. A mutant of Micromonospora purpurea blocked in
synthesis of gentamicin p r o d ~ c e d ‘ the
~ pseudodisaccharide paro-
mamine (see Scheme 16).
The bioconversion of several gentamicins and related compounds
isolated from the fermentation of M . purpurea was investigated to

(71) R. T. Testa and B. C. Tilley,J. Antibiot., 29, 140-146 (1976).

 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 113

determine which, if any, of these may be precursors of the gentamicin

C group. The products formed by these transformations indicated71that
the mutant was able to carry out the enzymic steps leading to the
gentamicin C components from gentamicin A (see Scheme 16).On the
basis of similar studies with a deoxystreptamine-negative mutant of
Micromonospora inyoensis, a biosynthetic scheme for sisomycin (45), a
dehydrogentamicin, was proposed.72 Biotransformation studies also
indicated that transformation of gentamicin X p (46) into antibiotic G-
418 (47) and gentamicin C,requires cobalt ions, and that, therefore, the
6’-methylation is a cobalt-ion-dependent step.73Furthermore, the step
in the biosynthetic pathway from gentamicin A (48) to gentamicin X p
(46), involving 4”-methylation and hydroxyl epimerization, is also
cobalt-ion-de~endent.’~
F2
R’ -6 -R3

/
kH2 0 1
I

OH OH

Sisomycin 46 Gentamicin X, R’ = H , R2 = H ,
R3 = O H , R4 = CH,, R5 = OH
45 47 Antibiotic G-418 R1 = O H , R2 = CH,, R3 = H ,
R4 = CH,, R5 = O H
48 Gentamicin A R’ = H , Rz = H , R3 = O H ,
R4 = O H , R5 = H

The data obtained in these biotransformation studies have to be

interpreted with some degree of caution. For example, the possibility
that gentamicin A might be broken down to paromamine, or even to 2-
deoxystreptamine, before incorporation into later products has not
been excluded. It is known that micro-organisms do degrade their own
secondary metabolites. Scheme 16 would, for instance, imply that L-
garosamine is not formed as a nucleotide-bound sugar as depicted in
Scheme 15, but by modification of the a-D-xylosyl group of gentamicin

(72) R. T. Testa and B. C. Tilley,]. Antibiot., 28,573-579 (1975).

(73) B. C. Tilley, M. Sc. Thesis, Department of Biology, Seton Hall University, South
Orange, N. J., 1976.
114 HANS GRISEBACH
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 115

t
Gentamicin X,
C-methylation, with
inversion of configura-
tion, at C-4"

C-methylation at C-6', with

l
Il-
I 1
inversion of configuration
I
t
Antibiotic G-418

t
amino substitution at C-6'
-4
t
Antibiotic JI-20-A Antibiotic JI-20-B

t- 3', 4 '-dehydroxylation

c
Gentamicin C,,
Gentamicin C,,
epimerization at C-6'

i
Gentamicin C,

Gentamicin C,,
N-methylation at C-6'

-I Gentamicin C,

SCHEME 16.-Postulated Pathway from the Building Units of Gentamicin A2 to

Various Gentamicins, as Concluded from Transformation Studies with Mutants.

A2. However, by analogy with the biosynthesis of other methyl-

branched sugars (see Section II,l), C-methylation of a nucleotide-
bound pentos-4-ulose seems more likely. From the results thus far
obtained, parallel pathways to gentamicins A and C , instead of the
sequential pathway shown in Scheme 16, cannot be excluded.
Labelling experiments and, even more important, studies of enzymic
transformations in cell-free systems are therefore needed in order to
clarify the biosynthetic relationships between the gentamicins.

3. Neomycin
The biosynthesis of neomycin (see Scheme 17) has been studied by
~ ~ incorporation studies with ~-[1-
Rinehart and ~ o w o r k e r s .Early
116 HANS GRISEBACH

14C]glucose,D-[6-14C]glucose, and 2-amino-2-deoxy-D-[l-14C]g~ucose

led to some ambiguities in interpretation of the biosynthetic pathway
due to the lack of a satisfactory degradation scheme for deoxystrep-
tamine (cited in Ref. 53). The incorporation O f D-glucose and 2-amino-
2-deoxy-D-glucose was, therefore, reexamined, employing carbon-13
labels.s3D-[6-13C]Glucoseand 2-amino-2-deoxy-D-[l-13~]glucose were
administered in separate experiments to cultures of Streptomyces
fradiae. Neomycin was isolated 5 days after addition of the precursor.
Neomycin B was separated from small proportions of neomycin C and
neamine, and was then converted into hexa-N-acetylneomycin. The
13C-n.m.r. spectrum of the acetyl derivative showed that [6J3C]glu-
cose had labelled C-6 of the neosamine moieties, C-5 of the D-ribosyl
residue, and C-2 of deoxystreptamine, whereas 2-amino-2-deoxy-[1-
1 3 C l g l ~ ~ olabelled
se C-1 of each of the subunits (see Scheme 17).

Deoxystreptamine

D -Ribose

I Neosamine B (C')
m*
Neomycin B
SCHEME 17.-Labelling pattern of Neomycin B After Feeding of ~-[6-'~C]Ghcoseand
2-Amino-2-deoxy-~-[l-'~C]glucose. (The incorporation of 2-amino-2-deoxy-D-gluco~e
probably occurs by way of D-[l-'3C]glucose; X = H, Y = CHzNHz.)

The results indicated a specific conversion of C-6 of D-glUCOSe into

C-5 of D-ribose, explicable by the hexose monophosphate pathway,
and suggested conversion of C-1 of 2-amino-2-deoxy-D-g~ucoseinto
C-1 of D-ribose, which could be explained by the conversion of 2-
amino-2-deoxy-D-glucose into D-glucose, and the operation, in S.
frudiue, of some version of the glucuronate pathway for the removal
of C-6 of D-glucose.
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 117

The labelling pattern in deoxystreptamine was unexpected, because

it differed from that found for ~ t r e p t i d i n e(see
~ ~ Section II1,l). The
labelling of C-2 (rather than of C-6) of deoxystreptamine by ~ - [ 6 -
'3C]glucose was explained by a new biosynthetic pathway (see Scheme
18, pathway B) involving cyclization of D-glucose to a cyclitol, perhaps
deoxyinosose (my0-[2-'~C]inositolwas not incorporated into neomy-
in^^), followed by amination at the carbonyl carbon atom. From this
stage on, the results could be explained if synthesis of deoxystrepta-
mine involves oxidation and subsequent amination at the &carbon
atom in the direction opposite to that reported for the biosynthesis
of streptamine.

___t

HO
OH OH OH
Streptamine Deoxystreptamine
SCHEME18.-Proposed Biosynthetic Pathway from an Intermediate Inosamine to (A)
Streptamine (compare Scheme 13) and (B) Deoxystreptamine. (For pathway A, the
intermediate inosamine has X = OH, and, for pathway B, X = H or OH.)

The incorporation of 2-amino-%deoxy-D-[l-'3C]glucose into deoxy-

streptamine is probably not direct, but takes place after deamination to
D-[ 1-W]glucose.
Experiments with mutant strains of S. fradiae that only produce
neomycin in the presence of added deoxystreptamine have been
reviewed b y Rinehart and S t r ~ s h a n eand
, ~ ~will not be discussed here.
[l-14C]Deoxystreptaminewas also incorporated into the deoxystrepta-
mine moiety of n e ~ m y c i nOn . ~ ~the other hand, no incorporation of [l-
14C]neosamineC into neomycin was found, probably because this sugar
cannot be activated as the ester of a n ~ c l e o t i d e . ~ ~
In a study on the effect of nitrogen compounds on the production of
neomycin, some phosphoramido-amino sugar antibiotics were isolated
from S . f r ~ d i a eThe
. ~ ~compounds were separated, and characterized as
neomycin B pyrophosphate, neomycin C pyrophosphate, and neomy-
cin C dipyr~phosphate.'~ From the analytical data, it may be concluded

(74) M. K. Majumdar and S. K. Majumdar,]. Antibiot., 32, 174-175 (1969).

(75) M. K. Majumdar and S . K. Majumdar, Biochem.]., 120,271-278 (1970).
118 HANS GRISEBACH

that the pyrophosphate groups are linked to nitrogen as pyrophos-

phoramido derivatives. Together with the finding that phosphatases
that can hydrolyze phosphoramidoneomycins are present in S . fradiae,
and that enzyme activity is related to antibiotic production,'6 it appears
that these phosphate derivatives might be biosynthetic intermediates.

4. Spectinomycin
Spectinomycin (see Scheme 19) has a tricyclic structure, and
contains actinamine. In aqueous solution, the carbonyl group is
present in hydrated form.
The biosynthesis of spectinomycin was examined by feeding
various, labelled precursors to Streptomyces flavopersicus.77 The
percent incorporation of some potential precursors is listed in Table I.

TABLEI
Incorporation of Potential Precursorsa into Spectinomycin

Time
Compound isolated Isotope
administered (h) incorporated (%)

~-[Methyl-'~C]methionine 48 39
~[6-~H]Glucose 48 3.5
rny0-[2-'~C]Inositol 24 47
[2-14C]Actinamine 72 6.6

aThe labelled precursors were added after 69 h, and the antibiotic was isolated after
the time period shown.

[2-14C]Acetate and D-[U-'4C]galactose gave no significant incorpora-

tion, although galactose considerably stimulated the production of
spectinomycin. After recrystallization of the antibiotic substance to
constant specific activity, it was degraded according to Scheme 19.
Label from my0-[2-'~C]inositol was found exclusively in the actina-
mine moiety. Label from ~-[methyl-'~C]methioninewas also found
only in the N-methyl groups of actinamine. ~-[6-~H]Glucose was an
excellent precursor, and its radioactivity was almost equally distrib-
uted between rings A and C of the molecule. Essentially all of the
tritium in ring C was associated with the C-methyl group of
spectinomycin.

(76) M. K. Majumdar and S. K. Majumdar, Biochern. J., 122, 397-404 (1971).

(77) L. A. Mitscher, L. L. Martin, D. R. Feller, J. R. Martin, and A. W. Goldstein, Chern.
Cornrnun., 1541-1542 (1971).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 119

Hu
NHMe NHMe NHMe
I I I

Actinamine
2 CH,NH,
+
CH,CH=CHCCH,OH
I1
I
0
CO,H +
"A

1
Spectinomycin Actinospectinoic acid L-2

CrO,

CH,CO,H
SCHEME19.-Chemical Degradation of Spectinomycin for Localization of the Label.
(Derivatization of degradation products is not shown.)

Radioactive actinamine prepared by cleavage of labelled spectino-

mycin formed biosynthetically from myo-[2-'*C]inositol was incorpora-
ted lnto spectinomycin to a considerably lower extent than myo-
inositol itself. The interpretation of the authors7?that this result argues
against actinamine's being a direct precursor of spectinomycin is not in
agreement with a mutant study in which it was shown that a mutant
unable to produce the antibiotic could do so in the presence of
a ~ t i n a m i n eEvidence
.~~ was also presented that N-methylation is the
last step in the biosynthesis of a ~ t i n a m i n e . ~ ~
As the label from D-[6-3H]glucose could not be located in actinamine
by the procedure of Scheme 19, Rinehart and coworkers undertook
feeding experiments with D-[6-13C]gl~~ose.79 In confirmation of the
results of Mitscher and coworkers,77 enrichment in carbon-13 was
found at (2-6' of actinospectose, and at C-6 of actinamine (see Scheme
20). The label at C-6 of actinamine is similar to the labelling44 of
streptidine with D-[6-'3C]glucose (see Scheme 13).
From the mother liquors of the crystallization of spectinomycin was
isolated a compound that was identified as dihydrospectinomycinsO
(see Scheme 21). The n.m.r. spectrum of the pentaacetyl derivative

(78) L. Slechta and J. H. Coats,AAbstr.Interscience Conf. Antirnicrob. Agents Chemo-

ther. 14th, Sun Francisco, Calif., Sept. 1974, 294.
(79) R. M. Stroshane, M. Taniguchi, K. L. Rinehart, Jr., J. P. Rolls, W. J. Haak, and
B. A. R ~ i f f , j Am.
. Chem. Soc., 98,3025-3027 (1976).
 120 HANS GRISEBACH

o’/
SCHEME20.-Carbon Atoms of Spectinomycin Labelled by ~-[6-~~C]Glucose.

proved that the epimer produced by Streptomyces spectabilis was that

represented by the structure depicted in Scheme 21. In Scheme 21, a
sequence of reactions is shown in which dihydrospectinomycin is the
precursor of spectinomycin.sOHowever, the conversion of dihydro-
spectinomycin into spectinomycin, or vice versa, has not yet been
shown.
5. Validamycin
Validamycins A, C, D, E, and F contain validoxylamine A (49) as a
common moiety in their molecule, but differ from one another in at least
one of the following characteristics: the configuration of the anomeric
center of the D-glucoside, the position of the D-glucosidic linkage, and
the number of D-glucosyl groups. Kameda and coworkerss1studied the
HOH,C, HOH,C

)=-7
HOQ
HO I
NH

OH
Validoxy lamine Validamycin A
49 50

(80) H. Hoeksema and J. C. Knight,./. Antibiot., 28, 240-241 (1975).

(81) Y. Kameda, S. Horii, and T. Yamano,J. Antibiot., 28, 298-306 (1975).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 121

"@OH +

OH J
by way
OH of enediol OH

NHMe
I
I
YHMe

HO OH

Spect inomy c in Dihydrospectinomycin

SCHEME2l.-Postulated Pathway for the Biosynthesis of Spectinomycin by way of
Dihydrospectinomycin.

incorporation of D-[U-14C]glucose into validamycin A (50). High

incorporation (- 10%) and approximately uniform distribution of
radioactivity were observed between the three moieties (validamine,
valienamine, and D-glucose) of the antibiotic. When [14C]validoxyl-
amine was added to a culture of Streptomyces hygroscopicus var.
122 HANS GRISEBACH

limonues, all radioactivity was found in the validoxylamine A moiety,

and virtually no label was found in the D-glucose portion. It was,
therefore, assumed that the D-glucose moiety is added last in the
biosynthesis of validamycin A.
Support for this assumption came from experiments in which the p-
D-glucosylation of validoxylamine A to give validamycin A was shown
by use of certain yeast strains in a medium containing cellobiose as the
p-D-glucosyl donor.s1 In this way, the semi-synthesis of several
validamycins was achieved by using validoxylamine A as the acceptor
and a disaccharide as the glycosyl donor. However, it is more likely
that, in the biosynthesis of validamycins, not a disaccharide but a
nucleoside 5‘-(a-D-glucosyldiphosphate) acts as the D-glucosyl donor.

I v . AMINOSUGARS NOT OCCURRING IN AMINOCYCLITOL

ANTIBIOTICS
1. Desosamine and Mycaminose
D-Desosamine (51) and D-mycaminose (52) occur in macrolide
antibiotics.2 Early studies with tracers showed that the hexose portion
of these sugars can be derived from D-glucose without inversion or
breakdown of the sugar chain.B2*83 The N-methyl groups are derived from
r n e t h i ~ n i n e .No
~ ~further
. ~ ~ work on the biosynthesis of these sugars has
been reported.

I I
OH OH
51 52

V. NUCLEOSIDE
ANTIBIOTICS
Nucleoside antibiotics contain a number of unusual sugar compo-
nents. The chemistry, biosynthesis, and biochemistry of these
antibiotics has been described in a book by Suhadolniks5in which the
literature is covered up to 1969. In Table 11, significant results on the
(82) H. Achenbach and H. Grisebach, Z. Naturforsch. Teil B , 19,561-568 (1964).
(83) J. C. Butte and J. W. Corcoran,Fed. Proc. Fed. Am. SOC. E x p . Biol., 21,89 (1962).
(84) H. Grisebach, H. Achenbach, and W. Hofheinz, Tetrahedron Lett., 234-237 (1961).
(85) R. J. Suhadolnik, “Nucleoside Antibiotics,” Willey-Interscience, New York, 1970.
 TABLEI1
Biosynthesis of Sugar Components from Nucleoside Antibiotics

Antibiotic Sugar moiety Experimental evidence for biosynthesis

Cordycepin (3'- ~-[6-'~C]glucose + incorporation into

deoxyadeno- sugar portion
sine) D-[ l-14C]ribose+ no incorporation
[U-'4C]adenosine + incorporation into
cordycepin without change in the
adeninekibose activity ratiosga
Enzymic mechanism by which reduction
of the 3'-hydroxyl group occurs is not
known
3'-Amino-3'- [U-"C]adenosine + incorporation into
deoxyadeno- 3'-amino-3'-deoxyadenosinewithout
sine cleavage of the ribosyl carbon-
nitrogen bondsgb

Psicofuranine ~-[U-'~C]glucose+ direct precursor for

[6-amino-g-(P- psicose moiety
D-pSiCOfUran0- ~-[1-'~C]glucosewas converted, in a
syl)purine] cell-free extract of S . hygroscopicus in
the presence of ATP and Mg2+, into
psicosesgc

Decoyinine [U-'4Cbsicofuranine was converted into

[6-amino-9-(6- [U-14C]decoyininewithout cleavage of
deoxy-P-& the N-glycos ylic bondsgd
erythro-&hex-
5-enulofurano-
sy1)purinel

Arabinos yl 9-p-D-arabinofuranosyladenine can b e

nucleosides formed by direct epimerization at C-2'
of adenosinesa (see text)

(85a) R. J. Suhadolnik, G. Weinbaum, and H. P. Meloche,J. Am. Chem. Soc., 86,948-

949 (1964).
(85b) B. M. Chassy and R. J. Suhadolnik, Biochim. Biophys. Acta, 182,316-321 (1969).
(85c) R. J. Suhadolnik and T. Sugimori, Fed. Proc. Fed.Am. SOC. E x p . Biol., 25, 525
(1966).
(85d) B. M. Chassy, T. Sugimori, and R. J. Suhadolnik, Biochim. Biophys. Acta, 130,12-
18 (1966).
(86) P. B. Farmer, T. Uematsu, H. P. C. Hogenkamp, and R. J. Suhadolnik,]. Biol. Chem.,
248, 1844-1847 (1973).
124 HANS GRISEBACH

biosynthesis of the sugar moieties of nucleoside antibiotics are

summarized. For each example, a leading reference is given.
Since the appearance of Suhadolnik’s book, very little new
information on the biosynthesis has become available. Farmer and
coworkers have further investigated the biosynthesis of ~-P-D-
arabinofuranosyladenine by Streptomyces antibioticus.86Nitrogen-15
and carbon- 14 from [6-amino-l5N,U-14C]adenosine were incorporated
into the arabinosyladenine with -17% efficiency. The 15N:14Cratios of
arabinosyladenine and of the precursor were the same. To show that
there is no hydrolysis of the N-ribosyl bond of adenosine, the percent
distribution of carbon-14 in the adenine of arabinosyladenine was
compared with that in the adenine of [U-14C]adenosineadded to the
culture filtrates. The percent of carbon-14 in the adenine residue of the
two compounds was the same. On the basis of these results, and others
not discussed here, it was thus proved that the amino nitrogen atom of
adenosine is retained as the amino nitrogen atom of arabinosyladenine,
and that the N-ribosyl bond is not cleaved during conversion into
arabinosyladenine. It was further shown that 2’-deoxy-[U-14C]adeno-
sine is not an intermediate in the conversion of adenine into
arabinosyladenine.
To determine the fate of the 2’- and 3’-hydrogen atoms of adenosine
in the conversion of adenosine into arabinosyladenine, experiments
with [2’-3H, U-14C]adenosine and D-[3-3H, U-14C]ribose were per-
formed. Arabinosyladenine derived from [2’-3H,U-14C]adenosine had
lost all of its tritium, whereas adenosine isolated from the RNA of the
organism had about the same 3H:14Cratio as the added adenosine. In
the D-[3-3H, U-14C]ribose experiment, the proportion of tritium in
arabinosyladenine was only 33% of that present in the adenosine from
RNA.
The results indicated that, during the epimerization of adenosine to
arabinosyladenine, oxidation to a 2’-keto-nucleoside takes place. The
low percentage of tritium in the arabinosyladenine from D-[3-3H, U-
14C]ribosemay be explained by the formation of an enolic intermedi-
ate. The reversibility of the adenosine-arabinosyladenine reaction
could not be shown in z)iz)o, because [U-l4C]arabinosy1adenine was
hydrolyzed by S. antibioticus to adenine.
Six novel, cytosine nucleosides were isolated from the fermentation
broth of S. griseochromogenes, which produces blasticidin S (55, see
Scheme 22). The structural relationship between these nucleosides
and cytosinine, the nucleoside moiety of blasticidin S, was investi-
gated.” One of these compounds was identified as l-(P-D-glucopyrano-
syluronic acid)cytosine (53). The presence of 53 was taken as evidence
(87) H. Seto, K. Furihata, and H. Yonehara,J. Antibiot., 29, 595-596 (1976).
 SUGAR COMPONENTS OF ANTIBIOTIC SUBSTANCES 125

that this compound is the direct precursor of a hypothetical

intermediate (54 in Scheme 22), and that oxidation of the hydroxy-
methyl group of the hexose moiety takes place prior to oxidation of the
hydroxyl group at C-4.

g glucose + Cytosine

-
HO RHN
OH OH
53 54 55

SCHEME22.-Postulated Pathway for Biosynthesis of Blasticidin S (55, R = blastidyl).

Interestingly, the sugar moiety of the postulated intermediate 54 is

identical to that of the intermediate in the biosynthesis of UDP-D-
apiose from UDP-D-glucuronic acid (27, see Scheme 11).
The polyoxins (56) formed by S. cacaoi contain a 5-amino-5-deoxy-D-
allofuranosyluronic acid residue. The biosynthesis of the polyoxins

0
II

O=CHN-CH

HCOH
I
HOCH
I HO OH
CH,OCNH,
8
Polyoxin L
56
126 HANS GRISEBACH

TABLEI11
Distribution of Carbon-14 in Polyoxin Having Various Potential Precursorss8
~~ ______ ______ ~~

Distribiition of Distribution of
Compound added 14C in nucleoside '4C in uronic acid

Pyrimidine Uronic C-1'-5' C-6'

base (%) acid (%) (%I (%)

[U-14C]Uridine 100 0 - -
&[l-'4C]Ribose 13 87 95 5
D[ l-'4C]Allose 0 0 - -
D-[l-'4C]Ghcose 31 69 78 22
D-[6-14C]Glucose 28 72 79 21
D[3,4-14C2]Glucose 11 89 98 2
[1,3-14C2]Glycerol 35 65 80 20

was investigated with 14C-labelled precursors.88 The distribution of

carbon-14 (from the labelled compounds) incorporated into the uracil
and 5-amino-5-deoxy-~-a~~ofuranosyluronic acid residues is shown in
Table 111. The results obtained make it unlikely that the glycosyluronic
acid residue is formed directly from either D-glucose, D-allOSe, or D-
ribose. Although the [14C]uracil residue from [U-14C]uridine was
incorporated into the polyoxins, the ~-['~C]ribosylresidue did not
contribute to the biosynthesis of the uronic acid by the addition of a
one-carbon unit. D-[l-'4C]Glucose and D-[6-'4C]glucose gave the same
distribution of carbon-14 in C-1 and C-6 of the uronic acid. This
distribution of carbon-14 favors the theory of an aldolase pathway,
followed by resynthesis of a hexose, and subsequent oxidation to the
uronic acid. However, further investigations are needed in order to
clarify the biosynthesis of the sugar portion of the polyoxins.

(88) K. Isono and R. J. Suhadolnik, Arch. Biochem. Biophys., 173,141-153 (1976).

 THE LECTINS: CARBOHYDRATE-BINDING PROTEINS OF
PLANTS AND ANIMALS*

BY IRWINJ . GOLDSTEIN E . HAYES

AND COLLEEN

Department of Biological Chemistry. The University of Michigan.

Ann Arbor. Michigan 48109; lmmunobiology Research Center.
University of Wisconsin. Madison. Wisconsin 53706

I . Introduction .......................................................... 128

1. Detection of Lectins ............................................... 133
2. Isolation and Purification of Lectins ................................. 136
3. Carbohydrate-binding Specificity of Lectins ......................... 139
4. Nomenclature of Lectins ........................................... 145
5. Function of Lectins ................................................ 146
11. D-Mannose(D-Glucose)-binding Lectins ................................ 150
1. Concanavalin A of the Jack Bean (Canaualia ensiformis) ............. 150
2. Lens culinaris syn . esculenta (Lentil) ................................ 190
3. Pisum sativum (Pea) .............................................. 196
4. Viciafaba (Fava Bean) ............................................. 201
I11. 2-Acetamido-2-deoxy-~-glucose-binding Lectins ........................ 206
1. Bandeiraea simplicifolia I1 ......................................... 206
2. Cytisus sessilifolius ................................................ 208
3. Solanum tuberosum (Potato) ........................................ 210
4. Triticum uulgaris (Wheat Germ) .................................... 214
5. Ulex europeus I1 (Gorse or Furze Seed) ............................. 224
IV . 2-Acetamido-2-deoxy-~-galactose-binding Lectins ....................... 226
1. Dolichos bijlorus (Horse Gram) ..................................... 226
2. Glycine max (Soybean) ............................................. 231
3. Helix pomatia (Edible Snail) ....................................... 239
4. Phaseolus lunatus syn. limensis (Lima Bean) ........................ 243
5. Sophora japonica (Japanese Pagoda Tree) ........................... 250
V. D-Galactose-binding Lectins ............................................ 254
1. Abrus precatorius (Jequirity Bean) .................................. 254
2. Arachis hypogaea (Peanut) ......................................... 257
3. Bandeiraea simplicifolia I .......................................... 262
4. Maclura pomifera syn. aurantica (Osage Orange) .................... 267
5. Ricinus communis (Castor Bean) .................................... 270
* Writing of this article was supported. in part. by a grant (AM-10171)
from the National
Institutes of Health . The authors are grateful for the assistance of the Editors and of
Paula Kane and Peggy Rogers in the preparation of this Chapter .

127
128 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

VI. L-Fucose-binding Lectins ...................... .......... 277

1. Anguilla anguilla (Eel Serum) ...................................... 277
2. Lotus tetragonolobus (Asparagus Pea) . ........................... 282
3. UZex europeus I (Gorse or Furze Seed) .............................. 289
VII. Other Lectins ....... ............................................. 291
1. Phaseolus vulgaris (Red Kidney-bean) ............. 291
2. Vicia graminea .................................................... 302
3. Miscellaneous Lectins ........ .................................. 304
VIII. Cell-surface, Lectin-reactive Glyc teins ............... ...... 317
1. Erythrocytes and Platelets .......................................... 3 18
2. Lymphocytes ............. .................................... 325
3. Neuronal Cells ............................................ 326
4. Tumor Cells ....................................................... 327
IX. Tables . . . ................................ ............... 334

I. INTRODUCTION
Stillmark’s discovery,’ in 1888, that castor-bean extracts agglutinate
animal erythrocytes, marked the initiation of studies on plant aggluti-
nins. The hemagglutinating activity of seed extracts was traced to the
presence of discrete proteins (glycoproteins), variously termed
agglutinins, hemagglutinins, phytohemagglutinins ,and, most recently,
lectins. Boyd and Shap1eigh2a3coined the term lectin (Latin, legere, to
select or pick out), based on their observation that some seed extracts
could distinguish among human b l o o d - g r ~ u p s . ~Specifically,
-~ Boyd
and Reguera4 discovered that lima-bean (Phaseolus lunatus) extracts
selectively agglutinate type A erythrocyte^.^,^ Almost simultaneously,
Renkoned made the same observation, reporting several blood-
group-specific seed-extracts among 57 species belonging to 28 different
genera. These discoveries stimulated renewed interest in the isolation
and characterization of lectins, and initiated a seemingly endless
number of applications of these fascinating plant-proteins to biological
studies.
Lectins have played an important role in the development of im-
munology. In 1891,Ehrlich7 showed that specific immunity to the toxic

(1) H. Stillmark, “Uber Rizin, ein gifiiges Ferment aus dem Samen von Ricinus com-
munis L. und einigen anderen Euphorbiaceen,” Inaug. Diss., Dorpat. 1888.
(2) W. C. Boyd and E. Shapleigh,]. Immunol., 73,226-231 (1954).
(3) W. C. Boyd and E. Shapleigh, Science, 119, 419 (1954).
(4) W. C. Boyd and R. M. Reguera,]. Immunol., 62,333-339 (1949).
(5) W. C. Boyd, “Introduction to Immunochemical Specificity,” Wiley-Interscience,
New York, 1962, pp. 1-158.
(6) K. 0. Renkonen, Ann. Med. Exp. Biol. Fenn., 26,66-72 (1948).
(7) P. Ehrlich, Dtsch. Med. Wochenschr., 17, 976-979, 1218-1219 (1891).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 129

lectins ricin (Ricinus communis) and abrin (Abrus precatorius) could be

achieved by repeated injection of small amounts of these antigens into
white mice. The first quantitative determination of an antibody in vitro
was performed by Ehrlich8 in 1897; immune serum inhibited ricin-red
blood-cell agglutination. Furthermore, abrin was employed by Land-
steinergin an early demonstration of the reversibility of the antibody-
antigen reaction; abrin was displaced from agglutinated erythrocytes
by incubating at 50°, and then centrifuging the cells.
Currently, lectins find application in serological laboratories for typ-
ing blood and determining secretor s t a t ~ s , ’ ~ separating
-*~~ leucocytes
from erythrocytes,1sand agglutinating cells from blood in the prepara-
tion of plasma.16 Moreover, lectins serve as reagents for the detection,
isolation, and characterization of carbohydrate-containing mac-
romolecules, including blood-group Enormous interest
centers at present on probing the nature and distribution of
membrane-bound, carbohydrate-containing structures by using lectins
of defined specificity. The distribution and mobility of cell-
16325-27

surface glycoproteins on normal and malignant cells has been inves-

tigated25by employing lectins labelled with ferritin, fluorescein, or

(8) P. Ehrlich, Fortschr. Med., 15,41-43 (1897).

(9) K. Landsteiner, Wien. Klin. Wochenschr., 14, 713-714 (1901).
(10) G. W. G. Bird, Br. Med. Bull., 15, 165-168 (1959).
(11) W. C. Boyd, Vox Sang., 8 , l - 3 2 (1963).
(12) W. C. Boyd and E. Shapleigh,J. Lab. Clin. Med., 44,235-237 (1954).
(13) W. C. Boyd and E. Shapleigh, Blood, 9, 1195-1198 (1954).
(14) R. R. Race and R. Sanger, “Blood Groups in Man,” Blackwell, Oxford and London,
6th Edition, 1975, pp. 1-659.
(14a) 0. Prokop and G. Uhlenbruck, “Human Blood and Serum Groups” (Translated by
J. L. Raven), Maclaren, London, 2nd Edition, 1965, pp. 1-891.
(15) J. G. Li and E. E. Osgood, Blood, 4,670-675 (1949).
(16) M. Dorset and R. R. Henley,J. Agric. Res. (Washington, D.C.), 6,333-339 (1916).
(17) I. J. Goldstein, Methods Carbohydr. Chem., 6, 106-119 (1972).
(18) H. Lis and N. Sharon, Annu. Rev. Biochem., 42,541-574 (1973).
(19) T. Osawa, Biochim. Biophys. Acta, 115, 507-510 (1966).
(20) 0. Makela, Nature, 184, 111-113 (1959).
(21) W. M. Watkins and W. T. J. Morgan, Nature, 169, 825-826 (1952).
(22) W. T. J. Morgan and W. M. Watkins, Br. J . E x p . Pathol., 34,94-103 (1953).
(23) W. T. J. Morgan and W. M. Watkins, Nature, 177,521-522 (1956).
(24) W. M. Watkins, in “Glycoproteins,” A. Gottschalk, ed., Elsevier, Amsterdam, 2nd
Edition, 1972, Part B, pp. 830-891.
(%a) T. Kristiansen, Methods Enzymol., 34 (Part B), 331-341 (1974).
(25) G. L. Nicolson, Znt. Rev. Cytol., 39, 89-190 (1974).
(26) N. Sharon and H. Lis, Methods Membrane B i d . , 3,147-200 (1975).
(27) N. Sharon, in “Extracellular Matrix Influences on Gene Expression,” H. C. Slavkin
and R. C. Greulich, eds., Academic Press, New York, 1975, pp. 479-487.
130 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

isotopes. Certain lectins distinguish normal from malignant cells18,25-31;

the possible use of lectins in cancer chemotherapy has been pro-

In their interaction with saccharides, lectins serve as models for

carbohydrate-specific antibodies, with the important advantage that it
is possible to purify lectins in gram quantities. Furthermore, lectin
combining-sites appear to be homogeneous and noninteracting, in con-
trast to those of immune antibodies.
Nowell’s that extracts of red kidney-bean stimulate lym-
phocyte division in vitro led to enormous interest in mitogenic lectins
as tools for studying the biochemical events accompanying cellular
growth, differentiation, and d i ~ i s i o n . ~ ~ - ~ ~ ~
Investigations using lectins to probe membrane-bound hormone-
receptors have been spurred by the observation that certain lectins (for
example, wheat-germ agglutinin, concanavalin A, and lima-bean and
lentil lectins) are as effective as insulin in enhancing transport of
D-glUCOSe and in inhibiting epinephrine-stimulated lipolysis in iso-
lated adipocytes. 38-40 Finally, lectins are being studied with respect to
toxic properties that may affect the nutritional value of b e a n ~ . ~ l - ~ ~
(28) J. C. Aub, C. Tieslau, and A. Lankester, Proc. Natl. Acad. Sci. U.S.A.,50,613-619
(1963).
(29) J. C. Auh, B. H. Sanford, and M. N. Cote, Proc. Natl. Acad. Sci. U.S.A.,54,396-399
(1965).
(30) M. M. Burger and A. R. Goldberg, Proc. Natl. Acad. Sci. U.S.A.,57,359-366 (1967).
(31) M. Inhar and L. Sachs, Proc. Natl. Acad. Sci. U.S.A., 63, 1418-1425 (1969).
(32) J.-Y.Lin,K.-Y.Tsemg, C.-C. Chen, L.-T. Lin, andT.-C. Tung,Nature, 227,292-293
(1970).
(33) H. Lin, W. R. Bruce, and M. J. Walcroft, Cancer Chemother. Rep., 59, 319-326
(1975).
(34) P. C. Nowell, Cancer Res., 20,462-466 (1960).
(33) C. K. Naspitz and M. Richter, Prog. Allergy, 12, 1-85 (1968).
(36) N. R. Ling, “Lymphocyte Stimulation,” North-Holland, Amsterdam, 1968, pp.
1-290.
(37) G. Moller, Transplant. Rev., 11, 1-216 (1972).
(37a) N. Sharon, in “Mitogens in Immunobiology,” J. J. Oppenheim and D. L.
Rosenstreich, eds., Academic Press, New York, 1976, pp. 31-41.
(37b) H. Lis and N. Sharon, in “The Antigens,” M. Sela, ed., Academic Press, New York,
1977, Vol. 4, pp. 429-529.
(38) P. Cuatrecasas and G. P. E. Tell, Proc. Natl. Acad. Sci. U.S.A.,70,485-489 (1973).
(39) M. P. Czech and W. S. Lynn, Biochim. Biophys. Acta, 297, 368-377 (1973).
(40) H. M. Katzen and D. D. Soderman, Biochemistry, 14,2293-2298 (1975).
(41) H. J. H. de Muelenaere, Nature, 206,827-828 (1965).
(42) I. E. Liener,]. Agric. Food Chem., 22, 17-22 (1974).
(43) A. T. Andrews and D. J. Jayne-Williams, Br. I . Nutr., 32, 181-188 (1974).
(44) W. G. J d b , in “Toxic Constituents of Plant Foodstuff,” I. E. Liener, ed., Academic
Press, New York, 1969, pp. 69-101.
(45) I. E. Liener, Annu. Reo. Plant Physiol., 27,291-319 (1976).
(46) R. H. Turner and I. E. Liener,]. Agric. Food Chem., 23,484-487 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 131

Definition of the term lectin is a difficult problem in limiting the

scope of this article. By employing the term lectin, Boyd and Shap-
leigh2-3,swished to call attention to the serological specificity of certain
plant-seed agglutinins. After it became apparent that the activity of
many hemagglutinins is inhibited by simple sugars,21*22 these proteins
were commonly referred to as plant-seed, carbohydrate-binding pro-
teins. However, the discovery of hemagglutinins in such diverse
sources as b a ~ t e r i a , ~ ~ * ~ * l i ~ h e n s ,fish
~ ~ r, ~ ~
e:~-~~
and even mammal^^'-^^^ has resulted in a broadening of the
eels,21*64-66
term “lectin” to include carbohydrate-binding proteins (glycoproteins)
without regard to their origin. For the purpose of this article,
carbohydrate-binding proteins of known enzymic function,
carbohydrate-transport proteins, and carbohydrate-specific antibodies
are not considered to be lectins. Moreover, we have included only
those lectins that have been purified to homogeneity, and studied with
regard to their biophysical, biochemical, and carbohydrate-binding
specificity. This arbitrary decision regrettably necessitated the exclu-
sion of an enormous number of interesting studies conducted with
crude seed-extracts.
(47) E. Neter, Bacteriol. Reu., 20, 16G-188 (1956).
(48) N. Gilboa-Garber, Biochim. Biophys. Acta, 273, 165-173 (1972).
(49) W. W. Ford,]. Pharmacol. E x p . Ther., 2,285-318 (1910-11).
(50) M. Coulet, J. Mustier, and J. Guillot, Rev. Mycol., 35, 71-89 (1970).
(51) H. J. Sage and S. L. Connett,]. Biol. Chem., 244, 4713-4719 (1969).
(52) Y. Fujita, K. Oishi, K. Suzuki, and K. Imahori, Biochemistry, 14,4465-4470 (1975).
(53) E. Estola and K. 0. Vartia, Ann. Med. E r p . Biol. Fenn., 33,392-395 (1955).
(54) M. L. Howe and J. T. Barrett, Biochim. Biophys. Acta, 215,97-104 (1970).
(55) 0. Prokop, D. Schlesinger, and G. Geserick, Z. Immunitaetsforsch. Allerg. Klin.
Immunol., 132,491-494 (1967).
(56) G. Uhlenbruck and 0. Prokop, Vox Sang., 12,465-466 (1967).
(57) D. J. Anstee, P. D. J. Holt, and G. I. Pardoe, Vox Sang., 25, 347-360 (1973).
(58) 0. Prokop, A. Rackwitz, and D. Schlesinger,]. Forensic Med., 12, 108-110 (1965).
(59) A. Rackwitz, D. Schlesinger, and 0. Prokop, Acta Biol. Med. Ger., 15, 187-190
(1965).
(60) 0. Prokop, D. Schlesinger, and A. Rackwitz, Z. Immun. Allergieforsch., 129,
402-412 (1965).
(61) G. Uhlenbruck and 0. Prokop, Vox Sang., 11,519-520 (1966).
(62) W. C. Boyd and R. Brown, Nature, 208, 593-594 (1965).
(63) S. Hammarstrom and E. A. Kabat, Biochemistry, 8,2696-2705 (1969).
(64) B. Jonsson, Acta Pathol. Microbiol. Scand., Suppl., 54,456-464 (1944).
(65) R. Grubb, Acta Pathol. Microbiol. Scand., Suppl., 84,5-72 (1949).
(66) P. R. Desai and G. F. Springer, Methods Enzymol., 28, Part B, 383-388 (1972).
(67) R. J, Stockert, A. G. Morell, and I. H. Scheinberg, Science, 186,365-366 (1974).
(67a) R. L. Hudgin, W. E. Pricer, Jr., G. Ashwell, R. J. Stockert, and A. G. Morell,]. Biol.
Chem., 249,5536-5543 (1974).
(6%) T. Kawasaki and G. Ashwell,]. Biol. Chem., 251, 1296-1302 (1976).
(67c) A. de Waard, S. Hickman, and S. Kornfeld,]. Biol. Chem., 251,7581-7587 (1976).
(67d) H. Den and D. A. Malinzak,]. Biol. Chem., 252,5444-5448 (1977).
132 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

The reader is referred to a number of r e ~ i e w s ~ . ~ ~(espe-

. ~ ~ * ~ * ~ ~
cially those published18*37b,45,72,75,76a,b
since 1970) and treatise~,~~-~Oas
well as to valuable compendia dealing with the screening and sero-
logical properties of thousands of plant-seed extracts.2,5*6,68*77-91a
Lec-
tins deriving from micro-organisms and invertebrates were discussed
in a volume.76The cell-binding and biological properties of lectins
have also been reviewed.18*25,26*g2.92a
Full proceedings of two confer-
ences on plant and invertebrate agglutinins%and ong4concanavalin A
have been published.

(68) G. W. G. Bird, Acta Chir. Belg., 53, 33-40 (1954).

(69) W. C. Boyd,J. Immunol., 85,221-229 (1960).
(70) I. E. Liener, Econ. Bot., 18,27-33 (1964).
(71) J. M. Dechary, Vox Sang., 15,401-409 (1968).
(72) N. Sharon and H. Lis, Science, 177,949-959 (1972).
(73) W. C. Boyd, Ann. N.Y. Acad. Sci., 169, 168-190 (1970).
(74) G. C. Toms and A. Western, in “Chemotaxonomy of the Leguminosae,” J. B.
Harbome, D. Boulter, and B. L. Turner, eds., Academic Press, London and New
York, 1971, pp. 367-462.
(75) J. A. Callow, Curr. Adu. Plant Sci., 18, 181-193 (1975).
(76) E. R. Gold and P. Balding, “Receptor-Specific Proteins, Plant and Animal Lectins,”
American Elsevier, New York, 1975, pp. 1-440.
(76a) E. Kottgen, Klin. Wochenschr., 55,359-373 (1977).
(76b) N. Sharon, Sci. Am., 236, 108-119 (1977).
(77) M. Kriipe, “Blutgruppenspezifische Pflanzliche Eiweisskorper (Phytagglutinine),”
Ferdinand Enke Verlag, Stuttgart, 1956, pp. 1-131.
(78) 0. Makela, Ann. Med. E x p . Biol. Fenn. Suppl. 1 1 , 35, 1-156 (1957).
(79) 0. Makela, “Studies on Hemagglutinins of Leguminosae Seeds,” Weilin and Goos,
Helsinki, 1957, pp. 1-133.
(80) J. Tobiska, “Die Phythaemagglutinine,” Akadeniie Verlag, Berlin, 1964,pp. 1-302.
(81) P. Cazal and M. Lalaurie, Acta Haematol., 8, 73-80 (1952).
(82) G. W. G. Bird,]. lmmunol., 69, 319-320 (1952).
(83) G. W. G. Bird, Br. J . E x p . Pathol., 35,252-254 (1954).
(84) G. W. G. Bird, Army Med. C o r p s J . , 11, 17-25 (1955).
(85) W. C. Boyd, E. Waszczenko-Zacharczenko, and S. M. Goldwasser, Transfusion
(Philadelphia), 1, 374-382 (1961).
(86) L. Brilliantine and N. K. Allen,]. Zmmunol., 86, 575-577 (1961).
(87) L. Brilliantine, L. H. Aranda, D. M. Foster, and N. K. Allen,]. Zmmunol., 92, 555-
558 (1964).
(88) T. Martin and G. Bomchil, Vox Sung., 11,54-58 (1966).
(89) A. A. Hossaini, Vox Sang., 15,410-417 (1968).
(90) N. K. Allen and L. Brilliantine,]. Immunol., 102, 1295-1299 (1969).
(91) G. W. G. Bird and J. Wingham, Vox Sang., 24,48-57 (1973).
(91a) J. Tobiska, Z . Immunitaetsforsch. E x p . Ther., 117, 156-163 (1959).
(92) K. D. Noonan, in “Virus Transformed Cell Membranes,” C. Nicolau, ed., Academic
Press, London, 1976.
(92a) A. M. C. Rapin and M. M. Burger, Ado. Cancer Res., 20,l-91 (1974).
(93) E. Cohen, Ann. N.Y. Acad. Sci., 234, 1-412 (1974).
(94) T. K. Chowdhury and A. K. Weiss, Ado. E x p . Med. Biol., 55, 1-360 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 133

In the ensuing discussion, we have classified lectins according to

their carbohydrate-binding specificity. Thus, D-mannose(D-glucose)-
binding lectins (concanavalin A and the agglutinins from the pea, lentil,
and fava bean) are dealt with first, followed by 2-acetamido-2-deoxy-
D-glucose-binding lectins. Although this grouping is arbitrary, it brings
a semblance of order to an extremely active and burgeoning field of
investigation. Moreover, it readily permits a comparison of properties
and data on carbohydrate-binding specificity within a group of aggluti-
nins. As already stated, emphasis will be placed on chemical and
physicochemical aspects. The literature on biological properties and
activity of lectins is now so enormous as to require special treatment.
The interested reader is referred to the many reviews cited that deal
with various cellular and molecular-biological aspects of lectins.
Finally, we have summarized (1) the physicochemical properties of
purified lectins, (2) their carbohydrate-binding and blood-group spec-
ificity, and (3) equilibrium-dialysis, binding data in the three Tables at
the end of this Chapter (Section IX). These Tables should provide the
reader with a synopsis of the most important and distinguishing charac-
teristics of lectins. We also present several glycopeptide structures
(Figs. 14-16) showing the carbohydrate-binding loci with which vari-
ous lectins interact.

1. Detection of Lectins
Inasmuch as the first biological activity to be recognized for lectins
was their capacity to agglutinate erythrocytes, most investigators have
detected lectins by hemagglutination, using a panel of freshly drawn,
animal or human erythrocytes, or both.78,9s-g7 Red blood-cells digested
with papain, trypsin, neuraminidase, or other enzymes have also been
employed; such treatment often renders cells more sensitive to
Other types of animal cells have also been
~ ~ e Blood-group-specific
d . ~ ~ ~ ~ lectins ~ , are~ identified
~ with the aid of a
panel of typed erythrocytes.
Hemagglutination tests are generally conducted at room temperature
by adding a drop of a 2 to 3% erythrocyte suspension to tubes or wells
containing serially diluted lectin. After incubation for 1.5-2 h, the

(95) E . A. Kabat and M. M. Mayer, “Experimental Immunochemistry,” Charles C.

Thomas, Springfield, Ill., 2nd Edition, 1961, pp. 1-905.
(96) M. M. Burger, Methods Enzymol., 32, 615-621 (1974).
(97) H. Lis and N. Sharon, Methods Enzymol., 28, Part B, 360-365 (1972).
(98) G. I. Pardoe and G. Uhlenbruck,]. Med. Lab. Technol., 27,249-263 (1970).
(99) J. A. Gordon, N. Sharon, and H. Lis, Biochim. Biophys.Acta, 264,387-391 (1972).
(100) 0 . Prokop, G. Uhlenbruck, and W. Kohler, Vox Sang., 14,321-333 (1968).
134 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

agglutination is observed by the naked eye or by microscope. Activity is

expressed as titer, the reciprocal of the maximal dilution of lectin that
gives visible aggregation. An error o f f 1 tube is commonly accepted,
that is, a titer of 64 could range from 32 to 128.
Occasionally, a lectin will not agglutinate red blood-cells unless they
are suspended in a medium more viscous than physiological saline, for
example, poly(vinylpyrrolidinone), serum albumin, and serum.78Such
agglutinins have been termed "incomplete" lectins, in that they re-
quire the presence ofaccessory factors in order to cause a g g l ~ t i n a t i o n . ~ ~
A spectrophotometric assay was introduced by Liener.'O' It is based
on the observation that rabbit erythrocytes sediment at a rate propor-
tional to the concentration of the hemagglutinin. Hemagglutinating
activity is calculated from absorbance measurements at 620 nm of the
unsedimented cell-suspension after a given time.97*'0'
Hemagglutinins may be of several types: (a) nonspecific lectins that
agglutinate cells without regard to their origin (either species, or blood
type), ( b ) lectins that preferentially agglutinate the cells of one or
several kinds of animals, or (c) blood-group-specific lectins.
As additional knowledge of the properties of lectins became avail-
able, more-refined screening-procedures for detecting them were de-
veloped. Naturally occurring g l y ~ o p r o t e i n s , ~ ~ ,synthetic
~~~*~~-~~~
carbohydrate-protein (and carbohydrate-phloroglucinol)
conjugates,112-1'6and polysa~charides'~~~~~~-'~~ have been employed in

(101) I. E. Liener, Arch. Biochem. Biophys., 54, 223-231 (1955).

(102) J. B. Sumner and S. F. Howell,]. Bacteriol., 32, 227-237 (1936).
(103) W. C. Boyd, E. Shapleigh, and M. McMaster,Arch. Biochem. Biophys., 55,226-
234 (1955).
(104) G. W. G. Bird, Vox Sang., 4,307-313 (1959).
(105) J. H. Morse, Immunology, 14,713-724 (1968).
(106) H. Harris and E. B. Robson, Vox Sang., 8,348-355 (1963).
(107) S. Nakamura, K. Tanaka, and S. Murakawa, Nature, 188, 144-145 (1960).
(108) M. E. Etzler and E. A. Kabat, Biochemistry, 9,869-877 (1970).
(109) N. M. Young and M. A. Leon, Biochim. Biophys. Acta, 365,418-424 (1974).
(110) S. Yachnin,]. Immunol., 108,845-847 (1972).
(111) S. Yachnin,]. Exp. Med., 141,242-256 (1975).
(112) I. J. Goldstein and R. N. Iyer, Biochim. Biophys. Acta, 121, 197-200 (1966).
(113) R. N. Iyer and I. J. Goldstein, Immunochemistry, 10,313-322 (1973).
(114) W. T. Shier, Proc. Natl. Acad. Sci. U.S.A.,68,2078-2082 (1971).
(115) J.-P. Privat, F. Delmotte, and M . Monsigny, F E B S Lett., 46, 224-228 (1974).
(116) A. E. Clark, R. B. Knox, and M. A. Jermyn,]. Cell Sci., 19, 157-167 (1975).
(117) J. A. Cifonelli, R. Montgomery, and F. Smith,]. Am. Chem. Soc., 78,2485-2488
(1956).
(118) G. F. Springer,]. Zmmunol., 76,399-407 (1956).
(119) G. W. G. Bird, Nature, 187,415-416 (1960).
(120) I. J. Goldstein and L. L. So, Arch. Biochem. Biophys., 111, 407-414 (1965).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 135

detecting lectins and in studying their specificity. Formation of a pre-

cipitate between a lectin and carbohydrate-containing mac-
romolecules, either in liquid (capillary tubes) or semisolid (agar gel)
media, is indicative of precipitating lectin activity. Moreover, it can
provide information regarding lectin specificity and anomeric prefer-
ence, as well as on the constituent sugars and glycosidic linkages ofthe
macromolecule.
Sugar-protein conjugates in which p-diazophenyl glycosides were
coupled to a protein or to phloroglucinol,116provided
model substrates for detecting lectins and studying their specificity. In
fact, a second phytohemagglutinin, a 2-acetamido-2-deoxy-~-
glucopyranosyl-binding protein, was first detected in Bundeiruea
simplicifolia seeds by using a p-azophenyl 2-acetamido-2-deoxy-P-~-
glucopyranoside-bovine serum albumin conjugate.12s
Precipitate formation between multivalent, model substrates and
lectins must be cautiously interpreted. Nonspecific interactions must
be distinguished by sugar inhibition-tests; for example, addition of
maltose to the precipitate formed between a maltose-protein conjugate
and seed extract should dissolve the precipitate, whereas lactose should
be ineffective.lZ0Furthermore, the carbohydrate-protein linkage of
the synthetic substrate may have a marked influence on the precipita-
tion reaction, and may even give rise to artifactual precipitates. It is,
therefore, preferable to employ a di- or tri-saccharide conjugate, in
which the nonreducing, terminal glycosyl group is glycosidically
bound to a second sugar residue rather than to an aromatic aglycon.
A new class of synthetic sugar-protein conjugates has been intro-
duced to circumvent this problem.1z6An aldonic acid was joined,
through peptide bonds, to a protein, and the product was used for
screening and specificity studies.
HofejSi and Kocourek devised a new procedure for the detection of
lectins which they termed l Z 7 “affinity electrophoresis.” This technique
(121) I. J. Coldstein, C. E . Hollerman, and J. M. Merrick, Biochim. Biophys. Actu, 97,
68-76 (1965).
(122) M. Paulovi, M. Tichi, G . Entlicher, J. KoStii, and J. Kocourek,FEBS Lett., 9,345-
347 (1970).
(123) N. M . Young, M. A. Leon, T. Takahashi, I. K. Howard, and H. J. Sage,]. Biol.
Chem., 246, 1596-1601 (1971).
(124) J. P. Van Wauwe, F. C. Loontiens, and C. K. D e Bruyne, Biochim. Biophys. Actu,
313,99-105 (1973).
(125) P. N. Shankar Iyer, K. D. Wilkinson, and I. J. Coldstein,Arch. Biochem. Biophys.,
177,330-333 (1976).
(126) J. Lonngren, I. J. Coldstein, and J. E. Niederhuber,Arch. Biochem. Biophys., 175,
661-669 (1976).
(127) V. Hoiejs’i and J. Kocourek, Biochim. Biophys. Actu, 336, 338-343 (1974).
136 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

combines the principles of affinity chromatography and elec-

trophoresis; proteins are subjected to electrophoresis on a matrix
formed by copolymerization of alkenyl glycosides with acrylamide.
Proteins having combining sites complementary to the ligand are re-
tarded, whereas other proteins undergo a normal separation. By de-
termining the electrophoretic mobility of a lectin in the presence of
several concentrations of its specific, sugar ligand, Hoiejii and col-
league~ extended
~ ~ ~ ~ their procedure to enable a determination of
the dissociation constant of the lectin-sugar complex to be made. Af-
finity electrophoresis has also been used to study sugar-binding het-
erogeneity of lectins and their chemically modified derivative^.'^'^

2. Isolation and Purification of Lectins

It is only within the past decade that lectins have been purified to a
homogeneity demonstrable by physicochemical and immunological
criteria. Although Sumner and Howell crystallized the jack-bean lectin
(concanavalin A or con A) over forty years ago, they were unable to free
it completely from carbohydrate contaminants.lo2
Isolation of lectins generally begins with a saline (or buffer) extrac-
tion of the finely ground, seed meal. Pre-extraction with organic
solvents (for example, methanol or diethyl ether) is often employed to
remove lipid or other interfering s ~ b s t a n c e s . ' ~ ~Ammonium
~ ' ~ ~ - ' ~ ~ sulfate
fractionation, centrifugation, and dissolution of the precipitate yields a
supernatant liquor containing the lectin(s). Plant agglutinins may be
isolated from saline extracts by conventional, protein-purification
techniques, affinity chromatography, or a combination thereof. Virtu-
ally all contemporary, lectin-purification schemes employ affinity
chromatography that exploits the specific, sugar-binding capacity of the
, ' ~ ~ stated, a carbohydrate ligand with which the lectin
l e ~ t i n . ' ~Simply
interacts is insolubilized, the lectin is adsorbed as the extract is perco-
lated slowly over the adsorbent, and displacement of bound lectin is
accomplished by elution, either with a sugar that competes for lectin
sites with the specific adsorbent, or by altering the nature of the eluant
(by lowering the pH, increasing the ionic strength, or adding denatur-
(127a) V. HoiejXi, M. Ticha, and J. Kocourek, Biochim. Biophys. Acta, 499, 290-300
( 1977).
(127b) V. Hoiejs'i, M. Ticha, and J. Kocourek, Biochim. Biophys. Acta, 499,301-308
(1977).
(128) A. K. Allen, A. Neuberger, and N. Sharon, Biochem. J., 131, 155-162 (1973).
(129) Y. Nagata and M. M. Burger,J. Biol. Chem., 249,3116-3122 (1974).
(130) J. H. Shaper, R. Barker, and R. L. Hil1,Anal. Biochem., 53, 564-570 (1973).
(131) C. E. Hayes and 1. J. Goldstein,J. Biol. Chem., 249, 1904-1914 (1974).
(132) H. Lis, R. Lotan, and N. Sharon, Ann. N.Y. Acad. Sci., 234,232-238 (1974).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 137

ants). Such insoluble, naturally occurring, or chemically modified sub-

stances as ~ h i t i n , ' ~S~e Jp ~h ~a d e ~ , ' ~ and
~-'~agarose
~ or S e p h a r ~ s e ' ~ ~ - ' ~ ~
have also been employed as affinity matrices.
Applications of affinity chromatography to lectin purification have
been s u m m a r i ~ e d . ~ *AJ ~few ~ examples of the diverse approaches
utilized for affinity-column synthesis and used in the isolation of lectins
follow; these are discussed in detail in subsequent Sections. HoPejSi
and K o c ~ u r e k copolymerized
'~~ a series of alkenyl glycosides with
acrylamide, whereas Matsumoto and O s a ~ aincorporated '~~ a variety of
sugar residues into starch; both materials have been used149as affinity
matrices in lectin isolation. Reaction of epoxy-activated Sepharose 6B
with 2-acetamido-2-deoxy-~-glucoseand -D-galactose afforded affinity
columns used for the purification of wheat-germ agglutinin and soy-
bean lectin, r e ~ p e c t i v e l y . Affinity
' ~ ~ ~ chromatography on aminoethyl
poly(acrylamide) gels containing reductively aminated disaccharide
residues (lactose, melibiose, maltose, andN,N'-diacetylchitobiose) was
used by Baues and Gray149b to isolate lectins from the B . simplicifolia
seeds, castor beans, jack beans, lentils, and wheat germ. Several pur-
ification schemes employed Sephadex affinity-chromatography; a - ~ -
glucopyranosyl- and a-D-mannopyranosyl-binding lectins from the
R. Bloch and M. M. Burger, Biochem. Biophys. Res. Commun., 58,13-19 (1974).
B. B. L. Agrawal and I. J. Goldstein, Biochem. ]., 96,23C-25C (1965).
B. B. L. Agrawal and I. J. Goldstein, Biochim. Biophys. Acta, 147,262-271 (1967).
M. 0. J. Olson and I. E. Liener, Biochemistry, 6, 105-111 (1967).
K. Aspberg, H. Holmen, and J. Porath, Biochim. Biophys. Acta, 160, 116-117
(1968).
I. K. Howard and H. J. Sage, Biochemistry, 8,2436-2441 (1969).
M. Tomita, T. Osawa, Y. Sakurai, and T. Ukita, Znt.]. Cancer, 6,283-289 (1970).
J. L. Wang, J. W. Becker, G. N. Reeke, Jr., and G. M. Edelman,]. Mol. B i d , 88,
259-262 (1974).
G. Entlicher, J. V. KodtiF, and J. Kocourek, Biochim. Biophys. Acta, 221,272-281
(1970).
M. Tichi, G. Entlicher, J. V. KoBtii, and J. Kocourek,Biochim. Biophys. Actu, 221,
282-289 (1970).
S. Toyoshima, T. Osawa, and A. Tonomura, Biochim. Biophys. Acta, 221,514-521
(1970).
M. Tomita, T. Kurokawa, K. Onozaki, N. Ichiki, T. Osawa, and T. Ukita, Experien-
tia, 28,84-85 (1972).
B. Ersson, K. Aspberg, and J. Porath, Biochim. Biophys. Acta, 310,446-452 (1973).
G. L. Nicolson, J. Blaustein, and M. E. Etzler, Biochemistry, 13,196-204 (1974).
S . Olsnes, E. Saltvedt, and A. Pihl,]. BioZ. Chem., 249,803-810 (1974).
V. Ho'rejSi and J. Kocourek, Biochim. Biophys. Acta, 297, 346-351 (1973).
I. Matsumoto and T. Osawa, Biochem. Biophys. Res. Commun., 46, 1810-1815
(1972).
(149a) P. Vretblad, Biochim. Biophys. Acta, 434, 169-176 (1976).
(149b) R. J. Baues and G. R. Gray,]. B i d . Chem., 252, 57-62 (1977).
138 IRWIN J . GOLDSTEIN AND COLLEEN E. HAYES

seeds of Canavalia ensiformis (con A),134-136 Vicia c r a ~ c a , ~Vicia

~'
faba,139*140 Pisum sativus (pea),14' and Lens culinaris (lenti1)'38,142~L43
were isolated in this way. Likewise, such P-D-galaCtOpyranOSyl-
binding proteins as ricin (Ricinus c ~ m m u n i s ) ' ~and ~ . ' abrin
~ ~ (Abrus
p r e c a t o r i ~ s ) have
' ~ ~ been purified on agarose. LecMns from soybean,
Wistaria jloribunda, Bauhinia purpurea alba, and Sophora japonica
could be isolated by affinitychromatography on acid-treated Sepharose
6B, but not on untreated Sepharose.LsoaCross-linked guaran was used
for isolating the lectins from Helix pomatia, Glycine max, Ricinus
communis, and Echinocystis lobata (wild cucumber), and the a-D-
galactopyranosyl-binding protein from Bandeiraea simplicifolia; how-
ever, the lectins from Dolichos bijlorus, Phaseolus lunatus, and
Sophora japonica failed to bind to this ab~orbent.'~"A further pro-
cedure involves entrapment of polysaccharide in poly(acry1amide)
gels. The lectins from the peanut (Arachis hypogaea) and Bandeiraea
simplicifolia seeds were isolated by using guaran-entrapped
bead^.^^^^,^^^^ The use of 1,4-butanediol diglycidyl ether to couple car-
bohydrate ligands to Sepharose has also afforded affinity matrices for
lectin p u r i f i ~ a t i o n . ' ~ ~ ~
Adsorption to insolubilized, hog-mucin type A + H substance, fol-
lowed by elution with 2-acetamido-2-deoxy-~-galactose afforded pur-
ified lectins from the seeds of Dolichos bijlorus,lo8Phaseolus lunatus
(lima bean),lS1and Helix pomatia (the edible snail).63Similarly, chitin
has been used for purifying the 2-acetamido-2-deoxy-~-glucose-
binding lectins from Triticum vulgaris (wheat germ),133Solanum
tuberosum (potato),152and Bandeiraea simplicifolia. 125 A soluble ad-
sorbent, the tris(p-azophenyl a-L-fucopyranoside) conjugate of
phloroglucinol, was employed by Yariv and to isolate the
L-fucose-binding protein from Lotus tetragonolobus.
An alternative approach involves insolubilized glycoproteins. Thus,
an affinity system for the isolation of the lectin of red kidney-bean
(Phaseolus v ~ l g a r i s ) ' ~involved~ * ' ~ ~ thyroglobulin-Sepharose, and, for

(150) S. Olsnes and A. Pihl, Eur. J . Biochem., 35, 179-185 (1973).

(150a) H. J. Allen and E. A. Z. Johnson, Carbohydr. Res., 50,121-131 (1976).
(150b) J. Lonngren, I. J. Goldstein, and R. Bywater, FEBS Lett., 68,31-34 (1976).
(150c) M. Horisberger, Carbohydr. Res., 53,231-237 (1977).
(150d) K. Sutoh, L. Rosenfeld, and Y. C. Lee, Anal. Biochem., 79,329-337 (1977).
(150e) R. Uy and F. Wold, Anal. Biochem., 81,98-107 (1977).
(151) W. Galbraith and I. J. Goldstein, FEBS Lett., 9, 197-201 (1970).
(152) M. A. Leon, personal communication.
(153) J. Yariv, A. J. Kalb, and E. Katchalski, Nature, 215,890-891 (1967).
(154) R. L. Felsted, R. D. Leavitt, and N. R. Bachur, Biochim. Biophys. Acta, 405,
72-81 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 139

the isolation of Limulus polyphemus, bovine submaxillary mucin-

Sepharose.’” Fetuin-Sepharose was employed for the isolation of the
agglutinins from wheat germ (Triticum vulgaris), horse-shoe crab
(Limulus polyphemus), jack bean (Canavalia ensiformis), and several
other sources.’56Wheat-germ agglutinin was also isolated by using
insolubilized ovomucoid.ls’ Finally, erythrocytes treated with formal-
dehyde and glutaraldehyde have been used as adsorbents for lectin
is~lation.’~~-’~~
Although seeds are the common source of lectin activity, there are
reports (some contradictory) of agglutinins occurring in leaves, stems,
and root^.^^,^^ For example, black-locust (Robinia pseudoaccacia) lec-
tin’60was isolated from its bark, potato (Solanum tuberosum) lectin16’
from its tubers, and poke-weed (Phytolacca americanum) mitogen162
from its roots. A new class of plant cell-wall and membrane-bound
lectins (“p-lectins”) has been d e ~ c r i b e d . ~A~treatise
~ , ’ ~ collected
~~~~~
several schemes for the purification of l e c t i n ~ . ’ ~ ~ J ~ ~ ~

3. Carbohydrate-binding Specificity of Lectins

The carbohydrate-binding specificity of lectins varies greatly with
respect to the binding of simple sugars, the precipitation of
carbohydrate-containing macromolecules, and the agglutination of
plant and animal cells. In order that a lectin may be a truly useful tool in
carbohydrate chemistry and other areas (for example, biology), it is
necessary that its carbohydrate-binding specificity be established; this
is generally accomplished by the Landsteiner hapten-inhibition tech-
nique.166Sugar-lectin complementarity is established by comparing

(155) J. D. Oppenheim, M. S. Nachbar, M. R. J. Salton, and F. Aull, Biochem. Biophys.

Res. Commun., 58, 1127-1134 (1974).
(156) B.-A. Sela, J. L. Wang, and G. M. EdelmanJ. B i d Chern., 250,7535-7538 (1975).
(157) S. Avrameas and B. Guilbert, Biochimie, 53,603-614 (1971).
(158) R. W. Reitherman, S. D. Rosen, and S. H. Barondes,Nature, 248,599-600 (1974).
(159) T. P. Now& and S. H. Barondes, Biochim. Biophys. Acta, 393,115-123 (1975).
(160) R. Bourrillon and J. Font, Biochim. Biophys. Acta, 154, 28-39 (1968).
(161) E. Gellhom, in “Handbuch der Biochemie des Menschen und Tiere von Op-
penheimer,” F. Schiff, ed., Gustav Fischer, Jena, 1925, Vol. 2, pp. 346-377.
(162) J. Borjeson, R. Reisfeld, L. N. Chessin, P. D. Welsch, and S. D. Douglas,J. E x p .
Med., 124,859-872 (1966).
(163) H. Kauss and C. Glaser, FEBS Lett., 45,304-307 (1974).
(164) D. J. Bowles and H. Kauss, Plant Sci. Lett., 4,411-418 (1975).
(165) Methods Enzymol., 38 (Part B), 313-388 (1972).
(165a) Methods Enzymol., 34 (Part B), 317-331 (1974).
(166) K. Landsteiner, “The Specificity of Serological Reactions,” Dover Publications,
New York, Revised Edition, 1962, pp. 1-330.
140 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

sugars on the basis of the minimal concentration required to inhibit

(I) the precipitin reaction between the lectin and a reactive mac-
romolecule,167-170 or (2) the hemagglutination reaction.10~19~21~22~77.7s.”71
Other procedures, including equilibrium dialysis,172ultraviolet spec-
t r o ~ c o p y ,elution ’ ~ ~solid-phase ads or bent^,^^".'^^ and precipita-
~ ~ ~ ~from
tion with na~Ura~63,102,104.106~117,120,121,124,170,176--178 and model
carbohydrate-protein conjugates113J26J79 have also been informative.
The use of alkyl a-and P-glycosides as hapten inhibitors, in addition
to free sugars, yields vital information on the anomeric specificity of the
lectin, but it is also necessary to test di- and higher oligo-saccharides, in
order to interpret studies on the structural features of polysaccharides
and glycoproteins correctly. (Sugars purchased commercially may con-
tain hemagglutinin substances, and must be purified prior to
Glycosides of aromatic aglycons provide information about the nature
of the protein site adjacent to the anomeric carbon atom ofthe sugar, but
can be misleading insofar as lectin specificity for oligosaccharides is
concerned.113J67J80-184 In effect, the interaction of the protein with an
aromatic moiety is explored, not the carbohydrate-binding specificity
of the lectin; this situation dictates caution in interpreting such experi-
ments.
Although the interaction of lectins with polysaccharides, glycopro-
teins, and glycolipids is, without dispute, more complex than with
simple sugars, the results of hapten-inhibition studies employing
(167) G. F. Springer and P. Williams, Biochem. J., 85,282-291 (1962).
(168) I. J. Goldstein, C. E. Hollerman, and E. E. Smith, Biochemistry, 4,876-883( 1965).
(169) L. L. So and I. J . Goldstein,J. lnzmunol., 99, 158-163 (1967).
(170) K. 0. Lloyd, E. A. Kabat, and S. Beychok,J. Immunol., 102, 1354-1362 (1969).
(171) A. H. Rule and W. C. Boyd, Transfusion (Philadelphia), 4,449-456 (1964).
(172) L. L. So and I. J. Goldstein, Biochim. Biophys. Acta, 165,398-404 (1968).
(173) G. S. Hassing and I. J. Goldstein, Eur. J. Biochem., 16, 545-556 (1970).
(174) W. Bessler, J. A. Shafer, and I. J. GoldsteinJ. B i d . Chem., 249,2819-2822 (1974).
(175) S. M. Chien, S. Singla, and R. D. Poretz,]. Immunol. Methods, 8,169-174 (1975).
(176) G. W. G. Bird, Vox Sang., 4, 307-313 (1959).
(177) H. Markowitz,J. lmmunol., 103, 308-318 (1969).
(178) S. Hammarstrom, A. A. Lindberg, and E. S. Robertsson, Eur.J.Biochem., 25,274-
282 (1972).
(179) I. J. Goldstein and R. N. Iyer, Biochim. Biophys. Acta, 121, 197-200 (1966).
(179a) M. D. Marquardt and J. A. Gordon, Nature, 252, 175-176 (1974).
(180) R. D. Poretz and I. J. Goldstein, Biochem. Pharmacol., 20,2727-2739 (1971).
(181) F. G. Loontiens, J. P. Van Wauwe, R. De Gussem, and C. K. D e Bruyne,
Carbohydr. Res., 30, 51-62 (1973).
(182) J.-P. Van Wauwe, F. G. Loontiens, H. A. Carchon, and C. K. De Bruyne,
Carbohydr. Res., 30,249-256 (1973).
(183) T. Irimura, T. Kawaguchi, T. Terao, and T. Osawa, Carbohydr. Res., 39,317-327
(1975).
(184) R. D. Poretz, H. Riss, J. W. Timberlake, and S. M. Chien, Biochemistry, 13,250-
256 (1974).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 141

mono- and oligo-saccharides and their derivatives have, in all cases,

been applicable to studies with complex saccharides. (A single exam-
ple, the unexpected reactivity of the pea lectin with a glycopeptide, has
yet to be explained.I6') The higher affinity-constants observed for bind-
ing of lectins to glycoproteins and cell-surface carbohydrates than to
simple sugars can be rationalized on the basis of multivalent interac-
tions between the lectin (almost all lectins contain two or more binding
sites; see, for example, Ref. 186) and the complex sa~charide.'~'-'~~
Furthermore, nonspecific interactions between the lectin and the mac-
romolecule, and the influence of steric factors, must be considered, in
addition to the specific interaction between the carbohydrate and the
le~tin.'~~*'~3
In early studies, Morgan and WatkinsZ2observed that sugars (for
example, L-fucose, 2,6-dideoxy-~-lyxo-hexose, L-galactose, and
6-deoxy-~-talose)that inhibit the L-fucose-binding lectin from Lotus
tetragonolobus all have the same configuration at C-3 and C-4. Simi-
larly, K r i i ~ enoted
~ ~ that sugars (2-acetamido-2-deoxy-~-galactose,
D-galactose, lactose, and melibiose) that inhibit the agglutinin from
Sophora japonica seeds all have the same configuration at C-3 and C-4.
Generalizing from these and his own studies, Make1a78suggested that
lectin-reactive monosaccharides may be divided into four classes,
based on their configuration at C-3 and C-4 of the pyranose forms (see

c)
Fig. 1).

no
noo

no
0 OH noo

I
no HO
I II 111 IV
FIG. 1.-Classification of Pyranoses into Four Croups, Based on the Configurations
of the 3- and 4-Hydroxyl Groups (after Makela7a).

(185) J. PospiSilovh, G. Entlicher, and J. Kocourek, Biochim. Biophys. Acta, 362,593-

597 (1974).
(186) M . D . Stein, I. K. Howard, and H. J. Sage,Arch. Biochem. Biophys., 146,353-355
(1971).
(187) M. D . Stein, H. J. Sage, and M. A. Leon, Arch. Biochem. Biophys., 150,412-420
(1972).
(188) P. W. Majerus and G. N. Brodie,J. B i d . Chem., 247,4253-4257 (1972).
(189) S. Hammarstrom, Scand. J. Immunol., 2, 53-66 (1973).
(190) C. L. Homick and F . b r u s h , Immunochemistry, 9,325-340 (1972).
(191) C. E. Hayes and I. J . Coldstein,J. B i d . Chem., 250, 6837-6840 (1975).
(192) M . W. Davey, J. W. Huang, E. Sulkowski, and W. A. Carter,J. Biol. Chem., 249,
6354-6355 (1974).
(193) S. K. Podder,A. Surolia,and B. K. Bachhawat,Eur.J. Biochem., 44,151-160(1974).
142 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

For a time, before extensive and systematic studies on pure lectins

were conducted, this classification served a useful purpose. Thus,
agglutinins derived from the pea(Pisurn sativum)20and the lentil (Lens
~ u l i n a r i s ) ' ~preferentially
~*'~ bound Makela's group I11 sugars, being
inhibited by D-glucose and D-mannose, whereas lectins from the seeds
of Sophora japonica,17 Ricinus c o r n m ~ n i s , ~and ~ * ~Bandeiraea
~~
~ i m p l i c i f o l i a ~were
~ ~ * inhibited
'~~ by group I1 carbohydrates. Sugars of
group I inhibit the lectins from Lotus tetragonolobusZ2 and the
L-fucose-binding lectin of Ulex e u r o p e u ~ .As '~~yet no lectin has been
isolated that interacts with Makela's group IV sugars (D-idose,
D-gulose, L-glucose, and L-xylose). The limitations of this classification
have already been mentioned.
Unlike human agglutinins, which generally require a disaccharide
hapten as the smallest fragment that will inhibit precipitation, the
binding site of most lectins accommodates a single glycosyl residue. In
a few cases, however, the nature of the glycosidic linkage plays a role in
determining the binding specificity. Thus, isomaltose and other
oligosaccharides containing nonterminal, a-D-( 1+6)-glucosidic bonds
have a higher affinity for con A than oligosaccharides having nonreduc-
ing a-D-( 1+2, -+3, or -4)-glucosidic linkages.'68J97 A more-
pronounced linkage-specificity was observed by O ~ a w a 'for ~ the
Laburnum alpinum and Cytisus sessilifolius lectins. Only disac-
charides containing a glycosidic bond to a secondary hydroxyl group
were bound by these lectins; for example, laminarabiose and cel-
lobiose are good inhibitors, whereas gentiobiose is a noninhibitor. A
similar specificity is exhibited by the B . simplicifolia 2-acetamido-2-
deoxy-D-glucose-binding 1 e ~ t i n . l ~ ~
The fact that several lectins bind di- and tri-saccharides'2J more
strongly than alkyl glycosides indicates that the reducing sugar may
contribute to the binding energy of the lectin-carbohydrate complex.
For example, a trisaccharide inhibits soybean lectin better than the
corresponding disaccharide lacking its "reducing" sugar residue.lg8
The same phenomenon was observed for the Dolichos bi.orus108 and
l i m a - b e a ~(Phaseolus
~'~~ lunatus) lectins. The precise position of a de-
terminant sugar in an oligo- or poly-saccharide is also important to
lectin reactivity. Pereira and KabatZo0observed that the location of
(194) G. L. Nicolson and J. Blaustein, Biochim. Biophys. Acta, 266, 543-547 (1972).
(195) 0.Makela, P. Makela, and M. Kriipe, Z. Immunitaetsforsch.Exp. Ther., 117,220-
229 (1959).
(196) I. Matsumoto and T. Osawa, Vox Sang., 21, 548-557 (1971).
(197) E. E. Smith and I. J. Goldstein, Arch. Biochem. Biophys., 121, 88-95 (1967).
(198) M. E. A. Pereira, E. A. Kabat, and N. Sharon, Carbohydr. Res., 37,89-102 (1974).
(199) W. Galbraith and I. J. Goldstein, Biochemistry, 11, 3976-3984 (1972).
(200) M. E. A. Pereira and E. A. Kabat, Biochemistry, 13,3184-3192 (1974).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 143

a-L-fucopyranosyl end-groups in various oligosaccharides is of

paramount importance in determining the reactivity with the Lotus
tetragonolobus lectin (compare the L-fucose-binding lectin of Eu-
onymus europeus; see Section VII,3). Examples in which the reducing
sugar residue does not contribute to the affinity of a lectin for a disac-
charide include the Bandeiraea simplicifolia lectin, for which, methyl
a-D-galactopyranoside, melibiose, and 2-0-a-D-ga~actopyranosyl-D-
glucose all bind to the same extent.l3I On the other hand, the peanut-
lectin combining-site interacts in a highly specific way with the di-
saccharide 2-acetamido-2-deoxy-3-O-~-D-ga~actopyranosy~-D-ga~ac-
tose201-203
Only for con A [which binds (1+2)-a-D-manno-oligosaccha-
ride^^^^*^^*^^^] and a series of lectins (wheat-germ agglutinin,128potato
and Ulex europeus lectinZo8)that bind N-peracetylated
chito-oligosaccharides has it been shown that the carbohydrate-
binding site is complementary to three or more sugar units (see later).
With but few exceptions, lectins interact with the nonreducing, ter-
minal glycosyl groups of polysaccharide and glycoprotein chain-ends.
Con A, on the other hand, in addition to its interaction with a - ~ -
mannopyranosyl and a-D-glucopyranosyl terminal groups, binds
internal 2-0-a-D-mannopyranosyl residues.209Similarly, three other
D-mannose-specificlectins (those of the pea, the lentil, and broad bean)
reportedly interact with the (reducing) D-mannose residue of 2-0-(2-
acetamido-2-deoxy-~-~-g~ucopyranosy~)-~-mannose.~'~ Wheat-germ
agglutinin (and, most probably, the 2-acetamido-2-deoxy-~-glucose-
binding lectins of the potato and Ulex europeus) interacts with
internal, 4-O-substituted, 2-acetamido-2-deoxy-&glucopyranosyl resi-
dues.128
Lectins differ markedly with respect to their anomeric specificity.
Some, such as con A (Refs. 168, 169, and 197), the lectins from Ban-
(201) G. Uhlenbruck, G. I. Pardoe, and G. W. G. Bird, Z . Immunitaetsforsch. Allerg.
Klin. Immunol., 138, 423-433 (1969).
(202) R. Lotan, E. Skutelsky, D. Danon, and N. Sharon,J. Biol. Chem., 250,8518-8523
( 1975).
(203) M. E. A. Pereira, E. A. Kabat, R. Lotan, and N. Sharon, Cnrbohydr. Res., 51, 107-
118 (1976).
(204) L. L. So and I. J. Goldstein,J. Biol. Chem., 243, 2003-2007 (1968).
(205) I. J. Goldstein, Ado. E x p . Med. Biol., 55, 35-53 (1974).
(206) G. I. Pardoe, G. W. G. Bird, and G. Uhlenbruck, Z. Immunitaetsforsch. Allerg.
Klin. Immunol., 137, 442-457 (1969).
(207) A. K. Allen and A. Neuberger, Biochem. J,, 135,307-314 (1973).
(208) I. Matsumoto and T. Osawa, Arch. Biochem. Biophys., 140,484-491 (1970).
(209) I. J. Goldstein, C. M. Reichert, A. Misaki, and P. A. J. Gorin, Biochim. Biophys.
Acta, 317, 500-504 (1973).
(210) R. Kaifu, T. Osawa, and R. W. Jeanloz, Carbohydr. Res., 40, 111-117 (1975).
144 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

deiraea simpli~ifolia,'~~ the pea,211and Lotus tetragonolobus,22.200 ex-

hibit pronounced, anomeric specificity, whereas other lectins, such as
those from soybean1g8~212 and castor bean'46(RCA,), appear to be almost
indifferent.
Many lectins tolerate some variation at C-2 of the sugars which they
bind. Thus, con A (Refs. 168 and 169) and the lectins from the pea
(Pisum sativum),211lentil (Lens ~ u l i n a r i s ) , ' ~and ~ , ' fava
~ ~ bean (Vicia
f a b ~ ) all
~ ' exhibit
~ a primary specificity for D-mannose, but will also
bind D-glucose and, to a lesser extent, 2-acetamido-2-deoxy-~-glucose.
A considerable number of lectins display a preferential affinity for
2-acetamido-2-deoxy-~-galactose, but also react, to different extents,
with D-galactose (for example, the lectins from Dolichos bi.orus,'08
~ ~ Helix p ~ m a t i a , ' ~ ~Conversely,
Phaseolus l ~ n a t u s , 'and * ~ ' ~ ) , there is a
series of lectins that display a primary specificity for D-galactose, and
cross-react to a varying degree with 2-acetamido-2-deoxy-D-galactose.
These include the a-D-galactopyranosyl-binding proteins from Ban-
deiraea ~implicifolia'~' and Ricinus c ~ m m u n i s . ' ~ ~
In general, lectins tolerate very little variation at C-3 of the sugars
they bind, although a report213indicates that the Viciafaba lectin has a
greater affinity for 3-O-methyl-D-glucosethan for D-glucose (compare,
the eel a g g l ~ t i n i n ' and
~ ~ *pea
~ ~lectin211).
~
The 4-hydroxyl group of carbohydrates is also critically involved in
lectin binding. In general, D-glucose(D-mannose)-binding lec-
do not interact with D-galactose, and vice versa.131~146
tins20,'23,'43,'68,213
Similarly, 2-acetamido-2-deoxy-~-glucose-binding lectins do not inter-
act with 2-acetamido-2-deoxy-~-galactose.'~~~~~~-~~~ Lectins that bind
2-acetamido-2-deoxy-~-galactose do not generally interact with
2-acetamido-2-deoxy-~-glucose,although the Helix pomatia agglutinin
interacts with 2-acetamido-2-deoxy-~-glucose,albeit with only 10% -
of the affinity of 2-acetamido-2-deoxy-~-galactose.~~
Certain sugars in their furanose form unequivocally bind to con A (for
example, D-fructose and ~ - a r a b i n o s e ) .There ~ ~ ~ ~is
' ~reason
~ ~ ~ ~to~expect
that the pea, lentil, and Vicia fuba lectins will also bind
D-fmctofuranosyl and D-arabinofuranosyl end-groups. A single report
that wheat-germ agglutinin binds 2-acetamido-2-deoxy-~-glucose in its
furanose form needs confirmation.lZ8

J. P. Van Wauwe, F. G. Loontiens, and C. K. D e Bruyne, Biochim. Biophys. Acta,

379,456-461 (1975).
H. Lis, B.-A. Sela, L. Sachs, and N. Sharon, Biochim. Biophys. Acta, 211,582-585
(1970).
A.K. Allen, N. N . Desai, and A. Neuberger, Biochem. J . , 155, 127-135 (1976).
G. F. Springer and P. Williamson, Vox Sang., 8, 177-195 (1963).
L. L. So and I. J. Goldstein, Carbohydr. Res., 10,231-244 (1969).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 145

Finally, a group of lectins having a complex, sugar specificity that has

yet to be established includes agglutinins from the red kidney-bean
(Phaseolus vulgaris),216 the spindle tree(Evonymus europaea L.),217 the
scarlet runner-bean (Phaseolus coccineus L.),218 Vicia graminea,219,220
and the meadow mushroom (Agaricus bisporus).221

4. Nomenclature of Lectins
In any rapidly developing area of science, it is necessary to establish
a systematic nomenclature, a problem not yet confronted in the already
extensive field oflectin biochemistry. The names of a few lectins derive
from their genus of origin (for example, ricin, abrin, concanavalin A,
and favin for the lectins from Ricinus communis, Abrus precatorius,
Canavalia ensiformis, and Viciafaba, respectively); others bear com-
mon names (for example, pea, soybean, and lima-bean lectins from the
seeds of Pisum sativum, Glycine max, and Phaseolus lunatus, respec-
tively). Particularly inappropriate is the accepted designation PHA for
the phytohemagglutinin of the red kidney-bean (Phaseolus vulgaris).
Employing the nomenclature ofblood-group ~ e r o l o g y ,Prokop
~ ~ ~ ,and
~~~
his coworker^^^^,^^^ proposed a nomenclature for plant and animal
agglutinins. As an example, the lectins from Dolichos bi$orus, Helix
hortensis, and Helix pomatia, all of which agglutinate type A erythro-
cytes were termed ADb,AHH, and AHP-the A standing for antigen, and
Db, HH, and HP, for the initial letters of the Latin name. At present,
most new lectins are cited by their genus and species names; for
example, the lectin from Wistaria Poribunda and the lectin from
Sophora japonica. (In a few cases, the sugar-binding specificity of the
lectin has also been included in the designation.)
At least two serious deficiencies are associated with all of these
designations. First, it is theoretically possible, and indeed known, that
several lectins having different sugar-binding specificities are present
in the same plant-seed (for example, Ulex europeus contains a-L-
fucopyranosyl-binding and 2-acetamido-2-deoxy-/3-~-glucopy-
ranosyl-binding lectins208,22s,,226; Bandeiraea simplicifolia con-
(216) R. Kornfeld and S. Kornfeld, J . Biol. Chern., 245,2536-2545 (1970).
(217) F. Pacak and J. Kocourek, Biochirn. Biophys. Acta, 400, 374-386 (1975).
(218) N. Nowakova and J. Kocourek, Biochirn. Biophys. Acta, 359,320-333 (1974).
(219) G. F. Springer, H. Tegtmeyer, and S. V. Huprikar, Vox Sang., 22,325-343 (1972).
(220) M. J. Prigent and R. Bourrillon, Biochirn. Biophys. Acta, 420, 112-121 (1976).
(221) H. J. Sage and J. J. Vasquez, J . Biol. Chern., 242, 120-125 (1967).
(222) A. S. Wiener, Transfusion (Philadelphia), 1, 308-320 (1961).
(223) A. S. Wiener, Blood, 27, 110-125 (1966).
(224) 0. Prokop, G. Uhlenbruck, and W. Kohler, Vox Sang., 14,321-333 (1968).
(225) L. L. Flory, Vox Sang., 11, 137-156 (1966).
(226) I. Matsumoto and T. Osawa, Biochirn. Biophys. Acta, 194, 180-189 (1969).
146 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

tains a-D-galaCtOpyranOSyl-’3’ and 2-acetamido-2-deoxy-D-glucosyl-

binding lectinslZ5).Second, these designations are generally quite un-
informative with regard to the sugar-binding specificity of the lectin.
There is certainly a need for an informative nomenclature based on a
distinguishing characteristic or feature, similar to that formulated for
enzymes.227 (Were a physiological hnction for lectins to be discovered,
however, it might be necessary to revise such a system; see Section 1,5.)
Makela classified7*lectins into four groups, based on the configura-
tion of C-3 and C-4 of the most complementary pyranose (see Fig. 1).
K r i i ~ elisted
~ ~ three groups. The rationale for this system was that
lectins can usually be assigned to one of the four groups. However, this
framework is severely limited, because it fails to stipulate the con-
figuration of C-2 or of the anomeric carbon atom. Thus, concana-
valin A, wheat-germ agglutinin, and the 2-acetamido-2-deoxy-~-
glucopyranosyl-binding protein from the seeds of Bandeiraeu
simplicifolia all conform to Makela’s group I11 lectins and yet, they
preferentially bind such disparate sugars as a-D-mannopyranose,
N,N’,N’’-triacetylchitotriose,and 2-acetamido-2-deoxy-~-glucopy-
ranose, respectively.
Inasmuch as new lectins are being isolated, characterized, and
utilized in a variety of systems, we now suggest a nomenclature that
cites the genus and species, followed by the sugar-binding specificity
in parentheses. If a lectin is not completely specific for a single sugar
and anomeric form, it would be necessary to cite two or three structures.
If, on the other hand, the lectin exhibits no anomeric preference, the
sugar would be cited without the a or /3 designation [for example, the
lectins from Canaualia ensiformis (a-D-Manp > a - ~ - G l c p> a-D-
GlcNAcp), Triticum vulgaris (N,N’,N”-triacetylchitotriose > N,N’-
diacetylchitobiose > > p-~-GlcNAcp),and Lotus tetrugono2obus (a-
L - F u ~].~Use) of this nomenclature would obviously require adequate
characterization of the sugar-binding specificity of each lectin so de-
scribed.

5. Function of Lectins
Although numerous physiological roles have been speculated for
plant lectins, there is a paucity of supportive data for any of them. The
interaction of lectins with animal cells and carbohydrates is almost
certainly the result of coincidental complementarity between the lectin
(227) “Enzyme Nomenclature,” Recommendations (1972) of the International Union of
Pure and Applied Chemistry and the International Union of Biochemistry;
Revision of the Recommendations of 1964. American Elsevier Publishing Co.,
New York, 1972, pp. 1-443.
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 147

combining-site and the sugar c ~ n f i g u r a t i o nOn . ~ the

~ ~ other
~ ~ ~ hand, it is
probably no accident that these carbohydrate-binding proteins are
adapted to bind certain carbohydrate^.^^ K r i i ~ efirst ~ ~suggested that
lectins might function as “Kohlenhydratfixieren” (carbohydrate catch-
ers), with roles in such physiological functions as sugar transport, stor-
age, or i m m o b i l i z a t i ~ n ~ Jbeing
~ , ~ ~a ~distinct
* ~ ~ ~ possibility.
An antibody-like role for the lectins, intended to counteract soil
bacteria, was postulated by P ~ n i n ~and ~ OSaint-PaulZ3* (compare, Ref.
232). Indeed, the lectin of Vicia cracca inhibits the growth of a soil
micro-organism known to metabolize the seed coats of this s p e ~ i e s . 2 ~ ~
Wheat-germ agglutinin inhibits synthesis of chitin, hyphal growth, and
spore germination by binding to the hyphal tips of Trichoderma viride
and Fusarium ~ o l a n iIn. ~ all~these
~ cases, a mechanism must be estab-
lished whereby the plant lectin is transported from the seeds to the
roots, and then secreted into the soil.
A more-interesting hypothesis predicted a relationship between the
agglutinin of a leguminous species and the bacteria of the genus
Rhizobium found in the root nodules of the and theorized that
the lectin might bind the bacteria to the roots. Although such a relation-
ship is not established by the demonstration that a lectin agglutinates
the rhizobial micro-organisms that inhibit that species of legume, there
have been several such reports, none of them experimentally convinc-
As evidence against such a relationship, Makela78found that
ing.2366-237e
extracts from the seeds of Trifolium pratense, Pisum sativum, Vicia
faba, and Phaseolus vulgaris agglutinate neither their own Rhizobium
leguminosarum strains nor the strains of one another. Similarly, Dazzo
and H ~ b b e 1 demonstrated
1~~~~ that con A combined strongly with vari-

(228) A. Ensgraber, Ber. Dtsch. Bat. Ges., 29, 349-361 (1958).

(229) W. C. Boyd, D. L. Everhart, and M. H. McMaster,J. Immunol., 81, 414-418
(1958).
(230) W. Punin, Z . Naturforsch., Teil B , 7,48-50 (1952).
(231) M. Saint-Paul, Transfusion (Philadelphia), 4, 3-37 (1961).
(232) P. Albersheim and A. J. Anderson, Proc. Natl. Acad. Sci, U.S.A., 68, 1815-1819
(1971).
(233) D . A. Jones, Heredity, 19,459-469 (1964).
(234) D. Mirelman, E. Galun, N. Sharon, and R. Lotan, Nature, 256,414-416 (1975).
(235) M. Krupe, Z . Immunitaetsforsch. E x p . Ther., 107,450-464 (1950).
(236) J. Hanlblin and S. P. Kent, Nature (London) New Biol., 245,28-30 (1973).
(237) B. B. Bohlool and E. L. Schmidt, Science, 185, 269-271 (1973).
(237a) F. Dazzo and D . Hubbell, Appl. Microbiol., 30,1017-1033 (1975).
(237b) J. S. Wolpert and P. Albersheim, Biocliem. Biophys. Res. Commun., 70,729-737
(1976).
(237c) F. Dazzo, C. Napoli, and D . Hubbell, Appl. Microbiol., 32, 166-171 (1976).
(237d) F. Dazzo and D . Hubbell, Plant Soil, 43,717-722 (1975).
148 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

ous nodulating and non-nodulating strains of Rhizobium regardless of

their respective hosts. O t t e n s ~ o s e P
speculated
~~ that the production of
agglutinins may be stimulated by infection with rhizobial micro-
organisms. More-convincing evidence for the participation of lectins
in the recognition processes between leguminous plants and rhizobial
bacteria comes from the studies of Dazzo and his colleagues.
These workers isolated, from clover seeds and roots, a lectin that
binds Rhizobium trifolii organisms to clover root-hairs. Of the sugars
tested, only 2-deoxy-~-arubino-hexose (“2-deoxy-D-glucose”) inhib-
ited this binding interaction; this 2-deoxy sugar was also shown to be
a constituent of the capsular heteropolysaccharide S unique to R.
trifolii.
The striking parallel between the simple sugars that interact with
jack-bean lectin (concanavalin A) and the monosaccharides that
serve as substrates for yeast hexokinase (see Fig. VI.3 in Ref. 239)
suggested a possible enzymic function for the l e ~ t i nHowever, . ~ ~ ~ in-
vestigation failed to detect any hexokinase, phosphatase, or phos-
phorylase activity. The more obvious possibility of a glycosidase activ-
ity for lectins must be examined critically, in view of the abundance of
glycosidases in most p l a n t - s e e d ~Unlike
. ~ ~ ~ glycosidases, the activity of
lectins does not increase after germination, and several purified-lectin
preparations do not hydrolyze the disaccharides that inhibit their
agglutination of animal cells.z0~z28 The observation that UDP-D-
galactose binds to the a-D-galactopyranosyl-binding lectin from Ban-
deiruea simplicifolia seeds signaled a possible D-galactosyltransferase
role for this lectin, although there is actually no evidence to support this
view.131
A few investigators have studied the appearance and disappearance
oflectins during the life cycle of the plant.z4z,243 J. M. Jones and cowork-
e r reported
~ ~ that ~ Maclura
~ pomiferu agglutinin begins to accumulate

(238) F. Ottensooser, Arch. Biol., 39,76-85 (1955).

(238a) F. B. Dazzo, C. Napoli, and D. Hubbel1,Appl. Enuiron. Microbiol., 32,166-171
(1976).
(238b) F. B. Dazzo and W. J. Bril1,Appl. Enuiron. Microbiol., 33, 132-136 (1977).
(239) M. Dixon and E. C. Webb, “Enzymes,” Academic Press, New York, 2nd Edition,
1964, p. 217.
(240) I. J. Goldstein, C. M. Reichert, and A. Misaki, Ann. N.Y. Acad. Sci., 234,283-296
( 1974).
(241) K. M. L. Agrawal and 0. P. Bahl,J. Biol. Chem., 243, 103-111 (1968).
(242) J. M. Jones, L. P. Cawley, and G. W. Teresa, J. Immunol., 98,364-367 (1967).
(243) I. K. Howard, H. J . Sage, and C. B. Horton,Arch. Biochem. Biophys., 149,323-326
(1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 149

during the early development of the seed, reaches a maximum as the

seed achieves maturity, and decreases slowly during germination, but
it could still be detected in low concentrations at a seedling age of six
months. All lectin activity was localized within the seed, none being
found in either the seed coat or the leaves of the mature tree. On
studying the same phenomenon in the lentil (Lens culinaris), Howard
and noted the appearance of lectin during early develop-
ment and differentiation ofthe seed embryo; a possible involvement in
the germination and maturation process was thus indicated. Con A
and the lectin from Phnseolus vulgaris were shown to stimulate the
germination of The castor-bean agglutinin, present only
in the matrix region of protein bodies of the seed e n d o ~ p e r m , 2 ~ ~ ~ . ~
undergoes slow degradation during germination, in contrast to storage
proteins .243b
Talbot and Etzler isolated a possible “prolectin” from the stems
and leaves of the Dolichos biflorus plant.243dPresent in microgram
quantities, the prolectin cross-reacts with antibodies to the seed lec-
tin. The cross-reacting material, which neither agglutinates human
blood-group A erythrocytes, nor binds to insolubilized, hog blood-
group (A + H ) substance (as does the seed lectin), has one subunit
identical to subunit I A of the seed lectin, and a unique subunit of
higher molecular weight. In common with the seed lectin, both “pro-
lectin” subunits have an N-terminal alanine residue, suggesting an
extension at the carboxyl end of one of the polypeptide chains. Mia-
lonier and also discovered the presence of a substance
in the leaves of Phaseolus vulgaris L. (var. red) that cross-reacted with
antiserum to the seed lectin.
Finally, it has been shownz4 that addition of black-bean (Phaseolus
vulgaris) phytohemagglutinin to the normal diet of the bruchid beetle
that can eat phytohemagglutinin-free cowpeas (Vigna unguiculata),
but not P . vulgaris seeds, kills the bruchid larvae. It was concluded2”
that “a major part of the adaptive significance of phytohemagglutinins
in black bean and other legume seeds is to protect tham from attack b y
insect seed predators.”

(243a) D. Southworth, Nature, 258, 600-602 (1975).

(24311) R. J . Youle and A. H. C. Huang, Plant Physiol., 58,703-709 (1976).
(243c) R . E. Tully and H. Beevers, Plant Physiol., 58,710-716 (1976).
(243d) C. F. Talbot and M. E. Etzler, Fed. Proc., 36,795 (1977).
(243e) G . Mialonier, J.-P. Privat, M . Monsigny, G . Kahlem, and R. Durand, Physiol.
Veg., 11,519-537 (1973).
(244) D. H. Janzen, H. B. Juster, and I. E. Liener, Science, 192, 795-796 (1976).
150 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

LECTINS
11. D-MANNOSE(D-GLUCOSE)-BINDING
1. Concanavalin A of the Jack Bean (Canavalia ensiformis)
(jack bean; a-D-Manp > a - ~ - G l c p> a-~-GlcNAcp)
a. Introduction; Physical and Chemical Characterization.-
Concanavalin A (con A) is without doubt the most notable of the plant
agglutinins. First isolated and crystallized by Sumner and H 0 ~ e 1 1 , ~ ~ ~ , ~
who identified it as the phytohemagglutinin of the jack bean
(Canuvulia this lectin exhibits a series of remarkable
properties. It has been employed as a structural probe for carbohy-
’ ~well as at cell surfaces,25as a lymphocyte mito-
drates in s ~ l u t i o nas
gen,248,249as a reagent for differentiating normal from malignant
ce11s,31~2so~251as an anticancer agent,33,252 and, in its interaction with
carbohydrates, as a model for the antibody-antigen reac-
tion.72*102,’21~168~253
Currently, its interaction with hormone-like receptors
(for example, insulin) is being studied i n t e n ~ i v e l y . ~Sepharose-
~-~~
coupled con A (con A-Sepharose) is being used to isolate membrane
glycoproteins, enzymes, cell-surface receptors, and viral glycoproteins,
all of which possess con A-reactive oligosac~harides.~~ A novel
immuno-electrode composed of con A covalently attached to a
poly(viny1 chloride) membrane has been constructed, and shown to be
capable of sensing yeast D - ~ ~ l a n n a nTwo
. ’ ~ ~volumes devoted entirely to
con A have been p u b l i ~ h e d , ~ ~ ~ ~ ~ ~
Sumner and HowellLo2 prepared con A by extracting jack-bean meal
(pre-extracted with aqueous acetone and ethanol) with 5% sodium
chloride solution, and dialyzing the salt extract against distilled water
containing toluene. From 100 g of meal, these investigators obtained
2.5-3.0 g of crystalline con A. The lectin, which could not be entirely
freed of carbohydrate,lo2agglutinated cat, dog, guinea-pig, horse, and
rabbit erythrocytes, but not the red blood-cells ofthe cow, goat, human,
pig, or sheep.10224462447
In all respects, con A exhibited the properties of a protein: it was
(245) J. B. Sumner,]. B i d . Chem., 37, 137-142 (1919).
(246) J. B. Sumner, S. J. Howell, and A. Zeissig, Science, 82, 65-66 (1935).
(247) J. B. Sumner and S. I?. Howell,]. Ztnmunol., 29, 133-134 (1935).
(248) M . Wecksler, A. Levy, arid W. G. Jaffh, Acta Cient. Venez., 19, 154-156 (1968).
(249) A. E. Powell and M. A. Leon, E x p . Cell Res., 62,315-325 (1970).
(230) M. Inbar and L. Sachs, Nature, 223, 710-712 (1969).
(251) H. Ben-Bassat, M. Inbar, and L. Sachs,]. Membr. Biol., 6, 183-194 (1971).
(252) J. Shoham, M. Inbar, and L. Sachs, Nature, 227, 1244-1246 (1970).
(253) E. J. Hehre, Bull. Soc. Chim. Biol., 42, 1581-1585 (1960).
(254) J. Janata,J. Am. Chem. Soc., 97, 2914-2916 (1975).
(255) H. Bittiger and H. P. Schnebli, “Concanavalin A as a Tool,” Wiley, New York,
1976, pp. 1-639.
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 151

denatured102by 0.1 M hydrochloric acid and 0.1 M sodium hydroxide,

readily digested by pepsin102(but it resisted tryptic digestionlo2),con-
tained 3.36% of tryptophan, 5.2% of tyrosine, and 0.4% of c y ~ t e i n e , ~ ~ ~
and had an isoelectric pointlo2of 5.5and a molecular weight of 96,000in
neutral, phosphate b ~ f f e r . ~A~metalloprotein,
'=~~~ the jack-bean lectin
contains Mn2+ and Ca2+; removal of the metal ions abolished the
hemagglutinating
In its ability to precipitate glycogen and starch from solution,
agglutinate erythrocytes from many animal species, and clump certain
bacteria (such as Mycobacterium and Actinomyces),con A was likened
to an antibody.'02Sumner and Howell102also suggested "that the chem-
ical component in stromata [of rabbit erythrocytes] with which con A
unites may be glycoprotein. . ." These same investigators observed102
that cane sugar "prevents the agglutination of starch as well as the
agglutination of erythrocytes," thus demonstrating the first hapten in-
hibition of the con A system by sugars of low molecular weight. In-
hibition of the turbidimetric reaction between con A and polysac-
charide has been used for the quantitative analysis of D - m a n n o ~ e . ' ~ ~ ~
In 1965,the isolation of con A in pure form was reported by Agrawal
~ ~ ~ ' ~ ~ the observation that con A precipitates
and C o l d ~ t e i n . ' Exploiting
dextrans,120,121~253,260 these investigators adsorbed the jack-bean lectin
on cross-linked dextran gel (Sephadex),and eluted it with D-glucose, a
competitive inhibitor of this interaction.120,168 Some time later, Olson
and L i e n e F also purified con A to homogeneity by adsorption to
Sephadex and elution with M acetic acid. These reports constituted
the first instance of lectin purification by affinity chromatography.
Physical characterization of con A revealed monodispersity in the
ultracentrifuge at pH 2-5, and a two-peak pattern at pH 7, suggesting
pH-dependent association of s ~ b ~ n i t s It . ~is ~
now
~ *known
~ ~ ~ * ~ ~ ~

(256) J. B. Sumner and V. A. Graham,]. B i d . Chem., 64,257-261 (1925).

(257) J, B. Sumner, N. Gralkn, and I.-B. Eriksson-Quensel, Science, 87,395-396 (1938).
(258) J. B. Sumner, N. GralBn, and 1.-B. Eriksson-Quensel,]. Biol. Chem., 125,45-48
(1938).
(259) J. B. Sumner and S. F. Howell,]. Biol. Chem., 115,583-588 (1936).
(259a) R. D. Poretz and I. J. Goldstein, Carbohydr. Res., 4,471-477 (1967).
(260) J. A. Cifonelli, B. A. Lewis, R. Montgomery, and F. Smith,ABstr. Pap. Am. Chena.
soc. Meet., 129,313 (1956).
(261) B. B. L. Agrawal and I. J. Goldstein, Biochim. Biophys. Acta, 133,376-379 (1967).
(262) B. B. L. Agrawal and I. J. Goldstein,Arch. Biophys. Biochem., 124,218-229 (1968).
(262a) K. D. Hardman and C. F. Ainsworth, Biochemistry, 11,4910-4919 (1972).
(263) J. L. Wang, B. A. Cunningham, and G. M. Edelman, Proc. N a t l . Acad. Sci. U.S.A.,
68, 1130-1134 (1971).
(264) A. B. Edmundson, K. R. Ely, D. A. Sly, F. A. Westholm, D. A. Powers, and I. E.
Liener, Biochemistry, 10, 3554-3559 (1971).
152 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

that con A consists of polypeptide subunits, ofmolecular weight 26,000.

At pH 5.6 and below, two protomers are associated in a dimer of
molecular weight 52,000; above pH 5.6, the dimers aggregate, forming
tetramers. As aggregation is pH- and temperature-dependent,2s4a
molecular weight determinations must be scrutinized in terms of the
conditions of measurement. The original determination of 96,000 (in
neutral, phosphate buffer) by Sumner and coworker^^^^*^^^ is in excel-
lent agreement with current estimates based on amino acid analyses
(dimer, 52,000; tetramer, 104,000). Molecular-weight determinations
vary between 50,000
from analytical, ultracentrifuge data136,261,262,265,266
and 120,000.
Con A holds the distinction of being the first lectin to have its amino
acid sequence determined,264.267-271 and the first for which the three-
dimensional crystal-structure has been solved. The
multiple isomorphous-replacement method, using a series of heavy-
metal derivatives (sodium mersalyl, Pb(NO,),, K,PtCl,, K2PtC16,and
mercuric phenylacetate), revealed the overall molecular architecture of
crystalline con A by low-resolution, X-ray d i f f r a ~ t i o n , 2 ~and
~ - ~later,
~~3~~
by high-resolution studies at267-269375*276
200 and262a*274 240 pm to be a
composite of four identical protomers (molecular weight 26,000).The
pseudotetrahedral cluster has an aggregate molecular weight of
(264a) M. Huet, Eur. J . Biochem., 59,627-632 (1975).
(265) A. J. Kalb and A. Lustig, Biochim. Biophys. Acta, 168,366-367 (1968).
(266) G. H. McKenzie, W. H. Sawyer, and L. W. Nichol, Biochim. Biophys. Acta, 263,
283-293 (1972).
(267) G. M. Edelman, B. A. Cunningham, G. N. Reeke, Jr., J. W. Becker, M. J. Waxdal,
and J. L. Wang, Proc. Natl. Acad. Sci. U.S.A., 69, 2580-2585 (1972).
(268) G. N. Reeke, Jr., J. W. Becker, B. A. Cunningham, G. R.Gunther, J. L. Wang, and G.
M. Edelman, Ann. N.Y. Acad. Sci., 234, 369-382 (1974).
(269) G. N. Reeke, Jr., J. W. Becker, B. A. Cunningham, J. L. Wang, I. Yahara, and G. M.
Edelman, Ado. E x p Med. Biol., 55, 13-33 (1975).
(270) J. L. Wang, B. A. Cunningham, M. J. Waxdal, and G. M. Edelman, J . Biol.
Chem., 250, 1490-1502 (1975).
(271) B. A. Cunningham, J. L. Wang, M. J. Waxdal, and G. M. Edelman, J . Biol.
Chem., 250, 1503-1512 (1975).
(272) K. D. Hardman, M. K. Wood, M. Schiffer, A. B. Edmundson, and C. F. Ainsworth,
Proc. Natl. Acad. Sci. U.S.A., 68, 1393-1397 (1971).
(273) K. D. Hardman, M. K. Wood, M. Schiffer, M. Edmundson, and C. F. Ainsworth,
Cold Spring Harbor Symp. Quant. Biol., 36,271-275 (1971).
(274) K. D. Hardman, Adu. E x p . Med. Biol., 40, 103-123 (1973).
(274a) J. Greer, H. W. Kauhan, and A. J. Kalb,J. Mol. Biol., 48, 365-366 (1970).
(275) J. W. Becker, G. N. Reeke, Jr., J. L. Wang, B. A. Cunningham, and G. M. Edelman,
J . Biol. Chem., 250, 1513-1524 (1975).
(276) G. N. Reeke, Jr., J. W. Becker, and G. M. EdelmanJ. Biol. Chem., 250,1525-1547
(1975).
(277) F. A. Quiocho, G.N. Reeke, Jr., J. W. Becker, W. N. Lipscomb, and G. M. Edelman,
Proc. Natl. Acad. Sci. U.S.A., 68, 1853-1857 (1971).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 153

104,000. The single polypeptide-chain protomers (however, see later)

are compactly folded, forming dome-shaped structures approximately
4.2 nm high x 4.0 nm wide x 3.9 nm thick. Each subunit contains one
Mn2+,one Ca2+,and one carbohydrate binding-~ite,'~~,~~~-~~~ believed,
until recently, to be situated in a prominent cavity in the molecular
Two such subunits are joined with their flattened bases prox-
imal, to form roughly ellipsoidal domes. The dimers pair across one
another, to form tetramers as depicted in Fig. 2.

FIG.2.-Schematic Representation of Con A T e t ~ a r n e r[Ca,

.~~~ Mn, and S indicate
the position of the Ca2+,Mn2+,and carbohydrate-specific binding-sites, respectively.
I localizes the position of the hydrophobic binding-site present in crystals of the I222
space group (reproduced by permission of Nature).]

The most striking feature of the con A secondary structure is its high
content (>50%) ofp-pleated sheet, a finding supported by c.d. and 0.r.d.
A theoretical study of 0-sheet-breaking residues was
studies.281-285a
(278) J. Yariv, A. J. Kalb, and A. Levitzki, Biochim. Biophys. Acta, 165,303-305 (1968).
(279) G. H. McKenzie and W. H. Sawyer, J . Biol. Chem., 248,549-556 (1973).
(280) J. W. Becker, G. N. Reeke, Jr., and G. M. EdelmanJ. Biol. Chem., 246,6123-6125
(1971).
(281) R. Zand, B. B. L. Agrawal, and I. J. Goldstein, Abstr. Pap. Am. Chern. SOC.Meet.,
156, BIOL-132 (1968).
(282) C. Kay, FEBS Lett., 9,78-80 (1970).
(283) M. N. Pflumm, J. L. Wang, and G. M. Edelman,J. Biol. Chem., 246,4369-4370
(1971).
(284) R. Zand, B. B. L. Agrawal, and I. J . Goldstein, Proc. Natl. Acad. Sci. U.S.A., 68,
2173-2176 (1971).
(285) W. D. McCubbin, K. Oikawa, and C. M. Kay, Biochem. Biophys. Res. Commun.,
43,666-674 (1971).
154 IRWIN J. GOLDSTEIN AND COLLEEN E . HAYES

conductedzE6 on con A. One extensive, @-pleatedsheet forms almost the

entire back surface of the protomer, and is responsible for most of the
noncovalent interactions bonding two protomers to form dimers, and
two dimers to form tetramers. There are two additional regions of
@-structure;the rest of the protein exists as random coil. The interested
reader is referred to extensive discussionsz6za~26E~~6g~z74~z76~zE7on the
molecular structure of con A.
A second remarkable structural feature is the presence in native con
A of several molecular species: an intact subunit of molecular weight
25,500 containing 237 amino acids, and three fragments of the subunit
termed Al, A‘,, and A2 (Ref. 263), B1, B2, and B2a (Ref. 288),and I, 11,
and I11 (Ref. 264). Disc gel-electrophoresis in dodecyl sodium sul-
fate,263,2EE
gel filtration in guanidiniuni chloride44 propanoic 8
M urea,zEEa or 40% acetic acid,zE6and ion-exchange chromatography on
Amberlite CG-50 (Ref.264) or D E A E - c e l l u l ~ sallowed
e ~ ~ ~ separation of
these species (compare, Ref. 28813). By end-group and amino acid
analyses of the fragments, it was found that the fragments are all de-
rived from the intact, polypeptide subunit; Al and A2 result from cleav-
age ofthe polypeptide chain between residues 118 (Asn)and 119 (Ser),
and At1 (present in low proportion) is derivedz63from A,. Tryptic,
peptide maps confirmed these Interestingly, there are no
obvious differences between the three-dimensional structures or
biological activities of con A tetramers made up entirely of intact sub-
units and of tetramers containing fragment^,^^^,"^ However, there are
some differences in the physical and chemical properties of the two
species. For example, incubation of native con A in 1% ammonium
hydrogencarbonate at 37” results in the formation of a precipitate that
contains the intact subunit and the naturally occurring fragments of the
molecule, whereas the solution phase contains only the intact sub-
A similar fractionation may be achieved by gel filtration on
Bio-Gel P-100 (Refs. 279 and 290) and by gradient elution from
Sephadex with D-g~ucosezE9; the fragmented molecules are eluted at

(285a) W. D. McCubbin, K. Oikawa, and C. M.Kay, F E B S Lett., 23, 100-104 (1972).

(286) E. A. Kabat and T. T. Wu, Proc. Natl. Acad. Sci. U.S.A., 70, 1473-1477 (1973).
(287) K. D. Hardman and I. J. Goldstein, in “Immunochemistry of Proteins,” M. Z.
Atassi, ed., Plenum Press, New York, 1977, Vol. 2, pp. 373-416.
(288) Y. Abe, M. Iwabuchi, and S.-I. Ishii, Biochem. Biophys. Res. Commun., 45, 1271-
1278 (1971).
(288a) M. 0. J. Olson and I. E. Liener, Biochemistry, 6, 3801-3808 (1967).
(289) B. A. Cunningham, J. L. Wang, M. N. Pflumm, and G. M. Edelman, Biochemistry,
11,3233-3239 (1972).
(290) W. H. Sawyer, S. Hammarstrom, G. Moller, and I. J. Goldstein,Eur.J.Immunol., 5,
507-510 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 155

lower concentrations of D-glucose than the intact molecules. The origin

of the fragments is unknown, but they could result from proteolytic
cleavage prior to, or during, the isolation.z63Moving-boundary elec-
trophoresisZ6' and isoelectric-focusing studies290aalso revealed the
heterogeneity of purified con A. An isoelectric point of 4.5-5.5 was
obtained by isoelectric f o c u ~ i n g , ~
electrophoresis
~~a on cellulose ace-
tate,290acataphoresis,lo2and the minimum-solubility method,z90b in con-
trast to pZ 7.1 k O . 1 obtained by free-boundary electrophoresisz6zand
electrophoresis on cellulose acetate.z62
The availability of the intact subunits was a prerequisite to sequence
determination. Tryptic and chymotryptic digestion ofthe intact subunit
produce both fragments of high molecular weight and smaller pep-
tides,zE9neither of which are the same as the naturally occurring con A
fragments. Three fragments (FlyFZ, and F3)were generated by degrada-
tion with cyanogen b r ~ m i d eas, was ~ be
~ ~ to ~ expected
~ ~ ~ from a polypep-
tide containing two methionyl residues. The order ofthe three peptides
was established by end-group analysis, and confirmed by sequence
analysis of the two methionine-containing, overlapping peptides iso-
lated from tryptic digests.z91Sequence analysis of the F, and Fz pep-
tidesZ7O and the F3peptideZ7* gave the complete sequence of 237 amino
acids.
As already indicated, each con A subunit contains one Caz+and one
Mnz+ ion. Both metals are needed for carbohydrate-binding, and,
hence, for all of the biological activities of the protein (compare Ref.
292), a dependence first established by Sumner and and
later verified by several investigators.'73~z7E~z93~z94The metals may be
removed from the protein by acidification with 0.1M hydrochloric acid
and dialysis against distilled water,zsgor by dialysis against 1.0M acetic
acidszs3Binding of Mnz+ to demetallized con A must precede Caz+
binding,173,z94the Mnz+being bound to a specific site, S1. It appears that
the binding of the transition-metal ion induceszg4the formation of a
specific Ca2+ion-binding site (S2). Shoham and showed

(290a) G. Entlicher, J. V. KoStii, and J. Kocourek,Biochim. Biophys. Acta, 236,795-797

(1971).
(290b) F. A. Czonka, J. C. Murphy, and D. B. Jones,]. Am. Chem. Soc., 48, 763-768
(1926).
(291) M . J. Waxdal, J. L. Wang, M. N . Pflumm, and G. M. Edelman, Biochemistry, 10,
3343-3347 (1971).
(292) C. F. Brewer, D. M. Marcus,A. P. Grollman, and H. Sternlicht,]. Biol. Chem., 249,
4614-4616 (1974).
(293) B. B. L. Agrawal and I. J. Goldstein, Can. J. Biochem., 46, 1147-1150 (1968).
(294) A. J. Kalb and A. Levitzki, Biochem. J., 109,669-672 (1968).
(295) M . Shoham, A. J. Kalb, and I. Pecht, Biochemistry, 12, 1914-1917 (1973).
156 IRWIN J, GOLDSTEIN AND COLLEEN E. HAYES

that the Ca2+site is specific; it binds Caz+or Cd2+,but not Ba2+,transi-

tion metals, or Sm3+.Ni2+ may be substituted for Mn2+without any
substantial change in carbohydrate-binding activity of the l e ~ t i n . ~ ~ ~
Mn2+and Ca2+appear to stabilize the conformation of the subunit; in
their absence, the four carboxyl groups clustered at the metal binding-
sites would strongly repel one another.274For this reason, it is, perhaps,
not surprising that the demetallized apoprotein cannot be crystallized
in the same space-group.2g6A U.V. difference-spectrum resulted when
Mn2+and Ca2+were added to the metal-free protein.2g7Potentiometric
titration, and chemical modification, of the demetallized protein indi-
cated that two histidine residues, pK, 6.8, are ligands for the tran-
sition-metal ion.2g8Titration of Caz+ion indicated involvement of two
carboxyl groups (pK, 4.2).2g9 The X-ray diffraction data revealed that the
two metal ions are -450 pm apart.262a*27s Edelman and coworker^^^^.^^^
identified Glu 8, Asp 10, Asp 19, His 24, and 2 water molecules as
coordinate ligands for Mn2+;and Asp 10, two water molecules, and the
backbone carbonyl groups of Tyr 12, Asn 14, and Asp 19 as ligands for
Ca2+.Hardman and AinsworthZBza and HardmaP4 obtained the same
results for Mn2+,but reported only one water molecule and Asp 208 in
addition to the four amino acid side-chains reported by Edelman’s
group. Con A prepared by Sephadex affinity-chromatography may not
contain its full complement of metal ions, giving rise to forms that differ
in their carbohydrate-binding ability.300,301 This deficiency may be al-
by addition of Mn2+and Ca2+.Gadolinium (Gd3+)binds to
a site (S3) distinct from the Mn2+( S l ) and Ca2+(S2)binding sites.
Alter and his showed that binding of Mn2+to con A is
cooperative in the presence, and noncooperative in the absence, of
Ca2+;the degree of cooperativity increases with pH. A kinetic study
of the binding of Mn2+and Ca2+to con A over the pH range of 5.3 to
6.4 has been conducted.304b
(296) A. Jack, J. Weinzierl, and A. J. Kalb,J. Mol. Biol., 58, 389-395 (1971).
(297) R. J. Doyle, D . L. Thomasson, R. D. Gray, and R. H. Clew, FEBS Lett., 52,185-187
(1975).
(298) G. Gachelin and L. Goldstein, FEBS Lett., 26,264-266 (1972).
(299) G. Gachelin, L. Goldstein, D. Hofnung, and A. J. Kalb, Eur. J . Biochem., 30,
155-162 (1972).
(300) T. Uchida and T. Matsumoto, Biochim. Biophys.Acta, 257,230-234 (1972).
(301) B. Karlstrom, Biochim. Biophys. Acta, 329,295-304 (1973).
(302) A. D. Sherry and G. L. Cottom, Arch. Biochem. Biophys., 156,665-672 (1973).
(303) A. D. Sherry, A. D. Newman, and C. G. Gutz, Biochemistry, 14,2191-2196 (1975).
(304) B. H. Barber, B. J. Fuhr, and J. P. Carver, Biochemistry, 14,4075-4082 (1975).
(304a) G. M. Alter, E. R. Pandolfino, D. J. Christie, and J. A. Magnuson, Biochemistry,
16,4034-4038 (1977).
(304b) R. D. Brown 111, C. F. Brewer, and S. H. Koenig, Biochemistry, 16,3883-3896
(1977).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 157

Koenig and coworkers30sreported the dependence on magnetic field

and temperature of solvent proton spin-lattice relaxation-rates for zinc
and manganese derivatives of con A in solution at pH 5.6; they con-
cluded that there is one rapidly exchanging water molecule ligand on
the Mn2+ion (residence lifetime, 2.5 psec at 25") and, further, that
monosaccharide binding to manganese-con A has little effect on the
relaxation rates of solvent protons. This suggests that sugars do not bind
directly to the transition metal. Essentially similar results were ob-
tained by Villafranca and Viola.306
Con A has also been isolated by Sephadex affinity-chromatography
from Canavalia gladiata306a,b*c and Canavalia maritima30ecbeans. Both
proteins were immunologically indistinguishable306C from C. ensifor-
mis con A. Interestingly, whereas C. ensiformis and C . gladiata con A
contain two metliionine residues per protein subunit, only one
methionine residue is found in the C. maritima l e ~ t i n Analysis
.~~~~
indicated that the methionine residue at position 129 in C. ensiformis
con A is from C. maritima con A.

b. The Concanavalin A-Saccharide Binding-site.-In 1971, Becker

and coworkersza0reported locating the con A saccharide-binding site.
They infused o-iodophenyl P-D-glucopyranoside, a glycoside inhibitor
of con A-levan precipitation,lEO into lectin crystals (1222)and analyzed
the crystalline, lectin-glycoside complex by X-ray crystallography. Dif-
ference electron-density projections at a resolution of 280 pm posi-
tioned the iodine atoms directly within the prominent surface cavities
of the protein (see Fig. 2). The site identified was 350-600 pin wide,
-750 pm high, and 1.8 nm deep. The iodine atom was deeply buried in
the cavity; the glycopyranoside ring was, assumedly, situated between
the iodine atom and the surface, with the anomeric carbon atom to-
wards the interior of the protein.267*268~280 Becker and coworkers also
speculated that, inasmuch as con A binds nonreducing termini of
polysaccharides, in order to bind at the same site as o-iodophenyl
P-D-ghcopyranoside, the polysaccharide terminus would have to be
oriented in the reverse direction, with the anomeric carbon atom to-
wards the surface of the protein, and the polysaccharide extending out
into the solution.268,280 In the case of D-glucose, having equatorial

(305) S. H. Koenig, R. D. Brown, 111, and C. F. Brewer, Proc. Natl. Acad. Sci. U.S.A., 70,
475-479 (1973).
(306) J. J. Villafranca and R. E. Viola, Arch. Biochem. Biophys., 165, 51-59 (1974).
(306a) S. Nakaniura and R. Suzuno, Arch. Biocheni. Biophys., 111,499-505 (1965).
(306b) H . Akedo, Y. Mori, Y. Tanigaki, K. Shinkai, and K. Morita, Biochim. Biophys.
Act(/., 271, 378-387 (1972).
(306c) D. R. Hague, Platit Physiol., 55, 636-642 (1975).
158 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

groups at C-2, C-3, and C-4, reversal would leave these three hydroxyl
groups in the same relative position. Each particular saccharide would
bind in one or the other orientation, depending on the restrictions
imposed by the site and the configurational features of the specific
sugar.z68*280
Brewer and coworkers studied the binding of I3C-enriched methyl a-
and /3-D-glucopyranoside to con A by carbon-13 nuclear magnetic
resonance s p e ~ t r o s c o p yAddition
. ~ ~ ~ ~ of
~~ Zn-con
~ A or Mn-con A to the
D-glucoside resulted in line broadening in the I3C-n.m.r. spectrum of
the sugar, without detectable changes in the chemical shifts of the
carbon resonances. [13C]Sugarcarbon-13 spin-lattice relaxation-times
(TJ, measured in the presence and absence of the two protein-
transition metal derivatives, was used to calculate the distance be-
tween Mnz+and each bound, sugar carbon atom. The results indicated
that methyl a-D-ghcopyranoside binds (to the protein) in the 4C, con-
formation, with C-3 and C-4 closest to the Mnz+,at a mean distance of 1
nm. This value contrasts with 2 nm proposed from the results of X-ray
crystallographic s t ~ d i e s Different
. ~ ~ ~ binding-orientations
~ ~ ~ ~ ~ ~ ~ ~for
methyl a-and /I-D-glucopyranosidewere postulated, in which the 2-, 3-,
and 4-hydroxyl groups of the p anomer bind at positions occupied by
the 6-, 4-, and 3-hydroxyl groups of the a anomer, respectively.30s
Two further studies, one by Villafranca and Viola,309who carried out
I3C spin-lattice relaxation measurements on the interaction of methyl
a-D-glucopyranoside with con A, and the other by Alter and Mag-
n u ~ o n , ~ ~who
O characterized the binding of 2-deoxy-2-
(trifluoroacetamido)-D-glucose to the lectin, placed the saccharide at
1.0 to 1.4 nm from the Mnz+ion, in agreement with the results of Brewer
and coworker^.^^^^^^^ Villafranca and Viola’s mode1309placed C-1, C-2,
and C-6 closest to the Mnz+(see Fig. 3). On the other hand, support for
the cavity proposed by Becker and coworkerszs0comes from p.m.r.-
spectral measurements of the longitudinal relaxation-times of the
methyl proton of methyl a-D-mannopyranoside with Znz+-and Gdz+-
substituted con A.311*31z

(307) C. F. Brewer, H. Sternlicht, D. M. Marcus, and A. P. Grollman, Proc. Natl. Acad.

Sci. U.S.A.,70, 1007-1011 (1973).
(308) C. F. Brewer, H. Sternlicht, D. M. Marcus, and A. P. Grollman, Biochemistry, 12,
4448-4457 (1973).
(309) J. J. Villafranca and R. E. Viola, Arch. Biochem. Biophys., 160, 465-468 (1974).
(310) G. M. Alter and J. A. Magnuson, Biochemistry, 13,4038-4045 (1974).
(311) B. H. Barber, A. Quirt, and J, P. Carver, Adu. E r p . Med. B i d , 55, 325 (1975).
(312) B. J. Fuhr, B. H. Barber, and J. P. Carver,Proc. N a t l . Acad. Sci. U.S.A.,73,322-326
(1976).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 159

FIG.3.-Distances from the Con A Mn2+Site to the Seven Carbon Atoms of Methyl
cu-DGlucopyranoside.811(Reproduced by permission of Arch. Biochem. Biophys.)

In an important paper, Hardman and Ainsworth313showed that con A

crystals contain a cavity that binds a number of relatively nonpolar
molecules (methyl p-hydroxybenzoate, phenyl phosphate,
o-iodobenzoic acid, o-iodoaniline, and dimethylmercury) and, most
important, o-iodophenyl P-D-galactopyranoside. As a noninhibitor of
the con A system, o-iodophenyl P-D-galactopyranoside serves as a con-
trol substance for o-iodophenyl P-D-glucopyranoside binding. Binding
of the galacto analog and several unrelated nonpolar substances at a
position identical with the purported D-glucoside binding-site strongly
suggested that this cavity is not the specific carbohydrate-binding site
of the jack-bean lectin. Rather, it appears to be a nonspecific bind-
ing site for hydrophobic molecules (perhaps lipid).314Interestingly,
native crystals of con A cracked and dissolved in methyl
a-D-mannopyranoside of 21 mM concentration, and in 5 mM
o-iodophenyl P-D-glucopyranoside, but not in 5 mM o-iodophenyl
P-D-galactopyranoside (which does not bind to the protein in solu-
ti~n).~ These
l~ results suggested that molecules binding at the
carbohydrate-specific site either sterically interfere with the crystal-
lattice packing, or produce in the protein conformational changes
which do so. Conformational changes may occur as a result of substrate
binding to con A, as monitored by U.V. difference spectroscopy,

(313) K. D. Hardman and C. F. Ainsworth, Biochemistry, 12,4442-4448 (1973).

(314) W. G. Jaffk and A. Palozzo, Actn Cient. Venez., 22,102-105 (1971).
(315) R. J. Doyle, E. P. Pittz, and E. E. Woodside, Cnrbohydr. Res., 8,89-100 (1968).
160 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

circular dichroism studies,283*285 n.m.r.-spectral measurements,316elec-

trophoretic mobility,317 and the use of a “reporter” group
[2-(bromomethyl)-4-nitrophenol].318 Alkali- and urea-induced confor-
mational changes in con A have also been studied.319The demonstra-
tion that ligands which bind to con A in solution [for example, methyl
a - ~ - m a n n o p y r a n o s i d e , 2MnZ+
~ ~ ~ (Refs. and 321),and
~ ~ ~ * ~173,285,297,
~~
Ca2+ (Refs. 173, 285, 297, 316, and 321)] stabilize and protect the
protein against aggregation,320heat denaturation,318enzymic digestion
(pronase),318precipitation of fragments during incubation289at 3 T , pH
7.9,and the time-dependent change in molecular weight observed in
the absence of ligand,266s318 also supports the possibility of ligand-
induced, conformational change.
Further evidence that there may be fundamental differences be-
tween the carbohydrate-binding properties of con A in solution and in
the crystalline state was provided by Bessler and coworkers174using
U.V. difference displacement spectroscopy. Certain ligands, such as
p-nitrophenyl a-D-mannopyranoside, bind to con A, thereby inducing
pronounced ultraviolet spectral changes. A second ligand binding to
the same site can displace the chromogenic ligand from con A, quench-
ing the difference spectrum. Bessler and coworkers found that all
saccharides inhibiting con A-polysaccharide interaction are able to
displace p-nitrophenyl a-D-mannopyranoside from con A, whereas no
quenching of the difference spectrum was noted with myo-inositol
(compare Ref. 322), o-iodophenyl P-D-galactopyranoside, and
p-nitrophenol, all of which are noninhibitors of con A-polysaccharide
i n t e r a ~ t i 0 n .The
l ~ ~ data were interpreted in terms of a single carbohy-
drate binding-site per monomeric unit for the protein in solution. In-
hibition indices of a number of glycosides were linearly related to their
association constants, and provided further support for a single,
carbohydrate-specific, b i n d i n g - ~ i t e . ~ ~ ~

(316)B. H. Barber and J. P. Carver, Can. J . Chem., 53,371-379 (1975).

(317)H. Akedo, Y. Mori, M. Kobayashi, and M. Okada, Biochem. Biophys. Res. Com-
mun., 49, 107-113 (1972).
(318)R. J. Doyle, S. K. Nicholson, R. D. Gray, and R. H. Glew, Carbohydr. Res., 29,
265-270 (1973).
(319)M. N. Pflumm and S. Beychok, Biochemistry, 13,4982-4987(1974).
(3u)) L. L. So and I. J. Goldstein,]. Biol. Chem., 242, 1617-1622 (1967).
(321)R. J. Doyle, D. L. Thomasson, and S. K. Nicholson, Carbohydr. Res., 46,111-118
(1976).
(322)K.D.HardmanandC. F.Ainsworth,Nature(London)New Blol., 237,54-55(1972).
(323)F.G. Loontiens, J. P. Van Wauwe, and C. K. De Bruyne, Carbohydr. Res., 44,
150-153 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 161

The carbohydrate-specific binding-site of con A has been identified

in crystals of the con A-methyl a-D-mannopyranoside (or its 2-deoxy-2-
iodo derivative) ~ o m p l e ~ It . ~was
~ ~necessary
* ~ ~ ~to* obtain
~ ~ ~a new
crystalline form of the carbohydrate-con A complex, as protein crystals
of the I222 space group cracked and dissolved upon addition of such
sugar derivatives as methyl a-D-mannopyranoside. Growth of the new
crystalline form was obtained by incubating con A with a large excess of
methyl a-D-mannopyranoside, followed by replacement of the
D-mannoside with o-iodophenyl P-D-glucopyranoside or methyl
2-deoxy-2-(2-iodoacetamido)-c~-~-glucopyranoside.32~
The gross conformations of the monomers, dimers, and tetramers in
the new space group C222, were found very similar to those of the
carbohydrate-free protein. The carbohydrate binding-site at 600-pm
resolution was 700 pm and 1.1nm from the Ca2+and Mn2+,respec-
t i ~ e l yThese
. ~ ~ ~results are in good agreement with those obtained by
n.m.r. s p e ~ t r o s c o p y . The
~ ~ ~amino
- ~ ~ ~acyl side-chains in closest prox-
imity to the bound carbohydrate appear to be324Tyr 12 and 100, Asp 16
and 208, Asn 14, Leu 99, Ser 168, and Arg 228. The finding of Asp 16 and
208 corroborates evidence from chemical modification, and titration
(see later).326Tyrosyl groups have similarly been implicated.327This
carbohydrate-specific site is 3.5nm from the iodophenyl (hydrophobic)
site originally proposed as the carbohydrate binding-site.280Essentially
similar results were obtained by Becker and coworkers32gin structural
studies on demetallized, inactive con A, and on glutaraldehyde cross-
linked crystals (space group 1222) complexed with methyl 2-deoxy-2-
iodo-a-D-mannopyranoside.Removal of bound Mn2+ and Caz+ from
native con A resulted in a loss of carbohydrate-binding activity,
suggesting an association between the metal-binding and saccharide-
binding sites.325Indeed, interpretation of electron-density maps lo-
calized significant structural differences between native con A and
metal-free con A (compare Ref. 296) in the vicinity of the
carbohydrate-binding site.325Studies at higher resolution will un-
doubtedly provide a more definitive structural solution to con
A-saccharide interaction.
Efforts to identify the amino acyl residues involved in

(324) K. D. Hardman and C . F. Ainsworth, Biochemistry, 15, 1120-1128 (1976).

(325) J. W. Becker, G. N. Reeke, Jr., B. A. Cunningham, and G. M. Edelman,Nature, 259,
406-409 (1976).
(326) G. S . Hassing, I. J. Goldstein, and M. Marini, Biochim. Biophys. Acta, 243,90-97
(1971).
(327) R. J. Doyle and 0. A. Roholt, Life Sci., 7 , 841-846 (1968).
162 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

carbohydrate-con A interaction have been r e p ~ r t e d . ~ These

~ ~ - in-
~~~
clude chemical-modification s t ~ d i e s , 3 ~ “affinity ~ ~ ~ labelling at-
tempts,326,332*333
and hydrogen-ion titration in the presence and absence
of methyl a-~-mannopyranoside.~~~ Approximately eight carboxyl
groups (pK, 3.8) per 100 kg of protein (2 carboxyl groups/subunit) are
not titratable when methyl a-D-mannopyranoside is present.326Fur-
thermore, the same glycoside protected con A carboxyl groups from
amidation with g l y ~ i n and
e~~ n i~t r o t y r ~ s i n emethyl
~ ~ ~ esters. Dean and
Homer334 measured the quenching of the fluorescence of
4-methylumbelliferyl a-D-mannopyranoside by con A as a function of
pH, thereby implicating a carboxyl group (pK, 3.5)in the binding
phenomenon. Carboxyl groups have also been implicated in the bind-
ing of saccharides to wheat-germ agglutinin.335
Treatment of con A with sodium acetate and acetic anhydride acety-
lated 84% of the amino groups and 31% of the phenolic The
modified protein was immunochemically indistinguishable from na-
tive con A, but could not be absorbed to S e p h a d e ~ even , ~ ~ ~in the
presence of 1 mM MnZf.Although retaining only-50% of its activ-
acetylated con A exhibited unaltered saccharide-binding spec-
ity,32s*330
ificity. Acetylation and succinylation reportedly converted tetrameric
into dimeric con A, thereby altering its biological properties.336These
results suggested that free amino groups and many tyrosyl residues are
neither directly involved in carbohydrate binding nor in maintaining
the structural integrity of the protein. Like acetylation, derivatization of
con A with maleic anhydride gave a stable, dimeric species having an
essentially unchanged binding-~ite.~~’ Maleylated con A required the
presence of high concentrations of salt to precipitate d e ~ t r a n .Doyle ~~]
and RoholPz7observed that reaction of con A with tyrosyl-modifying
reagents (tetranitromethane, N-acetylimidazole, and iodine) de-
creased, but did not abolish, glycogen-precipitating ability. Hassing
and G o l d ~ t e i obtained
n ~ ~ ~ similar results with N-acetylimidazole. The
combined evidence suggested that “accessible” tyrosyl residues prob-
(328) B. B. L. Agrawal, I. J. Goldstein, G. S . Hassing, and L. L. So, Biochemistry, 7,
4211-4218 (1968).
(329) G. S. Hassing and I. J. Goldstein, Eiochim. Biophys. Acta, 271,388-399 (1972).
(330) R. J. Doyle and D. C. Birdsel1,J. Bocteriol., 109, 652-658 (1972).
(331) N. M. Young, Biochim. Biophys. Acta, 336, 46-52 (1974).
(332) M. Beppu, T. Terao, and T. Osawa,]. Eiochem. (Tokyo), 78, 1013-1019 (1975).
(333) A. R. Frstser, J. J. Hemperly, J. L. Wang, and G. M. Edelman, Proc. Natl. Acad. Sci.
U.S.A.,73, 790-794 (1976).
(334) B. R. Dean and B. R. Homer, Biochim. Biophys. Acta, 322, 141-144 (1973).
(335) R. H. Rice and M. E. Etzler, Biochemistry, 14,4093-4099 (1975).
(336) G. R. Gunther, J. L. Wang, I. Yahara, B. A. Cunningham, and G. M. Edelman, Proc.
Natl. Acad. Sci. U.S.A., 70, 1012-1016 (1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 163

ably do not contribute significantly to stabilization of the

carbohydrate-protein complex. Other amino acid modifications gave
ambiguous results, or denatured the protein.329The effect of chemical
modification on the circular dichroism spectrum has been reported.285a
Fluorescence studies indicated that neither the specific binding of
methyl a-D-mannopyranoside nor alternation of the con A quaternary
structure changed the number or accessibility of the solvent-exposed
tryptophan residues in the l e ~ t i n . ~ ~ ‘
Affinity-labelling experiments employing methyl 2-deoxy-2-(p-
diazobenzamid0)-a-D-glucopyranosideand p-diazophenyl glycosides
of D-glucose, D-mannose, and maltose failed to label con A carbo-
hydrate-binding sites specifically.329However, partial success was
achieved by using a photoaffinity-labelling ligand, p-azidophenyl
a - ~ - m a n n o p y r a n o s i d eThe
. ~ ~lack
~ ~ ~ of
~ ~success in these experiments
can probably be attributed to the low association-constants of the
affinity-labelling reagents.
Kinetic parameters for the binding of p-nitrophenyl a-D-manno-
pyranoside, 3383339 methyl a-D-mannopyrano~ide,~~~ and the anomers
of methyl D-gl~copyranoside~~~ to con A have been reported. Con A
apparently binds its ligands of low molecular weight more slowly than
do other proteins. Specifically, most second-order rate-constants for
protein-ligand complexing are 107-108 sec-I M - l , whereas the con-
stants for the binding of con A to p-nitrophenyl a-D-mannopyranoside
or methyl a- and P-D-glucopyranosides are less than lo5 sec-’ M - l .
Brewer and postulated that complex-formation with con
A is not a simple, one-step, reversible reaction, but involves a ligand-
induced, conformational change such as that described by equation 1 ,
wherein PD, and PD, represent different conformers of the complex.

kl k2
P+D$PD,ePD,
k-2

Whereas a simple, one-step pathway for complex-formation requires

that, when one reactant is in large excess, the observed pseudo-first-

(337) R. Pelly and P. Horowitz, Biochim. Biophys. Acta, 427, 359-363 (1976).
(338) D. Gray and R. H. Glew,J. Biol. Chem., 248,7547-7551 (1973).
(339) S. D. Lewis, J. A. Shafer, and I. J . Goldstein, Arch. Biochem. Biophys., 172,
689-695 (1976).
(340) J. J. Grimaldi and B. D . Sykes,J. Biol. Chem., 250,1618-1624 (1975).
(341) C. F. Brewer, D. M . Marcus, A. P. Grollman, and H. Sternlicht, in “Lysozyme
Proceedings of the Lysozyme Conference,” E. Osserman, S. Beychok, and R.
Canfield, eds., Academic Press, New York, 1974, pp. 225-239.
164 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

order rate-constant (kobs) shall increase linearly with increasing excess

of reactant, the two-step pathway requires that kobs shall become inde-
pendent of [PI when k,[P] >> 1. Attempts by Lewis and coworkers339to
obtain such kinetic evidence for a two-step pathway were unsuccessful;
their results indicated that, if the two-step pathway obtains, k, <360 M-'.
Gimaldi and S y k e ~ ~observed
~O a substrate-induced, conformational
change in the binding of methyl a-wmannopyranoside to con A by using
stopped-flow n.m.r. spectroscopy. These workers340reported avalue for
k, of 6200 M-l. The data340also indicated that the rate of the conforma-
tional change is at least 100times lower with methyl a-D-mannopyrano-
side than the lower limit calculated for the corresponding confonnation-
a1change were the binding ofp-nitrophenyl a-D-mannopyranoside also
a two-step process. Unfortunately, these studies340were complicated by
the fact that metal binding and ligand binding were studied simultane-
ously. It is to be hoped that further studies of the kinetics of binding of
ligand to con A will reveal the nature of the interactions or conforma-
tional changes that accompany ligand binding, or both.

c. Preparation and Biological Activity of Mono-, Di-, and Tetra-

valent Concanavalin A.-Multivalence apparently accounts for many
biological properties of lectins, for example, cell agglutination, patch
and cap formation, mitogenesis, and polymer precipitation. The
influence ofvalence on biological activity may be investigated by using
lectins of various valences. For example, tetravalent, lima-bean lectin
is a more potent hemagglutinin, mitogen, and glycoprotein precipitant
than the divalent s p e ~ i e s . ' Glutaraldehyde
~~,~~~ cross-linked, soybean
agglutinin343and polymeric tetra- and multi-valent l e ~ t i are, n ~ simi-
~ ~ ~
larly, more potent than the native (divalent) agglutinin.343
Con A has also been manipulated in order to probe the biological
effects of multivalence. The equilibrium between naturally occurring,
dimeric (bivalent) and tetrameric (tetravalent) con A is affected by pH,
temperature, and chemical derivatization. The dimeric form at pH s5
retains specific, carbohydrate-binding ~apacity."~ However, equilib-
rium dialysis experiments suggested firstly, that dimer and tetramer
differ slightly in binding affinity, and secondly, that dimers containing
a fragmented protomer possess only one b i n d i n g - ~ i t e ?(C.d.
~ ~ studies

(342) R. W. Ruddon, L. M. Weisenthal, D. E. Lundeen, W. Bessler, and I. J. Goldstein,

P ~ o cNatl.
. Acad. S C ~ U.S.A.,
. 71, 1848-1851 (1974).
(343) R. Lotan, H. Lis, A. Rosenwasser, A. Novogrodsky, and N. Sharon, Biochem.
Biophys. Res. Commun., 55, 1347-1355 (1973).
(343a) B. Schechter, H. Lis, R. Lotan, A. Novogrodsky, and N. Sharon,Eur.J.Immunol.,
6,145-149 (1976).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 165

also revealed small, conformational differences between intact and

fragmented species.344)Below pH 5, predominantly dimeric con A does
not precipitate polysaccharides.173~204~290~309 Moreover, dimers isolated
by gel filtration at physiological pH failed to precipitate glycogen,
although the mitogenic activity was equivalent to that of the t e t ~ a m e r . ~ ~ ~
(It is important to distinguish between the properties of lectin in solu-
tion and at a cell surface.29o)Dimers induced by lowered temperatures
did not agglutinate erythrocytes, although electron microscopy showed
they could bind soluble g l y ~ o p r o t e i n s . ~ ~ ~
The mitogenic activities of native (tetravalent) and succinylated or
acetylated (bivalent) con A were Whereas the protein
concentration-dependent, DNA synthetic response of mouse spleno-
cytes to native con A exhibited a sharp peak, succinyl con A stimulated
DNA synthesis over a broad range ofprotein c o n ~ e n t r a t i o nMixing
. ~ ~ ~ of
native con A with its dimeric, succinylated derivative at pH 4.5 resulted
in the formation of hybrid molecules. A dimer species consisting of
equimolar amounts of native con A protomer and its succinyl derivative
was isolated, and s h o ~ n to ~ ~ a molecular weight of50,OOO at both
~ have
pH 5 and pH 7.
Treatment of con A with trypsin reportedly yielded nonagglutinat-
ing, monovalent con A, which restored normal growth to tumor
(Py 3T3) Monovalent con A (operationally defined as
protease-treated, nonagglutinating con A) has been prepared, and used
indiscriminately by cell biologists without extensive characteriza-
tion.347,348 Preparation of monovalent con A fragments from demetal-
lized lectin has been reported.349Alternatively, partial blocking of the
carbohydrate-binding sites was achieved by using a combination of
succinylation and photoaffinity labelling (with p-azidophenyl a - ~ -
m a n n o p y r a n o ~ i d e ) . ~Affinity
~ ~ , ~ ~ chromatography
~,~~~ of the derivatized
protein yielded a fraction consisting of dimers having, at pH 5, one free
and one blocked saccharide-binding site. Aggregation and protomer-
exchange ~ ~ ~ ~ r at r pHe d 7. A~ peptide
~ ~ reportedly
, ~ ~ ~capable, ~ of~ ~

(344) W.H. Sawyer,G. H. McKenzie, and L. W. Nicho1,Aust.J. Biol. Sci., 27,l-6 (1974).
(345) C. Huet, M.Lonchampt, M. Huet, and A. Bernadac,Biochim. Biophys. Acta, 365,
28-39 (1974).
(345a) A. R. Fraser, J. L. Wang, and G. M. Edelman,J. Biol. Chem., 251,4622-4628
(1976).
(346) M. M. Burger and K. D. Noonan, Nature, 228,512-515 (1970).
(347) M. S. Steinberg and I. A. Gepner, Nature, 241, 249-251 (1973).
(348) P. M.Evans and B. M. Jones, Exp. Cell Res., 88,56-62 (1974).
(349) D. L.Thomasson and R. J. Doyle, Biochem. Biophys. Res. Commun., 67, 1545-
1552 (1975).
(350) M. Beppu, T. Terao, and T. Osawa,J. Biochem. (Tokyo), 79, 1113-1117 (1976).
166 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

inhibiting the hemagglutinating and mitogenic activities of native con

A was obtained by limited digestion of the lectin with t r y p ~ i n . ~ ~ ~
d. Interaction of Concanavalin A with Po1ysaccharides.-Sumner
and Howell noted that con A precipitated glycogen and yeast mannan,
and agglutinated starch granule^.'^^*^^^ F. Smith and his colleagues later
measured glycogen and yeast mannan by a turbidimetric a . ~ s a y ~ ~1~ * ~ " ;
mg of a given polysaccharide incubated with an aliquot of con A pro-
duced a turbid solution quantitated by A,. Various polysaccharides
were differentiated on the basis of their "glycogen value," the turbidity
ratio comparing one mg of the given polysaccharide to one mg of
rabbit-liver glycogen in the standard assay."7~121~3s44-359
Turbidimetry indicated that precipitate formation depended on the
concentration of the protein as well as on that of the polysaccharide, the
time-period of incubation, and the ratio of protein to polysac-
har ride.'^^,^"^^^^ Precipitin curves determined turbidimetrically dif-
fered from those determined by analysis of precipitated nitrogen.360For
these reasons, early turbidimetric studies conducted on crude (rather
than purified) con A are difficult to interpret. The quantitative precipi-
tin method, wherein increasing amounts of polysaccharide or glycopro-
tein are added to standard aliquots of con A, and the resulting precipi-
tate is washed, and analyzed for nitrogen, is now considered the
method of choice for studying protein-polysaccharide i n t e r a c t i ~ n . ' ~ ~ * ~ ~ ~
A typical precipitin-curve for the reaction320between con A and
dextran B-1355-Sis presented in Fig. 4. Analogous to antibody-antigen
precipitin curves, there are three zones: lectin excess (all added dex-
tran is precipitated), equivalence (virtually all dextran and lectin are
precipitated), and polysaccharide excess (soluble complexes are
formed). Optimal precipitation of con A by dextran B-1355-Soccurred
in 24 h at 2 9 , between pH 6.1 and 7.2, and was unaffected320by (buf-
fered) concentrations of sodium chloride of up to 4.2 M .
From precipitin studies on a large number of polysaccharides, Gold-
(351) D . S. Seidl, A. Palozzo, A. Levy, V. Azavache, M. Jaff6, and W. G. Jaff6,Experientia,
31,37-38 (1975).
(352) J. B. Sumner and D. J. O'Kane, Enzymologia, 12,251-253 (1948).
(353) J. A. Cifonelli and F. Smith, Anal. Chem., 27, 1639-1641 (1955).
(354) J. A. Cifonelli, R. Montgomery, and F. Smith,J. Am. Chem. Soc., 78,2488-2489
( 1956).
(355) J. A. Cifonelli and F. Smith,]. Am. Chem. Soc., 79,5055-5057 (1957).
(356) R. Montgomery, Arch. Biochem. Biophys., 67, 378-386 (1957).
(357) D. J. Manners and A. Wright,J. Chem. Soc., 4592-4595 (1962).
(358) 0. Kjolberg, D. J. Manners, and A. Wright, Comp. Biochem. Physiol., 8,353-365
(1963).
(359) E. E. Smith, Z. H. Gunja Smith, and I. J. Goldstein, Biochem. J . , 107, 715-724
(1968).
(360) R. D . Poretz and I. J. Goldstein, Immunology, 14, 165-174 (1968).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 167

- 500

Dextran Img)
FIG. 4.-Quantitative, Precipitin Curve of Dextran B-13553 with Con A.309 [The
total amount of dextran in the precipitate is also illustrated. Con A, 42 pg of nitrogen.
(Reproduced by permission of the Journal of Biological Chemistry.)]

stein and coworkers120,121

suggested that all branched polysaccharides
having multiple, terminal (nonreducing) a-D-glucopyranosyl, a-D-
mannopyranosyl, or D-fmctofuranosyl groups would precipitate with
con A. The “chain-end mechanism” postulated,120~121~16s~1s7*zo4~359
in
which tetravalent con A interacts with specific glycosyl residues of
polysaccharide or glycoprotein chain-ends, is diagrammed in Fig. 5.

Mullivalent
polysaccharide # Ea!er

FIG. 5.-Diagranimatic Representation of the Chain-end Mechanism of Con A-

Polysaccharide (Glycoprotein) Interaction.
168 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

This chain-end mechanism has been modified209*240 to account for the

observation that 2-0-substituted a-D-mannopyranosyl (and, probably,
2-0-substituted a-D-ghcopyranosyl) residues in internal regions of
polysaccharides and glycoproteins also interact with con A. Branched
a~~~g~ucans,102,120,121.2S3,320.3S3-35918d61.362 a-~~mannans,102,120,121,204,3S2-3S4,363

and precipitate with con A, whereas linear

polymers may bind, but will not precipitate the protein.'21*204*3s9*362
These polysaccharides are discussed in detail in the following Sec-
tions.
Con A-polysaccharide complex-formation is probably stabilized
largely by hydrogen bonding and charge-dipole interaction.16g~310*320*365
As charged sugars are not bound, it is unlikely that charge-charge
interactions are involved. Nonpolar forces have, however, not been ex-
cluded.
Nonspecific precipitation between con A and various polyelectro-
lytes (such as fucoidan, RNA, heparin and phosphorylated dextran,
starch, and glycogen) has been ~ b ~ e r v e d High . ~ ~concentrations
~ , ~ ~ ~ * ~ ~ ~ ~
of salt (1.0 M NaC1) inhibited the con A-heparin reaction, suggesting
that nonspecific, charge-charge interaction had been involved.
Complete inhibition of specific precipitation would be expected to
result from addition of competing sugar hapten, but not by high con-
centrations of salt.320However, the results of inhibition studies are not
always decisive; sugar haptens occasionally produce partial inhibition
of con A-polyelectrolyte precipitation, but higher concentrations of
saccharides are generally necessary than those required for complete
inhibition of con A-neutral glycan interaction^.^^^ Perhaps, specific
saccharide interaction with con A modifies the surface charge-distribu-
tion on the protein as a result of conformational changes (see earlier).
Nonspecific, hydrophobic forces might also be involved in lectin-
polysaccharide or, more likely, l e c t i n - g l y c o p r ~ t e i n ~interaction
~~J~~
(see Section II,l,j).
Preliminary information regarding certain structural features of

(361) M. E. Preobrazhenskayaand E. L. Rosenfel'd, Biokhimiya, 33,784-791 (1968).

(362) I. J. Goldstein, R. D. Poretz, L. L. So, and Y. Yang, Arch. Biochem. Biophys., 127,
787-794 (1968).
(363) M. E. Slodki, R. M. Ward, and J. A. Boundy, Biochim. Biophys. Acta, 304,449-456
(1973).
(364) B. A. Lewis, M. J. S. Cyr, and F. Smith, Carbohydr. Res., 5,194-201 (1967).
(365) R. D. Poretz and I. J. Goldstein, Biochemistry, 9,2890-2896 (1970).
(366) R. J. Doyle, E. E. Woodside, and C. W. Fishel, Biochem. I., 106,35-40 (1968).
(366a) V. Buonassisi and P. Colburn, Arch. Biochem. Biophys., 183,399-407 (1977).
(367) N. DiFerrante and R. Hrgovcic, FEES Lett., 9,281-283 (1970).
(368) R. J. Doyle and T.-J.Kan, FEES Lett., 20,22-24 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 169

polysaccharides can sometimes be obtained by determining the

binding strength between the two components of a con A precipitate.
To a first approximation, this is proportional to (1) the “solubility” of
the precipitate,320and (2) the proportion of a sugar hapten that is
needed in order to inhibit precipitation320 by 50%. These two
operationally defined parameters have proved extremely useful in
comparative studies of a series of related polysaccharides (for example,
dextrans) permitting some tentative conclusions regarding the extent
and nature of branching, and the number of glycosyl residues in
exterior, oligosaccharide side-chains (see later).

e. Interaction of Concanavalin A with Glycogen and

Amy1opectin.-As already mentioned, Sumner and Howell’s discovery
of con A-polysaccharide precipitation102eventually led to the defini-
tion of “glycogen value” (g.v.), determined t u r b i d i m e t r i ~ a l l y .F.
~~~*~~~
Smith and coworker^,^^^^^^^ noting that the extent of reaction depended
on the origin of the glycogen and that amylopectin did not react,
suggested that con A could be used to distinguish between these two
classes of branched a-D-glucans.
Manners and his ~ o l l e a g u e s reported
~ ~ ~ ~ ~ a~gross
* relationship be-
tween the g.v. and the degree of branching for many glycogen speci-
mens; they also found no reaction with amylopectin and amylopectin
@limit dextrin. On the other hand, sweet corn (Zea mays) phytoglyco-
gen gave a g.v. similar to those of animal glycogens, affirming the
structural relationship between these two groups of p o l y s a c ~ h a r i d e s . ~ ~ ~
Inasmuch as periodate abolished, whereas beta-amylase enhanced, the
reactivity of glycogen with con A, Smith and coworkers suggested117
that the protein probably interacted with “intact inner branches of the
glycogen molecule,” in addition3s7 to “hydroxyl groups of the
molecule.”
Goldstein and coworkerslZ1determined the optimal concentration of
lectin for precipitation, and showed that potato amylopectin did pre-
cipitate with the jack-bean lectin. In a turbidimetric study of the in-
teraction of con A with amylopectin, glycogen, and their enzymically
and chemically degraded products, E. E. Smith and coworkers3s9
confirmed that the extent of interaction of polysaccharides with con A is
highly dependent on the concentration of protein. For example, florid-
ean starch gave no turbidity with con A at 1 mg/ml, whereas yeast
mannan, dextran B-13554,and rabbit-liver glycogen all formed pre-
c i p i t a t e ~ .However,
~~~ with con A at 5 mg/ml, floridean starch
exhibited3sgthe same reactivity as dextran B-1355s.
The molecular weight of the polysaccharide also affects con A-
170 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

polysaccharide i n t e r a ~ t i o n .Chemically
~ ~ ~ . ~ ~ ~ or enzymically degraded
polysaccharides of decreased molecular weight generally reacted more
extensively with jack-bean lectin than their undegraded counter-
p a r t ~ .Assuming
~ ~ ~ * that ~ A binds nonreducing termini, increased
~ ~ con
polysaccharide reactivity could be a consequence of increasing, by
degradation,3ssthe number of chain ends per unit weight.
Glycogens extracted from rabbit liver by three different procedures
were investigated by quantitative precipitation (see Fig. 6): glycogen
having the lowest molecular-weight distribution, obtained by extrac-
tion with potassium hydroxide solution, attained maximal precipitation
of nitrogen at the lowest concentration of glycogen, followed by the
trichloroacetic acid-extracted and the water-extracted glycogens.369
Glycogen extracted with potassium hydroxide solution also inhibited
precipitation more effectively than the trichloroacetic acid-extracted
glycogen (the glycogen-excess region of the precipitin curve).36sThe
same pattern was also observed for preparations of levan and dextran of
differing molecular weight.369A particularly interesting feature of the
con A-glycogen precipitation reaction is the very broad range of max-
imal precipitation (compare the dextran B-1355-S precipitin curve in
Fig. 4); this distinctive type of curve was also given by corn (Zea mays)
phytoglycogen, but not by corn amylopectin.

1 I I 1 I I I 1 I I I I I I 1 . 1 1 I I I I I 1
0 0.4 0.8 1.2 1.6 20 2.4 2 8 4.0 6.0 8.0 10.0
Glycogen (mg 1
FIG. 6.-Quantitative, Precipitin Curves of Rat-liver Glycogen with Con A. (0,
Potassium hydroxide-extracted glycogen; A, trichloroacetic acid-extracted glyco-
gen; U, water-extracted glycogen. Con A, 48 pg of nitrogen. After Ref. 369.)

(369) L. L. So and I. J. Goldstein,J. Immunol., 102,53-57 (1969).

 LECTINS: CARBOHYDRATE-BINDING PROTEINS 171

The influence of the molecular weight and the polydispersity of the

polysaccharide preparation on the con A-polysaccharide interaction
was also studied by two-dimensional, agar-gel diffusion.
Homogeneous polysaccharides generally gave single, sharp, precipitin
bands with con A, whereas polydisperse preparations gave diffuse
bands of variable breadth. Polysaccharides and glycoproteins of low
molecular weight usually have high diffusion coefficients, and form
bands close to the con A well. On the other hand, species having high
molecular weight form precipitin bands close to the polysaccharide
we11.120,359~369
Quantitative studies with glycogens and amylopectins of
known structure and molecular weight are needed in order that a full
understanding of these interactions may be achieved.

f. Interaction of Concanavalin A with De~trans~~~.-Dextran-

con A interaction is of special significance, because the isolation of
this lectin by affinity chromatography on cross-linked dextran gels
( S e p h a d e ~ ) ' ~ *resulted
-'~~ from the discovery that con A precipitated
with dextrans. The availability of a large number of dextrans containing
a variety of different a-D-glucosidic linkages370also makes this an ideal
system for investigating structure-activity relationships.
Cifonelli and Smith3s5reported that, although dextran B-512 did not
form a precipitate with con A, a "dextrin"-derived dextran synthesized
by Acetobacter capsulatum did precipitate. Furthermore, con A
displayed a specificity similar to that of type XI1 pneumococcus rabbit
antibody in its differential reactivity with dextrans, especially those
purportedly containing a high proportion of a - ~ -1+=2)-glucosidic
(
bond^.^^^,^^^*^^' Validating this relationship, Suzuki and Hehre372
showed that the proportion of kojibiose isolated following acetolysis of
dextrans paralleled both the extent of type-XI1 cross reactivity and,
broadly speaking, the degree of precipitation253*,260*371~372 with con A.
Each of 23 dextrans studied by Goldstein and coworkerslZ1formed
precipitates with purified con A, although seven of them [all containing
>90% of a - ~ 1+6)-like
-( D-glucosidic linkages]370required high con-
centrations of lectin for precipitate formation. Similar results were
reported by Preobrazhenskaya and R ~ z e n f e l ' d ~these
~ ' ; findings were
also corroborated by gel-diffusion studies.120Formation of precipitate
with con A was approximately related to the content of non a - ~ -1(4 6 ) -
like linkages, determined from periodate-oxidation data.I2'
The dextran from Streptococcus bovis (previously considered to be

(370) R. L. Sidebotham, Adu. Carbohydr. Chem. Biochem., 30,371-444 (1974).

(371) E. J. Hehre, K a g a k u No Ryoiki, 9,454-455 (1965).
(372) H. Suzuki and E. J . Hehre, Arch. Biochem Biophys., 104, 305-313 (1964).
172 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

linear)373afforded a typical precipitin curve when levels of con A higher

than those required to precipitate dextran B-13554 were employed.362
Nigerose was isolated from this dextran, indicating the possible origin
of branch points.362Dextran B-512(F) and the dextran from Betacoccus
arabinosaceous (L. rnesenteroides), Birmingham strain, likewise gave
precipitin curves with high levels of con A; a chemically synthesized,
linear (1+6)-a-D-gl~can~~~ did not precipitate the protein.362From
these studies, it was suggested that all native dextrans are branched.362
The solubility of the dextran-con A precipitates at equivalence is a
revealing parameter. Compared to the dextran B-1355-con A complex,
which had320a solubility of 1.5 p g of N/ml, the solubility of S. bovis-,
dextran B-512(F)-, and B . arabinosaceous dextran-con A precipitates
were 28, 19, and 11pg of N/ml, respectively.362Similarly, 0.024, 0.03,
and 0.22 pmoles of methyl a-D-glucopyranoside were needed in order
to inhibit the precipitation reaction between con A and dextrans S.
bovis, B-512(F), and B . arabinosaceous, respectively.362These two
parallel parameters indicate the affinity of the dextran for con A.
Methylation studies showed that B . arabinosaceous dextran had one
branch point per 6 to 7 D-gh2osyl residues375;18% of all branches
consisted of a single D-glucosyl Thus, this dextran is approxi-
mately three times as ramified as B-512(F), that has one branch per 23
Not surprisingly, the con A-B-512(F) dextran precipitate is
more soluble than, and -7 times as readily inhibited by methyl a - ~ -
glucopyranoside as, the con A-B. arabinosaceous dextran precipitate.
By these criteria, S . bovis dextran is more closely related to B-512(F)
dextran than to B . arubinosaceous d e ~ t r a nOne
. ~ ~additional
~ parameter
that correlates well with structure is the maximum amount of con A
nitrogen precipitated by each of these dextrans: 28,39, and 83% of con
A added to the dextrans from S. bovis, B-512(F),and B . arabinosaceous,
respectively.362Torii and his colleagues378employed fractional precipi-
tation with con A to demonstrate the microheterogeneity of dextran
B-1397.
Dextran B-1355-S was adsorbed to three different forms of im-
mobilized con A (con A-agaro~e,~’~ poly(L-leucy1)-con A,379and
(373) R. W. Bailey, Biochem. J., 71, 23-26 (1959).
(374) E. R. Ruckel and C. Schuerch,J. Am. Chem. Soc., 88,2605-2606 (1966).
(375) S. A. Barker, E. J. Bourne, G. T. Bruce, W. B. Neely, and M. StaceyJ. Chem. Soc.,
2395-2399 (1954).
(376) E. J. Bourne, D. H. Hutson, and H. Weigel, Biochem. J., 86,555-562 (1963).
(377) J. W. Van Cleve, J. W. Schaffer, and C. E. Rist,J. Am. Chena. Soc., 78,4435-4438
(1956).
(378) M. Torii, K. Sakakibara, B. P. Alberto, and A. Misaki, Biochem. Biophys. Res.
Commun., 72,236-242 (1976).
(379) K. 0. Lloyd, Arch. Biochem. Biophys., 137, 460-468 (1970).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 173

glutaraldehyde-insolubilized con A),380whereas dextran B-512 was not

adsorbed. These results were explained in terms of dextran structure.
Corroboration for the chain-end mechanism of con A-polysaccharide
interaction was obtained when Duke and coworkers381showed that
dextran N-4, similar in structure to dextran B-512, lost its con A pre-
cipitating capacity after nonreducing (terminal) a-D-glucopyranosyl
groups had been converted into glucopyranosyluronic acid groups by
catalytic oxidation. Although it is probable that 2-O-substituted a-D-
glucopyranosyl residues (such as those found in the internal regions of
dextrans) would also interact with con A, this has not been demon-
~trated.~~~
Until complete, structural information is available for each of these
dextrans, it is not possible to relate dextran structure to con A reactivity.
However, such factors as polydispersity, molecular fre-
quency and nature of branching, type of a-D-glucosidic linkages, and,
very important, the number of glycosyl residues in the exterior chains,
all contribute to the con A reactivity of the polymer.362The capacity for
precipitate formation with con A is now virtually a standard procedure
in studies on dextran s t r u c t ~ r e . ~ ~ ~ . ~ ~ ~

g. Interaction of Concanavalin A with Mannans.-Yeast mannan

(gum) was among the first polysaccharides reported by Sumner and his
colleague^^^^^^^^ to precipitate with con A. Sumner’s results were
confirmed by Cifonelli and F. Smith; the interaction of con A with yeast
mannan was distinguished by the very high “glycogen values” ob-
tained in a turbidimetric a s ~ a y . ’ ~ This
’ ” ~ ~behavior was attributed to the
extensive branching in the molecule (34%), and, hence, to the large
number of chain ends.lZ1
The interaction between con A and a-mannans from a variety of
micro-organisms was studied by So and G o l d ~ t e i n The .~~~ mannans
from Saccharomyces cerevisiae, Saccharomyces rouxii, and Sarcina
sp., and the phosphomannan from Saccharomyces pini Y-2579, all gave
classical precipitin curves similar to those obtained for d e x t r a n ~ , 3 ~ ~ , ~ ~ ~
but unlike the extended curves given by glyc0gen.3~~ A synthetic, linear
(1+3)-a-D-mannopyranan did not precipitatezo4with con A.
The greater reactivity of a-mannans compared to other con A-reactive
polysaccharides was indicated by several observations: (1)the con A-

(380) E. H. Donnelly and I. J. Goldstein, Biochem.], 118, 679-680 (1970).

(381) J. Duke, I. J. Goldstein, and A. Misaki, Biochim. Biophys. Actu, 271, 237-241
(1972).
(382) E. E. Smith, FEBS Lett., 12,33-37 (1970).
(383) M. Kobayoshi, K. Shishido, T. Kikuchi, and K. Matsuda, Agric. B i d . Chem., 37,
357-365 (1973).
174 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

excess region of the precipitin curve rose more steeply than for any
other polysaccharide or g l y ~ o p r o t e i n(2)
,~~the
~ con A-mannan precipi-
tates exhibited the lowest “solubility” of any carbohydrate-containing
macromolecule (approximately 1pg/ml),204 (3) the con A-mannan pre-
cipitation reaction was more difficult to inhibit with carbohydrate hap-
ten than with any other con A-reactive p ~ l y m e r , ’and(4)
~ ~ * ~D-mannose
~~
derivatives were the most potent inhibitors of the con A
These observations are corroborated by evidence that con A contains a
specific-binding locus for the 2-hydroxyl group of D-mannose, whereas
the 2-hydroxyl group of D-ghCOSe actually constitutes a destabilizing

By utilizing the Ouchterlony technique of two-dimensional, agar-gel

diffusion, Slodki and his colleagues investigated the interaction of con
A with O-pho~phonomannans.~~~ They suggested a possible correlation
between the extent of precipitate formation with con A and the content
of a-D-( 1+2)-mannosidic linkages.
Chemically synthesized D-mannans have also been investigated.384
The D-mannan prepared by polymerization of 1,6-anhydro-P-~-
mannopyranose reacted vigorously with the jack-bean lectin, indicat-
ing that it was most probably a highly branched polymer having multi-
ple a-D-mannopyranosyl chain-ends. A mannodextran, prepared by
grafting D-mannose units onto dextran B-512, precipitated a higher
proportion of con A than the parent d e ~ t r a n . ~ ~ ~
Hapten-inhibition studies employing D-mannose-containing oli-
gosaccharides led to the important discovery that a - ~ 1-+2)-linked
-(
D-mannopyranosyl residues were exceptionally reactive with con A,
the di- and tri-saccharides [a-D-Manp-(1+2)-D-Man and a-D-Manp-
(1+2)-a-D-Manp-( 1-*2)-~-Man]being 4 and 20 times as potent as
methyl a - D - m a n n o p y r a n o ~ i d eThese
.~~~~findings
~~~ suggest that the
con A binding-site could be complementary to a sequence of three,
or four, a-D-( 1-*2)-mannopyranosyl r e s i d ~ e s .On ~ ~the
, ~other
~ ~ hand,
internal a - ~ 1+3)-
-( or a-D-( 1+6)-linked D-mannosyl residues do not
interact with con A. The implication of a-D-( 1+2)-mannopyranosyl
residues as con A binding-sites in animal glycoproteins adds further
significance to these studies.240
Fluorescein- and mercury-labelled con A386-388have been used to
demonstrate the presence of D-mannan in yeast c e l l - ~ a l l s and ~ ~ ~in~ ~ ~ ‘
(384) R. Robinson and I. J . Goldstein, Carhohydr. Res., 13,425-431 (1970).
(385) B. Lindberg and S. Svensson, Actu C h m . Scand., 24, 711-713 (1970).
(386) J. S. Tkacz, E. B. Cybulska, and J. 0. Lampen,j. Bacteriol., 105, 1-5 (1971).
(387) M. Horisberger, H. Bauer, and D. A. Bush, FEBS Lett., 18,311-314 (1971).
(388) H. Bauer, M. Horisberger, D . A. Bush, and E. Sigarlakie, Arch. Mikrobiol., 85,
202-208 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 175

bud scars388 of Saccharomyces cerevisiae. By employing a con

A-Sepharose column, Lloyd379demonstrated adsorption of yeast (Sac-
charomyces cerevisiae) mannan, and fractionation of a galactomannan
from Cladosporium werneckii.

h. Interaction of Concanavalin A with D-Fructans.-Precipi-

tate formation between p-D-fructans (levans) and con A was first
demonstrated by Goldstein and So'zo~3sg; levans from NRRL B-512 and
B-1662, and from Erwinia ananas, Bacillus megatherium, and Aero-
bacter levanicum, formed precipitin bands with con A by Ouchterlony,
two-dimensional, agar-gel diffusion. A levan from Leuconostoc mes-
teroides, strain C , likewise364precipitated con A. In a comprehensive
study of D-fructans, So and Goldstein obtained typical precipitin curves
between levans from Aerobacter levanicum No. 15552, Bacillus
megatherium KM, and Bacillus subtilis and the l e ~ t i n . In ~ contrast
~'~~~~
to dextran B-1355-S and a variety of a-D-mannans that precipitated
95-98% of the total con A added, these levans maximally pre~ipitated"~
only 70-78% of added con A. A relatively weak precipitin reaction was
generated between con A and the D-gluco-D-fructanfrom the Hawaiian
ti plant; at equivalence, only 41% of added con A was precipitated."'
The con A-levan (A. levanicum) precipitate, moreover, was more s o h -
ble than the con A-dextran B-1355-S (Ref. 320) or con A-S.
cerevisiae a-D-mannanZo4 precipitates. A further indication of the low-
ered reactivity of D-fructans compared to those of D-mannans, dextrans,
and glycogens was the relative ease with which sugar haptens inhibited
precipitation. Only 16% of the sugar hapten was needed to inhibit, by
50%, the A . levanicum levan-con A precipitation reaction that was
necessary for the con A-dextran B-135543 system.215
Hapten inhibition studies showedz1' that sugars containing the
D-arabinofuranoid ring system bind to con A. Thus, methyl p-D-
fructofuranoside, methyl a-D-arabinofuranoside, 2,5-anhydro-~-
mannitol, and 2,5-anhydro-~-glucitolall inhibited the con
A-polysaccharide system. Interestingly, although inulobiose and in-
ulotriose both inhibited the dextran-con A precipitation reaction, in-
ulin itself inhibited the con A-dextran system but did not form a
precipitate with the lectin."' Further discussion of the binding of
furanoid sugars to con A is included in Section II,l,l.

i. Interaction of Concanavalin A with Teichoic Acids.-Con A spe-

cifically precipitates with teichoic acids containing a-D-g~ucopyranosy~
and 2-acetamido-2-deoxy-a-~-glucopyranosyl end-groups, but not the
corresponding p-glycosyl groups. Furthermore, both the organisms
176 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

and the isolated cell-walls derived therefrom are agglutinated by the

lectin.
Reeder and E ~ k s t e dstudied
t ~ ~ ~ the interaction of con A with teichoic
acids from Staphylococcus aureus and Staphylococcus epidermidis by
gel diffusion,and precipitation in a fluid system. The teichoic acid from
strain T, of S . epidermidis contains a-D-glucopyranosyl residues, and it
precipitated with con A, whereas strain Tz, which is p-D-glucosylated,
did not. Classical precipitin curves resulted when con A interacted
with strains T, and 412 (also a-D-ghcosylated); the precipitation was
specifically inhibited by D-glucose, 2-acetamido-2-deoxy-~-glucose,
and methyl a-and /3-D-glucopyrano~ides.~~~
Con A also precipitated, in agar gel, with teichoic acids from strains
Copenhagen and 416 of S. aureus, but not with strains Foggie and
Smith.389From the work of Sanderson and coworkers, it is known that S .
aureus Copenhagen teichoic acids contain both a- and p-linked
2-acetamido-2-deoxy-~-glucosyl resid~es.~ Interaction
~ ~ ~ ~ ~ 'of these
teichoic acids with con A resulted in the precipitation of only the
a-D-linked species.389Thus, not only was a fractionation achieved, but
Reeder and Eckstedt also demonstrated that the a-and p-linked amino
sugars occurred on separate teichoic acid molecules, instead of the
anomers being present in a single chain.389Failure ofcon A to react with
teichoic acid from the Foggie and Smith strains of S . aureus confirmed
the presence of p-D-linked 2-acetamido-2-deoxy-~-glucosyl residues.
Previous determinations of a- or P-specificity involved either hapten
inhibition of precipitation with specific a n t i b ~ d y ,enzymic
~ ~ ~ , ~degra-
~~
d a t i ~ n , ~or~ O
optical rotatory dispersion
Archibald and Coapes observed that both the intracellular and mem-
braneous teichoic acids isolated from Streptococcus fecalis 8191 gave,
with con A in agar gel, sharp precipitin lines that dissolved when
methyl a-D-glucopyranoside was added.3g5Moreover, Lactobacillus
plantarum wall teichoic acid, which also contains a-Dglucopyranosyl
substituents, precipitated with con A, whereas the Staphylococcus epi-
dermidis 12 teichoic acid, containing P-Dglucopyranosyl substituents,
did not.395Additionally, these authors demonstrated395cell-wall agglu-
tination by con A.

(389) W. J. Reeder and R. D. Ekstedt,]. ImmunoZ., 106,334-340 (1971).

(390) A. R. Sanderson, J. L. Strominger, and S. G. Nathenson,J. B i d . Chem., 237,
3603-3613 (1962).
(391) A. R. Sanderson, W. G. Juergens, and J. L. Strominger, Biochem. Biophys. Res.
Commun., 5, 472-476 (1961).
(392) S. I. Morse,J. Exp. Med., 117, 19-26 (1963).
(393) M. Torii, E. A. Kabat, and A. E. Bezer,]. E x p . Med., 120, 13-29 (1964).
(394) E. A. Kabat, K. 0. Lloyd, and S. Beychok, Biochemistry, 8,747-756 (1969).
(395) A. R. Archibald and H. E. Coapes, Biochem. J., 123,665-667 (1971).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 177

Interaction between con A and cell-wall digests of Bacillus subtilis

168 resulted in insoluble complexes.330Interestingly, the precipitation
reaction exhibited two pH optima, at pH 3.1 and 7.4. Similar to specific,
neutral polysaccharide-con A interaction, the precipitation reaction
that was optimal at pH 7.4 was completely inhibited by methyl a - ~ -
mannopyranoside, whereas the precipitation reaction that peaked at
pH 3.1 was only partially inhibited. By using a purified, teichoic acid
preparation, virtually no inhibition with sugar hapten could be dem-
o n ~ t r a t e d Only
. ~ ~ ~the precipitation reaction optimal at pH 3.1 was
sensitive to increase in salt concentration, indicating that a charge-
charge interaction had probably occurred between positively charged
con A and negatively charged teichoic a ~ i d .Doyle ~ ~ and
~ *coworkers
~ ~ ~
utilized a con A-Sepharose column for the large-scale isolation of B .
subtilis 168 teichoic from 45 mg of autolyzate added to a con
A-Sepharose column, they recovered 21 mg of pure teichoic acid by
elution with methyl a-D-glucopyranoside solution.
Bacteriophage $125 infects B. subtilis 168 by binding to CY-D-
glucopyranosyl residues of the cell-wall teichoic acids. Prior treatment
of B . subtilis 168 with con A completely inhibited phage
The inhibition depended on the concentration of con A, and was rever-
sible in the presence of methyl ~~-D-glucopyranoside.~~~ The organiza-
tion of the surface teichoic acid was investigated by electron micros-
copy after complexing cells with con A.399Binding studies with 14C-
labelled con A revealed teichoic acids on the external, but not internal,
face of the cell wall.400The number of con A-reactive sites increased
significantly following limited enzymic digestion of the cell, indicating
that some teichoic acids are probably embedded within the peptido-
glycan matrix.400

j. Interaction of Concanavalin A with Glyc~proteins.~~~-For many

years, the reaction of jack-bean extract with the components of various
animal secretions and body fluids, notably plasma102*10"107~402,403 and
gastric juice,404has been recognized. Kabat and coworkers dem-
(396)T.-J. Kan, R. J. Doyle, and D. C. Birdsell, Carbohydr. Res., 31,401-404 (1973).
(397)R. J. Doyle, D. C. Birdsell, and F. E. Young, Prep. Biochem., 3, 13-18 (1973).
(398)D. C.Birdsell and R. J. Doyle,]. Bacteriol., 113, 198-202 (1973).
(399)D. C.Birdsell, R. J. Doyle, and M. Morgenstem,]. Bacteriol., 121,726-734(1975).
(400)R. J. Doyle, M. L. McDannel, J. R. Helman, and U. N. Streips,]. Bacteriol., 122,
152-158 (1975).
(401)A. Surolia, S.Bishayee, A. Ahmad, K. A. Balasubramanian, D. Thambi-Dorai, S. K.
Podder, and B. K. Bachhawat, Ado. E x p . Med. Biol., 55,95115 (1975).
(402) S. Murakawa and S. Nakamura, Bull. Yamaguchi Med. Sch., 10,11-29 (1963).
(403) S. Nakamura, S. Tominaga, A. Katsuno, and S. Murakawa, Cornp. Biochem.
Physiol., 15,435-444 (1965).
(404) A. E. Clark and M. A. Denborough, Biochem.]., 121,811-816 (1971).
178 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

onstrated the interaction of con A with blood-group substances iso-

lated from hog-stomach linings and from a specimen of human-stomach
lining.170A product formed by alkaline, borohydride degradation of hog
gastric-mucin, namely, 4-0-(2-acetamido-2-deoxy-a-~-glucopyrano-
sy1)-D-galactitol, inhibited con A-mucin precipitation, suggesting a
basis for this interaction.405Human ovarian-cyst mucins having B, I,
and i activities reacted similarly with jack-bean lectin.406-408
Complex-formation between con A and various glycoproteins has
now been elucidated. For example, con A specifically binds several
providing a basis for isolation
purified, glycoprotein lectins,10s*,240,409-411
of lectins by affinity chromatography on con A - S e p h a r ~ s e .Certain
~~~
myeloma proteins and immunoglobulins,409~412 ribonuclease B (Ref.
409),o v a l b ~ r n i nand
, ~ ~carcinoembryonic
~ antigen4I3also precipitated
with con A. Additional con A-reactive glycoproteins and glycopeptides,
identified by inhibition of hemagglutination, mitogenesis, or polysac-
charide precipitation, include serum proteins ( t r a n ~ f e r r i n ) ,im-
~~~
munogl~bulins,~ ~ ~thyr~globulin.~~'
and ~ ~ ~ ~ - ~ ~ ~ Con A also precipitates
mannan-protein complexes such as occur in extracellular /3-D-
glucanases from Saccharomyces c e r e v i ~ i a eLectin-reactive
.~~~~ glyco-
proteins from cell membranes are discussed in Section VIII.
Occasionally, con A-glycoprotein complexes are stabilized by
nonspecific, hydrophobic interaction, rendering them insoluble even
in the presence of known, carbohydrate inhibitors (for example, methyl
a-D-mannopyranoside). In these instances, organic solvents (for exam-
ple, ethylene glycol) may dissociate the complex. The experi-

(405) C. Moreno and E. A. Kabat,J. lmmunol., 102, 1363-1367 (1969).

(406) F. Maisonrouge-McAuliffe and E. A. Kabat, Arch. Biochcm. Biophys., 175,71-80
(1976).
(407) F. Maisonrouge-McAuliffe and E. A. Kabat, Arch. Biochem. Biophys., 175, 81-
89 (1976).
(408) F. Maisonrouge-McAuliffe and E. A. Kabat, Arch. Biochem. Biophys., 175, 90-
113 (1976).
(409) I. J. Goldstein, L. L. So, Y. Yang, and Q. C. Callies,J. Immunol., 103, 695-698
(1969).
(410) W. Bessler and I. J . Goldstein, F E B S Lett., 34, 58-62 (1973).
(411) W. G. Jaffb, A. Levy, and D . I. Gonzilez, Phytochemistry, 13,2685-2693 (1974).
(412) M. A. Leon, Science, 158, 1325-1326 (1967).
(413) S. Hammarstrom, E. Engvall, B. G. Johansson, S. Svensson, G. Sundblad, and I. J.
Goldstein, Proc. Natl. Acad. Sci. U.S.A.,72, 1528-1532 (1975).
(414) B. R. Andersen, lmmunochemistry, 6, 739-749 (1969).
(415) P. S. Chase and F. Miller, Cell. lmmunol., 6, 132-139 (1973).
(416) R. Kornfeld and C. Ferris,J. Biol. Chem., 250,2614-2619 (1975).
(417) S. Toyoshima, M. Fukuda: and T. Osawa, Biochemistry, 11,4000-4005 (1972).
(417a) P. Biely, Z. Kritkjr, and S. Bauer, Eur. J. Biochem., 70, 75-81 (1976).
(418) M. W. Davey, E. Sulkowski, and W. A. Carter, Biochemistry, 15,704-713 (1976).
(419) E. F. Plow and H . Resnick, Biochim. Biophys. Acta, 221,657-661 (1970).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 179

mental systems studied concerned con A and i n t e r f e r ~ n ' and

~ ~ ,ri-
~~~
cin.lg3Thus, superimposed upon primary, specific interaction between
con A and glycoproteins (by carbohydrate binding) may be secondary,
nonspecific, nonpolar interactions; the latter may be of pronounced
importance in considering the reactions between con A and cellular
surfaces.

k. Interaction of Convanavalin A with Miscellaneous Poly-

saccharides.-In addition to branched a-D-glucans, a-D-mannans,
and p-D-fructans, con A also interacts with numerous other polysac-
charides that contain the proper determinant sugars. These polysac-
charides include lipopolysaccharides (for example, those isolated from
Salmonella typhimurium, S . bredeney, and S. senftenberg) which con-
tain terminal a-D-glucopyranosyl certain pneumococ-
cal, capsular polysaccharides (for example, S XII, which contains
kojibiosyl e n d - g r o u p ~ ) ' ~ arabinogalactans
~ , ~ ~ ~ * ~ ~ ~ ; (which possess non-
reducing, terminal D-arabinofuranosyl g r o ~ p s ) ~ arabinoman- ~~-~~~;
nans42ss427;and polysaccharides from Histoplasma capsulatum (all of
which reportedly contain D-mannose, D-galactose, and traces of
D-gh~CoSe'~~). The precipitin reaction between con A and certain Kleb-
siella polysaccharides possessing 2-0-substituted a-D-mannopyranosyl
residues has already been noted.209*240

1. Carbohydrate-binding Specificity of Concanavalin A.-The

carbohydrate-binding specificity of con A has been studied in great
detail by a variety of approaches: ( a ) con A precipitation by natu-
rally occurring, and model, carbohydrate-containing macromole-
cu1es,I02,113,117,120,121,126,170,177,253,3j5,420,423 ( b ) quantitative, hapten inhibi-
tion of precipitation,168-170*197*204,215*365 ( c ) hemagglutination and
lymphocyte stimulation,249( d ) displacement of the lectin from
S e p h a d e ~ ,(' e~)~equilibrium d i a l y s i ~ ,(f) ' ~ U.V.
~ ~ ~spectroscopy,
~~
and (g) n.m.r. ~ p e c t r o s c o p yThese .~~~ results
~ ~ ~permit
~ ~ ~ a~ mapping
~ of

(420) I. J. Goldstein and A. M. Staub, Immumchemistry, 7, 315-319 (1970).

(421) J. A. Cifonelli, P. Rebers, M. B. Perry, and J. K. N. Jones, Biochemistry, 5,3066-
3072 (1966).
(422) I. J. Goldstein, J . A. Cifonelli, and J. Duke, Biochemistry, 13, 867-870 (1974).
(423) I. J. Goldstein and A . Misaki,]. Bacteriol., 103, 422-425 (1970).
(424) T. M. Daniel and J. J . Wisnieski, Am. Reo. Respir. Dis., 101, 762-764 (1970).
(425) T. M. Daniel, Am. Rezj. Respir. Dis., 110, 634-640 (1974).
(426) T. M . Daniel and L. S. Todd, Am. Reu. Respir. Dis., 112,361-364 (1975).
(427) T. M. Daniel and A. Misaki, A m . Rev. Respir. Dis., 113,705-706 (1976).
(428) G. Betail, L. Genaud, and M. Coulet, Ann. Inst. Pasteur, 123, 731-740 (1972).
(429) R. D . Brown, 111, C. F. Brewer, and S. H. Koenig,Ado. E x p . Med. Biol., 55,323-324
(1975).
180 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

the con A binding-site in terms of the various saccharide hydroxyl

groups with which the protein interacts.
The simplest known sugar derivatives that interact with the jack-
bean lectin are 1,5-anhydro-2-deoxy-~-arabino-hexitol and 1,4-
anhydro-~-arabinitol.l~~~~~ These alditols contain the minimal struc-
tural features required for binding to con A, namely, unmodified hy-
droxyl groups of the D - W U ~ ~ configuration
~ O at C-3, C-4, and C-5 of the
1,5-anhydrohexitol, or C-2, C-3, and C-4 of the 1,4-anhydropentitol ring
systems. Introduction of an axial 2-hydroxyl group of the L-gZycero
configuration into 1,5-anhydro-2-deoxy-~-arabino-hexitol enhances
binding affinity to the lectin, whereas an equatorial 2-hydroxyl group of
the D-gly Cero configuration diminishes its complementar-
Configurationally related sugars will also interact
ity.168,169,'97,204*215,365
with the protein. For example, a-L-sorbopyranose and p-D-
fructopyranose can formally be related to 1,5-anhydro-D-glucitol and
1,5-anhydro-~-mannitol, respectively, and a-and p-D-arabinofuranose,
a- and p-D-fructofuranose, 2,5-anhydro-~-mannitol,and 2,5-anhydro-
D-glucitol to 1,4-anhydro-~-arabinitol.~~~~~~~ Fig. 7 outlines these
configurational relationships.

CH,OH

FIG. 7.-Configurational Relationships Among Furanoid and Pyranoid Sugars that

Interact with Con A. (Hydroxyl groups critical for binding to con A are underlined.
Note the relationship between the hydroxyl groups in the furanoid and pyranoid
systems.) (a) R,, R,, R,, R4, R, = H : 1,5-anhydro-2-deoxy-~-arabino-hexitol; R,, R3,
€&, R, = H; R, = OH : 2-deoxy-a-~-arabino-hexopyranose; R,, R,, R3. R, = H; R4 =
OH : 1,5-anhydro-~-glucitol;R,, R,, R4,R, = H; R3 = O H : 1,5-anhydro-~-mannitol; R,, R3,
& = H; Rz = R4 =OH : a-D-glucopyranose; R,, €&, €& = H; R, = R, = OH : a - ~ -
mannopyranose; R,, R,, RI = H; R4 = R5 = OH : a - ~sorbopyranose;
- R,, R,, R, = H; R3 =
R5 = OH : P-D-fructopyranose; (b) R,, R, = H : 1,4-anhydro-~-arabinitol; R, = H; R2 =
OH : a-D-arabinofuranose; R, = H; R, = OH : p-Darabinofuranose; R, = CH,OH; R2 =
OH : a:D-fructofuranose; R, = CH,OH; R, = OH : p-Dfructofuranose.

Con A exhibits a pronounced preference for the a anomer of

D-manno- and D-gluco-pyranose. In fact, a-D-mannopyranosyl resi-
dues are most complementary to the carbohydrate-binding
site168-170,197,204,21.5,365 of con A; there are loci in the carbohydrate-
specific, binding site of con A that interact with each hydroxyl group, as
well as with the anomeric oxygen atom of this residue.365Epimeric
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 181

TABLEI
Inhibition of Concanavalin A-Dextran 13554 Precipitation by Mono- and
Oligo-saccharides” and Some of Their Derivatives

Micromoles giving
Saccharide 50% inhibition

D-Glucose 20
DMannose 4.4
D-Fructose 8.8
D-Galactose 104 (8%)b
D-Allose 97 (4%)

Methyl a-D-glucopyranoside 2.5

Methyl P-Dglucopyranoside 67
Methyl a-Dmannopyranoside 0.6
Methyl 2-acetamido-2-deoxy-a-~-glucopyranoside 5.0

Methyl P-D-fructopyranoside 0.85

Methyl a-D-hctopyranoside 107 (40%)
Methyl a-D-fructofuranoside 16
Methyl p-D-fructofuranoside 5.7
Methyl a-L-sorbopyranoside 3.1
Maltose 4.3
Isomaltose 2.2
Cellobiose 100 (2%)
Laminarabiose 102 (0%)
Gentiobiose 100
Sucrose 23

Sophorose 7
2-O-~-~Galactopyranosy~-D-g~ucose 7
Methyl a-sophoroside 1.2
Methyl P-sophoroside 25
~ ~~

“Data taken from Refs. 168,169,197,215,365, and 430.*Numbers in parentheses refer

to percentage inhibition given by the pmoles of saccharide noted.

a-D-glucopyranosy~residues have %th to V5th of the affinity of methyl

a-D-mannopyranosyl residues, whereas a-D-galactopyranosy1residues
exhibit no affinity.16B-170,197.2”.215,s65
Hapten-inhibition studies with a
series of deoxy, deoxyfluoro, and 0-methyl derivatives of D-glucose and
D-mannose permit a more-definitive analysis of the binding mechanism
(see Table I).168,169~197”40~36s
The observation that 6-deoxy-~-glucose,
(430)I. J. Goldstein, R. N. Iyer, E. E. Smith, and L. L. So, Biochemistry, 6,2373-2377
(1967).
182 IRWIN J. GOLDSTEIN A N D COLLEEN E. HAYES

6-O-methyl-~glucose, and methyl 6-deoxy-6-fluoro-a-D-g~uco-

pyranoside all poorly inhibit the con A-dextran precipitation reac-
tion suggests that it may be the hydrogen atom of the 6-hydroxyl
group that interacts1ss~240*365 with con A. A report that D-glucuronic acid
is an inhibitor of the con A system could not be s u b ~ t a n t i a t e d . ~ ~ ' * ~ ~ ~
Similarly, the failure of methyl 4-deoxy-a-~-xylo-hexopyranoside,
methyl 4-O-methyl-a-~-glucopyranoside, and 4-deoxy-4-fluoro-~-
glucose to bind to con A implicates the hydrogen atom ofthe 4-hydroxyl
group of D-gIucose and D-mannose in the binding phenomenon.240In
contrast, although methyl 3-deoxy-a-~-arubino-hexopyranoside and
methyl 3-O-methyl-a-~-glucopyranosideare poor inhibitors, 3-deoxy-
3-fluoro-D-glucose binds to the protein almost as well as D-glucose,
s ~ g g e s t i n ga~role * ~the
~ ~ for ~ ~oxygen atom of the 3-hydroxyl group in
binding to con A. Finally, inasmuch as the 2-methyl ether of D-mannose
and 2-deoxy-2-fluoro-D-mannose inhibit con A-polysaccharide interac-
tion to the same extent, we suggest that the oxygen atom on C-2 of
D-mannose participates in hydrogen bonding to the [It is
of some interest that, whereas 2-acetamido-2-deoxy-~-mannose is a
noninhibitor, the ortho ester 1,2-0-(1-methoxyethy1idene)-fi-D-manno-
pyranose, as well as the disaccharide 2-O-fi-D-glucopyranosyl-D-
mannose, binds to con A approximately as well as D-mannose
Methyl a-D-glycopyranosides of D-mannose and D-glu-
cose are six to seven times as inhibitory as their correspond-
ing 1,8anhydrohexitol derivatives, prompting the suggestion that the
a-D-glycosidic oxygen atom contributes to the binding energy of the
~ ~ ~ * ~ ~ of~ this view, a - ~ -
con A-carbohydrate ~ o m p l e ~In. substantiation
glucopyranosyl fluoride inhibits the con A system to the same extent as
. ~ ~ ~ data are summarized in Tables I and I1 and Fig. 8.
D - g l u c o ~ eThese

H H
FIG.8.-Carbohydrate-binding Specificity of Con A. (The hydroxyl groups of the
a-D-mannopyranosyl group that are most critically involved in binding to con A are
italicized. Hydrogen and oxygen atoms believed to participate in hydrogen bonding
are overscored.)

Con A will tolerate considerable modification at C-2 of D-glucose. In

fact, 2-deoxy-~-arabino-hexoseis bound more strongly than
~ - g l u c o s e .Whereas
~ ~ ~ * ~the
~ ~2-methyl and 2-ethyl ethers of D-glucose
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 183

TABLEI1
Inhibiting Power of Various Monosaccharide Derivatives"
~

Quantity needed
for 5w0inhibition0
Compound (pmoles)
~

Methyl 2-deoxy-a-~-urubino-hexopyranoside 0.95

Methyl a-D-mannopyranoside 0.34
Methyl a-D-glucopyranoside 1.30

D-Mannose 2.0
2-0-Methyl-D-mannose 1.9
p-Nitrophenyl a-D-mannopyranoside 0.17
p-Nitrophenyl2-O-methyl-a-~-mannopyranoside 0.19
2-Deoxy-2-fluoro-D-mannose 1.5

DGlucose
3-Deoxy-~-ribo-hexopyranose
3-Deoxy-3-fluoro-D-glucose

Methyl a-Dglucopyranoside
Methyl a-D-galactopyranoside
Methyl 4-deoxy-a-~xyZo-hexopyranoside
Methyl 4O-methyl-a-~-glucopyranoside
D-Glucose
4-Deoxy-4-fluoro-~-g~ucose
Methyl a-Dglucopyranoside
Methyl a-D-xylopyranoside

Methyl 6-deoxy-a-~-glucopyranoside
Methyl 6-deoxy-6-fluoro-a-~-glucopyranoside

1,5-Anhydro-~-glucitol 7.2
Methyl a-D-ghcopyranoside 1.3
Methyl P-Dglucopyranoside 37.0

DGlucose 25
a-DGlucopyranosyl fluoride 25

"Because these data were collected in a series of experiments over a long period of
time, only the numerical values within each set may be directly compared. a m e r e num-
bers are given in parentheses, 50% inhibition was not attained; the numbers in paren-
theses indicate the percent inhibition given by the pmoles of saccharide noted. Re-
printed, with permission, from Ann. N.Y. Acud. Sci.140
are equivalent to D-glUCOSe in inhibitory potency, 2-acetamido-2-
deoxy-D-glucose is only 50% as active, and the 2-chloro-2-deoxy and
2-deoxy-2-iodo derivatives are '/loth and +oth as potent as ~ - g l u c o s e . ~ ~ ~
Interestingly, 2-amino-2-deoxy-D-glucose itself does not bindlg7to con
A, probably owing to the positively charged amino group (pK 7.8).
184 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

A comprehensive examination of furanoid sugars allows the

generalization that 1,4-anhydro-~-arabinitol(see Fig. 7) possesses2"
the minimal structural requirements for interaction with con A. The
disposition of the hydroxyl groups in this sugar derivative is similar to
the orientation of the hydroxyl groups at C-3, (2-4, and C-5 of the
D-mannopyranosyl residue.21sOf the pentoses, only D-arabinofuranose
contains the required binding loci2l5;con A interacts with Mycobac-
terium bovis cell-wall arabinogalactan (containing nonreducing, ter-
minal, a-Darabinofuranosyl (compare, Ref. 424).
Both 2,5-anhydro-~-mannitoland 2,5-anhydro-~-glucitolcontain the
requisite 1,4-anhydro-~-arabinitol ring system.21s*240
The D-mannitOl
derivative is about 3.5 times as potent an inhibitor as the D-glUCitOl
derivative, perhaps by virtue of the fact that it may bind to con A either
through the hydroxyl groups at C-1, (2-3, and C-4, or (2-3, C-4, and C-6
(compare Ref. 429). Finally, the important observation that methyl a-
and @-D-fructofuranosideare inhibitors (the @ anomer being approxi-
mately three times as active as the a anomer) provides an explanation
for the interaction of con A with plant and bacterial levans (see
Section II,l,h).
Hapten-inhibition studies with a group of disaccharides indicated
that, with but one exception, namely, s o p h o r ~ s (see e ~ ~ later),
~ only
those disaccharides containing nonreducing, terminal a - ~ -
glucopyranosyl or a-D-mannopyranosyl groups i n t e r a ~ t e d ~ ~with ~*"~~'~~
con A. Thus, isomaltose, kojibiose, maltose, and nigerose (in order of
potency) inhibited the con A-dextran precipitation reaction.lsa In con-
trast, laminarabiose and cellobiose were noninhibitors, gentiobiose
was a very poor inhibitor, and, unexpectedly, sophorose was a strong
i n h i b i t ~ r ' ~ ~(see ~ * ~ ~I).~ Although a more limited number of
* ' ~Table
D-mannose-containing disaccharides was investigated, the same pat-
tern emerged: a d i n k e d D-mannobioses were good inhibitors,
whereas 4-0-@-D-mannopyranosyl-D-mannose was a n0ninhibit0r.l~~
Later, it was found that, like sophorose, 2-O-P-D-mannopyranosyl-D-
mannose also binds209to con A.
Investigation of higher oligosaccharide series revealed several im-
portant feature^.'^' On a molar basis, all members of the homologous
"maltodextrin" series of oligosaccharides (maltose to maltodecaose, in-
clusive) inhibited the con A-dextran precipitation reaction to the same
extent.lg7The same was true of isomalto-oligosaccharides (isomaltose
to isomaltoheptaose, inclusive), and the methyl isomaltoside series
(methyl a-isomaltoside to methyl a-isomalto-octaoside, i n c l ~ s i v e ' ~ ~ ) .
These data indicated that the con A combining-site was complementary
to a single, nonreducing, terminal a - ~ 1+4)- -( or a - ~ 1+6)-
-(
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 185

glucopyranosyl group.Is7 The failure of the cyclo-hexa- and -hepta-

amyloses to inhibit is consistent with the absence of nonreducing,
terminal D-glucosyl groups from their ring-like molecules.1s7
In sharp contrast to the malto- and isomalto-oligosaccharide series,
the inhibiting power of a series of manno-oligosaccharides containing
(1+2)-linked a-D-mannopyranosyl residues increased as the number of
units was i n ~ r e a s e d .Thus,
~ ~ ~ mannobiose
~~~' was five, and mannotriose
and mannotetraose were twenty, times as potent, respectively, as
methyl a-D-mannopyranoside.2".43lThese results raised the possibility
that the con A combining-site may be composed of a series of subsites
similar to that found for wheat-germ agglutinin'28 (see Section 111,4);
in this case, the subsite specificity would be complementary to se-
quences of a-D-( 1-*2)-mannopyranosyl residues Further support
.2043431

for an extended binding-site derives from spectrophotometric studies

that indicated that con A can interact simultaneously with hydroxyl
groups on both a-D-mannopyranosyl units of the chromogenic ligand
p-nitrophenyl 2-O-a-D-mannopyranosy~-a-D-rnannopyran0side.~~~~
Examination of a series of branched oligosaccharides and of those
containing several different kinds of glycosidic linkages was informa-
tive.lS7All gluco-oligosaccharides containing terminal, nonreducing
isomaltosyl groups (for example, panose) inhibited to the same extent
as isomaltose; the same phenomenon was observed for oligosac-
charides containing maltosyl e n d - g r o u p ~ .These
' ~ ~ data lend support to
the hypothesis that it is mainly the terminal, nonreducing, a - ~ -
glycopyranosyl groups of simple and complex a-D-glycans with which
con A interactsls7(however, see later). The surprisingly low inhibitory
activity of the trisaccharide 4,6-di-O-a-~-g~ucopyranosy~-D-g~ucose, al-
though it possesses two nonreducing a-D-glucosyl groups, was attrib-
uted to steric hindrance to the close approach of the protein.1s7
The discovery that sophorose (2-O-@-D-g~ucopyranosy~-D-g~ucose)
was a good inhibitor of the con A system430forced a re-examination of
the concept of an exclusive "chain-end mechanism" for con
@-D-Ghcopyranosyl residues
A-carbohydrate interaction.121~1e7~zo4~z40~365
as they occur in oligo- and poly-saccharides do not i n t e r a ~ t ' ~with
~*'~~
con A. Therefore, the 3-, 4-, and 6-hydroxyl groups of the reducing
D-glUCOSe residue of sophorose were implicated as jack-bean lectin
binding-loci (see Fig. 9). A comprehensive study of the interaction of
con A with sophorose and its derivatives supported this view430(see
Table I).
(431) I. J. Goldstein, Adu. E n p . Med. Biol., 55, 35-53 (1975).
(431a) T. J. Williams, J. A. Shafer, and I. J. Goldstein,Abstr. Pap. Am. Chem. SOC. Meet.,
174, CARB-64 (1977).
186 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Sophorose ~2-O-~-o-glucopyronosyl-o-glucose)
FIG. 9.-Structure of Sophorose (2-0-p-DGlucopyranosyl-D-glucose). (The site of
interaction with con A is outlined. From Ref. 240. Printed by permission of Ann.
N.Y. Acad. Sci.)

Over ten years ago, Hehre and his pointed out that
2-0-substituted a-D-glucopyranosyl residues, which occur in many
dextrans, possess the configurational features necessary for interaction
with con A (see earlier). Glycosylated a-D-mannopyranosyl residues
should also interact with con A; this has been shown conclusively by
inhibition and precipitation s t ~ d i e s . ~Thus, , ~ ~ ~ ,the
~ ~ although ~ ~trisac-
~
charides 0-a-D-galactopyranosyl-( 1+2)-0-a-~mannopyranosy~-( 142)-
D-mannose and 0-a-D-galactopyranos yl-( 1+6)-O-a-D-mannopyrano-
syl-( 1+2)-D-mannose each possesses a terminal, a-D-galactopyranosyl
group, they nevertheless are more potent inhibitors of the con A-
dextran system than methyl a - D - m a n n o p y r a n o ~ i d e .Nonspecific
2~~~~~~~~~~
interaction of a-D-galactopyranosyl residues with the .protein could
contribute to enhanced activity. Reduction of the first-mentioned tri-
saccharide with sodium borohydride yielded the corresponding
alditol, which still possessed considerable inhibitory potency owing to
its internal, (1+2)-substituted, a-D-mannopyranosyl residue.209
Similarly, 2-0-~-D-g~ucopyranosy~-D-mannose~09 2-0-P-D-manno-
pyranosy~-D-mannose,209and 2-0-(2-acetamido-2-deoxy-~-~-gluco-
pyranosy~)-D-mannose210 all inhibited the con A system. The last-
named disaccharide is of particular interest, because it occurs
commonly in erythrocyte and lymphocyte surface-glycoproteins, as
well as in plasma proteins (for example, immunoglobulin^).^^^^^^^^^^^ BY
hemagglutination inhibition, this disaccharide and its methyl
a-glycopyranoside were shown to be as effectively bound as methyl
a-D-mannopyranoside."O On the other hand, the same investigators

(432) R. Kornfeld,J. Keller, J. Baenziger, and S. Kornfeld,]. B i d . Chem.,246,3259-3268

(1971).
(433) F. Miller, Immunochemistry, 9,217-228 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 187

observed that the disaccharide methyl furanoside was inactive, sup-

porting the hypothesis that it is the reducing 2-0-substituted a - ~ -
mannopyranose residue that is the locus21ofor interaction with con A.
Kornfeld and Ferris416studied intact, and partially degraded, im-
munoglobulin glycopeptides for their ability to interact with con A. The
most potent inhibitor was the branched glycopeptide 1containing two
terminal @-~-GlcNAcp-( 1+2)-a-D-Manp units. Kornfeld and Ferris4I6

R
t
p-o-GlcNA~p-(I-- 2)- a - ~ - M a n p (1-
- 3)- p-o-Manp
6
t
a-~-Manp
2
t
1
~-D-G~cNAcP
1

conjectured that “in glycopeptides, such terminal P - N -

acetylglucosamine residues [may] assume a more favorable conforma-
tion for interaction with the saccharide binding site of concanavalin A
than the same residue can assume in smaller glycosides.” Although the
results of inhibition studie~’~’make it highly unlikely that
2-acetamido-2-deoxy-~-~-glucopyranosyl residues interact in a speci$c
way with con A, nonspecific interaction with the protein is possible.
Alternatively, glycosyl substitution at 0-2 may enhance the interaction
of a-D-mannopyranosyl residues with con A; this possibility is sup-
ported by the observation that the methyl 2-O-a-(and @)-D-gluco-
pyranosyl-a-D-glucopyranosideshave a greater affinity for con A than
has methyl a - ~ - g ~ u c o p y r a n o s i d e . ~ ~ ~ ~ ~ ~ ~
Further support for the reactivity of internal, 2-0-linked7 a-D-
mannopyranosyl residues with con A comes from precipitation studies
with macromolecules of known structure. A lipopolysaccharide from
Klebsiella 0 group 5 and the capsular polysaccharide from Klebsiella
K-24 (after deacetylation) generatedzo9classical precipitin curves in
their reaction with con A. Several investigators have also reported that
glycopeptides and glycoproteins lacking con A-reactive sugars at
glycosyl chain-ends have the capacity to interact with the lectin.416,433 In
all these cases, there are internal, 2-0-substituted a-D-mannopyranosyl
residues that may act as loci for with con A.
(434)S. Komfeld, J. Rogers, and W. Gregory,]. B i d . Chem., 246,6581-6586 (1971).
(435)C. A. Presant and S. Kornfeld,]. Biol. Chem., 247,6937-6945 (1972).
188 IRWIN J. GOLDSTEIN A N D COLLEEN E. HAYES

The influence of the aglycon on the inhibitory potency of a large

number of alkyl /3-D-glucopyranosides and aryl a- and p-Dglucopy-
ranosides and -mannopyranosides has been investigated.'80*181*323,365*436
The inhibiting power of all of the P-D-glucopyranosides of primary
alcohols that were examined was virtually identical, regardless of the
degree of substitution at the p-carbon atom of the a g l y c ~ n On .~~~
the other hand, those /3-D-glucopyranosides that possess aglycons
branched at the a-carbon atom displayed inhibiting powers in-
versely related365to the degree of substitution at this carbon atom
(see Table 111).The lack of any linear correlation between the Taft

TABLEI11
Inhibition of Concanavalin A Precipitation by Aliphatic and Aromatic
P-D-Glucopyranosides"

Micromoles required
Inhibitor (aglycon) for 50% inhibition

Methyl 37
Ethyl 37
Butyl 36
Neopentyl 37
Isobutyl 35
Benzyl 37
Isopropyl 100
Cyclohexyl 65
Cyclopentyl 46
tert-Butyl 165

Phenyl 7.0
3-Methylphenyl 3.9
3-Eth ylphenyl 3.0
3-Isopropylphenyl 2.4
3-tert-Butylphenyl 1.8
3,5-Dimethylphenyl 2.4
3,5-Di-tert-butylphenyl 0.70

"Data abstracted from Refs. 180 and 365.

substituent constant of the aglycon and the inhibiting power of the

saccharide indicated that polar effects are not involved in the binding of
the aglycon to the protein, and is consistent with the concept that the
P-D-glucopyranosidic oxygen atom is not involved365in the binding of
the glycoside to con A.
In contrast, data have accumulated that suggest the presence of a
(436) R. D. Poretz and I. J. Goldstein,Arch. Biochem. Biophys., 125,1034-1035 (1968).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 189

region on the con A molecule, adjacent to the specific, saccharide-

binding site, which is capable of interacting specifically with the
aromatic ring of phenyl ~-D-glucopyranosides.160~161~3z3*436 On investi-
gating the inhibition potency of a variety of ortho-, metu-, and paru-
substituted-phenyl P-D-glucopyranosides, Poretz and Goldstein ob-
served no electronic substituent effect (m)for p-substituents and only
moderately greater effects foro- a n d m - s u b s t i t u e n t ~ .Nor
' ~ ~was
~ ~ ~there
~
any correlation with the Van der Waals radii or the surface area of the
substituents. However, 2,6-disubstitution of the phenyl group (for
example, 2,6-dimethylphenyl and pentafluorophenyl p-D-
glucopyranoside) drastically lessened the inhibiting power of the sac-
charide.lsO These results suggest that a specific orientation of the
phenyl ring with respect to the pyranose ring is required in order that
the saccharide may interact maximally with a corresponding region on
the protein molecule.1s0
In contrast to the relative insensitivity ofthe con A system to substitu-
tion of the aromatic ring of phenyl p-D-glucopyranosides by polar
groups, there was a very good correlation between the Hammett
substituent-constants and the inhibiting activity of p-substituted-
phenyl a-D-glucopyranosides and a-D-mannopyranosides.
The influence of hydrophobic substituents on the inhibition potency
of aryl a- and p-D-gluco- and a-D-manno-pyranosides was also investi-
gated. A linear correlation was demonstrated between the hyd-
rophobicity of the substituent and the inhibiting power of the glycoside
for 0-,m-, and p-substituted-phenyl p-D-glucopyranosides.
In his studies on the interaction of organic compounds with proteins,
Hansch and coworkers introduced the concept ofrr, a parameter indica-
tive of the hydrophobicity of an atom or group of a t o r n ~ . ~11~has ' , ~been
~~
defined as log P J P , where P , is the partition coefficient ofa substituted
solute in a water-lipophile binary system, and PH is the partition
coefficient of the parent compound. In this case, the partition of a
substituted-aromatic glycoside in 1-octanol-water compared to that of
the parent compound'60*436 was determined. As indicated, there was an
excellent correlation between the hydrophobicity (rr) of
m-substituted-aryl P-D-glucosides and their inhibition potency. In fact,
the most potent inhibitor tested was 3,5-di-tert-butylphenyl p-D-
g l u c o p y r a n ~ s i d e (see ~ ~ ~ 111).Loontiens and colleagueslsl ex-
' ~ ~ ~Table
tended this relationship by demonstrating a good correlation between
the hydrophobicity of p-alkyl-substituted-phenyl p-D-
glucopyranosides and their binding to con A.

(437)T. Fujita, J. Iwasa, and C. Hansch,J. Am. Chem. Soc., 86,5175-5180 (1964).
(438)K. Kiehs, C.Hansch, and L. Moore, Biochemistry, 5,2602-2605 (1966).
190 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

In summary, on the con A molecule (in solution), there is a hydro-

phobic region adjacent to the carbohydrate-binding site that interacts
specifically with the phenyl groups of P-D-glucopyranosides. The
phenyl group must be linked directly to the glycosidic oxygen atom.
(Benzyl P-D-glucopyranoside has only 20%of the inhibition potency of
phenyl P-~-glucopyranoside~~"). Hydrophobic substituents in the
meta- and para-positions are highly effective in enhancing the inhibi-
tion potency of the glycosides.

2. Lens culinaris syn. esculenta (Lentil)

(lentil; a-D-Manp > CY-D-GIC~, CY-D-G~CNAC~)
Hemagglutinating activity in the common lentil (Lens culinaris or
Lens esculenta) was first reported by Landsteiner and R a ~ b i t s c h e k . ~ ~ ~
Several laboratories have since isolated the lentil lectin in pure form,
and studied its physical-chemical properties and interaction with car-
bohydrates.138,142,143,440-442
The lentil lectin consists of two proteins that differ in their elec-
trophoretic mobility. These proteins, termed LcH-A and LcH-B by
Howard and and I and I1 by TichC and colleague^,'^^ may
be isolated as their natural mixture by iso-ionic precipitation and
DEAE-cellulose chr~matography,'~~ or by specific adsorption to
Sephadex G-50, G-100, or G-150 [but not G-25 (compare Refs. 134 and
440)], followed by elution with 0.1 M D - ~ ~ u c o s ~ ' ~ ~ or ~'~~~'~
glycine-hydrochloric acid buffefl4Oof pH 2. Both proteins were found
in all individual seeds examined; however, their relative proportions
varied with the lentil source.441 (Toyoshima and coworkers143 reported a
single protein; nevertheless, two components are evident in their elu-
tion pattern of the lectin from Sephadex by 0.1 M D-glucose.)
The two lectins differ slightly in their affinity for Sephadex G-150,
and can be separated by careful pooling of peak fractions eluted by 0.1
M D - g l u ~ o s e . 'Alternatively,
~~ they can be separated by vertical,
starch-gel e l e c t r o p h o r e ~ i sor,
~ ~ most
~ conveniently, by 0-(carboxy-
methy1)cellulose c h r ~ m a t o g r a p h y . ~The ~ ' , ~ ~isolectins have iden-
tical molecular weights, values reported ranging from 42,000 to
63,000, with 52,000 being the best estimate.'42,'43,440*441 The proteins
(439) K. Landsteiner and H. Raubitshek, Zentrulbl. Bukt. Parusitenk. Infektionskr.
Hyg., Abt. 1. Orig., 45, 660-667 (1907).
G. Entlicher, M. Tichi, J. V. KoStiF, and J. Kocourek,Experientiu, 25,17-I9 (1969).
I. K. Howard, H. J. Sage, M . D. Stein, N . M. Young, M. A. Leon, and D. F. Dyckes,
J. Biol. Chem., 246, 1590-1595 (1971).
(442) M. Paulovi, M. Tichi, G. Entlicher, J. V. Koztif, and J. Kocourek, Biochim.
Biophys. Actu, 252, 388-395 (1971).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 191

were immunochemically identical, as measured by double immunodif-

fusion using an antiserum prepared against the natural mixture,441and
displayed identical hemagglutinating properties.441Each protein gave
a single, symmetrical peak in the analytical ultracentrifuge, and a
single, sharp, protein band by disc-gel electrophoresis.
The lectins have almost identical amino acid cornpositi~ns,~~' and
contain a preponderance of aspartic acid (and/or asparagine),
threonine, serine, and valine142*143,441; 1.8 residues of cysteineM1per
molecular weight of 49,000; but no m e t h i ~ n i n e . ' (Toyoshima
~ ~ ? ~ ~ ~ and
coworkers'43reported 4 residues of methionine per 63 kg of protein;
this was perhaps, due to examination ofa different variety of lentil.) The
difference in electrophoretic behavior and adherence to
O-(carboxymethy1)cellulose has been attributed441to the higher con-
tent of lysine in LcH-B. End-group analysis gave and
t h r e ~ n i n eas ~ N-terminal amino acids, and serine as the C-terminal
' ~the
amino The lectins contain -1.5-3% of neutral carbohy-
drate'43.441(principally D-glucose, as well as 2-amino-2-deoxy-D-
glucose)
Tryptic peptide mapping of each lectin revealed half of the expected
number ofpeptides (15-16 peptides ofthe expected 32-37). The lectins
share 15 identical, or closely similar, tryptic peptides"'; this suggests
that each protein is a molecule consisting of identical halves.441How-
ever, a unique peptide was identified for LcH-A which stained gray
with ninhydrin. The amino terminal sequence of the first 25 amino
acids of the a- and p-subunits of the pea and the lentil lectins has been
determined. It was found that, of the 25 residues analyzed,"za*b~c the
N-terminal a-chain of the lentil and pea lectins differed only at three
positions, and the p-chains at two positions.
Metal analysis revealed one atom of Mn2+ and six atoms of Ca2+
associated with each 66.5 kg of ~ r 0 t e i n .Addition
l~~ of 5 mM Ca2+and
mM Mn2+to the native lectins enhanced both hemagglutination and
mannan precipitation. This indicates that either some metal loss occur-
red during purification, or that the metal binding-sites are never com-
pletely occupied442(compare Refs. 300 and 301). Mn2+stimulated the
hemagglutinating activity of all lentil-lectin preparations more than did
Ca2+,whereas Caz+promoted mannan precipitation more effectively.442
Demetallization, by dialysis against 0.1 M EDTA followed by M acetic
(442a) E. Van Driessche, A. Foriers, A. D. Strosberg, and L. Kanarek, FEBS Lett., 71,
220-222 (1976).
(442b) A. Foriers, E. Van Driessche, R. De Neve, L. Kanarek, and A. D. Strosberg,FEBS
Lett., 75,237-240 (1977).
(442c) A. Foriers, C. Wuilmart, N. Sharon, and A. D. Strosberg, Biochem. Biophys. Res.
CO~WZU 75,~ 980-986
., (1977).
192 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

acid, decreased the Mn2+content and hemagglutination activity by

-95%, without altering the Ca2+content.442The demetallization treat-
ment may have caused partial denaturation, as Mn2+addition failed to
reactivate the protein.442The electron paramagnetic resonance spec-
trum of the lentil phytohemagglutinin was to be similar to that
of con A.
A detailed study of the subunit structure of the lentil lectin (Lens
culinaris Moench; var. Hrotovicka) was conducted by Fliegerovh and
coworkers.444Both isolectins dissociated readily in acidic solution (0.1
M glycine-hydrochloric acid buffer, pH 2.2, or M acetic acid) or in 8 M
urea, and were resolved into their component polypeptide chains by
gel filtration on columns of Bio-Gel P-100 or Sephadex G-200. There are
two distinct species of subunits: heavy chains (H)and light chains (L) of
approximate molecular weights 18,000 and 8,000, respectively (deter-
mined by gel filtration). Howard and confirmed dissocia-
tion of an LcH mixture in 6 M guanidinium chloride-0.1 M
2-mercaptoethanol by sedimentation-equilibrium ultracentrifugation
and by disc-gel electrophoresis in dodecyl sodium sulfate. The molecu-
lar weights obtained by dodecyl sodium sulfate disc-gel elec-
trophoresis were identical with the H and L chains isolated as already
described. The H subunits tended to aggregate in dilute buffer solu-
t i o n ~ . ~ "Because dissociation into subunits did not require
2-mercaptoethanol, the possibility that subunits are linked by disulfide
bonds was excluded.444
Dissociation of the phytohemagglutinin subunits was accompanied
by a complete release of Mn2+from the protein.444Carbohydrate and
Ca2+remained bound to both subunits. Very low hemagglutinating
activity was associated with the heavy subunit; the light subunit was
inactive. Addition of Mn2+and Ca2+to a mixture of H and L chains
failed to restore hemagglutinating
Amino acid analysis of the isolated subunits showed that only the H
chains contained cysteine (one residue per molecule).444Unfortu-
nately, this amino acid analysis did not include methionine, reported
absent by one groupM1and present by a second group.143End-group
analysis showed N-terminal valine and threonine for the H and L
chains, respectively. Treatment with carboxypeptidase gave serine as
the C-terminal amino acid for both subunits.
On the basis of these data, it appears that both lentil isolectins consist
of 2H and 2L subunits giving a noncovalently linked aggregate of
(443) M. Tichjr, M. Tichi, and J. Kocourek, Biochim. Biophys. Acta, 229,63-67 (1971).
(444) 0. Fliegerovi, A. Salvetovi, M. Tichl, and J. Kocourek, Biochim. Biophys. Acta,
351,416-426 (1974).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 193

rnolecular weight 52,000-in good agreement with the determinations

of Howard and and of Hayman and C r ~ m p t o nAs . ~ the
~~
isolated subunits differed both in amino acid composition and
N-terminal amino acid, the L subunit is probably not a fragment of the
H chains. The precise relationship between the two lentil lectins re-
mains unclear, but it would appear from the available data that there is
considerable homology between them, and that one could have arisen
from the gene coding for the other b y minor m ~ t a t i o n ( s ) . ~ ~ l
The lentil lectins are nonspecific with regard to hemagglutination of
human erythrocyte^,'^^,'^^ but they do exhibit a distinct species spec-
ificity. They agglutinate erythrocytes of some species of
rabbit,138,446
and horse,138but not sheep,138goat,138,446 or
The importance of accurately identifying the seed variety is illus-
trated by the report of FialovL and coworkers447They isolated a new
type of hemagglutinin from Lens esculenta Moench, subspecies mi-
crosperma (Baumg.)Barulina, termed “small-seed” lentil as opposed to
the “large-seed” lentil (Lens esculenta Moench. Lens culinaris Med.).
In contrast to the purified, large-seed lentil isolectins, the small-seed
lentil hemagglutinin was retarded but not adsorbed to Sephadex. [Ad-
dition of Mn2+and Ca2+did not change this property (compare Refs. 300
and 301)]. The purified, small-seed lectin consisted of a single
homogeneous protein which corresponded in electrophoretic mobility
to phytohemagglutinin I1 from the large-seed lentil [poly(acrylamide)
gel, pH 8.91. The small-seed lectin was likewise composed of two
subunits (molecular weight 8,000 and 18,000), and had an aggregate
molecular weight of 53,300 *2,500. Although similar in its amino acid
and carbohydrate composition to the large-seed lectins, the small-seed
protein contained only one-fourth of the 2-amino-2-deoxy-~-g~ucose
residues. Analysis revealed the same two N-terminal amino acids (val-
ine and threonine) as in the two large-seed lentil i ~ o l e c t i n s . ~ ~
Hemagglutination inhibition with D-glucose, D-mannose, and their
methyl a-D-glycosides showed no difference between the two varieties
of lentil seeds.447However, the small-seed lectin showed a lower
agglutinating titer against human group B erythrocytes compared to Al,
Az,and 0 cells, and a higher electrophoretic mobility on starch gel than
the large-seed l e ~ t i n s . ~ ~ ~

(445) M. J. Hayman and M . J. Crumpton, Biochem. Biophys. Res. Commun., 47,923-930

(1972).
(446) L. Bures’, G. Entlicher, M. TichP, and J. Kocourek, Experientia, 29, 1546-1547
(1973).
(447) D. Fialovi, M. Tichi, and J. Kocourek, Biochim. Biophys. Acta, 393, 170-181
(1975).
194 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Using equilibrium dialysis, Stein and coworkers186found that there

are approximately two saccharide binding-sites per LcH-A molecule
with K, = 230+30 M-' for D-mannose and K, = 100+24 M-' for methyl
a-Dglucopyranoside. These are extremely low binding-constants (see
Section IX, Table XXV).
The interaction of the lentil lectins with p ~ l y s a c c h a r i d e s ' ~ ~ ~ ~ ~ ~
and g l y c o p r ~ t e i n s ~has ~ ' ~ ~ studied by many investigators.
~ ~ been
Precipitation of muscle glycogen and yeast(Sacchar0myces cereuisiae)
D-mannan by the lentil lectins was reported by TichL and cowork-
e r ~ . On ~ the~ ~other * ~hand,~ ~ sixteen dextrans failed to precipitate
with the leQtil lectins, although interaction with two of these (B-1299-L
and B-1299-S) was demonstrated by their inhibition of lectin-human
erythrocyte agglutinati~n.'~~ In view of their very low D-glUCOSe and
D-mannose binding-constants,'86 it is, perhaps, not surprising that the
lentil lectins do not precipitate dextrans as readily as does con A.'23*'38
The 0-phosphonomannan from Pichia pinus effectively inhibited
hemagglutination and also precipitated with the lectin both in agar-gel
double-diffusion and in solution.123The flocculation profile was very
similar to an antibody-antigen precipitin curve, and proceeded best at
20°,pH 6-8,and an ionic strength >0.1.
The lentil lectin p r e ~ i p i t a t e d ' ~several~ , ' ~ ~ glycoprotein components
of human serum: a,-macroglobulin, IgM, P,-glycoprotein, as well as
traces of IgA and IgG. Transferrin, ceruloplasmin, and haptoglobin
were unreactive. Glycopeptides from transferrin and IgM, but not
ovalbumin, inhibited the Lens culinaris-glycoprotein precipitation
system.'0g Young and coworkers stated that, whereas the lentil lectin
discriminated less well between simple sugars than does con A, it was
superior in distinguishing the aforementioned glycopeptide~.'~~
Classified with lectins inhibited by Makela's group I11 sugars, the
lentil lectin is primarily specific for a-D-mannopyranosyl r e s i d u e ~ . ' ~ ~ J ~ ~
Inhibition studies are summarized in Table IV. D-Glucose and
2-acetamido-2-deoxy-~-glucoseinhibited to the same extent (display-
ing about twice the potency of D-fructose),whereas D-galactose was a
n o n i n h i b i t ~ r . ' ~ ~ ~ ' ~D-Glucose
~ . ' ~ ~ . ' ~ protected
~ the lentil lectins from
heat denaturation, whereas D-galactose did a-D-Linked
D-glUCOSe disaccharides interacted with the lentil lectin, whereas gen-
tiobiose and cellobiose were, re~pectively,'~~ a poor inhibitor and a
noninhibitor.
Aromatic aglycons (phenyl, p-nitrophenyl, and benzyl groups) en-
hanced binding of the a-D-linked glycosides of D-glucose and
2-acetamido-2-deoxy-~-glucose.'~~ Replacement of the hydroxyl
groups at C-3, C-4, or C-6 of methyl a-D-glucopyranoside by hydrogen
atoms abolished the inhibiting capacity, thereby indicating the impor-
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 195

TABLEIV
Comparative Inhibitory Data on the Lentil Lectina

Inhibitor A" B' C"

D-Mannose 1.0 1 1.0

DGlucose 3.8 4 1.0
D-Fructose 5.4 8
2-Acetamido-2-deoxy-~-glucose 2.1 4
Methyl a-D-glucopyranoside 1.1 8 0.7
Methyl P-D-glucopyranoside 2.1 X
Methyl a-D-mannopyranoside 0.5
Methyl 2-acetamido-2-deoxy-a-~-glucopyranoside 1.0
Methyl 2-acetamido-2-deoxy-~-~-glucopyranoside 8' >4.0
3-0-Methyl-D-glucose 0.25
Maltose 1.4 8 1.o
Isomaltose 4 0.4
Cellobiose X >4.0
Gentiobiose 16 4.0

"All data normalized to D-mannose = 1.0."From Ref. 123;molarity required to produce

50% inhibition of precipitation of lentil lectin with phosphonomannan or glycoprotein.
"From Ref. 143;minimum amounts (mg/ml) completely inhibiting 4 hemagglutinating
doses of lentil lectin. x, no inhibition at 20 mg/ml. dFrom Ref. 213;millimolarity needed
to produce 50% inhibition of agglutination.

tance of these groups for binding to the 1 e ~ t i n .That

l ~ ~ the configuration
of the 3- and 4-hydroxyl groups is of critical importance was also indi-
cated by the fact that D-allOSe and D-galactose, the C-3 and C-4 epimers
O f D-glUCOSe, were noninhibitor~.'~~ On the other hand, there is some
latitude in substitution at C-2; D-mannose, 2-deoxy-arabino-hexose,
and D-glucose (in order of potency) all inhibited lentil hemagglutina-
tion. In a comparative study of D-mannose(D-glucose)-binding lectins,
Allen and coworkers213discovered that 3-0-methyl- and 3-0-benzyl-
D-glucose were better inhibitors ofthe lentil lectin than D-glUCOSe itself
(see Table IV).
Like the jack-bean and pea lectins, the lentil agglutinin interacted
with 2-0-(2-acetamido-2-deoxy-~-~-g~ucopyranosy~)-D-mannose~~~
providing evidence that the lentil lectin binds to internal 2-0-
substituted D-mannopyranosyl residues that occur in animal glycopro-
teins.
Van Wauwe and coworkers182studied the effect of various para-
substituents on the binding of phenyl a-D-mannopyranosides to the
lentil lectin. As with con A and the pea lectin, binding of
p-substituted-phenyl a-D-mannopyranosides correlated fairly well
with the Hammett substituent constant uH'in which electron-releasing
196 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

substituents favored binding. The hydrophobicity of the substituents

contributed little or nothing to the binding.
Inasmuch as extensive maleylation of free amino groups did not
abolish the hemagglutinating and poly saccharide-precipitating activity
of the lentil lectin, it is probable that amino groups are not involved in
the carbohydrate-binding site of the l e ~ t i nTreatment
. ~ ~ ~ ~ of the lentil
lectin with 1-acetylimidazole resulted in the modification of 3-5 tyrosyl
and 7-17 amino The modified proteins had lost their
hemagglutinating activity, but retained their capacity to interact with
polysaccharides and to bind to
In summary, the lentil lectins exhibit a carbohydrate-binding spec-
ificity similar to those of the pea lectin and con A; primary specificity is
towards a-Dmannopyranosyl residues (2), and secondary is to a - ~ -

glucopyranosyl and 2-acetamido-2-deoxy-a-~-glucopyranosy~ res-

idues, Furthermore, 2-0-substituted a-D-mannopyranosyl residues
interact with the protein.210
Many unanswered questions remain with respect to the lentil lectins.
These include the chemical relationship between the two lectins, a
comparison of their carbohydrate-binding specificity, a study of the
binding of furanoid sugars (for example, a- and p-arabinofuranosides)
to the lectins, and a determination of the H or 0 atoms of each hydroxyl
group which may be involved in the binding phenomenon. There is
also the question as to the size of the binding site; will the lentil-lectin
site be more complementary to sequences of a-D-(1+2)-
mannopyranosyl residues than is con A (compare Ref. 204)?

3. Pburn sativurn (Pea)

(garden pea; a-D-Manp > u - D - G ~ c a-~-GlcNAcp)
~,
Although there are several reports on the extraction, partial purifica-
tion, and properties of a blood-group ABO nonspecific agglutinin from
only recently have the lectins of the garden pea (Pisum
the pea,141.448
(4474 D. VanEurovi, M. Tichi, and J. Kocourek, Biochim. Biophys. Acta, 453,301-310
( 1976).
(448) S . V. Huprikar and K. Sohonie, Enzymologia, 28, 333-345 (1965).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 197

sativum) been isolated in pure form. Conventional purification-

technique^,^^^,^^^ or affinity chromatography on fol-
lowed by elution with D-glucose solution or acidic b ~ f f e r ~ yield, ~ ~ ~ , ~ ~
pure preparations.
The hemagglutinating activity of the pea resides in two closely re-
lated agglutinins. The isolectins [termed I and I1 (Ref. 141), and A and B
(Ref. 453)], separable by DEAE-cellulose ion-exchange chromatog-
raphy at pH 8.4 and 8.8, have141*453 identical molecular weights of
approximately 50,000, and remarkably similar amino acid composi-
tions. Aspartic acid and threonine preponderate, whereas methionine
and half-cystine are a b ~ e n t . ~Only ~ ~ *traces
~ ~ ~of* carbohydrate
~ ~ ~
(<0.5%)are present.141*453*454 The two pea-agglutinins differ in elec-
trophoretic m ~ b i l i t y .The
~ ~acidic
~,~~ ~ 4.1) component A and neutral
(PI
(PI 6.5) component B occur as a mixture in the ratio453of 1:2. The
isoelectric points of A and B were later found to vary with buffer
composition; Entlicher and K o c o ~ r e k ~ ~ ~ PI 5.9 and 7.0 for the
reported
two pea-agglutinins; a hybrid form had PI 6.35 (see Ref. 455).
Approximately one atom of Mn2+and 2.5 atoms of Ca2+are bound per
lectin molecule.141 (Ethylenedinitri1o)tetraacetate (EDTA) inhibits
both hemagglutination and precipitation of yeast m a n n a r ~ . ~Succes-
’~
sive dialysis against EDTA andM acetic acid removed most ofthe Ca2+,
but did not change the Mn2+content. The apoprotein failed to precipi-
tate D-mannan, and the hemagglutination titer was decreased456by
75%.
Subunit structural studies demonstrated two types of polypeptide
chains. The smaller subunit (a),molecular weight 7,000-10,500, has
N-terminal valine, whereas the larger subunit (p), molecular weight
12,000-18,000, has threonine in the N-terminal
Kocourek and coworkers4s4separated the subunits on Biogel P-100 in 5
M guanidinium hydrochloride buffer containing EDTA [monitoring by
dodecyl sodium sulfate-poly(acry1amide) electrophoresis]. The sub-
units a and p occurred in the same relative proportion in each purified

(449) T. Shinohara, Proc. J p n . Acad., 47,331-336 (1971).

(450) G. Betail, J. Guillot, and M. Coulet, C. R. Soc. Biol., 163, 150-152 (1969).
(451) J. Guillot, M. Mustier, G. Betail, A. M. Chabanier, and M. Coulet, C . R. Soc.
Biol., 163, 152-154 (1969).
(452) K. Onodera and T. Shinohara, Agric. Biol. Chem., 37, 1661-1666 (1973).
(453) I. S. lrowbridge,J. Biol. Chem., 249, 6004-6012 (1974).
(454) T. Maiik, G. Entlicher, and J. Kocourek, Biochim. Biophys. Acta, 336, 53-61
( 1974).
(455) G. Entlicher and J. Kocourek, Biochim. Biophys. Acta, 393, 165-169 (1975).
(456) M. Paulova, G. Entlicher, M. Tichi, J. V. KoStif, and J. Kocourek, Biochim.
Biophys. A c ~ u237,
, 513-518 (1971).
(457) G. Betail, M. Coulet, and J. Guillot, C. R. Soc. B i d , 163, 1771-1775 (1969).
198 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

component, as well as in the native mixture. N-Terminal sequences

of the a- and @-subunitsof the pea lectin were determined442aand
shown to exhibit a remarkable degree of homology with the analogous
subunits of the lentil lectin."2b*c
Trowbridge's results using isoelectric focusing in 8 M urea differed
slightly.453 Components A and B of affinity-purifiedpea-lectin gave two
major protein-bands. The more-basic band (p) was common to both
proteins, but there were two types ofa subunits: an acidic species came
from the A component, and a more-basic species from the B component.
Entlicher and K o c ~ u r e kconfirmed
~~~ these results, adding that the
common subunit (p)was larger than the unique subunits (a). The amino
acid compositions of the a and /? subunits show significant differ-
e n c e ~ .The ~ * ~ ~ ~that one is generated from the other by
~ ~possibility
proteolytic cleavage was excluded on the basis of tryptic peptide map-
ping.4s3The two pea-lectins are evidently tetrameric molecules consist-
ing of two light (a)and two heavy ( p ) polypeptide chains united by
noncovalent forces. They may be represented as a2p2and az'p2,as their
a subunits are unique. A hybrid molecular species aa'p2(PI6.35) might
be generated in alkaline media.454
Equilibrium-dialysis studies showed two equivalent, sugar-binding
sites on each isolectin: intrinsic association-constants (Ka') were 1450
f 2 3 0 M-' for binding to D-mannose and 773 f126 M-' to methyl a - ~ -
glucopyrano~ide.~~~
Photo-oxidation of the pea lectin at pH 8.2 led to stepwise inactiva-
tion of the protein.458Tryptophanyl, histidyl, arginyl, and tyrosyl res-
idues were progressively decomposed, along with a loss of protein-
bound Mn2+.The conformation of the protein, namely, a mixture of
p-structure and random coil, remained unchanged during photo-
oxidation (as measured by circular dichroism). Kocourek and cowork-
e r suggested
~ ~ that ~ certain
~ tryptophan residues might be of impor-
tance in the carbohydrate-binding mechanism of the pea lectin.
Under defined, mild conditions, the reaction of the pea lectin with
(2-nitropheny1)sulfenylchloride results in sulfenylation of only 2 of the
10 tryptophan residues of the lectin molecule, with simultaneous loss
of biological activity. Both sulfenylated tryptophan residues belong to
the two heavy subunits of the lectin. Enzymic hydrolysis, and separa-
tion of the tryptic peptides, gave one homogeneous, yellow, octapep-
tide containing the modified tryptophan residue. The octapeptide
either is directly a part of the pea-lectin binding-site, or plays an
important role in maintaining the tertiary structure of the binding site.
According to the amino acid composition and amino acid sequence, the
(458) L. BureH, G. Entlicher, and J. Kocourek, Biochim. Biophys. Acta, 285,235-242
(1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 199

octapeptide isolated from the pea lectin is almost identical4sgwith a

corresponding peptide in con A.
The carbohydrate-binding specificity of the pea lectins has been
studied by inhibition of hemagglutination,20~1zz*213~4z8~440~44g~452 precipita-
tion,122,21fand equilibrium d i a l y s i ~ . Makela
~ ~ ~ , ~demonstrated
~~ that
sugars conformingto his group I11 sugars were the best inhibitors ofthe
hemagglutination of human erythrocytes .20 In decreasing order of
effectiveness they were: D-mannose, trehalose, and turanose > 2-
amino-2-deoxy-D-glucose, maltose > 2-acetamido-2-deoxy-~-glucose,
D-fiXCtOSe, and sucrose. He suggested that the pea lectin could dis-
tinguish between the a-and P-D-glycosides of D-glUCOSe, D-mannose,
and 2-acetamido-2-deoxy-~-glucose.~~ Other investigators obtained
similar results by inhibition of h e r n a g g l ~ t i n a t i o n ,and
~ ~ ~by
~ ~pre-
~~
cipitation studies.211
Inhibiting sugars enhanced the electrophoretic mobility of the lectin
on starch gels in 0.03 M acetate buffer at pH 5.0 in proportion to the
sugar concentration and the inhibitory power.440
A thorough study of pea-lectin, carbohydrate-binding specificity was
conducted by Van Wauwe and coworkers211(see Table V). By hapten-

TABLEV
Inhibition of the P. satiuum Lectin-P. pinus Phosphonomannan
Precipitation Reaction by Various Carbohydrates‘

Concentration giving
Inhibitor inhibition (mM)

D-Mannose 0.7
D-Glucose 1.7
D- Fructo se 3.0
L-Sorbose 6.2
Methyl a-D-mannopyranoside 0.3
Methyl a-D-glucopyranoside 0.84
Methyl P-D-glucopyranoside 7.4
Phenyl a-D-glucopyranoside 0.44
Phenyl P-D-glucopyranoside 3.5
Maltose 0.75
Isomaltose 0.9
Cellobiose 15.0 (O%)*
Gentiobiose 11.0
Melibiose 15.0 (0%)

“Taken from Ref. 21 1. *Numbers in parentheses represent the percentage inhibition

given by the concentration of carbohydrate noted.

(459) M. Cennakovi, G. Entlicher, and J. Kocourek, Biochim. Biophys. Acta, 420,236-

245 (1976).
200 IRWIN J. GOLDSTEIN A N D COLLEEN E. HAYES

inhibition measurements on the precipitin reaction between the Pisum

sativum lectin and Pichia pinus 0-phosphonomannan, these inves-
tigators showed that configurationally related monosaccharides,
namely, D-mannose,D-glucose, D-fructose, and L-sorbose, bound to the
lectin. Unmodified hydroxyl groups at C-4 and C-6 of the
D-glucopyranose ring were essential for protein binding. Methyl a-D-
mannopyranoside was 2.8 times as potent as the corresponding a-D-
glucoside, indicating a positive contribution of the axial 2-hydroxyl
group of D-mannose to the binding energy of the protein-carbohydrate
complex (compare Ref. 452). Pea lectin, unlike con A, appears to be
relatively insensitive to variations at c - 2 of D-ghOSe; 2-deoxy-D-
arabino-hexose, D-glucose, 2-deoxy-2-fluoro-D-glucose, and
2-acetamido-2-deoxy-~-g~ucose all inhibit to the same extent. Although
2-deoxy-2-fluoro-D-mannose and D-mannose are equally effective,
2-acetamido-2-deoxy-D-mannose failed to inhibit hemagglutination211
(compare, Ref. 197).
In contrast to changes at C-2, modification at the 3-hydroxyl group
notably affects pea-lectin binding. Although 3-deoxy-3-fluoro-~-
glucose has about one-third the activity of D-glUCOSe, the 3-0-methyl
and 3-0-benzyl derivatives of D-glucose are -10 times as potent as
D-glUCOSe.211.213 Likewise, methyl 2,3-di-O-methyl-a-D-
glucopyranoside is 18 times as inhibitory as methyl a - ~ -
glucopyranoside. In this respect, the pea lectin is dissimilar to con A,
which does not tolerate substitution at C-3 of D-glucose, except for the
3-deoxy-3-fluoro derivative.240
The pea lectin binds a-D-linked D-glucobioses, but not p-D-linked
disaccharides (except for gentiobiose, which has one-eleventh the in-
hibitory effect of isomaltose). Like con A, it interacts with 2-0-
(2-acetamido-2-deoxy-~-D-g~ucopyranosyl)-~-mannose, suggesting
that it may be capable of binding to glycoproteins containing internal
2-0-substituted a-D-mannopyranosyl residuesS2l0
Binding of several fluoro sugars by the pea lectin allowed elucidation
of the specific atom of the hydroxyl groups that is recognized by the
lectin.211 Inasmuch as 4-deoxy-4-fluoro- and 6-deoxy-6-fluoro-~-
glucose do not bind, it is probable that the hydrogen atoms of the 4- and
6-hydroxyl groups are hydrogen-bonded to the pea lectin. By the same
reasoning, the oxygen atoms of the 2- and 3-hydroxyl groups of
D-mannose are implicated in lectin interaction; this is precisely the
pattern observed for con A (see Fig. 8).169,240-365
Significant differences were found for the binding of methyl a- and
P-D-xylopyranoside to the pea lectin.211These glycosides, lacking hy-
droxymethyl groups on (2-5,are less effective, by factors Of V240 and 6/20,
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 201

than the corresponding methyl a- and P-D-glucopyranosides; methyl

P-D-xylopyranoside is -4-6 times as potent as the corresponding a
anomer. These results conform to a hypothesis advanced by Brewer
and colleagues for the con A binding-site308:the 5-(hydroxymethyl)
group of methyl P-D-glucopyranoside binds in the same position oc-
cupied by the 2-hydroxyl group of methyl a-D-glucopyranoside.
The effect of para-substitution on the affinity of phenyl a - ~ -
mannopyranoside (and phenyl a-D-glucopyranoside) for the pea lectin
was investigated by Van Wauwe and coworkers.'82 Affinities were re-
lated to the electronic properties of the substituents, expressed as the
Hammett substituent-constants. Thus, electron-releasing substituents
favored the binding of p-substituted-phenyl a-D-mannopyranoside
(and a-D-glucopyranoside) by an increase in electron density either at
the anomeric oxygen atom or in the phenyl ring. The same phenome-
non applied to con A (Refs. 180 and 181) and the lentil lectin.182The
contribution of the hydrophobic character of para-substituents was
small. On the other hand, the corresponding aryl P-D-glucosides bound
to the pea lectin independent of both the polar and hydrophobic nature
of the substituent.182
The pea lectin precipitated with muscle glycogen, yeast mannan
from Saccharom yces cereuisiae, and 0-phosphonomannan from Pichia
pinus. 122 All of these reactions were inhibited by specific, sugar haptens
(D-mannose, D-glucose, and D-fructose).'22
A comparative study of the hemagglutinating activities of the pea
(Pisum satiuum), the lentil (Lens culinaris), and the jack-bean
(Canaualia ensiformis) lectins towards the erythrocytes of fourteen
different species that the pea and lentil lectins were, in all
instances, more active than con A. Of great interest was the finding that,
among these three lectins, con A alone was completely inactive against
human erythrocytes (blood group, unspecified).446
The pea lectin, like con A, agglutinated normal, embryonic fibro-
blasts of human and rat origin at high lectin concentrations (1.500 mg/
ml), whereas, various, rat tumor-cells transformed in vitro (spontane-
ously, and by Rous sarcoma virus) were agglutinated at very low con-
centrations (5 Fg/ml) of l e ~ t i n . ~ ~ ~

4. Viciu fuba (Fava Bean)

(fava or broad bean; a-D-Manp > a - ~ - G l c p>> a-~-GlcNAcp)
Apart from its capacity to agglutinate human erythrocytes nonspe-
cifically,the fava bean(Vicia faba) has been known for many years as the
(460) P. Vesel?, G . Entlicher, and J. Kocourek, Erperientia, 28, 1085-1086 (1972).
202 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

cause of favism among certain population groups, notably those from

countries bordering the eastern Mediterranean.461Favism is an ac-
quired, hemolytic anemia associated with metabolic abnormalities of
the erythrocyte (for example, instability to glutathione, and low
D-glucose 6-phosphate dehydrogenase activity.) Although aggravated
by ingestion of the raw or partially cooked fava bean, the disease does
not appear to be due to the presence of the fava-bean agglutinin.
In their pioneering paper on the blood-group specific hemaggluti-
nins, Boyd and Reguera5 observed that a saline extract of the fava bean
agglutinated all types of human erythrocytes. Both K r i i ~ eand ~~
M&ela78included the fava bean in their compendia of plant aggluti-
nins.
Cregar and GiffordM2confirmed that fava-bean extracts agglutinated
human, red blood-cells without regard to blood group, and also aggluti-
nated guinea-pig, rabbit, and albino-rat erythrocytes. Erythrocyte
aggregation was greatly enhanced by incorporating gum acacia (but not
gum ghatti or gum tragacanth) in the agglutination medium, but was
inhibited by human serum (apparently the y-globulin f r a ~ t i o n ) . ~ ~ , ~ ~ ~
Treatment of erythrocytes with proteolytic enzymes464(for example,
ficin, and an exopolypeptidase from Streptomyces griseus) greatly in-
creased agglutinability by fava-bean extracts, whereas treatment of the
cells with tannic acid inhibited a g g l u t i n a t i ~ nVarious
. ~ ~ ~ murine and rat
tumor-cells were also agglutinated by fava-bean extracts.13QThe
agglutination and toxicity of a purified preparation of the lectin were
tested against Yoshida sarcoma cells.466Hashem and K a b a r i t ~re- ~~~
ported that fava-bean extracts were mitogenic for human, peripheral
lymphocytes.
The Viciafuba lectin was first purified by Tomita and by
affinity chromatography of crude extracts on a column of Sephadex G-50;
however, no physical-chemical properties were reported. Also employ-
ing Sephadex affinity-chromatography (elution with 0.1 M D-glucose),
Wang and coworkers140isolated the lectin in pure form. Gel elec-
trophoresis in dodecyl sodium sulfate gave a major protein-band (mol.
wt. 18,000), constituting 80-85% of the total material, and two minor
bands with molecular weights of -16,000, and 9,000 (Ref. 140). As

(461) A. Luisada, Medicine, 20,229-250 (1941).

(462) W. P. Creger and H. Gifford, Blood, 7,721-728 (1952).
(463) K. L. Roth and A. M. Frumin,J. Lab. Clin. Med., 56, 695-700 (1960).
(464) E. Suescun and A. M. Frumin, Am. J . Clin. Pathol., 49,602-605 (1968).
(465) E. Suescun, E. A. Pachtman, and A. M. Frumin, Vox Sang., 18,77-80 (1970).
(466) M. Tomita, T. Kurokawa, K. Onozaki, T. Osawa, Y. Sakurai, and T. Ukita, 1nt.J.
Cancer, 10,602-606 (1972).
(467) N. Hashem and A. Kabarity, Lancet, (1) 1428-1429 (1966).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 203

identical patterns were obtained in the presence and absence of

2-mercaptoethanol, it appeared that there were no interchain disulfide
l ~ ~components of molecular weight 18,000and
bonds in the 1 e ~ t i n .The
9,000 stained with the periodic acid-Schiff reagent, indicating bound
carbohydrate. Molecular-weight determinations by sedimentation
equilibrium and X-ray diffraction gave values of 51,000 and 53,000,
respectively. These data are consistent with several possible subunit
structures, which include a tetramer (consisting of two chains of 18,000
and two of 9,000 molecular weight) and a more-unusual structure com-
posed of three polypeptide chains140of molecular weight 18,000. The
possibilities of other molecular fragments (for example, polypeptide of
molecular weight 16,000) cannot be excluded (compare Ref. 263). A
tetrameric structure composed of two polypeptide chains of molecular
weight 17,300 and two of molecular weight 14,300 was also proposed
by Allen and Rechromatography on Sephadex removed a
polypeptide (mol. wt. 9,000), believed to be a contaminant (compare
Ref. 140). Notable differences in the proportions of aspartic and
glutamic acids, and isoleucine and lysine were observed in the two
subunits.467aPreliminary, X-ray crystallographic data140indicated that
the fava lectin crystallized in the orthorhombic space group P212121.
Proceeding from the observation that 2-acetamido-2-deoxy-3-0-
methyl-D-glucose was a potent inhibitor of the fava-bean lectin, Allen
and purified the Vicia fubu phytohemagglutinin by
affinity chromatography, using 2-amino-2-deoxy-3-O-methyl-~-
glucose covalently attached through the amino group to CH-Sepharose
(an w-hexanoic acid derivative of agarose). The lectin, whose molecular
weight was determined by gel filtration to be -47,500, is believed to be
composed of two apparently identical subunits of molecular weight
24,000. Fragmented subunits, similar to those in con A and
soybean agglutinin, were found in active preparations of this lectin. A
glycoprotein, the agglutinin contains 2-acetamido-2-deoxy-~-glucose
and mannose. The carbohydrate moiety is presumably bound by way of
an N-asparaginyl linkage, as treatment with alkali failed to release
carbohydrate.
Studies on the carbohydrate-binding specificity of the fava-bean lec-
tin, as determined by hapten inhibition of hemagglutination,140.2’3.46*~469
showed that the lectin is inhibited by Makela’s group 111 sugars (see
Table VI). Hemagglutination by the lectin was inhibited by

(467a) H . J. Allen and E. A. Z. Johnson, Biochim. Biophys. Acta, 444,374-385 (1976).

(468) C. B. Perera and A . M. Frumin, Science, 151,821 (1966).
(469) J . K. N. Lee, E. A. Pachtman, and A. M. Frumin, Proc. Natl. Acad. Sci. U.S.A.,234,
161-169 (1974).
204 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEVI
Comparison of the Inhibitory Effect of Various Saccharides on the Agglutinating
Activity of Four D-Mannose/D-Glucose-specific Lectinsa

Concentration (mM) needed to produce

50% inhibition of agglutination

Lectin of
Inhibitor Concanavalin A L. culinaris P. sativum V .faba

D-Mannose 9 50 25 6.3
D-Glucose 25 50 25 25
3-O-Methyl-~-glucose >200 12.5 1.6 1.6
3-O-Benzyl-D-glucose 67 33 4.2 2.1
Methyl a-D-glucopyrano-
side 6.3 33 12.5 25
Methyl 2,3-di-O-methyl-a-
D-glucopyranoside 16 2.1 0.5 0.5
Methyl 2-acetamido-2-
deoxy-a-D-glucopyrano-
side 6.3 50 25 25
Methyl 2-acetamido-2-
deoxy-3-O-methyl-a-~-
glucopyranoside 200 4.6 1.6 2.3
Methyl 2-acetamido-2-
deoxy-4-O-methyl-a-~-
glucopyranoside 100 100 >200 >200
Methyl 2-acetamido-2-
deoxy-6-O-methyl-a-~-
glucop yranoside 200 200 >200 >200
Methyl 2-acetamido-2-
deoxy-P-D-glucopyrano-
side >200 >200 200 >200
Methyl 2-acetamido-2-
deoxy-3-O-methyl-P-~-
glucopyranoside >200 50 25 50
Trehalose 3.1 50 12.5 12.5
Kojibiose 2.1 8.3 2.1 4.2
Nigerose 8.3 100 67 33
Maltose 3.1 50 12.5 12.5
Isomaltose 1.2 19.5 4.9 9.8
Sophorose 4.2 50 17 17
Laminarabiose 50 67 67 33
Cellobiose >200 >200 200 200
Gentiobiose 100 200 100 100

“From Ref. 213. Published by permission of the Biochemical ]oumaZ.

 LECTINS: CARBOHYDRATE-BINDING PROTEINS 205

D-mannose, D-glucose, 2-acetamido-2-deoxy-~-glucose, D-fructose,

L-sorbose, maltose, and sucrose, but not by L-arabinose, D-XylOSe,
D-ribose, D-galactose, L-fucose, 2-amino-2-deoxy-D-galactose,
2-acetamido-2-deoxy-~-galactose, 2-amino-2-deoxy-~-g~ucose, lactose,
D-mannitol, or D-glucit01.~~**~~~ Methyl a-D-mannopyranoside, a,a-
trehalose, and melezitose were also strong inhibitors of agglutina-
tion.140A rough titration indicated that D-mannose is approximately four
times as inhibitory as D-glucose or maltose.'39 These data establish a
specificity for a-D-mannopyranosyl residues (2) similar to that of con A,
and place the lectin in the same class as those of the pea (Pisum sati-
vum) and the lentil (Lens culinaris). Sugar-inhibition of yeast-cell
agglutination gave similar results.469a
It has been shown that substitution of 0-3 of D-glucose and 2-acet-
amido-2-deoxy-~-glucoseby methyl or benzyl groups greatly en-
hances the binding of these sugars to the V. f a b a lectin213;this is also
true of the lentil and pea lectins, but not of con A. The best inhibitor of
all of the monosaccharide derivatives for the pea, the lentil, and the V.
f a b a lectins (but not con A) is methyl 2,3-di-O-methyl-a-~-
glucopyranoside, which is 25-50 times as inhibitory for the V. f a b a and
P . sativum lectins as methyl a-~-glucopyranoside.~~~ These results
suggest that there is a hydrophobic area in the lectin binding-site that
interacts with the methyl group of the 3-0-methyl derivatives and the
methylene (or benzyl) group of 3-0-benzyl-~-glucose.~'~ The 3-0-
substituted D-mannose derivatives have not yet been tested.
Table VI presents comparative inhibitory data on four
D-mannose(D-glucose)-binding l e c t i n ~Such . ~ ~ ~studies are exception-
ally useful in defining fine differences in the specificity of a group of
lectins having similar, but not identical, sugar-binding specificity.
A comparative study of the binding of four D-mannose(D-glucose)-
binding lectins (pea, lentil, con A, and V. f a b a ) to 6C3HED murine
ascites tumor-cells indicated that the pea, lentil, and V. f a b a lectins
bind to a group of cell-surface glycoproteins different from those to
which con A binds. Cell-surface glycoprotein heterogeneity was dem-
onstrated for con A and the pea l e ~ t i n . ~ ~ ~ , ~ ~ ~ ~
From the interest expressed in the fava-bean lectin, we may expect to
receive more-complete information on both its structure and
carbohydrate-binding specificity in the near future.

(469a) P. Ziska, Acta B i d . Med. Germ., 35, 1575-1576 (1976).

(46913) H. J. Allen and E. A. Z. Johnson, Biochim. Biophys. Acta, 436,557-566 (1976).
206 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

LECTINS
111. 2-ACETAMIDO-2-DEOXY-D-GLUCOSE-BINDING
1. Bandeiraea simplicqolia I1
(p-~-GlcNAcp= a-~-GlcNAcp)
A lectin having high affinity for nonreducing 2-acetamido-2-deoxy-
a- and -P-D-ghcopyranosyl groups was isolated from Bandeiruea
sirnplicifolia seed-extracts by affinity chromatography on chitin.125Elu-
tion of the chitin column with 2-acetamido-2-deoxy-~-glucose gave
virtually pure lectin, designated BS 11. Traces of the a-D-
galactopyranosyl lectin, also present in these seeds and designated BS I
(see Section V), were removed by passage of purified BS I1 through a
melibionate-Biogel P-300 c01urnn.l~~ The 2-acetamido-2-deoxy-~-
glucose-binding lectin is a tetrameric structure composed of four ap-
parently identical subunits of molecular weight -30,000; an aggregate
molecular weight of 113,000was determined by gel filtration. A glyco-
protein (4% of carbohydrate), BS I1 contains a high proportion of hy-
droxylic and acidic amino acids, and two methionine and three cysteine
residues per subunit (a disulfide bridge links two subunits). Each sub-
unit contains one carbohydrate-binding
The BS I1 lectin does not agglutinate A, B, or 0 erythrocytes, but will
agglutinate acquired-B, T-activated, and Tk polyagglutinable
ce11s.125,470BS I1 gave precipitin-like curves with p-azophenyl
2-acetamido-2-deoxy-a- and -p-D-glucopyranoside-bovineserum al-
bumin conjugates-the model substrates first employed in its detec-
tion, Shier's antigen A [N,N'-diacetylchitobiosyl-poly(L-aspartate)
p~lymer"~] also precipitated BS I1 1 e ~ t i n . Glycogen
l~~ and dextran
formed precipitates at high concentrations of 1 e ~ t i n . l ~ ~
Carbohydrate-binding specificity-studies were performed by sugar
inhibition of the lectin-p-azophenyl 2-acetamido-2-deoxy-~-~-gluco-
pyranoside-bovine serum albumin precipitating system. The most
effective monosaccharide inhibitor found (see Table VII) was 2-
acetamido-2-deoxy-~-glucose; it was over 400 times as inhibitory as
D-glucose. Of the common amino sugars examined, only the aforemen-
tioned 2-acetamido-2-deoxy-~-glucose exhibited lectin reactivity, es-
tablishing the necessity for an equatorial acetamido group at C-2 of the
D-hexopyranosyl ring. The failure of D-galactose and 2-acetamido-2-
deoxy-Dgalactose to inhibit the precipitin reaction also established a

(469c) S. Ebisu, P. N. Shankar Iyer, and I. J. Goldstein, Carbohyd. Res., 61, 129-138
(1978).
(470) W. J. Judd, M . L. Beck, B. L. Hicklin, P. N. Shankar Iyer, and I. J. Goldstein, Vox
Sang., 33,246-251 (1977).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 207

TABLEVII
Carbohydrate-binding Specificity of Bandeiraea simplicifolia
2-Acetamido-2-deoxy-~-glucose-binding Lectin"

Micromoles of inhibitor
Sugar inhibitors required for 50% inhibition
~ ~~

2-Acetamido-2-deoxy-~-glucopyranose 0.017
Methyl 2-acetamido-2-deoxy-cr-~-glucopyranoside 0.010
Methyl 2-acetamido-2-deoxy-fl-~-glucopyranoside 0.080
p - Nitrophenyl 2-acetamido-2-deoxy-a-D-glucopyranoside 0.005
p - Nitrophenyl2-acetamido-2-deoxy-fl-~-glucopyranoside 0.03
2-Acetamido-2-deoxy-D-mannose 17.0
2-Acetamido-2-deoxy-D-galactose 0% at 50 pmol
D-Glucose 7.4
Methyl a-Dglucopyranoside 1.3
Methyl 0-D-glucopyranoside 16.0
Methyl a-D-galactopyranoside 0% at 100 pmol
3-0-( 2-Acetamido-2-deoxy-a-~-glucopyranosy~)-~-g~uco-
pyranose 0.01
Maltose 1.8
Cellobiose 1.6
Gentiobiose 16.0
N, N'-Diacetylchitobiose 0.0045
N, N ', N"-Triacetylchitotriose 0.0062
~~

"Data taken from Ref. 125.

requirement for an equatorial 4-hydroxyl group. Comparison of the

inhibiting capacity of methyl or phenyl 2-acetamido-2-deoxy-a-D-
glucopyranoside with those of their respective anomers indicated that
the a anomer is bound six to eight times as avidly as the corresponding /3
anomer. The situation becomes complex on considering disaccharides.
Maltose and cellobiose are equivalent inhibitors (gentiobiose is a poor
inhibitor) and N,N'-diacetylchitobiose is about twice as potent as the
disaccharide having a nonreducing a-D-linked 2-acetamido-2-deoxy-
D-glucosyl group [a-~-GlcNAcp-( 1+3)-~-Glc]. The data suggest that
the BS I1 sugar-binding site is complementary to a nonreducing
2-acetamido-2-deoxy-a- or -~-D-g~ucopyranosyl group (3). BS I1 is the
first lectin described as having a primary specificity for both a- and
p-D-linked 2-acetamido-2-deoxy-~-glucose.Furthermore, a p-D-
(1+6)-glycosidic linkage seemingly destabilizes the lectin-carbohy-
drate complex; in this regard, BS I1 resembles the lectins from
Laburnum alpinum and Cytisus sessifoZius.'9~471

(471)W. M. Watkins and W. T. J. Morgan, Vox Sang., 7, 129-150 (1962).

208 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

NHAc
3

It has now been established that the BS I1 lectin interacts with the
same determinant in blood-group substances derived from hog and
human stomach linings as con A (see Section II,1, j), namely, nonreduc-
ing, terminal 2-acetamido-2-deoxy-a-~-glucopyranosyl groups ,471a

2. Cytisus sessilifolius
[P-D-GlcNAcp-( 1+4)-P-~-GlcNAc]
Renkonen originally discovered the strong anti-H(0) and anti-A,
serological activity of Cytisus sessilifolius extrack6 K o u l ~ m i e s , ~ ~ ~ , ~ ~
Kriipe7', and Make1a78later confirmed his findings. Saliva from secre-
tors (but not nonsecretors) was shown to inhibit the hemagglutination
a finding substantiated by precipitin band-formation in
Ouchterlony plates.474Although the agglutination of 0 cells by C. ses-
silijolius seed-extracts was neutralized by high dilutions of human and
hog H ( 0 ) substances,22~77~78~196~473~475
the agglutinin cannot be classified
with eel serum, Lotus tetragonolobus, and Ulex europeus I, because it
is not inhibited by L-fucose or its derivatives. Rather, it resembles Ulex
europeus I1 (Refs. 77, 208, and 225)and Laburnum alpinums*22*77~471;
N,N'-diacetylchitobiose best inhibits 0 erythrocyte hemagglutination
induced by these l e ~ t i n s . *The~ ~erythrocyte
' ~ ~ ~ ~ ~structure
~ with which
"Cytisus-type" (Ref. 208) anti-H(0) agglutinins react is not clear. The
complementary 2-acetamido-2-deoxy-P-D-glucosyl group may not
occur in a terminal, nonreducing position on the blood-group deter-
minant; in fact, internal 2-acetamido-2-deoxy-/3-~-glucopyranosyl
residues were demonstrated in soluble blood-group substance.24
Moreover, purified Bacillus fulminans a-L-fucosidase destroyed
0-erythrocyte agglutinability by Cytisus-type anti-H(0) lectins con-
comitantly with blood-group 0 reactivity.,08
(471a) C. Wood, E. Kabat, S. Ebisu, and I. J. Goldstein, manuscript submitted.
(472) R. Koulumies, Ann. Med. E x p . Bid. Fenn., 27, 185-188 (1949).
(473) R. Koulumies, Ann. Med. E x p . BioZ. Fenn., 28, 160-167 (1949).
(474) F. J. Grundbacher, Science, 81,461-463 (1973).
(475) V. P. Rege, T. J. Painter,W. M. Watkins,and W. T. J. Morgan,Nature, 203,360-363
(1964).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 209

Matsumoto and Osawa synthesized an affinity adsorbent for Cytisus

sessilifolius lectin by cross-linking N,N',N''-triacetylchitotriose to in-
soluble starch.'49Gel filtration of a fraction precipitated with ammonium
sulfate was followed by adsorption of the hemagglutinin to the affinity
r n a t r i ~ . ' ~Glycine-hydrochloric
~,~~~ acid buffer (pH 3.0) eluted a
homogeneous protein, as judged by ultracentrifugation and poly(ac-
rylamide) disc-gel electrophoresis (pH 8.9).The chemical composition
of this protein (molecular weight 110,000;~,~,,=6.8 S) was not deter-
mined, nor was its subunit structure investigated.
Carbohydrate-binding specificity studies have been conducted on
crude C.sessilifolius extracts by hemagglutination inhibition. K r i i ~ e ~ ~
first reported inhibition by salicin [o-(hydroxymethy1)phenyl p-D-
glucopyranoside], a finding confirmed by other^.'^^^^^'^^ The p-D-linked
D-glucobioses cellobiose and laminarabiose were also good in-
h i b i t o r ~ , ' ~ *whereas
~ ~ * ~ ~gentiobiose
' and s o p h ~ r o s e were
~ ~ ' poorly in-
hibitory. N,N'-Diacetylchitobiose conformed most closely to the lectin
saccharide binding-site (4) as judged by hemagglutination inhibi-

I NHAc

NHAc
4

tion.'g*471,475,477All a-D-linked disaccharides of D-glucose tested were

noninhibitor~.'~ A trisaccharide, believed to be 0-a-L-fucopyranosyl-
( 1+2)-0-~-D-galactopyranosyl-(1-*4)-2-acetamido-2-deoxy-D-g1ucose,
was also a good inhibitor, suggesting that an internal (1-*4)-2-
acetamido-2-deoxy-~-glucosylresidue could interact with the C.ses-
silifolius b i n d i n g - ~ i t ePhenyl
. ~ ~ ~ and p-nitrophenyl (but not m e t h ~ l ) ' ~ , ~ ~ ~
P-D-glycosides of D-glucose and 2-acetamido-2-deoxy-~-glucose were
comparable to (or somewhat better than) salicin as inhibitor^.'^^'^^
(476) I. Matsumoto and T. Osawa, Biochemistry, 13,582-588 (1974).
(477) T. J. Painter, V. P. Rege, and W. T. J. Morgan, Nature, 199,569-570 (1963).
2 10 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Phenyl and p-nitrophenyl P-D-galactopyranoside, as well as lactose,

were poorly i n h i b i t ~ r y . ' ~
The ~ lectin-reactivity reported for 2'-
, ~ ~high
0-L-fucosyllactose is interesting, but inconsistent with the observed
carbohydrate-binding specificity of the lectin (the authors might well
have tested the very active trisaccharide just noted).ls6
Makelazoand Osawal9 suggested the C. sessilifolius lectin as a rea-
gent for preliminary examination of a,p-anomerism in disaccharides of
D-glucose and 2-acetamido-2-deoxy-~-glucose. It should, however, be
noted that p-D-linked oligosaccharides must possess a P-D-(143 or
4)-glycosidic linkage in order to bind to the lectin ( ~ o p h o r o s eand ~~l
g e n t i o b i o ~ e lare
~*~ noninhibitors).
~~
In summary, both the physical-chemical and carbohydrate binding-
specificity studies on this interesting and important lectin are incom-
plete; additional characterization is required before the lectin can be
employed as a meaningful, structural probe.

3. Solanum tuberosum (Potato)

[potato; p-~-GlcNAcp -( 1+4)-[P-~-GlcNAcp-( 1+4)1,-/3-~-GlcNAc >
P-D-G~cNA -( 1+4)-P-~-GlcNAcp
~~ -( 1+4)-D-GkNAc]
The presence of a nonspecific, erythrocyte agglutinin in potato tub-
ers (Solanum tuberosum) was first recognized by Gelhorn16' in 1925.
Using a crude preparation from potatoes, M a r c u s s ~ n - B e g u ninvesti-
~~~
gated the agglutinability of a variety of animal red-cells and the inhibi-
tory effect of several different sera on the hemagglutination reaction.
Later, Kriipe77*480 reported that the potato lectin agglutinated erythro-
cytes of several mammalian species, including guinea pig, human (ir-
respective of blood-group type), mouse, rabbit, and sheep. Further-
more, he found that blood-group substances, but not simple sugars,
inhibited a g g l u t i n a t i ~ n . ~ ~ ~
Although Singh and coworkers481 obtained an active, agglutinin prep-
aration following acetone precipitation, M a r i n k o v i ~ hfirst~~~ seriously
attempted to purifjl the potato lectin by employing acetone fractiona-
tion and cellulose ion-exchange chromatography. Homogeneity, by
electrophoresis in starch gel and by ultracentrifugation, was dem-
onstrated. He determined the glycoprotein nature of the lectin;
chemical analysis showed arabinose, a preponderance of acidic amino
acids, and a high molar content of cysteine residues.48zThe agglutinin

(478) G. BBtail and M. Coulet, C. R. Acad. Sci., 168,295-299 (1974).

(479) H. Marcusson-Begun, 2. Immunitaetsforsch. Exp. Ther., 45,49-73 (1926).
(480) M. Krupe, Hoppe-Seyler's Z . Physiol. Chem., 299,277-282 (1955).
(481) G. Singh, S. D. Verma, and G. W. G. Bird, Z. Immunitaetsforsch.Exp. Ther., 121,
181-184 (1961).
(482) V. A, Marinkovich,]. Immunol., 93, 732-741 (1964).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 211

was inactivated when treated with 0.2M 2-mercaptoethanol for 90 min

at 37". No specificity for A, B, 0, MN, or Rh antigens was exhibited.
Inhibition of hemagglutination could not be demonstrated, using a
variety of mono- and di-saccharides and saliva specimens.482Despite an
earlier reporP on the heat stability of potato extract, Marinkovich
that his preparation rapidly lost activity on heating to 80".
Pardoe and coworkerszo6observed that treatment of a series of ten
different, animal erythrocytes with pronase or neuraminidase had vir-
tually no effect on the agglutination titer of the potato lectin.
By employing conventional, protein-purification techniques (am-
monium sulfate fractionation, and DEAE-cellulose, CM-cellulose, and
Sephadex chromatography),Allen and NeubergerZo7 purified the potato
lectin to homogeneity (yield: 38 mg from 4.5 kg of potato tubers), and
showed it to be a glycoprotein containing -50% (by weight) of car-
bohydrate. The authors suggested that the lectin was composed of two
(or possibly more) identical subunits (molecular weight -46,000),
bound noncovalently, giving a molecular weight of 85,000to 100,000.
Attempts to purify the potato lectin on ovomucoid- and N,N'-
diacetylchitobiose-substituted Sepharose failed; the lectin was so
strongly bound that it could not be removed from the adsorbent by
displacement with urea.483To circumvent this difficulty, Delmotte and
employed p-aminobenzyl2-acetamido-2-deoxy-l-thio-~-
D-glucoside-substituted Sepharose, and eluted the lectin with 0.1 M
acetic acid (yield: 58 mg from 128 g of the protein in potato-tuber
extract).
Arabinose was the preponderant sugar in the glycosyl part of the
potato lectin, along with smaller proportions of galactose, glucose, and,
possibly, 2-acetamido-2-deoxy-~-glucose.~~~ Because the carbohydrate
residue survived dialysis against 0.5M sodium hydroxide solution, the
presence of alkali-labile,484glycosidic linkages involving the hydroxyl
groups of serine and threonine was excluded. Allen and Neuberger207
suggested that arabinosyloxy-L-proline constituted the major carbo-
hydrate-protein linkage; this linkage occurs in the cell wall of many
plants, including the potato.485,486
The amino acid composition of the potato lectin is unusual in a
number ofrespects: the most abundant amino acid is hydroxy-L-proline
(the first lectin reported to contain this amino acidzo7),11.5% of the
residues are half-cystine (compare wheat-germ agglutinin, with 20%of

(483) F. Delmotte, C. Kieda, and M. Monsigny, FEBS Lett., 53, 324-330 (1975).
(484) R. D . Marshall and A. Neuberger, Adv. Carbohydr. Chem. Biochem., 25,407-478
(1970).
(485) D. T. A. Lamport, Biochemistry, 8, 1155-1163 (1969).
(486) M. F. Heath and D. H. Northcote, Biochem. J., 125, 953-961 (1971).
212 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

h a l f - c y ~ t i n e ' ~ ~L-phenylalanine
~'~~), is absent, and ornithine is a possi-
ble, although disputed, constituent of the lectin.204s483
The carbohydrate-binding specificity of the potato lectin has been
studied by examining the extent to which carbohydrates of known
structure inhibit hemagglutination of rabbit or human erythrocytes (see
Table VIII). The Solanum tuberosurn agglutinin is specifically inhib-
ited by oligosaccharides containing 2-acetamido-2-deoxy-~-glu-
cose,20G*207,483 but not by 2-acetamido-2-deoxy-~-glucose itself77*207*480
(see, however, Ref. 483), or its methyl a- and p-glycopyrano~ides.~~~
Several earlier investigators were similarly unable to demonstrate in-
hibition of hemagglutination by monosaccharides .77*480 A series of di-
saccharides containing a 2-acetamido-2-deoxy-~-glucoseresidue in the
reducing position, namely, /3-D-Galp-(1+3)-~-GlcNAc, a-D-Galp-
(1+3)-~-GlcNAc,and P-D-Galp-(1+4)-~-GlcNAc,also failed to inhibit
the hemagglutinin reaction between potato lectin and animal erythro-
cytes treated with pronase and neuraminidase.20G
Interestingly, Matsumoto and OsawaIg6reported that p-nitrophenyl
(and phenyl) 2-acetamido-2-deoxy-~-~-glucopyranoside, at a concen-
tration of 10 mg/ml, did not inhibit the agglutination reaction between
the potato lectin and human erythrocytes, whereas both p-nitrophenyl
2-acetamido-2,6-dideoxy-~-~-glucopyranoside and phenyl 2-acet-
amido-3-0-(D-l-carboxyethyl)-2-deoxy-~-~-glucopyranoside, at a level
of 5 mg/ml, completely inhibited the hemagglutinin reaction. The

TABLEVIII
Inhibition, by Various Sugars, of the Hemagglutinating Activity of the
Potato (Solanum tuberosum) Lectin"

Concentration required
Compound for 50% inhibition (mM)
~~ ~~

2-Acetamido-2-deoxy-~-glucose 0% at 200 mM
Methyl 2-acetamido-2-deoxy-a-~-glucopyranoside 0% at 200 mM
Methyl 2-acetamido-2-deoxy-~-~-glucopyranoside 0% at 200 mM
Benzyl 2-acetamido-2-deoxy-ar-~-glucopyranoside 40
N, N'-Diacetylchitobiose 0.1
N, N',N"-Triacetylchitotriose 0.05
N,N', N", N"'-Tetraacetylchitotetraose 0.005
N,N',N",N"', N""-Pentaacetylchitopentaose 0.002
p-~-GlcNAcp-(1+4)-MurNAcb 8
p-~-GlcNAcp-(1+4)-p-MurNAcp-( 1+4)-P-~-GlcNAcp-
( 1+4)-MurNAc 0.1
P-D-Galp-(1+4)-~-GkNAc O%at3mM
p-~-Glcp-( 1-+4)-D-Glc 0% at 200 mM

"Data taken from Ref. 207. "MurNAc = N-acetylmuramic acid.

 LECTINS: CARBOHYDRATE-BINDING PROTEINS 2 13

interpretation of these data requires further clarification, especially in

view of the observation that benzyl 2-acetamido-2-deoxy-a-~-g1uco-
pyranoside was reported by Allen and NeubergerZo7to be a weak
inhibitor.
The inhibitory power of the N-peracetylated chito-oligosaccharides
increases with increase in chain length up to the tetraose; this suggests
an extended b i n d i n g - ~ i t e ~similar
~ ~ ~ ~to~ that
~ * ~proposed
*~ for wheat-
germ agglutinin.lZ8However, the combining site of the potato lectin
may be somewhat larger, as it is complementary to a tetrasaccharide
rather than a trisaccharide. Delmotte and coworkers483observed that
the aromatic ring in p-nitrophenyl glycosides of N-peracetylated
chitobiose and chitotriose enhanced the binding affinity of these
ligands to the lectin. Like wheat-germ agglutinin, cellobiose and
N-acetyllactosamine206~zo7 [P-D-Galp-(1+4)-~-GlcNAc]were not inhib-
itors of the potato lectin. It is noteworthy that the bacterial cell-wall
oligosaccharides, [p-~-GlcNAcp-( 1+4)-P-~-MurNAc], (n = 1,or 2) (see
Table VIII), are weaker inhibitors than the corresponding chito-oligo-
saccharide h o m o l ~ g s . ~ ~ ~
In summary, the lectin from potato tubers (Solanurn tuberosum) is a

h;'
I
CH,OH

NHAC

NHAc 5
2 14 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

nonspecific, blood-group agglutinin that exhibits a carbohydrate-

binding specificity similar to that of wheat-germ agglutinin, in that it
binds to /3-~-(1+4)-linked oligomers of 2-acetamido-2-deoxy-D-
glucose. Its binding site can accommodate a tetrasaccharide (5) (possi-
bly, a pentasaccharide). Finer details of this interesting glycoprotein
are still lacking, and it is hoped that further studies will be forthcoming
so that its potential as a structural probe may be realized.

4. Triticum uulgaris (Wheat Germ)

[wheat germ; j3-~-GlcNAcp-( 1+4)-/3-~-GlcNAcp-(1+4)-GlcNAc >
p-D-GlcNAcp-( 1+4)-/3-D-GlcNA~]
One of the first reports of the interaction between carbohydrate-
binding proteins and tumor cells was made by Aub-andcolleaguesz8in
1963; a wheat-germ, lipase preparation agglutinated malignant cells.
Subsequently, Burger and Goldberg30reported that virally transformed
cells were agglutinated to a greater extent than their respective, parent
cell-lines.
Wheat-germ agglutinin (WGA) is not a blood-group-specific
hemagglutinin; it will agglutinate all types of human erythrocytes, as
well as a variety of normal and neoplastic animal cells.
Some years later, the agglutinating principle of wheat germ was
isolated, and crystallized, by Nagata and BurgeP' and LeVine and
coworkers48s;both studies reported a glycoprotein having a molecular
weight of 23,500. Others purified WGA b y conventional
techniques,'z8~'z9~48g*4go and by affinity chromatography on synthetic ad-
sorbents: [6-aminohexyl 2-deoxy-/3-~-urubino-hexopyranoside-Seph-
arose 4B (Ref. 130), p-aminobenzyl 2-acetamido-2-deoxy-l-thio-/3-~-
g l u c o p y r a n 0 s i d e , 4 ~2-acetamido-N-(6-aminohexanoyl)-2-deoxy-/3-
~~~~~~~
D-glucopyranosylamine-Sepharose 4B (Ref. 491)], immobilized
o v o m ~ ~ o iand d chitin.'=
, ~ ~ ~ Careful
~ ~ ~analysis
~ ~ ~ revealed
~ ~ the protein

(487) Y. Nagata and M. M. Burger,]. Biol. Chem., 247,2248-2250 (1972).

(488) D. LeVine, M. J. Kaplan, and P. J. Greenaway, Biochem. J . , 129,847-856 (1972).
(489) B. Ozanne and J. Sambrook, Nature (London) New Biol., 232,156-160 (1971).
(490) Y. Nagata, A. R. Goldberg, and M. M. Burger, Methods Enzymol., 32, 611-615
(1974).
(490a) M. E. Rafestin, A. Obrenovitch, A. Oblin, andM. Monsigny,FEBS Lett., 40,62-66
(1974).
(490b) P. Bouchard, Y. Moroux, R. Tixier, J.-P. Privat, and M. Monsigny, Biochimie, 58,
1247-1253 (1976).
(491) R. Lotan, A. E. S.Gussin, H. Lis, and N. Sharon, Biochem. Biophys. Res. Commun.,
52,656662 (1973).
(492) M. M. Burger, Proc. Natl. Acad. Sci. U.S.A., 62, 994-1001 (1969).
(493) V. T. Marchesi, Methods Enzymol., 28, Part B, 354-356 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 215

to be devoid of covalently bound c a r b ~ h y d r a t e ' ~ ~ -hence,

~ ~ ~ . ~WGA
~~~;
is not a glycoprotein. Hemagglutinating activity was associated with
several fractions eluted from columns of SE-Sephadex and SP-
S e p h a d e ~ . ' ~Rice
~ . ~and
~ ~ E t ~ l e 1 3isolated
~~ 200-250 mg of purified
WGNkg of raw wheat-germ (Bouchard and coworker^^^^^ reported 500
mg of WGNkg of wheat germ); this was distributed among four active
fractionP5 as follows: 35% of I, 50% of 11,, 5% of IIb, and 10% of 111.
WGA fractions I, 11,, and I11 were indistinguishable, focusing in a
sharp band at pH 8.7k0.3 in isoelectric focusing experiments; WGA IIb
focused at pH 7.7 k0.3. The amino acid compositions of the four,
chromatographically distinct, forms indicated considerable similarity
(although fraction I contained no histidine), and were comparable to
related results of others.128,12s,133,4s1 There was a very high content of
half-cystine (but no free sulfhydryl groups) and glycine. A
genetic basis for the occurrence of multiple forms of WGA has been
advanced.494a
There is now general concurrence that WGA is a dimeric protein, of
molecular weight 36,000, composed of two identical subunits (molecu-
lar weight 18,000).129,335*495 Denaturants (8 M urea), pH extremes (pH
1-11), or high concentrations of salt promoted subunit exchange among
WGA I, 11,, and 111, or with chemically modified (N-acetyl and
N-methyl derivatives), electrophoretic variants to give hybrid aggluti-
nins. These experiments provided further evidence for the structural
similarity and dimeric nature of the three WGA forms.335All four isolec-
tins of wheat-germ agglutinin exhibited similar c.d.
Analysis of the c.d. spectrum in the ultraviolet region indicated -12%
of p-structure, and no evidence of an a-helical The
X-ray crystallographic analysis of the agglutinin has been solved at
~ ~ * ~ ~ ~ ~agglutinin dimers (4 x 4 x 7
200-pm r e ~ o l u t i o n . Wheat-germ
nm) are composed of two closely associated protomers, each consist-
ing of 164 amino acids. Each protomer consists of four structurally
homologous and spatially distinct domains.49saCrystals of the agglu-
tinin containing the 4-methylumbelliferyl P-glycoside of N,N'-di-
acetylchitobiose have been prepared.

(494) J.-P. Privat, F. Delmotte, G. Mialonier, P. Bouchard, and M. Monsigny, Eur. J .

Biochem., 47,5-14 (1974).
(494a) R. H. Rice, Biochim. Biophys. Acta, 444, 175-180 (1976).
(495) R. H. Rice and M. E. Etzler,Biochem. Biophys. Res. Commun.,59,414-419(1974).
(495a) M. W. Thomas, E. F. Walborg, Jr., and B. Jirgensons,Arch. Biochem. Biophys.,
178,625-630 (1977).
(496) C. S. Wright, C. Keith, R. Langridge, Y. Nagata, and M. M. Burger,J.Mol. Biol., 87,
843-846 (1974).
(496a) C. S. Wright,J. Mol. Biol., 111,439-457 (1976).
216 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Equilibrium-dialysis studies on the number of carbohydrate-

binding sites per subunit have been conducted, but the results are
somewhat discordant. Thus, LeVine and coworkers,488 using
2-acetamido-2-deoxy-~-glucose, reported one binding site per subunit
of molecular weight 23,000, whereas Nagata and BurgerlZ9found two
binding sites per molecular equivalent of subunit. In order to exclude
the possibility that two molecules of 2-acetamido-2-deoxy-~-glucose
might bind to one subunit within the same binding site (see later),
Privat and coworker^"^ studied the binding of NaBT4-reduced
N,N”N”,N’”-tetraacetylchitotetraose to WGA. In agreement with the
results of Nagata and Burger,lz9 two moles of tetrasaccharide were
bound per mole of polypeptide chain. By means of a fluorescence-
quenching technique, the same investigators demonstrated that two
moles of the 4-methylumbelliferyl P-glycosides of 2-acetamido-2-
deoxy-D-glucose, N,N’-diacetylchitobiose, and N , N ’ ,N ’ ’-triacetylchito-
triose were similarly bound to a WGA monomeric unit? (compare
Ref. 497a). Inasmuch as the fluorescence of all of these glycosides was
quenched to the same extent (as opposed to the findings of Van Land-
schoot and it was proposed that the aglycon occupied
the same subsite in the three cases, and that the lectin binding-site
contains only three subsites. If it is confirmed that there are two iden-
tical and noninteracting subsites per subunit, it would be expected
that there is an element of symmetry in each polypeptide chain; for
example, the subunit might consist of two (identical) halves resulting
from gene duplication. Indeed, X-ray crystallographic data suggest the
possibility of gene quadruplication followed by divergent e v o l ~ t i o n . ~ ~ ~ ~
Studies on the specific precipitation of WGA by glycoproteins,
glycolipids, bacterial cell-wall peptidoglycan, and model substrates
have been reported. Shier1I4chemically linked N,N ’-diacetyl-p-chito-
biosylamine to poly(L-aspartic acid), to afford a model substrate termed
antigen A [poly(N,N’-diacetyl-p-chitobiosylaspartate)] which formed a
precipitin band1I4 with WGA. A saturated solution of N,N’-diacetyl-
chitobiose completely inhibited the reaction. Antigen A gave4s8a pre-
cipitin curve with WGA. Similarly, Watanabe and Hakomori dem-
o n ~ t r a t e da ~precipitin
~~ reaction by agar-gel diffusion between WGA
and purified glycosphingolipids from human adenocarcinoma and
blood-group A erythrocytes. The latter, blood-group A sphingolipid

(497) J.-P. Privat, F. Delmotte, and M. Monsigny, FEBS Lett., 46,229-232 (1974).
(497a) A. Van Landschoot, F. G . Loontiens, R. M. Clegg, N. Sharon, and C. K. De
Bruyne, Eur. J . Bfochem., 79,275-283 (1977).
(498) I. J. Goldstein, S. Hammarstrom, and G . Sundblad, Biochim. Biophys. Acta, 405,
53-61 (1975).
(499) K. Watanabe and S. Hakomori, FEBS Lett., 37,317-320 (1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 217

was composed of 1.0 fucose, 0.9 glucose, 2.1 galactose, 0.92-amino-2-

deoxyglucose, and 1.7 2-amino-2-deoxygalactoseunits. Smith degrada-
tion caused decomposition of fucose and 2-amino-2-deoxygalactose,
whereas the reactivity with WGA intensified, suggesting that the site
recognized b y WGA may be in the interior chain.499A ceramide
trisaccharide [O-2-acetamido-2-deoxy-~-~-glucopyranosyl-( 1+3)-0-@-
D-galactopyranosyl-(1+4)-p-~-glucopyranosylceramide]isolated from
degraded, blood-group glycolipid was also strongly WGA-reactive,
suggesting that WGA can bind nonreducing (terminal) 2-acetamido-2-
deoxy-D-glucosyl groups as In order to form a lattice with the
lectin, the glycolipid is most probably present in the micellar form, in
which case, multiple 2-acetamido-2-deoxy-D-glucosyl groups would be
available for interaction with the protein.
Privat and coworker^"^ diazotized p-aminophenyl N,N'-diacetyl-
p-chitobioside to bovine serum albumin, and demonstrated that it
precipitated with WGA in agar gel; N,N'-diacetylchitobiose inhibited
this interaction.
Carcinogenic embryonic antigen (CEA) precipitated with WGA in
agar gels4B9and in solution.498CEA that had undergone three Smith-
degradation cycles also gave a precipitin curve.498As may be seen from
Fig. 10, each subsequent degraded product precipitated with WGA,
including the products after periodate oxidation and reduction, but
prior to mild, acid hydrolysis.

- 6

-m
3.5
'EI

.-.-'04
c
a
g 3

g 2
L
e
21

10 20 30 40 5 0 60 70 00
CEA antigen ( p g )

.,
FIG. 10.-Precipitation of Carcinoembryonic Antigen (CEA) and its Smith-degraded
Products by Wheat-germ Agglutinin.*8(0,Native CEA; 0 ,periodate-oxidized, borohy-
dride-reduced CEA; V, Smith-degraded CEA, stage I; V, periodate-oxidized, boro-
hydride-reduced, Smith-degraded CEA I; two cycles of Smith-degraded CEA,
stage 11; 0 , periodate-oxidized, borohydride-reduced, stage 11; A, three cycles of
Smith-degraded CEA. Conditions: 30 pg of wheat-germ agglutinin per tube; total
volume, 200 pl. Reproduced by permission of Biochim. Biophys. Acta.)
218 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

p-Azophenyl 2-acetamido-2-deoxy-~-~-glucopyranoside-bovine
serum albumin conjugate precipitated with WGA, and this system was
used to investigate the sugar-binding specificity of the agglutinin.498
Interaction of WGA with a soluble, linear peptidoglycan secreted by
Micrococcus luteus, and with the teichoic acid of Staphyloccus aureus
H , was demonstrated by agar-gel diffusion, quantitative precipitation,
and inhibition of trypsinized rabbit-erythrocyte h e m a g g l u t i n a t i ~ n . ~ ~ ~
No interaction was observed with the teichoic acid from a phage-
resistant mutant (S. aureus 52A2) that lacked 2-acetamido-2-deoxy-~-
glucose residues.
Burger and Goldberg30 first reported that, of the sugars found in
glycoproteins, only 2-acetamido-2-deoxy-~-glucoseand N,N’-diacetyl-
chitobiose inhibited the agglutination of polyoma virus-transformed
BHK cells and chemically-induced L1210 leukemia cells (compare
Refs. 501-504).
Sugar inhibition of rabbit erythrocyte agglutination by WGA dem-
onstrated that, of the common amino sugars, 2-acetamido-2-deoxy-~-
glucose alone inhibited the agglutination reaction.lZ8Neither 2-amino-
2-deoxy-D-glucose,2-acetamido-2-deoxy-~-galactose, nor S-acetamido-
2-deoxy-~-mannosewas inhibitory, demonstrating the necessity for an
equatorial hydroxyl group at C-4 and the D-glyCerO configuration at C-2
(compare Ref. 498).WGA exhibited little or no anomeric preference for
methyl glycopyranosides of 2-acetamido-2-deoxy-~-glucose.~~~~~~~
Employing a precipitation reaction between WGA and p-azophenyl
2-acetamido-2-deoxy-~-~-glucopyranoside-bovine serum albumin
conjugate, Goldstein and coworkers showed that p-nitrophenyl
2-acetamido-2-deoxy-~-~-glucopyranoside was a poor inhibitor, and
the methyl a- and P-glycopyranosides of D-glUCOSe, D-mannose, and
D-galactose were noninhibit01-s.~~~ (Earlier, the same p-azophenyl
2-acetamido-2-deoxy-~-~-glucopyranoside-bovine serum albumin
conjugate was shown to inhibit WGA-induced h e m a g g l u t i n a t i ~ n . ~ ~ )
Like 2-amino-2-deoxy-~-glucose, 2-acetamido-l-(~-aspart-4-oyl)-2-
deoxy-P-D-glucopyranosylaminewas a noninhibitor of both hemag-
glutination128and p r e c i p i t a t i ~ n .These
~ ~ ~ data suggest that an un-
charged amide group (for example, N-acetyl or N-propionyl group)

(500)R. Lotan, N. Sharon, and D. Mirelman, Eur. J . Biochem., 55,257-262 (1975).

(501) R. Liske and D. Franks, Nature (London), 217,860-861 (1968).
(502)G.Uhlenbruck, G.I. Pardoe, and G . W. G.Bird, Nuturwissenschuften, 55,347
(1968).
(503)G.Uhlenbruck,W.Gielen, and G . I. Pardoe,Z. Krebsforsch., 74,171-178(1970).
(504)R.Lotan and N. Sharon, Biochem. Biophys. Res. Commun., 55,1340-1346(1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 219

must be present at C-2 in the D-glycero configuration; neighboring, free

carboxyl and amino groups interfere with the binding phenomenon.
On the other hand, uncharged 2-acetarnido-N-acetyl-2-deoxy-~-~-
glucopyranosylamine inhibited to the same extent as the a- or p-
glycopyranosides of 2-acetamido-2-deoxy-~-glucose,indicating that
nitrogen may replace the glycosidic oxygen atom without any sacrifice
in inhibitory potency.128Nonspecific binding of N-acetylneuraminic
acid, possibly to the same site as that to which 2-acetamido-2-deoxy-~-
glucose binds, was also reported505(compare Refs. 505a-d).
Methylation at 0 - 4 or 0 - 6 (for example, 2-acetamido-2-deoxy-4- or
-6-O-methyl-~-glucose)had no effect on inhibitory activity, whereas
similar substitution at 0 - 3 gave an inactive derivative.128These results
suggest that oligosaccharides containing 4- and 6-0-substituted
(perhaps, also, 4,6-di-0-substituted) 2-acetamido-2-deoxy-~-glucose
residues may bind to WGA; this has been substantiated for certain 4-0-
substituted 2-acetamido-2-deoxy-~-glucoseoligosaccharides,'28*4Qsbut
has yet to be demonstrated for the 6-0-substituted derivative (see
later).
Of the oligosaccharides examined for their capacity to inhibit WGA,
only those containing 2-acetamido-2-deoxy-~-glucoseresidues proved
effective. Lactose, cellobiose, cellotriose, and N-acetyllactosamine (2-
acetamido-2-deoxy-4-O-~-D-ga~actopyranosyl-~-g~ucose) were all in-
active as inhibitors of the p-azophenyl 2-acetamido-2-deoxy-P-~-
glucopyranoside-bovine serum albumin-WGA reaction.4Q8Table IX
presents representative inhibition data.
By far the most interesting series of WGA sugar inhibitors are the
N-peracetylated chito-oligosaccharides. Confirming Burger and
Goldberg's results,3OWGA showed a much higher affinity forN,N'-diace-
tylchitobiose than for 2-acetamido-2-deoxy-~-glucose.
In fact, this affinity increased as the homologous series was ascended,
such that the biose was several hundred and the triose several thousand
times more tightly bound than 2-acetamido-2-deoxy-D-glucose. Com-
parative data, from studies in five different systems, on WGA binding of
N-peracetylated chito-oligosaccharides have been tabulated.4g8Al-
though, in some systems, N-peracetylated chitotetraose and chito-

(505) P. J. Greenaway and D. LeVine, Nature (London) New Biol., 241,191-192 (1973).
(505a) P. Cuatrecasas, Biochemistry, 12, 1312-1323 (1973).
(505b) W. R. Redwood and T. G. Polefka, Biochim. Biophys. Acta, 455,631-643 (1976).
(505c) D. H. Boldt, S. F. Speckart, R. L. Richards, and C. R. Alving, Biochem. Biophys.
Res. Commun., 74,208-214 (1977).
(505d) V. P. Bhavanandan, J. Umemoto, J. R. Banks, and E. A. Davidson, Biochemistry,
16,4426-4437 (1977).
220 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEIX
Carbohydrate-binding Specificity of Wheat-germ Agglutinin

Concentration required
for 50% inhibition
Inhibitor nmoV2OO pla mM*

N,N', N", N"', N""-Pentaacetylchitopentaose 1.0 0.005

N, N ',N", N"'-Tetraacetylchitotetraose 1.3 0.01
N,N',N"-Triacetylchitotriose 1.6 0.01
N , N'-Diacetylchitobiose 45 0.5
N,N'-Diacetylchitobiitol 500
Chitobiose >4,500
a-D-Manp-(1+3)-,9-D-Manp-( 1+4)-~-GlcNAc >700
P-DManp-( 1+4)-P-~-GlcNAcp-(1 + 4 ) - p ~ -
GlcNAcp-(l+N)-Asn 33
p-~-ClcNAcp-(1+4)-/3-~GlcNAcp-(l+N)-Asn 39
2-Acetamido-N-(4-~-aspa1toyl)-2-deoxy-/3-~-gluco-
pyranos ylamine 350 0% at 35 mM
p-Nitrophenyl 2-acetamido-2-deoxy-/3-~-gluco-
pyranoside 1,100
Methyl 2-acetamido-2-deoxy-/3-~-glucopyranoside 3,300 10
Methyl 2-acetamido-2-deoxy-a-~-glucopyranoside 5,000 10
2-Acetamido-2-deoxy-~-glucose 5,900 30
2-Amino-2-deoxy-~glucose >63,000 0% at 200 mM
2-Acetamido-2-deoxy-~-galactose 31,000 200
Methyl 2-acetamido-2-deoxy-3-O-methyl-a-~-
glucopyranoside 200
Methyl 2-acetamido-2-deoxy-4-O-methyl-a-~-
glucopyranoside 10
Methyl 2-acetamido-2-deoxy-6-O-methyl-a-~-
glucopyranoside 10
Cellobiose >2,000
Cellotriose >3,000

"Inhibition of wheat-germ agglutinin-p-azophenyl 2-acetamido-2-deoxy-/3-~-gluco-

pyranoside-bovine serum albumin precipitation by various saccharides. Data taken
from Ref. 498. bInhibitory effect of various sugars on the agglutinating activity of wheat-
germ agglutinin against rabbit erythrocytes. Data from Ref. 128.

pentaose showed a somewhat higher affinity for WGA than did

chitotriose, the WGA combining-site appears to be complementary to a
sequence of three @-D-( l-+4)-linked 2-acetamido-2-deoxy-~-glucose
units (6), with additional glycosyl residues adding little to the free
energy of binding.
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 221

0
CH,OH

0‘ I
NHAc

NHAC
6

Studies on the homologous, N-peracetylated chito-oligosaccharides

led Allen and coworkerslZ8to postulate that the WGA binding-site
consists of a system of adjacent subsites similar to that assumed for
hen’s egg-white lysozyme506(see Fig. 11).They observedlZ8that the
bacterial cell-wall disaccharide [P-DGlcNAcp-(1+4)-MurNAc] was a
weaker inhibitor than N,N’-diacetylchitobiose, whereas the tetrasac-
charides (N-peracetylated chitotetraose) and [p-D-GlcNAcp-(1-4)-
MurNAcplz were about equally as active as inhibitors. Subsites A, B,

Subs it e : A B C D
D-G~cNAcp - 0 - R
8- D - GlCNAC p - (1- 4) - D -GlCNAC
6-D - GlCNACp - (1- 4 ) -@- D - GlCNACp - (1- 4)- D - GlCNAC
8- D - GlcNAc p - (1-4LMurNAc
4-D - GlcNAc p - (l+ 4)-p-MurNAcp-(l-c 4)-p-D -GlcNAc p- (l--l)-MurNAc
FIG. 11.-System of Subsites Proposed for the Binding Site of Wheat-germ Agglu-
tinin.lZ8

(506) C. C. F. Blake, L. N. Johnson, G. A. Mair, A. C. T. North, D. C. Phillips, andV. R.

Sarma, Proc. R. Soc. London Ser. B , 167,378-388 (1967).
222 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

and C were envisaged as binding 2-acetamido-2-deoxy-~-glucose

residues, and subsite D as accommodating aglycons of the glycosides.
They also suggestedlZ8that 2-acetamido-2-deoxy-~-glucose itself is
bound only to subsite C, and that its 3-hydroxyl group must be
unsubstituted. On the other hand, oligosaccharides composed of 2-
acetamido-2-deoxy-~-glucoseresidues may bind to WGA even if the
residue that occupies subsite B contains a substituent at (2-3, as in N -
acetylmuramic acid. Thus, as indicated previously, bacterial cell-wall
tetrasaccharide binds as well as N-peracetylated chitotetraose.lZ8
The relative affinity of WGA for the mono-, di-, and tri-saccharide
differs according to the system employed. Although the reason for these
differences is not yet clear, it seems probable that they are related to the
ability of the “substrate” used in the inhibition assays to occupy one or
more subsites in WGA. Whereas fluorescence enhancement and
equilibrium dialysis are direct measurements of binding strengths,
inhibition assays depend on the ability of the hapten to displace the
“substrate” from the site. In the case of inhibition ofprecipitation using
p-azophenyl P-D-GlcNAc-bovine serum albumin, it may be assumed
that only one subsite is When 2-acetamido-2-deoxy-~-
glucose is added as an inhibitor, both competitive and noncompetitive
binding of 2-acetamido-2-deoxy-~-glucose to the agglutinin may occur,
thus increasing the concentration needed for 50% inhibition. In con-
trast, only competitive binding will occur with N , N ’ , N ’ ’ -
triacetylchitotriose, assuming that WGA has three subsites. On the
other hand, N,N’-diacetylchitobiose may bind both noncompetitively
and competitively. However, as compared to 2-acetamido-2-deoxy-~-
glucose, the relative contribution of noncompetitive binding to total
binding will be smaller with the b i o ~ e . ~ O ~
Like 2-amino-2-deoxy-~-glucose,the chito-oligosaccharides having
unsubstituted 2-amino groups do not bind to WGA, due, presumably, to
repulsion of their positively charged -NH3+groups.128*408
The ability of WGA to interact with P-D-Manp-(1+4)-p-~-ClcNAcp-
(1+4)-P-~-GlcNAcp-(l+N)-Asn and p - ~ - G l c N A c p -1-+4)-P-~- (
GlcNAcp-(l+N)-Asn with approximately the same affinity as N , N ’ -
d i a c e t y l c h i t o b i ~ s efurther
~ ~ ~ supports the observations of Privat and
coworker^"^ and Shier114;macromolecules containing multiple N,N’-
diacetylchitobiosyl residues precipitate with the agglutinin.
The carbohydrate-binding sites of WGA are probably situated at the
surface of the protein molecule. The pH dependences of association
constants for lectin-chitotriose binding indicate that an ionizable
group, pK = 3.9, is probably involved in complex-formation.404*504 This
observation is especially noteworthy, inasmuch as protein carboxyl
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 223

groups have also been implicated in the carbohydrate-binding sites of

con A (see Ref. 326). Reductive methylation (treatment with HCHO in
the presence of NaBH4)had no detectable effect on the carbohydrate-
binding ability of the protein.33sSubsequent 0-acetylation with acetic
anhydride prevented binding of the protein to ovomucoid-Sepharose
and greatly lowered its erythrocyte-agglutinating ability; inhibition
was reversible by treatment with h y d r o ~ y l a m i n eSimilarly,
.~~~ acetyla-
tion of native WGA, modifying available amino and phenolic groups,
greatly lowered the agglutinating ability.335Regeneration of free
tyrosine by brief treatment with hydroxylamine restored the ability of
WGA to agglutinate erythrocytes. These experiments suggest that free
amino groups (for example, the 6-amino groups of lysine residues) are
not involved in carbohydrate binding, whereas the hydroxyl groups of
tyrosine residues may make contact with bound ~ a r b o h y d r a t eMod-
.~~~
ification ofcarboxyl groups with glycine methyl ester or glycinamide in
the presence of a water-soluble carbodiimide abolished the erythro-
agglutinating activity of the protein, supporting the role of carboxyl
groups of the lectin in the carbohydrate-binding m e c h a n i ~ m . ~ ~ ~ . ~ ~ ~
Fluorescence studies led to the conclusion that at least two of the
three tryptophan residues in WGA are highly accessible to solvent
molecules.494Inasmuch as the fluorescent, tryptophan residues were
not fully protected by N,N',N''-triacetylchitotriose from quenching by
iodide, it was postulated that they were not directly in the WGA
binding-site."' Treatment of WGA with N-bromosuccinimide in 0.1 M
acetic acid-8 M urea modified all of the tryptophan residues, whereas
only two of three tryptophan residues were modified when the reaction
was conducted in 0.1 M citrate (pH 6.0) buffer.508Oxidation (in acetic
acid-urea) of one tryptophan residue per subunit led to almost com-
plete (97%)loss of hemagglutination activity and a 3.5-fold decrease in
the affinity constant for N,N',N"-triacetylchitotriose, and rendered the
subunits unable to form the native dimer.s08No significant changes in
the circular dichroism spectrum of WGA were observed after oxidation
of three tryptophan residues, suggesting that no gross conformational
changes had occurred.508
The presence of 2-acetamido-2-deoxy-~-glucose, -D-galactose,
-D-mannose, or the N-acetylated chito-di-, -tri-, and -tetra-saccharides
induced conformational changes in the l e ~ t i n . @ The~ ~changes at 270-
290 nm of the c.d. spectrum were attributed to perturbations of tyrosine
and, possibly, tryptophan and disulfide c h r o m o p h ~ r e s . ~ ~ ~ ~
(507)J.-P, Privat and M. Monsigny, Eur. J . Biochem., 60,555-567 (1975).
(508)J.-P.Privat, R. Lotan, P. Bouchard,N. Sharon, and M. Monsigny,Eur.]. Biochem.,
68,563-572 (1976).
224 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

It has now been found that WGA will form a specific precipitate with
keratan from cornea and nasal cartilage50ea;these glycosaminoglycans
contain alternate, (1+4)-linked 2-acetamido-2-deoxy-~-D-gluco-
pyranosyl residues. Similarly, pneumococcus S 14 capsular poly-
saccharide (containing 2-acetamido-2-deoxy-~-~-glucosyl residues
substituted by glycosyl groups at both 0-4 and 0-6) gave a precip-
itin CUrve508a.508b with WGA (compare Ref. 502). S 14 polysaccharide
that had been subjected to controlled, Smith degradation also
precipitated with WGA; this derived polysaccharide contains
2-acetamido-2-deoxy-~-~-glucosyl residues linked solely at 0-6. These
results demonstrate unequivocally that 2-acetamido-2-deoxy-p-D-
glucosyl residues linked ( 1 4 6 ) or (1+4, 1 4 6 ) are capable of interact-
ing50sa,50sb with WGA.
Elution of a chitin column, charged with wheat-germ extract, with
0.01 M acetic acid (as opposed to 0.05 M hydrochloric gave
protein(s) mitogenic for purified, human peripheral-lympho~ytes.~~~~
This discovery is contrary to several past reports (see, for example, Ref.
156)that showed that WGA had no mitogenic activity. The relationship
between WGA and the newly described protein(s),termed wheat-germ
mitogen, has yet to be established.

5. Ulex europeus I1 (Gorse or Furze Seed)

[gorse or furze seed; ~-~-GlcNAcp-(1+4)-D-GlcNAcl
The extract of Ulex europeus seeds has long been used in serological
1aboratorie~'~J~ as a reagent for typing blood-group 0 erythrocytes,
determining subgroups of blood types A and AB, and assessing secretor
status (the occurrence of H-active substance in saliva). This use was
based upon the initial observation of Cazal and Lalauriesl that three
species of Ulex seeds (Ulex provincialis Lois, Ulex jussiaei Webb, and
Ulex europeus L.) contained strong hemagglutinating activity against
blood-group 0 erythrocytes, with weaker activity against Az cells, and
occasional reactivity with B cells. These results were subsequently
confirmed by Boyd and Shapleigh,12*13 F l ~ r yand
, ~ K~ ~~+ p e Subgroup
.~~
AzB reacted with the extract, whereas AIB did not.lZSaliva of secretor
individuals inhibited 0 erythrocyte agglutination by Ulex europeus
extract, regardless of the individual's blood type.13

(508a) H. E. Carlsson, J. Lonngren, I. J. Goldstein, J. E. Christner, and G. W. Jourdain,

FEBS Lett., 62, 38-40 (1976).
(508b) S.Ebisu, J. Lonngren, and I. J. Goldstein, Carbohydr. Res., 58,187-191 (1977).
(508c) J. M. Brown, M. A. Leon, and J. J. LightbodyJ. Immunol., 117,1976-1980 (1976).
(508d) G. W. G. Bird and J. Wingham, Vox Sung., 19,132-139 (1970).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 225

Flory fractionated U. europeus extracts with ethan01.2~~ The 30%-and

50%-alcohol precipitates agglutinated type 0 erythrocytes, a reaction
that was readily inhibited by P-D-glycosides (cellobiose, lactose, or
salicin), but unaffected by L - ~ U C O S On ~ ~ other hand, 0 erythrocyte
~ . ~the
agglutination by the 70%-alcohol precipitate was inhibited by L-fucose
(slightly by D-arabinose, D-ribose, and D-lyxose), but not by any of the
aforementioned glycosides. Furthermore, the 30%- and 50%-alcohol
precipitates reacted with all human, buccal (epithelial cheek) cells,
whereas the 70%-alcohol precipitate agglutinated only the buccal cells
of secretors. On the basis of these observations, Flory suggestedzz5that
Ulex extract contained at least two agglutinins having different
carbohydrate-binding specificity.
Apparently, lectins of two alternative specificities interact with type
0 erythrocytes. Eel serum-type anti-H(0) agglutinins (Anguilla an-
guilla, Lotus tetragonolobus, and Ulex europeus I) are best inhibited
by L-fucosides; they are considered in Section VI. Cytisus-type anti-
H ( 0 ) agglutinins (Cytisus sessilifolius, Laburnum alpinum, and Ulex
europeus 11) are inhibited by N,N’-diacetylchitobi~se.~~~ Purified,
H-decomposing enzyme from Bacillus fulminans, an L-fucosidase, de-
stroyed the agglutinability of 0 erythrocytes not only by L-fucose-
binding anti-H(0) lectins but also by N,N‘-diacetylchitobiose-binding
anti-H(0) lectins as Although L-fucose does not inhibit the
N,N’-diacetylchitobiose-binding lectins directly, it evidently affects
the complementarity of complex oligosaccharides to the binding sites
of these lectins. A structure with 2-acetamido-2-deoxy-~-~-glucosyl
residues occurs in an internal position of blood-group substance
oligosaccharide chains.z4
Proceeding from Flory’s observations, Matsumoto and Osawa pur-
ified two lectins of distinctly different specificities from Ulex europeus
extracts196~08~2z6~50s;
an L-fucose-binding protein, Ulex I (Ref. 226), and a
2-acetamido-2-deoxy-~-glucose-binding protein, Ulex I1 (Ref. 208).
Ulex I1 was obtained in pure form from the 40-70%-saturated am-
monium sulfate precipitate by cellulose ion-exchange, poly(viny1
chloride) block electrophoresis, and Biogel P-200 gel-filtration.z08The
lectin was homogeneous in the analytical ultracentrifuge, and by
poly(acry1amide) disc-gel electrophoresis. An Szo,, value of 6.5 S, but
no value for the molecular weight, was reported. Amino acid analysis
revealed Ulex I1 to be rich in acidic and hydroxy amino acids, and low in
those containing sulfur. Carbohydrate analysis gave 21.7% (by weight)
of carbohydrate (mannose, galactose, arabinose, and 2-amino-2-
deoxyglucose). Ulex lectin I1 has also been purified14sby adsorption to
(509) T. Osawa and I. Matsumoto, Methods Enzymol., 28, Part B, 323-327 (1972).
226 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

N,N"N"-triacetylchitotriose-substituted starch, followed by elution

with 0.05 M glycine hydrochloride buffer, pH 3.0.
Characterization of the carbohydrate-binding specificity of Ulex
europeus I1 is still very rudimentary. Studies by inhibition of
hemagglutination have been completed, using only a few of the ap-
propriate sugars necessary to define specificity of the binding.
Makela found7*that crude extracts from Ulex seeds were inhibited by
L-fucose and salicin, whereas Kriipe7' did not include L-fucose among
the sugars which inhibited Ulex-0 erythrocyte agglutination (compare
Ref. 508d). Oligosaccharides isolated from blood-group substances,
and containing (nonreducing) p-D-linked 2-acetamido-2-deoxy-~-
glucosyl groups inhibited such Cytisus-type anti-H(0) lectins as U.
europeus I1 (see Ref. 475).
The investigation of Matsumoto and Osawa10s~500 is the most complete
to date; it reveals a binding specificity for (1+4)-linked 2-acetamido-
2-deoxy-~-~-glucosyl derivatives. The binding site appears to be an
extended one, capable of accommodating N,N'-diacetylchitobiose (4)
or N,N',N"-triacetylchitotriose (6). Salicin, phenyl p-D-glucopyrano-
side, and cellobiose were poor inhibitors; maltose, lactose,N-acetyllac-
tosamine [@D-Galp-(1-*4)-D-GlcNAc], 2-acetamido-2-deoxy-~-galac-
tose, and 2-acetamido-2-deoxy-~-glucosewere noninhibitors.

Iv. 2-ACETAMIDO-2-DEOXU-D-GALACTOSE-BINDING
LECTINS
1. Dolichos biflorus (Horse Gram)
(horse gram; a-D-GalNAcp >> a-D-Galp)
The phytohemagglutinin of Dolichos bi$orus is one of three well-
characterized, blood-group A-specific, plant lectins (see also Phaseolus
lunatus and Helix pomatia), and one of two phytohemagglutinins to
find application in routine serology (see also Ulex e u r ~ p e u s ) . The
'~
agglutination of human A, eyrthrocytes by Dolichos extracts was first
reported by Bird.510-512 When the lectin was compared to human anti-A
sera with respect to agglutination of erythrocytes from representative
blood-group types, complete correspondence was observed.512A, cells
were agglutinated by the lectin at titers between 8,192 and 32,768, and
A2 cells at titers of 4-16. Furthermore, Dolichos lectin specificity was
retained when tested with papain-treated r e d - ~ e l l s . ~
Subsequently,
'~
Boyd and Shapleigh2observed that saliva of type A, secretor individu-
(510) G. W. G. Bird, Curt-. Scf.,20,298-299 (1951).
(511) G. W. G. Bird, Indian]. Med. Res., 40,289-293 (1952).
(512) G.W. G . Bird, Nature (London), 170,674 (1952).
(513) G. W. G. Bird, Nature (London), 174, 1015 (1954).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 227

als formed a precipitate with Dolichos bijlorus extracts, and also inhib-
ited A, erythrocyte agglutination. Saliva of Az secretors gave very weak
reactivity in both assays; B and 0 secretor saliva were inactive. Similar
observations were made by
From agar-gel diffusion and precipitin studies of ovarian cyst A, and
Az substances with Dolichos extract, Bird'O~'~~ concluded that the lectin
made no qualitative distinction between the two subgroups of A. Hog
gastric-mucin A substance reacted in a manner identical to the cyst
substances.
In addition to its reactivity with A, erythrocytes, secretor saliva, and
ovarian-cyst substances, Dolichos bijorus lectin agglutinated strep-
tococci of the serological group C, but failed to react with a group C
variant whose cell-wall polysaccharide, unlike normal group C or-
ganisms, lacked terminal 2-acetamido-2-deoxy-cu-~-galactosyl side-
chains.s14A precipitin band with peptone A substance was also dem-
onstrated in agar plates.51sStepwise fractionation of Dolichos seed-
extracts with ammonium sulfate and Sephadex G-200 or CM-cellulose
chromatography was conducted by Kiihnemund and coworkers.516
They reported that the streptococcus-agglutinating activity was as-
sociated with a glycoprotein of molecular weight -130,000, and they
conducted preliminary amino acid and carbohydrate analyses.
Etzler and Kabat purified the agglutinating principle from Dolichos
seed extracts by adsorption to insoluble polyleucyl hog A H sub- +
~tance.'~~ Specific
* ~ ' ~ elution of the adsorbent with 2-acetamido-2-
deoxy-D-galactose gave a protein that appeared homogeneous by im-
munodiffusion, immunoelectrophoresis, disc-gel electrophoresis
under acid and alkaline conditions, and sedimentation analysis. The
amino acid composition of the protein of molecular weight 141,000
reflected a high content of aspartic acid and serine, little methionine,
and no cysteine. Carbohydrate analysis showed a sugar content of 2.4%
of hexose, 1.6% of hexosamine, and 1.5% of 2-acetamido-2-deoxy-
hexose.
An alternative purification of Dolichos bijorus lectin was published
by Font and coworkers.S18Fractional precipitation with ammonium

(514)W.Kohler and 0. Prokop, Z. Immunitaetsforsch. Allerg. Klin. Immunol., 133,

171-179 (1967).
(515) G . Uhlenbruck, I. Sprenger, A. J. Leseney, J. Fontand, and R. Bourrillon, Vox
Sung., 19,488-495 (1975).
(516)V. 0. Kuhnemund, W. Kohler, and 0. Prokop, Hoppe-Seyler's Z. Physiol. Chem.,
349,1434-1436 (1968).
(517)M.E. Etzler, Methods Enzymol., 28 (Part B), 340-344 (1972).
(518) J. Font, A. M. Leseney, and R. Bourrillon, Biochim. Biophys. Acta, 243,434-446
(1971).
228 IRWIN J. GOLDSTEIN AND COLEEN E. HAYES

sulfate was followed by gel filtration and ion-exchange chromatog-

raphy, yielding a homogeneous preparation [by starch-gel and
poly(acry1amide)-gel electrophoresis, as well as by ultracentrifugal
analysis]. They reportedSl8a sedimentation coefficient of 6.3 S, an
isoelectric point of pH 5.5, and a carbohydrate content of 3.75% by
weight. The amino acid analysis was in good agreement with that of
Etzler and Kabat.lo8The glycosyl portion of the lectin was composed
largely of mannose residues, with small proportions of glucose, fucose,
xylose, and arabinose residues. Periodic acid oxidation resulted in
rapid loss of agglutinating activity, accompanied by only minor decom-
position of neutral sugars, and no loss of hexosamine.
Subsequent studies by PBre and coworkerss19 demonstrated a
molecular weight of 30,000 for a subunit. Gel filtration in 6 M
guanidinium chloride, dodecyl sodium sulfate electrophoresis, and
peptide mapping of the tryptic digest confirmedSl9that four noncova-
lently associated subunits comprise the native lectin of molecular
weight 120,000. Only N-terminal alanine was found, suggesting that
the four subunits are identical.
Carter and E t ~ l e r ~confirmed
~O the presence of one methionyl residue
per subunit by cyanogen bromide cleavage ofDolichos lectin into two
fragments. The peptide fragments, separated by anion exchange and
gel filtration, were distinguished, in that the amino terminal segment,
of molecular weight 15,000, contained carbohydrate, whereas the car-
boxyl terminal segment, of molecular weight 12,000, did not. Acid
hydrolysis, and analysis of the pronase-digested glycopeptide, gaveszoa
mixture of serine, aspartic acid, mannose, and 2-acetamido-2-deoxy-D-
glucose in the ratios of 1:5:20-25:5-10. 2-Acetamido-N-(~-aspart-4-
oyl)-2-deoxy-~-D-glucopyranosy~amine was also identified in a partial,
acid hydrolyzate.”O
What had, in earlier ~ t u d i e sappeared, ~ ~ ~to ~be~a homogeneous,
~ ~ ~ ~ ~ ~
protein preparation was further fractionated into electrophoretically
distinguishable isolectins A and B by chromatography on concanavalin
A-Sepharo~e.~ The~ ~B* form,
~ ~ ~ which comprised less than 12% of the
original lectin sample, was not bound, whereas the A form was bound
and was specifically elutedsZ1as a biphasic peak with a gradient of
methyl a-D-glucopyranoside solution. Analysis, for carbohydrate con-
tent, of fractions obtained from different portions of the elution profile
revealed considerable heterogeneity in the relative proportions of
D-mannose and 2-acetamido-2-deoxy-~-glucose per molecule of pro-
(519)M.PBre, J. Font, and R. Bourrillon, Biochim. Biophys. Acta, 365,40-46 (1974).
(520)W. G.Carter and M. E. Etzler, Biochemistry, 14,5118-5122(1975).
(521)W. G. Carter and M. E. Etzler,J. Biol. Chem., 250,2756-2762 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 229

tein. Forms A and B, with respective apparent molecular weights of

113,000 and 109,000, were present as native proteins in the seed,
instead of arising from modification or degradation during the course of
the purification.521The two species showed considerable similarity in
amino acid composition, were indistinguishable in immunodiffusion
against antisera to the seed extract, and exhibited identical amino-
(alanine) and carboxyl-terminal (leucine and valine) amino acids.
Studies on the carbohydrate specificity of forms A and B showed essen-
tially no difference. Dodecyl sodium sulfate-urea gel-electrophoresis
ofthe purified isolectins revealed four distinguishable types of subunit,
IA and IIA in form A, IB and IIB in form B, whereas dodecyl sodium
sulfate gel-electrophoresis revealed only two types of subunit, a band
of material of molecular weight 26,500 from form A, and a band of
molecular weight 26,000 from the B form.521A so-called “prolectin”
has been isolated from the leaves and stems of 6-weeks-old Dolichos
bijlorus plants243a(see Section 1,5).
Subunits IA and IIA were isolated by ion-exchange chromatography
on DEAE-cellulose in 8 M urea.s22Sedimentation-equilibrium analysis
in 8 M urea gave subunit molecular weights of 27,700 and 27,300 for IA
and IIA, respectively. The isolated subunits did not differ substantially
in amino acid composition, or antigenicity. Although they both con-
tained N-terminal alanine, digestion of IA with carboxypeptidase A
released leucine and valine simultaneously, whereas IIA was not
hydrolyzed by this enzyme under identical conditions. On the basis of
their results, Carter and EtzleP2 proposed a stoichiometry of IA211A2
for the principal form of Dolichos bijlorus lectin. The first 30 amino-
terminal amino acids of the IA and IIA subunits of the D . bijlorus lectin
were shown to be supporting the suggestion that the two
subunits may differ from one another only at their carboxyl-terminal
ends .520,522
The carbohydrate-binding specificity of purified Dolichos bijlorus
lectin has been studied in detail by Etzler and Kabat.’08 In precipitin
studies conducted on the Dolichos lectin, 87-99% was precipitated by
ovarian cyst A substance, whereas only 60-70% was precipitated by
hog A substance; the latter result was attributed to the formation of
soluble, lectin-oligosaccharide complexes. A, substances were more
reactive than A2 substances; however, no reaction was observed with B
or 0 substances, even following mild hydrolysis with acid or two
stepwise, Smith degradations. N-Deacetylation of A substance with

(522) W. G. Carter and M. E. Etzler, Biochemistry, 14,2685-2689 (1975).

(522a) M. E. Etzler, C. F. Talbot, and P. R. Ziaya, FEBS Lett., 8 2 , 3 9 4 1 (1977).
230 IRWIN J. GOLDSTEIN AND COLEEN E. HAYES

Clostridium tertium N-deacetylating enzyme abolished Dolichos lec-

tin reactivity; restoration was achieved by re-N-acetylation.
Sugar inhibition of human A substance-lectin precipitation
suggested that the lectin combining-site was more complementary to
2-acetamido-2-deoxy-a-~-galactopyranosyl end-groups. The A-active
disaccharide, a-~-GalNAcp-(1+3)-~-Gal,and trisaccharide, a - D -
GalNAcp-(1+3)-@-D-Galp-(1+3)-D-GlcNAc7 were equivalent in in-
hibitory capacity to methyl 2-acetamido-2-deoxy-a-~-galactopyrano-
side (0.55 pmole for 50% inhibition). The methyl a-glycoside was 2.5
times as effective as 2-acetamido-2-deoxy-~-galactose,whereas the
ethyl @-glycosidewas less effective than the latter. The following com-
pounds were noninhibitory: 2-acetamido-2-deoxy-~-glucose, 2-acet-
amido-2-deoxy-~mannose,2-amino-2-deoxy-~galactose,2-amino-2-
deoxy-D-glucose, 2-acetamido-2-deoxy-~-galactitol,Dgalactose, L-fu-
cose, D-mannose, and D-glucose. Two disaccharides in which 0-3 of
2-acetamido-2-deoxy-~-galactose was substituted by a @-Dgalactosyl
or a 2-acetamido-2-deoxy-a-~-glucosylgroup were also inactive. The
best inhibitor tested, needing only 0.32 pmole for 50% inhibition, was
the A-active, reduced pentasaccharide shown in formula 7. The authors

U-D -
-GalNAc p- (l+ 3)- 0-0 -Calp- (1- 4)- 8-D-GlcNAcp ( l - d ) - R
2
t1
U-L -Fuc~
7

suggested that L-fucosyl substitution either confers a more favorable

conformation upon the nonreducing 2-acetamido-2-deoxy-a-~-
galactosyl group, as compared to the di- and tri-saccharides mentioned
previously, or itself contributes to stabilization of lectin-oligosac-
charide binding, implying an extended carbohydrate-binding site on
the Dolichos lectin molecule.
PBre and coworkers investigated the circular dichroic spectrum of
Dolichos lectin in the presence and absence of 2-acetamido-2-deoxy-
D-gala~tose.~'~The far-ultraviolet, c.d. spectrum of Dolichos lectin dis-
played weak, negative bands at 217 and 230 nm, with a positive band at
197 nm. By analogy to the spectrum of con A, whose complete, three-
dimensional structure is known, the authors concluded that the
Dolichos lectin has a preponderance of the aperiodic, bent structure
stabilized by hydrophobic interactions, and a significant content of
(523) M.PBre, R. Bourrillon, and B. Jirgensons, Biochim. Biophys. Acta, 393, 31-36
(1975).
 LECTINS: CARBOHYDFMTE-BINDING PROTEINS 231

p-pleated-sheet conformation. The addition of 2-acetamido-2-deoxy-

Dgalactose caused a significant diminution in the amplitude of the c.d.
spectrum at 280-286 nm and at 290 nm, at pH 6.8-7.5. No alteration was
observed at pH 8.25. As the 280-300-nm spectral zone is related to the
tryptophanyl and tyrosyl chromophores, Pkre and coworkers sugges-
ted523that these residues may be perturbed upon sugar binding. Addi-
tion of dodecyl sodium sulfate to the lectin disrupted its tertiary struc-
ture, and induced some a-helix formation.
Purified Dolichos bijlorus lectin has been applied to the study of
glycoproteins and glycolipids from mammalian cell^.^^^-^^^ Fluores-
cein isothiocyanate-conjugated lectin was employed in a study of the
rat stomach and d ~ o d e n u m . " From ~ both organs could be isolated
A-active material by adsorption to Dolichos lectin-agarose columns,
and elution with 2-acetamido-2-deoxy-~-ga~actose.Fluorescein-
labelled lectin was also used to examine sections from various regions
of rat s m a l l - i n t e ~ t i n eDifferential
.~~~ fluorescent staining of epithelial
cells lining the crypts and villi of the intestine was observed from the
proximal to the distal end, suggesting differential localization of cell-
surface and secretory components in these regions. Furthermore, a
Dolichos lectin-Sepharose column, in tandem with a Lotus tet-
rugonolobus lectin-Sepharose column, proved effective in separating
hog gastric-mucin into A-substance and H-substance devoid of cross-
. ~ ~ ~ the interaction between Dolichos lectin and
c o n t a r n i n a t i ~ nFinally,
chick embryonic fibroblasts as a function of development was studied
by Roguet and B o u r r i l l ~ n The . ~ ~ ~number of lectin-binding sites per
cell remained constant, but the apparent association-constant de-
creased from day 8 to day 16. The effect of lectin on [3H]thymidine
incorporation was age-dependent in the chick fibroblast.

2. Glycine max (Soybean)

(soybean; a-D-GalNAcp 5 0-D-GalNAcp >> a-D-Gab)
The nutritional superiority of heated soybean meal in contrast to raw
was initially thought to reflect the presence of a growth-
repressive, heat-labile s u b ~ t a n c e . ~
Thus,
~ ~ -Liener
~ ~ ~ and P a l l a n s ~ h ~ ~ ~
(524) M.E. Etzler, Ann. N.Y. Acad. Sci., 234,260-275 (1974).
(525)M.E. Etzler and M. L. Branstrator,J.Cell B i d , 62,329-343 (1974).
(526)M.E. A. Pereira and E. A. Kabat,J. E x p . Med., 143,422-436 (1976).
(527)R. Roguet and R. Bourrillon, Biochim. Biophys. Acta, 389,380-388(1975).
(528)T. B. Osborne and L. B. Mende1,J. Biol. Chem., 32,369-387 (1917).
(529) I. E. Liener and M. J. Pallansch,J. Biol. Chem., 197,29-36 (1952).
(530)I. E. Liener,J. Nutr., 49,527-539 (1953).
(531) I. E. Liener and T. A. Seto, Cancer Res., 15,407-409 (1955).
232 IRWIN J. GOLDSTEIN AND COLEEN E. HAYES

and LienePO fractionated a toxin, from defatted soybean flour, that

exhibited some growth-repressive activity in feeding studies, strong
hemagglutinating activity, and lability to heat and to pepsin diges-
tion.532However, they observed no specific effect of the substance on
tumor growth, or on the urease, lipoxidase, or anti-tryptic activities
known to be present in soybean extracts. Although Liener and Pal-
l a n s ~ found
h ~ ~ no
~ statistically significant correlation between toxicity
and hemagglutinating activity, they implied that the two activities were
due to a single entity.
These early studies were conducted on a preparation partially pur-
ified by fractional precipitation with ammonium sulfate and
Several purification schemes have since been reported to yield
homogeneous preparations. Wada and coworkers534used the technique
of recycling, moving-boundary electrophoresis to obtain a hemaggluti-
nating protein that manifested homogeneity in ultracentrifugation,
starch-gel electrophoresis, and anion-exchange chromatography. Stead
and applied an ammonium sulfate- and acetone-
precipitated fraction to a DEAE-cellulose column, and obtained six
peaks with a stepwise gradient of sodium chloride. They reported no
correlation between the toxicity, hemagglutinating activity, and trypsin
inhibition of the six peaks, confirming the earlier work of Rackis and
An alternative purification, in which ammonium sulfate
fractionation was followed by chromatography on calcium phosphate,
was reported by Lis537and Lis and to give a protein identi-
cal to a sample obtained from Liener.530Hemagglutinating activity was
confined to a single peak of the calcium phosphate, gel-elution profile.
Rechromatography of this fraction on calcium phosphate, CM-
cellulose, or Sephadex G-50 yielded a single, protein peak. Further
analysis by chromatography on Sephadex G-50 in 4 M guanidinium
chloride, ultracentrifugation, and gel electrophoresis at both acidic and
alkaline pH did not reveal significant contamination. However, when
the purified agglutinin was chromatographed on DEAE-cellulose,
four distinct peaks of hemagglutinating activity were f o ~ n d .Each ~ ~ * ~ ~ ~
(532)I. E. Liener,J. Bid. Chem., 233,401-405 (1958).
(533)M.J. Pallansch and I. E. Liener, Arch. Biochem. Biophys., 45,366374(1953).
(534) S. Wada, M.J. Pallansch, and I. E. Liener,J. B i d . Chem., 233,395-400 (1958).
(535)R. H.Stead, H. J. H. de Muelenaere, and G. V. Quicke, Arch. Biochem. Biophys.,
113,703-708 (1966).
(536)J. J. Rackis, H. A. Sasame, R. L. Anderson, and A. K. Smith,]. Am. Chem. Soc., 81,
6265-6270 (1959).
(537)H. Lis, N. Sharon,and E. Katchalski,Biochim. Biophys. Acta, 83,376-378(1964).
(538)H. Lis, N. Sharon, and E. Katchalski,]. Biol. Chem., 241,684-689 (1966).
(539)H. Lis, C. Fridman, N. Sharon, and E. Katchalski,Arch. Biochem. Biophys., 117,
301-309 (1966).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 233

pooled peak was rechromatographed on DEAE-cellulose, followed by

calcium phosphate, and, finally, again on DEAE-cellulose. The four
hemagglutinins were not separable on calcium phosphate or CM-
cellulose at pH 4.0; they migrated identically in gel electrophoresis at
pH 4.5, and had almost identical amino acid compositions, but varied
slightly in their content of mannose and 2-amino-2-deoxyglucose. In
verification of these results, Catsimpoolas and Me~e15~O demonstrated
the presence of four, immunochemically indistinguishable,
hemagglutinating proteins by isoelectric focusing of soybean extract.
Thus, it appears that soybean agglutinin exists as multiple, highly
similar forms.
Two affinity systems for the purification of soybean agglutinin have
been developed. Gordon and coworker^^^^,^^^ coupled N-(6-amino-
hexanoy1)-P-D-galactopyranosylamineto cyanogen bromide-activated
Sepharose, to afford a specific adsorbent for the agglutinin. Elution
with D-galactose gave, in 90% yield, a major hemagglutinin from which
minor agglutinins could be removed by DEAE-cellulose, according to
Lis and Sharon.g7Soybean agglutinin so prepared was homogeneous
by gel electrophoresis, and identical, with respect to electrophoretic
mobility, and amino acid and carbohydrate analyses,538to agglutinin
prepared by previous methods. A second, simpler, affinity adsorbent
was prepared by Allen and N e ~ b e r g e Pby ~ ~reaction of 2-amino-2-
deoxy-D-galactose with CH-Sepharose 4B in the presence of a car-
bodiimide. The adsorbent bound 12 mg of agglutinin per ml, which
was 1% of the total protein applied. Elution with Dgalactose solution
gave one major and several minor agglutinins which were separated by
anion-exchange chromatography. The electrophoretic mobility and
chemical composition were in accord with those for soybean agglutinin
prepared by other method^.^^,^^^
The biophysical characteristics of soybean agglutinin were first in-
vestigated by Pallansch and Liener533with a protein preparation known
to contain minor contaminants. They determined a sedimentation
coefficient of 6.4 S , a diffusion coefficient of 5.72 X cm2 sec-', an
extinction coefficient E,l& = 15.7, a molecular weight (by sedimenta-
tion) of 105,000, and an isoelectric point at pH 6.1. Catsimpoolas and
MeyeP40 confirmed that the isoelectric point is pH 6.0, and reported
that dissociation of the lectin in phenol-acetic acid in the presence of

(540) N. Catsimpoolas and E. W. Meyer, Arch. Biochem. Biophys., 132,279-288 (1969).

(541) J. A. Gordon, S. Blumberg, H. Lis, and N. Sharon, FEBS Lett., 24,193-196 (1972).
(542) J. A. Gordon, S. Blumberg, H. Lis, and N. Sharon, Methods Enzymol., 28, Part B,
365-368 (1972).
(543) A. K. Allen and A. Neuberger, FEBS Lett., 50,362-364 (1975).
234 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

2-mercaptoethanol and urea gave two subunits. Lis and coworkers538

analyzed a highly purified, lectin sample, and estimated a sedimenta-
tion coefficient of 6.0 S, a diffusion coefficient of 5.0 x lo-' cm2.sec-1,
and calculated the molecular weight of the protein to be 110,000.
Without question, the most complete, biophysical characterization of
soybean agglutinin is that of Lotan and The lectin,
affinity-purified according to Gordon and was
homogeneous by poly(acry1amide) gel-electrophoresis at pH 4.5, 8.9,
and 7.2 in the presence of a detergent. Furthermore, electrophoresis in
an acrylamide gradient of 4 to 8% with 8 M urea yielded a single,
protein band, as did isoelectric focusing. Sedimentation-velocity
studies in buffers of pH 2.2 to pH 10.8 at protein concentrations of 3 to
10 mg per ml resulted in a single, symmetrical peak of s&,w = 6.0L0.125
S. The molecular weight calculated was 122,000 k1,300, using an
experimentally determined, partial specific volume of 0.745mug. This
value was in good agreement with the molecular weight of 120,000
L 10,000 determined by gel filtration. The protein formed aggregates
having high molecular weight if stored for a long time in the lyophilized
state."5 A subunit molecular weight of 30,000 was obtained by poly(ac-
rylamide) gel-electrophoresis in dodecyl sodium sulfate, by gel filtra-
tion in detergent, and, finally, by sedimentation equilibrium in 6 M
guanidinium the authors reported an extinction coefficient
of = 12.8 cm-'.
Although, on the basis of end-group analysis (1mole of N-terminal
alanine per 30,000 g of protein in 8 M urea) and electrophoretic
s t ~ d i e s , 5the
~ ~soybean agglutinin appeared to be a tetrameric protein
composed of identical subunits, Lotan and coworkers546later reported
resolution of two types of subunit in the ratio of 1:1,either by electro-
phoresis at alkaline pH in the presence of urea or detergents, or by
chromatography on DEAE-cellulose in Tris buffer, pH 7.3, with 8 M
urea.
Several groups have investigated the chemical composition of soy-
bean agglutinin. The early amino acid analysis of Wada and cowork-
e r ' differs
~ ~ ~considerably from the later analyses by Lis and co-
w o r k e r ~ .They ~ ~ ,found
~ ~ almost twice the content of serine, leucine,
and lysine, and substantially increased proline than those reported by
Wada and whereas the contents of methionine,

(544)R. Lotan, H.W. Siegelman, H. Lis, and N. Sharon,]. Biol. Chem., 249,1219-1224
(1974).
(545)R. Lotan, H.Lis, and N. Sharon, Biochem. Biophys. Res. Commun., 62,144-150
(1975).
(546) R.Lotan, R. Cacan, M. Cacan, H. Debray, W. G . Carter, and N. Sharon,FEES Lett.,
57, 100-103 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 235

isoleucine, and arginine were comparatively diminished. The results

of Lis and c o ~ o r k e r sconcur
~ ~ ~ with
, ~ ~ each
~ other. Wada and cowork-
e r calculated
~ ~ six ~ residues
~ of cysteine per molecule of lectin on the
basis of elemental analysis, although no free sulfhydryl groups were
found. Both of the later amino acid analyses suggested that soybean
agglutinin is devoid of c y ~ t e i n e . " ~Like
, ~ ~ several
~ other lectins, soy-
bean lectin is comparatively rich in acidic and hydroxylic amino acids.
Of the nonidentical subunits separated by Lotan and coworkers,s46
subunit I1 contains two more aspartic acid residues, one additional
glutamic acid residue, and one lysyl residue fewer, as compared to
subunit I. End-group analysis revealed only N-terminal
alanine.534*544*546
The amino-terminal residues of the soybean agglu-
tinin have been s e q ~ e n c e d . ~ ~Eleven
~~,"~ of the first amino-terminal
residues are identical with that of the peanut lectin and of the /+chain
of the lentil l e ~ t i nAmong
. ~ ~ ~the 14 nonidentical residues, 9 could
have resulted from a single nucleotide substitution for the lentil lectin,
and 8 for the peanut lectin.442c These data indicate a rather high level
of sequence homology among the three lectins, and suggest a common
ancestry for the genes coding for these three plant-lectins.
Chemical modification of soybean agglutinin by acetylation of its
amino groups resulted in little loss of agglutinating activity, whereas
the protein was quite sensitive to modification of its tyrosyl
Failure of the protein to react with 2-iodoacetamide or p-(chloro-
mercuri)benzoate in 6 M urea confirmed that it was devoid of
sulfhydryl groups. A metalloprotein containing151Ca2+and Mn2+,the
soybean lectin is inactivated by A13+, Fe3+,and Pb2+,whereas MnZ+,
Ba2+,Mf+, Ag+, Li+, and K+ are without
The soybean lectin is a glycoprotein containing -7% (by weight) of
carbohydrate comprised of mannose and 2-amino-2-
Each subunit of the agglutinin carries an
deoxyglucose.534=538~539~s44~546
oligosaccharide chain composed of nine mannosyl and two 2-amino-2-
deoxyglucosyl residues.s48A glycopeptide of molecular weight 4,600
was isolated from pronase-digested, native agglutinin. Di-
gestion with glycosidase revealed that mannose occurs in three, dis-
tinct regions of the oligosaccharide, separated by 2-acetamido-2-
deoxy-D-glucosyl residues, Exhaustive digestion with a-D-
mannosidase and 2-acetamido-2-deoxy-~-~-glucosidase allowed Lis
and coworkerssmto isolate the carbohydrate-protein linkage region of
the glycopeptide ,and to characterize it as 2-acetamido-(~-aspart-4-oyl)-

(547) I. E. Liener and S. Wada,J. B i d . Chem., 222,695-704 (1956).

(548) N. Sharon, H. Lis, and R. Lotan, CoZZoq. Int. C. N . R. S., 221,693-710 (1974).
(549)H.Lis, Isr. J . Chem., 6, 114~(1968).
(550)H.Lis, N.Sharon,andE. Katchalski,Biochim.Biophys. Acta, 192,364-366(1969).
236 IRWIN j. GOLDSTEIN AND COLLEEN E. HAYES

2-deoxy-P-D-glucosylamine. Oxidation of five of the nine mannosyl

units per subunit of soybean agglutinin with periodic acid caused no
diminution of hemagglutinating activity.ss1Reduction of the oxidized
lectin with sodium borotritide afforded a radioactive product indis-
tinguishable from the native protein by several criteria.551
The carbohydrate-binding specificity of soybean agglutinin appears
to be directed towards both anomers of 2-acetamido-2-deoxy-~-
galactose.z1z~ss52
By inhibition of hemagglutination, Lis and coworkers21z
found that four disaccharides, in which 2-acetamido-2-deoxy-~-
galactose was linked P-D-(1+6), a - ~ 1+3),
-( P-D-(1+3), and P-D-(1+4),
respectively, to D-galactose, were approximately equivalent to
2-acetamido-2-deoxy-~-galactose inhibition. In a comprehensive, im-
munochemical study, Pereira and coworkersszzextended these obser-
vations by measuring mono- and oligo-saccharide inhibition of soybean
agglutinin-human blood-group precursor substance precipitation (see
Table X).Several conclusions emerge from their results. Firstly, the
lectin exhibited greatest affinity for 2-acetamido-2-deoxy-~-galactose,
its glycosides, and oligosaccharides in which this was the nonreducing,
terminal, sugar group; the reaction was inhibited to a lesser extent by
D-galaCtOSe and its derivatives. Secondly, a slight preference for a-over
P-glycosidically linked sugars was evident, as was a preference for
aromatic over alkyl aglycons. Thirdly, substitution of blood-group
A-active oligosaccharides by L-fucosyl residues greatly diminished
their binding capacity, although L-fucose was linked to the penultimate
D-galactosyl residue, leaving nonreducing, terminal 2-acetamido-2-
deoxy-D-galactosyl groups unsubstituted. An analogous "blocking ef-
fect" occurred upon L-fucosyl substitution of lactose derivatives and
blood-group B-active oligosaccharides. Fourthly, on comparing P-D-
galactosyl disaccharides, it was evident that 6-0-linked sugars were
more reactive than their 4-0- or 3-0-substituted counterparts. Finally
D-glucose, D-mannose, L-fucose, L-rhamnose, 2-acetamido-2-deoxy-
D-glucose, and 2-acetamido-2-deoxy-~-mannose were ineffective. In
summary, soybean agglutinin is most complementary to 2-acetamido-
2-deoxy-a-~-galactopyranosylend-groups (8).

NHAC
8
(551)€7. Lotan, H.Debray, M. Cacan, R. Cacan, and N. Sharon,J. B i d . Chem., 250,
1955-1957 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 237

TABLEX
Inhibition of Soybean Agglutinin-Human Blood-group Precursor Substance
Precipitation by Mono- and Oligo-saccharidesSs2

Concentration giving
sugar 50% inhibition ( p M )

Phenyl 2-acetamido-2-deoxy-a-~-galactopyranoside 0.014

2-Acetamid0-3-0-(2-acetamido-2-deoxy-a-~-galactopyranosyl)-
2-deoxy-6-O-~~ga~actopyranosy~-D-g~ucose(A-active
trisaccharide) 0.019
3-0-(2-Acetamido-2-deoxy-a-~-galactopyranosy~)-~-ga~actose
(A-active disaccharide) 0.025
Methyl 2-acetamido-2-deoxy-cY-Dgalactopyranoside 0.025
Ethyl 2-acetamido-2-deoxy-~-~-galectopyranoside 0.037
2-Acetamido-2-deoxy-~-galactose 0.10
p-Nitrophenyl a-D-galactopy-ranoside 0.48
Mono-L-fucosyl A-active pentasaccharide 0.60
6~-P-D-Gdactopyranosyl-D-glucose 0.70
2-Acetamido-2-deoxy-6-O-~-~galactopyranosyl-~-glucose 0.70
Methyl a-Dgalactopyranoside 0.71
p-Nitrophenyl P-D-galactopyranoside 0.71
Mnose 0.71
Stachyose 0.71
3-O-a-D-Gdactopyranosy~-Dga~actose 0.71
Methyl P-Dgalactopyranoside 1.40
2-Acetamido-2-deoxy-4-O-~-~g~actopyranosyl-~-glucose 1.40
2-Acetamido-2-deoxy-3-O-~-~galactopyranosy~-~-glucose 1.40
Lacto-N-tetraose 1.40
Di-L-fucosyl A-active pentasaccharide 1.40
Lactose 1.60
D-Galactose 2.40
2'-O-~-Fucosyllactose(2-O-~-fucopyranosyl-4-O-~-t1-gdacto-
pyranosyl-D-glucose) 2.40

Noninhibiting sugars
DGlucose
(and its methyl a-and P-glycosides)
2-Acetamido-2-deoxy-~-glucose
D-Mannose
%Acetamido-2-deoxy-~-mannose
Methyl a-L-fucopyranoside
L-Rhamnose
&Acetamido-2-deoxy-3-O-~-~-glucopyranosyl-~g~lactose
3-0-(2-Acetamido-2-deoxy-~-~glucopyranosyl)-~-gdactose
4-O-~-~Glucopyranosyl-~galactose
Lactodifucotetraose
Difucosyl B-active oligosaccharides
~~

(552) M. E. A. Pereira, E. A. Kabat, and N. Sharon, Carbohydr.Res., 37,89-102 (1974).

238 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Equilibrium-dialysis experiments revealed two 2-acetamido-2-

deoxy-D-galactose binding-sites per molecule of soybean agglutinin.544
This value was confirmed by gel filtration of the lectin on a column
equilibrated with a radioactive ligand. The noninteracting, identical
sites exhibited an association constant of K, = 3.0 x lo41iter.mole-' for
this 2-acetamido-2-deoxyhexose.
Soybean agglutinin precipitated several, purified, blood-group sub-
s t a n c e ~ Maximal
.~~~ precipitation was achieved with type A, sub-
stances; on a weight basis, Lea substances were -60-70% as active as
A, substances, whereas Az substances were considerably less active.
I-active blood-group precursor substances gave good precipitin reac-
tions. However, B-active substances reacted poorly, despite their con-
tent of terminal a-D-galactosyl groups. The attributed this
finding to the L-fucosyl blocking-effect already noted. H-substances
were unreactive. Two streptococcal polysaccharides differed in their
reaction with the soybean agglutinin. A group C polysaccharide (non-
reducing, terminal 2-acetamido-2-deoxy-~-galactosyl groups) formed a
precipitate, whereas a group A polysaccharide (terminal, nonreducing
2-acetamido-2-deoxy-~-glucosyl groups) did not. In another study,
Irimura and observed inhibitory activity of
neuraminidase-treated, porcine thyroglobulin, porcine submaxillary
mucin, and bovine submaxillary mucin in a hemagglutination assay.
Based on the structures proposed for these glycopeptides, their results
are in agreement with those of Pereira and The reactivity
of porcine-thyroglobulin glycopeptide Byfrom which sialic acid and
galactose have been enzymically removed, thereby exposing non-
reducing (terminal)2-acetamido-2-deoxy-~-~-glucosyl groups, remains
~nexp1ained.l~~
Soybean agglutinin exhibits several striking, biological activities.
The lectin agglutinates both rabbit and human erythrocytes (type A >
type 0 > type B),212rabbit red-cells binding five to six times as much
iodinated agglutinin as do human cells.QQTrypsinized erythrocytes
exhibited dramatically increased agglutinability, and the amount of
lectin bound increased only slightly.QQ Binding is reversible by addi-
tion of 2-acetamido-2-deoxy-~-galactose.The lectin also agglutinates
neuraminidase-treated, murine splenocytes, causing them to undergo
blast transformation and to exhibit an accelerated rate of DNA syn-
thesis.553Stimulation is inhibitable by 2-acetamido-2-deoxy-D-galac-
tose. Chemical cross-linking of native, soybean agglutinin into dimers
and higher oligomers greatly enhanced its hemagglutinating and
(553) A. Novogrodsky and E. Katchalski, Proc. Natl. Acad. Sci. U.S.A., 70,2515-2518
(1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 239

lymphocyte-transforming activities.343Lyophilization of soybean ag-

glutinin gives rise to polymeric forms of the lectin that are two orders of
magnitude more mitogenic towards pig lymph-node cells than dimeric
le~tin.~~~a
In further biological studies, Sela and coworkersss4found that, al-
though normal hamster, human, mouse, and rat cell-lines were not
agglutinated by soybean lectin, transformed counterparts of the cell
lines from mouse, human, and rat were agglutinable. Furthermore,
mild, proteolytic digestion rendered normal cells agglutinable. The
change in agglutinability could not be explained in terms of altered
receptor-density.sssAgglutination was again reversed by the compet-
ing sugar. The major lectin from a soybean cultivar (Glycine max cv
D68-127) was purified by chromatography on hydroxylapatite and
DEAE-cellulose. A tetrameric glycoprotein (molecular weight 92,000),
the lectin is composed of four subunits, molecular weight 23,000, and
is specific for 2-acetamido-2-deoxy-~-galactose.~~~~

3. Helix pomutiu (Edible Snail)

(edible snail; a-D-GalNAcp > a-~-GlcNAcp>> a-~-G:alp)
The edible snail, Helix pomatia, contains a lectin that specifically
agglutinates human type A, but not types B or 0, erythro-
The albumin gland, a part of the sexual ap-
cytes.s8~60~61~63~100~ss6~s63
paratus, contains rather large amounts of the agglutinin (8% of the
soluble protein).562Several other species of snails (for example, Helix
hortensisYs8 and Euphadra periomphala56s)also con-
Otala lactea,6z*ss4
tain specific agglutinins. (See Refs. 76 and 566 for a discussion of snail
lectins.)
(554)B.-A. Sela, H. Lis, N. Sharon, and L. Sachs,]. Membr. Biol., 3,267-279 (1970).
(555)B.-A. Sela, H.Lis, N. Sharon, and L. Sachs, Biochim. Biophys. Acta, 249,564-568
(1971).
(555a) D. W. Fountain and W.-K. Yang, Biochim. Biophys. Acta, 492, 176-185 (1977).
(556) Z.Kim, G.Uhlenbruck, 0. Prokop, and D. Schlesinger, Z. Immunitaetsforsch.
Allerg. Klin. Immunol., 130,290-295(1966).
(557)I. Ishiyama and T. Yomaguichi,Jpn. J . Leg. Med., 20,285-288 (1966).
(558)0. Kiihnemund and W. Kohler, Experientia, 25,1137-1138 (1969).
(559)T. Takatsu, M.Mukaida, and I. Ishiyama,Jpn. J. E x p . Med., 41,411-421 (1971).
(560)I. Ishiyama, M. Mukaida, and A. Takatsu, Ann. N.Y. Acad. Sci., 234,7594(1974).
(561) S. Hammarstrom, Ann. N.Y. Acad. Sci., 234, 183-197 (1974).
(562)S. Hammarstrom, Methods Enzymol., 28, Part B, 368-383 (1972).
(563)W. Knobloch, I. Knobloch, W.-E. Vogt, S. Schnitzler, and M. Bottger, Z. Im-
munitaetsforsch. Allerg. Klin. Immunol., 139, 119-128 (1970).
(564)H. M.Bhatia, W. C. Boyd, and R. Brown, Transfusion (Philadelphia), 7, 53-59
(1967).
(565)I. Ishiyama and A. Takatsu, Vox Sang., 19,522-526 (1970).
240 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Purification of the snail hemagglutinin from Helix pomatia has been

accomplished by adsorption to insolubilized human, or hog, blood-
group A s u b s t a n ~ [also, e ~ to
~ poly(acry1amide)
~ ~ ~ ~ ~ ~ ~ gel polymerized in
the presence of blood-group substance from human A m e c ~ n i u m ~ ~ ~ ]
followed by elution with 2-acetamido-2-deoxy-~-galactose (5-15 mM).
Purification by DEAE- and CM-cellulose chromatography has also
been Despite its specificity for a-D-linked 2-acetamido-2-
deoxy-D-galactose, the snail agglutinin can also be purified by adsorp-
tion to Sephadex G-100 and G-200, followed by elution with
2-acetamido-2-deoxy-~-galactose, D-galactose, or D-glucose, or at an
acid pH.558,559,567,568 Although the agglutinin has a very low affinity for
D-glucose, the large number of (nonreducing) a-D-glucosyl end-groups
in Sephadex must provide for the hexavalent protein enough binding
loci to cause it to be adsorbed to the matrix.
The purified hemagglutinin was homogeneous by gel filtration and
immuno-electrophoresis, and in the analytical ultracentrifuge. A
molecular weight of 79,000 was determined by the sedimentation
equilibrium method561*56g (100,000by sedimentation and velocity mea-
s u r e m e n t ~and
, ~ ~ 53,000 by gel filtration563).
Analysis showed that the snail agglutinin contained a preponderance
of acidic and hydroxylic amino acids and a large proportion of proline
r e s i d ~ e s . ~ Uncharacteristic
~ * ~ ~ ~ - ~ ~ ~of lectins from leguminous-plant
seeds, the hemagglutinin contained 18 half-cystine residues and 10
molecular proportions of methionine per molecule of protein.
About 8%(by weight) ofcovalently bound carbohydrate was found; this
was principally D-galactose and D-mann~se.'~
Free sulfhydryl groups were shown by Hammarstrom and coworkers
to be absent.56gReduction of the protein in 6 M guanidinium chloride
with an excess of l,Pdithiothreitol, followed by alkylation with
iodoacetate, led to the introduction of 18 moles of acetate per mole of
agglutinin. The mean molecular weight of the reduced, alkylated sub-
unit was determined to be 13,000 by gel filtration on a calibrated
column.569Digestion with trypsin, followed by peptide mapping, was
consistent with the presence of a single type of When the
snail hemagglutinin was treated with 6 M guanidinium chloride, either
alone, or at pH 4.0 (for 0.25-48 h), it gave, upon gel filtration on a
calibrated column, a single species, of molecular weight 26,000-

(566) R. T.Pemberton, Ann. N.Y.Acad. Sci., 234,95-121 (1974).

(567)I. Ishiyama and G . Uhlenbruck, Comp. Biochem. Physiol. A, 42,269-276 (1972).
(568)I. Ishiyama and G . Uhlenbruck, 2. Immunitaetsforsch. Allerg. Klin. Immunol.,
143, 147-155 (1972).
(569) S. HammarsWm, A.Westoo, and I. Bjork, Scand.J. Immunol., 1,295-309(1972).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 241

31,000. Equilibrium dialysis against a blood-group A-active, reduced

pentasaccharide revealed the presence of one carbohydrate binding-
site per molecular weight of 17,000, with K&= 5 x lo31iter.mole-' (at
25"), and570AGO - 21.1 kJ.mole-' (-5.04 kcal.mole-'). Scatchard plots
were linear, indicating homogenous, noninteracting sites.s70
On the basis of these data, the following model of the subunit struc-
ture is proposed: the snail hemagglutinin consists of 6 identical,
polypeptide chains (subunits), each containing one intrachain disulfide
bond and a carbohydrate binding-site. Furthermore, two subunits are
linked by an intrachain, disulfide bond to form subunit dimers of
molecular weight 26,000, and three dimers (mol. wt. 26,000) are held
together by noncovalent interactions.569
The utility of the Helix pomatia lectin as a probe for the detection of
terminal (nonreducing) 2-acetamido-2-deoxy-a-~-galactosyl groups in
biopolymers and cell surfaces (blood cells, tumor cells, and micro-
organisms) has been noted.60,61.100,s62,567 Studies of precipitation be-
tween the Helix pomatia agglutinin and a wide range of polysaccharides
and glycoproteins have been conducted by many investigators, includ-
ing Uhlenbruck and Prokop,61 Prokop and coworkers,100and Ham-
marstrom and his colleagues.63J78*s61~s62~570Human blood-group A sub-
stance (and, to a lesser extent, B, H, and Lea blood-group substances),
desialized ovine submaxillary mucin, group C streptococcal polysac-
charide (group C and H streptococci are specifically a g g l ~ t i n a t e d ~ ~ ' )
and hog group A + H substance, all form precipitates with the aggluti-
nin by virtue of their content of nonreducing, a-D-linked
2-acetamido-2-deoxy-~-ga~actosyl terminal groups.61,63,'00~'78~560-562,570
A
synthetic, carbohydrate-protein conjugate, p-azophenyl2-acetamido-
2-deoxy-~-~-galactopyranoside-bovine serum albumin, was also shown
to precipitate the H. pomatia l e ~ t i n Tay-Sachs
.~~~ ganglioside was
reported to precipitate with the snail lectin.61It was also observed that
guaran and Staphylococcus aureus teichoic acid, containing nonre-
ducing (terminal) a-D-galactopyranosyl and 2-acetamido-2-deoxy-a-~-
galactopyranosyl groups, respectively, also interacted with the ag-
glutinin, whereas teichoic acids containing p-D-linked 2-acetamido-2-
deoxy-D-galactosyl end-groups, or macromolecules having p-D-linked
D-galactopyranosyl end-groups, were ina~tive.~~'"O
Hammarstrom and c o ~ o r k e r s ' ~also
~ investigated a series of

(570) S . Hammarstrom and E. A. Kabat, Biochemistry, 10,1684-1692 (1971).

(571) W. Kohler and 0. Prokop, Z . Immunitaetsforsch. Allerg. Klin. Immunol., 133,
50-53 (1967).
(572) G. Uhlenbruck and W. Gielen, Hoppe-Seyler's 2. Physiol. Chem., 348,1693-1696
(1967).
242 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

lipopolysaccharides for their capacity to precipitate the snail aggluti-

nin. Lipopolysaccharides isolated from Salmonella typhimurium
rough mutants of chemotype Ra and Rb precipitated the Helix pomatia
lectin, whereas no precipitation was obtained with lipopolysaccharides
of the parent, smooth strain, or of Rc, Rd, and Re mutants. It was
suggested that nonreducing 6-O-a-D-galactopyranosy~-D-glucosyl
groups in the lipopolysaccharides provided the binding sites for the
agg1~tinin.l~~
Extensive information on the carbohydrate-binding specificity of the
Helix pomatia lectin was obtained by examining the extent to which a
large number of sugars inhibited the precipitation reaction between
agglutinin and human blood-group A substance or Salmonella
typhinurium SH 180 1 i p o p o l y s a c ~ h a r i d e ~(see ~ ~ ~ ~ ~XI).
~ J ~Table ~~~~

TABLEXI
Inhibition of HeZix pomatia Hemagglutinin-Blood-group A Substance
Precipitation by Various Saccharidesss2

Micromoles required
Saccharide for 50% inhibition

2-Acetamido-2-deoxy-~-galactose 1.65
2-Acetamido-2-deoxy-~-glucose 9.0
DGlucose >18.4
DGalactose >15.8
D-Mannose >22.3
2-Amino-2-deoxy-~galactose >9.7
2-Amino-2-deoxy-~-glucose >18.8
Methyl 2-acetamido-2-deoxy-a-~-galactopyranoside 0.76
Ethyl 2-acetamido-2-deoxy-/3-~-galactopyranoside >3.60
Methyl 2-acetamido-2-deoxy-a-~-glucopyranoside >4.0
Ethyl 2-acetamido-2-deoxy-/3-~-glucopyranoside 10.0
Methyl a-Dgalactopyranoside >10.9
Phenyl 2-acetarnido-2-deoxy-a-~-galactopyranoside 1.65
cY-D-GalNAcp-(1+3)-/3-D-Galp-( 1+3)-~-GlcNAc 0.96

Carbohydrate-specificity studies involving hemagglutination in-

h i b i t i ~ n ~ ~ J (and
" ' * ~sugar
~ ' displacement from S e p h a d e ~ , 5
and
~ ~an im-
munosorbent of human blood-group A substance,567gave essentially
the same results.
Methyl 2-acetamido-2-deoxy-a-~-galactopyranoside was the best in-
hibitor tested.63~556*56',562~567 The observation that a blood-group Type A
pentasaccharide inhibited to approximately the same extent as this
glycoside led to the conclusion that the combining site accommodates a
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 243

single, a-D-linked glycosyl unit.63 Preference for the a anomer was

indicated by the 4-fold higher affinity of the methyl a-glycoside of
2-acetamido-2-deoxy-~-galactose over the parent amino sugar, and the
fact that the ethyl P-D-glycoside was noninhibitory, even at very high
concentration^.^^ 2-Amino-2-deoxy-D-galactosewas a noninhibitor; ap-
parently, a positively charged 2-amino group destabilizes the
carbohydrate-protein complex.63*s6',s62 Methyl 2-acetamido-2-deoxy-
a-D-galactopyranoside is bound four times as avidly as the C-4 epimer
(methyl 2-acetamido-2-deoxy-a-~-glucopyranoside), indicating a pref-
erence for an axial 4-hydroxyl group.63J00~s61*562 On the other hand,
2-acetamido-2-deoxy-~-mannose, the C-2 epimer, does not bind to the
snail agglutinin. The presence of an equatorially oriented N-acetyl or
2-0-acetyl group is essential for strong binding to the hemagglutinin.
D-Galactose is a poor inhibitor, showing less than one percent of the
activity of 2-acetamido-2-deoxy-~-galactose. Melibiose, although a
poor inhibitor, displaced the snail agglutinin from S e p h a d e ~ . ~ ~ ~
The data indicate that H. pomatia lectin exhibits a specificity for
2-acetamido-2-deoxy-cr-~-galactopyranosyl end groups (8),but will also
interact with 2-acetamido-2-deoxy-c~-~-glucopyranosyl and, to a more
limited extent, a-D-galactopyranosy1groups. The hexavalent nature of
the snail lectin must also be considered a factor in its Such
multivalence enhances the affinity of the lectin for multivalent,
configurationally related structures, for example, Sephadex and
g ~ a r a n(See
. ~ ~p. 249 for further binding-studies.)
Cell-binding studies on human erythrocytes, several human,
urinary-bladder, carcinoma cell-lines, and an osteogenic-sarcoma cell-
line have been conducted.189After treatment with neuraminidase,
-80% of human lymphocytes will bind the H . pomatia lectin.
Neuraminidase-treated lymphocytes can also be fractionated on Helix
pomatia hemagglutinin coupled to Sepharose beads.s72b

4. Phaseolus lunatus syn. limensis (Lima Bean)

(lima bean; a-D-GalNAcp > a-D-Gab)
The lima-bean agglutinin holds the distinction of being the first
lectin shown to exhibit blood-group specificity. Although Boyd initially
observed, in 1945, that the lima-bean lectin specifically agglutinated
type A erythrocyte^,^,^ he did not publish his observation until 1949,

(572a) S. Hammarstrom, U. Hellstrom, P. Perlmann, and M.-L. Dillner,J.Exp. Med., 138,

1270-1275 (1975).
(572b) U. Hellstrom, S. Hammarstrom, M.-L. Dillner, H. Perlmann, and P. Perlmann,
Scand. J . lmmunol., 5, Suppl. 5, 45-55 (1976).
244 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

when it appeared together with a report on the agglutinins from a large

variety of plants5
The lima-bean lectin specifically agglutinated type A erythrocytes in
the following order of decreasing activity: A, > AIB > A2 > A2B. Some
varieties of lima bean showed a slight agglutination of type B
ce11s.2*5*22*573The proportion of lectin activity in lima beans differed
from variety to variety, and genetic studies suggested the presence of a
locus for lectin Between 1945 and 1970, investigators study-
ing the lima-bean lectin employed crude, saline extracts or, at most,
partially purified preparation^.^^^.^^^ By fractional precipitation with
alcohol, Boyd and coworkers103achieved partial purification of the
Sieva lima-bean lectin. One-third of the nitrogen present in this prep-
aration was precipitated by hog-mucin, type A substance; ultracen-
trifugal and electrophoretic studies revealed the presence of two, and
four, components, respectively.
Lima-bean lectin precipitated blood-group A and B secretor saliva,
but not 0; it did not precipitate the saliva of any n o n ~ e c r e t o r s . ~ , ~ ~ ' ~ ~
K r i i ~ e substantiated
?~ Boyd's results, using secretor saliva as an in-
hibitor of lima-bean lectin-erythrocyte agglutination. Types AIB, A1A2,
A1O, and A 2 0 saliva all inhibited the lima-bean lectin, whereas type
00 saliva and type 00 ovarian-cyst material were noninhibitory.
In one of the first applications of the Landsteiner hapten-inhibition
technique to lectin studies, Morgan and Watkins22demonstrated that
purified, group A substance, at a dilution of 1:500,000, inhibited lima-
bean-extract agglutination of Al cells, as did 2-acetamido-2-deoxy-~-
galactose, whereas preparations of human ByH, and Lea substances,
and of material extracted from 0 stroma, were without activity at a
dilution of 1:lOO. K r i i ~ verified
e ~ ~ inhibition of the lima-bean lectin by
2-acetamido-2-deoxy-~-galactose,and reported that the lectin did not
react with chicken, guinea pig, mouse, rabbit, or sheep erythrocytes.
Make1a78showed that the lima-bean lectin reacted equally well in
saline, serum, or poly(vinylpyrro1idinone) media with normal, or
papain-treated, human, type A erythrocytes, and he added bovine
erythrocytes to the list of unreactive, red blood-cells. Scheinberg and
c o l l e a g ~ e sstudied
~ ~ ~ *the
~ ~binding
~ of lima-bean lectin to types A, By
and AB cells, and observed that the lectin could be partially purified by
adsorption to A2 cells, followed by elution at 60".
(573)A. Chattoraj,]. Irnrnunol., 98, 757-763 (1967).
(574)K. F. Schertz, W. Jurgelsky, and W. C. Boyd, Proc. Natl. Acad. Sci. U.S.A., 46,
529-532 (1960).
(575)H.M.Bhatia, Y. C. Kim, and W. C. Boyd, Vox Sang., 15,278-286 (1968).
(576)K. C. Atwood and S. L. Scheinberg, Science,l29,963-964(1959).
(577)S. L.Scheinberg and D. T. 0. Wong, J . Immunol., 92,520-528 (1964).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 245

Bhatia and further purified the lectin preparation of

Boyd and coworkers103by using Bio Gel P-200 molecular sieve. The
protein peak first eluted contained the “specific,” anti-A hemaggluti-
nin. However, this preparation gave 3 peaks in the ultracentrifuge, and
hog A substance precipitated only 35% of the protein p r e ~ e n t . “Al-~
though treatment of the “specific” lectin with 2-mercaptoethanol and
2-iodoacetamide had no effect on its activity, the reduced, alkylated
protein gave two peaks in gel chromatography, with the activity resid-
ing in the peak of material of low molecular weight. This experiment
suggested breakdown of the lectin into smaller subunits by reduction of
disulfide bonds.s7sRetention of hemagglutinating activity after reduc-
tion and alkylation is surprising, in view of subsequent results indicat-
ing the necessity of free thiol groups for lectin activity (see later).
Investigating lima-bean extracts for cytoagglutinin activity,
found that sarcoma 180 cells were agglutinated. Inasmuch as adsorp-
tion with type A, red blood-cells removed both the hemagglutinin and
the cytoagglutinin activity of the lima-bean extract, the two activities
appeared to reside in the same molecule; this conclusion is in contrast
to that for the Phaseolus vulgaris lectin (see later). Tunis further
noted579that lima-bean, navy-bean, and kidney-bean activities were
inhibited by (ethylenedinitrilo)tetraacetate,an observation later attrib-
uted to the presence of metal ions in these lectin~.’~’
Affinity labelling of the lima-bean lectin was attempted by Matsubara
and Boyd.580*s81 Diazotized p-aminophenyl a-glycosides of D-glucose,
D-galactose, 2-amino-2-deoxy-D-glucose, 2-amino-2-deoxy-D-
galactose, 2-acetamido-2-deoxy-D-g~ucose, and 2-acetamido-2-
deoxy-D-galactose were used as the affinity labelling compounds. The
diazotization reaction failed to decrease the anti-A activity of the lectin,
but actually increased the ability of the lectin to agglutinate type B
erythrocytes and, in some cases, even promoted the agglutination of 0
erythrocytes. Although these results are interesting, they must be
scrutinized in terms of the experimental conditions. The combination
of using an impure lectin preparation (of 30% purity), glycosides having
unreported properties (some not even synthesized at the time), and a
prolonged reaction-time (22 h), together with the lack of adequate
controls (for example, protection with 2-acetamido-2-deoxy-~-
galactose and other sugars), militates heavily against specific, active-
site labelling; more probably, nonspecific labelling occurred. The

(578) M . Tunis,]. Zmmzcnol., 92, 864 (1964).

(579) M. Tunis,J. Zmmunol., 95,876-879 (1965).
(580) S. Matsubara and W. C. Boyd,J. Zmmunol., 91, 641-643 (1963).
(581) S. Matsubara and W. C. Boyd,J. Zmmunol., 96,25-28 (1966).
246 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

same authorssE2reported that (p-phenylazobenzoy1)ation of the lima-

bean and Sophora japonica lectins enhanced their respective specific
activities.
Boyd and coworkers5E3 also labelled lima-bean lectin with 13'1(indi-
rectly, by coupling of ['311]-p-iodoanilineto protein by d i a z o t i ~ a t i o n ) ~ ~ ~
and visible dyes,s84~sEs and quantitated the binding to type A erythro-
cytes.
The isolation of the lima-bean lectin(PhaseoZus Zunatus) in pure form
was r e p ~ r t e d ' in ~ ~1970.
~ ' On
~ ~employing
~ ~ ~ ~ salt~ ~ fractionation
~ ~ in
conjunction with pH adjustment, followed by gel filtration on Bio-Gel
A-0.5, Gould and ScheinbergsE6isolated two active components from
the lima bean (Phaseolus Zunatus, var. thorogreen). Components I1 and
I11 (designated in order of elution from the gel-filtration column) were
essentially pure by poly(acry1amide) gel-electrophoresis and by ul-
tracentrifbgation; the molecular weights computed were 269,000 and
138,000, respectively. Galbraith and G ~ l d s t e i n ' ~ used ' ~ ' ~specific
~~~~~
adsorption to insolubilized, type A blood-group substance, followed by
elution with 2-acetamido-2-deoxy-~-galactose and recycling, Sephadex
G-200 chromatography, to obtain the same two proteins. Components
I1 and I11 were pure, as shown by poly(acry1amide) gel-elec-
trophoresis, and their molecular weights were 247,100 and 124,400,
respectively. (Pole lima-beans, PhaseoZus Zunatus, var. Carolina or
Sieva, were used for these s t ~ d i e s . ' ~Although
' ~ ~ ~ ~ ) an inactive precipi-
tate slowly formed during prolonged storage at 4", there was no indica-
tion that either component was transformed into the other. A new,
affinity-chromatographic procedure has been employed to purify the
lima-bean lectin~.~lO
On treatment with 1,4-dithiothreitol or 2-mercaptoethanol in
dodecyl sodium sulfate, components I1 and I11 both yielded'g9,5E6 sub-
units of molecular weight 31,000. Poly(acry1amide) gel-
electrophoresis, in the presence of 1%of dodecyl sodium sulfate alone,
gavesE6a component in the range of 60,000. Amino acid analysis (see
later) showed two half-cystine residues per subunit. Direct titration of
components I1 and I11 with 5,5'-dithiobis(2-nitrobenzoic acid), in the
absence or presence of 8 M urea or dodecyl sodium sulfate, gave'99*sE6
(582) S. Matsubara and W. C. Boyd, J . Zmmunol., 96,829-831 (1966).
(583) W. C. Boyd, H. M. Bhatia, M. A. Diamond, and S. Matsubara,]. Zmmunol., 89,
463-470 (1962).
(584) J. T. Miller, W. C. Boyd, and M. A. Diamond, Vox Sang., 13,449-460 (1967).
(585) H. M. Bhatia, C. K. Yang, J. Jaumatte, and W. C. Boyd, ZndianJ. Med. Res., 56,
1525-1530 (1968).
(586) N. R. Gould and S. L. Scheinberg, Arch. Biochem. Biophys., 137, 1-11 (1970).
(587) W. Galbraith and I. J. Goldstein, Methods Enzymol., 28, Part B, 318-323 (1972).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 247

one sulfhydryl group per subunit of molecular weight 31,000. How-

ever, after reduction with 1,4-dithiothreitol, two sulfhydryl groups
could be titrated.1Bs~s86
These data indicated that both components con-
sist of apparently identical subunits of molecular weight 31,000; two of
these polypeptides are linked by an interchain, disulfide bridge, to
form a subunit of molecular weight 62,000, with two thiol groups
remaining free. Component I1 contains four dimers (or eight polypep-
tide chains), and component 111,two dimers (or four chains); the dimers
are held together by strong, noncovalent forces that require strong
detergents or dissociating agents to disaggregate them.
The hemagglutinating activity of the lima-bean lectins is strongly
dependent on the integrity of the free sulfhydryl groups.
N-Ethylmaleimide, 5,5'-dithiobis(2-nitrobenzoic acid), and p-(chloro-
mercuri)benzoate inhibited the activity of the proteins.1BB,s88 2-Acet-
amido-2-deoxy-~-galactose,a specific inhibitor of the hemagglutina-
tion of type A erythrocytes by lima-bean lectin, offered protection
against inactivation by the aforementioned sulfhydryl reagents.s88
Complete, immunological cross-reactivity was demonstrated be-
tween components I1 and I11 with rabbit anti-component I11 antiserum,
thereby providing strong evidence that the two molecular species are
closely related.lBB
Analysis similar amino acid distributions for the lima-
bean lectin components I1 and 111. Neither contained methionine, but
each contained two half-cystine residues per subunit. Both compo-
nents were rich in aspartic acid, serine, and leucine.1s1J0B~s86
The lima-
bean lectins, which are glycoproteins,1s1~1gB~s86 contain 3 4 % of car-
bohydrate consisting of mannose, fucose, and 2-amino-2-deoxyglucose,
and traces of arabinose and x y l o ~ e . ' ~Carbohydrate
~*~~~ analysis of
the glycosyl moiety of the lima-bean lectin gave a structure consist-
ing of four residues of mannose, two of 2-amino-2-deoxyglucose,
and 0.5 molecule of f ~ c o s e Three
. ~ ~ ~ of the D-mannosyl residues
have a-Dglycosidic bonds, which accounts for the precipitation re-
action1s1*'gs,410*588a
between the lima-bean lectin and con A.
Ions of Mn2+and Ca2+were bound to the purified, lima-bean lec-
Removal of Mn2+lowered the hemagglutination titer by 75%.
tins.lS1*lBB
(Ethylenedinitri1o)tetraacetate completely inhibited the precipitin
reaction between lima-bean lectin component I11 and type A blood-
group substance (compare Ref. 579). Several divalent-metal cations
restored activity to the demetallized protein or
(ethylenedinitri1o)tetraacetate-treatedlectin; the addition of Ca2+,

(588) N. R. Could and S. L. Scheinberg,Arch. Biochern. Biophys., 141,607-613 (1970).

(588a) A. Misaki and I . J. Goldstein,]. Biol. Chem., 262,6995-6999 (1977).
248 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Co2+,Fez+,Mg2+,Mn2+,Ni2+, Sr2+,or Zn2+gave equivalent results.'s9

The specific-titer activity of lima-bean lectin components I1 and I11
towards type A human erythrocytes was 5,100 and 1,300, respectively,
and, towards type B erythrocytes, 20 and 5.1, r e s p e c t i ~ e l y .The~~~*~~~
hemagglutinating activity of component I1 is, thus, four times the ac-
t i ~ i t y ' of~~ , ~ ~ ~ 111. Neither component reacted with type 0
component
human, red blood-cells, or native or trypsinized, rabbit erythro-
~ytes.'~~~'~~
Equilibrium-dialysis studiessss using I4C-labelled methyl 2-acet-
amido-2-deoxy-a-~-galactopyranoside gave straight-line, Scatchard
plots for both components, thereby suggesting homogeneous, non-
interacting binding-sites. Component I11 has two binding sites per
mole of protein, and component 11, four binding sites per mole of
protein, with association constants of 1.01 x lo31iter.mole-' and 0.93 x
103 liter.mole-', respectively. Interestingly, tetravalent component I1
was much more m i t o g e n i ~than ~~~divalent component 11.
The reaction of partially purified, lima-bean lectin with hog gastric-
mucin type A substance exemplified the precipitin-like curve obtained
for lectin-polysaccharide or -glycoprotein reactions.103 Classical
precipitin-curves between purified components I1 and I11 and type A
blood-group substance were also Maximal precipitation
of component I1 (equivalence) occurred at a lower ratio of A substance
per mole of protein than for component 111. Under conditions where
type A substance precipitated 90% of the lectin, types A2 and B pre-
~ i p i t a t e d ' ~66 ' ~ ~13%, respectively, of component 11, and 21 and
~ *and
0% of component 111.Neither ofthe lima-bean lectins precipitated with
type 0 blood-group substance.
The specificity of the binding site of the lima-bean lectin has been
probedlS9by hapten inhibition of the precipitation reaction between
components I1 and I11 and blood-group A substance, and by inhibition
of h e m a g g l u t i n a t i ~ n . ~ ~*~~~
2-Acetamido-2-deoxy-~-galactose was the
best monosaccharide inhibitor tested, being respectively 20 and 4
times as potent as 2-acetamido-2-deoxy-~-glucoseand D-galactose. The
preference of the lectin for the a-anomeric linkage shown in formula 8
was established by the three- to four-fold greater inhibitory potency of
methyl 2-acetamido-2-deoxy-a-~-galactopyranoside and methyl a - ~ -
galactopyranoside over the respective
Yosizawa and Miki"" similarly observed that 2-acetamido-2-deoxy-~-
galactose and 0-(2-acetamido-2-deoxy-a-D-galactopyranosyl)-( 1-3)-

(589) W. Bessler and I. J. Goldstein, Arch. Biochern. Biophys., 165, 444-445 (1974).
(590) Z. Yosizawa and T. Miki, Proc. Jpn. Acad., 39, 187-192 (1963).
(591) L. A. Murphy and I. J. Goldstein, unpublished results.
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 249

D-galactose were good inhibitors of the hemagglutination reaction,

whereas 0-(2-acetamido-2-deoxy-~-D-galactopyranosy1)-( 1+3)-~-ga-
lactose was a poor inhibitor. Moreover, melibiose was approximately
twice as good an inhibitor as lactose. Methyl 4-deoxy-Pfluoro-a-D-
galactopyranoside had one-third of the effect of methyl a-D-galaCtO-
pyranoside, whereas D-fucose was very similar to D-galactose in
inhibitory potency."'
Interestingly, methyl 2-deoxy-2-(p-nitrobenzamido)-and -(pamino-
benzamido)-a-D-galactopyranosidewere the most potent inhibitors
tested.lSsThese data suggest that, on the protein, there may be a region
that can interact specifically with an aromatic moiety at C-2 of 2-amino-
2-deoxy-~-galactose.Table XI1 presents some representative, inhibi-
tion data.

TABLEXI1
Inhibition of Lima-bean Lectin by Saccharides"

Micromoles of inhibitor
Inhibitor for 50% inhibition

Methyl 2-deoxy-2-(p-nitrobenzamido)-cr-~-galactopyranoside 0.5

Methyl 2-acetamido-2-deoxy-a-~-galactopyranoside 2.6
Methyl 2-acetamido-2-deoxy-~-~-galactopyranoside 17.0
2-Acetamido-2-deoxy-~-galactose 8.0
2-Acetamido-2-deoxy-~-glucose 158
Methyl a-Dgalactopyranoside 34
Methyl P-D-galactopyranoside 200
Melibiose 28
Lactose 51
D-Galactose 38
D-Fucose 45
Methyl 4-deoxy-4-fluoro-c~-D-galactopyranoside 107

"Inhibition of the precipitation reaction between lima-bean lectin component I11

and human blood-group A s u b s t a n ~ e . ~ ~ ~ . ~ ~ '

A comparative study of the carbohydrate-binding specificity of four

2-acetamido-2-deoxy-~-galactose-binding lectins (from Dolichos
bi$orus, Glycine max, Helix pomatia, and Phaseolus lunatus) has been
conducted by using a series of model macromolecules for direct pre-
cipitation and a variety of mono- and oligo-saccharides as hapten in-
h i b i t o r ~The
. ~ ~data
~ indicated that the combining site of all four lectins

(592)S . Hammarstrom, L. A. Murphy, I. J. Goldstein, and M. E. Etzler, Biochemistry,

16,2750-2755 (1977).
250 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

corresponds to the size of a monosaccharide unit. The contact groups in

2-acetamido-2-deoxy-~-galactose that are probably involved in
hydrogen-bond formation with the four lectins are: for Helix pomatia,
the carbonyl oxygen atom of the 2-acetamido group and the oxygen
atom of the 4-hydroxyl group; for soybean agglutinin (Glycinemax), the
carbonyl oxygen atom of the 2-acetamido group and the hydrogen
atoms of the 4- and 6-hydroxyl groups; for the Dolichos biflorus and
lima bean (PhaseoZus Zunatus) lectins, only the 2-acetamido group was
identified as a probable binding-locus.
In view of its discovery as the first blood-group-specific lectin, it is
surprising that the lima-bean lectin is still one ofthe least studied, plant
agglutinins. Undoubtedly, this lectin merits closer scrutiny.

5. Sophora japonica (Japanese Pagoda Tree)

(Japanese pagoda tree; P-D-GalNAcp > P-D-Galp)
The agglutinin of Sophora japonica seeds was first described by
Kriipe and B r a ~ n . "Shortly
~ thereafter, Morgan and Watkins22obtained
preliminary evidence that the lectin in the seed extract, which aggluti-
nated A and B cells more strongly than 0 erythrocytes, interacted with
both type A and type B blood-group substance. Absorption of the
extract with type A erythrocytes left neither A nor B agglutinating
activity in the supernatant liquor; absorption with B erythrocytes gave
the same result. Furthermore, type A substance inhibited lectin
agglutination of B cells, as well as of A cells, as did type B substance.
Sugar inhibition ofhemagglutination demonstrated the reactivity of the
lectin towards 2-acetamido-2-deoxy-D-galactose, lactose, D-galaCtOSe,
and melibiose (in order of decreasing reactivity), whereas L-arabinose
and D-hcose showed weak inhibiting activity. Later studies of
hemagglutination inhibition employing crude extract^'^ or partially
purified preparation^^^^-^^^ confirmed the work of Morgan and Wat-
kins.22
Osawa and Akiya partially purified S . japonica extract by precipita-
tion with an organic solvent and ammonium sulfate f r a c t i o n a t i ~ n . ~ ~ ~
Ultracentrifugation and starch gel-electrophoresis revealed multiple

(593) M. Kriipe and C. Braun, Naturwissenschaften, 39,284-285 (1952).

(594) T. Osawa and S. Akiya, Bull. Tokyo Med. Dent. Unio., 8,299-305 (1961).
(595) T. Osawa and S. Akiya, Bull. Tokyo Med. Dent. Univ., 8,287-298 (1961).
(596) Z. Yosizawa and T. Miki, Proc. Jpn. Acad., 39, 182-186 (1971).
(597) T. Terao and T. Osawa,J. Biochem. (Tokyo), 74, 199-201 (1973).
(598) R. D. Poretz, Methods Enzymol., 28, Part B, 349-354 (1972).
(599) P. Balding and E. R. Gold, Z. Immunitaetsforsch. Allerg. Klin. Immunol., 145,
156-165 (1973).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 251

components. The protein was heat-labile, agglutinated both A and B

erythrocytes, and contained 17.3% (by weight) of carbohydrate (D-
galactose, D-arabinose, and D-xylose).
The lectin of Sophora japonica has now been purified to
hom~geneity.'*~ Poretz
* ~ ~ ~and coworkers184specifically adsorbed the
saline-extracted protein onto insolubilized A plus H-active hog
gastric-mucin, and eluted the bound lectin with D-galactose. The
physicochemical purity of the D-galactose-eluted protein was assessed
by electrophoretic, immunochemical, and centrifugal methods. Amino
acid analysis revealed a relatively high content of acidic amino acids,
no methionine, and 5 half-cystinyl residues per molecule of molecular
weight 132,800 (determined by gel filtration). The presence of car-
bohydrate was established by chemical analysis [7.8%(by weight) of
mannose, xylose, and 2-acetamido-2-deoxy-glucose], and by reaction
with con A. An unusual, pH profile was obtained by the hemagglutina-
tion assay: activity rose rapidly with increasing pH, to a maximum at pH
8.5.
Early sugar-specificity studies on partially purified S . japonica lectin
suggested that 2-acetamido-2-deoxy-~-galactose was complementary
to the lectin b i n d i n g - ~ i t eD-Galactose,
.~~~ D-fUCOSe, D-gUlOSe, D-talOSe,
2-amino-2-deoxy-D-galactose, and 2-amino-2-deoxy-D-glucose also in-
hibited lectin-B erythrocyte a g g l ~ t i n a t i o nP-D-Galactopyranosides
.~~~
were better inhibitors than a-D-galactopyranosides. Whereas methyl
groups at 0-2,-3,and -6of D-galactose did not diminish the inhibitory
capacity, 4-0-methyl-D-galactose was a noninhibitorVSg4 On a weight
basis, blood-group A and B substances were far more reactive than any
of the simple sugars tested.594
Sugar inhibition of highly purified, lectin-type B blood-group sub-
stance precipitation was achieved by Poretz and coworker^.^^^,^^
Hemagglutination inhibition studies were conducted by Irimura and
coworkers.lm This work is summarized in Table XIII. A comparison of
2-acetamido-2-deoxy-~-galactose(and its glycosides) with D-galaCtOSe
(and its glycosides) confirmed a preferential binding of the lectin to the
2-acetamido-2-deoxy-~-galactosyl structure. Moreover, on contrasting
the P-glycoside of either sugar with the a-linked anomer, it is evident
that S.japonica exhibits a stronger affinity for the p-anomeric configura-
tion. Among p-D-linked D-galactopyranosides, the aromatic aglycon
contributed importantly to stabilization of the lectin-saccharide com-
plex in comparison to the aglycon of alkyl glycosides. N-Acetyl-
lactosamine is of special interest, in that it inhibits hemagglutination to
the same extent as phenyl 2-acetamido-2-deoxy-P-D-
galactopyranoside, regardless of the apparent preference of the lectin
for 2-acetamido-2-deoxy-~-galactose over D-galaCtOSe, and aromatic
252 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXI11
Sugar Inhibition of Sophora japonica Lectin

Micromoles required Concentration required

for 50% inhibition for B erythrocyte
of B-substance hemagglutination
Sugar p r e c i p i t a t i ~ n ' ~ ~ ~ ' ~inhibitionla
~ (mM)

2-Acetamido-2-deoxy-D-galactose 4.6
p-Nitrophenyl2-acetamido-2-
deoxy-P-D-galactopyranoside 0.21
Phenyl2-acetamido-2-deoxy-P-~-
galactopyranoside 0.8
Phenyl2-acetamido-2-deoxy-a-~-
galactopyranoside 1.6
Methyl 2-acetamido-2-deoxy-P-D-
galactopyranoside 2.7
Methyl 2-acetamido-2-deoxy-a-~
galactopyranoside 3.5 2.7
D-Galactose 25
p-Nitrophenyl P-D-gdactopyrano-
side 1.2
Phenyl P-Dgalactopyranoside 3.9 6.3
Phenyl a-D-galactopyranoside 13
Methyl P-D-galactopyranoside 11
Methyl a-Dgalactopyranoside 18
2-Acetamido-2-deoxy-4-O-P-D-
galactopyranos yl-D-glucose
(N-acetyllactosamine) 0.81
Lactose 6.6
Melibiose 20

over alkyl glycosides.'@ This result suggests the possibility of an ex-

tended binding-site complementary to a unit larger than a monosac-
charide unit. Indeed, Balding and Gold599found that the strongest
hemagglutination inhibitors were sugars in which a
D-galactopyranosyl group was linked 3-0-p- or 4-0-p- to the penulti-
mate 2-acetamido-2-deoxy-~-glucopyranosylresidue: N-acetyl-
lactosamine, lacto-N-biose I, and lacto-N-tetraose. The data suggest a
specificity for 2-acetamido-2-deoxy-~-~-galactopyranosyl groups (9).

NHAC
9
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 253

The observation that S. japonica lectin agglutinates B erythrocytes

slightly more strongly than A cells, whereas 0 cells are agglutinated
only at high concentrations of l e ~ t i n , together ~ ~ ~with ~ ~sugar
J ~ ~ ~ ~
inhibition data, suggests that the lectin may react with non-ABO
blood-group structures. Some investigators have reported agglutina-
tion of erythrocytes regardless of blood type594*596; others have found
variable results, depending on the source of the seeds.5g9
In order to clarify the nature of the erythrocyte structure responsible
for S.japonica lectin binding, Chien and coworkersa0'investigated the
relationship between cell agglutination and the I antigenic determin-
ant. By comparison of the reactivity of purified lectin with human, red
blood-cells of various I phenotypes (adult I, adult i, cord blood i) and
that of human anti-B or anti-A sera with the same cells, theya0' dem-
onstrated that lectin agglutination was closely related to the presence
of the I antigen. Furthermore, the hemagglutination inhibiting capacity
of Smith-degraded, type B blood-group substance having decreased
nonreducing (terminal) a-D-galactosyl and a-L-fucosyl groups, but in-
creased P-D-galactosyl termini, in comparison to that of native B sub-
stance was not diminished, nor was its capacity to precipitate the lectin
from solution substantially lessened. However, this chemical degrada-
tion completely destroyed the B determinant, as evidenced by the low
inhibitory activity of degraded B substance in the B erythrocyte-anti-B
serum agglutination reaction. In addition, digestion of B substance
with a-D-galactosidase, which diminished the nonreducing (terminal)
a-D-galactosylgroups, without concomitantly increasing the number of
P-D-galactosyl termini, increased the concentration of oligosaccharide
required for hemagglutination inhibition without altering that needed
for inhibition of lectin agglutination. Enzymic digestion substantially
lessened the capacity of B-active blood-group substance to precipitate
the lectin from solution. Chien and coworkersao1concluded that S.
japonica lectin reacts with P-D-galactosyl residues (believed to repre-
sent the I antigenic determinant)602common to both A and B erythro-
cytes of I phenotype.
Balding and Gold599studied the agglutination of various, red blood-
cell types by extracts ofSophora japonica seeds from varied geographi-
cal sources. Whereas a purified lectin sample from R. D. Poretz (seeds
from a supplier in the United States) and lectin from Japanese seeds did
not agglutinate 0 cells, a sample from Portugal did. All three lectin

(600) J. T. Miller and W. C. Boyd, Vox. Sang., 13,209-217 (1967).

(601) S. M. Chien, T. Lemanski, and R. D. Poretz, Immunochemistq, 11, 501-506
(1974).
(602) T. Feizi, E. A. Kabat, G. Vicari, B. Anderson, and W. L. Marsh,]. E x p . Med., 133,
39-52 (1971).
254 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

samples reacted strongly with OEn(a-) cells, which contain a lowered

proportion of surface sialic acid due to a genetic abnormality. If it is
assumed that the lectin can react with a- or P-D-galactosides or
2-acetamido-2-deoxy-~-galactosides, the lectin may interact not only
with blood-group A and B substances, but also with nonreducing
(terminal) P-D-galactosyl groups on (a) I-precursor substance (incom-
pletely biosynthesized A or B substance), (b)incompletely biosynthe-
sized, or partially degraded, M or N substances, (c) desialized,
myxoviral receptors, and (d) OEn(a-)
Irimura and coworkerslmemployed a series of native and degraded
glycopeptides as inhibitors of S. japonica lectin hemagglutination.
Desialized, porcine thyroglobulin gave good inhibition by virtue of its
O-P-D-galactopyranosyl-(l-*4)-2-acetamido-2-deoxy-~-~-glucopyrano-
syl nonreducing termini. On the other hand, desialized bovine sub-
maxillary-mucin did not react, despite the presence of O-(e-acetamido-
2-deoxy-a-D-galactopyranosyl)-serine (or -threonine), which may have
been sterically unavailable to the lectin. Finally, acid hydrolysis of
nonreducing (terminal) a-L-fucosyl groups from degraded, porcine
submaxillary-mucin exposed unsubstituted 0-0-D-galactopyranosyl-
(1-*4)-2)-acetamido-2-deoxy-a-~-galactopyranosyl-serine (or -threo-
nine), resulting in a two-fold diminution of inhibitory capacity. This
result implies that a-L-fucose either represented a binding locus, or
conferred a more favorable conformation on the disaccharide to which
it was attached.

V. BGALACTOSE-BINDING
LECTINS
1. Abrus precatorius (Jequirity Bean)
(jequirity bean; P-D-Gab > cw-D-Gab)
In his very early study of Ricinus communis and Abms precatorius
seed-extracts, Stillmark observed toxic and hemagglutinating activities
in both.603The strongly toxic effects of the extracts were originally
attributed to the hemagglutinating a ~ t i v i t y , 6but
~ ~ it, ~has
~ ~since been
demonstrated that these activities are associated with distinct proteins
present in Abrus precatorius e x t r a ~ t s . ' ~Interest
~ ' ~ ~ *in~purifying
~~ the
toxic protein, abrin, was greatly enhanced by reports of its anti-tumor
activity.32Olsnes and coworker^^^'*^^^ purified abrin by a combination of
ion-exchange and affinity chromatography, and studied the mechanism

(603)H.Stillmark, Arb. Phannakol. Inst. Dorpat, 3,59-151 (1889).

(604)H. Hellin, Ph.D. Thesis, Universitkt zu Dorpat (1891).
(605)A. H.Kahn, B. Gul, and M. A. Rahman,]. Immunol.,96,554-557(1966).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 255

of its toxic action. The protein gave a single band in disc-gel elec-
trophoresis, and exhibited a molecular weight of 65,000 by gel filtra-
tion. Abrin was split into two polypeptide chains of molecular weights
35,000 and 30,000 on reduction with 2-mercaptoethan01.~~~ The larger
nontoxic subunit, designated the B chain, bound to carbohydrate, as
evidenced by its ability to agglutinate erythrocytes in the presence of
anti-abrin serum.147,150 The smaller chain, designated the A chain, was
incapable of agglutinating erythrocytes in the presence of anti-abrin
serum, but manifested considerable activity as an inhibitor of protein
synthesis in a cell-free system. Olsnes and coworkersao6concluded that
the mechanism of action of the toxins abrin and ricin was virtually
identical: the toxin binds to the cell surface by virtue of the B chain; and
the A chain is pinocytosed, whereupon it inhibits the chain-elongation
step of protein synthesis.
An alternative purification of abrin, involving chromatography in
three separate, ion-exchange systems, gave two toxic proteins desig-
nated607abrin A and abrin C. The two toxins were homogeneous by
sedimentation and electrophoretic analysis, and similar in molecular
weight (abrin A, 60,100; abrin C, 63,800), but differed slightly in amino
acid composition. However, the proteins differed markedly in their
affinity for Sepharose 4B:abrin A was not bound, whereas abrin C was
bound, and was eluted specifically with D-galactose. Wei and cowork-
ers8O7suggested that abrin C may be identical to the toxin studied by
Olsnes and Pihl,lS0and abrin A to that reported by Lin and coworkers."*
The effect of utilizing widely different seed-sources in the three inves-
tigations is difficult to assess. On reduction with 2-mercaptoethanol,
the abrin A of Wei and coworkers607gave subunits of molecular weights
32,000,29,500, and 28,000 in the ratios of 2: 1:1, whereas abrin C gave
equal proportions of subunits of molecular weight of 33,000 and 28,000.
Abrin C was crystallized by Wei and Einstein,609and preliminary, X-ray
crystallographic studies were made.
Limited data are available with respect to the carbohydrate-binding
specificity of abrin. Olsnes and coworkers147demonstrated that the
agglutination of erythrocytes by abrin in the presence of anti-abrin
antiserum was inhibited best by D-galactose. Lactose, melibiose, and
D-fucose (in the order of decreasing potency) gave weaker inhibition,
whereas D-glucose, D-mannose, D- and L-arabinose, D-xylose,

(606) S. Olsnes, K. Refsnes, and A. Pihl, Nature, 249,627-631 (1974).

(607) C. H. Wei, F. C. Hartman, P. Pfuderer, and W.-K. Yang,J. Biol. Chern., 249,
3061-3067 (1974).
(608) J.-Y. Lin, Y . 4 . Shaw, and T.-C. Tung, Toxicon, 9,97-101 (1971).
(609) C. H. Wei and J. R. Einstein,J. B i d . Chem., 249,2985-2986 (1974).
256 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

D-fructose, D-ribose, and maltose were without effect. Equilibrium-

dialysis analysis147revealed one binding site for lactose on the abrin
molecule, with an apparent association constant of 8 x lo3 it4-l.
The hemagglutinin of A. precatorius was also purified and charac-
terized, by both Olsnes and coworkers'47and Wei and coworkers.610
The latter authors reported that a lectin preparation, homogeneous by
electrophoresis, isoelectric focusing, and sedimentation analysis, gave
a molecular weight of 130,000&5,000 by gel filtration."O A molecular
weight of 125,400 to 126,000 was obtained by equilibrium sedimenta-
tion and calculation from the chemical composition. The protein exhib-
ited an isoelectric point at pH 5.0, a carbohydrate content (mannose)
of5% by weight, and subunits ofmolecular weight 33,800 and 32,200 in
the presence of 2-mer~aptoethanol.~~~
Olsnes and c o ~ o r k e r s ' ~concluded
' that their lectin preparation
probably contained two highly similar proteins that differed in their
subunit structure. After fractionation on DEAE-cellulose, the active
fraction was specifically adsorbed to Sepharose 4B, and was eluted with
D-gahCtOSe. The resultant preparation, after chromatography on CM-
cellulose, had a very high hemagglutinating activity and very little
toxicity, and migrated as a single, symmetrical peak when sedimented
in a sucrose density-gradient. The protein of molecular weight 134,000
split into two components having molecular weights of 68,000 and
69,000 in the presence of dodecyl sodium sulfate, and further into
subunits of molecular weight 36,000, 35,000, and 33,000 in the
presence of 2-mercaptoethanol and a detergent. The relative amounts
of the two larger polypeptide chains was variable, but that of the small
chain was constant. The authors suggested that the two agglutinins
differ in their heavy chain but share the subunit of molecular weight
33,000. Again, a disparity in the seed source used by Olsnes and co-
w o r k e r ~and ~ ~by
~ Wei and coworkers610may account, in part, for the
differences noted by the two groups.
Roy and coworkers6'l purified both the toxin and the lectin ofAbrus
precatorius Linn. by using a combination of ammonium sulfate frac-
tionation, CM-cellulose ion-exchange, and affinity chromatography on
Sepharose 4B. The two proteins were obtained in crystalline form,
exhibiting homogeneity by immunochemical and ultracentrifugal
criteria. The molecular weight reported for the lectin, namely, 132,000,
agreed with previous values. 147m0 Biophysical characterization gave
s Z O ~ , =~ 6.7 S, DZOo,w= 4.7 x 10-7cm2.sec-1,andF = 0.73for the protein.611
(610) C. H. Wei, C. Koh, P. Pfuderer, and J. R. Einstein,J. B i d . Chern., 250,4790-4795
( 1975).
(611) J. Roy, S. Som, and A. Sen, Arch. Biochern. Biophys., 174,359-361 (1976).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 257

A pH-dependent dissociation of the tetramer to dimers of molecular

weight 64,000 occurred between pH 2 and 7 .Analysis ofthe C-terminal
amino acid showed both alanine and leucine, and the N-terminus was
valine. The possibility of multiple isolectins was indicated by multiple
band-formation of isoelectric-focused hemagglutinin.611
A. precatorius lectin agglutinates human type B and 0 erythrocytes
more effectively than type A cells.'47The agglutination reaction was
inhibited by D-galactose, lactose, melibiose (in order of decreasing
activity), and, to a lesser extent, by L-arabinose and D-fUCOSe. Although
limited, the data suggest a specificity for P-Dgalactopyranosyl groups
(10). Two binding-sites for lactose were observed by Olsnes and co-

CH,OH

OH
10

workers14' in equilibrium-dialysis experiments; the calculated associa-

tion constant was 8 x 103M - l , which was identical to that of abrin. Wei
and coworkers obtained crystals ofA. precatorius lectin, and conducted
preliminary X-ray crystallographic studies.610Whereas fresh prepara-
tions of abrin were extremely toxic and nonmitogenic, storage of the
glycoprotein for several months at 4"rendered it relatively nontoxic and
highly mitogenic.612The chemical relationship between the subunits of
the toxin and the agglutinin has not yet been established, nor have the
stereochemical features involved in carbohydrate binding to either
protein been studied in detail.

2. Arachis hypogaea (Peanut)

[peanut;P-D-Gab-(1+3)-~-GalNAc > D - G ~ N H =
, a-D-Galp]
Polyagglutinability, the agglutination of erythrocytes irrespective of
blood type by a high percentage of adult, human sera, is a problem
encountered in routine serology.613T-Polyagglutinability usually arises
as a result of in vitro contamination of blood specimens (although
authentic cases of in vivo T-polyagglutinability have been reported)
through the degradative action of bacterial neuraminidase. Until re-

(612) S. J. Kaufman and A. McPherson, Cell, 4,263-268 (1975).

(613) G. W. G. Bird and J. Wingham, Scand. J. Haematol., 8,307-308 (1971).
258 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

cently, complex and time-consuming absorption studies were needed

in order to establish the presence of the T-antigen. With the discovery
of the T-antigen-reactive peanut lectin, determination of
T-polyagglutinability was greatly ~ i m p l i f i e d . ~Agglutination
'~,~'~ of A, B,
or 0 erythrocytes by peanut lectin occurs only after digestion of these
cells with neuraminidase, which exposes the T determinant.20',202
Two affinity-purification schemes for the lectin in peanuts have been
reported.202*61sLotan and coworkers chromatographed a solubilized,
ammonium sulfate fraction of the seed extract on Sepharose-N-(6-
aminohexanoyl)-~-D-galactopyranosylamine.202 The bound lectin was
eluted with D-galactose as a single peak in 87%yield. Homogeneous by
disc-gel electrophoresis (pH 4.3, 7.2, and 8.9), gel filtration, and
sedimentation analysis, the protein exhibited a molecular weight of
110,000by gel filtration. A molecular weight of 111,000 was calculated
from biophysical data ( S Z " ~ , ~= 5.7 S, D&,w= 5.0 x lo-' cm2.sec-', ij=
0.73). In the presence of detergent, the protein dissociated into
subunits of molecular weight 27,000-28,000, The amino acid composi-
tion reported by Lotan and coworkers showed a high content of acidic
and hydroxylic amino acids, relatively little methionine, tryptophan,
and histidine, and the complete absence of cysteine. The finding of a
unique sequence for the five NH2-terminal amino acids of the peanut
lectin suggested four identical subunits.202Sequence homology with
the p-chain of the lentil (see Section II,2) and pea (see Section II,3)
lectins and the soybean lectin (see Section IV,2) has been S ~ O W I I . ~ ~ ~
Peanut lectin has also been isolated on poly(acry1amide)-entrapped
guaran beads and on poly(acry1amide) copolymerized with ally1 a - ~ -
galactopyrano~ide.'~"
Terao and coworkers fractionated the seed extract on Sepharose 6B;
peanut lectin was retarded with respect to contaminating proteins.816
The homogeneous protein (disc gel-electrophoresis at pH 4.3 and 7.1,
and by ultracentrifugation) had a molecular weight of 106,500, calcu-
lated from sedimentation data ( S L = , ~6.05 S). Detergent-dissociated
subunits had molecular weight 27,000. The chemical composition re-
ported was at variance with the results of Lotan and coworkers.202Terao
and coworkersrn3reported considerably less serine, threonine (due,
perhaps, to their using unextrapolated values), tryptophan, and ar-
ginine; moreover, they found no methionine, but reported 16.6 moles

(614)W. C. Boyd, D. M. Green, D. M. Fujinaga, J. S. Drabik, and E. Waszczenko-

Zacharczenko, Vox Sung., 4,456-467 (1959).
(615)G. W. G. Bird, Vor Sung., 9,748-749 (1964).
(616)T.Terao, T.Irimura, and T. Osawa,Hoppe-Seyler's Z. Physiol. Chem., 356,1685-
1692 (1975).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 259

of cysteine per mole of lectinS6l6Neither group found covalently bound

carbohydrate.202*616
Early studies on the carbohydrate-binding specificity of peanut lec-
tin suggested that P-glycosidically linked D-galactosyl residues might
be an important part of the T-antigen to which the lectin bound.
Uhlenbruck and coworkerszo1found particularly strong inhibition of
hemagglutination by 2-acetamido-2-deoxy-3-O-~-~-galactopyranosyl
D-galactose, and by glycoproteins and gangliosides carrying this
disaccharide in a nonreducing (terminal) position. Later studies
confirmed these results.202,203,604 The investigations of Dahr and
coworker^^'^,^*^ suggested that the peanut-lectin receptor was part of
the base-labile oligosaccharide 11 that is responsible for MN blood-

a-AcNeu - (2-3)-p-o-Galp - (1- 3)-GalNAc- (l+O)-Ser, Thr

6
t
2
a-AcNeu
11

group activity. Desialization by neuraminidase or acid hydrolysis re-

vealed the peanut-reactive structure, which was labile to digestion
with periodate, D-galactose oxidase, and P-D-galactosidase.
An extensive, immunochemical investigation of the specificity of
peanut agglutinin was conducted by Pereira and coworkers.z03Cor-
roborating previous work, only derivatives of D-galactose gave sig-
nificant inhibition of peanut agglutinin-blood-group precursor-
substance precipitation (see Table XIV). Of the monosaccharides
tested, 2-amino-2-deoxy-~-galactosewas the most effective inhibitor,
whereas 2-acetamido-2-deoxy-~-galactoseand S-deoxy-~-lyxo-
hexoseZo2were inactive, suggesting the importance of a hydrogen-
bonding substituent on C-2. By comparison of methyl a-D-
galactopyranoside with its 6-0-methyl-substituted counterpart, and of
D-galactose with D-fucose, it is evident that the 5-(hydroxymethyl)
group is also implicated in lectin-saccharide interaction. Lotan and
coworkers further reported that D-galaCtOSe 6-sulfate7 D-galacturonic
acid, and L-arabinose are noninhibitory, confirming the importance of
an unsubstituted, 5-(hydroxymethyl) group.z02Monosaccharides lack-
ing inhibitory activity were D-glucose,2-acetamido-2-deoxy-~-glucose,
D-mannose, 2-acetamido-2-deoxy-D-mannose, L-fucose, L-rhamnose,
and maltose.z0z~203
(617) W. Dahr, G. Uhlenbruck, and G. W. G . Bird, Vox Sung.,27,29-42 (1974).
(618) W. Dahr, G. Uhlenbruck, and G. W. G. Bird, Vox Sung., 28, 133-148 (1975).
260 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXIV
Inhibition of Peanut Agglutinin-Blood-group Precursor Substance
Precipitation by saccharide^'^^

Micromoles required
Sugar for 50% inhibition

D-Galactose 6.0
2-Amino-2-deoxy-D-ga1actose 2.7
Methyl a-D-galactopyranoside 2.7
p-Nitrophenyla-D-galactopyranoside 2.7
Methyl P-D-galactopyranoside 4.0
p-NitrophenylP-Dgalactopyranoside 4.0
Methyl 6-O-methyl-a-~-galactopyranoside 6.0
DFucose 10.0
2-Acetamido-2-deoxy-~-galactose >60
2-Acetamido-2-deoxy-3~-~-D-galactopyranosy~-~-ga~actose 0.11
2-Acetamido-2-deoxy-3-O-~-galactopyranosyl-~-galactito~ 2.7
3-0-a-D-Galactopyranos yl-D-galactose 7.0
6-O-a-~-Galactopyranosy~-6-0-c~-~galactopyranosyl-P-D-g~uco-
pyranosyl D-fmctoside 7.0
3-0-(2-Acetamido-2-deoxy-c~-D-ga~actopyranosy~)-D-ga~actoseinactive at 4.0
4-O-P-D-Galactopyranosyl-D-g~ucose 2.7
6-O-~-D-Galactopyranosyl-D-glucose 2.7
6-O-a-~-Galactopyranosyl-D-glucose 10.0
2-Acetamido-2-deoxy-3-O-~-~-ga~actopyranosyl-D-glucose 10.0
2-Acetamido-2-deoxy-4-O-~-~-galactopyranosyl-~-glucose 1.5

The disaccharide 2-acetamido-2-deoxy-3-O-~-D-ga~actopyranosy~-D-

galactose (12)shows the corresponding group was, without question,

OH
12

the most complementary to the lectin binding-site. Reduction of this

disaccharide resulted in a 25-fold decrease in reactivity. Substitution at
0 - 2 or -4of the nonreducing (terminal) D-galactosyl group with an L-
fucosyl or a 2-acetamido-2-deoxy-~-glucosyl group, respectively,
greatly diminished the inhibiting capacity. This finding suggests that
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 26 1

peanut lectin does not recognize (internal) D-galaCtOSyl residues. The

results of Irimura and coworkers183are at variance with these findings,
in that substitution of an L-fucosyl group on 0-2 of the nonreducing
(terminal) D-galaCtOSyl group of 2-acetamido-2-deoxy-O-~-~-galacto-
pyranosyl-D-galactopyranosylserine/threonine(the oligosaccharide
core of porcine, submaxillary mucin) increased the binding to peanut
lectin by a factor of 2.5. Further substitution by a 2-acetamido-2-deoxy-
D-galaCtOSyl group at 0-3of the same D-galactosyl residue did not alter
the binding characteristics. These discrepancies remain unexplained.
With respect to the anomeric carbon atom of nonreducing (terminal)
D-galactosyl groups, a discrepancy exists between monosaccharides
and disaccharides. Whereas, methyl a-D-galactopyranoside was 1.5
times as effective as methyl P-D-gdactopyranoside, 6-0-P-D-galacto-
pyranosyl-D-glucose was 3 times as potent as its a-D-linked counter-
part; lactose gave better inhibition than melibiose,zOzand other
oligosaccharides having nonreducing (terminal) a-D-galactosyl groups
were relatively poor inhibit01-s.~~~ A preference for 4-0-P-D-linkage
over 3-0-p-D-linkage was suggested by comparison of the activity of
2-acetamido-2-deoxy-4-0-~-~-ga~actopyranosyl-~-g~ucose with that of
2-acetamido-2-deoxy-3-O-~-~-galactopyranosyl-~-glucose.
The precipitin reaction with peanut agglutinin revealed consider-
able heterogeneity among blood-group substances having the same
Thus, some samples of A-, B-, and H-active substances
precipitated almost 100% of the lectin at equivalence, whereas others
failed to react, or gave an intermediate result. In no case did A,
substances give precipitation. Strikingly, all A-, B-, and H-active
substances gave very strong precipitin reactions after mild, acid
hydrolysis, or one-step Smith degradation. Moreover, all precursor
blood-group substances (I-active) gave strong precipitin reactions. The
authorszo3suggested that the peanut agglutinin reacts with a deter-
minant other than that which accounts for A, Byor H specificity, which
although normally concealed, can be exposed by chemical degradation.
Incomplete biosynthesis of oligosaccharide chains, and variable, steric
hindrance of the access of lectin to reactive, short chains by unreactive,
large oligosaccharide chains were explanations invoked by Pereira and
coworkers to account for the blood-group substance heterogeneity
observed with peanut l e ~ t i n . ~ ~ ~
A major disagreement exists with respect to the biological activity of
peanut lectin.fi16*619 Terao and coworkers616could not demonstrate
mitogenicity of peanut lectin towards either normal, or neuraminidase-

(619) A. Novogrodsky, R. Lotan, A. Ravid, and N. Sharon,J. Immunol., 115,1243-1248

(1975).
262 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

treated, human, peripheral-blood lymphocytes. On the other hand,

Novogrodsky and coworkers observed stimulation of DNA synthesis in
neuraminidase-treated, rat or human lymphocytes by this l e ~ t i n . ~ l ~
Untreated lymphocytes did not respond, nor did normal or neurami-
nidase-digested lymphocytes of the mouse or guinea pig.61gThe ques-
tion of the mitogenicity of peanut lectin, therefore, remains unresolved.

3. Bandeiraea simplicifolia I
(cY-D-Gab > a-D-GalNAcp)
A human blood-group B-specific phytohemagglutinin, first observed
in Bandeiraea simplicifolia seeds by the Makelas,620has now been
purified by Hayes and G~ldstein.'~' An ammonium sulfate fraction of
the seed extract was chromatographed on a melibionyl-Bio-gel P-300
column (compare Refs. 148,149b, and 150c). Elution ofthe specifically
adsorbed protein was effected with D-galactose. The protein appeared
homogeneous by disc gel-electrophoresis (pH 4.3), immunoelectro-
phoresis, gel filtration, and sedimentation analysis. Electrophoresis at
pH 9.5, and isoelectric focusing, revealed multiple bands. Later studies
revealed the presence of multiple, lectin specificities in Bandeiraea
simplicifolia ~ e e d ~The . a-D-galactos
~ ~ ~ * ~yl-binding~ ~ lectin, designated
BS I, was completely precipitated from solution by the galactomannan
guaran. A molecular weight of 114,000 was calculated from sedimen-
tation velocity centrifugation ( S O Z ~ , ~ = 7.52 x S, Dio,w= 5.2 x
lo-' cm2.sec-', ij= 0.69 cm3.g-'), and was corroborated by gel filtration
and calculation based on the chemical composition. The glycoprotein
lectin (9.0%, by weight, of carbohydrate, namely, mannose, fucose,
xylose, and 2-amino-2-deoxyglucose) dissociated in the presence of
detergent into subunits having molecular weight 28,500. Amino acid
analysis demonstrated the abundance of hydroxylic and acidic amino
acids, -0.5 molecule of methionine, and one free-thiol cysteine resi-
due per subunit.
Chemical modification, with 5,5'-dithiobis(2-nitrobenzoic acid), of
2-3 molar proportions of cysteine residues per mole of protein de-
stroyed the hemagglutinating activity.I3l However, both native and
demetallized lectin failed to react with methyl methanethiosulfonate
(MeSS02Me), indicating that the thiol groups were most probably
"buried," and not directly involved in carbohydrate- or metal-
binding.622Amidation of 9 carboxyl groups per subunit with glycine
methyl ester hydrochloride lowered the precipitating capacity of the
(620) 0. Makela and P. Makela, Ann. Med. E x p . Biol. Fenn., 34,402-404 (1956).
(621) I. J. Goldstein, L. A. Murphy, and T. Ebisu, Pure Appl. Chem., 49, 1095-1103
(1977).
(622) J. Lonngren and I. J. Goldstein, Biochim. Biophys. Acta, 439, 160-166 (1976).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 263

polysaccharide622by 80%. Inclusion of methyl a-D-galactopyranoside

during amidation afforded some protection; 8 residues per subunit
were modified, giving a protein that had 80% of the activity of native
lectin. On the other hand, acetylation of 8 lysyl side-chains and 2
hydroxyl or thiol groups per subunit caused virtually no perturbation of
the carbohydrate-binding activity.622These studies suggest that car-
boxyl groups participate in carbohydrate binding, either directly, or
indirectly by way of stabilization of the conformation or metal binding,
whereas free amino and sulfhydryl groups do not.622
The BS I lectin requires bound calcium for activity.131Two moles of
calcium and 1.25moles of magnesium per mok of protein were found by
atomic absorption spectroscopy. Inactive, metal-free lectin, obtained
by exhaustive dialysis, could be reconstituted by addition of calcium,
cadmium, or strontium (magnesium restored 80% of the activity). Al-
though bound-calcium was not removed by dialysis against EDTA,
inclusion of this chelating agent in the precipitin reaction resulted in
complete inhibition.
Conformational analysis of BS I by c.d. spectroscopy indicated that
the protein contained 30-40% of structure in its native conforma-
t i ~ nThe. ~c.d.
~ ~spectrum was relatively insensitive to alteration in pH,
removal of bound metal, and addition of methyl a-D-galactopyranoside.
However, addition of dodecyl sodium sulfate or 2,2,2-trifluoroethanol
resulted in the formation of some a-helical structure, and was accom-
panied by the loss of polysaccharide-precipitating capacity. Urea (8 M )
irreversibly denatured the lectin.
BS I agglutinates human, type B and AB erythrocytes strongly, and Al
cells weakly, and does not agglutinate Az or 0 ce11s.131,195-624
(Old seed
samples were reported to be more specific for B erythrocytes than fresh
seeds, which contained some anti-A activity.6zo)Polysaccharides and
glycoproteins having terminal (nonreducing) a-D-galactopyranosy1
groups gave131*6zs a precipitin reaction with BS I. Thus, type B blood-
group substance and a-D-galactopyranosyl-substituted, branched
polysaccharides precipitated all of the lectin from solution under opti-
mal conditions, whereas type A blood-group substance gave a much
diminished precipitate. Type Az and H ( 0 ) substances did not react, nor
did fetuin or orosomucoid (either native or de~ialized).'~~ A series of
polysaccharides was s t ~ d i e d 'by ~ ~Ouchterlony-diffusion
~ ~ ~ ~ and
quantitative-precipitin analysis with BS I. The shape of precipitin
curves with six leguminous-seed galactomannans correlated with their

(623) J. Lonngren, I. J. Goldstein, and R. Zand, Biochemistry, 15,436-440 (1976).

(624) W. J. Judd, E. A. Steiner, B. A. Friedman, C. E. Hayes, and I. J. Goldstein, Vox
Sang., 30,261-267 (1976).
(625) C. E. Hayes, L. A. Murphy, and I. J. Goldstein, to be published.
264 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

galactose to mannose ratio (C yamopsis tetragonolobus, Cassia alata,

Ceratonia siliqua L., Caesalpinia spinosa, Crotalaria mucronata, and
Leucaena glauca). Weaker precipitin-reactions were obtained with
Tsuga canadensis galactoglucomannan and Torulopsis gropengiesseri
galactomannan. Larch egalactan and sugar-beet L-arabino-P-D-
galactan did not react.131A synthetic melbionate-bovine serum albu-
min conjugate also precipitated with the lectin.lz6Cell-surface galacto-
mannan from Schixosaccharomyces pombe cells was localized by gold-
labelled BS I lectin,BZsaand a-D-galaCtOSyl-COntining components on
mouse neuroblastoma cells by 1311-labelledlectin.62sbThe BS I lectin
was also shown to give rise to an immediate-type skin-reaction in a
sensitized, laboratory worker.62sc
Makela and colleagues investigated the specificity ofB. simplicifolia
seed-extracts by hemagglutination i n h i b i t i ~ n . ~ ~ The
, ' ~ ~a - ~ -
galactosides melibiose and raffinose gave good inhibition, as did
D-galactose and 2-acetamido-2-deoxy-~-galactose. D-Fucose and
L-arabinose were weakly inhibitory.
The carbohydrate-binding specificity of BS I has been studied in
detail by sugar inhibition of lectin-galactomannan p r e c i p i t a t i ~ n . ' ~ ~ , ~ ~ ~
Representative data are summarized in Table XV. That BS I preferen-
tially binds a-D-linked D-galactopyranosides was readily established
by comparison of glycosides having nonreducing (terminal) a - ~ -
galactopyranosyl groups with those which are P - ~ - l i n k e d . ~Fur- ~"~~
thermore, a-glycosides of D-galactose were bound with equivalent
affinity, regardless of the nature of the aglycon, whereas P-glycosides
having aromatic aglycons were more avidly bound than those having
saccharide or hydrocarbon agly~ons.'~' A relatively important binding-
locus resides in the 2-hydroxyl group of the D-ga~actopyranosy~
configuration, as deduced by a comparison of methyl 2-acetamido-2-
deoxy-D-galactopyranoside, 2-deoxy-~-lyxo-hexose,and D-talose, all of
which inhibited poorly. Likewise, the importance of the 3-hydrogen
atom in stabilizing the lectin-saccharide complex was suggested by
comparing D-galactose with D-gulose, and p-nitrophenyl P-D-
galactopyranoside with its 3-deoxy-3-fluoro d e r i v a t i ~ e . ' ~Examina-
'*~~~
tion of analogs of D-galactose altered at C-5 (D-fUCOSe, L-arabinose,
methyl 6-0-methyl-a-D-galactopyranoside,and 6-deoxy-6-fluoro-~-
galactose) led to the conclusion that the oxygen atom of the

(625a) M. Horisberger and J. Rosset, Arch. Microbiol., 112, 123-126 (1977).

(625b) S . Basu, J. R. Moskal, and D. A. Gardner, in "Ganglioside Function: Biochemical
and Pharmacological Implications," G. Porcellati, B. Ceccarelli, and G . Tetta-
manti, eds., Plenum, New York, 1976. PP.45-63.
(625c) T. M. Kanellakes and K. P. Mathews:j. Allergy Clin. lmmunol., 56,407-410
(1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 265

TABLEXV
Inhibition of Bandeiraea simplicifolia I Lectin-Galactomannan Precipitation'zss1gs

Concentration required
Sugar for 50% inhibition (mM)

Methyl a-D-galactopyranoside, melibiose, p-nitrophenyl

a-D-gdactopyranoside, 2-O-a-D-galactopyranosyl-D-g1LI-
cose, O - ~ - D - g ~ a C t O p y r ~ O S y l ~ y O - i n O S i t o l 0.62
p-Nitrophenyl P-D-galactopyranoside 0.84
Raffinose 1.20
D-Galactose 1.90
Methyl P-D-galactopyranoside 5.40
Methyl 6-O-methyl-a-~-galactopyranoside 5.50
Methyl 2-acetamido-2-deoxy-a-D-galactopyranoside 9.20
6-Chloro-6-deoxy-D-galactose 10.4
6-Deoxy-6-fluoro-D-galactose 10.4
DTalose 20.0

Inhibition at 25 mM
2-Deoxy-D-lyxo-hexose 40"
DFucose 18
4-Deoxy-~-xylo-hexose 17b
3-O-P-~-Galactopyranosyl-D-arabinose 15
Methyl a-wglucopyranoside 15
Lactose 14
L-Arabinose 8
4-Deoxy-4-fluoro-D-ga~actose 0'
Galactitol 1
D-LyXOSe 1
Methyl a-D-lyxofuranoside 0
Methyl a-D-xylofuranoside 0
Methyl a-D-mannopyranoside 0
DGulose 0
L-Fihamnose 0

" Inhibition at 20 mM. Inhibition at 4.2 mM. Inhibition at 10.4 mM.

5-(hydroxymethyl) group was most probably involved in hydrogen

bonding to the lectin. The important binding-loci for BS I-saccharide
interaction are italicized in formula 13.

OH
13
266 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Equilibrium-dialysis studies on the binding of methyl a-D-galacto-

pyranoside to BS I revealed the existence of one carbohydrate-binding
site per subunit for the tetrameric protein, withlgl an intrinsic associa-
tion constant of 8.6 x lo4M-' at 2", and 3.3 x lo4M-' at 20". These
values correspond to a free energy of binding, AGO' (pH 7.2), of -26
kJ.mole-' and -25.36 kJ.mole-' at 2" and 20", respectively. The sites
appeared to be homogeneous and n~ninteracting.'~'
The BS I that had been lZ5I-labeled by a diazonium coupling
technique626was employed in a study of lectin-receptor density on the
B erythro~yte.'~' Scatchard analysis of the binding data demonstrated a
variation in receptor density among type B individuals from 7.2 x lo4to
13.4 x lo4 sites per erythr~cyte.'~' Likewise, apparent association-
constants for lectin-cell interaction varied from 1.1 x lo' to 2.9 x lo7
M-1.
A specific adsorbent has been prepared by covalent coupling of BS I
to cyanogen bromide-activated Sepharose 4B, and its interaction with
model, carbohydrate-protein conjugates has been studied.627The
carbohydrate-binding specificity of the lectin was retained. The au-
thorssZ7further demonstrated the utility of the specific adsorbent by
effecting a single-step purification of the Cassia alata seed-
galactomannan.
Further investigation revealed that BS I consists of five isolectins:
these are tetrameric structures composed of two, unique subunits (A
and B) in various proportions.621*628 They are designated BS I(A4),BS
I(&B), BS I(AZ&),BS I(A&), and BS I(B4). All are glycoproteins, as
revealed by a fluorescent, glycoprotein reagent.s2gThe subunits are
indistinguishable on the basis of size and immunochemical reactivity,
Subunit A is devoid of methionine,
but differ in isoelectric point.621,s28
whereas B contains one methionyl residue per polypeptide chain. The
single cysteinyl residue of subunit B was titrated with 5,5'-dithiobis(2-
nitrobenzoic acid).621,628 On the other hand, the cysteinyl residue of
subunit A could only be titrated in 6 M guanidinium chloride.
Perhaps the most interesting difference between the subunits is
their carbohydrate-binding specificity.621*62sBS I(&)agglutinated type
A erythrocytes (specific titer, 170) but not type B cells, and precipi-
tated A substance, indicating an affinity for a-D-linked 2-acetamido-2-
deoxy-D-galactosyl residues. Cross-reactivity with a-D-galactosyl res-

(626) C. E. Hayes and I. J. Coldstein, And. Biochem., 67,580-584 (1975).

(627) T. T. Ross, C. E. Hayes, and I. J. Goldstein, Carbohydr. Res., 47,91-97 (1976).
(628) L. A. Murphy and I. J. Goldstein,J. Biol. Chem., 252,47394742 (1977).
(629) A. E. Eckhardt, C. E. Hayes, and I. J. Goldstein, And. Biochem., 73, 192-197
(1976).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 267

idues was evidenced by precipitation of BS I(&) with type B sub-

stance and guaran, but not larch arabinogalactan. Sugar inhibition of BS
I(A4)-guaran precipitation confirmed this conclusion; methyl
2-acetamido-2-deoxy-a-~-galactopyranoside was 20 times as inhibitory
as methyl a-D-galactopyranoside. In contrast to BS I(A4), BS I(AB,,B,) is
highly blood-group B specific (specific titer, 120); these forms did not
agglutinate A erythrocytes, precipitated only B substance and guaran,
and showed no reactivity towards A substance or larch arabinogalactan.
Moreover, methyl a-D-galactopyranoside inhibited BS I(AB3,B4)-
guaran precipitation, whereas methyl 2-acetamido-2-deoxy-a-~-
galactopyranoside was one-hundredth as potent.
In many respects, the finding of related isolectins (of BS I) closely
parallels structural studiess30 on Phaseolus vulgaris agglutinin (see
Section V11,l). It will be interesting to determine whether there is
sequence homology between BS I subunits A and B, like that between
the two P . vulgaris and furthermore, to discover a structural
basis for the observed differences in specificity.

4. MacZura pomgera syn. aurantica (Osage Orange)

(Osage orange; D-Gab, D-GalNAcp)
Osage-orange seeds (Maclura pornifera) contain a nonspecific,
blood-group h e m a g g l ~ t i n i n . " ~Ulevitch
.~~~ and coworkerss33reported
the isolation of Maclura pornifera lectin by affinity chromatography of
seed extracts on a column prepared by condensing 2-amino-2-deoxy-
D-galaCtOSe with an aminosuccinyl-Sepharose. The bound-protein frac-
tion, eluted with melibiose, gave a single peak when chromatographed
on Sephadex G-200, and a single band on disc-gel electrophoresis, and
appeared homogeneous in the analytical ultracentrifuge. A molecular
weight of 80,000 to 100,000 was determined by gel filtration. Amino
acid analysis revealed a high content of glycine and aspartic acid, with
relatively little cysteine, methionine, and histidine.
Bausch and Poretz purified M.pornifera agglutinin on insolubilized
polyleucyl, hog-gastric m ~ c i nThe immunochemically homogene-
. ~ ~
ous protein migrated as a single component at acidic or alkaline pH in
poly(acry1amide) gel electrophoresis, and focused as a single band

(630) J. B. Miller, R. Hsu, R. Heinrikson,andS.Yachnin,Proc.Natl.Acad.Sci. U.S.A.,72,

1388-1391 (1975).
(631) J. M. Jones, L. P. Cawley, and G. W. Teresa, Vox Sang., 12,211-214 (1967).
(632) G. I. Pardoe, G. W. G. Bird, G. Uhlenbruck, I. Sprenger, and M. Heggen, Z.
Immunitaetsforsch. Allerg. Klin. Zmmunol., 140,374-394 (1970).
(633) R. J. Ulevitch, J. M. Jones,and J. D. Feldman, Prep. Biochem., 4,273-281 (1974).
(634) J. N. Bausch and R. D. Poretz, Fed. Proc., 35, 1530 (1976).
268 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

having an isoelectric point PI 4.75.In sharp contrast to the results of

Ulevitch and Bausch and Poretz determined a molecular
weight of 43,000 b y gel filtration.634Subunits of molecular weight
11,000 were observed by gel filtration in dodecyl sodium sulfate or
guanidinium
Regrettably, the only studies on the sugar-binding characteristics of
the MacZura pornifera lectin were conducted by hemagglutination in-
hibition of crude seed-extracts rather than on a purified, protein
Consequently, disparate results were obtained.
Representative hemagglutination inhibition are presented in
Table XVI, notwithstanding our serious reservations concerning the

TABLEXVI
Hemagglutination Inhibition of Crude Macluru pornifera Lectin by
Simple Sugars83s

Concentration required
Sugar for 100% inhibition (mM)

Melibiose 4.0
Stachyose 5.0
2-Acetamido-2-deoxy-~-galactose 5.0
2-Deoxy-D-lyxo-hexose 10
D-Galactose 150
2-Amino-2-deoxy-~-galactose 300
Raffi nos e 370
L-Rhamnose 400
D-Ribose 400
Maltose 400

significance of studies conducted on crude seed-extracts. Melibiose

and stachyose, having terminal (nonreducing) a-D-galactopyranosyl
groups, were effective inhibitors, whereas raffinose, having the same
terminal (nonreducing) group, was a very poor inhibitor. Equivalent to
melibiose and stachyose, 2-acetamido-2-deoxy-~-galactosewas 60
times as inhibitory as the related 2-amino-2-deoxy-~-galactose,and
twice as effective as 2-deoxy-~-Zyxo-hexose.Noninhibitory sugars in-
cluded D-arabinose, L-arabinose, 2-acetamido-2-deoxy-~-glucose,
D-glucose, 2-amino-2-deoxy-~-glucose,D-fUCOSe, L-fucose, cellobiose,
lactose, D-lyxose, D-mannose, melezitose, L-sorbose, sucrose, a,a-
trehalose, D-XylOSe, and ~ - x y 1 o s e . These
~ ~ ~ findings - ~ ~be~ con-
~ ~ ~ ~ must
sidered preliminary, until they are confirmed with purified lectin.
(635) L. P. Cawley, J. M. Jones, and G. W. Teresa, Transfusion (PhiludeZphia), 7,343-
346 (1967).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 269

In disagreement with the data presented in Table XVI, Pardoe and

coworkers632reported that D-galactose is a better inhibitor than
melibiose, whereas 2-acetamido-2-deoxy-D-galactose does not inhibit
this lectin. Sprenger and coworkers636found that stachyose inhibits
rather poorly, its activity being of the same order of magnitude as that of
2-amino-2-deoxy-~-galactose. Chuba and coworkers637 stated that
methyl a-D-mannopyranoside inhibited hemagglutination, although
no data were presented; this result is difficult to reconcile with previ-
ous reports.
It may tentatively be concluded that the lectin binds sugars related to
D-galactose. No clear difference has been established between the
inhibitory potency of 2-acetamido-2-deoxy-~-galactoseand
D-galactose. Although lactose was found noninhibitory in two
s t ~ d i e s ,several
~ ~ ~ ,reports
~ ~ ~ of lectin interaction with terminal, non-
reducing, p-D-linked D-galactopyranosyl groups of glycopeptides
e ~ i s t . ~Thus,~ ~ Pardoe
* ~ ~ and
~ * coworkers632
~ ~ ~ found neuraminidase-
treated, horse-erythrocyte glycopeptide (having terminal p-D-
galactopyranosyl groups) the most potent hemagglutination-inhibitor
of an extensive series of polysaccharides and glycoproteins tested.
Similarly, Sprenger and coworkers636demonstrated inhibition by a
number of mucins having the same, nonreducing terminus. Finally,
Chuba and coworkers637and Ahmend638studied the interaction of the
lectin with antarctic-fish, antifreeze glycoproteins. The glycopeptides
are comprised of repeating glycotripeptide units; the disaccharide
side-chains of 2-acetamido-2-deoxy-3(or 4)-O-P-D-ga~actopyranosyl-D-
galactose interact strongly with Mucluru pornifera lectin. On the other
hand, melibiose and a-D-galactopyranosyl-substitutedglycoproteins
were also shown to inhibit h e m a g g l u t i n a t i ~ n . ~Results
~ ~ * ~ ~using
~,~~~
2-acetamido-2-deoxy-~-galactosyl-substitted glycoproteins are some-
what puzzling in view of the sugar inhibition studies (see Table XVI).
Nonreducing (terminal) 2-acetamido-2-deoxy-a-~-galactopyranosyl
groups occur in peptone A substance and in horse-erythrocyte
glycoprotein digested sequentially with neuraminidase and p-D-
galactosidase; both were good inhibitor^.^^,^^^ Globoside I (terminal
2-acetamido-2-deoxy-~-~-galactosyl groups) and p-aminophenyl 2-
acetamido-2-deoxy-p-~-galactoside, diazotized and coupled to oval-
bumin, were without inhibitory ~apacity."~
In summary, despite a considerable amount of investigation, the
~

(636) I. Sprenger, G. Uhlenbruck, and G. I. Pardoe, Haematologia, 4,373-378 (1970).

(637) J. V. Chuba, W. J. Kuhns, R. F. Nigrelli, J. R. Vandenheede, D. T. Osuga, and R. F.
Feeney, Nature, 242, 342-343 (1973).
(638) A. I. Ahmed, D. T. Osuga, and €4. F. Feeney,]. B i d . Chem., 248,8524-8527 (1973).
270 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

sugar-binding requirements of purified M . pomifera lectin have not yet

been delineated. Furthermore, there are few biophysical data concern-
ing the molecular weight of this lectin, and no data regarding its subunit
structure, metal requirements, or presence of a glycosyl moiety; it may
be hoped that further studies will soon become available.

5. Ricinus communis (Castor Bean)

(castor bean; P-D-Gab > cu-D-Galp)
The (toxic)castor bean has been associated with folk medicine since
antiquity; the Ricinus communis bean was first described in the sixth-
century Sanskrit work on medicine, Susruta A y u r u e d ~Dixons40
. ~ ~ and,
later, Stillmark603attributed the toxicity of these seeds to an extractable
protein. As seed extracts agglutinated erythrocytesYso3 it was assumed
that toxicity was the result of agglutination. It is now recognized that
two, chemically distinct, carbohydrate-binding proteins are present in
castor beans: a toxin and a h e m a g g l ~ t i n i n . Extensive
~ ~ ~ * ~ ~ bibliog-
~
raphies on the lectins of R. communis have been compiled by Olsnes
and PihP9 and Balint.643
Multiple purification-schemes have been applied to the separation of
the toxin and the hemagglutinin.144J46~147~150*194*641-s“1 Fractionation
using salt and ethanol precipitation led to c r y s t a l l i ~ a t i o nof~the
~~~~~~
toxin known as ricin or ricin D. The hemagglutinin was isolated, free
from toxic activity, by ion-exchange chromatography and gel filtra-
tion.642*s46-s4*
With the introduction of affinity chromatography on
Sepharose 4B, to which both proteins bind, purification of the two R.

(639)S. Olsnes and A. Pihl, in “Receptors and Recognition Series: The Specificity and
Action of Animal, Bacterial and Plant Toxins,” Chapman and Hall, London, 1976.
(640)T. Dixon, Aust. Med. Gaz., 6, 137-155 (1887).
(641)T.Takahashi, G. Funatsu, and M. Funatsu,]. Biochem. (Tokyo), 52,50-53 (1962).
(642)M. Ishiguro, T. Takahashi, G. Funatsu, K. Hayashi, and M. Funatsu,]. Biochem.
(Tokyo), 55,587-592 (1964).
(643)G. A. Balint, Toxicology, 2, 77-102 (1974).
(644)E.A. Kabat, M. Heidelberger, and A. E. Bezer,]. B i d . Chem., 168,629-639(1947).
(645)M. Kunitz and M. R. McDonald,]. Gen. Physiol., 32,25-31 (1948).
(646)E.Waldschmidt-Leitz and L. Keller, Hoppe-Seyler’s Z. Physiol. Chem., 350,503-
509 (1969).
(647)E.Waldschmidt-Leitzand L. Keller, Hoppe-Seyler’s Z. Physiol. Chem., 351,990-
994 (1970).
(648) L. G. Gurtler and H. J. Horstmann,Biochim. Biophys. Acta, 295,582-594 (1973).
(649)S. Olsnes and A. Pihl, Biochemistry, 12,3121-3125 (1973).
(650)M. Lhermitte, G.Lamblin, P. Degand, and P. Roussel, Biochimie, 57,1293-1299
(1975).
(651) S.Olsnes, K. Refsnes, T. B. Christensen, and A. Pihl, Biochim. Biophys. Acta, 405,
1-10 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 271

communis lectins was greatly simplified.144~'46~'47Jso~1g4~649 The

D-galactose eluate of Sepharose columns was further separated into
toxic and agglutinating components by gel filtration146*194,650 or ion-
exchange chromatography.147~1s0~649~650 Alternatively, Nicolson and co-
workers reported selective elution of ricin from Sepharose with
2-acetamido-2-deoxy-D-galactose.146 Subsequent D-galactose elution of
the hemagglutinin effected a single-step p u r i f i ~ a t i o n . ' ~ ~
Ricin, which has a molecular weight of 60,000, consists of two
nonidentical, disulfide-bonded, polypeptide chains.'46J50Jg4~64E~64g The
subunits A and B have14s~64E~651
respective molecular weights of -28,000
and 32,000. The amino acid composition of the isolated subunits has
been determined, and they were found to be remarkably similar.B4E~6s1
The B subunit has relatively more aspartic acid and cysteine, but rela-
tively less glutamic acid and phenylalanine than the A subunit. Both
subunits contain low proportions of basic amino acids and methionine.
No reduced cystinyl residues occur in the native toxin."' The A and B
subunits gave almost identical, tryptic peptide maps, but differed in
their N-terminal (isoleucine-A, alanine-B) and C-terminal amino acids
(serine-A, phenylalanine-B).642,64E In contrast, Lhermitte and coworkers
observed650 N-terminal glycine and serine for the intact toxin. Sequenc-
ing studies were conducted on ricin subunits by Li and coworkers.652
Circular dichroism studiesss3showed that the ricin A chain contained
0.3%of a-helix, and the B chain, -25%. Both subunits contain cova-
lently bound carbohydrate: the A chain, mannose, and the B chain,
mannose and 2-amin0-2-deoxyglucose.~~~ Ricin has been crystal-
lized,844*645*654*s55
and its crystal structure studied.656
Ricin is extremely toxic to eukaryotic cells.
The experiments of Olsnes and Pih1150*639*s49 and Pappenheimer and co-
worker~ demonstrated
~~~ that one of the two subunits binds to the cell
membrane, presumably by way of a carbohydrate structure, whereas
the second subunit inhibits protein synthesis by a catalytic mechanism
in a cell-free system. This suggests that toxicity may result from the

(652)S. S.-L. Li, C. H. Wei, J.-Y.Lin, andT.-C.Tung,Biochem. Biophys. Res. Commun.,

65,1191-1195(1975).
(653)M. Funatsu, G.Funatsu, M. Ishiguro, and K. Hara,Jpn.J. Med. Sci. Biol., 30-32
(1973).
(654)1.-Y. Lin, Y . 4 . Shaw, and T.-C. Tung, Toxicon, 9,97-101 (1971).
(655) M.Funatsu, G.Funatsu, S. Ischiguro, S. Nanno, and K. Hara, Proc.J p n . Acad., 47,
718-723 (1971).
(656)C. H. Wei, J . Biol. Chem., 248,3745-3747 (1973).
(657)J.-Y. Lin, K.Lin, C.-C. Chen., and T . C . Tung, Cancer Res., 31,921-924 (1971).
(658)A. M. Pappenheimer, S. Olsnes, and A. A. Harper,J. Immunol., 113, 835-841
(1974).
272 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

binding of the toxin to mammalian cells by way of its B subunit,

followed by ingestion of toxin (or A chain), and, finally, by A chain
inhibition of protein synthesis.639
R . communis agglutinin, like the toxin, is composed of two types of
subunits bound by disulfide bonds. In the presence of
2-mercaptoethanol and a detergent, the protein of molecular weight
120,000 breaks down into subunits; Olsnes and coworker^'^' reported
subunits of molecular weight 31,000 and 34,000; Nicolson and cowork-
e r ~ 29,500
, ~ ~and ~ 37,000; and Giirtler and Horstmann,M827,500,
30,000, and 33,000. The smaller subunit is reportedly an inhibitor of
protein synthesis.639The amino acid composition revealed a polypep-
tide rich in aspartic and glutamic acids, serine, threonine, leucine, and
isoleucine, with low proportions of cysteine, methionine, or
phenylalanine.M8*6so No reactive sulfhydryl groups were detected in the
presence or absence of urea.639Threonine and alanine were reported648
as theN-termini of the subunits; serine, lysine, and phenylalanine were
found in the C-terminal position.646On the other hand, Lhermitte and
reported N-terminal alanine and glycine. The agglutinin
contains about 12%(by weight) of carbohydrate (mannose, glucose, and
2-amin0-2-deoxyglucose)~~~ not required for its carbohydrate-binding
a ~ t i v i t y .The
~ ~ failure ~ EDTA to inhibit R. communis agglutinin
~ , ~ ~ of
suggested its lack of dependence on metal cations for its activity.I8The
lectin is relatively stable to freeze-thawing, alteration of pH, deter-
gents, and
Immunochemical studies, tryptic peptide mapping, and end-group
analysis suggested that the toxin and agglutinin may have one common
and one unique type of subunit e a ~ h . ~Thus, ~ a compari-
~ ~ ~ ~ ~
son of tryptic peptide maps gave146a ratio of identical to unique pep-
tides of 1:l. Furthermore, antisera to either the toxin or the agglutinin
cross-reacted with the other protein.146,147,6s8,660,662
Shimazaki and coworkers663analyzed the c.d. spectrum of both toxin
and agglutinin in the presence and absence of lactose. The spectra of
the two proteins were similar, but there were differences in band
-
strength between them. The toxin contains 15%of a-helix and 52%of
@pleated sheet; the agglutinin, 13%of a-helix and 51% of P-pleated

(659) A. Surolia, A. Ahmad, and B. K. Bachhawat,Biochin~Biophys. Acta, 371,491-500

(1974).
(660) M. Jacoby, Beitr. Chern. Physiol. Pathol., 1, 51-77 (1902).
(661) M. Jacoby, Beitr. Chern. Physiol. Pathol., 2, 535-544 (1902).
(662) S . Olsnes and E. Saltvedt,J. Imrnunol., 114, 1743-1748 (1975).
(663) K. Shimazaki, E. F. Walborg, Jr., G. Neri, and B. Jirgensons, Arch. Biochern.
Biophys., 169,731-736 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 273

sheet. Conformational transitions occurred in both spectra upon the

addition of lactose.
The hemagglutinin agglutinates human erythrocytes without regard
to type. The toxin, on the other hand, agglutinates cells only after
addition of antiricin antiserum, or at high concentrations, where dimer
formation 0 c c ~ r r e d . l ~ ~
The carbohydrate-binding specificity of purified agglutinin has been
studied by sugar inhibition of h e m a g g l u t i n a t i ~ n , ~ inhibition
~ ~ ~ * ~ J ~ ~of
lectin-hog A + H substance p r e ~ i p i t a t i o n ,and
' ~ ~ inhibition of lectin-
alfalfa galactomannan precipitation. lZ4The results obtained by sugar
inhibition studies in precipitating systems are in very close agreement.
Table XVII presents representative data from the work of van Wauwe

TABLEXVII
Sugar Inhibition of R. communis Hemagglutinin-Alfalfa Galactomannan
PrecipitationlZ4

Inhibitor required for

Sugar 50% inhibition (mM)

p-Nitrophenyl P-D-galactopyranoside 0.026

Lactose 0.05
Methyl P-D-galactopyranoside 0.16
p-Nitrophenyl 2-acetamido-2-deoxy-~-~-galactopyranoside 0.225
p-Nitrophenyl a-D-galactopyranoside 0.26
Methyl a-D-galactopyranoside 0.29
Melibiose 0.32
Raffinose 0.35
D-Galactose 0.39
D-Fucose 0.51
L-Arabinose 1.42

and c o w ~ r k e r s . ' The

~ ~ following sugars were reported to be
noninhibitors of lectin-alfalfa galactomannan interaction: D-mannOSe,
D-ghOSe, L-fucose, and D - r i b o ~ e .Nicolson
'~~ and found
L-rhamnose comparable to methyl P-D-galactopyranoside as an in-
+
hibitor of lectin-hog A H blood-group substance precipitation, and
extended the list of noninhibiting sugars to include galactitol,
2-acetamido-2-deoxy-~-galactose,2-amino-2-deoxy-D-galactose,
D-lyxose, 2-acetamido-2-deoxy-~-glucose,and L-glucose. These results
are, with some exceptions, in accord with those of hemagglutination
inhibition studies. Olsnes and coworkers'47 found methyl a - ~ -
galactopyranoside to be a noninhibitor of lectin-human type B erythro-
cyte agglutination. Furthermore, Irimura and reported
274 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

that the phenyl and methyl glycosides, both a-and p-, of 2-acetamido-
2-deoxy-D-galactopyranose were noninhibitors of hemagglutination,
whereas 2-acetamido-2-deoxy-4-O-~-D-ga~actopyranosy~-D-g~ucose
(N-acetyllactosamine) was as effective as phenyl p-D-
galactopyranoside. Finally, Nicolson and B l a ~ s t e i n 'reported
~~ that
L-arabinose does not inhibit hemagglutination, whereas it does inhibit
hog A + H substance pre~ipitati0n.I~~ Hemagglutination studies are
complicated, in that the receptor structure(s) is not known, quantitation
is imprecise, and nonspecific interaction between lectin and cell can-
not be excluded. These considerations may, in part, explain the dis-
crepancies noted.
It may be concluded from these studies that sugars of the
D-galactopyranose configuration are bound most effectively by R .
communis agglutinin. The lectin exhibits some preference for p-D-
galactosides (lo),although the a anomer is also bound. The binding site
of the lectin may recognize more than a simple monosaccharide deter-
minant. Thus, lactose is three times as effective as methyl p-D-
galactopyranoside and, according to Irimura and coworkers,lS
N-acetyllactosamine, but neither phenyl nor methyl 2-acetamido-2-
deoxy-P-D-galactopyranoside,inhibits hemagglutination.
Equilibrium-dialysis experiments indicated that R. communis
agglutinin binds two moles of lactose per mole of protein, with an
association constant of 1.5 x lo4M - l , whereas ricin binds one mole of
lactose per mole of protein with an identical association ~0nstant.I~'
Two binding sites for p-nitrophenyl P-D-galactopyranoside,association
constant 1.65 x lo4M-I, were observed by van Wauwe and cowork-
erstZ4on studying the agglutinin. The equilibrium-dialysis experiments
of Podder and coworkers193differed slightly: an association constant of
2.2 x lo3 M-' was reported for the two lactose-binding sites of the
hemagglutinin.
The carbohydrate-binding specificity of the toxin is very similar to
that of the agglutinin, with one important difference; the toxin was
inhibited by 2-acetamido-2-deoxy-D-galactose, whereas the agglutinin
was The order of decreasing inhibitory capacity by sugars in
+
the toxin-hog A H substance precipitating system is lactose > methyl
P-D-galactopyranoside, methyl a-D-galactopyranoside > D-galactose,
melibiose > raffinose, D-fUCOSe, 2-acetamido-2-deoxy-~-galactose >
L-rhamnose, ~ - a r a b i n o s e . ' ~ ~
The interaction of R. communis agglutinin with polysaccharides has
been i n v e ~ t i g a t e d . ' The
~ ~ *lectin
~ ~ ~ precipitates galactomannans of al-
(664) J. P. Van Wauwe, F. G . Loontiens, and C. K. de Bruyne, Biochim. Biophys. Actu,
354, 148-151 (1974).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 275

falfa, fenugreek, and guar in a very similar way (compare Ref. 659). The
reaction between the agglutinin and locust-bean galactomannan or
larch arabinogalactan, as compared to that of the galactomannans, is
considerably weaker, reflecting a lower ga1actose:mannose ratio in the
former polysaccharide, and the presence of nonreducing (terminal)
L-arabinosyl groups (which bind poorly) in the latter. The linear, par-
tially sulfated carrageenans did not precipitate with the lectin, nor did
beef-lung galactan. A series of amyloids (tamarind, balsamine, and
capucine) gave precipitin reactions proportional to their galac-
tose:glucose:xylose ratios; increased substitution of the main chain
with reactive 2~-~-D-galactopyranosyl-a-D-xylosy~ side-chains was
correlated with more-complete precipitation of the lectin at lower con-
centrations of the polysaccharide. Prior treatment of capucine amyloid
with P-D-galactosidase resulted in the destruction of precipitating
capacity, confirming the presence of nonreducing (terminal) p-I>-
galactopyranosyl groups.
Codington and studied the interaction of fragments
obtained from epiglycanin, the major membrane-glycoprotein of a
murine, mammary carcinoma, with a series of lectins, including
Ricinus communis. The fragments only became inhibitors of
hemagglutination after neuraminidase treatment. This finding together
with chemical analysis suggested the presence of penultimate p-D-
galactopyranosyl residues.
A comparative study of D-galactose-binding lectins was made by
Irimura and who reported that porcine thyroglobulin
glycopeptide B, both untreated and neuraminidase-digested, was a
good inhibitor of hemagglutination, whereas porcine and bovine
submaxillary-mucin glycopeptides were not inhibitory, even after re-
moval of sialic acid, 2-acetamido-2-deoxy-~-galactose,and L-fucose
residues. The latter result is difficult to explain, in that removal of these
sugar residues would be expected to leave the core structure
pGal+aGalNac+ Ser/Thr, which would be complementary to the spec-
ificity of the lectin.
Adair and Kornfeld666studied the binding of '251-labelledR . com-
munis lectins to human erythrocytes, isolated a lectin-reactive, eryth-
rocyte glycoprotein by affinity chromatography on a lectin-Sepharose
column, and compared several glycoproteins, glycopeptides, and sim-
ple sugars in a standard inhibition-assay (see Table XVIII). The lectin-
reactive, erythrocyte glycoproteins were 1,200 times as inhibitory as

(665) J. F. Codington, K. B. Linsley, R. W. Jeanloz, T. Irimura, and T. Osawa, Carbohydr.

Res., 40, 171-182 (1975).
(666) W. L. Adair and S. Kornfeld,]. Biol. Chem., 249,4696-4704 (1974).
276 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXVIII
Inhibitory Activity of Various Carbohydrates, Glycopeptides, and Glycoproteins
in a Standard, r251-Lectin-Erythrocyte Ghost-binding Assayess

Concentration required
for 50% inhibition"
Rich R. communis
Substance (pM) Hemagglutinin ( p M )

Affinity-purified erythrocyte receptor for ricin 0.5 1.4

Affinity-purified erythrocyte receptor for R . communis
hemagglutinin 0.5 0.6
Erythrocyte sialoglycoprotein 18 15
Erythrocyte sialoglycoprotein digested with
neuraminidase 4.5 8
IgG glycopeptide 50 67
Fetuin glycopeptide 108 25
Transferrin glycopeptide 185 180
Lactose 108 100
D-Galactose 650 830
2-Acetamido-2-deoxy-~-galactose 280 3000

"Normalized to content of D-galactose.

simple sugars in this assay. Multivalent interaction and a possibly

extended, carbohydrate-binding site may account for this finding. Clear
differences in specificity between ricin and the agglutinin were ob-
served with fetuin glycopeptide, affinity-purified erythrocyte receptor
for ricin, and 2-acetamido-2-deoxy-~-galactose.
Extracts of R. communis seeds formed p r e c i p i t a t e ~ ' ~ ~
with
~''~~~~~
salivary mucins, ovarian-cyst, blood-group substances, and
pneumococcal polysaccharide XIV. Furthermore, the Sepharose-
coupled agglutinin reacted with each of fifty monoclonal, IgM im-
munoglobulins by way of a site on the Fc fragment,667and with IgG,
immunoglobulins.66sAs IgG3does not differ from IgG,, IgG2,and IgG,
in its carbohydrate content, this result suggests that IgG3 differs con-
formationally from the other immunoglobulins.668
Surolia and also coupled the R. communis lectin to
Sepharose, and demonstrated its utility by separating guaran from

(667)M.Harboe, E. Saltvedt, 0. Closs, and S. Olsnes, Scnnd. J. lmmunol., 4,Suppl. 2,

125-134 (1974).
(668)E.Saltvedt, M.Harboe, I. Folling, and S. Olsnes, Scand. J. Immunol., 4,287-294
(1975).
(669)A. Surolia, A. Ahmad, and B. K. Bachhawat, Biochim. Biophys. Acta, 404, 83-92
(1975).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 277

glycogen (the guaran was specifically bound, and eluted with 0.1 M
lactose). Desialated ceruloplasmin and fetuin were readily bound by
the immobilized agglutinin.66sSmall fractions of “native fetuin” and
“
native ceruloplasmin” were retarded, and, upon analysis, it was found
that the retarded material had a lower content of sialic acid compared to
the native g l y c o p r o t e i n ~It ~ ~ suggested that columns of R . com-
. ~was
rnunis lectin might be used as an analytical tool for the separation of
partially sialated glycoproteins.

VI. L-FUCOSE-BINDING
LECTINS
1. Anguilla anguilla (Eel Serum)
(eel; CX-L-FU~~,
3-0-Me-~-Fucp,3-0-Me-o-Gab)
An extremely potent source of natural anti-blood-group H ( 0 )activity
was discovered in the serum of the eel, Anguilla ~ n g u i l l a . ~ ’ .Eel
~~,~”~~~
anti-H(0) agglutinins, which some investigators consider to be primi-
tive antibodies, are found in approximately 50% of eels.21However,
important differences exist between eel-serum lectins and conven-
tional immunoglobulins; the antigenic stimulus (if any) responsible for
anti-H(0) activity in eel serum has yet to be identified.671Eel sera
agglutinate human 0 erythrocytes strongly, and A2 cells to some extent,
but do not react with types A, or B ce11s.21,64,65~670The agglutination of 0
cells by eel sera was readily neutralized by saliva from secretor indi-
v i d u a l ~ , ’ by
~ , ~soluble
~ blood-group H ( 0 ) substance, and by sugars
related to the type 0 oligosaccharide structure. Early
hemagglutination inhibition studies revealed that the carbohydrate-
binding site of the eel agglutinin was most complementary to methyl
a-~-fucopyranoside.~’*”~~~’*~~~ Specifically, studies showed that
L-fucose was strongly inhibitory, whereas D-fucose was not, that the
pyranose form was essential, and that an a-L-glycosidic linkage, in
contrast to the p-L-linkage, contributed positively to the binding of
L-fucose. On comparing difucosides with milk oligosaccharides, Wat-
kins and Morgan concluded4” that 2-O-a-~-linkedL-fucosyl residues
were bound more avidly than 3- or 4-O-a-~-fucosylresidues.
This fairly straightforward concept of the stereochemical features
defining the specificity of eel serum was, however, complicated by an
unexpected finding: extremely low concentrations (for example, 0.3
puglml) of an L-fucose-free polysaccharide from Taxus cuspidata com-

(670) S. Sugishita, juzenkwai-Zasshi, 40(5), 1938 (1935).

(671) G. F. Springer and P. R. Desai, Vox Sang., 18, 551-554 (1970).
(672) R. Kuhn and H. G. Osman, Hoppe-Seyler’s Z . Physiol. Chem., 303, 1-8 (1956).
278 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXIX
Eel Serum-0 Erythrocyte Agglutination and Eel Serum-H(O) Substance
Precipitation: Sugar Inhibition Studies167*674*676

Minimum amount Amount giving

completely inhibiting 50% inhibition
4 hemagglutinating of precipitation
Sugar doses (mg/ml) (pmoles)

L-Fucose and related sugars

L-Fucose 0.1-0.2 1.2-1.5
Methyl a-L-fucopyranoside 0.02 0.3-0.5
p-Aminophenyl a-L-fucopyranoside 0.02 0.25
Methyl P-L-fucopyranoside 0.3 13% at 2.0
2-0- Meth yl-~-fucose 0.05 0.6
Methyl 2-O-methyl-a-~-fucopyranoside 0.02 0.6
Methyl 2-O-methyl-P-~-fucopyranoside 0.15 30% at 2.0
3-O-Methyl-~-fucose 0.05-0.1 0.6
2,3-Di-O-methyl-~-fucose 0.05 0.4
Methyl 2,3-di-O-methyl-a-~-fucopyranoside 0.02 0.25
Methyl 2,3-di-O-methyI-P-~-fucopyranoside 0.3 2.7
2,3,4-Tri-O-methyl-~-fucose 5.0
2,3,5-Tri-O-methyl-~-fucose >5
2-Acetamido-2-deoxy-~-fucose 1.75

D-Fucose and related sugars

D-Fucose inactive inactive
2-O-Methyl-D-fucose 2.5
3-O-Methyl-D-fucose 0.05 0.5
Methyl 3-0-methyl-a-D-fucopyranoside 0.02-0.05 0.5
Methyl 3-0-methyl-/3-D-fucopyranoside 0.02 0.3
2,3-Di-O-methyl-D-fucose 0.05 0.5
Methyl 2,3-di-O-methyl-a-D-fucopyranoside 0.05 0.4
Methyl 2,3-di-O-methyl-~-~-fucopyranoside 0.15 0.7
2,3,4-Tri-O-methyl-D-fucose 1.2-2.5 1.2-2.5
2,3,5-Tri-0-methyl-D-fucose >5
Methyl a-Bfucopyranoside inactive inactive
Methyl P-D-fucopyranoside inactive inactive
Methyl 2-O-methyl-a-(or P-)D-fucopyranoside inactive inactive
Galactose and related sugars
D-Galactose inactive
L-Galactose inactive
2-O- Methyl-D-galactose 2.5-5
3-O- Methyl-D-galactose 0.1
2,3-Di-O-methyl-~galactose 0.1
3,4-Di-O-rnethyl-D-galactose >5
2,3,4-Tri-O-methyl-~-galactose 0.3
3-O-a-D-Galactopyranosyl-D-galacl
:ose >5
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 279

pletely inhibited the agglutination of O-erythrocytes by eel serum.'18

The serological activity of Taxus cuspidata twig polysaccharide was
shown to be due to its content of 2-O-methyl-~-fucose.~~~ As a conse-
quence of this observation, Springer and his associates systematically
s t ~ d i e d ' the ~ ~ interaction
,~~~ of D- and L-fucose derivatives with eel
serum (see Table XIX). Unexpectedly, 2- or 3-O-methylation conferred
inhibitory activity on the otherwise totally inactive, parent sugar,
D-fucose. High inhibitory activity was also exhibited by 2,3-di-0-
methyl-D-fucose. However, further methylation at 0-4 or 0-5 greatly
lessened the binding capacity. Interestingly, 2-0-, 3-0-, or 2,3-di-0-
methyl-L-fucose derivatives were slightly more inhibitory than the
free sugar. As in the D series, further substitution at 0-4or 0-5 dimin-
ished the lectin binding. Whereas 3-O-methyl-D- and -L-fucose dis-
played equivalent activity, the 2-O-methyl enantiomorphs differed
substantially. These findings were difficult to reconcile in terms of the
concept of stereospecificity in protein-ligand interaction. Based on his
observation of space-filling models, Kabat675suggested that rotation of
3-O-methyl-D-fucoseby 180"about its major axis would align this sugar
with methyl a-L-fucopyranoside in such a way that centers of electro-
negativity, hydrogen-bonding capability, and hydrophobicity would
become virtually superposable.
Another feature ofthe eel-serum specificity that is unlike that ofother
lectins is the existence of anomeric preference only within the L-hcose
' , ~ ~ ~a-L-hcopyranosides are two to five times as efficient
~ e r i e s . ' ~ Thus,
as inhibitors as their parent compounds, whereas their p-L-
glycosidically linked counterparts have half to one-third the effect. On
the other hand, the anomers of methyl 3-0-methyl-~-fucopyranoside
are of equivalent activity, being approximately twice as active as the
parent saccharide. In the D series, over 90% of the binding energy is
evidently accounted for by parts of the sugar molecule other than the
anomeric carbon atom and the aglycon.
A second, plant polysaccharide devoid of L-fucose, that from Sassa-
fras albidum, also displayed strong H ( 0 ) activity when tested with eel
serum (2.0 puglml gave complete inhibition of h e m a g g l ~ t i n a t i o n ~Of
~~).
its constituent sugars, 3-O-methyl-~-galactoseis apparently responsi-
ble for blood-group H ( 0 ) activity. An ensuing study of methylated
D-galactoses revealed that 3-0-methyl-D-galactose and 2,3-di-0-
(673) G . F. Springer, N. Ansell, and H. W. Ruelius, Naturwissenschuften, 43,256-257
(1956).
(674) G. F. Springer, P. R. Desai, and B. J. Kolecki, Biochemistry, 3,1076-1085 (1964).
(675) E. A. Kabat, Biochem. I., 85,291-293 (1962).
(676) G. F. Springer, T. Takahashi, P. R. Desai and B. J. Kolecki, Biochemistry, 4,
2099-2113 (1965).
280 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

methyl-D-galactose were as active as L-fucose, although D-galactose

itself was n ~ n i n h i b i t o r y . ~ ? ~
Springer and reported an extensive list of saccharides
that were weakly inhibiting or noninhibiting. These included the
2-acetamido-2-deoxy derivatives of D-galactose, D-glUCOSe, D-ribose,
D-talose, and D-arabinose, 3,6-dideoxy-~-and -D-xylo-hexose, L-lyxo-
hexosulose, L-arabinose, D-fructose, and 2-deoxy-D-erythro-pentose.
Three cardiac glycosides, namely, panstrosid, strospesid, and char-
treusin, were effective inhibitors of eel-serum agglutinin by virtue of
the content of p-D-linked 3-O-methyl-~-fucose(D-digitalo~e).'~~ A p-D-
glycoside of 2,3-di-O-methyl-~-fucose,namely, streblosid, also gave
good i n h i b i t i ~ n . ' ~ ~ , ~ ~ ~
Perhaps the most unusual aspect of eel serum-carbohydrate interac-
tion, discovered during the course of precipitin inhibition studies, was
the demonstration that 3-0-methyl-D-fucose could itself function as a
specific precipitinogen of eel-serum a g g l ~ t i n i n . ~Over ~ ' . ~90%
~ ~ of the
eel serum anti-H(0) activity was precipitated by addition of 4 to 8
pmoles of this sugar per ml of serum. This precipitin reaction, which
was also obtained with 3-O-methyl-D-galactose,was specifically inhib-
ited by L-fucose and other known inhibitor^.^^^,^^^ The precipitinogen,
D-digitalose, could be converted into an inhibitor of precipitation by
methylation at 0-1or 0-2, or reduction at C-2. Other common, anti-
H ( 0 ) reagents, including the lectins ofLotus tetragonolobus, were not
precipitated by D-digitalose. By assuming a lattice theory of precipita-
tion, Springer and Desai6??suggested the novel idea that the com-
plementary grouping for binding to eel-serum agglutinin might be
smaller than a monosaccharide. The minimal requirements for inhibit-
ing sugars (see Fig. 12A) were stated66to be "a methyl substituent

1 1
HO HO

(A) (B)
FIG. 1 2 . 4 A ) 3-O-Methyl-~-fucose,an Inhibitory Hapten for the Eel Agglutinin,
and (B) 3-O-Methyl-~fucose,a Precipitinogen of Eel Agglutinin. [The inhibitor
(A) is inverted by 180". Areas reactive with the agglutinin are marked by an arrow
(t).Note the similarity between the two molecules. (From Springer and De~ai.~'?)]

(677) G. F. Springer and P. R. Desai, Biochemistry, 10, 3749-3761 (1971).

 LECTINS : CARBOHYDRATE-BINDING PROTEINS 28 1

attached equatorially to a pyranose ring, an ether oxygen adjoining this

methyl group, and an axial, oxygen-carrying substituent cis to the
methyl group on a contiguous C atom.” Additional specific require-
ments for precipitinogens includess (see Fig. 12B) “three vicinal oxy-
gens protruding from a C1 pyranose ring. The oxygen at C-3 must carry
an apolar group, and the two oxygens flanking this group must be cap-
able of hydrogen bonding. One of these latter oxygens must be equa-
torial and trans to the oxygen at C-3 and the other axial and cis.”
Digitalose-precipitated, eel-serum agglutinin was quantitatively re-
covered by dialysis of the precipitate, thereby affording a novel, affinity
purification of the protein.66*671*677*678 The homogeneity of the isolated
agglutinin was assessed by disc gel-electrophoresis (pH 5 to lo),
moving-boundary electrophoresis (pH 8.6), zone electrophoresis on
paper and on cellulose acetate, immunodiffusion, and sedimentation in
the ultra~entrifuge.~~**~~~~~~~ From the biophysical constants measured
~ 7.2 S, DZ”o,w
( ~ 2 0 0 ,= = 5.0 X lo7 cm2.sec-’, ij = 0.705), Springer and
coworkers calculated a molecular weight of 123,000 for this almost
spherical, globular protein.679Dissociation by detergent, or succinyla-
tion, gave material having a molecular weight of 40,000, and treatment
with 2-mercaptoethanol yielded subunits of molecular weight 10,000.
From these data, the authors67gproposed, for eel-serum agglutinin, a
tertiary structure in which three subunits, each composed of four
disulfide-bonded, polypeptide chains of molecular weight 10,000, are
associated noncovalently.
Chemical analysis of the protein revealed substantial proportions of
aspartic and glutamic acid, alanine, glycine, serine, and threonine, with
small proportions of methionine, tryptophan, and phenylalanine. Very
little carbohydrate was found: 0.39% (by weight) comprised of
2-amin0-2-deoxyglucose.~~~ End-group analysis demonstrated equimo-
lar amounts of N-terminal serine and alanine, and C-terminal serine
and g l y ~ i n eThe. ~ ~molar
~ ratios77of monosaccharide precipitinogen to
protein at equivalence was 5.73: 1.
A comparative, c.d.-spectral analysis of individual, eel-serum anti-
H ( 0 ) specific proteins showed virtually identical patterns.680A strong,
positive band centered at 197 nm suggested the /3-conformation. Fur-
thermore, a negative band at 213 nm, weak positive bands at 235 nm,

(678) G. F. Springer, P. R. Desai, and J. C. Adye, Ann. N.Y. Acad. Sci., 234, 312-
331 (1974).
(679) A. Bezkorovainy, G. F. Springer, and P. R. Desai, Biochemistry, 10, 3761-
3764 (1971).
(680) B. Jirgensons, G. F. Springer, and P. R. Desai, C o m p . Biochem. Physiol., 34,
721-725 (1970).
282 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

-290 nm, and 295 nm, and a strong, positive band at 270 nm gave no
indication of a-helical structure. The c.d. spectra were markedly dif-
ferent from those of human 7 s immunoglobulin.680
An alternative purification of eel-serum agglutinin was reported by
Matsumoto and 0 ~ a w a . L-Fucose
l~~ was coupled to starch by use of
epichlorohydrin. Eel serum was chromatographed on the resulting
adsorbent, and bound protein was in two peaks with glycine
hydrochloride buffer, pH 3.0. Peak B (60% of the eluate) was
homogeneous by gel electrophoresis at pH 8.9, and by analytical ul-
tracentrifugation. In agreement with Springer and coworkers,679Mat-
sumoto and O ~ a w a reported
'~~ a sedimentation coefficient of 7.2 S.
However, they estimated a molecular weight of 140,000 by gel filtra-
tion, a value considerably higher than that reported earlier.679
Cell-binding studies conducted with [1251]-eel-serumagglutinin
1.7 x 106receptorsites per 0 erythrocyte, with K, = 2.7 x lo6
A4-l. Binding of eel-serum agglutinin to red cells of types A, and B gave
a biphasic curve, suggesting two types of receptor. Linear curves were
obtained in the presence either of human anti-A, or of anti-B serum.
Competitive inhibition in cell-binding studies on 0 erythrocytes was
observed between eel serum, Ulex europeus I, Ulex europeus 11, and
Cytisus sessil$olius lectins, suggesting that they interact with the
same, or closely associated, cell-surface structures.476
In addition to H ( 0 ) substances, Matsumoto and OsawalS6reported
hemagglutination inhibition by B and A substances and
neuraminidase-digested, porcine submaxillary-mucin. Although Lea
substance did not inhibit, a closely related milk oligosaccharide,
lacto-N-fucopentaose 11, did exhibit activity in this assay. On screening
22 invertebrate extracts, Baldo and coworkers found eel-serum-
reactive material in 15 species.681
A fucoxylomannan from the fruiting bodies of Fomes annosus (Fr.)
Cook was isolated by precipitation with purified, eel-serum aggluti-
nin.682

2. Lotus tetragonolobus (Asparagus Pea)

(asparagus pea; (Y-L-Fuc~,2-O-Me-~-Fucp)
The blood-group H ( 0 ) specific hemagglutinating activity of Lotus
tetragonolobus extracts was originally documented in the 1948 report
of Renkonen6, and this was confirmed by other.^.^^,^','^ In a landmark

(681)B. A. Baldo, G. Uhlenbruck, and G. Steinhausen, Vor Sang., 25,398-410 (1973).

(682)K.Axelsson, H.Bjomdal, S. Svensson and S. Hammarstrom, Acta Chem. Scand.,
25,3645-3650(1971).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 283

study, Morgan and Watkins22attempted to establish the nature of the

serological specificity of seed extracts by investigating the ability of
blood-group substances and simple sugars to inhibit hemagglutination.
Included in this investigation were extracts prepared from Lotus tet-
rugonolobus. In agreement with the results of Renkonen,6they22found
that the extract agglutinated type 0 cells considerably better than A2
cells, whereas types A1, B, and AB were not agglutinated. Lotus-0 cell
agglutination was inhibited by purified, human, or porcine, H(0)-
active blood-group substance, and by L-fucose, alone of the four con-
stituent monosaccharides of the ABO blood-group substances.22Fur-
thermore, methyl a-L-fucopyranoside was twice as active as the parent
sugar, whereas the /3 anomer and 2-deoxy-~-Zyxo-hexosewere half as
active. These early inhibition-studies were instrumental in establish-
ing that a-L-fucosyl residues are important, immunodominant struc-
tures of H-active blood-group substances. Kuhn and 0sma1-1~'~
confirmed, and extended, the work of Morgan and Watkins,22 and
suggested that serologically active fucosides must be ofthe L configura-
tion and in the a-1;-linked, pyranose form.
The hemagglutinin of Lotus seeds has been purified by three, dif-
ferent, affinity techniques based on its binding capacity for a - ~ -
fucopyranosides. Yariv and specifically precipitated the
lectin with the trifunctional fucosyl dye, 1,3,5-tri-(p-a-~-fucosyloxy-
phenylazo)-2,4,6-trihydroxybenzene. Following dissolution of the
precipitate in 2% L-fucose, an ion-exchange resin removed the dye,
leaving a protein preparation that appeared homogeneous by sedi-
mentation analysis. A molecular weight of -107,000 was determined
by sedimentation equilibrium.
Kalbs= subsequently separated the aforementioned protein prepara-
tion into three components by ion-exchange chromatography on
DEAE-cellulose. Sedimentation equilibrium established the follow-
ing molecular weights: component A, 120,000; component B, 58,000;
and component C, 117,000 (designated in order of salt elution from
DEAE-cellulose at pH 7.6).
A refined, affinity purification for L-fucose-binding Lotus proteins
was reported by Blumberg and ~ o w o r k e r s . N-(6-Amino-l-oxo-
~~~,~~~
hexy1)-/3-L-fucopyranosylamine was coupled to cyanogen bromide-
activated Sepharose 4B. The three proteins, A, B, and C, were bound by

(683) A. J. Kalb, Biochim. Biophys. Acta, 168, 532-536 (1968).

(684) M. Blumberg, J. Hildesheim, J. Yariv, and K. Wilson, Biochim. Biophys. Acta,
264,171-176 (1972).
(685) J. Yariv, A. J. Kalb, and M. Blumberg, Methods Enzymol., 28 (Part B), 356-360
(1972).
284 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

the adsorbent at 4", whereas only A and C were bound at 25". Proteins A
and C were further separated by ion-exchange chromatography. Each
of the three gave a single band in disc gel-electrophoresis.
Pereira and KabatZo0adsorbed the L-fucose-binding proteins on
+
polyleucyl hog A H substance. The dialyzed, L-fucose eluate gave a
single band against antiserum to the crude seed-extract, both by double
diffusion in agar and by immunoelectrophoresis. However, disc gel-
electrophoresis, at pH 9.3 and 2.7, revealed three components; these
were subsequently separated by preparative, isoelectric focusing.
Fraction I, isoelectric point pH 5.4, was identified as the component B
of Kalbsm;fraction 11, isoelectric point pH 6.2, as component C; and
fraction 111, isoelectric point pH 7.1, as component A.
Pereira and KabaPooinvestigated the subunit structure of the three
components by dodecyl sodium sulfate disc-gel electrophoresis. With
or without 2-mercaptoethanol, proteins A (111)and B (I) each exhibited
a single band, ofmolecular weight 27,800, whereas protein C (11)gave a
band of material having molecular weight 27,000. Equivalent fucose-
binding weights of 31,000,31,000, and 32,000 for A, B, and C, respec-
tively, were reported by Blumberg and coworkers.6s4KalbsS3deter-
mined equivalent binding-weights of 28,0OO,29,000, and 30,000 for A,
B, and C. It appears likely that A and C are tetrameric, whereas B is a
dimeric protein.
Chemical analysis of the three, related, L-fucose-binding proteins
revealed certain similarities, but sufficient differences to establish
their unique character. Thus, component A contains 8%of hexose and
1.4% of 2-amino-2-deoxyglucose (by weight); component B, 4% of
hexose and 0.8% of 2-amino-2-deoxyglucose; and component C, 8% of
hexose and 1.2% of 2-amino-2-deoxyglucose.683Each protein is devoid
of cysteine, extremely low in methionine content, and relatively rich in
aspartic acid (asparagine), threonine, and serine. However, they differ
noticeably in the relative content of the dibasic amino acids.683
The capacity of simple sugars to inhibit Lotus extract-0 erythrocyte
agglutination was an important observation in early studies concerning
the chemical basis of serological s p e ~ i f i ~ i t yThese . ~ ~initial
~ ~ ~re-
* ~ ~ ~ ~ ~
ports prompted Springer and c o w o r k e r ~ ' ~ to
' , ~synthesize,
~~ and test,
methyl ethers of fucose and their methyl glycosides, as well as an
extensive list of unrelated sugars, as inhibitors of Lotus extract-0 cell
agglutination. In accordance with earlier work, methyl a - ~ -
fucopyranoside elicited the greatest inhibition. The following sugars
were of approximately equivalent potency (half as active as methyl
a-L-fucopyranoside): p-aminophenyl a-L-fucopyranoside, L-fucose,
2-O-methyl-~-fucose, methyl 2-O-methyl-a-~-fucopyranoside, 3-0-
methyl-L-fucose, 2,3-di-O-methyl-~-fucose, and methyl 2,3-di-0-
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 285

methyl-a-L-fucopyranoside. The p-L-linked methyl glycosides of

L-fucose and 2-O-methyl-~-fucosehad one-tenth the effect of methyl
a-L-fucopyranoside. Fucoses methylated at 0-4, D-fUCOSe, S-deoxy-~-
xylo-hexose (colitose), and fucofuranosides were inactive. Although
these experiments were conducted on a crude, lectin preparation, they
illuminate important features of Lotus lectin-carbohydrate interaction,
namely, the common requirement of the three constituent components
for the pyranoid (as opposed to the furanoid) form, the 0-2 and 0-3
atoms, and the hydrogen atom of the 4-hydroxyl group of the L-fucose
configuration.
During the course of their work, Springer and W i l l i a r n ~ o nmade
'~~
the startling observation that, although D-fucose was totally inactive,
2-0-methyl-D-fucose possessed considerable inhibiting capacity.
Methyl glycosidation eliminated the binding of this D-fucose deriva-
tive. Kabat675explained this apparent anomaly introduced by 2- or
3-0-methylation on the basis of structural similarities between D- and
L-fucose derivatives as noted in three-dimensional models. Thus, a
similarity in profile and in regions of low hydrogen-bonding capability
is evident between %O-methyl-~-fucoseand L-fucose, if the latter is
rotated 180" about its horizontal axis. This phenomenon is even more
apparent in the binding of D-fucose derivatives to the L-fucose-binding
lectin of the eel167(see Section V1,l).
Pereira and KabatZo0undertook a careful study of Lotus lectin
carbohydrate-binding specificity employing blood-group-active sub-
stances, related oligosaccharides, and derivatives of L-fucose as in-
hibitors of precipitation of affinity-purified lectin with blood-group H
substance. The data (see Table XX) indicated that the sugar-binding
sites of components A, B, and C are essentially homogeneous with
respect to specificity, although they differ in affinity. The ratios of
relative inhibitory capacity for three L-fucose derivatives were constant
for the three proteins and the unfractionated mixture, thereby dem-
onstrating their complementarity. However, the published associa-
tion constants for methyl a-L-fucopyranoside binding are different: 1.2
x lo4 M - l , 0.6 x lo4M - * , and 3.7 x lo4M-' for A, B, and C, respec-
ti~ely.~~~
The immunochemical studies of Kabat and coworkers200,686 revealed
some striking subtleties in Lotus tetragonolobus sugar-specificity. Pre-
cipitin reactions between the lectin and several blood-group H sub-
stances demonstrated strong affinity of the lectin for this determinant.
H substances precipitated 83438% of the purified Lotus lectin at

(686) L. Rovis, E . A. Kabat, M. E.A. Pereira, and T. Feizi, Biochemistry, 12, 5355-
5360 (1973).
286 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXX
Sugar Inhibition2O0of the Precipitin Reaction Between Human H Substance
and Lotus tetragonolobus Lectins A, B, and C
~~ ~

Relative
inhibiting power
compared to that
Lectin Sugar Nanomoles for of methyl a-L-
fraction inhibitor 50% inhibition fucopyranoside

A (111) Methyl a-L-fucopyranoside 30 100

2'-O-~-Fucosyllactose 90 33
L-Fucose 110 27

B (1) Methyl a-L-fucopyranoside 13 100

2'-O-~-Fucosyllactose 31 39
L-Fucose 41 29
c (11) Methyl a-L-fucopyranoside 90 100
2'-O-~-Fucosyllactose 240 38
L-Fucose 330 27
Unfractionated Methyl a-L-fucopyranoside 50 100
2'-O-~-Fucosyllactose 150 33
L-Fucose 180 28

equivalence. Interestingly, precipitin studies of human blood-group A2

substances showed them to be -33% to 70%as active as H substances,
but human Al blood-group substances showed no reaction at all. Lea
substances precipitated the lectin, whereas B substances gave
relatively poor precipitation, and I-active precursor (as well as i-active)
substances did not react.
Of all oligosaccharides investigated by Pereira and Kabat,20° an
H-active difucosyl oligosaccharide, H JS RIMS2.5, exhibited the
strongest inhibition (see Table XXI). It was, in fact, superior to methyl
a-L-fucopyranoside and two monofucosyl oligosaccharides tested ( H JS
RL 0.75, 2-fucosyllactose). All of these oligosaccharides have a main
chain of 4-O-@-D-galactopyranosyl-D-g~ucoseor 2-acetamido-2-deoxy-
D-glUCOSY1 residues, which is representative of the type 2 chains
found in blood-group substances.ss7The L-fucosyl residues are substi-
tuted at the 2-hydroxyl group of galactose and, in difucosyl derivatives,
the 3-hydroxyl of glucose (or 2-acetamido-2-deoxy-~-glucose).How-
ever, if the fucosyl residue(s) is attached to a type l chain (~-O-@-D-
galactopyranosyl-D-glucose or 2-acetamido-2-deoxyglucose),the in-
(687) E. A. Kabat, Ado. Chem. Ser., 117,334-340 (1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 287

TABLEXXI
Inhibition of Human Blood-group H Substance-Lotus tetragonolobus Lectin
Precipitation by Sugars and Oligosaccharides200of Low Molecular Weight"

Nanomoles for
Sugar 50% inhibition

LX-L-FUC (Y-L-FUC
1 1
.1 .1
2 3
P-D-Gal-( 1+4)-P-~-GlcNAc-(1-+6)-R
(H JS RIMS2.5) 40

Methyl mL-hcopyranoside 50
Lacto-difucotetraose* 80
(Y-L-FUC
1
.1
2
P-D-GaI-(1+4)-P-~-GlcNAc-(1+6)-R
(H JS RL0.75) 105

2'-0-~-Fucosyllactose 150
L-Fucose 180
Lacto-N-fucopentaose 111 450
3-0-L-Fucosyllactose 575
Lacto-N-fucopentaose 11 700
Lacto-N-fucopentaose I inactive
Lacto-N-difucohexaose I inactive
3-O- Methyl-L-fucose inactive
Fucitol inactive
Urine-B oligosaccharide inactive
Urine-A oligosaccharide inactive
~~

"Data of Ref. 200, using affinity-purified, unfractionated lectin. bFor chemical struc-
tures of oligosaccharides, see Ref. 200.

hibiting capacity is abolished (for example, lacto-N-fucopentaose I,

lacto-N-difucohexaose I). Likewise, substitution of the type 2 chain
galactose at C-3 with 2-acetamido-2-deoxy a-D-galaCtOSe (A-active
urine oligosaccharide), or at 0 - 3 with a-D-galactose (B-active
urine oligosaccharide) negates the inhibitory activity of the oligo-
saccharide. The authorszo0concluded that Lotus lectin manifests a
high degree of specificity for unsubstituted, type 2 chains of blood-
group H-active substances. The structural constraints demonstrated for
Lotus lectin-oligosaccharide interaction were cited by Pereira and
KabaPoO in their re-examination of A, and Az blood-group specificity. As
288 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Lotus lectin does not react with A,, they suggested that all type 2 chains
of A, substance must be substituted by 2-acetamido-2-deoxy-a-~-
galactosyl groups, thereby blocking their reactivity. Similarly, reaction
of Lotus lectin with type A2 material indicated the existence of some
unsubstituted, type 2 chains on A2 substances. The chemical studies of
Kabat and coworkersasaestablished two types of type 2 determinant in
H substance. If a difference in 2-acetamido-2-deoxy-~-
galactosyltransferase substrate-specificity between A, and A2 individu-
als is postulated, the authors suggested that simultaneous reactivity of
A2 substances with anti-A antisera and Lotus lectin may be explained
by substitution of one type 2 chain-structure with the blood-group A
determinant sugar, whereas the other unsubstituted, type 2 chain-
structure would remain available to react with the lectin. A qualitative
difference between Al and A2 substances would then exist, together
with a quantitative difference in A-determinants. A fraction enriched in
type 2 chains was obtained by fractionating A2 substance (cyst 14) on a
Lotus-Sepharose column.526Fractionation of pooled, hog gastric-
mucin A + H blood-group glycoproteins on the same immobilized
Lotus column yielded fractions showing only A, only H, or AH activ-
ity.526
In studies of secretor and nonsecretor saliva antigens, Grundbacher
and C O W O ~ ~ observed ~ ~ S ~ a ~strong,
~ * ~but ~ diffuse,
~ precipitin band
between nonsecretor saliva (A, By or 0 donors) and the Lotus lectin,
whereas secretor saliva of all types formed a strong, diffuse band, as
well as a weaker band, by Ouchterlony gel-diffusion. The two bands
were poorly separated, complicating interpretation of the patterns. The
weaker band appeared identical to the antigen reactive with Ulex
europeus, and was presumed to be the H antigen. However, the
stronger band appeared, by Ouchterlony diffusion, to differ from the H,
Lea, and Leb antigens. The antigen described was not present in saliva,
or on red cells of the Oh (Bombay) type.
Inbar and coworkersas9determined that r3H]-acetylatedlectin did not
bind to the surface of normal cell-lines (hamster and rat embryo, and
mouse 3T3 cells) or to the transformed counterparts of these cells.
Trypsinization did not alter these results. The authors concludedssgthat
L-fucosyl residues were not exposed on the cell surface. In view of the
specificity studies of Kabat and coworkers,200*686 this interpretation may
be incorrect; L-fucosyl residues may be rendered sterically inaccessi-
ble by glycosylation of neighboring sugar residues.
(688) P. W. Napier, D. L. Everhart, and F. J. Gmndbacher, Vox Sang., 27, 447-458
(1974).
(689) M. Inbar, I. Vlodavsky, and L. Sachs, Biochim. Biophys. Acta, 255, 703-708
( 1972).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 289

3. Ulex europeus I (Gorse or Furze Seed)

(gorse or furze seed; C X - L - F U ~ ~ )
Since Cazal and Lalaurie's discoverys1of anti-H(0) hemagglutinat-
ing activity in extracts of three Ulex species (Ulex provincialis L., Ulex
jussiaei Webb, and Ulex europeus L.), U . europeus extract has become a
standard, serological reagent.14 It is used in typing 0 blood, in distin-
guishing Az from Al, and in the assessment of secretory status (the
occurrence of H-active substance in s a l i ~ a ) . ' ~ -0' ~cells are strongly
agglutinated, Az and AzB react more weakly, whereas A, and AIB are not
aggl~tinated.'~-'~~~~,~~*~~~ Secretor saliva, irrespective of type, inhibits
O-erythrocyte agglutination by U . europeus e ~ t r a c t . ' ~
There are two lectins of distinct, sugar-binding specificity in Ulex
extract^.'^^^^^^ Using ethanol fractionation, FloryZz5initially discrimi-
nated between Ulex I, which is readily inhibited by L-fucose deriva-
tives, and Ulex 11, which binds P-D-glucosides. (A comprehensive dis-
cussion of Ulex I1 purification and properties is presented in Section
111,5.)Flory's 50-70% ethanol precipitate enriched in Ulex I, was
slightly inhibited by D-arabinose, D-ribose, and D-lyxose, in addition to
~ - f u c o s eThis
. ~ ~ ~preparation agglutinated human, epithelial, cheek-
cells of secretory individuals
Matsumoto and Osawa196~208~226~509 isolated and studied both Ulex I
and I1 (see Section 111,5).Ulex europeus hemagglutinin I, which pre-
cipitated between 0 and 40% saturation with ammonium sulfate, was
purified by CM-cellulose chromatography followed by gel filtration.226
(The same authors were unable to adsorb U . europeus I to insoluble
L-fucose-substituted starch.149)Homogeneity of the preparation was
assessed by ultracentrifugation and gel filtration. Hemagglutination of
0 erythrocytes by hemagglutinin I was completely inhibited by
L-fucose and by secretor saliva. These authors149reported an szo,w value
of 6.5S,a molecular weight of 170,000,and an amino acid content rich
in aspartic acid and serine, but relatively poor in methionine and
cysteine. The protein had 3.8%(by weight) of neutral sugar and 1.4% of
a hexosamine.
By using an a-L-fucosyl-poly(acry1amide) adsorbent, with
L-fucose elution, Hoiejii and Kocourek obtainedag0a pure preparation
of U. europeus I, as judged by cellulose acetate electrophoresis and
sedimentation analysis. A major and a minor protein component were
observed on poly(acry1amide)disc gel-electrophoresis, pH 8.9. In con-
trast to the protein reported by Matsumoto and Osawa,226Hor'ejii and
K o ~ o u r e determined
k~~~ a molecular weight of 46,000by sedimentation

(690) V. HolejG and J. Kocourek, Biochim. Biophys. Acta, 336,329-337 (1974).

290 IRWIN J, GOLDSTEIN AND COLLEEN E. HAYES

equilibrium centrifugation, 40,000 by poly(acrylamide) gel elec-

trophoresis in dodecyl sodium sulfate, and 60,000-65,000 by gel filtra-
tion. No observable dissociation of the protein occurred upon reduc-
tion, or dissolution in 4 M urea. Although the amino acid analysis was in
good agreement with earlier results, the neutral sugar determined by
Ho?ejs’i and Kocourekago(7.2%) was twice that determined previ-
ously.226A single polypeptide chain was implicated by the finding of
only N-terminal serine. Interestingly, EDTA had no effect on the
hemagglutinating activity, although the lectin contained bound metals
(two calcium ions and one zinc or manganese ion per 46 kg of pro-
tein).690
A second, affinity-purification scheme was developed by Frost and
coworkers691; they synthesized the mixed 6-aminohexyl ~ , P - L -
fucopyranosides, and coupled these to CNBr-activated Sepharose 4B.
Elution, by L-fucose solution, ofthe resin-adsorbed lectin gave 13mgof
purified protein from 100 g of seeds. The sample was homogeneous by
gel electrophoresis at pH 4.5 and pH 8.1, and by immunodiffusion
against specific, rabbit antiserum.691As evidence of its biological pur-
ity, 0 erythrocyte agglutination was 97% inhibitable by 10 mM
L-fucose. In contrast to the results of Matsumoto and OsawazZaand
HoZejBi and K o c o ~ r e kFrost
, ~ ~ ~and coworkersss1reproducibly observed,
in dodecyl sodium sulfate gel-electrophoresis, two closely similar
bands that had molecular weights of 31,000 and 32,000. (U. europeus
lectin prepared by elution of formalized erythrocytes with L-fucose also
gave two bands in dodecyl sodium sulfate e l e c t r o p h ~ r e s i s . ~ ~ ~ )
Characterization of the carbohydrate-binding specificity of Ulex lec-
tin I is still very rudimentary. All reported studies applied the
hemagglutination inhibition technique, but used only a few of the
appropriate sugars necessary to the defining of binding specificity.
Make1a78found that a crude extract from Ulex seeds was inhibited by
L-fucose and salicin, whereas Kriipe7’did not include L-fucose among
the sugars that inhibited Ulex-0 erythrocyte agglutination.
Inhibition analysis of purified U . europeus I revealed a binding site
complementary to a - ~ - f u c o s i d e s .Ally1
~ ~ ~a-L-fucopyranoside
~~~~-~~~ and
2’-O-~-fucosyllactosegave strong inhibition of hemagglutination,
whereas L-fucose, lacto-N-fucopentaose I, and lacto-N-fucopentaose I1
were slightly less e f f e c t i ~ e . ’The
~ ~ *monosaccharides
~~~ L-arabinose,
D-ribose, D-galactose, D-glucose, D-mannose, 2-acetamido-2-deoxy-~-
glucose, and 2-acetamido-2-deoxy-~-galactose, and the disaccharides

(691) R. G. Frost, R. W. Reitherman, A. L. Miller, and J. S. O’Brien, A n d . Biochem.,

69, 170-179 (1975).
(692) T. Osawa and I. Matsumoto, Methods Enzymol., 28, Part B, 323-327 (1972).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 29 1

lactose, maltose, cellobiose, and gentiobiose were noninhibit-

In addition to 1mM L-fucose, Frost and coworkers noted6**
ory.1g63690-692
that 50 mM D-mannose and 10 mM methyl a-D-mannopyranoside gave
50% inhibition of 0 cell agglutination. We note a possible explanation
in the stereochemical relationship between L-fucose and D-mannose:
rotation of L-fucopyranose by 180"about its horizontal axis results in its
4-, 3-, and 2-hydroxyl groups becoming superposable on the 2-, 3-, and
4-hydroxyl groups of D-mannopyranose.
Complex, blood-group-active substances have been employed as
inhibitors of h e m a g g l u t i n a t i ~ n . 'Osawa
~~~~~ ~ Matsumoto found6s2
and
inhibition of both Ulex lectins by human A-, B-, and H-active sub-
stances, whereas Lea substance and porcine, submaxillary mucin
bound only to Ulex 11.Chuba and coworkers693reported that substances
from porcine stomach (A-active), equine stomach (B-active), and ba-
boon stomach (A-, B-, and O-active substances), as well as human A-, B-,
and O-active substances, bound equally well to both U . europeus lec-
tins. However, porcine, submaxillary mucin and nonsecretor saliva
inhibited only Ulex I1 activity.

VII. OTHERLECTINS
1. Phaseolus vulgaris (Red Kidney-bean)
Lymphocyte mitogenicity, an intriguing and distinctive biological
activity of several lectins, was first discovered in red kidney-bean
(Phaseolus vulgaris) extract^.^^^^^^ (For reviews of lectin-induced
mitogenesis, see Refs. 35-37a.) Biochemical studies on this
phytohemagglutinin (regrettably termed PHA, see Section I,4) have
been complicated by the large number of varieties of P . vulgaris inves-
tigated; red kidney-bean, black kidney-bean, wax bean, pinto bean, and
navy bean are all members of this genus and species. Further complex-
ity is introduced by the multifarious biological properties that various
species exhibit. They possess leuco- and erythro-agglutinating activity
and the ability to induce lymphocyte blast transformation with accom-
panying stimulation of DNA, RNA, and protein synthesis that culmi-
nates in cell division.
In the early 1 9 0 0 ' ~Landsteiner
~ and Raubitschek discovered the
hemagglutinating properties of kidney-bean Dorset and
HenleyI6 later used navy-bean extracts to separate cellular elements
(693) J. V. Chuba, W. J. Kuhns, J. D. Oppenheim, M . S. Nachbar, and R. F. Nigrelli,
Immunology, 29, 17-30 (1975).
(694) D. A. Hungerford, A. J. Donnelly, P. C. Nowell, and S. Beck, Am. J . Hum.
Genet., 11, 215-236 (1959).
292 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

from blood, in preparing an antiserum against hog cholera. Goddard

and MendelGg5partially purified the navy-bean hemagglutinin, and
characterized it as a protein that clumped the erythrocytes of chicken,
dog, duck, man, mouse, rabbit, and rat. The demonstration by Li and
Osgood’s that Phaseolus vulgaris and Phaseolus communis bean-
extracts could be used to separate leucocytes from erythrocytes pro-
moted isolation of the proteins in pure form. Rigas and OsgoodGg6 ob-
tained from P . vulgaris a purified “mucoprotein” which they freed of
polysaccharide by dissociation at pH 1.0. The protein was homogene-
ous in electrophoresis (pH 2.0 to 8.0),had PI 6.5, and contained 3.4% of
reducing substance.696 It agglutinated human erythrocytes irrespective
of the blood-group type.6g6
Further purification of the red kidney-bean lectin was conducted by
Rigas and Johnson by using ammonium sulfate fractionation, anion
exchange, and gel-permeation c h r o m a t ~ g r a p h y .The
~ ~ ~homogene-
,~~~
ous glycoprotein, molecular weight 128,000, contained alanine as the
only detectable N-terminal amino acid.697*6gs The lectin was rich in
aspartic acid, leucine, serine, and threonine, but low in sulfur-
containing amino acids .697*698 Prolonged incubation of the agglutinin in
8.0 M urea, followed by starch gel-electrophoresis in 8.0 M urea, gave
eight discernible components which were interpreted as eight unique
subunits. The hemagglutinating and mitogenic activities of the lectin
were associated with a single molecule, as shown by adsorption to
erythrocyte ~ t r o m a . ” ~Other
* ~ ~ ~investigators speculated that the
lymphocyte-stimulating and agglutinating activities of the kidney bean
might reside on separate r n ~ l e c u l e s . ~
Repeated
~ ~ - ~ ~adsorption
~ of par-
tially purified, Phaseolus vulgaris extracts with erythrocytes gave a
supernatant solution that exhibited mitogenic activity and was solely
leucoagglutinating; both activities were removed by leucocyte adsorp-

(695)V. R. Goddard and L. B. Mende1,J. Biol. Chem., 82,447-463 (1929).

(696)D . A. Rigas and E. E. Osgood,J. Biol. Chem., 212,607-615 (1955).
(697)D. A. Rigas and E. A. Johnson, Ann. N.Y. Acad. Sci., 113,800-818(1964).
(698)D.A. Rigas, E. A. Johnson, R. T. Jones, J. D. McDermed, and V. V. Tisdale,
in “Chromatographie et Methodes de Separation Immediate,” G. Parissakes,
ed., Association of Greek Chemists, Athens, Greece, 1966,Vol. 11, pp. 151-223.
(699)J . Bojeson, R. Bouveng, S . Gardell, b. Nordbn, and S. Thunell, Biochim.
Biophys. Acta, 82,158-161 (1964).
(700)T.Punnett and H. H. Punnett, Nature (London), 198,1173-1175(1963).
(701) P. Barkhan and A. Ballas, Nature (London), 200, 141-142 (1963).
(702)A. Rivera and G. W. Mueller, Nature (London), 212, 1207-1210 (1966).
(703)C. T.Mordman, A. de la Chapelle and R. Grasbeck, Acta Med. Scand. Suppl.,
412,49-58(1964).
(704)T. Weber, C. T. Nordman, and R. Grasbeck, Scand. J . Haematol., 4, 77-80
(1967).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 293

t i ~ n . ~By
Ousing
~ chromatography on SE-Sephadex C-50, Weber and his
colleagues704separated P . vulgaris preparations into two lymphocyte-
stimulating fractions: one that was solely leucoagglutinating and
lymphocyte-stimulating, and a second that was both strongly eryth-
roagglutinating and capable of agglutinating mixtures of leucocytes
and erythrocytes.
The discovery that erythroagglutinating activity was inactivated by
8.0 M urea at a moderately high rate, whereas mitogenic activity was
relatively stable, reinforced the suggestion that these two activities
might reside on different proteins (or Indeed, Rigas
and Head705resolved the P . vulgaris lectin into eight components by
poly(acry1amide) gel-electrophoresis in 8.0 M urea-borate buffer, pH
8.7. The eight bands (numbered in reverse order of migration towards
the anode) stained for both protein and carbohydrate. Band 2, and, to a
somewhat lesser extent, band 3, were the most potent erythroaggluti-
nins, displaying no detectable mitogenicity. Bands 5 and 6 exhibited
the highest mitogenic activity towards human peripheral lymphocytes,
with only marginal erythroagglutinating It was
hypothesized that different ratios of mitogenic to erythroagglutinating
activities might reflect combinations of a mitogenic and an eryth-
roagglutinating subunit in various ratios .697*705 Analysis of tryptic
glycopeptides from a highly purified, lectin sample suggested that at
least eight heteropolysaccharide chains of two or more different types
were p r e ~ e n t . ~ ~ ~ . ~ ~ ~
Commercially available phytohemagglutinin (PHA-P) prepared
from red kidney-beans (Phaseolus vulgaris) gave 17 different protein
bands when analyzed by poly(acry1amide) gel-electrophore~is.~~~ The
same sample, chromatographed on CM-Sephadex and Sephadex
G-150, yielded several, distinct, mitogenic proteins having differing
hemagglutinating capacity.708The most potent mitogen isolated (L-
PHA) was homogeneous by several criteria; it also had considerable
leucoagglutinating activity.?08A mixture of (at least two) closely related
proteins (H-PHA) possessing 250 times the hemagglutinating activity
of L-PHA was also isolated. Slightly less mitogenic than L-PHA, this
material possessed some leucoagglutinating activity.708The amino acid
and carbohydrate compositions of the two molecular species were
(705) D. A. Rigas and C. Head, Biochem. Biophys. Res. Commun., 34, 633-639
( 1969).
(706) D. A. Rigas, C. Head, and C. Eginitis-Rigas, Physiol. Chem. Phys., 4, 153-165
(1972).
(707) E. A. Johnson and D. A. Rigas, Physiol. Chem. Phys., 4,245-256 (1972).
(708) L. W. Allen, R. H. Svenson, and S. Yachnin, Proc. Natl. Acad. Sci. U.S.A.,63,
334-341 (1969).
294 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

similar, but H-PHA contained about twice as much carbohydrate as

L-PHA (both contained 2-amino-2-deoxyglucose and mannose, with
lower proportions of xylose and arabinose or fucose), and had a slightly
higher molecular weight, as determined by molecular sieving.708
Two distinct subunits in the P . vulgaris lectins were reported b y
Allan and C r u m p t ~ nand ,~~confirmed
~ by Oh and Conrad,710Weber and
coworkers,711and Yachnin and his c o l l e a g ~ e s . ~In~ ~fact, - ~ ~red
~
kidney-bean lectins comprise a family of five mitogenic glycopro-
t e i n ~ ~ each
' ~ ; contains four subunits held together by noncovalent
The individual isolectins contain various proportions of L
and R s u b ~ n i t s , "in~a manner reminiscent oflactic acid dehydrogenase
i s ~ z y m e sThe. ~ ~subunits
~ can be distinguished by N-terminal amino
acid sequence, isoelectric point, and biological properties.712
Leucoagglutinin (L-PHA) consists of four identical L-subunits, with
serine as the N-terminal amino acid and having PI 5.25; it exhibits
strong affinity for lymphocyte receptors, but little for those of red
b l o o d - ~ e l l sThe . ~ ~hemagglutinin
~~~~~ (H-PHA) contains four identical,
R subunits having an N-terminal amino acid sequence beginning with
alanine, and712 PI 5.95; this form exhibits a strong affinity for
erythrocyte-membrane receptors.714The erythroagglutinin (H-PHA),
which is probably the molecular species isolated by Rigas and cowork-
e r ~ ,was ~ reported711
~ ~ * ~to ~have ~ a somewhat higher molecular weight
(150,000)than the leucoagglutinin (L-PHA). The three intermediate,
mitogenic glycoproteins are tetramers that contain various proportions
of L and R subunits (LR,, LzRz,L3R).A schematic representation of the
five P . vulgaris isolectins is depicted in Fig. 13. The mixed,

FIG. 13.-Schematic Representation of the Tetrameric Structure of the Five Iso-

lectins from Phaseolus vulgaris. [Each form consists of various proportions of L
(N-terminal Ser) and R (N-terminal Ala) subunits. From Ref. 630. (Published by per-
mission of Proc. Natl. Acad. Sci. U.S.A.)]

(709) D. Allan and M. J. Crumpton, Biochem. Biophys. Res. Commun., 44, 1143-1148
( 1971).
(710) Y. H. Oh and R. A. Conard, Arch. Biochem. Biophys., 152,631-637 (1972).
(711) T. W. Weber, H. Aro, and C. T. Nordman, Biochim. Biophys. Acta, 263,94-105
(1972).
(712) J. B. Miller, C. Noyes, R. Heinrikson, H. S. Kingdon, and S. Yachnin,J. E x p .
Med., 138,939-951 (1973).
(713) S. Yachnin and R. H. Svenson, Immunology, 22, 871-883 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 295

erythrocyte-lymphocyte agglutination activity (mixed agglutination)

displayed by these isolectins reflects their hybrid structures. These
hybrid species also evoke a lymphocyte mitogenic response in propor-
tion to their content of L subunits. Interestingly, treatment of the
agglutinins with sodium metaperiodate affected only slightly their
agglutinating activity while abolishing their lymphocyte-stimulating
activity.?'
The L and R subunits have been isolated in homogeneous form by
isaelectric focusing in 8 M urea.630,712 They have identical molecular
weights, and both lack methionine and cysteine. The subunits differ in
amino acid sequence from residues 1to 7 (from theN-terminus), but are
identical in positions 8-24 and in their three C-terminal residues.630
The twelfth residue in each subunit is a glycosylated asparagine res-
idue; the carbohydrate composition of the R and L subunits is identi-
cal.630Thus, there are striking similarities between the subunits, and it
would appear that the differences in biological properties between the
two species are the result of relatively restricted differences in their
primary structure.630It is, indeed, interesting that a similar set of five
tetrameric isolectins composed of two different subunits (A and B) has
also been discovered628for the Bandeiraea simplifdia I lectin (Section
V73).
L e ~ c o a g g l u t i n i n ~ ~ ~ ~(L4)
~ ~ was
' , ~p~ ~ .r ~i ~f "i ~e ~don
~~ a~preparative
~,~~~
scale by fractional precipitation with ethanol, followed by ion-
exchange chromatography on DEAE-cellulose and SP-Sephadex, and
exclusion chromatography on Sephadex G-150. The crystalline glyco-
protein was homogeneous by numerous physicochemical and im-
munochemical riter ria.^^^,^'^ It is composed of four L-subunits, molecu-
lar weight 31,000 (compare 34,000 given in Ref. 712), and has an
aggregate molecular weight7I8of126,000( s & =~6.87 S). This glycopro-
tein lacks sulfur-containing amino acids, but contains high proportions
of aspartic acid, leucine, serine, threonine, and valine; it also contains
2-amino-2-deoxyglucose and mannose as the only carbohydrates718
(compare Ref. 629). In addition, leucoagglutinin also contains Mn2+and
Ca2+, which are essential both for lymphocyte-stimulating and
leucoagglutinating activities718(compare Ref. 151).
(714) S. Yachnin, L. W. Allen, J. M. Baron, and R. H. Svenson, Cell. Immunol., 3,
569-589 (1972).
(715) T. H. Weber, Scand. J . Clin. Lab. Inoest., 24, S u p p l . 111, 1-80 (1969).
(716) T. Weber and R. Grasbeck, Scand. /. Clin. Lab. Inoest., 21, S u p p l . 101, 14
( 1968).
(717) N. 0. Kaplan, Brookhawen Symp. Biol., 17, 131-153 (1964).
(718) V. Risanen, T. H. Weber, and P. Grisbeck, Eur. J . Biochem., 38, 193-200
(1973).
296 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Resolution of affinity-purified, red kidney-bean lectin by ion-

exchange chromatography afforded71Bafive distinct, homogeneous
glycoproteins, each of molecular weight 115,000 (&4,130).The tetra-
meric isolectins (termed E4, LIE3,L2E2,L3E1, and L4)differed in prop-
erties in the expected way: E4 was the most potent erythroagglutinin
(active at nanogram concentrations), L4was the most potent lympho-
cyte mitogen, and intermediate forms displayed both properties.71sa
Dissociation in vitro of the native isolectins in 6 M guanidinium chlo-
ride, followed by removal of dissociation agents, led to reconstitution
of the original, isolectin
The P . vulgaris lectins have now been isolated by affinity chromatog-
raphy on columns of Sepharose conjugated with porcine thyroglobu-
lin14BJs4 and fetuin,lS6both of which glycoproteins bind the lectin.
In light of the evidence, it is difficult to reconcile it with a report that
the RNA- and DNA-stimulating activity of Phaseolus vulgaris extracts
is not due to proteins.71B
A hemagglutinin from the wax bean (Phaseolus vulgaris; var. Sure
Crop Stringless Wax) was purified by Takahashi and coworkers720;its
amino acid distribution was similar to that of the red kidney-bean
lectins (no sulfur-containing amino acids, and high proportions of as-
partic acid, serine, and thre~nine).~*O The oligosaccharide chain(s) re-
portedly contained (in addition to mannose and 2-amino-2-
deoxyglucose) glucose, arabinose, galactose, fucose, and xylose. In
view of the difficulty in removing polysaccharides from the lectin, some
of the carbohydrate may result from an impurity. Supporting this
suggestion, the same investigators isolated, from the lectin, a glycopep-
tide of carbohydrate composition grossly different from that of the
original g l y c o p r ~ t e i nOn
. ~ ~the
~ other hand, Sela and coworkers722iso-
lated from the wax bean (Phaseolus uulgaris; var. Brittle-Wax) two
hemagglutinins that also contained arabinose, glucose, galactose, and
fucose, in addition to mannose. Moreover, the lectins were tetramers
(subunit molecular weight 30,000)of molecular weight 125,000+5,000.

(718a) R. D. Leavitt, R. L. Felsted, and N. R. Bachur,]. Biol. Chem., 252,2961-2966

(1977).
(718b) R. L. Felsted, M. J. Egorin, R. D. Leavitt, and N. R. Bachur,]. Biol. Chem., 252,
2967-2971 (1977).
(719) M. L. Goldberg, W. Rosenau, and G. C. Burke, Proc. Natl. Acad. Sci. U.S.A.,
64,283-289 (1969).
(720) T. Takahashi, P. Ramachandramurthy, and I. E. Liener, Biochim. Biophys. Acta,
133, 123-133 (1967).
(721) T. Takahashi and I. E. Liener, Biochim. Biophys. Acta, 154,560-564 (1968).
(722) B.-A. Sela, H. Lis, N. Sharon, and L. Sachs, Biochim. Biophys. Acta, 310,
273-277 (1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 297

Isolectins from the “haricot” kidney-bean (Phaseolus vulgaris) were

fra~tionated.~’~ They displayed723two types of glycoprotein subunit
(molecular weight 30,000 +1,000 and 35,000 +1,000). Although these
glycoproteins were shown to be agglutinins of red and white blood-
cells, they had negligible effects on lymphocyte transformati~n.’~~
Two lectins from Phaseolus vulgaris (Blue Lake) were isolated by
Dahlgren and his colleagues.724One of the lectins (PHA-a”)was pur-
ified to homogeneity. It had a molecular weight of 83,000, which is
considerably lower than that of other P . vulgaris lectins, although it was
supported by electron-microscope data.725PHA-a” exhibited eryth-
roagglutinating, leucoagglutinating, and mitogenic properties; at-
tempts to separate these activities failed.724
The lectin from Phaseolus vulgaris L. (var. red), isolated by conven-
tional protein-purification procedures, was shown to be localized in
the cytoplasm of the cotyledon and embryo of the A glycopro-
tein (6.6%carbohydrate), the lectin had molecular weight 128,000,
and contained high proportions of aspartic acid, serine, and tryptophan,
but no c y ~ t e i n e . ~ ~ ~ ~
Proteins from the black kidney-bean were fractionated.726Analysis
showed species that had both hemagglutinating and mitogenic proper-
ties; these agglutinins proved to be g l y c o p r o t e i n ~Rabbit
. ~ ~ ~ antiserum,
prepared against proteins from black kidney-beans, cross-reacted with
the water-soluble proteins from white and red kidney-beans (although
each variety of bean gave very different immunoelectrophoretic pat-
tern~).~~~,~~~
The lectins from Phaseolus vulgaris are not readily inhibited by
simple gar^,^^^^^^^,^^^.^^^ despite reports that 2-acetamido-2-deoxy-D-
galactose selectively inhibited both agglutinating and mitogenic ac-
tivities of a partially purified, P . vulgaris It is now
fairly well accepted that a complex, as yet ill-defined, saccharide struc-
ture is required for binding to the P . vulgaris lectin. To elucidate the
carbohydrate-binding specificity of the lectin, glycoproteins and
glycopeptides have been used. Specifically, lectin-reactive structures

(723) A. Pusztai and W. B. Watt, Biochim. Blophys. Acta, 365, 57-71 (1974).
(724) K. Dahlgren, J. Porath, and K. Lindahl-Kiessling, Arch. Biochem. Biophys., 137,
306-314 (1970).
(725) S. Hoglund and K. Dahlgren, Eur. J . Biochem., 17,23-26 (1970).
(726) W. G. Jaffi, and K. Hannig, Arch. Biochem. Biophys., 109,BO-91 (1965).
(727) W. G. Jaffi,, 0. Briicher, and A. Palozzo, Z. Immunitaetsforsch. Allerg. Klin.
Immunol., 142,439-447 (1972).
(728) H . Borberg, J. Woodruff, R. Hirschhorn, B. Gesner, P. Miescher, and R. Silber,
Science, 154, 1019-1020 (1966).
(729) H. Borberg, I. Yesner, B. Gesner, and R. Silber, Blood, 31, 747-757 (1968).
298 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

(assayed by inhibition of hemagglutination, or precipitate formation)

are subjected to sequential, chemical modification (for example, Smith
degradation), or treatment with purified, specific glycosidases, or both;
the degraded product is re-assayed. The validity of these studies de-
pends on careful monitoring of the chemically or enzymically released
sugars, and on a rigorous characterization of the product; unfortunately,
these imperatives have not always been appreciated.
Treatment of human erythrocytes with trypsin released a soluble
glycopeptide (molecular weight 10,000)that bound to purified P . vul-
garis lectin and thereby abolished its erythroagglutinating and
lymphocyte-stimulating properties.730Pronase-digested glycopeptide
was chromatographed on D E A E - c e l l u l ~ s e . Carbohydrate
~ ~ ~ * ~ ~ ~ and
amino acid analyses, followed by sequential cleavage ofsugars from the
nonreducing end(s) with specific glycosidases, to the
branched structure proposed in formula 14. The derived glycopeptide

a-AcNeu
2
1
6

t
334
p- D - GlcNAc p p- D -GlcNAc p
?t ?t
a ?
a - D- M a p - (1- a)-@-D-Manp - (l--?)-D-G1cNAcP-c Asn
14

was an excellent inhibitor of P . vulgaris lectin-induced cell-

agglutination. Neuraminidase cleavage of the sialic acid residue did
not affect the activity, whereas treatment of the desialized product with
P-D-galactosidase essentially abolished its ability to inhibit
h e m a g g l u t i n a t i ~ n ~ ' ~(see
~~~Table
~~~~ XXII).
~ - ~ ~Interestingly,
~ removal
ofthe single, free P-D-galactosylgroup from glycopeptide I resulted in a
loss of about 50% of its inhibitory potency.2'6,432*730-732 A second erythro-
cyte glycopeptide has been described in which both P-D-galactosyl
residues were penultimate to sialic acid.730Neuraminidase treatment

(730) S. Komfeld and R. Kornfeld, Proc. Natl. Acad. Sci. U.S.A., 63, 1439-1446 (1969).
(731) A. M. Leseney, R. Bourrillon, and S. Kornfeld, Arch. Biochem. Biophys., 153,
831-836 (1972).
(732) R. Kornfeld and S. Kornfeld, Ann. N.Y. Acad. Sci., 234,276-282 (1974).
(733) R. Kornfeld, W. T. Gregory, and S. A. Komfeld, Methods Enzymol., 28, Part B,
344-349 (1972).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 299

TABLEXXII
Inhibition of PhaseoEus oulgaris Lectin-Erythrocyte Agglutination by Simple
and Complex C a r b ~ b y d r a t e s ~ ' ~ * ' ~ ~

Nanomoles needed to
Compound give 1 inhibitory unir"

Erythrocyte trypsin fragmenP 0.8

Erythrocyte glycopeptide I" 0.3
Immunoglobulin G glycopeptide I b 0.35
Immunoglobulin G glycopeptide 11* 1.6
Fetuin glycopeptideb 5.0
Transferrin glycopeptide 15.0
Lacto-N-tetraose [P-D-Galp-(1+3)-P-~-GlcNAcp-(1+3)-p-~-
Galp-(1+4)-~-Gk] >500
Lacto-N-neotetraose p-D-Galp-( 1+4)-P-~-GlcNAcp-(1+3)-P-
D-Galp-(1-+4)-~-Glc] >500
N- Acetyllactosamine [P-D-Galp-(1+4)-~-GlcNAc] >500
Lactose >500
2-Acetamido-2-deoxy-~-galactose 22,000
2-Acetamido-2-deoxy-~-glucose >20,000
D-Galactose >20,000

"Amount of material necessary to inhibit erythrocyte agglutination completely in the

standard system for three minutes. *All contain the sequence A c N e u + p - G a l p +
P-D-GlcNAcp+D-Manp in their structure.

also did not affect its activity as an inhibitor.730Porcine erythrocyte

glycoproteins isolated by phenol-saline extraction gave precipitin
bands (immunodiffusion in agar gel) against commercial (Difco) P.
vulgaris preparation^.^^**^^^ Only neuraminic acid-containing fractions,
resolved by isoelectric focusing, inhibited P. vulgal-is-type 0 erythro-
cyte h e m a g g l ~ t i n a t i o n . ~ ~ " ~ ~ ~
A series of linear, model oligosaccharides having the terminal
sequence P-D-galaCtOpyranOSyl-(1-+3,4)-2-acetamido-2-deoxy-~-glu-
cose (see Table XXII) had less than 0.1%of the inhibitory capacity of
glycopeptide I, whereas D-galactose, lactose, 2-acetamido-2-deoxy-~-
glucose and 2-acetamido-2-deoxy-~-galactosewere n o n i n h i b i t ~ r y . ~ ~ ~ ~ ~ ~ '

(734) G. Uhlenbruck, U. Reifenberg, and R. Oyen, 2. Naturforsch., Teil B , 24, 147

(1969).
(735) G. Uhlenbruck, G. Wintzer, K. Schumacher, H. Oerkermann, G. I. Pardoe,
W. D. Hirschmann, G. Alzer, and R. Gross, in "The Role of Lymphocytes and
Macrophages in the Immunological Response," D. C. Dumonde, ed., Springer-
Verlag, Berlin, Heidelberg, New York, 1971, pp. 87-90.
(736) G . Uhlenbruck, G. Wintzer, B. Salfner, K. Schumacher, H. Oerkermann, W. D.
Hirschmann, and G. Alzer, K2in. Wochenschr., 48, 1369-1370 (1970).
300 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

On the other hand, two glycopeptides isolated from432 immunoglobulin

G, and glycopeptides derived from f e t ~ i and n ~ t~r a~n ~ f e r r i nwhich
,~~~
have mannose residues in their cores as well as GaldGlcNAc se-
quences in their outer chains, proved to be good inhibitors.732From
these studies, it would appear that a branched structure containing at
least two p-D-galactopyranosylgroups that are terminal or residues that
are penultimate to sialic acid are required for efficient binding to PHA.
The core region of the oligosaccharide chain is probably involved in
maintaining the proper spatial orientation. Multiple p-D-galaCtOSyl
groups were also implicated as P. vulgaris lectin-reactive sites by
Pardoe and her colleagues, who found that orosomucoid and erythro-
cyte glycopeptides from the hog, the horse, and the human were potent
inhibitors of P. vulgaris-induced h e m a g g l u t i n a t i ~ n . ~ ~ ~
A glycoprotein that reacted with P. vulgaris lectin is secreted in all
human saliva and ovarian-cyst fluid.739 Binding to P . vulgaris lectin was
abolished by proteolysis (with trypsin or pronase), or mild, alkaline
hydrolysis, but was unaffected by sequential digestion with
neuraminidase (all detectable sialic acid was liberated) and p-D-
galactosidase (less than 1%of the galactose was released).739
Osawa and his colleagues examined the binding of porcine-
thyroglobulin glycopeptides to P . vulgaris l e ~ t i n . ' ~Although
~ . ~ ' ~ bind-
ing of the lectin to thyroglobulin glycopeptide B can be rationalized on
the basis of the presence of three p-D-galactopyranosyl units (one ter-
minal group and two residues penultimate to sialic acid), binding to
glycopeptide A [(Man),+( GlcNAc),+Asn] is difficult to e ~ p l a i n . ' ~ ~ , ~ ' '
Sequential treatment of glycopeptide B with neuraminidase and p-D-
galactosidase diminished its inhibitory capacity to approximately one-
~ , ~ ' ~ activity could be explained by incomplete re-
t ~ e l f t h . ' ~Residual
lease of galactose. Additional studies on the isolation and inhibitory
activity of glycopeptides from human erythrocytes have been
made.740,741
Red kidney-bean extracts formed precipitates with several, normal,
animal sera, as well as with individual ~ e r u m - c o m p o n e n t s . ' ~ ~ * ~ ~ ~ ~ ~

(737) R. G . Spiro,J. Biol. Chem., 239,567-573 (1964).

(738) G. A. Jamieson, in "Protides of the Biological Fluids," H. Peefers, ed., American
Elsevier, New York, 1966, pp. 14 and 71.
(739) J. A. Strauchen, C. F. Moldow, and R. Silber, J. Zmmunol., 104, 766-768 (1970).
(740) Y. Akiyama and T. Osawa, Proc. J p n . Acad., 47, 104-109 (1971).
(741) Y. Akiyama and T. Osawa, Hoppe-Setller's Z. Physiol. Chem., 353, 323-331
(1972).
(742) L. Beckman, Nature, 195,582-583 (1962).
(743) W. H. Marshall and L. C. Norins, Aust. J. E x p . Biol. Med., 43, 213-228 (1965).
(744) N. H. Holland and P. Holland, Nature, 207, 1307-1308 (1963).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 301

The interaction of partially purified, commercial preparations of P .

vulgaris with a large number of serum glycoproteins was studied by
Morse.'osThe lectin precipitated a,-macroglobulin, @lipoproteins, and
immunoglobulin M. Preparations of a,-glycoprotein, orosomucoid, and
several immunoglobulin A myeloma proteins also reacted.lo5 Chon-
droitin 4-sulfate, dermatan sulfate, and heparin also precipitated a par-
tially purified preparation of red kidney-bean; the precipitation reac-
tion was completely inhibited by 0.5 M sodium chloride solution.366a
Yachnin demonstrated"O that it was the Phaseolus vulgaris
hemagglutinin (H-PHA), not leucoagglutinin (L-PHA), that formed a
precipitate in agar gel with both normal serum and the a-globulin
derived therefrom. He also showed'll that fetuin, which possesses
oligosaccharide units similar in structure to those of red blood-cells,
bound and precipitated H-PHA, but not L-PHA. Gel-filtration studies
revealed that H-PHA formed soluble complexes with fetuin, whereas
the affinity of L-PHA for fetuin was much lower, and any molecular
complexes that formed dissociated rapidly. Despite these differences,
fetuin is able to inhibit both H- and L-PHA-induced, lymphocyte trans-
formation."' Pusztai and Watt723similarly observed that fetuin would
inhibit hemagglutination of red cells by PHA. According to their
increasing content of R (hemagglutinating) subunits, affinity-purified
and resolved P . vulgaris isolectins exhibited an increasing capacity to
precipitate with 14 different animal sera.74s
Purified, carcinoembryonic antigen (CEA), which has sugar se-
quences similar to those in erythrocyte glycoprotein and fetuin, pre-
cipitated with purified l e u ~ o a g g l u t i n i n .Periodate
~~~ oxidation de-
stroyed the capacity of CEA to interact with l e u ~ o a g g l u t i n i n . ~ ~ ~
An early study implicated bound-sialic acid as a site of reaction with
wax-bean h e m a g g l ~ t i n i n . ~ ~ ~
In summary, although the carbohydrate-binding specificity of the P .
vulgaris lectins has still not been defined, it would appear that a
complex structure consisting of at least two fi-D-ga~actopyranosy~
groups that are terminal, or residues that are subterminal to sialic acid,
is required. The D-mannose residues in the core region of P . vulgaris
lectin-reactive glycoproteins appear to be involved in binding to the
In fact, it has been proposed that a trisaccharide
lectin.216*417-733*73s,747

(745) R. L. Felsted, R. D. Leavitt, and N. R. Bachur, Comp. Biochem. Physiol., B , 55,

417-421 (1976).
(746) R. L. Northrup and I. E. Liener, Proc. Soc. E x p . Biol. Med., 100, 105-108
(1959).
(747) S. Kornfeld and R. Kornfeld, in "Glycoproteins of Blood Cells and Plasma," G. A.
Jamieson and T. J. Greenwalt, eds., Lippincott, Philadelphia, 1971, pp. 50-67.
302 IRWIN J, GOLDSTEIN AND COLLEEN E. HAYES

unit [P-D-Galp-(1+3,4)-P-~-GlcNAcp-(1+2)-D-Man] is an essential

component of glycoproteins reactive with the red kidney-bean
l e ~ t i nKaifu
. ~ ~ and
~ Osawa have748synthesized the trisaccharide cited
[containing a P-D-( 1+4)-galactosidic linkage], and reported that it
inhibited 0-erythrocyte-P. vulgaris lectin hemagglutination. If con-
firmed, this will rank as an important discovery in lectin chemistry.

2. Vicia graminea
Historical interest in the Vicia graminea lectin has closely paralleled
interest in the chemical relationship between M and N blood-group
specificity. This blood-group N-specific hemagglutinin was discovered
by Ottensooser and S i l b e r s ~ h m i d twho
, ~ ~ ~noted a striking correspon-
dence between human anti-N sera (which is rare) and Vicia graminea
extracts. Although it agglutinated only NN and M N cells, the lectin
could be adsorbed with MM erythrocytes,750and it agglutinated751MM
cells stored in citrate buffers ofhigh pH. It had not yet been established
whether M and N substances were cross-reactive products of two al-
leles utilizing a common precursor substance, or were characterized by
a precursor-product relationship such that incomplete biosynthesis
would lead to the expression of both antigens.
Nagai and Springer isolated752blood-group M substance from eryth-
rocyte stroma, and demonstrated that this unreactive glycoprotein
became a potent inhibitor of Vicia graminea hemagglutination follow-
ing acid-catalyzed hydrolysis of sialic acid, a result confirmed by Lis-
k o w ~ k a Neuraminidase
.~~~ digestion likewise converted M-substance
into a product indistinguishable, by Vicia graminea reactivity, from
N - s ~ b s t a n c e , whereas
~ ~ ~ * ~ N-active
~~ substances were unaffected by
neuraminidase d i g e ~ t i o n ~and ~ ~mild
- ~ ~ hydrolysis
~ with a ~ i d . ' ~ ~ , ~ ~ '
Once sialic acid had been removed from M-substance, Vicia graminea

(748) R. Kaifu and T. Osawa, Carbohydr. Res., 52, 179-185 (1976).

(749) F. Ottensooser and K. Silberschmidt, Nature, 172,914 (1953).
(750) P. Levine, F. Ottensooser, M. J. Celano, and W. Pollitzer, Am.]. Phys. Anthropol.,
13,29-36 (1955).
(751) T. van Wageningen and L. E. Nijenhuis, Vox Sang., 5,572-573 (1960).
(752) Y. Nagai and G. F. Springer, Fed. Proc., 21,67(d) (1962).
(753) E. Lisowska, Nature, 198,865-866 (1963).
(754) M. Kriipe and G . Uhlenbruck, Z . Immun. Allergieforsch., 126,408-414 (1964).
(755) G. Uhlenbruck and M. Kriipe, Vox Sang., 10,326-332 (1965).
(756) G . Uhlenbruck and M. Kriipe, 2. Immunitaetsforsch. E x p . Ther., 124, 342-345
(1962).
(757) G. F. Springer and K. Hotta, Fed. Proc., 22,2261 Abs. (1963).
(758) G. F. Springer, Y. Nagai, and H. Tegtmeyer, Biochemistry, 5,3254-3272 (1966).
(759) E. Lisowska, Eur. J . Biochem., 10,574-579 (1969).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 303

reactivity could be abolished by treatment with D-galactose oxidase or

P-D-galactosidase,as was also true of native N-substance.219~678~754~7s5*7s8~76
Deacetylation of erythrocyte N-substance also resulted in the loss of
lectin rea~tivity."~ The oligosaccharide with which Vicia graminea
lectin reacts was labile to alkaline borohydride reduction-hydrolysis
after desialization, suggesting that it was attached by an 0-glycosyl
linkage to the polypeptide chain.7s92-Acetamido-2-deoxy-O-~-galac-
tosylgalactitol was released by this reaction. These studies suggested
that M-substance resulted from sialic acid substitution on the non-
reducing (terminal)P-D-galactosyl group of N-substance by the product
of the M-blood-group gene, a sialyltransfera~e.~~~~~~~
Vicia graminea lectin has been purified to homogeneity by Prigent
and Bourrillon.220An ammonium sulfate fraction of the extract of the
ground seed was chromatographed on DEAE-cellulose, and then proc-
essed by gel filtration on Sephadex G-150. The eluted protein (yield: 50
mg/kg of seeds) was homogeneous by sedimentation analysis, elec-
trophoresis, immuno-electrophoresis, and isoelectric focusing.
Sedimentation equilibrium studies gave a calculated molecular weight
of 105,000 ( s ; ~ ,=~ 5.3 S). Subunits of molecular weight 25,000were
obtained by electrophoresis in the presence of a detergent.
The chemical composition of Vicia graminea lectin was similar to
that of many other lectins, being high in its content of serine, and
aspartic and glutamic acids, and relatively deficient in cysteine and
methionine. In addition, three amino acids having hydrophobic side-
chains, namely, glycine, isoleucine, and leucine, were found in abun-
dance. The protein contained 7.3% of carbohydrate, principally man-
nose and 2-acetamido-2-deoxyglucose, along with lesser proportions of
fucose, galactose, and glucose.
Most investigators have reported that Vicia graminea extracts do not
bind any known mono- or oligo-saccharide, as assayed by hemaggluti-
Among those tested were sialic acid,
nation inhibition.21S.220*678*7s8~761,762
D-galactose, methyl a-D-galactopyranoside, methyl P-D-galactopyrano-
side, o-nitrophenyl P-D-galactopyranoside, 2-acetamido-2-deoxy-D-
galactose, lactose, melibiose, 3-O-a-D-galactopyranosy~-D-galactose,
6-0-P-D-ga~actopyranosyl-D-gdactose, 2-acetamido-2-deoxy-3-0-P-D-
galactopyranosyl-D-galactose (the disaccharide determinant of desia-
lyzed N-substance), 2-acetamido-2-deoxy-3-(and4-)O-P-D-galacto-
pyranosyl-D-glucose, lacto-l\r-tetraose, ~acto-l\r-neotetraose,D-glUCOSe,

(760) E. Romanowska, Vox Sang., 9,578-588 (1964).

(761) G . Uhlenbruck and W. Dahr. Vox Sang., 21,338-351 (1971).
(762) J. F. Codington, A. G. Cooper, M. C. Brown, and R. W. Jeanloz, Biochemistry,
14,855-859 (1975).
304 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

2-acetamido-2-deoxy-~-glucose,L-fucose, D-fUCOSe, and D-mannose.

On the other hand, Springer and his colleagues678noted inhibition of
Vicia graminea-induced agglutination of N N erythrocyte by 1acto-N-
tetraose, lacto-N-neotetraose, 2-acetamido-2-deoxy-3-(and4 - ) 0 - p - ~ -
galactopyranosyl-D-glucose, and D-galactose (in decreasing order of
potency). This discrepancy awaits resolution through continued
analysis. It is conceivable that Vicia graminea lectin has an extended
binding-site complementary to an oligosaccharide more complex than
those tested (compare Refs. 183, 216, 276, and 417). None of thirty
amino acids were inhibit01-y.~~~
The lectin reacts with bovine subma~illary-mucin,~~~ as well as with a
large glycoprotein released by proteolysis from the surface of TA3-
Ha cells of a mouse rnamrnary-~arcinorna.~~~

3. Miscellaneous Lectins
Numerous other lectins have been isolated; however, information
regarding their physical and chemical properties and carbohydrate-
binding specificity is incomplete. We present here, short descriptions
of several lectins that have interesting biological and carbohydrate-
binding properties. Although many of these substances are already
being used in a variety of studies, it is probable that more-complete
characterization will enhance their value considerably.
Extracts of Vicia cracca seeds were among the first plant-materials
shown to exhibit anti-A hemagglutination a ~ t i v i t y . ~ Gel
, ~ ~filtration
,~~,~~~
of a seed extract on Sephadex G-100 led to the discovery of two lectins
in V. cracca seeds.137The column eluate exhibited an increased, anti-A
~pecificity'~~; elution of the Sephadex column with 0.1 M D-glUCOSe
solution displaced a second, blood-group, nonspecific lectin with
carbohydrate-binding properties similar137to those of con A. Some
mitogenic activity was demonstrated in the latter, nonspecific 1e~tin.I~'
The anti-A lectin was isolated on an affinity adsorbent prepared by
conjugating ovarian-cyst A substance to agarose beads.764The anti-A
lectin was displaced with 0.1 M acetate buffer, pH 4.0. A substance
having blood-group A activity was isolated from hog gastric-mucin by
precipitation with V. cracca anti-A l e ~ t i nSubsequently,
.~~~ ovarian-cyst
A-substance was isolated on Vicia cracca lectin immobilized on
a g a r o ~ e Some
. ~ ~ ~ subfractionation of the A-substance was achieved.

(763) M. Kriipe, Z. lmmunitaetsforsch. E x p . Ther., 111, 22-31 (1954).

(764) T. Kristiansen, L. Sundberg, and J. Porath, Biochim. Biophys. Acta, 184, 93-98
(1969).
(765) T. Kristiansen and J. Porath, Biochirn. Biophys. Acta, 158, 351-357 (1968).
(766) T. Kristiansen, Biochim. Biophys. Acta, 388, 246-253 (1974).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 305

2-Acetamido-2-deoxy-~-galactose was the best inhibitor ofVicia cracca

agglutination of A, erythrocyte^^^,^^,^^; melibiose and raffinose were
weak inhibitor^.^^ (A, cells are more subject to agglutination than Az
~ e l l s . ~Interestingly,
~,~~) D-mannose, D-glucose, and maltose were the
best inhibitors of rabbit and pig cells, reflecting reaction of the second,
D-mannose-binding lectin with these animal cells.78Two blood-group
A lectins were isolated from Vicia cracca seed-extracts by affinity
chromatography on an adsorbent containing matrix-bound
2-acylamido-2-deoxy-~-galactose.~~~~ Continuous, pH-gradient elution
gave two fractions, each of which consisted of several agglutinating
species. Both lectin fractions had a molecular weight of 125,000and a
subunit weight of 33,000. D-Galactose was only 1% as potent as
2-acetamido-2-deoxy-~-galactose in inhibiting hemagglutination of A,
erythrocytes by the V. cracca l e c t i n ~ At ~ ~ 8,
. ~ pH ~ 2-acetamido-2-
deoxy-D-galactose binds to the lectins with K,,,, = 6 x lo3A 4 - l .
Extracts from the seeds of Laburnum alpinum were first shown to
possess anti-H(0) activity by Renkonen,6 and this was confirmed by
Morgan and Watkins22and other^.^^,^^, Although the L. alpinum lec-
tin(s) has not been purified, hemagglutination-inhibition studies con-
ducted on seed extracts reveal a specificity towards N,N'-
diacetylchitobiosyl residues.19*196~471 Human A, H, and neuraminidase-
treated human Lea blood-group substances were also extremely good
inhibitors of the Laburnum l e ~ t i n . ' ~ ~ ~ ~ ' ~
The hemagglutinin from Bauhinia purpurea alba seeds was purified
by specific adsorption on Sepharose 4B, and subsequent displacement
with 0.1 M The purified lectin, homogeneous in the ultracen-
trifuge and by electrophoresis on poly(acry1amide)gel had a molecular
weight of 195,000,and contained 11% of carbohydrate (principally
mannose and 2-amino-2-deoxyglucose, with smaller proportions of
xylose, glucose, galactose, and f u c o ~ e ) . The
' ~ ~ most notable feature of
the amino acid analysis was the absence of methionine, and the high
proportions of aspartic acid, serine, and t h r e ~ n i n e Carbohydrate-
.~~~
binding, specificity studies, performed by hemagglutination inhibition
analysis, revealed the lectin to be most reactive towards
2-acetamido-2-deoxy-~-galactose.~~'~~~~ D-Galactose, lactose,
melibiose, and N-acetyllactosamine were all -50% as a ~ t i v e . ~De- ~~,'~~
sialized, ovine submaxillary-mucin [which contains terminal (non-
reducing) 2-acetamido-2-deoxy-a-~-galactopyranosyl groups] was ex-
ceptionally The purified lectin agglutinated erythrocytes, in-
dependent of their ABO and MN blood-group types767;this nonspecific

(766a) H. Riidiger, Eur. J . Biochem., 72,317-322 (1977).

(767) T. Irimura and T. Osawa, Arch. Biochem. Biophys., 151, 475-482 (1972).
306 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

activity is in contrast to the anti-N reactivity described by others for

crude extracts of Bauhinia purpurea alba s e e d ~ . ~Perhaps,
~ , ~ ~ ' a second
lectin exhibiting anti-N specificity is present in B . purpurea extracts, or
anti-N specificity is present in only certain varieties of B . purpurea.
Ersson and isolated a D-galactose-binding lectin from
sunn-hemp seeds (Crotalariaj ~ n c e aby ) ~affinity
~ chromatography on
acid-treated, galactan gel. The lectin, eluted with 0.1 M lactose solu-
tion, had molecular weight 120,000 by gel filtration, contained 5% of
covalently bound carbohydrate, and was inhibited best by lactose'45
(compare Ref. 78). D-Galactose and 2-amino-2-deoxy-~-galactosehad
12%, and melibiose, -50%, of the activity of 1 a ~ t o s e .The l ~ ~ C. juncea
lectin was especially mentioned by Makelazofor its ability to distin-
guish between a-and /3-D-galactosides. C. juncea lectin, immobilized
by coupling to agarose-gel beads, was employed as an adsorbent for
serum proteins.7s8The C. juncea lectin was also purified on a matrix
prepared by coupling D-galactose to Sepharose 6B activated with di-
vinyl ~ u l f o n eHomogeneous
.~~~~ by several physical criteria, the lectin
was shown to be a tetrameric protein, subunit molecular weight
-31,000. It is a glycoprotein (9.8%of carbohydrate: mannose, 2-amino-
2-deoxyglucose, fucose, and xylose), contains bound metallic ions
( Ca2+and Mgz+),and has a high content of aspartic acid and serine, but
no methionine or ~ y s t i n e . ' ~ ~ ~
A sialic acid-binding lectin has been isolated from horse-shoe crab
(Limulus polyphemus) hemolymph by both c o n ~ e n t i o n a 1 and ~~~*~~~
affinity chromatographic (insolubilized, bovine subma~illary-mucin,'~~
and formalinized, horse erythrocyte^'^^) procedures. The purified lec-
tin (molecular weight, -400,000)1sg*769-771 is composed of 18 to 20, non-
covalently bound subunits (molecular weight, -20,000)15B~76s9-771 in the
form of a ring-shaped structure,772contains covalently bound carbohy-
drate (-4%, by weight, of 2-amino-2-deoxyglucose and neutral
s ~ g a r ) , 'and
~ ~ a. preponderance
~~~ of acidic amino acids, as well as high
proportions of glycine and leucine.17B~76s*770 Three moles of cysteine and
2 to 3 moles of methionine per mole of subunit were also reported to be
present in the Limulus a g g l ~ t i n i n . ~The ~ ~ ,sequences
~~~ of the

(768) B. Ersson and J. Porath, FEBS Lett., 48, 126-129 (1974).

(768a) B. Ersson, Biochim. Biophys. Acta, 494,51-60 (1977).
(769) J. J. Marchalonis and G . M. Edelman,J. MoZ. Biol., 32,453-465 (1968).
(770) A.-C. Roche and M. Monsigny, Biochim. Biophys. Acta, 371, 242-254 (1974).
(771) C. L. Finstad, R. A. Good, and G . W. Litman, Ann. N.Y. Acad. Sci., 234, 170-
182 (1974).
(772) H. Femindez-Moran, J. J. Marchelonis, and G. M. Edelman, J . MoZ. Biol.,
32,467-469 (1968).
(772a) R. Kaplan, S . S.-L. Li, and J. M. Kehoe, Biochemistry, 16,4297-4303 (1977).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 307

N-terminal-24 (Ref. 771) and -50 amino acids (Ref. 772a) have been
determined. The Limulus lectin agglutinates e r y t h r o ~ y t e s (. ~irre-
~~-~~~
spective of blood type), human l e u c ~ c y t e splatelets,776
,~~~ and tumor
Neuraminidase-treated cells could no longer be agglutinated
by Limulus serum.777Calcium ions are required for, and sometimes
enhance, the agglutination reaction,155J59,769*770J75 and they induce a
change in the c.d. spectrum of the l e ~ t i n . ~ N-Acetylneuraminic
~l
a ~ i d and ~ ~ D-glucur~nic'~~
~ J ~ ~ (but not D-galacturonic) acid specifically
inhibited agglutination of horse erythrocytes b y the LimuZus lectin.
2-Acetamido-2-deoxy-~-glucosewas also reported to be an inhibitor
by some investigator^,^^^,^^^ but not by others.159Human orosomucoid770
and carcinoembryonic antigen775inhibited agglutination b y the Lim-
ulus lectin. A precipitin reaction in immunodiffusion gels was ob-
served with purified, bovine submaxillary-mucin, but not against the
desialated m ~ c i n . These ' ~ ~ studies indicated that the Limulus poly-
phemus lectin possesses the capacity to react with biopolymers and
cells containing terminal sialic acid residues. As such, it should prove
to be an exceptionally usefbl reagent for identification and isolation
purposes. Labelled with ferritin and fluorescein, the lectin should be
capable of localizing neuraminic acid residues on cell surfaces. Lim-
ulin has also been shown to be mitogenic toward human, peripheral
lymphocytes .777a
An L-rhamnose-binding protein has been isolated from the culture
filtrate of Streptomyces 27S5 by Fujita and his c o l l e a g ~ e s ~they ~*~~~;
used conventional procedures, and affinity chromatography on insol-
ubilized gum arabic. (It was necessary to use M D-galaCtOSe in order to
displace the lectin.") The purified lectin, a protein of molecular weight
11,000, is d i s t i n g u i ~ h e dby
~ ~its high content of alanine, glycine, and
valine (corresponding to 47% of the total amino acid residues); low
content of carbohydrate (1.8%,corresponding to one residue ofhexose,
which could represent an impurity); and an unusual, c.d. spectrum
(large positive peak at 226 nm). The lectin yielded a typical

(773) H. Noguchi, Zentralbl. Bakteriol. Parasitenkd. Infektionskr. H y g . Abt. 1. Orig.,

34,286-288 (1903).
(774) E. Cohen, A. W. Rose, and F. C. Wider, Life Sci., 4,2009-2016 (1965).
(775) E. Cohen, M. Rozenberg, and E. J. Massaro, Ann. N.Y. Acad. Sci., 234, 28-33
( 1974).
(776) G. I. Pardoe, G. Uhlenbruck, and G . W. G. Bird, Immunology, 18, 73-83 (1970).
(777) E. Cohen, Truns. N.Y. Acad. Sci., 30,427-443 (1968).
(777a) A.-C. Roche, Y. Perrodon, B. Halpern, and M. Monsigny, Eur. /. lmmzcnol., 7 ,
263-267 (1977).
(778) Y. Fujita, K. Oishi, and K. Aida, Biochem. Biophys. Res. Commun., 53,495-501
(1973).
308 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

precipitation-curve with gum arabic, showed a high, rather specific,

titer against type B erythrocytes (256 times that for A or 0 cells), and
was strongly inhibited in its agglutination of B erythrocytes by blood-
group B substance from human saliva. Hemagglutination inhibition
analysis indicated a specificity for L-rhamnose. D-Galactose, methyl
a-D-galactopyranoside, melibiose, and L-arabinose had lo%, D-fUCOSe
2.5%, and phenyl a-and P-D-galactopyranoside -40% of the potency of
~ - r h a m n o s e . ~Galactitol,
~ , ~ ~ ~ D-ga~actono-1,4-~actone, 2-amino-2-
deoxy-D-galactose, and 2-acetamido-2-deoxy-~-galactose were essen-
tially n ~ n i n h i b i t o r s . ” *On
~ ~the
~ other hand, guar gum, locust-bean
gum, and gum arabic, all of which contain terminal (nonreducing)
a-D-galactopyranosyl groups, were very good inhibitors. Once it is
more completely characterized, this interesting lectin should prove a
useful structural probe for biopolymers containing L-rhamnose resi-
dues.
A nonspecific lectin was isolated from the meadow mushroom
(Agaricus campestris) by chromatography on DEAE-cellulose.si~221
The lectin, pure by ultracentrifugation, electrophoresis on cellulose
acetate, and immunoelectrophoresis, had molecular weight 64,000, and
contained 4% of ~ a r b o h y d r a t eA. ~four-chain
~ structure (AzB,) was pro-
posed for the hemagglutinin, based on gel-filtration studies in dis-
sociating solvents, and tryptic peptide-analysis.sl Four “buried” sulf-
hydryl groups were detectedSs1Erythrocytes from all of the major
blood-types gave equal titers against the A. compestris lectin; most of
the animal erythrocytes were also agglutinated. Of the many sugars
tested, none inhibited human, red blood-cell agglutination. However, a
sonic suspension of red-cell ghosts did produce inhibition of lectin-
induced hemagglutination.221
Presant and K ~ r n f e l disolated ~ ~ ~ two lectins (PHA-A and PHA-B)
from the commercial mushroom (Agaricus bisporus) by DEAE-
cellulose and 0-phosphonocellulose chromatography, and studied
their cell-binding properties. Glycopeptides from erythrocyte-
membranes, fetuin, and immunoglobulin A were potent inhibitors of
the hemagglutination reaction. A complex, carbohydrate-binding spec-
ificity similar to that for the Phaseolus vulgaris isolectins was indicated.
Isolation of a 2-acetamido-2-deoxy-~-galactose-binding protein from
a slime mold, Dictyostelium discoideum, was reported by Rosen and
Affinity chromatography on Sepharose 4B, followed by

(779) S. D. Rosen, J. A. Kafka, D. L. Simpson, and S. H. Barondes, Proc. Natl. Acad.

Sci. U.S.A., 70,2554-2557 (1973).
(780) D. L. Simpson, S. D. Rosen, and S . H. Barondes, Biochemistry, 13, 3487-3493
(1974).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 309

elution with D-galactose solution, gave the pure lectin, called dis-
A tetramer having molecular weight 100,000, the lectin
consists of four, apparently identical, subunits (molecular weight
25,000). A pure protein, discoidin is rich in aspartic acid, glutamic acid,
and 3-hydroxy amino acids, and it also contains a considerable propor-
tion of half-cysteine residues.780It is believed that discoidin may
mediate intercellular adhesion and aggregation during starvation of the
organism. The lectin is a powerful agglutinin of sheep erythro-
c y t e ~ . Carbohydrate-binding
~ ~ ~ * ~ ~ ~ specificity studies, monitored by
hemagglutination inhibition of formalinized sheep-erythrocytes, re-
vealed a primary specificity for 2-acetamido-2-deoxy-~-galactose.~~~
Lactose and D-galactose were approximately %th and l/icth as active.
Interestingly, 3-O-methyl-D-gluCOSe was equivalent to S-acetamido-
2-deoxy-~-galactosein its inhibitory potency (compare Ref. 213).
A similar, carbohydrate-binding protein was isolated from another
cellular slime-mold, PoZysphondyZium pallidurn, by adsorption to for-
malinized, human erythrocytes followed by elution with D-galactose
The purified lectin, subunit molecular weight 25,000, has a
primary specificity for D-galactose; lactose is 4 times as effective as an
inhibitor, and 2-acetamido-2-deoxy-D-g~ucosehas one quarter of the
potency O f D-galaCtOSe.78'This lectin is also believed to be involved in
cellular a g g r e g a t i ~ n . ~ ~ '
Potent, mitogenic activity was reported by Farnes and his col-
l e a g u e ~to~be
~ ~present in the extracts of pokeweed (pigeon berry;
PhytoZncca nmericnna). Extracts of whole (ripe and unripe) berries,
seed, pulp, and stem produced slight, and variable, erythrocyte
agglutination, and were mitogenic. High erythrocyte-agglutinating ac-
tivity and mitosis ofleucocytes were found with root and leaf extracts.771
Subsequently, Borjeson and coworker^'^^,^^^ reported the isolation from
pokeweed-root extracts of a homogeneous substance (a single band in
disc gel-electrophoresis, and a single line in immunoelectrophoresis)
that possessed three biological properties: hemagglutination,
leukoagglutination, and mitogenesis. Adsorption studies with red cells
or stroma removed the hemagglutinating activity, without altering the
mitogenic activity, whereas adsorption with leukocytes resulted in loss
of both the mitogenic and the leukoagglutinating activities (compare

(781) S. D . Rosen, D. L. Sinipson, J. E. Rose, and S. H. Barondes, Nature, 252, 128,

149-151 (1974).
(782) P. Farnes, B. E. Barker, L. E. Brownhill, and H . Fanger, Loncet, (2) 1100-1101
( 1964).
(783) L. N. Chessin, J . Borjeson, P. D. Welsh, S. D. Douglas, and H. L. Cooper,
J. E x p . Mcd., 124, 873-884 (1966).
310 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Ref. 703). Reisfeld and colleagues784 employed preparative,

acrylamide-gel electrophoresis to isolate pokeweed mitogen as a
homogeneous glycoprotein. It had molecular weight 32,000, an unusu-
ally high content of half-cysteine residues [but no free sulfhydryl
groups, as determined by titration with 5,5-dithiobis(2-nitrobenzoic
acid)], and contained 3.2% of carbohydrate (4residues of mannose, 1-3
each of fucose and glucose, and 2.8 of hexosamine).
W a ~ d a fractionated
1~~~ a saline extract of pokeweed roots on hy-
droxylapatite, followed by Sephadex-gel filtration. Five glycoproteins
were obtained, designated Pa-1 through Pa-5. Each glycoprotein had
distinctive, physical and chemical properties, and biological activities,
and each appeared in the plant at different times of the year.78sAll
components had molecular weights in the range of 19,000-31,000,
contained 1.8-12.5% of covalently bound carbohydrate, and 8 to 50
residues of half-cystine. Glycoprotein Pa-1 was the most potent
hemagglutinin (no significant difference in hemagglutination titer for
ABO cells was noted). All of these glycoproteins were mitogenic over
an unusually wide range of concentration of protein compared to other
mitogenic l e ~ t i n s .However,
~ ~ ~ , ~only
~ ~ Pa-1 was mitogenic for both the
B and the T classes of murine lymphocytes.786This difference in
mitogenic specificity is directly referable to the physicochemical prop-
erties of the glycoproteins. Only Pa-1 behaves as a multimeric protein
in gel filtration in nondissociating solvents. Furthermore, of the
mitogenic P . americana glycoproteins, Pa-1 has the fewest disulfide
bonds.785,788 Yokoyama and coworkers787performed a similar fractiona-
tion of P . americana extracts by employing DEAE-cellulose and
affinity chromatography on immobilized, desialized, human-
erythrocyte glycopeptides. Five mitogens (Pa-1 through Pa-5) were
isolated, and characterized. In agreement with the results of Waxdal
and B a ~ h a monly ~ was found to be mitogenic both for murine
, ~ ~Pa-1
B-cells and T-cells, whereas the other glycoproteins were T-cell mito-
gens.787Two phytomitogens were isolated from saline extracts of
Phytolacca esculenta (shoriku) roots by salting out with ammonium
sulfate and chromatographing on DEAE-cellulose and Sephadex
G-100 columns.788Fraction E-2 had properties (molecular weight
32,000; 18 residues of half-cystine; 5.3% of carbohydrate) similar to
(784) R. A. Reisfeld, J. Borjeson, L. N. Chessin, and P. A. Small, Jr., Proc. Natl. Acad.
Sci. U.S.A., 58, 2020-2027 (1967).
(785) M. J. Waxdal, Biochemistry, 13,3671-3677 (1974).
(786) M. J. Waxdal and T. Y. Basham, Nature, 251, 163-164 (1974).
(787) K. Yokoyama, 0. Yano, T. Terao, and T. Osawa, Biochim. Biophys. Acta, 427,
443-452 (1976).
(788) H. Tokuyama, Biochim. Biophys. Acta, 317,338-350 (1973).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 311

those of the pokeweed mi tog en^.^^^ The carbohydrate-binding spec-

ificity of the lectins from the genus Phytolacca is still unknown. This
system is of outstanding interest and importance to cellular im-
munologists; it could materially assist in unravelling the differential
requirements for B- and T-cell activation.
A phytohemagglutinin (termed Robin) from Robinia pseudoacacia
(black locust) seed-extracts was isolated by conventional techniques.l6O
The lectin had a molecular weight of -100,000 ( s ; ~ ,=~ 4.39 S), con-
tained 17% of covalently bound carbohydrate (mannose, 2-amino-2-
deoxyglucose, fucose, and xylose, and a trace of glucose), and was rich
in acidic amino acids, hydroxyl amino acids, and leucine, but lacked
cysteine.160The R. pseudoacacia lectin possesses a significant content
of @pleated, sheet conformation, but little, if any, a - h e l i ~ . "Treatment
~
of the glycoprotein with pronase and trypsin released a glycopeptide
which, after purification on Sephadex, was found to contain essentially
all of the activity of the h e m a g g l ~ t i n i n . ' ~The
~ , ' ~purified
~ glycopeptide
contained 65% of neutral sugar (mainly mannose) and 2-amino-2-
d e o x y g l u c ~ s e Fifteen
.~~ minutes after treatment with 0.05 M perio-
date,'89 half of the hemagglutinating activity was lost. Simple sugars do
not inhibit the R. pseudoacacia lectin, but a sialoglycoprotein isolated
from the urine of a pregnant woman was found to be a potent inhibitor
. ~ ~ ~R . pseudoacacia lectin has been
of lectin h e m a g g l u t i n a t i ~ nThe
used as a membrane probe.731
A D-mannose-Sepharose 6B column791was used to isolate a blood-
group nonspecific, D-mannose-binding lectin from Vicia ervilia
extract^.^^*^'^ Gel-filtration studies in dissociating solvents, and
ultracentrifugation analysis, led to the suggestion that the molecule of
the V . ervilia lectin is composed of four subunits, two of type A
(molecular weight 4,700) and two of type B (molecular weight 2l,OOO),
to give an aggregate molecular weight of 53,000,in agreement with the
amino acid composition and the ultracentrifuge data. The agglutination
reaction was inhibited by D-glucose, D-mannose, D-fructose, methyl a-
D-mannopyranoside, maltose, melezitose, and a,a-trehalose. This
carbohydrate-binding specificity places the V .ervilia lectin in the same
class as con A and the lectins from the lentil, the pea, and the broad bean
(V.faba).
Two lectins, one a potent m i t ~ g e n the , ~ ~other
~ displaying strong

(789) J. Font and R. Bourrillon, Biochim. Biophys. Acta, 243, 111-116 (1971).
(790) M. Lemonnier, Y. Goussault, and R. Bourrillon, Carbohydr. Res., 24, 323-331
(1972).
(791) N. Fornstedt and J. Porath, FEBS Lett., 57, 187-191 (1975).
(792) B. E. Barker and P. Farnes, Nature, 215,659-660 (1967).
312 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

hemagglutinating and leucoagglutinating activity, were separated from

seed extracts of Wistaria floribunda by SE-Sephadex chromatog-
raphy.793,794 The mitogen was a glycoprotein of molecular weight
-70,000. It contained 11.4% of carbohydrate (5.8% of mannose, 3.5% of
2-amino-2-deoxyglucose,and smaller proportions of arabinose, fucose,
and xylose), and had an amino acid composition rich in acidic and
hydroxylic amino acids and low in sulfur-containing amino
Both the purified, mitogenic lectin and the strongly hemagglutinating
fraction A (compare Ref. 794) were nonspecific with regard to blood-
group types ABO, and were strongly inhibited in their agglutination of
erythrocytes by 2-acetamido-2-deoxy-~-galactose(the hemagglutinin,
more (The mitogen was also reported to be inhibited
by N,N'-diacetylchit~biose.~~~) The hemagglutinin was inhibited
somewhat more strongly by lactose than by melibiose, whereas these
disaccharides were equally potent with respect to the m i t ~ g e n . ~ ~ ~
Mitogenic and cell-binding studies were reported for the mitogen and
the h e m a g g l ~ t i n i n . ~ The
~ ~ , hemagglutinin,
~~~*'~~ purified by adsorption
onto insolubilized, hog (A + H active) gastric-mucin, followed by
elution with D-galaCtOSe solution, appeared to be homogeneous by
isoelectric focusing (PI, 5.5), but actually consisted of a mixture of di-,
tetra-, and octa-meric forms of the l e ~ t i nRecycling
. ~ ~ ~ chromatography
on Sephadex G-200 showed the tetramer (mol. wt. 125,000) to consti-
tute 85% of the mixture. The lectin agglutinated human and murine
lymphocytes, but was nonmitogenic towards these cells.794Kurokawa
and his colleagues796also isolated a hemagglutinin from W. floribunda
seeds. A homogeneous glycoprotein of molecular weight 68,000, the
lectin contained 3.2% of carbohydrate (mannose, galactose, and
2-amino-2-deoxyglucose)in the molar ratios of 1:2:1.The lectin, shown
to contain two equivalent binding-sites for 2-acetamido-2-deoxy-~-
galactose (K, = 1.28 x lo4M-'),dissociated into two identical, mono-
valent subunits upon reduction with 2-mercaptoethanol, indicating the
presence of two subunits of mol. wt. 32,000 linked through a disulfide
bridges7% Hemagglutination-inhibition studies showed the W .
floribunda lectin to be complementary to 2-acetamido-2-deoxy-p-~-
galactopyranosyl units. A more complete investigation of the W.

(793)S. Toyoshima, Y. Akiyama, K. Nakano, A. Tonomura, and T.Osawa, Biochem-

istry, 10, 4457-4463 (1971).
(794)G. Cheung, A. Haratz, M. Katar, and R. D. Poretz, Abstr. P a p . Chem. Congr.
North Am. Continent I , BMPC 19 (1975).
(795)T. Osawa and S. Toyoshima, Methods Enzymol., 28, Part B, 328-332 (1972).
(796)T. Kurokawa, M. Tsuda, and Y. Sugino, /. Biol. Chem., 251, 5686-5693 (1976).
(797)S. Toyoshima and T. Osawa,J. Biol. Chem., 250, 1655-1660 (1975).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 313

Joribunda lectins should be rewarding, in that these substances will be

useful probes for both the chemist and the molecular biologist.
Two lectins were also isolated from Maackia amurensis seeds, one a
potent mitogen (MAM), the other a strong hemagglutinin (MAH).798
Boyd and coworkersssfirst described a hemagglutinin in M . amurensis
seeds. Purification involved affinity chromatography on procine-
thyroglobulin glycopeptides-Sepharose, and gel filtration.798Both lec-
tins were glycoproteins (mannose and 2-amino-2-deoxyglucose) of
-
molecular weight 130,000. Complex, carbohydrate-binding patterns
are apparent. Among the simple sugars, only 2-acetamido-2-deoxy-~
galactose and maltose had moderate inhibitory activity against both
MAH and MAM. Glycopeptide B from porcine thyroglobulin was a
potent inhibitor of MAM hemagglutination, but had no effect against
MAH. On the other hand, glycopeptide B and its sequential, enzymic-
degradation products inhibited the mitogenic activity of both M .
amurensis lectins (MAH also has a weak mitogenic
Multiple agglutinins were isolated from the hemolymph of the male
lobster (Homarus a m e r i c a n u ~ )One ~ the agglutinins (LAg-2) was
. ~ ~ of
reported to possess a site complementary to 2-acetamido-2-deoxy-o-
galactose; a second (LAg-1) was complementary to N-acetylneuraminic
acid.s00
Two lectins were purified by Bloch and his colleaguesso1from the
seeds of the pea tree (Caragana a b o r e s c e n ~ )One
. ~ ~of the blood-group
nonspecific glycoproteins (I) was purified by affinity chromatography
on a 2-acetamido-2-deoxy-~-galactose-substituted-Sepharose column;
the lectin is composed of two types of polypeptide chain (molecular
weight 30,000), cross-linked by disulfide bonds to form dimers that
appear to be in equilibrium with tetramers.sO1A strong agglutinin for
ABO erythrocytes, the lectin also agglutinates Ehrlich ascites cells,
and has a specificity for 2-acetamido-2-deoxy-~-galactose. Lectin 11,
a minor component, binds to underivatized Sepharose, and exhibits
low hemagglutinating activity.s01
Seeds from Euonymus europeus contain a mixture of lectins that
exhibit anti-(B + H) and A2 activity but no anti-Al activity,77~80~s0*~803 a

(798) T. Kawaguchi, I. Matsumoto, and T. Osawa, J . B i d . Chem., 249, 2780-2792

(1974).
(799) J. L. Hall and D. T. Rowlands, Jr., Biochemistry, 13,821-827 (1974).
(800) J. L. Hall and D. T. Rowlands, Jr., Biochemistry, 13, 828-832 (1974).
(801) R. Bloch, J. Jenkins, J. Roth, and M. M . Burger,J. Biol. Chem., 251,5929-5935
(1976).
(802) F. Pacik and J. Kocourek, Biochim. Biophys. Acta, 400,374-386 (1975).
(803) J. Petxyniak, M. E. A. Pereira, and E. A. Kabat, Arch. Biochem. Biophys., 178,
118-134 (1977).
314 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

finding that prompted the suggestion that these seed extracts could be
employed for distinguishing between subgroups A, and Az. Pacik and
KocourekEo2purified one of the E . europeus lectins (termed
phytohemagglutinin I) by preparative electrophoresis, and showed it to
be a homogeneous glycoprotein (1.9% of neutral sugar) of molecular
weight 127,000; the lectin contained bound Caz+(and traces of Mg2+,
Zn2+,and Cu2+),and aspartic acid as the sole, N-terminal amino acid.
None of the simple sugars tested inhibited lectin-induced hemaggluti-
nation. Petryniak and his colleaguesso3purified the E . europeus lectin
by adsorption onto insoluble polyleucyl hog A + H blood-group sub-
stance, followed by elution with lactose. The lectin, heterogeneous by
several criteria, had a molecular weight of 166,000 (compare Ref. 802)
and gave subunits of molecular weight 17,000 and 35,000 by elec-
trophoresis in dodecyl sodium sulfate on poly(acry1amide) gel. The
purified product, a glycoprotein (4.8% of D-galaCtOSe, 2.9% of
D-glucose, and 2.8% of 2-acetamido-2-deoxy-~-glucose),precipitated B
and H, but not Al, blood-group substances. Quantitative, hapten-
inhibition studies revealed theE. europeus lectin to be most specific for
an a-D-Galp-[cx-DFucp-(1+2)]-( 1+3)-p-D-Galp-( 1+3 or 4)-&~-GlcNAc
structure.803
A 2-amino-2-deoxy-~-hexose-binding lectin has been isolated from
barley.804A homogeneous protein of molecular weight 31,000, the bar-
ley lectin is devoid of carbohydrate and half-cysteine residues. The
lectin binds to the coat glycoprotein of barley stripe-mosaic virus by
way of 2-amino-2-deoxy-~-glucoseand -D-galactose units.E04Only the
free amino sugars, namely, 2-amino-2-deoxy-~-glucose,-D-galactose,
and -D-mannose inhibit barley-lectin-induced aggregation of barley
stripe-mosaic virus.804
The hemolymph from the elongate clam Tridacna maxima (Roding)
contains a lectin that agglutinated human type O - e r y t h r o c y t e ~ . ~ ~ ~ - ~ ~ ~
Purification of the agglutinin was accomplished by adsorption onto
polyleucyl-arabinogalactan followed by elution with a solution of
2-acetamido-2-deoxy-~-galactose.*~~ The isolated lectin (termed
Tridacnin) gave a single band in gel diffusion, but several bands in
disc-gel electrophoresis and isoelectric focusing.*05P-D-Galactans from
(804) J. Partridge, L. Shannon, and D. Gumpf, Biochim. Biophys. Acta, 541,470-483
(1976).
(805) B. A. Baldo and G . Uhlenbruck, FEBS Lett., 55,25-29 (1975).
(806) G. Uhlenbruck, B. A. Baldo, and G . Steinhausen, Z . Immunitaetsforsch. Allerg.
Klin. Immunol., 150,354-363 (1975).
(807) B. A. Baldo and G. Uhlenbruck, Carbohydr. Res., 40, 143-151 (1975).
(808) K. Eichmann, G. Uhlenbruck, and B. A. Baldo, Immunochemistry, 13, 1-6
(1976).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 315

Helix pomatia (L-galactose is also present), bovine lung, Larix occiden-

talis, and Lymnaea stagnalis L., as well as Pneumococcus type XIV
polysaccharide and hog H-substance, all precipitate the T . maxima
agglutinin.E06-E0E Hapten-inhibition studies revealed a primary spec-
ificity for p-linked 2-acetamido-2-deoxy-~-galactosyl units, although
the lectin also interacted strongly with nonreducing (terminal) p-D-
galactopyranosyl groups; oligosaccharides containing a - ~ -
galactopyranosyl residues were poor inhibitors or n o n i n h i b i t o r ~A. ~ ~ ~ ~ ~ ~ ~
comparative study of the interaction of the T . maxima lectin, Axinella
polypoides (sponge)agglutinin, and mouse-myeloma protein 5539 with
(1+6)-P-D-galactans has also been reported.E0E
A blood-group A lectin was isolated from frog(Rana catesbiana) eggs
by gel filtration and poly(acry1amide)-gel e l e c t r o p h o r e ~ i sA. ~glyco-
~
protein containing 1.8%of carbohydrate, the lectin had a molecular
weight of 210,000 and a carbohydrate-binding specificity that appears
to be directed towards nonreducing trisaccharide units [a-D-GalNAcp-
(1+3,4)-~-~-Galp-(1+4,3)-~-D-GlcNAcp-(l+R)], as well as internal
p-~-Galp-( 1+4)-P-D-GlcNAc residues.809The possibility exists that the
lectin preparation, although showing a single band in electrophoresis,
may be a mixture of two proteins having different s p e c i f i c i t i e ~ . ~ ~
A P-D-galactopyranosyl-binding protein was isolated from extracts of
electric-organ tissue of the electric eel (Electrophorus electricus) by
affinity chromatography on desulfated agarose (ECD-Sepharose).810
Termed “electrolectin,” the protein has a molecular weight of 33,000,
agglutinates trypsinized rabbit erythrocytes, and is specifically in-
hibited by P-Dgalactopyranosides (lactose or o-nitrophenyl p-D-
galactopyranoside), but not by a-D-galactopyranosides (melibiose or
raffinose).E1O
Commencing with the discovery by Dodd and his colleaguesE” that
sponges also contain hemagglutinins, several sponge lectins have been
studied.E12-E14 Most sponge lectins are considered to be relatively
nonspecific, in that they agglutinate A, By0,and AB erythrocytes to the
same extent. Differential agglutination of animal erythrocytes has,
however, been observed.E11

(809) F. Sakakibara, G. Takayanagi, H. Kawauchi, K. Watanabe, and S. Hakomori,

Biochim. Biophys. Actu, 444, 386-395 (1976).
(810) V. I. Teichberg, I. Silman, D. D. Beitsch, and G. Resheff, Proc. Nutl. Acad.
Sci. U.S.A., 72, 1383-1387 (1975).
(811) R. Y. Dodd, A. P. MacLennan, and D. C. Hawkins, Vox Sung., 15, 386-391
( 1968).
(812) S. Khalap, T. E. Thompson, and E. R. Cold, Vox Sung., 18,501-526 (1970).
(813) S. Khalap, T. E. Thompson, and E. R. Cold, Vox Sang., 20,150-173 (1971).
(814) H. Bretting, Z. Immunitaetsforsch. Allerg. Klin. lmmunol., 146,239-259 (1973).
3 16 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

Brettingsl*adsorbed to human A, B, and 0 erythrocytes two proteins

from crude extracts of the sponge Aaptos papillata. One of the proteins
exhibited hemagglutinating activity, and was isolated by gel filtra-
tion.81sOn using a column of polyleucyl blood-group A + H substance
followed by a column of DEAE-cellulose, and preparative, disc-gel
electrophoresis, Bretting and coworkersaI6isolated three Aaptos lectins
(I, 11, and 111). Poly(acry1amide)gel electrophoresis of lectin I in the
presence of dodecyl sodium sulfate gave two bands, of molecular
weight 12,000 and 21,000 respectively; Aaptos lectins I1 and I11 each
gave only one band, of molecular weight 16,000. The last two lectins
had similar amino acid compositions, whereas Aaptos lectin I had a
distinctive amino acid composition.816Lectin I precipitated blood-
group substances having terminal (nonreducing) 2-acetamido-2-
deoxy-D-glucosyl groups; N , N', N", N" '-tetraacetylchitotetraose was
the best inhibitor ofAaptos lectin I, being 2,000 times as active as
2-acetamido-2-deoxy-~-glucose. Lectins I1 and I11 precipitated
blood-group substances containing nonreducing (terminal)
2-acetamido-2-deoxy-~-glucosyl groups, 2-acetamido-2-deoxy-~-
galactosyl groups, or sialic acid residues, and were best inhibited by
N,N',N"-triacetylchitotriose. Aaptos lectin I was also precipitated by
the monovalent hapten p-nitrophenyl2-acetamido-2-deoxy-a-~-galac-
topyranoside.s'6
A lectin from the sponge Axinella spec. was purified by Gold and
colleague^,^'^ and was found to be an acidic protein having a molecular
weight of 15,000-18,000. It contained a negligible proportion of carbo-
hydrate, had an isoelectric point of 3.9, and contained bound Ca2+and
Fe3+. Lacto-N-tetraose was the best inhibitor of Axinella-induced
hemagglutination. Bretting and KabaP' isolated, and resolved, three
hemagglutinins from Axinella polypoides. The two main lectins (I and
11) were studied. A. polypoides lectins I and I1 had molecular weights
21,000 and 15,000, Both agglutinins contained -0.5%
of carbohydrate, and they were shown to be unrelated immunochemi-
cally and by amino acid composition.818Both lectins precipitated Al, Az,
B, and Lea blood-group substances, and were inhibited best by termi-
nal (nonreducing) P-D-galactopyranosyl groups.81a
Finally, although they are not classified as lectins, the interesting and
(815) H. Bretting and L. Renwrantz, Z. Immunitaetsforsch. Allerg. Klin. Immunol.,
147,250-261 (1974).
(816) H. Bretting, E. A. Kabat, J. Liao, and M. E. A. Pereira, Biochemistry, 15,
5029-5038 (1976).
(817) E. R. Gold, C. F. Phelps, S. Khalap, and P. Balding, Ann. N . Y . Acad. Sci., 234,
122-128 (1974).
(818) H. Bretting and E. A. Kabat, Biochemistry, 15,3228-3236 (1976).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 317

useful interactions between myeloma immunoglobulins and polysac-

charides should be mentioned (see, for example, G l a u d e m a n ~ ~ ' ~ ) .

VIII. CELL-SURFACE, GLYCOPROTEINS

LECTIN-REACTIVE
The nature of those cellular structures with which lectins interact has
been probed by (a) light- and electron-microscope analysis of tissue
sections following reaction with appropriately derivatized lectins, ( b )
mono- and oligo-saccharide inhibition of radiolabelled-lectin binding,
(c) competitive binding between two lectins of known specificity, (d)
the effect of glycosidase digestion on lectin reactivity, and, most con-
clusively, ( e ) isolation of the reactive structures themselves. Although
extensive use has been made of lectins (derivatized with ferritin,
fluorescein, and radioisotopes) as histological stains, this topic is be-
yond the scope of the present article; a limited bibliography is pre-
sented for the convenience of interested readers.92a,820-833
Several interesting discoveries have been made as a result of the
investigation of lectin-cell interaction. Many normal and transformed
cells exhibit differential agglutinability when tested with con A (Refs.
31, 250, 834, and 835), wheat-germ agglutinin (Refs. 28-30, 492, and
834-837), R. cornrnunis a g g l ~ t i n i n ,soybean
~ ~ ~ , ~agg1utiniqss4
~~ and len-
til l e ~ t i n .In~ ~many
~ cases, proteolytic digestion increases the
agglutinability of normal ce11s.492~836,837The binding of a lectin to a cell is
(819) C. P. J. Glaudemans, Adu. Carbohydr. Chem. Biochem., 31, 313-346 (19735).
(820) W. Bernhard and S. Avranieas, E x p . Cell Res., 64, 232-236 (1971).
(821) N. K. Gonatas and S. Avrameas,j. Cell Biol., 59,436-443 (1973).
(822) C. Huet and J. Garrido, E x p . Cell Res., 75, 523-527 (1972).
(823) G. L. Nicolson and S. J. Singer, Proc. Natl. Acad. Sci. U.S.A., 68, 942-945
(1971).
(824) R. M. Pratt, Jr., and W. A. Gibson, J. Histochem. Cytochem., 21,229-232 (1973).
(825) J . Roth and K. Thoss, Experientia, 30,414 (1974).
(826) J . Roth and K. Thoss, E x p . Pathol., 10, 258-267 (1975).
(827) J . Roth, K. Thoss, M. Wagner, and H . W. Meyer, Histochemistry, 43, 275-282
(1975).
(828) J. Roth, M. Wagner, and K. Thoss, E x p . Pathol., 11, 67-72 (1975).
(829) S. B. Smith and J.-P. Revel, Deu. Biol., 27, 434-441 (1972).
(830) J . D. Stobo and A. S. Rosenthal, E x p . Cell Res., 70,443-447 (1972).
(831) R. W. Stoddart and J. A. Kienian, Histochemie, 33, 87-94 (1973).
(832) K. Thoss and J. Roth, E x p . Pathol., 11, 155-161 (1975).
(833) M . J . Cline and D. C. Livingston, Nature (London) New Biol., 232, 155-156
(1971).
(834) A. A. Moscona, Science, 171,905-907 (1971).
(835) J . Roth, G. Neupert, and K. Thoss, Exp. Pathol., 10, 309-317 (1975).
(836) M. M. Burger, Fed. Proc., 32, 91-101 (1973).
(837) R. R. Gantt, J . R. Martin, and V. J. Evans, J. Natl. Cancer Iizst., 42, 369-374
( 1969).
318 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

a necessary, but not sufficient, condition for cell agglutination. For

example, certain normal cells bind con A and wheat-germ agglutinin,
but are not agglutinated, whereas their highly agglutinable, trans-
formed counterparts bind equivalent amounts of these lectins .489*833
(These findings have been c o n t e ~ t e d . ~Extended,
~ ~ * ~ ~ or
~ )complex, lec-
tin saccharide-binding sites were demonstrated for the P. vulgaris
and the Agaricus bisporus l e ~ t i nby
agglutinin’83,216~730*747 ~ ~using
j pur-
ified, membrane-derived glycopeptides as inhibiting structures in
place of simple oligosaccharides.216~435~730~747
The investigation of lectin-cell interaction by means of binding
studies with labelled lectins is informative, but limited, in that the
carbohydrate-binding sites of most lectins are still incompletely
characterized. A comprehensive understanding of lectin-cell interac-
tion requires, in addition to pure lectins of known properties, highly
purified, lectin-reactive molecules that are also well characterized. In
very few (if, indeed, any) cases has this ideal been achieved. It is to be
hoped that such more-complete analyses will be forthcoming.
In this Section, we consider in four sections, based on their cellular
origin, lectin-reactive, membrane glycoproteins and glycopeptides that
have been at least partially purified and characterized: 1 , erythrocytes
and platelets; 2, lymphocytes; 3, neuronal tissue; and4, tumor cells. In
reviewing lectin-reactive, cell-membrane glycoproteins, we have
avoided the term “lectin receptor.” The term “receptor” should, we
believe, be reserved for those unique, membrane structures that bind
external molecules in a highly specific way, thereby transmitting sig-
nals from the environment to the interior of the cell. Hormones and
drug receptors are examples of such unique structures. The interaction
between lectins and cell surfaces is more general, in that lectins will
react with a n y cell-surface, glycosyl moiety that is complementary to its
binding site. It is, of course, possible for a lectin to interact with a
unique “receptor” as defined; but this is only one of the cell-surface,
carbohydrate-containing structures available. Several reviews deal
with lectin-reactive, membrane g l y c o p r o t e i n ~ . A ~valuable
~~~~~~~~~~~
study on the activities of lectins and their immobilized derivatives in
detergent solutions has been

1. Erythrocytes and Platelets

Lectin-reactive glycoproteins of the erythrocyte membrane have
been studied extensively. Of these, the major sialoglycoprotein has
(838) K. D. Noonan and M. M. Burger,J. Biol. Chem., 248,4286-4292 (1973).
(838a) R. Lotan, G . Beattie, W. Hubbell, and G. L. Nicolson, Biochemistry, 16, 1787-
1794 (1977).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 319

been most completely characterized. This glycoprotein has a molecular

weight of 53,000-55,000,and contains 55-60% of ~ a r b o h y d r a t e . It ~~~-~~~
carries the immunodeterminant structure for blood-group MN, but not
ABO, specificity.842 The carbohydrate part of the major sialo-
glycoprotein occurs in two, distinct, oligosaccharide chains, an
0-glycosylically linked t e t r a s a c ~ h a r i d e and
~ ~ ~ an N-glycosylically
linked octa~accharide,~~"."~in the ratio435,841,843 of -6:1. Thomas and
W i n ~ l e rdetermined
~~~ the structure of the 0-glycosylically linked
chains to be those depicted in formula 15. 2-Acetamido-2-
0-AcNeu
2
16
0-AcNeu- (2- J)-p-D-Galp- (I-- 3 ) - -GalNAcp-
~ (l+O)-Ser ,Thr

15

deoxygalactose was found only in 0-linked chains; 2-acetamido-2-

deoxyglucose and mannose were constituents of the N-linked oligosac-
.216,730,842

The K ~ r n f e l disolated ~ ~ ~a P~. vulgnris

~ ~ ~ ~lectin-reactive
* ~ ~ ~ glycopep-
tide from human erythrocytes; it derived horn the major sialoglycopro-
tein. The trypsin-released glycopeptides were reduced with alkaline
borohydride (without loss of lectin reactivity), digested with pronase,
and chromatographed repeatedly on Sephadex gels and DEAE-
cellulose. The purified glycopeptide represented a 3.7% yield of the
initial activity. The single, N-glycosylically bound chain was composed
of sialic acid, D-galactose, D-mannose, and 2-acetamido-2-deoxy-~-
glucose residues linked to asparagine. Digestion with glycosidase re-
vealed a branched structure; one of two nonreducing (terminal) p - ~ -
galactopyranosyl groups was terminally substituted by sialic
a ~ i d ~ as ~ shown
~ , ~ ~ ~ * 14.
in formula ~ Although
~ ~ , desialization
~ ~ ~ did not
alter kidney-bean lectin reactivity, removal of D-galaCtOSe abolished
the inhibitory capacity. The chloroform-methanol-extracted, major
sialoglycoprotein was 90 times as effective as its glycopeptide fragment
in the kidney-bean lectin-erythrocyte hemagglutination inhibition as-
say, an effect ascribed to its m u l t i ~ a l e n c e . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
(839) J. P. Segrest, R. L. Jackson, E. P. Andrews, and V. T. Marchesi, Biochem.
Biophys. Res. Commun., 44, 390-395 (1971).
(840) J. P. Segrest, 1. Kahane, R. L. Jackson, and V. T. Marchesi, Arch. Biochem.
Biophys., 155, 167-183 (1973).
(841) M. Fukuda and T. Osawa,J. Biol. Chem., 248,5100-5105 (1973).
(842) D. B. Thomas and R. J. Winzler,J. Biol. Chem., 244,5943-5946 (1969).
(843) R. L. Jackson, J. P. Segrest, I. Kahane, and V. T. Marchesi, Biochemistry, 12,
3131-3138 (1973).
320 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

The lectins of the and Robinia p ~ e u d o a c a c i aappear

~ ~ ~ to
interact with the N-linked oligosaccharide of the major, erythrocyte
sialoglycoprotein. The trypsin-released glycopeptide already men-
tioned inhibited hemagglutination by these two lectins. Alkaline
borohydride removal of serine- and threonine-bound oligosaccharides
had no effect on lectin reactivity. Furthermore, P . vulgaris lectin com-
petitively inhibited binding of radiolabelled R. p ~ e u d o a c a c i aor~len-
~~
til l e ~ t i n ~
to3erythrocytes,
~ Conversely, R. pseudoacacia and lentil
lectins partially blocked kidney-bean lectin binding. These lectins may
interact with unique portions of a single oligosaccharide.
The lectin of commercial mushroom (Agaricus bisporus), on the
other hand, apparently binds to 0-glycosylically linked chains.435
Whereas trypsin-released glycopeptide I (with both types of chains)
inhibited lectin-induced hemagglutination, neither fragment of the
alkaline borohydride-reduced sialoglycoprotein exhibited activity.
However, glycopeptide B, a pronase-digested fragment of glycopep-
tide I possessing only 0-linked oligosaccharide chains, was strongly
inhibitory. Presant and K ~ r n f e l dsuggested
~~~ that the mushroom (A.
bisporus) lectin displays an extended binding-site, part of which rec-
ognizes the 0-glycosylic linkage to serine or threonine. Chemical or
enzymic removal of D-galactose from desialized glycopeptide B
abolished the mushroom-lectin reactivity.
Jackson and coworkerss43also studied glycopeptides derived from
the major sialoglycoprotein of the erythrocyte membrane. Purified
glycoprotein was degraded with trypsin; a combination of ion-
exchange chromatography and gel filtration then separated glycopep-
tides a-1, a-2, a-3, and p. Only glycopeptide a-1 contained
2-acetamido-2-deoxy-~-glucose residues, in which it accounted for
40% of the total hexosamine and was found only in N-glycosylic link-
age, presumably to asparagine. 2-Acetamido-2-deoxy-~-galactose and
sialic acid were constituents of each glycopeptide. Reduction with
alkaline borohydride destroyed virtually all of the 2-acetamido-2-
deoxy-D-galactose, suggesting that it occurred only in 0-glycosylic
linkage. Approximately half of the serine and threonine units of each
glycopeptide were substituted by 2-acetamido-2-deoxy-~-galactose
residues.843 Only the a-1 glycopeptide reacted with wheat-germ
agglutinin, as determined by a hemagglutination inhibition assay. In-
asmuch as the a-1 glycopeptide alone contains N-linked oligosac-
charide chains, these chains must account for the wheat-germ reactiv-
ity. It is more difficult to explain the finding that the P. vulgaris lectin
was inhibited by both the a-1and the p glycopeptide, but not by the a-2
or a-3 glycopeptide, as all have 0-linked carbohydrate chains. This
observation is at variance with the Kornfelds’ demonstration216that P .
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 32 1

vulgaris lectin reacted only with the N-linked chains. It is possible that
the reactive, 0-linked oligosaccharide of the P-glycopeptide differs
from the tetrasaccharide structure elucidated by Thomas and Winz-
ler.842In fact, Jackson and coworkers843demonstrated, by trypsin treat-
ment of red cells, that the a-3 and P-glycopeptides are derived from
buried portions (of the major sialoglycoprotein) which are unavailable
to trypsin and, therefore, would not have been among the glycopep-
tides studied by Thomas and W i n ~ l e r . * ~ ~
KubLnek and colleagues844isolated a glycopeptide reactive with pea
agglutinin from human, B erythrocyte ghosts. The purification scheme
involved delipidation of ghosts with chloroform-methanol, followed
by pronase digestion. Released glycopeptides were precipitated with
ethanol, dissolved in 0.1 M acetic acid, and separated into two neutral,
sugar-containing fractions (I and 11) by Sephadex G-25 gel-filtration.
Fraction I, reactive with pea lectin, was further separated by vertical-
paper electrophoresis in pysidine-acetate buffer, pH 5.6, into 5 pep-
tides. Peptide 1.3 exhibited 79% of the pea-lectin reactivity present in
unseparated fraction I, 14.3%of the total, neutral sugar of the erythro-
cyte ghost, and considerable blood-group B activity. Moreover, it inhib-
ited both pea and lentil lectin-induced hemagglutination 32 times as
effectively as D-glucose; the reactivity of the glycopeptide with con A
was 5 times that with D-glucose. Glycopeptide 1.3 was comprised of 1
sialic acid, 4 D-galactose, 2 D-mannose, 1 L-fucose, 8 2-acetamido-2-
deoxy-D-glucose, 1 aspartic acid, 1 serine, and traces of threonine,
glutamic acid, glycine, and alanine residues, giving a calculated
molecular weight of -4,000. By virtue of its content of 2-acetamido-2-
deoxy-D-glucose, this glycopeptide may contain N-glycosylically
linked chains derived from the major sialoglycoprotein.216~43s~730~747
PospiSilovL and coworkers'8s investigated the chemical structure of
the pea lectin-reactive glycopeptide isolated by KubLnek and cowork-
e r ~After
. ~ refining
~ the chemical analysis, and applying glycosidase

-
digestion, they proposed the structure depicted in formula 16. Despite

- -
Gal GlcNAc
\
-
Man GlcNAc Asn

Gal GlcNAc
/
16

the fact that pea hemagglutinin is not inhibited by D-galactose, 90% of

the pea reactivity of this glycopeptide is reportedly due to the two
(844) J. KubBnek, G. Entlicher, and J. Kocourek, Biochim. Biophys. Acta, 304, 93-
102 (1973).
322 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

nonreducing (terminal) D-galactosyl groups.1ssThis result is difficult to

reconcile with those from the sugar inhibition analysis of pea aggluti-
nin specificity (see Section 11,3), The possible presence of two
D-mannose units, one being a 2-0-a-D-mannopyranosyl residue, would
account for the reactivity of this glycopeptide with the pea lectin (com-
pare Refs. 209 and 210).
Adair and Komfeld investigated the behavior of detergent-extracted,
~ ~ ~ X-100-
erythrocyte glycoproteins with a series of l e ~ t i n s .Triton *~~~
sodium borate buffer solubilized 40-50% of erythrocyte ghost protein
and 75-85% of sialic acid, leaving undissolved between 10 and 30% of
the binding capacity for the following lectins: Abrus precatorius,
Agaricus bisporus, Lens culinaris, Phaseolus vulgaris, Ricinus com-
munis, and Triticum vulgaris. Inhibition by cell extracts was assayed
by reduction of [1251]-lectinbinding to erythrocyte ghosts. Affinity-
column chromatography employing Sepharose-coupled R . communis
agglutinin (RCAIzo)or wheat-germ agglutinin resulted in further pur-
ification of the detergent extract.aaaBoth columns were deliberately
overloaded in order to provide maximal yields of reactive glycoproteins
and to lessen nonspecific adsorption; this procedure should result in
the selective adsorption of the glycoprotein of highest affinity in the
event that several lectin-reactive species should be present. The lac-
tose eluate of RCAlz0-Sepharose columns gave three protein bands
(plus minor bands), but no clear bands staining with periodic acid-
Schiff reagent.aaaOne broad band, encompassing the major, sialo-
glycoprotein region, and probably representing several species, re-
sulted when gel slices were assayed for hexosamine. The glycoproteins
reactive with the lectins from A. bisporus, P. vulgaris, and L. culinaris
all failed to bind to RCA120-Sepharose.66a The combined evidence
suggests that the R. communis agglutinin does not bind the major,
erythrocyte sialoglycoproteinaaawith which the mushroom, kidney-
bean, and lentil lectins r e a ~ t . ~ The ~ ~ R.
, ~communis,
~ ~ * ~ affinity-
~ ~ * ~ ~ ~
isolated glycoprotein was 1,200 times as active as D-galactose in the
standard inhibition assay.6a6Moreover, it inhibited A . precatorius lec-
tin, but not the mushroom, wheat-germ, or kidney-bean agglutinins.
In contrast, when erythrocyte-ghost extract was applied to a wheat-
germ agglutinin-Sepharose column, the 2-acetamido-2-deoxy-~-
glucose eluate contained only the major sialoglycoprotein, as judged by
electrophoretic mobility and chemical composition.66aThis prepara-
tion inhibited wheat-germ agglutinin 15,000 times as effectively as
2-acetamido-2-deoxy-~-glucose,and blocked A . bisporus and P . vul-
garis lectin binding to erythrocyte ghosts as well. Pronase degradation
of the native glycoprotein decreased its inhibitory activity by 98%; the
authorsaa6ascribed this observation to the conversion of a multivalent
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 323

into a monovalent hapten. Neuraminidase treatment decreased

wheat-germ agglutinin reactivity to one-ninth of its previous level.
In agreement with Adair and Kornfeld,666Kahane and coworkersw5
reported a single-step purification of the major, erythrocyte sialogly-
coprotein from detergent-extracted ghosts by means of wheat-germ
agglutinin-Sepharose affinity chromatography. Furthermore, R. com-
munis and P . vulgaris hemagglutinin-Sepharose columns also bound
the major sialoglycoprotein, whereas con A and P . vulgaris
leukoagglutinin-Sepharose columns did
A slightly different, membrane-extraction procedure was employed
by find la^^^^ in a study of erythrocyte glycoproteins. (Ethylenedini-
tri1o)tetraacetate- and Triton X-100-extracted, membrane glycoprotein
was chromatographed on columns of Sepharose-coupled con A or lentil
lectin. Eight percent of the applied material bound to con
A-Sepharose, and was specifically eluted with methyl a - ~ -
mannopyranoside solution, whereas the corresponding figure for the
lentil column was 30%.There was no apparent difference between the
two elution-profiles, or between the protein-stained, electrophoretic-
gel patterns. However, staining with the periodic acid-Schiff reagent
showed that lentil-Sepharose had retained both the major sialoglyco-
protein and a minor glycoprotein [PAS 2, component a (Ref. 847); com-
ponent I11 (Ref. 848)],whereas con A-Sepharose retained only PAS 2.
Of the total, minor glycoprotein(s), 20-30%bound to con A-Sepharose,
60-80% bound to lentil lectin-Sepharose, and a fraction did not bind to
either column. Moreover, a portion of the glycoprotein unreactive with
con A-Sepharose did adsorb to a lentil lectin-Sepharose column. On
the basis of these findings, find la^^^^ postulated heterogeneity in the
oligosaccharide of PAS 2. In accordance with the results of Kahane and
coworker^,^^ there was no evidence of interaction between con A and
the major sialoglycoprotein.
Fukuda and Osawawl utilized detergent extraction, ion-exchange
chromatography, and gel filtration to isolate the major sialoglycoprotein
of human, 0-erythrocyte membrane in pure form. Purity was dem-
onstrated in ultracentrifugal studies and b y dodecyl sodium sulfate
acrylamide-gel electrophoresis. The glycoprotein was devoid of tryp-
tophan and cysteine, and contained very little methionine,
phenylalanine, or tyrosine. Sialic acid and D-galactose were abundant;

(845) I . Kahane, H. Furthmayr, and V. T. Marchesi, Biochim. Biophys. Acta, 426,

464-476 (1976).
(846) J. B. C. Findlay,J. B i d . Chem., 249,4398-4403 (1974).
(847) M. S. Bretscher,J. Mol. Biol., 59, 351-357 (1971).
(848) G. Fairbanks, T. L. Steck, and D. F. H. Wallach, Biochemistry, 10, 2606-2617
(1971).
324 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

2-acetamido-2-deoxy-D-galactose, 2-acetamido-2-deoxy-~-glucose,
D-mannose, and L-fucose were also identified. By hemagglutination
inhibition assay, kidney-bean lectin and wheat-germ agglutinin in-
teracted strongly with the glycoprotein, and C . sessilifolius lectin and
the Ulex europeus lectins [both blood-group H ( 0 ) specific] interacted
with moderate affinity. Inhibition by the glycoproteins of the lectins
from S . japonicu, eel serum, B . purpureu, and V. gruminea (blood-
group N specific) was weak; R . communis lectin, con A, and lima-bean
lectin were not inhibited. Alkaline borohydride cleavage of
O-glycosylically linked chains severely lowered the reactivity with the
blood-group N and H ( 0 ) specific lectins, but increased the reactivity
towards several blood-group, nonspecific lectins (con A, R . communis,
lentil, pea, V. fuba).This suggested that, in some way, O-linked chains
mask the reactivity of N-linked chains. In contrast to Kahane and co-
w o r k e r ~and
~ ~FindlayYM6
~ Fukuda and 0sawas4l demonstrated inter-
action between the major sialoglycoprotein and con A. Furthermore,
theys4*reported interaction between R. communis agglutinin and the
major sialoglycoprotein, as did Kahane and coworkerss45;however, this
finding is contested by the results of Adair and Kornfeld.666 The resolu-
tion of these discrepancies awaits further investigation.
Human A erythrocyte stroma, pre-extracted with 26% potassium
chloride solution, yielded soluble glycopeptides upon repeated diges-
tion with a - c h y m ~ t r y p s i n . ~After
~ ~ J ~precipitation
l by 90%-saturated
ammonium sulfate, and extraction with phenol, this preparation con-
tained inhibitory activity toward several lectins. Fractionation was
performed with DEAE-cellulose, Sephadex G-100, and Sepharose 6B.
Numerous fractions were obtained, none of which were characterized.
Several fractions exhibited slight preferential reactivity towards lentil,
lima-bean, pea, wheat-germ, and V. faba lectins. The use of digestion
with a-chymotrypsin during the course of extraction precludes direct
comparison of these glycopeptides with the tryptic peptide fragments
derived from the major, erythrocyte glycoprotein studied previ-
ous~y,216.730,747,S43

Akedo and purified two rat-erythrocyte glycoproteins

which interacted with con A of the nata bean, Canavalia gladiata.
Rat-erythrocyte stroma were solubilized with dodecyl sodium sulfate
and 2-mercaptoethanol. The acetone precipitate of the supernatant
liquor was redissolved in detergent, and separated by preparative,
acrylamide-gel e l e c t r o p h o r e ~ i s . ~
Two
~ ~ electrophoretically ho-
mogeneous, con A-reactive glycoproteins, BPI and BP2 (molecular
weight 200,000 and 300,000, respectively), were eluted. They were
comprised of protein, neutral sugar (25% by weight, BPI; 19% by
weight, BPJ, and small proportions of sialic acid; neither had sig-
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 325

nificant hexosamine contents. P. vulgaris agglutinin did not react with

either glycoprotein.
Human-platelet membranes have been extracted with lithium 3,5-
diiodosalicylate, and the soluble glycoproteins separated on
O-phosphonocellulose.a4pThe purified, membrane glycoprotein,
molecular weight 100,000, was immunochemically identical to a sam-
ple purified from platelet-membrane extract by means of con
A-Sepharose affinity chromatography. No further characterization was
reported.

2. Lymphocytes
A mixture of lentil lectin-reactive glycoproteins from pig
lymphocyte-plasma membrane was isolated by lentil lectin-Sepharose
chromatography of sodium deoxycholate-solubilized membrane^."^
Eighty-seven percent of the protein applied (17% of hexose) passed
through unretarded, and 13% of the applied protein (83% of hexose)
was bound, and eluted with methyl a-D-glucopyranoside solution. Re-
covery was 95% of the material applied, in contrast to the recovery in
similar experiments conducted on con A-Sepharose columns (80%
recovery).850The eluate from the lentil column, which contained at
least ten glycoproteins, blocked lymphocyte transformation induced by
lentil or kidney-bean l e ~ t i n s . " ~
A highly purified, con A-reactive glycopeptide was obtained from
calf thymocytes by PospiiilovL and colleaguess5' by using essentially
the same protocol developed by KubLnek and coworkersa44to isolate
the pea lectin-reactive glycoprotein from erythrocytes. Pronase diges-
tion of delipidated membranes yielded soluble g l y c ~ p e p t i d e sThese
.~~~
were separated by gel filtration, and repeated, vertical, descending
paper-electrophoresis in pyridine-acetate buffer, pH 5.6, yielding 12
mg of con A-active glycopeptide from 16 g of membranous material,8s1
The glycopeptide represented 0.15% of the membrane hexose, and
contained 10.4% of carbohydrate. Analysis revealed the presence of
D-mannose, D-galactose, D-glucose, L-fucose, 2-amino-2-deoxy-D-
glucose, and sialic acid, as well as many amino acids (glycine and
alanine preponderated). A minimum molecular weight of 12,000 was
calculated. The glycopeptide inhibited [1311]conA binding to calf
thymocytes 200 times more effectively than methyl CX-D-
(849) R. L. Nachman, A. Hubbard, and B. Ferris, J . Biol. Chem., 248, 2928-2936
(1973).
(850) D. Allan, J. Auger, and M. J. Crumpton, Nature (London) New Biol., 236, 23-
25 (1972).
(851) J. PospiHilovi, C. HaSkovec, G . Entlicher, and J. Kocourek, Biochim. Biophys.
A c ~ u373,444-452
, (1974).
326 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

mannopyranoside (compared on a molar basis), and, in addition,

blocked con A-induced, DNA synthesis by calf thymocytes.8s1

3. Neuronal Cells
Lectin affinity columns have been used to isolate, and purify, glyco-
proteins and glycopeptides from neuronal tissue, especially brain.
Gombos and coworkersa52employed glutaraldehyde-insolubilized con
A to isolate a series of glycopeptides from the pronase-treated, micro-
soma1 fraction of rat brain. These glycopeptides were rich in mannose
and 2-acetamido-2-deoxyglucose,and stimulated neurite growth in
tissue culture.8s2 Similar results were obtained by Susz and col-
l e a g u e ~ who
, ~ ~ ~chromatographed whole-brain extracts, solubilized
with deoxycholate and dodecyl sodium sulfate, on con A-Sepharose. In
a continuation of these studies, it was shown that 25-30% of the total
glycopeptides, obtained by papain treatment of delipidated, whole
brain, bound to a con A-Sepharose column.8s4Elution with 5% methyl
a-D-glucopyranoside solution gave a series of glycopeptides that con-
tained mainly mannose and 2-acetamido-2-deoxyglucose,as well as
small proportions of galactose and f ~ c o s eTreatment
.~~~ of a purified,
glycopeptide fraction with a-D-mannosidase drastically lowered its
affinity for con A, suggesting the presence of terminal (nonreducing)
a-D-mannopyranosyl groups,8s4
A water-soluble, 50%-methanol-soluble, acidic glycoprotein was iso-
lated from rat brain by affinity chromatography on con A-Sepharose.8ss
The glycoprotein, pure by poly(acry1amide)gel-electrophoresis at pH
8.8, had an apparent molecular weight of 14,500 +1,400 by dodecyl
sodium sulfate gel-electrophoresis, No analysis for carbohydrate was
reported.8ss
Employing columns of immobilized, lentil lectin and wheat-germ
agglutinin, Gurd and MahlersS6isolated a series of glycopeptides from
deoxycholate-extracted, synaptic-plasma membranes; no analytical
data were provided. In a similar study, Zanetta and coworkersss7
chromatographed the dodecyl sodium sulfate-solubilized extract from

(852) G. Gombos, J. C. Hermetet, A. Reeber, J.-P. Zanetta, and J. Treska-Ciesielski,

FEBS Lett., 24, 247-250 (1972).
(853) J. P. Susz, H. I. Hof, and E. G. Brunngraber, FEBS Lett., 32, 289-292 (1973).
(854) J. I. Javaid, H. I. Hof, and E. G. Brunngraber, Biochim. Biophys. Acta, 404,
74-82 (1975).
(855) G. Ramirez, K. G. Haglid, B. Karlsson, and L. Ronnback, FEBS Lett., 38,
143-146 (1974).
(856) J. W. Gurd and H. R. Mahler, Biochemistry, 13,5193-5198 (1974).
(857) J,-P. Zanetta, I. G. Morgan, and G. Gombos, Bruin Res., 8 3 , 3 3 7 4 4 8 (1975).
 LECTINS: CARBOHYDRATE-BINDING PROTEINS 327

delipidated rat-brains on a con A-Sepharose column. A portion of the

extract (fraction C1) bound to the affinity column, and was eluted with
methyl a-D-glucopyranoside solution.857Fraction C 1 was markedly en-
riched with respect to D-mannose and 2-acetamido-2-deoxy-D-glucose;
its electrophoretic profile was complex, showing multiple

4. Tumor Cells
The experiments of Walborg and his colleagues have focused on two
rat-hepatoma lines that grow in ascitic form and exhibit different
agglutinability patterns with con A and wheat-germ a g g l ~ t i n i n . ~ ~ ~ - ~ ~
Whereas Novikoff hepatoma cells are readily agglutinated by both
lectins, AS-SOD hepatoma cells are much more s u ~ c e p t i b l e ~ to ~~~~~’-~
agglutination by wheat-germ agglutinin than by con A. Papain diges-
tion of either tumor line not only rendered the cells agglutinable by low
concentrations of either wheat-germ agglutinin or con A, but also re-
leased sialoglycopeptides capable of inhibiting con A- or wheat-germ-
agglutinin-induced, guinea-pig-erythrocyte h e m a g g l ~ t i n a t i o n . ~ ~ ~ . ~ ~ ~
A crude, sialoglycopeptide fraction (representing 65-80% of
neuraminidase-labile sialic acidEs8)was obtained by Sephadex G-50
chromatography of the papain
By employing Sephadex chromatography of the pronase digest, Wray
and W a l b ~ r gpartially
~ ~ ~ resolved the crude mixture of glycopeptides
from Novikoff cells into a fraction of molecular weight >3,300 that
inhibited both lectins, and a fraction of molecular weight 2,000-3,300
that inhibited only con A. Each fraction was further resolved by ion-
exchange c h r o m a t ~ g r a p h yAlternatively,
.~~~ the crude papain-digest of
either tumor line was separated into sialic acid-containing fractions A,
B, and C by chromatography on Sephadex G-50 equilibrated with 0.1 M
acetic a ~ i d . AS30D - ~ ~ A
~ ~ ~ fraction ~ demonstrated wheat-germ aggluti-
nin and con A reactivity, fraction C contained only con A reactivity, and
fraction B inhibited neither lectin.861Fraction A from AS30D cells was
degraded with pronase, and resolved861into A1 and A11 on Sephadex
G-200. A1 (excluded from Sephadex G-200)appeared to be resistant to

(858) E. F. Walborg, Jr., R. S. Lantz, and V. P. Wray, Cancer Res., 29, 2034-2038
(1969).
(859) V. P. Wray and E. F. Walborg, Jr., Cancer Res., 31,2072-2079 (1971).
(860) D. F. Smith and E. F. Walborg, Jr., Cancer Res., 32,543-549 (1972).
(861) D. F. Smith, G. Neri, and E. F. Walborg, Jr., Biochemistry, 12, 2111-2118
(1973).
(862) G. Neri. D. F. Smith, E. B. Gilliam, and E. F. Walborg, Jr., Arch. Biochem.
Biophys., 165,323-330 (1974).
(863) G. Neri, D. F. Smith, E. B. Gilliam, and E. F. Walborg, Jr., in “Comparative
Biochemistry and Physiology of Transport,” L. Bolis, K. Bloch, S. E. Luria, and
F. Lynen, eds., North-Holland Publishing Co., Amsterdam, 1974.
328 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

proteolysis, and exhibited both wheat-germ agglutinin- and con

A-inhibiting activity, whereas A11 was nonreactive by hemagglutina-
tion inhibition.861A1 was separated into three components by ion ex-
change on DEAE-cellulose in pyridine-acetic acid buffer. Component
DAI-2, representing 12% (by weight) of the crude sialoglycopeptide
mixture, strongly inhibited wheat-germ agglutinin; it was composed of
relatively high proportions of aspartic and glutamic acids, and serine
and threonine, with no methionine or arginine. Furthermore, this com-
ponent was comprised of 18.8% of 2-amino-2-deoxy-D-glucose and
12.7% of D-gahCtOSe, with lesser proportions of 2-amino-2-deoxy-~-
galactose, D-mannose,L-fucose, D-glucose, and sialic acid. On the other
hand, the fraction most reactive with con A was DC-2, obtained by ion-
exchange chromatography of fraction C. DC-2 represented 7% (by
weight) of the original sialoglycopeptide mixture, but was not chemi-
cally analyzed.861
A similar resolution of Novikoff cell fraction A was achieved by
using862*sm Sephadex G-200. Fraction A1 had a high molecular weight
(100,000 to 200,000) and possessed lectin reactivity, whereas A11 did
not. DEAE-cellulose ion-exchange in pyridine-acetic acid buffer
distinguished eight component^.^^^*^^^ Con A reactivity was associated
with DAI-2 (which was ten times as active as the crude mixture); peak,
wheat-germ-agglutinin reactivity was associated with DAI-4 (four
times as active as the crude mixture). Chemical analysis of a DAI-1,2
mixture gave results very similar to those of AS-SOD fraction DAI-2;
glutamic and aspartic acids, serine, and threonine were abundant, and
30% of the carbohydrate was hexosamine; D-mannose, D-glucose,
D-galactose, L-fucose, and sialic acid were also present.862
Inhibitory activity of fraction A1 towards wheat-germ agglutinin and
con A paralleled the agglutinability of the tumor cell from which it
derived: Novikoff A1 exhibited potent reactivity towards both lectins;
AS-SOD A1 reacted strongly with wheat-germ agglutinin, and
minimally863with con A. The lectin-reactive component of A1 appeared
to be a peptide highly substituted with oligosaccharide side-chains, as
indicated by its exclusion from Sephadex G-200, its high content of
amino acids known to be involved in protein-carbohydrate linkages,
and its resistance to proteo1ysksa The structure of the reactive
oligosaccharide was not determined.
Mouse L1210 leukemia cells were the source of a fraction containing
wheat-germ-agglutinin-reactive g l y c o p r o t e i n ~ . ~Insoluble
~~-~~~
(864)M. M. Burger, Nature (London),219,499-500 (1968).
(865) V. K. Jansons and M. M. Burger, Biochim. Biophys. Acta, 291, 127-135 (1973).
(866) V. K. Jansons, C. K. Sakamoto, and M. M. Burger, Biochim. Biophys. Acta, 291,
136-143 (1973).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 329

material obtained by hypotonic incubation of L1210 cells, when

extracted with either phenol, guanidinium chloride, pyridine, or
lithium 3,5-diiodosalicylate, yielded a supernatant solution con-
taining at least four components (by gel electrophoresis), each of
which contained 2-acetamido-2-deoxy-~-glucose.~~~ The mixture
inhibited wheat-germ lectin agglutination of L1210, Py3T3, and
PyBHK transformed cells,x6s but not agglutination by con A or
Phaseolus vulgaris hemagglutinin.866Antiserum to the mixture of wheat-
germ-agglutinin-reactive glycoproteins showed specificity towards
L1210 cells, but it did not react with normal-mouse lymphocytes or
erythrocytes, with which wheat-germ agglutinin does react.x66The
antiserum blocked wheat-germ-agglutinin, but not con A agglutination
of Py3T3 cells.866Resolution of the supernatant solution on Sephadex
G-200 in 33% pyridine with 2-mercaptoethanol gave two broad peaks;
activity was associateds6' primarily with a peak having a molecular
weight of 40,000 to 60,000.
A lithium 3,5-diiodosalicylate extract of mouse L929 cells,
prelabelled in vitro with 2-amino-2-deoxy-~-~H]glucose and amino
[14Clacids,was separated by chromatography on Sephadex G-200 into
two fractions.s67Fraction A was excluded from the gel, and contained
one third of the applied 2-amino-2-deoxy-~-[~H]g~ucose; fraction B,
retarded by Sephadex G-200, contained two-thirds of the applied
2-amino-2-deoxy-~-[~H]g~ucose. Fraction A was applied to a column of
con A-Sepharose from which a glycoprotein of molecular weight
-100,000 was eluted with methyl a-D-mannopyranoside solution. The
glycoprotein migrated as a single component in dodecyl sodium sulfate
gel-electrophoresis and represented 51% of the 2-amino-2-deoxy-~-
[3H]glucoseapplied. Similar results were obtained when plasma mem-
brane, labelled by means of D-galactose oxidase and potassium boro-
tritide, was extracted and separated in an analogous manner.867Amino
acid analysis showed a relative abundance of serine, glycine, glutamic
acid, and alanine. Valine was found in the N-terminal position. Cleav-
age of the intact glycoprotein with cyanogen bromide resulted in five
fragments; each contained bound 2-amino-2-deoxy-~-[~H]glucose,
suggesting that the L929 membrane glycoprotein was multi-substi-
tuted by oligosaccharide chains.s67
Nachbar and coworkers investigated the composition of Ehrlich
ascites, tumor-cell plasma membrane by using a series of
lectin-Sepharose adsorbents ( R . communis hemagglutinin, con A,
wheat-germ agglutinin, and soybean agglutinin).s68The Ehrlich cells
(867) R. C. Hunt, C. M. Bullis, and J. C. Brown, Biochemistry, 14, 109-115 (1975).
(868) M. S. Nachbar, J. D. Oppenheim, and F. Aull, Biochim. Biophys. Acta, 419,
512-529 (1976).
330 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

were agglutinated by RCAIzo,wheat-germ agglutinin, con A, and

soybean agglutinin in decreasing order of agglutinability; Ulex
europeus I did not agglutinate the cells. For membrane protein from
Ehrlich cells labelled in vivo with ~-[~H]2-amino-2-deoxyglucose,
80-90% was extracted by 10% deoxycholate in Tris [2-amino-2-(hy-
droxymethyl)-1,3-propanediol] buffer. When the extract was separated
by means of one of the four lectin-Sepharose columns, electrophoretic
profiles of the specifically bound, sugar-eluted fractions showed both
common and unique peaks for each lectin. The fractions were neither
purified, nor characterized with regard to chemical composition.868
Kim and coworkers examined glycopeptides from normal, and
cancerous, colonic m ~ c o s a .Pronase
~~~ digestion of disrupted
cell-membranes gave a soluble glycopeptide fraction. The fractions
obtained from normal tissues inhibited Dolichos biflorus, but not
Ricinus communis, hemagglutination, whereas the reverse was true of
fractions obtained from malignant, colonic mucosa. No further analysis
was carried O U ~ . ~ ~ ~
Carcinoembryonic antigen (CEA), a glycoprotein (molecular weight
200,000) purified to homogeneity from malignant, human,
gastrointestinal-tract tumors,870 contains 50% (by weight) of
carbohydrate.413Hammarstrom and coworkers413 listed a series of
lectins that precipitated CEA, including con A, and P. vulgaris, R .
communis, and wheat-germ agglutinins. Helix pomatia, Dolichos
biflorus, and B . simplicifolia I did not precipitate CEA. One-cycle,
Smith-degraded CEA, in which all L-fucose, sialic acid, 70% of the
D-galaCtOSe, and 15-30% of the D-mannose had been destroyed,
showed strong reactivity with H . pomatia lectin, but the reactivity with
R. communis and con A was abolished. After two cycles of Smith
degradation, the H . pomatia reactivity was lost. Wheat-germ-agglu-
tinin-reactivity was unimpaired throughout four cycles of Smith
degradation (see Fig. 10).The authors postulated413that most of the
wheat-germ-agglutinin-reactive 2-acetamido-2-deoxy-~-glucoseis sit-
uated in the interior of the carbohydrate chain, in N-glycosylic linkage
to asparagine.
Two sublines of the TA3 mouse mammary carcinoma, St and Ha,
were studied by Codington and coworkers762with respect to their
reactivity with the Vicia graminea lectin. The Ha subline, which grows
in allogeneic, as well as syngeneic, mice, adsorbed 100 to 400 times as
much V. graminea lectin as subline St, which grows only in syngeneic

(869) Y. S. Kim, R. Isaacs, and J. M. Perdomo, Proc. Natl. Acad. Sci. U S A . , 71,4869-
4873 (1974).
(870) P. Gold and S. 0.Freedman,J. E x p . Med., 121, 439-462 (1965).
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 331

mice. Furthermore, proteolysis of Ha cells lowered the lectin

adsorption by 90%, and released glycopeptides that completely
inhibited V. gruminea-N erythrocyte agglutination at levels of 5-10
puglml. Proteolysis of subline St cells did not release lectin-reactive
material. Carbohydrate analysis of the papain-solubilized
glycopeptides revealed the presence of D-galactose,
2-acetamido-2-deoxy-~-galactose, 2-acetamido-2-deoxy-~-glucose,
and sialic
In summary, many investigations of lectin-reactive, membrane
glycoproteins and glycopeptides have been initiated. The molecules
have been obtained both by proteolysis of whole cells and by detergent
extraction of plasma membranes. In many instances, affinity
chromatography on lectin-Sepharose adsorbents has proved extremely
useful in separating and purifying lectin-reactive glycoproteins.
Unfortunately, analyses of purified material have most often included
compositional, but not structural, characterization (with the exception
of the work of Osawa and c o ~ o r k e r s , ' ~Koriifeld~ , ~ ~ ~ and
and PospiBilovii and coworker^'^^) and
colleagues,216*434*435*666,730*731,747

. .
H(0)-Determmant

~ - ( 1 - 3 ) - ~ - ~ - G l c N A ~ p - (- l- --- - - - - - - - Ser,Thr.

\ J
I

FIG. 14.-Hypothetical, Composite "Megalo-oligosaccharide" Structure for Blood-

group Substances, Showing the Carbohydrate-binding Loci for Various Lectins. [a,
Dolichos bijlorus; b, Phaseolus lunatus; c, Helix pomatia; d, Glycine max; e ,
Bandeiraea simplic$olia I (A4); f, Bandeiraea simplicifolia I (BJ; g, Abrus preca-
torius; h, Sophora japonica; i, Triticum vulgaris; j, Cytisus sessilifolius; k, Canuvalia
ensiformis; 1, Bandeiraea simplicifolia 11; m, Ricinus communis; n, Ulex europeus;
0, Arachis hypogaea; p, Lotus tetragonolobus; q, Anguilla anguilla; and r, Vicia
graminea.]
332 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

must therefore be considered incomplete. It will be both interesting

and valuable to compare lectin-reactive, membrane glycoproteins from
normal cells and transformed cells, erythrocytes and lymphocytes,
platelets, and other sources when information regarding their
structures becomes available.
Figs. 14 to 16 present the structures of a series of naturally occurring
glycopeptides, and summarize the carbohydrate-binding specificity of
a number of lectins in terms of the carbohydrate units with which they
interact.

+ 4,6) - P - D - Manp
1
I
4
p - D - Glc NAcp
1
I b, f , m,n
4
p-D-GlcNAcp
1
1
N
Asn
FIG. 15.-Human IgA, Glycopeptides7' Showing the Carbohydrate-binding Loci for
Various Lectins. [a. Limulus polyphemus; b, Triticum vulgaris; c, Ricinus communis;
d, Sophora japonica; e, Abrus precatorius; f, Cytisus sessilifolircs; g , Phuseolus
vulgaris; h, Canavalia ensiformis; i, Lens culinaris; j, Pisum sativus; k, Vicia faba;
1, Bandeiraea simplicifolia 11; m, Solanum tuberosum; and n, Ulex europeus 11.1

(871) J. Baeiiziger and S. Kornfeld,J. Biol. Chem., 249, 7260-7269 (1974).

 LECTINS: CARBOHYDRATE-BINDING PROTEINS 333

I
1

-
8- D- XJrlp (1---t
4 1
6)-p -D- GlcNACp

h, i , j , k

i
N
Asn
FIG. 16.-Glycopeptide from Pineapple-stem Bromelins7z~873 Showing the Carbohy-
drate-binding Loci for Various Lectins. [a, Canavalia ensijormis; b, Lens culinaris;
c, Pisum sativum; d, Vicia faba; e, Lotus tetragonolobus; f, Anguilla anguilla; g, Ulex
europeus I; h, Triticum vulgaris; i, Solanum tuberosum; j, Cytisus sessilifolius; and
k, Ulez europeus 11.1

A lectin from the seeds of Duturu strumonium L. Cjimson weed) was

isolated by affinity chromatography on the insoluble polysaccharides
from Aspergillus r ~ i g e r . ~A’ ~glycoprotein (28% neutral sugar, pre-
ponderantly arabinose), the lectin is blood-group ABO-nonspecific,
contains large proportions of cystine and glycine, and 6.3%of hydroxy-
proline, and has a molecular weight of 120,000. Chito-oligosaccharides
bind to the Daturu l e ~ t i n . ~ ~ , ~ ~ ~
z ~ ~ ~ the Muclura pomifera seed-lectin (see
Bausch and P ~ r e t purified
Section V,4) to homogeneity on insolubilized polyleucyl hog-gastric
mucin. The lectin, a glycoprotein containing 0.8% of neutral sugar,
has a molecular weight of 40,000, and is “composed of two pairs of
dissimilar polypeptide chains stabilized by noncovalent interactions”
(Ref. 875).Relatively rich in acidic and hydroxy amino acids, the lectin
contains a very small proportion of methionine, but is devoid of cys-
teine. Studies on sugar-inhibition of hemaggl~itination~~~ indicated
that the Macluru lectin has a high specificity for a-D-galactopyranosyl
end-groups and for 2-acetamido-2-deoxy-~-galactose.
(872) Y. Yasuda, N. Takahashi, and T. Murachi, Biochemistry, 9, 25-32 (1970).
(873) Y . 4 . Lee and J. R. Scocca,J. Biol. Chem., 247,5753-5758 (1972).
(874) V. HoiejGi and J. Kocourek, Biochim. Biophys. Acta, 532, 92-97 (1978).
(875) J. N. Bausch and R. D. Poretz, Biochemistry, 16,5790-5794 (1978).
334 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

IX.TABLES
TABLEXXIII
Physical and Chemical Properties of Purified Lectins"

Subunit structure

Latin name
(common name) Molecular weight Molecular weight Number

Abrus precatorius hemag- 132,000(Ref. 611)

glutinin (jequirity bean) 134,000(Ref. 147) 36,000,35,000,and
33,000
125,000-132,000 33,800and 32,200 2 1 4
(Ref. 610) (Ref. 610)
toxin 63,800(Ref. 607) 33,000 and 28,000

(abrin)
(Ref. 607)
65,000(Ref. 150) 35,000and 30,000
(Refs. 147,150)
1 l2
Anguilla anguilla (eel) 123,000 10,000 12
Aruchis hypogaea (peanut) 110,000 27,000-28,000 4
106,500
Bandeiraea simplicifolia
I 114,000 28,500 4
(5isolectins: &, A3B,
AdL BJ
I1 113,000 30,000 4
Canaualia ensiformis (jack 104,000(pH 7) 26,000 4 (PH 7)
bean) 52,000 (pH 5) 2 (PH 5)
Cytisus sessilifoliolizrs 110,000 n.d.
Dolichos bijlorus (horse
gram)
A 113,000 26,500 4
B 109,000 26,000 4
Glycine max (soybean) 120,000-122,000 30,000 4
Helix pomatia (edible snail) 79,000 13,000 6
Lens culinaris (lentil) 42,000-63,000 18,000
8,000 2 1 4
Limulus polyphemus 335,000 19,000 18-20
(horse-shoe crab) 400,000 20,000
Lotus tetragonolobus
(asparaguspea)
A 120,000 27,800-28,000 4
B 58,000 27,800-29,000 2
C 117,000 27,000-30,000 4
Phaseolus lunatus limensis
(lima bean)
I1 269,000,247,000 31,000 8
I11 138,000,124,000 31,000 4
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 335

Amino acid
Glycopmtein composition’’

Carbohydrate
(per cent by Methio-
weight) Sugars Cysteine nine Metals References

Man (Ref. 610)

8 Man, Glc, GlcNAc 21.6 18.3 147,611,651

5 Man, Glc, GlcNAc n.d. 610,651

3 Man, Glc, GlcNAc 4.2 6.7 147,150,607

Man, Glc, GlcNAc 10 11-13 612,651

0.39 GlcNAc n.d. + n.d. 679

- - 3 n.d. 202
16.6 0 n.d. 616

6.7 Man, Fuc, Xyl, 1 0.4 Ca2+,Mgz+ 131

GlcNAc

4.0 Man, Fuc, GlcNAc 1 3 Ca2+,Mg2+ 125

- - - 2 Ca2+,Mn2+ 136,258,261,
262,262a,265,
267,271,324
n.d. n.d. n.d. 149,476

3.8-4.7 Man, GlcNAc - 4 n.d. 108,519-521

1.3 Man, GlcNAc - 4 n.d. 108,520,521
7 Man, GlcNAc - 1 Ca2+,MnZ+ 151,538,544,546
8 Gal, Man 3 2 - 561,569,570
1.5-3 Glc, GlcNAc 2 - (4) Ca2+,Mn2+ 138,141-143,
441,442,444
>4 neutral sugars, 3 2-3 n.d. 159,769-771
GlcNAc

9.4 Gal, GlcNAc 0 trace n.d. 200,683,684

4.8 Gal, GlcNAc 0 trace n.d. 200,683,684
9.2 Gal, GlcNAc 0 trace n.d. 200,683,684

3-5 Man, GlcNAc, Fuc 2 0 Ca2+,Mn2+

151,199,586,588
3-5 Man, GlcNAc, Fuc 2 0 Ca2+,Mn2+
10-11 Man, GlcNAc 0 0 Ca2+,MnZ+ 630,697,698,
705-716,718
(Continued)
336 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXXIII (Continued)

Subunit structure

Latin name Molecular

(common name) Molecular weight weight Number

Phaseolus vulgaris (red 126,000-136,000 29,000-34,000 4

kidney-bean) (5 isolectins: L,,
L& LR,, L% R J
Pisum sativum (pea) 49,000-53,000 CY,7,000-10,500
p, 12,000-18,000 2 1 4
Ricinus comniunis hemag- 120,000 29,500 and 37,000
glutinin (castor bean) (Ref. 146)
31,000 and 34,000 4
(Ref. 147)
27,500 and 30,000
(Refs. 648, 651)
toxin (ricin) 60,000 28,000 and 32,000 2
(Refs. 146, 147, 150,
194, 648, 649)
Solanum tuberosum 100,000 46,000 2
(potato)
Sophora japonica (pagoda 132,800 n.d.
tree)
Triticum vulgaris (wheat 36,000 18,000 2
germ)
Ulex europeus (gorse, h n e ,
whin seed)
I 43,000, 45,000 n.d.
(Ref. 158)
31,000, 32,000
(Ref. 691)
40,000-65,000
(Ref. 690)
170,000 (Ref. 226)
I1 n.d. n.d.

Vicia faba (horse, broad, or 47,500-53,000 24,000 2 (Ref.

fava bean) 213)
18,000
9,000 R
17,300
13,300 2
Vicia graminea 105,000 25,000 4
 LECTINS : CARBOHYDRATE-BINDING PROTEINS 337

Amino acid
Glycoprotein compositionL

Carbohydrate
(per cent by Methio-
weight) Sugars Cysteine nine Metals References

trace (<0.5) Glc, Xyl 0 0 Ca2+,MnZ+ 141,453-455

4.7 Man, Glc, GlcNAc 1.69, 1.05, 146,147,648,

1.77 1.23 650,651

5.5-6 Man, Glc, GlcNAc 9.81 0.78 146,147,150,

1.32, 0.71, 194,648-650
1.82 0.95
50 Ara, Gal, Glc, 25 1 n.d. 207
GlcNAc
7.8 Man, Xyl, GlcNAc 5 0 n.d. 184,598

0 17-18 2 n.d. 128-130,335,

495

5.2-7.2 Man, Fuc, Glc, Xyl, 2 2-3 Ca2+,Mn2+, 158,196,225,

Ara, Glc, GlcNAc Zn2+ 226,509,690,
691

21.7 Man, Gal, Ara, + + n.d. 194,208,225,

GlcNAc 226,509
3 GlcNAc, Man 0 0 n.d. 140,2 13,467a

7.3 Man, Glc, Fuc, Gal, 8 8 n.d. 220

GlcNAc

“+,present; -,absent; n.d., notdetermined. *Molesofamino acid permole ofsubunit.

 0
0
00

TABLEXXIV

Latin name
Blood-group Specificity and Carbohydrate-binding Specificity of Purified Lectins

Blood-group specificity References Carbohydrate specificity References

1z
Y
Abrus precatorius 147,603 P-D-Galp > a-DGalp 147 Q
Anguilla anguilla 21,64 (Y-L-FUCP 21,22 0
65,670 3-0-Me-DFucp 167,471 r
U
3-O-Me-~Galp 672,674,676 M
4
Arachis hypogaea neuraminidase-digested 202,710 P - ~ G a l p -1+3)-~-GalNAcp
( 20 1,617-6 19
A, B, 0, or T antigen z
Bandeiraea simplicifolia
BS I B >> A, 131,195,622,
625
a - ~ - G a l p> a-DGalNAcp 20,131,195,
626
5
BS I1
Canaoalia ensiformis
Tk
nonspecific
125,470
102,246,247
P-D-GICNACP= a-DGlcNAc
a-DManp > a - ~ - G l c p> a-DGlcNAcp
125
168-170,
5r
204,215,365
r
M
Cytisus sessilifolius 0 > A2 > A, 6,77,78,472, P-wGlcNAcp-( 1+4)-P-~GlcNAc > cellobiose 19,20,471, M
475,477
z
473,476
M
Dolichos bijlorus Al >> As 108,510-512 a-DGalNAcp 108
Glycine mar
Helix pomatia
A>O>B
A
212
60,63,100,
a-D-GalNAcp > P-DGalNAcp
a-DGalNAcp >> a-DGlcNAcp
212,552
61,63,100, 5*
Lens culinaris
189
138,142 a-DManp > a-DGlcp, a-DGlcNAcp
178,561,562
109,123,138,
!2
nonspecific
143,213
Lotus tetragonolobus 0 >> A2 6,22,77,78 ~Y-L-Fuc~,
2-O- Me-nFucp 22,77,78,
167,672,674
Maclura pomifera nonspecific 631,632 D-Galp, DGalNAcp 632,635-637
Phaseolus lunatus syn. A, > A2 >> B 2,5,22,199 a-DGalNAcp > a-DGalp 22,151,199,
limensis 591
Phaseolus vulgaris nonspecific 696 730,731
Pisum sativum nonspecific 141,448 a-D-Manp > a - ~ G l c p> a-D-GlcNAc 122,211,213,
428,440,448,
449,452
Ricinus communis nonspecific 147,603,641, P-DGalp > a - ~ G a l p 124,146,147, 0
642 183,194 3
Solantcni tuberosum nonspecific 77,480,482 P-D-GlcNAcp-( 1+4)-M-D-GlcNAcp-(1+4)&-
P-DGlcNAc
206,207,483 ..5
Sophora japonica A>B>>O 22,78,184, P-D-GalNAcp > P - ~ G a l p> a-D-Galp 175,183
n
I-antigen 593,599,601 184
Triticum vulgaris nonspecific 501 P-D-GlcNAcp-( 1+4)-P-D-GIcNAcp-( 1+4)-P-D- 30,128,498 0
GlcNAcp > P-DGlcNAcp-(1+4)-GlcNAc
Vicia faba nonspecific 5,462 a-DManp > a-~-GlcNAcp 140,213,468,
469
Vicia graminea N )" 750,752-758
Ulex europeus
Ulex I 0 >> A2 12-14,77,81, CY-L-FLIC~
225
Ulex I1 0 >> A2 12-14,77,81, P-DGlcNAcp-(1+4)-P-~-GlcNAc
225

n P ~ G a l p 1+4)-P~GlcNAcp-(
-( 1+2)-a-~Manp

P ~ - G a l p -1+4)-P-~-GlcNAcp-(
( 1+2)-w~-Manp

0
0
(D
340 IRWIN J. GOLDSTEIN AND COLLEEN E. HAYES

TABLEXXV
Lectin-Carbohydrate Association as Studied by Equilibrium Dialysis

Carbohydrate & ("C) Refer-

Latin source ligand (M-') na ences

Abrus precotorius lactose 8x lo3(----)* 2/4 147

(jequirity bean)
Bandeiraea simplici- methyl a-mgalacto- 8.6 x lo4 (2") 4/4 626
foliu I pyranoside 3.3 x 104 (207
Cunavalia ensiforrnis methyl a-D-manno- 2.06 x 104 (20) 44 172,278
(concanavalin A from pyranoside
jack bean)
Glycine mas (soybean) 2-acetamido-2-deoxy-~- 3.0 x lo4 (4") 2/4 544
galactose
Helix pomatia (snail) A substance pentasac- 5x lo3 (25") 6/6 570
charide
Lens culinnris (lentil) D-mannose 2.3 x lo2 (4") 2/2 186
methyl a-mgluco- 1.0 x 102 (4") 2/2
pyranosicle
Lotus tetragonolobus
(asparagus pea)
A L-fucose 1.2 x 104 (37 44 683
B 0.6 x 104 (37 2/2
C 3.7 x 104 (3") 414
Phaseolus lunatus
(lima bean)
I11 methyl 2-acetamido-2- 1.01 x 103 (20) 2/4 589
I1 deoxy-a-D-galacto- 0.93 x lo3 (2') 4/8
pyranoside
Ricinus conzniunis lactose 1.5 x 104(----)b 2/4 147
(castor bean) p-nitrophenyl p-D- 1.65 X lo4 (25") 2/4 124
galactopyranoside
Pisum satiuum (pea) D-mannose 1.4 X lo2 (4") 214 453
methyl C X - D - ~ ~ L W O - 8 x lo2 (4") 2/4
pyranoside
Triticum vulguris 2-acetamido-2-deoxy-~- 1.3 x 10" (4") 4/2 129
(wheat-germ glucose
agglutinin) N,N',N",N"'-tetraacetyl- 5.3 x 104 (20") 4/2 115
chitotetraitol
Wistaria floribundn 2-acetamido-2-deoxy-~- 1.28 x 104 (----)I' 2/2 796
galactose

"Binding sites per subunit = n. !'(----), temperature not reported.

 BIOCHEMISTRY OF PLANT GALACTOMANNANS

BY PRAKASH
M. DEY*

Department of Biochemistry, Royal Holloway College, University of London,

Egharn Hill, Egham, Surrey, TW20 OEX, England

I. Introduction ... .. ... ........ . . .. . . .. . . .. . .............. . .. .. ..... ... .. . 341

1. Occurrence .... .. .................. .......... .... ... .. .... .... ...... 343
2. Location in uiuo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , , , . . . 345
3. Isolation . . . .... .. ...... ... .. .... ... .. .. . . . . , .. . .... . .......,. .. ... .. 345
4. Structure.. . . . . . . . . . . . . . . . . . . . . . . . . . , . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
11. Biosynthesis ......... .. .. .... ........... ..... . ... .... .. ...... ... .. ..... 352
111. Biochemical Degradation . . . . . . . . . . . . . . . . . . . , . . , . . . . . . . . . . . . . . . . . . . . . . . . 356
1. General Considerations . . . . . . , , . . , , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . , . 356
2. Enzymes Involved.. . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . .. 361
IV. Function ............................................................... 375

I. INTRODUCTION
Plant galactomannans are reserve polysaccharides composed of
linear chains of (1*4)-linked P-D-mannopyranosyl residues having
single stubs of a-D-galactopyranosyl groups joined by (1+6)-linkages
along the chain. These polysaccharides are also known as gums; their
use as substances for mummification can be traced back to 3000 B.C. in
ancient Egypt, and, hence, they are humorously called “Pharaoh‘s”
polysaccharides. The importance of these polysaccharides can be seen
in their wide use in industry, notably, food,*-” pharmaceutical^,'^-^^
*The author is indebted to Professors J. B. Pridham and J. E. Courtois, and Dr.
D. R. Davies, for their valuable suggestions.
(1) D. A. Rees, Biochem. J., 126,257-273 (1972).
(2) M. Glicksman and E. H. Farkas, Fr. Pat. 2,119,365 (1972);Chern. Abstr., 78,96,287
(1973).
(3) M. H. Yueh and E. D. Schilling, Gel. Pat. 2,104,743 (1971);Chem. Abstr., 76,35,4339
(1972).
(4) P. Kovacs, Food Technol. (Chicago),27, 26-30 (1973).
(5) H. R. Schuppner, Can. Pat. 824,635 (1969); Dairy Sci. Abstr., 32, 1488 (1970).
(6) A. J. Leo and E. Bielskis, U.S. Pat. 3,396,039 (1968); Chem. Abstr., 69, 66,286
(1968).

34 1
342 PRAKASH M. DEY

cosmetics,20--21
paper p r o d ~ c t s , ~paints
~ - ~ ~and plasters,26-28well-
drilling and mining,2e-33and explosives and fire-fighting.34-37

(7) B. Weinstein, U.S. Pat. 2,856,289 (1958);Chem. Abstr., 53, 1583e (1959).
(8) L. L. Little, U.S. Pat. 3,370,955 (1968);Chem. Abstr., 68, 94,742 (1968).
(9) H. Burton, H. R. Chapman, and D. J. Jayne-Williams, Proc. Int. Dairy Congr.
16th, Copenhagen, 3,82 (1962).
(10) M. Glicksman, “Gum Technology in the Food Industry,” Academic Press, New
York, 1969, p. 130.
(11)W. A. Carlson, E. M. Ziegenfuss, and J. D. Overton, Food Technol. (Chicago),
16,5044 (1962).
(12) E. Nuemberg and E. Rettig, Pharm. Ind., 36, 194-198 (1974).
(13) E. Nuernberg and E. Rettig, Drugs Made Ger., 17,26-28,28-31 (1974).
(14) E. Nuernberg, Ger. Pat. 1,290,661 (1969);Chem. Abstr., 70, 118,096 (1969).
(15) Laboratories Dausse S. A,, Fr. Pharm. Pat. M. 7794 (1970);Chem Abstr., 76,131,509
(1972).
(16) E. Nuernberg, E. Rettig, and H. Mueller, Ger. Pat. 2,130,545 (1972); Chem.
Abstr., 78, 62,171~(1973).
(17) E. Nuernberg, H. Mueller, H. Nowak, and P. Luecker, Ger. Pat. 2,017,495
(1971);Chem. Abstr., 76, 17,808 (1972).
(18) Synthelabo S. A., Fr. Pat. 2,073,254 (1971);Chem. Abstr., 77,39,247 (1972).
(19) E. Merck A. G., Neth. Pat. 6,504,974 (1965);Chem. Abstr., 64, 14,042g (1966).
(20) R. J. Chudzikowski,J. Soc. Cosmet. Chem., 22,43-60 (1971).
(21) J. I. Gonzales, Fr. Pat. 2,067,649 (1971);Chem. Abstr., 78,20,125 (1973).
(22) R. Nordgren, U.S. Pat. 3,225,028 (1965);Chem. Abstr., 64, 6891c (1966).
(23) J. W. Opie and J. L. Keen, U.S. Pat. 3,228,928 (1966); Chem. Abstr., 64, 11,430~
(1966).
(24) D. J. Chrisp, U.S. Pat. 3,301,723 (1967);Chem. Abstr., 67, 92,485~(1967).
(25) M. H. Yueh and E. D. Schilling, Fr. Pat. 2,080,462 (1971); Chem. Abstr., 77,
103,610 (1972).
(26) J. Fath and M. Rosen, U.S. Pat. 3,700,612 (1972); Chem. Abstr., 78, 59,856
(1973).
(27) D. J. Pettitt, U.S. Pat. 3,658,734 (1972); Chem. Abstr., 77, 35,846 (1972).
(28) G. Benz, Ger. Pat. 1,206,777 (1965); Chem. Abstr., 64, 629Oc (1966).
(29) R. E. Walker, U.S. Pat. 3,215,634 (1965);Chem. Abstr., 64, 382Oc (1966).
(30) N. H. Black and L. L. Melton, U.S. Pat. 3,227,212 (1966); Chem. Abstr., 64,
7941b (1966).
(31) V. V. Horner and R. E. Walker, U.S. Pat. 3,208,524 (1965); Chem. Abstr., 64,
503b (1966).
(32) W. C. Browning, A. C. Perricone, and K. A. Eking, U.S. Pat. 3,677,961 (1972);
Chem. Abstr., 77, 128,458 (1972).
(33) F. B. b o p , S. Afr. Pat. 69,06946 (1970);Chem. Abstr., 73, 111,681 (1970).
(34) J. J. Yancik, R. E. Schulze, and P. H. Rydlund, U.S. Pat. 3,640,784 (1972);
Chem. Abstr., 76, 101,828 (1972).
(35) P. R. Goffart, Fr. Pat. 1,533,471(1968); Chem. Abstr., 71, 23,404 (1969).
(36) E. I. du Pont de Nemours and Co., Inc., Fr. Pat. 1,537,625(1968); Chem. Abstr.,
72,45,652 (1970).
(37) W. W. Morgenthaler, US. Pat. 3,634,234 (1972); Chem. Abstr., 76, 143,009
(1972).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 343

Earlier reviews on g a l a c t o m a n n a n ~ ~mainly

~ - ~ ~ described their
general chemistry. Dea and Morrison43discussed the latest inves-
tigations on plant galactomannans with respect to their intermolecular
interactions and their interactions with other polysaccharides.

1. Occurrence
The distribution of galactomannans in the plant kingdom is limited;
the rich sources are the members of the family Leguminoseae. They
have also been found in the species of Annonaceae (custard apple
family),44 Convolvulaceue (morning glory family),44*4s Ebenaceue
(ebony family),46Loganiaceae (Buddleia family),46and PaZmae (palm
Amongst the leguminous plants, galactomannans are
located in the endospemic part of the seeds; Dea and Morrison43
summarized the sources in which these polysaccharides have been
detected. Baileys7 suggested that legume-seed galactomannans may

(38) R. L. Whistler and C. L. Smart, “Polysaccharide Chemistry,” Academic Press,

New York, 1953, p. 291.
(39) B. N. Stepanenko, Bull. Soc. Chim. Biol., 42, 1519-1536 (1960).
(40) F. Smith and R. Montgomery, “Chemistry of Plant Gums and Mucilages,”
Reinhold, New York, 1959.
(41) P. A. J. G r i n and J. F. T. Spencer, Adu. Carbohydr. Chem., 23,367-417 (1968).
(42) See “Carbohydrate Chemistry,” R. D. Guthrie, ed., Specialist Report, The
Chemical Society, London, 1970, Vol. 3, p. 236
(43) I. C. M. Dea and A. Morrison, Ado. Carbohydr. Chem. Biochem., 31, 241-312
(1975).
(44) P. Kooiman, Carbohydr. Res., 20,329-337 (1971).
(45) S. N. Khanna and P. C. Gupta, Phytochemistry, 6,605-609 (1967).
(46) P. Kooiman and D. R. Kreger, K. Ned. Akad. Wet. Proc. Ser. C, 63, 634-645
(1960).
(a7) G. 0. Aspinall, E. L. Hirst, E. G . V. Percival, and J. R. Williamson, J . Chem.
SOC.,3184-3188 (1953).
(48) F. Klages, Ann., 509, 159-181 (1934).
(49) F. Klages, Ann., 512, 185-194 (1935).
(50) A. K. Mukherjee, D. Choudhury, and P. Bagchi, Can. J . Chem., 39, 1408-1418
(1961).
(51) V. Subrahamnyan, G. Bains, C. Natarajan, and D. Bhatia,Arch. Biochem. Biophys.,
60,27-34 (1956).
(52) C. V. N. Rao and A. K. Mukherjee,J. Indian Chem. Soc., 39,711-716 (1962).
(53) C. V. N. Rao, D. Choudhury, and P. Bagchi, Can. J . Chem., 39,375-381 (1961).
(54) H. S. El Khadem and M. A. E. Sallam, Carbohydr. Res., 4,387-391 (1967).
(55) V. K. Jindal and S. Mukherjee, Cum. Sci., 38,459-460 (1969).
(56) V. K. Jindal and S. Mukherjee, IndianJ. Chem., 8,417-419 (1970).
(57) R. W. Bailey, in “Chemotaxonomy of the Leguminoseae,” J. B. Harborne, D.
Boulter, and B. L. Turner, eds., Academic Press, London, 1971, p. 503.
344 PRAKASH M. DEY

have taxonomic importance, and put forward two main reasons for this:
firstly, that different species have various levels of the polymer, and
secondly, that the ratio of D-mannose to D-galaCtOSe (expressed as M/G
in this text) also varies. However, the yield of galactomannans may vary
with the method of extraction. It has been generally experienced that
the solubility of the polymer in water increases as the ratio M/G de-
creases. The efficiency of aqueous extraction therefore depends upon
the in vivo composition of the galactomannan. The total time involved
in the extraction procedure may also alter the yield, because of enzymic
degradation (for example, by a-D-galactosidases), which may alter the
composition of the polysaccharide and hence affect its solubility.
a-D-Galactosidases are generally present in leguminous seeds.s8
Other galactomannan-degrading enzymes may have similar effects on
the solubility of the polysaccharide. Therefore, the taxonomic impor-
tance of these polysaccharides should be considered with caution.
The levels of galactomannan are very high (15-38% of the
dry weight of seed) in some members of the sub-families Caesal-
- ~ *Indiofereae.59,61m Some members of the sub-
p i n i ~ i d e a e ~ ~and
families Astragaleae,6' C r ~ t a l a r i e a e , ~ ~M* i~r' n- ~o ~s e ~ e , ~ ~ * ~ ~ *
~ ~ T r i f ~ l e a e , have
S o p h o r e ~ e ,and ~ ~ * ~15-25%
~ of the polysaccha-
ride, whereas those of De~rnoieae,~',~' Genisteae,72*73 G l y ~ i n e a e , ~ ~
and Loteae6'*7sa76 have only 1-15%. On the other hand, the seeds
of several kinds of legumes do not contain any g a l a c t ~ m a n n a n . ~ ~
(58) D. Barham, P. M. Dey, D. Griffiths, and J. B. Pridham, Phytochemistry, 10,
1759-1763 (1971).
(59) E. Anderson, Ind. Eng. Chem., 41,2887-2890 (1949).
(60) J. Y. Morimoto, I. C. J. Unrau, and A. M. Unrau, J . Agric. Food Chem., 10,
134-137 (1962).
(61) H. L. Tookey, R. L. Lohmar, I. A. Wolff, andQ. JonesJ. Agric. FoodChem., 10,131-
133 (1962).
(62) H. L. Tookey, V. F. Pfeiffer, and C. R. Martin, J. Agric. Food Chem., 11,
317-321 (1963).
(63) J. Y. Morimoto and A. M. Unrau, Hawaii F a r m Sci., 11, 6-8 (1962).
(64) S. A. I. Rizvi, P. C. Gupta, and R. K. Kaul, Planta Me& 20,24-32 (1971).
(65) A. M. Unrau and Y. M. Choy, Carbohydr. Res., 14, 151-158 (1970).
(66) V. P. Kapoor, Phytochemistry, 11, 1129-1132 (1972).
(67) A. S. Cerezo,J. Org. Chem., 30,924-927 (1965).
(68) C. Leschziner and A. S. Cerezo, Carbohydr. Res., 15,291-299 (1970).
(69) A. M. Unrau,J. Org. Chem., 26,3097-3101 (1961).
(70) J. S. G. Reid and H. Meir, Z. Pfanaenphysiol,, 62,89-92 (1970).
(71) M. P. Sinha and R. D. Tiwari, Phytochemistry, 9, 1881-1883 (1970).
(72) J . E. Courtois and P. Le Dizet, Bull. Soc. Chim. Biol., 45, 731-741 (1963).
(73) Z. F. Ahmed and A. M. Rizk,J. Chem. U.A.R., 6,217-226 (1963).
(74) R. L. Whistler and J. Saamio,J. Am. Chem. Soc., 79, 6055-6057 (1957).
(75) R. Somme, Acta Chem. Scand., 20, 589-590 (1966).
(76) R. Somme, Acta Chem. Scand., 21,685-690 (1967).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 345

In addition to the seed galactomannans, the leaf and stem tis-

sues of red clover (Trifolium pratense) have been shown to contain
a galactoglucomannan; this has a main chain of (1+4)-linked
p-D-glucopyranosyl and P-D-mannopyranosyl residues, to which are
attached single stubs of (1+6)-linked a-D-galactopyranosyl g r o ~ p s . ~ ~ , ’ ~
Such polysaccharides have also been isolated from the stem of the trop-
ical legume Stylosanthes h ~ m i l i sand
, ~ ~from some softwoods.80*81

2. Location in vivo
In general, galactomannans do not exist together with the
starch granules, but are present in seeds that are rich in oligosac-
charides of the &nose family. In such seeds, they are usually
located in the endosperm tissues (which lie outside the embryo
and are surrounded by the testa). They serve as reserve carbohy-
d r a t e ~ . By
~ ~light-microscopy,
, ~ ~ ~ ~ ~ Reid and Meiers3 examined the
location of galactomannan in fenugreek seeds (Trigonella
foenum-graecum); Fig. 1 shows that the endosperm cells are
completely filled with galactomannan. There are two known excep-
tions to the finding of a galactomannan in the endosperm: (a) Gymno-
cladus dioica,@ in which the polysaccharide lies in the inner side
of the seed coat, and (b) Glycine max,8sin which it occurs in the hull.

3. Isolation
Isolation43 involves pulverizing the seeds into a flour which is
then extracted with cold or hot water,65s76*8688 alkali,89 or dilute
The solubility of galactomannans has been discussed in
detail by Dea and Morrison.43The galactomannan from the crude
extract may be purified by various methods, including the use of
(77) B. D. E. Gaillard and R. W. Bailey, Phytochemistry, 7,2037-2044 (1968).
(78) A. J. Buchala and H. Meier, Carbohydr. Res., 31,87-92 (1973).
(79) M. Alam and G. N. Richards, Aust. J . Chem., 24,2411-2416 (1971).
(80) T. E. Timell, Ado. Carbohydr. Chem., 19,247-302 (1964);20,409-483 (1965).
(81) J. E. Courtois, Bull. Soc. Bot. Fr., 115, 309-344 (1968).
(82) A. Tschirch, “Angwandte Pflanzenanatomie,” Urgan and Schwarzenburg,
Vienna, 1889.
(83) J. S. G. Reid and H. Meier, Caryologin, 25,219-222 (1973).
(84) E. B. Larson and F. Smith,J. Am. Chem. Soc., 77,429-432 (1955).
(85) G. 0. Aspinall and J. N. C. Whyte, J. Chem. Soc., 5058-5063 (1964).
(86) K. F. Horvei, and A. Wickstrflm, Acta Chem. Scand., 18,833-835 (1964).
(87) N. R. Krishnaswami, T. R. Seshadri, and B. R. Sanna, Cum. Sci., 35, 11 (1966).
(88) R. Somme, Acto Chern. Scand., 22,870-876 (1968).
(89) P. V. Subbarao and M. V. L. Rao, IndianJ. Chern., 3,361-363 (1965).
346 PRAKASH M. DEY

FIG. 1.-A Cross-section of the Outer Part of a Seed of Trigonella foenum-graecum

Before Mobilization of the Galactomannan, Showing the Three-layered Seed-coat (S)
and a Small Part of the Cotyledon (C), with the Endosperm in Between. [The aleurone
layer (A) is the outer cell-layer of the endosperm, and the rest of the endosperm is com-
posed of large cells that have thin, primary walls and are completely filled with the
dark-stained galactomannan (G). Stained with the periodic acid-Schiff reagent; x300
(reproduced, by permission, from Ref. 199).]

primary alcohols,so copper ~ o m p l e ~ barium , ~ hy-

~ ~ ~ ~
d r o ~ i d e , a~ c~e*t ~y ~l a t i ~ n , and
~ ~ *acetylpyridinium
~~~~ bromide com-
p l e ~ The
. ~ ~ method of copper complexing, however, has been
shown to cause chain cleavageag7
As with most polysaccharides, galactomannans isolated chem-
ically may be heterogeneous and polydisperse with respect to both

(90) C. M. Rafique and F. Smith,]. Am. Chem. SOC., 72,4634-4636 (1950).

(91) A. M. Unrau and Y. M. Choy, Can.]. Chem., 48,1123-1128 (1970).
(92) P. Andrews, L. Hough, and J. K. N. Jones,/. Am. Chem. Soc., 74, 4029-4032
(1952).
(93) P. Andrews, L. Hough, and J. K. N. Jones,]. Chem. SOC., 2744-2750 (1952).
(94) A. J. Erskine and J. K. N. Jones, Can.J . Chem., 34,821-826 (1956).
(95) H. Meier, Methods Carbohydr. Chem., 5,45-46 (1965).
(96) R. G. Morley, Ph.D. Thesis, University of Salford (1972).
(97) N. Sugiyama, H. Shimahara, T. Andoh, M. Takemoto, and T. Kamata, Agric.
B i d . Chem., 36, 1381-1387 (1972).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 347

molecular weight and composition. In such cases, the yield and

nature of the polysaccharide depend upon the method of extraction
employed.8s

4. Structure
The fundamental structure of the plant galactomannan is the
following.
[P-D- Manp-( 1+4)-In-P-D- M anp-( 1 4 ) -
6
t1
a-D-Galp
This structure has been elucidated by extensive work over the past
four decades. The main techniques used were r n e t h y l a t i ~ n , ~ ~par- *~~-'~~
tial h y d r ~ l y s i s , ' ~ ~periodate
-'~~ o x i d a t i ~ n ,and
~ ~ Jspecific,
~ enzymic hy-
d r ~ l y s i s . ' ~ ~A- "more
~ detailed account of the structure is given in a re-
view43(see also, Refs. 46 and 113-115). The frequency ofsubstitution by
D-gdactosyl groups along the main chain of the D-mannan varies ac-
cording to the source of the p o l y ~ a c c h a r i d e ~ ~some * ' ~ ~examples
; are
shown in Table I.
It is evident that, in galactomannans having M/G ratios of 1.0, all
of the D-mannosyl residues carry a D-galaCtOSyl group; this is shown by
the inability of P-D-mannanases to hydrolyze the D-mannan backbone
(98)V. P. Kapoor, Indian J . Chem., 11, 13-16 (1973).
(99)F.Smith,J. Am. Chem. Soc., 70,3249-3253 (1948).
(100)E.L. Hirst and J. K. N. Jones,]. Chem. Soc., 1278-1282 (1948).
(101)D. S. Gupta and S. Mukherjee, Indian]. Chem., 11,505-506(1973).
(102)Z.F.Ahmed and R. L. Whistler,J. Am. Chem. Soc., 72,2524-2525 (1950).
(103)D. S. Gupta and S. Mukherjee, IndianJ. Chem., 13,1152-1154(1975).
(104)R. L. Whistler and J. Z. Stein,]. Am. Chem. Soc., 73,4187-4188 (1951).
(105)R. L. Whistler and D. F. Durso,]. Am. Chem. SOC.,73,4189-4190 (1951).
(106)0. E. Moe, S. E. Miller, and M. H. Iwen,]. Am. Chem. SOC., 69, 2621-2625
(1947).
(107)J. E.Courtois and P. Le Dizet, Carbohydr. Res., 3, 141-151 (1966).
(108)J. E.Courtois and F. Petek, Methods Enzymol., 28,565 (1966).
(109)P. Hui, Ph.D. Thesis, Juris, Zurich (1962).
(110)J. E. Courtois and P. Le Dizet, Bull. SOC. Chim. B i d , 52, 15-22 (1970).
(111) I. C. M. Dea, C. Hitchcock, S. Hall, and A. Morrison, unpublished results.
(112)J. E. Courtois and P. Le Dizet, Bull. SOC. Chim. B i d . , 50, 1695-1710 (1968).
(113)P. S. Kelkar and S. Mukherjee, Indian J . Chem., 9, 1085-1087 (1971).
(114)D.S. Gupta and S. Mukherjee, Indian ]. Chem., 11, 1134-1137 (1973).
(115)E. L.Richards, R. J. Beveridge, and M. R. Grimmett, Aust. J . Chem., 21, 2107-
2113 (1968).
(116)B. V. McCleary, N. K. Matheson, and D. M. Small, Phytochemistry, 15, 1111-
1117 (1976).
348 PRAKASH M. DEY

TABLEI
M/G Ratios" of Some Plant Galactomannans

Source M/Ga References

Medicago satiua L. 1.0- 1.25 70,92,117,118

(alfalfa, lucerne)
Trifolium repens L. 1.0-1.3 72,86
(white clover)
Trigonella foenum-graecum 1.1-1.2 70,93,119
(fenugreek)
Ceratonia siliqua 1.2-5.25 59,99,100,107,120,121
(carob, locust bean)
Cyamopsis tetragonoloba 1.3-7.0 59,61,62,120-122
bar)
COCOS nucifera 2.0 53
(coconut)
Gleditsia amorphoides 2.5 67
Cassia absus 3.0 98,123- 125
Cleditsia triacanthos 3.2-3.8 59,68,107
(honey locust)

"Ratio of D-mannopyran0se:D-galactopyranose.

(see Section 111,2,b).However, this does not hold if the galactomannan

(M/G = 1.O) possesses 6-O-a-D-ga~actopyranosy~-a-D-ga~actopyranosy~
groups instead of single D-galactosyl groups.
It has been suggested that galactomannans having an M/G ratio
of >1.0 have a regular distribution of D-galaCtOSyl groups along the
main chain of the D-mannan.126p127On the other hand, Courtois and Le
Dizet110,112demonstrated the formation of D-manno-oligosaccharides,
and a galactomannan having an M/G ratio of almost 1.0, following the
hydrolysis of the polysaccharide (M/G = 4)from GZeditsia ferox and
carob (M/G = 4) by a P-D-mannanase. This observation supports the
view that, in these polymers, there is an alternation of zones in which
(117) J. E. Courtois, C. Anagnostopoulos, and F. Petek, Bull. SOC. Chim. Biol., 40,
1277-1285 (1958).
(118) R. J. McCredie, Diss. Abstr., 19, 432 (1958).
(119) K. M. Daoud, Biochem. J . , 26,255-263 (1932).
(120) P. A. Hui and H. Neukom, Tappi, 47, 39-42 (1964).
(121) R. D. Jones and A. Morrison, unpublished results.
(122) E. Heyne and R. L. Whistler,J. Am. Chem. SOC.,70,2249-2252 (1958).
(123) V. P. Kapoor and S. Mukherjee, Curr. Sci., 38, 38 (1969).
(124) V. P. Kapoor and S. Mukherjee, Phytochemistry, 10,655-659 (1971).
(125) V. P. Kapoor and S. Mukherjee, IndianJ. Chem., 10,155-158 (1972).
(126) H. Deuel, R. Leunberger, and G. Huber, Helw. Chlm. Acta, 33,942-946 (1950).
(127) K. J. Palmer and M. J. Ballantyne, J . Am. Chem. SOC.,72, 736-741 (1950).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 349

each D-mannosyl residue carries a D-galaCtOSyl residue and zones

devoid of these residues. Dea and Morrison43favored the block (or
zone) structure, and discussed this aspect in relation to X-ray analysis
and the interaction of galactomannans with other polysaccharides.
To study the consecutive or the block structure of galactomannans,
C. W. Baker and W h i ~ t l e r ' ~applied ~ , ' ~ ~ a technique whose principle
was based on the fact that 6-deoxy-6-C-p-tolylsulfonylhexopyranosides
are susceptible to alkali-catalyzed, glycosidic hydrolysis. 130-132 They
followed the sequence of p - e l i m i n a t i ~ n ,m ' ~e~t h y l a t i ~ n ,acid
' ~ ~ hydrol-
y s i ~ ,r'e~d~~ c t i o n , and
' ~ ~ acetylation. This resulted in the formation of
1,5-di-0-acetyl-2,3,4,6-tetra-O-methyl-~-mannitol when the D-
mannosyl residues were substituted at 0-6 with a-D-galactosyl groups
and at 0 - 4 by a D-mannosyl group which was itself not substituted
at 0-6.But, when the latter D-mannosyl group was substituted at 0 - 6
with an a-D-galactosyl group, 1,4,5-tri-O-acetyl-2,3,6-tri-O-methyl-~-
mannitol was formed. By comparing the ratios of the two products, the
author^'^^^^^^ suggested that guar galactomannan (M/G = 2) has a reg-
ular structure, with alternate D-mannosyl residues substituted with
D-galaCtOSyl groups, whereas carob galactomannan (M/G = 4) has
blocks of 25 D-mannosyl residues which are substituted with D-galac-
tosyl groups. They further s ~ g g e s t e d ' ~that, ~ * 'assuming
~~ a molecular
weight of 210,000 for this galactomannan, the average length of each
unsubstituted block is 85 linearly ( 1+4)-linked, p-D-mannopyranosyl
residues. This unsubstituted chain-length is in agreement with the
requirement observed for the association of the polysaccharide with
carrageenan. 136*137
Lindberg and performed methylation analyses on the
products resulting when guar and carob galactomannans were oxidized
with periodate and then reduced with borohydride. The results indi-
cated that a simple, alternating structure for guaran, and a simple, block
structure for carob gum, are not possible. The authors suggested an

(128) C. W. Baker and R. L. Whistler, Carbohydr. Res., 45,237-243 (1975).

(129) C. W. Baker and R. L. Whistler, Methods Carbohydr. Chem., 7 , 152-156 (1976).
(130) B. Lindberg and H. Lundstrem, Actu Chem. Scand., 20,2423-2426 (1966).
(131) H. Bjorndal and B. W&ngstr@m,ActaChem. Scand., 23,3313-3320 (1969).
(132) 0. Lann, B. Lindberg, and S. Svensson, Carbohydr. Res., 20, 39-48 (1971).
(133) H. E. Conrad, Methods Carbohydr. Chem., 6,361-364 (1972).
(134) G. A. Adams, Methods Carbohydr. Chem., 5, 269-276 (1965).
(135) P. D. Bragg and L. Hough,J. Chem. SOC.,4347-4352 (1957).
(136) I . C. M. Dea, A. A. Mekinnon, and D. A. Rees,J. Mol. Biol., 68, 153-172 (1972).
(137) T. F. Child and N. G. Pryce, Biopolymers, 11,409-429 (1972).
(138) J. Hoffman, B. Lindberg, and T. Painter, Actu Chem. Scand., 30, 365-366
(1976).
350 PRAKASH M. DEY

almost random distribution of the D-galaCtOSyl groups. A closer exam-

ination of Baker and Whistler's results'28 reveals that they do not, in
fact, exclude the possibility of an almost random structure for both
polysaccharides.
It has been generally presumed that the results of structural studies
on polysaccharides may be interpreted as indicating the presence of
regular, repeating units. This view should, however, be accepted with
c a u t i ~ n . ~Galactomannans
'*~~ having similar M/G ratios may have
markedly different patterns of distribution of D-galactosyl groups along
the D-mannan main-chain. This is often reflected in the ''anomalous''
pattern of the action of a-D-galactosidase on the polymers'07~'09~110 (see
also, Section 111,2,a). Interpretation of the results of structural investi-
gations is difficult, because of variable extractability of the polymer
from the same source, and because each extract may have a different
M/G r a t i ~ , leading~ ~ ~ to~ a~ batch-to-batch
~ . ~ ~ ~ variation in the fine
structure.43
The products formed by the action of the endo-acting, white-clover
p-D-mannana~e'~'on the galactomannans of guar (M/G = 2), carob
(M/G = 4),and Gleditsia ferox (M/G = 4) have been examined'39 by
t.1.c. (see Fig. 2). For guar gum, none of the products of the enzymic
hydrolysis migrated from the base line (see Fig. 2C), despite a linear
increase, during the first 50 min, in the reducing power. This could
mean that the polysaccharide had been cleaved, presumably at the
points where a sufficient number of D-mannosyl residues were unsub-
stituted, to produce a mixture of fairly large, oligomeric fragments.
Thus, the distribution of D-galactosyl side-chain groups is unlikely to
follow a regular, alternating pattern in guaran. The Gleditsia digest
contained four D-manno-oligosaccharides of the same homologous
series (see Fig. 2B,c,d,e,f) and three that contained both D-mannose
and D-galactose (indicated as * in Fig. 2,B). On the other hand, the
carob digest had five members of the D-manno-oligosaccharide series
(Fig. 2A,c,d,e,f,g) and only two sugars consisting of both monosac-
charides (indicated as * in Fig. 2A). These results tended to suggest
that neither of the two polysaccharides possesses uniform, block struc-
tures, and that the D-galactosyl groups are distributed more randomly
in Gleditsia gum than in carob gum. The action of a-Dgalactosidase on
these polysa~charides~~ also shows that fewer, isolated a-D-galactosyl
groups are interspersed in the unsubstituted regions of carob gum than
in Gleditsia galactomannan; this is in agreement with the general
structural scheme proposed by Courtois and Le Dizet."O
(139) P. M.Dey, unpublished results.
(140) J. Williams, H.Villarroya, P. M. Dey, and F. Petek, Int. Congr. Biochern., IOth,
Hamburg, 623 (1976).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 351

FIG. e.-Thin-layer Chromatographic Separation of the Products of the Action of

White-clover P-D-Mannanase on Various substrate^.'^^ [A, Carob galactomannan; B,
Gleditsia ferox galactomannan; C, guar galactomannan; D, ivory-nut mannan; and E,
salep glucomannan. a, DMannose; b, ~ g l u c o s e ;c, mannobiose; d, mannotriose; e,
mannotetraose; f, mannopentaose; and g, mannohexaose. The homologous series of D-
manno-oligosaccharides consist of (1+4)-linked P-D-mannopyranosyl residues. In A and
B, the symbol * represents the unidentified, D-galactose-containing oligosaccharides,
and, in E, the symbol t represents the unidentified, D-glucose-containing oligosac
charides.]

Deviations from the main structure with respect to “anomalous”

glycosidic linkages have also been reported in some galactomannans.
These are P-D-(1+3)- and P-D-( 1+2)- linkages in the D-mannan back-
with substitution by a-D-galactosyl groups at 0 - 2 of some
bone,91.100*141
of the D-mannosyl residues.lz4The presence O f D-galaCtOSe stubs made
up of more than one D-galactose unit, linked together by a - ~ 1+6)-
-(
linkages, has also been ~ h o ~ n . ~ ~ , ~ ~ * ~ ~ ~

(141) V. P. Kapoor and S. Mukherjee, Can.J . Chem., 47,2883-2887 (1969).

352 PRAKASH M. DEY

11. BIOSYNTHESIS
The biosynthesis of plant galactomannans has attracted relatively
little attention and, as yet, no general pathway has been defined. It
has been generally concluded that D-galactose is stored in the seeds
in the form of galactomannan, by random attachment of D-galaCtOSyl
groups to a main-chain of D-mannan.92*'00 Hough and coworkers'42
pointed out that a plausible way whereby a galactomannan may be
synthesized is by trans-D-galactosylation as shown.

G
I
Raffinose + .-..M-M-M.-.-. * .....M-M-M.... + sucrose
D-Mannan polymer galactomannan

where M = D-mannosyl residue and G = D-galactosyl group.

In 1970, Reid and Meier'43 studied the biosynthesis of galacto-
mannan in fenugreek seeds at different stages of maturation. The poly-
saccharide is present at all stages, except in the very young seeds;
the maximum yield was obtained from 9-10-week-old seeds (at al-
most the fully mature stage). The most interesting feature of the poly-
saccharide is that the M/G ratio (1.2) remains the same during various
stages of development until maturation. In addition, both free D-
mannose and D-manno-oligosaccharides, which might act as primers
for the synthesis, are absent. It therefore seems unlikely that the
D-galactosyl groups are attached randomly to a D-mannan polymer.
Instead, the results suggest that D-galactose and D-mannose units are
deposited simultaneously, to form the galactomannan. Courtois and
Le D i ~ e also t ~ ~found unchanged M/G ratios for galactomannans in
maturing seeds of Gleditsia ferox and white clover.
The immature seeds of fenugreek also contain sucrose, myo-inositol,
D-glucose, galactinol (1L-~-O-a-D-galactopyranosy~myo-inositol), and
stachyose, in addition to ga1a~tornannan.I~~ In the mature seeds, when
the deposition of the polysaccharide is complete, stachyose is still
found in abundance. This oligosaccharide is known to be formed
through the transfer of a-D-galaCtOSyl groups from galactinol to raf-
f i n 0 ~ e . Courtois
l~~ and proposed the following pathway
for formation of stachyose.

(142) M. E. Henderson, L. Hough, and T. Painter,J. Chem. Soc., 3519-3522 (1958).

(143) J. S. G . Reid and H. Meier, Phytochemistry, 9, 513-520 (1970).
(144) W. Tanner and 0. Kandler, Plant Physiol., 41, 1540-1542 (1966).
(145) A. Sioufi, F. Percheron, and J. E. Courtois, Phytochemistry, 9, 991-999 (1970).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 353

UDP-Dgalactose + myo-inositol --* galactinol UDP +

+
UDP-D-galactose sucrose + raffinose UDP +
+ +
Galactinol raffinose + stachyose myo-inositol
It has been suggested that a similar mechanism of D-galactosyl trans-
fer may be involved in the biosynthesis of galactomannan, with the
participation of G D P - ~ - m a n n o s e . ' ~ ~
Courtois
*'~~ and coworkers 14s
have, in fact, detected UDP-D-galactose and GDP-D-mannose in fenu-
greek seeds.
In the absence of a well-worked-out, biosynthetic pathway for plant
galactomannans, a useful comparison may be made with that of a re-
lated polysaccharide, namely, glucomannan. This plant polysac-
charide mostly consists of D-mannosyl and D-glucosyl residues joined,
in the form of a linear chain, by P - D - ( 1 ~ 4 ) - l i n k a g e ~ . 'Hassid
~ ~ - ' ~ ~and
coworkers153isolated from mung-bean (Phaseolus aureus) seedlings
a particulate-enzyme fraction that catalyzes the transfer of D-['4c]glu-
cose from UDP-D-['~]glucose into cellulose. This incorporation was
stimulated in the presence of GDP-D-mannose, but the product formed
under these conditions differed from cellulose.'s4 It was further found
that the D-mannosyl group of GDP-~-['~C]mannoseis, in fact,
incorporated into a polymer characterized as g l u ~ o m a n n a n . ' ~ ~
~ ~ , ' ~ the
E l b e i ~ ~ ' studied ~ properties of the enzyme system respon-
sible for the biosynthesis of glucomannan, and showed that GDP-D-
['Cjglucose is utilized only in the presence of GDP-Dmannose.
However, incorporation of radioactivity from GDP-~-['~C]mannose
was inhibited by GDP-D-glucose. Structural analysis showed that the
polymer contains 3 to 4 D-mannose units per D-glucose molecule. This
is evidence for the existence of a biosynthetic pathway dependent on
a nucleoside 5'-(glycosyl diphosphate).
(146)J. E. Courtois, in "Plant Carbohydrate Biochemistry," J. B. Pridham, ed.,
Academic Press, London, 1974,p. 1.
(147)C. T. Bishop and F. P. Cooper, Can. J. Chem., 38,793-804 (1960).
(148)A. Tyminski and T. E. Timel1,J. Am. Chem. Soc., 82,2823-2827 (1960).
(149)M. 0. Gyaw and T. E. Timell, Can. J. Chem., 38,1957-1966 (1960).
(150)T. E. Timell, Methods Carbohydr. Chem., 5, 137-138 (1965).
(151)H.J. Rogers and H. R. Perkins, "Cell Walls and Membranes," Spon, London,
1968.
(152)0.Perila and C. T. Bishop, Can. J. Chem., 39, 815-826 (1961).
(153)A. D.Elbein, G . A. Barber, and W. Z . Hassid,]. Am. Chem. Soc., 86, 309-310
(1964).
(154) G.A. Barber, A. D. Elbein, and W. Z . Hassid,J. Biol. Chem., 239, 4056-4061
(1964).
(155)A. D. Elbein and W. Z. Hassid, Biochem. Biophys. Res. Commun., 23, 311-
318 (1966).
(156)A. D.Elbein,]. Biol. Chem., 244, 1608-1616 (1969).
(157)A. D.Elbein, Methods Enzymol., 28,560 (1972).
354 PRAKASH M. DEY

Some micro-organisms also contain galactomannans, O-phosphono-

galactomannans, O-phosphonomannans, and peptidoOphosphono-
galactomannans .158-162 The galactomannans from Trichophyton and
Microsporum species have been ~ h a r a c t e r i z e d .Gander
~~ and co-
workers183-166 characterized the peptido-O-phosphonogalactomannan
from Penicillium charlesii, and studied its biosynthesis (see also, Ref.
-
167).The polysaccharide molecule contained 90 Dmannopyranosyl
residues linked through a - ~1+-2)-
( and a - ~1+6)-linkages,
( and to 0-3
of D-mannose units of this backbone were attached 8 to 10 poly-
(Dgalactofuranosyl) chains of variable length. The D-galaCtOfUranOSyl
residues in these chains were linked together through @D-( b 5 ) -
linkages.ls3It was shown that P . charlesii produced -20% of the total
peptido-O-phosphonogalactomannan before the depletion of NH4+
from the growth medium, and the rest was formed during NH4+
starvation. Incorporation studies with exogenous precursors, namely,
D-['~]ghCOSe,~-['~C]threonine, and NaH32P04,showed that the prod-
uct was the result of a biosynthetic process, and was not derived by
the autolysis of cell wa11s.164*166The presence of UDP-D-galactofur-
anose has also been demonstrated in P . charlesii168;Fobes and
G a n d e F detected an enzyme [uridine 5'-(a-~-galactopyranosyl
diphosphate): NAD 2-~-hexosyloxidoreductase] that oxidized UDP-D-
galactopyranose to a D-lyxo-hexos-2-dose derivative. The authors
suggested that a possible role of this enzyme might be the conversion
of the D-hexosyl group of UDP-D-galactose from the pyranose into the
furanose form. The complete, biosynthetic situation is not yet clear.
In addition to GDP-D-mannose as the D-mannosyl d o n ~ r , * ~ ~ * ' ~ ' D -
(158) S. A. Barker, 0. Basarab, and C. N. D. Cruickshank, Carbohydr. Res., 3, 325-
332 (1967).
(159) A. Jeanes and P. R. Watson, Can.]. Chem., 40, 1318-1325 (1962).
(160) M. E. Slodki, Biochirn. Biophys. Acta, 57, 525-533 (1962).
(161) K. 0. Lloyd, Biochemistw, 9,3446-3453 (1970).
(162) T. R. Thieme and C. E. Ballou, Biochemistry, 10,4131-4129 (1971).
(163) J. E. Gander, N. H. Jentoft, L. R. Drewes, and P. D. Rick, J . Biol. Chem., 249,
2063-2072 (1974).
(164) J. E. Gander, Annu. Reu. Microbiol., 28, 103-119 (1974).
(165) L. R. Drewes and J. E. Gander,]. Bacteriol., 121,675-681 (1975).
(166) L. R. Drewes, P. D. Rick, and J. E. Gander, Arch. Microbiol., 104, 101-104
(1975).
(167) F. A. Troy, F. E. Frerman, and E. C. Heath, Methods Enzymol., 28,602 (1972).
(168) G. Trejo, J. W. Haddock, G. J. F. Chittenden, and J. Baddiley, Biochem. ]., 122,
49-57 (1971).
(169) W. S. Fobes and J. E. Gander, Biochem. Biophys. Res. Commun., 49, 76-83
(1972).
(170) N. H. Behrens and E. Cabib,]. Biol. Chem., 243,502-509 (1968).
(171) L. P. Kozak and R. K. Bretthauer, Biochemistry, 9, 1115-1122 (1970).
 BIOCHEMISTRY O F PLANT GALACTOMANNANS 355

mannosyl-lipid intermediate^'^^-^^^ have also been shown to partici-

pate in the biosynthesis of D-mannose-containing oligo- and poly-
saccharides in plants, animals, and microbial systems. However,
further research is needed in order to establish the biosynthetic path-
way of plant galactomannans (see also, Section 111,2,d).
Reid and Meier- performed a histological examination of the
process of deposition of galactomannan in the endosperms of ma-
turing fenugreek-seeds. There was no galactomannan in the very
young seeds (2 weeks after anthesis), and the endosperm was not

(172) I. M. Weiner, T. Higuchi, L. Rothfield, M. Saltmarsh-Andrew, M. J. Osborn,

and B. L. Horecker, Proc. Natl. Acad. Sci. U.S.A., 54,228-235 (1965).
(173) A. Wright, M. Dankert, and P. Robbins, Proc. Natl. Acad. Sci. U.S.A., 54, 235-
241 (1965).
A. Wright, M. Dankert, P. Fennessey, and P. Robbins, Proc. Natl. Acad. Sci.
U.S.A., 57, 1798-1803 (1967).
M. Scher, W. J. Lennarz, and C. C. Sweeley, Proc. Natl. Acad. Sci. U.S.A.,59,
1313-1320 (1968).
M. Scher and W. J. Lennarz,]. Biol. Chem., 244,2777-2789 (1969).
M. Lehar, T. H. Chiu, and W. J. Lennarz, J . Biol. Chem., 244, 5890-5898
(1969).
A. Wright,J. Bacteriol., 105,927-936 (1971).
K. Nikaido and H. Nikaido,]. Biol. Chem., 246, 3912-3919 (1971).
F. A. Troy, F. E. Frerman, and E. C. Heath, J . Biol. Chem., 246, 118-133
(1971).
K. Takayama and D. S. Goldman,J. Biol. Chem., 245,6251-6257 (1970).
P. Babczinski and W. Tanner, Biochem. Biophys. Res. Commun., 54, 1119-
1124 (1973).
R. M. Barr and F. W. Hemming, Biochem. J., 126, 1193-1202, 1203-1208
( 1972).
J. F. Caccam, J. J. Jackson, and E. H. Eylar, Biochem. Biophys. Res. Commun.,
35,505-511 (1969).
W. Tanner, P. Jung, and N. H. Behrens, FEBS Lett., 16,245-248 (1971).
C . J. Waechter, J. J. Lucas, and W. J. Lennarz, J . Biol. Chem., 248, 7570-
7579 (1973).
J. L. Strominger, Y. Higashi, H. Sandermann, K. J. Stone, and E. Willoughby,
in “Biochemistry of the Glycosidic Linkage,” R. Piras and H. G. Pontis, eds.,
Academic Press, London, 1972, p. 135.
J. B. Richards, P. J. Evans, and F. W. Hemming, In “Biochemistry of the Gly-
cosidic Linkage,” R. Piras and H. G. Pontis, eds., Academic Press, London,
1972, p. 207.
H. Kauss, in “Biochemistry of the Glycosidic Linkage,” R. Piras and H. G.
Pontis, eds., Academic Press, London, 1972, p. 221.
W. Tanner, P. Jung, and J. C. Linden, in “Biochemistry of the Glycosidic
Linkage,” R. Piras and H. G. Pontis, eds., Academic Press, London, 1972,
p. 227.
W. Z. Hassid, in “Biochemistry of the Glycosidic Linkage,” R. Piras and H. G.
Pontis, eds., Academic Press, London, 1972, p. 315.
M. Scher and W. J. Lennarz, Methods Enzymol., 28,563 (1972).
356 PFtAKASH M . DEY

cellular at this stage. In the fifth week after anthesis, while the seeds
were green and immature, deposition of galactomannan started in the
cells next to the embryo. The polysaccharide was seen at the per-
iphery of the cells and encroaching inwards into the cytoplasm. In the
mature, green seeds (8 weeks after anthesis), endosperm cells were
completely filled with galactomannan that had spread up to the border
of the aleurone layer.
Meier and Reidlg3conducted a detailed, electron-microscope study
of the deposition of galactomannan in the endospermic cells of fenu-
greek. They observed that the cells had stacked rough, endoplasmic
reticulum (ER) at the initial stage of deposition of galactomannan;
this was followed by the appearance of vacuoles in the intracisternal
space of the ER, and their membranes entrapped cytoplasmic pockets.
It was suggested that the rough E R makes contact with the plasma-
lemma, and discharges the galactomannan-containing enchylema,
outside the cell, with local disruption of the cell membrane. At the
advanced stage of development, a larger deposition of galactomannan
was seen within the protoplast, inside the ER vacuoles. This mode of
secretion, although not very common, has been shown to take place
in a few instance^.^^^^^^^ The general mode of secretion is, however,
through the Golgi v e s i ~ l e s . ' ~ ~ ~ ' ~ ~

111. BIOCHEMICALDEGRADATION

1. General Considerations

It has been known for some time that the albuminous seeds of the
family Leguminosae contain, in the endosperm, galactomannan as the
reserve polysaccharide. The exalbuminous seeds, on the other hand,
have the main reserve in the cotyledon in the form of starch, oil, and
p r ~ t e i n s . ' The
~ ~ ~seed
' ~ ~ galactomannan-reserves are generally mobil-
ized during the process of germination.43~57~8'*82*1sg~zos Detailed study
(193) H. Meier and J. S. G . Reid, Planta, 133,243-248 (1977).
(194) T. Rochmilevitz and A. Fahn,Ann. Bot. (London),37, 1-9 (1973).
(195) J. M. Unzelman and P. L. Healey, Protoplasma, 80, 285-303 (1974).
(196) J.-C. Roland and D. Sandoz,]. Microsc. (Paris), 8,263-268 (1969).
(197) M. Rougier,]. Microsc. (Paris), 10, 67-82 (1971).
(198) F. N. Howes, "The Nature and Uses of Gums," Chronica Botanica, Waltham,
Massachusetts, 1949.
(199) J. S. G . Reid, Planta, 100, 131-142 (1971).
(200) K. Sehgal, H. S. Nainawatee, and B. M. Lal, Biochenz. Physiol. Pflanz., 164,
423-428 (1973).
(201) J. S. G . Reid and H. Meier, Planta, 106,44-60 (1972).
(202) J. S. G . Reid and H. Meier, Planta, 112,301-308 (1973).
 BIOCHEMISTRY O F PLANT GALACTOMANNANS 357

of the metabolism of galactomannan during seed germination has been

limited to only a few species.
Working with fenugreek seeds soaked in water for 24 h at 24-
26", Reidlg9 defined the start of germination as the penetration of
the seed coat by the radicle; this required a further 6 h. A his-
tological study revealed that there was no morphological change
in the endosperm or the cotyledon during the first phase of germina-
tion (18 h after the appearance of the radicle; compare Fig. l).The
level of galactomannan remained constant in the endosperm, and that
of stachyose and verbascose decreased, but that of sucrose increased;
the levels of sucrose and free D-galactose were approximately equal.
On the other hand, cotyledons were free from starch and D-galactose,
and there was a decrease in the total carbohydrates of low molecular
weight, except sucrose, which increased. The dissolution of galac-
tomannan occurred (see Fig. 3) in the second phase (lasting for 24 h,

FIG.3.-A Cross-section Similar to Fig. 1, Showing the Seed During Mobilization of

Galactoinannan (S, Three-layered Seed-coat). [The galactomannan (G) in the endosperm
is in the process of being dissolved, and the dissolution zone (D) begins at the aleurone
layer (A) and spreads inwards towards the cotyledons (C). Stained with the periodic
acid-Schiff reagent; ~ 4 0 (reproduced,
0 by permission, from Ref. 199).]

(203) A. Seiler, Planta, 134,209-221 (1977).

(204) J. S. G. Reid, C. Davies, and H. Meier, Planta, 133,219-222 (1977).
(205) B. V. McCleary and N. K. Matheson, Phytochemistry, 13,1747-1757 (1974).
358 PRAKASH M. DEY

following the first phase). The decrease in the galactomannan level

(there was no change in the M/G ratio) was due to secretion of hydro-
lytic enzymes from the active, aleurone layer into the endosperm. This
is comparable to starch hydrolysis in cereal grains during germina-
tion.20sThe galactomannan reserve and the raffinose family of sugars
completely disappeared from the endosperm at the end of this phase.
The hydrolysis products, such as sucrose, D-galaCtOSe, D-mannose, and
P-D-(1+4)-linked mannobiose, reached their highest levels during
6-12 h of the second phase, and were lowered to traces by the end of
this phase. These sugars appear to be translocated to the cotyledons,
as evidenced by the concomitant synthesis of starch (see Fig. 4)and
the presence of a high level of sugars of low molecular weight. A trans-
itory synthesis of starch in the cotyledons of fenugreek as the endo-
sperm dissolved had been indicated by NadelmannZo7in 1889. The
sugars taken up by the cotyledons are likely to have been converted
into starch through the established p a t h ~ a y s . ~ Interestingly,
~~,~~~
although the cotyledons are capable of starch synthesis, such synthesis

--- phase 1 phase 2

Germinotion time ( h )
phase 3

FIG. 4.-Disappearance of Galactomannan from the Endosperm, and Appearance

of Starch in the Cotyledon, During Germination of Seeds of Trigonella foenum-gruecum
that had Already Been Soaked for 24 Hours at 24-26”. (Adapted from Figs. 5 and 6 of
Ref. 199.)
(206) T. A. Villiers, “Dormancy and Survival of Plants,” Arnold, London, 1975.
(207) H. Nadelmann, Ber. Dtsch. Bot. Ges., 7,248-255 (1889).
(208) D. R. Davies, in “Plant Carbohydrate Biochemistry,” J. B. Pridham, ed., Aca-
demic Press, London, 1974, p. 61.
(209) M. A. R. De Fekete and H. G. Vieweg, in “Plant Carbohydrate Biochemistry,”
J. B. Pridham, ed., Academic Press, London, 1974, p, 127.
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 359

FIG.5.-A Cross-section Comparable to Figs. 1 and 3. [Here, no galactomannan is

left in the endosperm (E), only a remnant of which remains between the seed coat (S)
and the cotyledon (C). PAS stained; ~ 4 1 (reproduced,
5 by permission, from Ref. 199).]

does not occur until the seeds germinate. On the other hand, in exal-
buminous legume seeds in which the endosperm disappears during
maturation, the reserve starch is stored in the cotyledons; this may be
due to environmental adaptation undergone by the galactomannan-
containing seeds which enables them to absorb and retain water.
During the third, and final, phase (24 h following the second phase),
galactomannan is absent (see Fig. 5 ) , and the endospermic carbohy-
drates of low molecular weight also disappear completely and rapidly.
Cotyledons, however, have a high level of starch (see Fig. 4).
The pattern of disappearance of galactomannan was also studied in
dry-isolated, fenugreek endosperms, which were aseptically incu-
bated under “germination” conditions. The process, and the results,
were identical to those found in whole-seed experiments, except that
the levels of free D-galactose and D-mannose were higher, probably
because these sugars could not be translocated and metabolized in the
absence of the cotyledons. In addition, the decomposition of galacto-
mannan began slightly earlier in the isolated endosperms; this may be
due to (a) the ease of diffusion of oxygen into the endosperm, causing
aerobic respiration to start, and, possibly, (b) the ease of removal of
360 PRAKASH M. DEY

any metabolic inhibitor from the tissues. ShiroyaZ1Ohas also shown that
oxygen is required for the disappearance of the raffinose family of
sugars in germinating cotton-seeds. As the absence of embryo in the
isolated-endosperm experiment had no effect on the mobilization of
galactomannan, this would indicate that, unlike the process in cer-
eals,211-214the embryo does not participate in the catabolic process. It
was also confirmedm1that the composition of the galactomannan of the
isolated endosperms of fenugreek, crimson clover (Trifolium in-
carnatum), and lucerne were unchanged (the M/G ratios were
-1) throughout the “germination” period. On the other hand,
Courtois and coworkers145showed a change of M/G ratio from 1.05 to
7.45 in the galactomannan of 24hour-germinated7 fenugreek seeds.
However, the latter experiment was conducted with whole seeds (see
also Section 111,2,b). It was also suggested that, during germination of
the seeds of Gleditsia ferox and G . triacanthos, there was an initial
removal of Dgalactosyl groups from the polysaccharides, followed by
the cleavage of the D-mannan backbones, giving rise to D-manno-
oligosaccharides (see also, Section 111,2,a).
Reid and MeierZo1 suggested that the presence of free D-galactose in
the endosperm during the early stages of germination might be due to
the non-living nature of the storage tissues. Cotyledons, on the other
hand, consist of living cells where Dgalactose is rapidly metabo-
l i ~ e dPridham
. ~ ~ ~ and coworkers216have actually detected a D-gdaCt0-
kinase in Viciafuba seeds, and related this enzyme to the metabolism
Of D-galaCtOSe through phosphorylation (see also, Refs. 145, 217, and
218). A similar observation was made by Shiroya,210who examined the
breakdown of raffinose and stachyose in germinating cotton-seeds; by
infiltration of D-galactose into the seeds, he showed that an effec-
tive mechanism for the utilization of D-galactose exists.
Reid and MeierZo1 found that the breakdown of galactomannan in iso-
lated, fenugreek endosperms is strongly inhibited by cycloheximide
(210)T. Shiroya, Phytochemistry, 2, 33-46 (1963).
(211)M. Black, “Control Processes in Germination and Dormancy,” Clarendon
Press, Oxford, 1972.
(212)H. Yomo, Hukko Kyokaishi, 16,444-448 (1958).
(213)D.Cohen and L. G . Paleg, Plant Physiol., 42, 1288-1296 (1967).
(214)J. E. Varner and G . R. Chandra, Proc. Nutl. Acad. Sci. U.S.A., 52, 100-106
(1964).
(215)H. Coring, E.Reckin, and R. Kaiser, Flora (Jena),Abt. A, 159,82-103 (1968).
(216)J. B. Pridham, M. W. Walter, and H. G. J. Worth,J. E r p . Bot., 20, 317-324
(1969).
(217)S. Clermont, M. J. Foglietti, and F. Percheron, Compt. Rend., 276, 843-847
(1973).
(218) M.J. Foglietti and F. Percheron, Biochimie, 56,473-475 (1974).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 361

{3-[2-(3,5-dimethyl-2-oxocyclohexyl)-2-hydroxyethyllglutarimide} and
by 2,4-dinitrophenol. They related this to inhibition of the synthe-
sis of protein, and oxidative phosphorylation, respectively, in the
aleurone layer. An electron-microscope examinationzo1showed the
presence of a large number of ribosomes in the aleurone cells, even
at the beginning of germination. There were large numbers of poly-
ribosomes in the first phase of germination, and these were membrane-
bound to small vesicles and to flat cisternae which were probably
newly formed and derived from ER. As the germination progressed,
the content of aleurone cells vanished, and vacuoles appeared. It
was, therefore, concluded that the galactomannan-mobilizing enzymes
are synthesized in the aleurone layer, and that the dissolution of the
polysaccharide is not controlled by the embryo.

2. Enzymes Involved
Hylin and Sawai2I9 isolated, and crystallized, an enzyme from
Leucaena leucocephala seeds which they termed “galactomannan
depolymerase.” This enzyme catalyzed a rapid fall in the viscosity of
solutions of galactomannan (obtained from the same source), at pH 5.3,
to a low value, after which, no further change in the viscosity or in-
crease in the reducing power occurred. The maximum depolymeriz-
ing activity was isolated from the 7-day-germinated seeds. It was
suggested that, in the initial stages of germination, the polysaccharide
is depolymerized into fragments of considerable size, which are then
further degraded into D-galactose and D-mannose units. The galacto-
mannan of L. leucocephala seeds consists of 57% of D-mannose and
43% of D-galactose, and the degree of polymerization (d,p.) is 150. The
main backbone of the D-mannan has some @-D-(1+3)-linkages, in ad-
dition to the usual @-D-(1+4)-linkages, and some D-galaCtOSyl residues
occupy intra-chain positions.69If an almost random, general structure
of the polysaccharide is considered (see also, Ref. 112), a few unsub-
stituted areas along the main chain would be expected. These areas
would, undoubtedly, be the points of attack by an endo-@-D-mannan-
ase; this enzyme would, however, be unable to by-pass the points
where the intra-chain D-galactose units occur. It would be of interest
to examine the substrate specificity of the depolymerase, in order to
establish its identity. A similar galactomannanase has also been iso-
lated from germinated guar seeds.z00

(219) J. W. Hylin and K. Sawai,J. Biol. Chem., 239, 990-992 (1964).

362 PRAKASH M. DEY

Whistler and coworkers220showed that, on incubation with guaran

solution, enzyme preparations from guar seeds increase the reducing
power and decrease the viscosity, but the ratio of these two effects
differed for different enzyme preparations. Thus, they predicted that
two or more guaran-hydrolyzing enzymes must be present in guar
seeds. These may be (a) an enzyme that would hydrolyze the /3-D-
mannosidic bonds in the main chain (for example, an endo-P-D-
mannanase), and (b)an enzyme that would hydrolyze a-D-galactosidic
bonds (for example, a-D-galactosidase).
According to Reese and Shibata,22‘the biodegradative hydrolysis
of galactomannan requires at least three different enzymes, namely,
a-D-galactosidase, endo-p-D-mannanase, and p-D-mannosidase, In
addition to these enzymes, there is also a report222of a phosphorylase,
identified as oligo-/3-~-mannosyl-(1+4)-phosphorylase, which may
take part in the degradation. These enzymes will now be discussed
in detail.
a. a-D-Ga1actosidase.-This enzyme (melibiase, a-D-galactoside
galactohydrolase, EC 3.2.1.22)catalyzes the following reaction.

“0 CH,OH

OH
OR + q0 “>..H
CH20H

OH
+ ROH

An ff-D-galactoside D - galactose

where R is an alkyl or aryl group, or a glycosyl (mono or oligo) residue

or group.
Under suitable conditions, the enzyme can also catalyze de nouo
syntheses of oligosaccharides and transfer reaction^.^^^,^^^
a-D-Galactosidases have been isolated from animal, plant, and
microbial sources; a review223includes details of the isolation, proper-
ties, and characterization of these enzymes. Amongst plants, they have

(220) R. L. Whistler, W. E. Eoff, and D. M. Doty, J . Am. Chem. SOC.,72, 4938-4949

(1950).
(221) E. T. Reese and Y. Shibata, Can. J . Microbial., 11, 167-183 (1965).
(222) M. J. Foglietti and F. Percheron, Compt. Rend., 274, 130-132 (1972).
(223) P. M. Dey and J. B. Pridham,Ado. Enzymol., 36,91-130 (1972).
(224) J. B. Pridham and P. M. Dey, in “Plant Carbohydrate Biochemistry,” J. B.
Pridham, ed., Academic Press, London, 1974, p. 83.
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 363

been detected in leguminous and other species, especially those

which contain or utilize a-D-galactosidic substrates.43~s7~81~z0s~,223~zzs~
It is well established that, in plants, one of the functions of WD-
galactosidases is to cleave a-D-galactosyl groups from a-D-galac-
tose-containing oligo- and poly-saccharides. The degradation products
thus formed serve as a ready source of energy and of cell metabolites.
These enzymes have also been associated with the metabolism of
galactolipids in plantszz8andwith chloroplast-membrane f ~ n c t i o n . ~ ~ ~ ~
The specificities of a-D-galactosidases vary widelyzz3;the enzymes
from Vicia s a t i ~ anda~~ Mortierella
~ v i n ~ c e ahydrolyze
~~~ substrates
of small molecular weight, such as phenyl a-D-galactopyranoside, but
do not act on larger substrates, such as galactomannans. However, the
enzymes from Phaseolus ~ u l g a r iand s ~ coffee
~ ~ beans108hydrolyze both
groups of substrates. A more detailed account of substrate specificity
is given in a review.z23It seems apparent that the specificity and func-
tion of an a-D-galactosidase are related to the nature of the a-D-galacto-
sidic compounds that may occur in a given source. Thus, a galacto-
mannan from Gleditsia ferox (M/G = 4) was attacked by the a - ~ -
galactosidase from that source, but was unaffected by the enzyme from
Penicillium p a x i l l u ~Courtois
. ~ ~ ~ and Le D i ~ eshowedt ~ ~ ~that the mode
of action of a-D-galactosidases on galactomannans may vary with the
source of the enzyme, even though these may come from galacto-
mannan-containing species. However, the enzymes were able to lib-
erate D-galactose more readily from a galactomannan having a low
content of D-galactose than from one having a high content thereof.
Hui and Neukom120reported that, despite lengthy incubation, a - ~ -
galactosidase from coffee beans liberated only 20% of the total D-
galactose residues from guar galactomannan (M/G = 2), and 70% from
that of carob (M/G = 4),and that complete removal of =galactose by
this enzyme was not possible; this conclusion was in agreement with
similar, earlier findings."' Agrawal and Bahl, using enzyme prepara-
tions from both Aspergillus nigerZ3"and Phaseolus v ~ l g a r i s also ,~~~

(225) D. French, Adu. Carbohydr. Chem., 9, 149-184 (1954).

(226) J. E. Courtois and F. Percheron, Mem. SOC. Bot. Fr., 29-39 (1965).
(227) P. M.' Dey and K. Wallenfels, Eur. J . Biochem., 50, 107-112 (1974).
(228) P. S. Sastry and M. Kates, Biochemistry, 3, 1271-1280 (1964).
(229) E. S. Bamberger and R. B. Park, Plant Physiol., 41, 1591-1600 (1966).
(230) S. Gatt and E. A. Baker, Biochim. Biophys. Acta, 206, 125-135 (1970).
(231) F. Petek, E. Villarroya, and J. E. Courtois, Eur. J . Biochem., 8, 395-402 (1969).
(232) H. Suzuki, S . C. Li, and Y. T. Li,J. Biol. Chem., 245, 781-786 (1970).
(233) K. M. L. Agrawal and 0. P. Bah1,J. Biol. Chem., 243, 103-111 (1968).
(234) J. E. Courtois and P. Le Dizet, Bull. SOC. Chim. Biol., 45,743-747 (1963).
(235) 0. P. Bahl and K. M. L. Agrawa1,J. Biol. Chem., 244,2970-2978 (1969).
364 PRAKASH M. DEY

obtained only a 30-40% removal of the total D-galaCtOSe from these

two polysaccharides. Interestingly, the a-D-galactosidase isolated from
guar seeds removed all ofthe D-galactose residues from guaran, leaving
a residue of water-insoluble D-mannan.236
Coffee-bean a-D-galactosidase has been extensively studied with
respect to its specificity for various galactomannans. The e n ~ y m e ~ ~ . ' ~ ~
liberated D-galactose at comparable rates during the initial stages of
reaction with the galactomannans from white clover, lucerne, fenu-
greek (all with M/G = l),Genista scoparia (M/G = L6), Caesalpinia
spinosa (M/G = 3),Gleditsia ferox, and honey locust (both having
M/G = 4).The results of prolonged action were, however, different.
The hydrolysis of the polysaccharide from white clover gave a prod-
uct having an M/G ratioIo7 of 1.7, whereas hydrolysis of lucerne
galactomannan yielded a product with an M/G ratiol10 of 3.0. The
products from those of G. ferox and carob (M/G = 4)had M/G ratios
of -30 and 5.3,r e s p e ~ t i v e l y . This
~ ~ ~ 'indicates
~~ a difference in the
fine structures of the two groups of polysa~charides.~~ Courtois and
Le Dizet"O suggested that the a-D-galactosidase initially removes
the randomly distributed D-galactosyl groups which are situated next
to the unsubstituted D-mannosyl residues in the polymer. This is
followed by a slower action on the extremities of the uniform
blocks of D-galactosyl groups along the D-mannan backbone. The
enzyme is, however, unable to remove all the D-galactosyl groups
from the polymer. The authors further suggested1I0that the action of
a-D-galactosidase would thus produce longer blocks of unsubstituted
D-mannosyl residues which would be relatively hydrophobic in
nature. These parts must then fold back, more or less over, the
blocks carrying D-gdactosyl groups, thereby making the latter in-
creasingly inaccessible to further enzymic cleavage. It is also
tempting to envisage a form of secondary structure having hydro-
phobic interactions between inter- or intra-chain, unsubstituted
D-mannose blocks; this may also make the D-galactosyl groups in-
accessible to the enzyme. This situation can be compared to that
of "buried" sulfhydryl groups in some proteins.
a-DGalactosidase activity has been shown to increase during the
germination of seeds of various species, notably, fenugreek, 145*202
g ~ a r , 'Gleditsia
~~ f e r o ~ , ~lucerne,205
~' carob,20s soybean,20s runner

(236) S. R. Lee, Ph.D. Thesis, Univ. of Minnesota, Minneapolis (1965); Chem. Abstr.,
68, 111,694~(1968).
(237) J. E. Courtois and P. Le Dizet, Bul. SOC. Chim. Biol., 48, 190-191 (1966).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 365

cotton,210coffee,240and various other^.^^^^^^' In most

bean,233,238*239
of these, a concomitant depletion of a-D-galactosidic, reserve carbo-
hydrates takes place.
Dry fenugreek seedszo2have a high level of a-D-galactosidase in
the embryo, but only negligible amounts in the endosperm. How-
ever, during germination, whereas the enzyme level in the embryo
remains constant, it increases sharply in the endosperm; this process
parallels the degradation of galactomannan. Similar observations
have been made on the incubation of dry-isolated embryos, and of
endosperms under “germination” conditions. In the in vivo experi-
ments, the increase in enzyme level was strongly diminished by
cycloheximide. The inhibitor acted by preventing protein synthesis
in the aleurone tissues. SeilerZo3has actually shown the synthesis of
a-D-galactosidase by demonstrating the in vivo incorporation of
~-[‘~C]serine into the enzyme protein during the germination of
carob beans. Reid and MeierzoZsuggested that galactomannan is not
the natural substrate for the embryo a-D-galactosidase. Rather, the
enzyme may be responsible for the hydrolysis of the raffinose fam-
ily of sugars. It has been found139that a-D-galactosidase from germ-
inated fenugreek seeds could be resolved into two active peaks by
chromatography on CM-cellulose, whereas the separation is not
possible by gel filtration. It will be of interest to examine the sub-
strate specificities of the enzymes from the embryo and the endo-
sperm.
McCleary and MathesonZosshowed the presence of multimolecular
forms of a-D-galactosidase (designated A, B, and C) in germinated
seeds of carob, guar, lucerne, and soybean. They achieved the sep-
aration of the isoenzymes by DEAE-cellulose chromatography.
Forms A and C were common to all species examined; carob and
soybean had a third form, B. By column chromatography on Sepha-
dex G-200, enzyme C from soybean was further resolved into two
peaks of activity, I, and 11. Form A was also present in the dry seeds,
and its level did not increase significantly on germination. It was
suggested that A and B, which are confined to the cotyledon-embryo
part of the seeds, are mainly responsible for the hydrolysis of a-D-
galactosidic oligosaccharides. On the other hand, form C is an endo-
spermic enzyme, and it increases rapidly on seed germination,

(238) D. Lechevallier, Compt. Rend., 258,5519-5522 (1964).

(239) D. Lechevallier, Compt. Rend., 250,2825-2827 (1960).
(240) M. Sadaksharaswami and G. Ramachandra, Phytochemistry, 7, 715-719 (1968).
(241) D. Lechevallier, Compt. Rend., 255, 3211-3213 (1962).
366 PRAKASH M . DEY

except in soybean, which has a very low level of galactomannan.

The enzyme C was shown to be highly specific for the hydrolysis
of galactomannan.
Two a-D-galactosidases from Viciu fubu have been resolved by
gel filtration.223During germination, form I of high molecular weight,
(preponderant in the dry seeds) declined, with a slow increase in
the level of form I1 (of low molecular weight). The changes in the
enzyme levels that occur during maturation and germination have
been described in There are indications242that some com-
partmentation of a-D-galactosidase activity in organelles, in the form
of spherosomes, occurs in V. fuba seeds; this may explain the simul-
taneous increase in the levels of both a-D-galactosidase and its
substrates during seed maturation.
b. P-D-Mannanase.-This enzyme [endo-p-D-mannanase, (1+4)-p-
D-Inannan mannohydrolase, EC 3.2.1.781 catalyzes the hydrolysis of
p-D-( 1+4)-mannopyranosyl linkages of mannans, galactomannans,
glucomannans, galactoglucomannans, and D-manno-oligosacchar-
ides.243In addition, some p-D-mannanases have also been
to hydrolyze such linkages as those in p-D-Manp-(1+4)-p-D-Glcp-
and p-~-Glcp-( 1+4)-P-D-Manp-.
P-D-Mannanases have been detected in r n i c r ~ - o r g a n i s m s , ~ ~ ~ - ~ ~ '
and animals.243~256-25s
plants,220~204~243~2"2-255 Microbial P-D-mannanases

(242) P. M. Dey, M. T. Fordom, and J. B. Pridham, unpublished results.

(243) R. F. H. Dekker and G. N. Richards, Ado. Carbohydr. Chem. Biochem., 32,
277-352 (1976).
(244) S. Innawi, Agric. Biol. Chem., 25, 155-163 (1961).
(245) H. Lyr and E. Novak, 2. Allg. Mikrobiol., 2,86-98 (1962).
(246) E. Ahlgren, K. E. Eriksson, and 0. Vesterberg, Acta Chem. Scand., 21, 937-
944 (1967).
(247) Y. Misawa, M. Matsubara, M. Hatano, M. Hara, and T. Inuzuka, Nippon Sho-
kuhin Kogyo Gakkai Shi, 14,286-291 (1967); Chem. Abstr., 68,86,259e (1968).
(248) K. E . Eriksson and W. Rzedowski, Arch. Biochem. Biophys., 129, 683-688
(1969).
(249) Y. Hashimoto, Nippon Nogei Kagaku Kaishi, 44,287-292 (1970); Chem. Abstr.,
73, 108,395d (1970).
(250) G. Keilich, P. J. Bailey, and W. Liese, Wood Sci. Technol., 4,273-283 (1970).
(251) S. Emi, J. Fukumoto, and T. Yamamoto, Agric. Biol. Chem., 36, 991-1001
(1972).
(252) B. V. McCleary and N. K. Matheson, Phytochemistry, 14, 1187-1194 (1975).
(253) H. Villarroya and F. Petek, Biochim. Biophys. Acta, 438, 200-211 (1976).
(254) P. Halmer, J. D. Bewley, and T. A. Thorpe, Planta, 130, 189-196 (1976).
(255) S. Clermont-Beaugiraud and F. Percheron, Bull. SOC. Chim. Biol., 50, 633-639
(1968).
(256) H. Bierry and J. Giaja, Biochem. Z., 40,370-378 (1912).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 367

are known to be both inductive and constitutive, occurring as extra-

cellular or intracellular enzymes.243It has been suspected that the
p-D-mannanases detected in the digestive tracts of various animals
may be due to symbiotic r n i c r o - ~ r g a n i s m s Dekker
. ~ ~ ~ and
characterized various p-D-mannanases.
Studies on the properties of plant p-D-mannanases have been
hampered by the difficulty in obtaining the enzyme free from inter-
fering glycosidases, notably, a-D-galactosidase, P-D-glucosidase, and
p-D-mannosidase, Beaugiraud and PercheronZ6O had only limited
success in obtaining, from germinated, fenugreek seeds, a p-D-man-
nanase free from a-D-galactosidase. They found that continuous-flow
electrophoresis gave an a-D-galactosidase preparation lacking detect-
able @-D-mannanase activity, but that the reverse was not
possible. A similar result was obtained when separation by CM-
cellulose chromatography was attempted.139By using ion-exchange
chromatography and gel-filtration, S. R. Lee236 resolved multiple
forms of p-D-mannanase from germinated guar-seeds; one of the forms
was free from a-D-galactosidase activity. Extensive purifications ofp-D-
mannanases have been achieved, the sources being lucerne,253white
clover,'40 and konjaP1; the enzyme from Bacillus subtilisZs1has
been crystallized.
Most 6-D-mannanases are endo-acting enzymes, as evidenced by a
rapid fall in viscosity when they act on polymeric substrates, pro-
ducing a series of Dmanno-oligosaccharides of different d.p. val-
ues, 110.1 12,243,251,285,262-265
Amongst the D-manno-oligosaccharides, those having a d.p. > 2 are
hydrolyzed, but the minimum d.p. requirement of a readily hy-
drolyzable substrate is14032433253-4. In a detailed study, Villarroya and
PetekZs3found that the lucerne p-D-mannanase is unable to hydro-

(257) P. J. Kooiman,J. Cell. Comp. Physiol., 63, 197-201 (1964).

(258) F. L. Meyers and D. H. Northcote, Biochem. J . , 6 9 , 5 4 ~(1958).
(259) J. E. Courtois, F. Petek, and T. Dong, Bull. SOC. Chim. Biol., 44, 11-21 (1962).
(260) S. Beaugiraud and F. Percheron, Compt. Rend., 259,3879-3881 (1964).
(261) H. Shimihara, H. Suzuki, N. Sugiyama, and K. Nisizawa, Agric. Biol. Chem., 39,
301-312 (1975).
(262) Y. Hashimoto and J. Fukumoto, Nippon Nogei Kagaku Kaishi, 43, 317-322
(1969); Chem. Abstr., 72,53,856b (1970).
(263) S. Clermont-Beaugiraud, M. Charpentier, and F. Percheron, Bull. SOC. Chim.
Biol., 52, 1481-1495 (1970).
(264) Y. Tsujisaka, K. Hiyama, S. Takenishi, and J. Fukumoto, Nippon Nogei Kagaku
Kaishi, 46, 155-161 (1972); Chem. Abstr., 77, 44,667s (1972).
(265) N. Sugiyama, H. Shimahara, T. Andoh, and M. Takemoto, Agric. Biol. Chem., 37,
9-17 (1973).
368 PRAKASH M. DEY

lyze p a - (1+4)-linked mannobiose (M,) and mannotriose ( M3). The

hydrolysis of M4 is slow, and results in M3 (lo%), M, (80%), and
D-mannose (10%). On the other hand, M5 is rapidly hydrolyzed,
producing equal amounts of M2 and M3, and only -2% of D-mannose.
The enzymic action on Me liberated M3 (50%), M4 (25%), and M2
(25%);on further incubation, M4 was degraded to a mixture of M3
(55%), M2 (41%), and D-mannose (4%). Based on these results,
Villarroya and Petek2= suggested the following mode of action.

(MA M-M
(M3) M-M-M
(M4) M-M-&M-&M
(M,) M-M-M-bM-M
(Me) M-M-M+M-$M-M

where M = P-D-(1+4)-linked D-InannOpyranOSyl residues, 4 =

preferential points of attack, J, = secondary points of attack, and f =
slow attack.
Beaugiraud and Percheron2ss~z60 showed that the enzyme from
germinated fenugreek seeds separately hydrolyze both M4 and Ms,
yielding M3, M,, and traces of D-mannose from each. To explain
the hydrolytic products from M4, the authors suggested that some of
the initially formed M, units are linked to M4 molecules, pre-
sumably by transglycosylation, producing Me; this could then be
cleaved into two M3 units.
The action of white-clover p-D-mannanase" on ivory-nut mannan
produced139 a mixture of D-manno-oligosaccharides (Mz-M6), as
shown in Fig. 2,D. On the other hand, the enzyme from fenugreek
seeds mainly produced M,, M3, and traces of D-manno~e.'~~ The
enzyme from white clover also hydrolyzes salep (Orchis militaris)
glucomannan (D-mannose/D-glucose = 3), yielding a mixture of ohgo-
saccharides, some of which contain D-glucose, and also traces of free
D-glUCOSe (see Fig. 2,E). The presence of free D-glucose and a disac-
charide containing D-mannose and D-glUCOSe units may indicate the
ability of the enzyme to act upon /3-~-(1+4)-linkages between these
two sugar residues. It will be of interest to investigate whether the
D-glUCOSe units of the oligosaccharides are situated terminally or in
the middle of the chain. However, the enzyme from lucerne2s3does
not produce any D-glucose-containing oligosaccharides, even
following a long incubation with the glucomannan; M, and M3 are the
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 369

only products. On the other hand, a purified enzyme from konjac

rapidly decreased the viscosity of konjac glucomannan, but the
reducing power liberated was very This shows the presence
of only a few points of attack on the polysaccharide.
The ability of p-D-mannanases to hydrolyze galactomannans is gen-
erally related to the M/G ratio of the polysaccharide; a higher value
facilitates the degradation.l16 It has been found that, during the
initial period of reaction (20 min),140the white-clover enzyme produces
almost equal amounts of reducing power from the galactomannans of
carob (M/G = 4) and Gleditsia ferox (M/G = 4), but the reducing
power was almost half with that of guar gum (M/G = 2),and one-tenth
with that of white clover (M/G = 1.7); the K , values for the polysac-
charides were, respectively, 0.16, 0.08, 0.83, and 1.1%(the concen-
trations are expressed as w/v). Despite a similar, overall M/G ratio,
Gleditsia galactomannan was more strongly bound to the enzyme
than that of carob. Analysis of the hydrolysis products showed that
the former yielded a greater number of D-galactose-containing oligo-
saccharides than the latter; the Gleditsia polymer also yielded some
free D-mannose (see Fig. 2,A and B). This is further evidence for
a more random structure for Gleditsia galactomannan (see also Sec-
tion 1,4),thus providing a greater number of sites for enzymic attack
along the main chain, and a higher affinity for the enzyme. Dea and
Morrison43reported that, in accordance with the random structure,
the action of a-D-galactosidase on the Gleditsia polysaccharide
yields a greater amount of insoluble products than its action on
carob galactomannan. The action of p-D-mannanase should, there-
fore, produce the opposite effect.
It was shown43that, whatever the initial M/G ratio (1.7-5.3) of
the galactomannan, the action of P-D-mannanase always produced
final, nondialyzable products having M/G ratios of -1. Thus, lucerne
galactomannan (M/G = 1) is resistant to the enzyme. However, the
a-D-galactosidase-treated polymer is attacked by the enzyme. It was
that the presence of a D-galactosyl group on each
of the Dmannosyl residues of the main chain of Dmannan hinders
the enzymic cleavage of endo-p-D-mannosidic linkages; for hydroly-
sis to occur, at least one unsubstituted D-mannosyl residue was
required. On the other hand, Reese and Shibata"' found that two
such units are necessary for the microbial P-D-mannanase; a similar
observation was made by McCleary and coworkers.116It has been
found139that, although guar gum (M/G = 2) demonstrates reducing
power when it is treated with white-clover @-D-mannanase,oligo-
saccharides of low molecular weight could not be detected (see
370 PRAKASH M. DEY

Fig. 2,C); this indicates that there are very few points of enzymic
attack on the polysaccharide, and, hence, the high K,value.
Bourquelot and Hi.rissey266-26s were the first to show the develop-
ment of p-D-mannanase, then referred to as seminases, during the
germination of seeds, notably those of carob, fenugreek, lucerne, and
white clover. The activity of this enzyme was also monitored in the
germination of the seeds of ivory nut,271barley,272guar,220*236
and konjac.26sIt is now generally established that, in seeds, p-D-
mannanase activity increases on germination, with a concomitant
depletion of D-mannan and related polysaccharides.'ss~201*204~20s~219~233~
236.252.254

McCleary and Matheson2s2found that the level of p-D-mannanase

reaches a maximum after 2-3 days of germination in lucerne, and
after 7-8 days in carob. Similar maxima occurred for guar and honey
locust. The authors purified, and resolved, these P-D-mannanaSe
activities by ion-exchange chromatography and gel filtration. Honey
locust has only one form of the enzyme, namely B; carob and soy-
bean have three forms, and lucerne, four; form B is predominant in
all these species. In guar, S . R. Lee23sdetected two forms of the
enzyme; these were found only in the endosperm. It was ob-
served252that the enzymes preferentially and rapidly hydrolyze the
galactomannans of carob and locust bean, which have a relatively
low content of D-gakiCtOSe. However, when the seeds contain poly-
saccharides having a high content of D-gahCtOSe (lucerne and guar),
they also possess a very active a-D-galactosidase. The physiological
significance of this enzyme is that an initial, partial removal of
D-galaCtOSyl groups could then lead to a rapid depolymerization by
p-D-mannanase; this, perhaps, explains why the maximal activity of
the latter enzyme is reached at a much later stage in germinating
lucerne, compared to carob. The M/G ratio of the polysaccharide
also increases in guar and lucerne during germination, but it re-
mains constant in carob and honey locust.2s2
Reid and coworkers204showed that, in fenugreek, P-D-mannanase
activity reaches a maximum at 42-48 h of germination, paralleling
(266) E. Bourquelot and H. Hhrissey, Compt. Rend., 129,614-616 (1889).
(267) E. Bourquelot and H. Hbrissey, Compt. Rend., 130, 42-44, 340-342, 731-733,
1719-1721 (1900).
(268) H. HBrissey, Compt. Rend., 133, 49-52 (1901).
(269) H. HBrissey, Compt. Rend., 134,721-723 (1902).
(270) J. Griiss, Ber. Dtsch. Bot. Ges., 20, 3 6 4 1 (1902).
(271) F. J. Paton, D. R. Nanji, and A. R. Ling, Biochem J . , 18, 451-454 (1924).
(272) H. Pringsheim and A. Genin, 2.Physiol. Chem., 40,229-234 (1924).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 371

the breakdown of storage galactomannans; this is in accordance with

our observations. The enzyme activity increases even when the dry-
isolated endosperms are incubated under “germination” condi-
t i o n ~The .~~ synthesis
~ of the enzyme takes place in the active, aleur-
one layer, and can be inhibited by the inhibitors of protein synthe-
sis. 199,201,202,204
The relationship between P-D-mannanase and the germination of
lettuce seeds has been examined in detail by Halmer and co-
w o r k e r ~ . ~ The
” * ~ lettuce-seed
~~ embryo is known to be surrounded
by a tough, endospermic wall whose major constituent s ~ g a r s ~ ~ ~ a ~
are D-mannose (58%), D-glucose (lo%), D-galaCtOSe (9.5%), and L-
arabinose (9.5%).The wall polysaccharide is similar to galactomannar~~~~
(compare, Refs. 80 and 274). The emerging radicle has to penetrate
this wall during germination, and an ordered degradation of the wall
has, in fact, been demonstrated during this process.273,27s*276 It was
suspected that cellulase was p r o d ~ ~ e dwhich ~ ~ might
~ * ~weaken
~ ~ , ~ ~ ~
the endosperm wall, but this would also affect the wall of the
emerging radicle. Halmer and coworkers273have, however, shown
the production, in lettuce endosperm, of a p-D-mannanase which
could degrade this wall; a similar observation was made for ivory-
nut e n d o ~ p e r m Alternatively,
.~~~ it was suggested that the grow-
ing radicle accumulates enough mechanical thrust to penetrate the
endosperm wa11.280~281 It is possible that the emergence of the radicle
is assisted by a combination of the thrust factor and pectinase,
cellulase, and P-D-mannanase activities. A low, basal level of the
latter enzyme is always present in the endosperm. It is not currently
known whether the endosperm wall in the vicinity of the growing
radicle-tip is initially degraded because of a local increase in the
P-D-mannanase activity. However, the high level of P-D-mannanase
produced after the radicle has emerged demonstrates the importance
of the enzyme in degrading D-mannose-containing polysaccharides
that serve as an energy source in the growing embryo.2s4

(273) P. Halmer, J. D. Bewley, and T. A. Thorpe, Nature, 258,716-718 (1975).

(274) G. 0. Aspinall, “Polysaccharides,” Pergamon Press, Oxford, 1970.
(275) R. L. Jones, Planta, 121, 133-146 (1974).
(276) W. M. Park and S. S. C. Chen, Plant Physiol., 53, 64-66 (1974).
(277) H. Ikuma and K. V. Thimann, Plant Cell Physiol., 4, 169-185 (1963).
(278) A. D. Pavlista and J. G. Valdovinos, Plant Physiol., 56 (suppl.),83 (1975).
(279) H. Meier, Biochim. Biophys. Acta, 28, 229-240 (1958).
(280) M. W. Nabors and A. Lang, Planta, 101,l-25 (1971).
(281) M. W. Nabors and A. Lang, Planta, 101,26-42 (1971).
372 PRAKASH M. DEY

c. P-D-Mannosidase.-This enzyme (p-D-mannoside mannohydro-

lase, EC 2.3.1.25) catalyzes the following reaction.
CH20H CH20H

HO0
A p-~-mannoside
+ H,O
H o O

D-mannose
O H + ROH

where R may be an alkyl, aryl or, glycosyl group.

P-D-Mannosidases are exo-hydrolases that have been shown to re-
move p - ~ 1+4)-linked
-( D-mannosyl groups from the nonreducing
end of their substrates, for example, D-manno-ohgosaccharides and
D-mannose-containing g l y ~ o p e p t i d e s . ~ ~The ~ - ~ ~ ~ of p-D-
* * ~presence
mannosidases has been shown in a range of animal t i s s ~ e s and ~ ~ - ~ ~ ~
Amongst plants, the enzyme has
in some micro-organisms.22'*23s~2g8~299
been extracted from fenugreek,2°2,260guar,236*252 pineapple,3"O lu-
cerne,2= honey l o ~ ~ s tPhoenix, ~ ~ canariensis,282
~ , ~ ~ ~ wheat
and malted barley.302On the other hand, extracts of a num-
ber of galactomannan-containing leguminous seeds had very little
or no p-D-mannosidase
There have been only a few kinetic studies on plant p-D-manno-
sidases, because of the apparently limited distribution of this en-
zyme in various species, and also because of the paucity of attempts
at purification. The enzyme from malted barley has been purified
41-fold by fractionation with ammonium sulfate, and chromatog-
raphy302on Biogel P-100, DEAE-cellulose, and CM-cellulose. Some
properties of p-D-mannosidases from various tissues are summarized
in Table 11.
(282) J. E. Courtois and P. Le Dizet, Bull. Soc. Chim. Biol., 46,535-542 (1964).
(283) K. Sugahara, T. Okumura, and I. Yamashina, Biochim. Biophys. Acta, 268,488-496
(1972).
(284) B. A. Bartholemew and A. L. Perry, Biochim.Biophys. Acta, 315,123-127 (1973).
(285) S. Toyoshima, M. Fukuda, and T. Osawa, Biochem. Biophys. Res. Commun., 51,
945-950 (1973).
(286) K. Sugahara and I. Yamashina, Methods Enzymol., 28, 769 (1972).
(287) G . L. E. Koch and C. A. Marsh, Comp. Biochem. Physiol. B , 42,577-590 (1972).
(288) T. Nagaoka,]. E x p . Med., 51, 131-138 (1949).
(289) T. Muramatsu and F. Egami,]. Biochem. (Tokyo), 62,700-709 (1967).
(290) J. C. Steigerwald and B. A. Bartholomew, Biochim.Biophys. Acta, 321,256-261
(1973).
(291) T. Sukeno, A. L. Tarentino, T. H. Plummer, Jr., and F. Maley, Biochem. Biophys.
Res. Commun., 45,219-225 (1971).
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 373

TABLEI1
Kinetic Parameters of Some P-D-Mannosidases

pH Km"
Source optimum (mM) Inhibitor References

Carob
cotyledon 5.0 0.29 - 252
endosperm 5.0 0.29 - 252
Hen oviduct 4.6 4.50 Ag+,no inhibition at 0.6 mM 305
~mannono-1,5-lactone
(competitive, Ki=17 p M )
Human skin 3.5 - - 290
Human synovial 3.5-4.0 3.4 p-nitrophenyl P-D- 284
fluid galactopyranoside
Lucerne
cotyledon 5.0 0.29 Hg2+,45% inhibition at 1mM 252
endosperm 5.0 0.83 Hg2+,42% inhibition at 1mM 252
Malted barley 5.5 0.32 2-amino-2-deoxy-D- 302
mannose (competitive,
KI = 0.18 mM)
Pineapple 3.5 - - 300
Snail 4.0-5.0 6.5 - 286

"The K, values were estimated with respect to p-nitrophenyl P-D-mannopyranoside,

except for the snail enzyme, where phenyl P-Dmannopyranoside was used.

(292) T. Sukeno, A. L. Tarentino, T. H. Plummer, Jr., and F. Maley, Biochemistry, 11,

1493-1501 (1972).
(293) J. Conchie and T. Mann, Nature, 179, 1190-1191 (1957).
(294) T. Muramatsu, Arch. Biochem. Biophys., 115,427-429 (1966).
(295) H. B. Bosmann, Biochim. Biophys. Acta, 258,265-273 (1972).
(296) A. Ohkawara, K. M. Halprin, J. R. Taylor, and V. Levine, Br.1. Dennatol., 87,450-
459 (1972).
(297) G. A. Levvy, A. J. Hay, and J. Conchie, Biochem. J., 91,378-384 (1964).
(298) Y. Hashimoto and J. Fukumoto, Nippon Nogei Kagaku Kaishi, 43,564-569 (1969);
Chem. Abstr., 72,51,167d (1970).
(299) M. Adams, N. K. Richtmyer, and C. S. Hudson,J. Am. Chem. Soc., 65,1369-1380
(1943).
(300) T. T. Li and Y. C. Lee, J. Biol. Chem., 247,3677-3683 (1972).
(301) J. W. Lee and J. A. Ronalds, J . Sci. Food Agric., 23, 199-205 (1972).
(302) C. W. Houston, S. L. Latimer, and E. D. Mitchell, Biochim. Biophys. Acta, 370,
276-282 (1974).
(303) R. Somme, Actu Chem. Scand., 24, 72-76 (1970).
(304) R. Somme, Acta Chem, Scand., 25,759-761 (1971).
(305) T. Sukeno, A. L. Tarentino, T. H. Plummer, Jr., and F. Maley, Methods Enzymol.,
28,777 (1972).
374 PRAKASH M . DEY

There are only limited reports on the substrate specificity of p-D-

mannosidases. Reese and Shibata2,1 showed that the fungal enzyme
hydrolyzes M3 more rapidly than M2.The reduction of M3 to D-
mannotriitol had no effect on the initial rate of hydrolysis, whereas
reduced M, was hydrolyzed much more slowly than M, itself. From
Rhi-zopus n i v e u was
~ ~ obtained
~~ a purified p-D-mannosidase that was
free from p-D-mannanase activity and showed the relative hydrolyses
ofthe substrates as: M4 = M3 > M2 > M5 > M6.
An enzyme preparation from g ~ a r , described
,~~ as eXO-p-D-( 1 4 ) -
mannanase, hydrolyzed ivory-nut mannan almost completely to
D-mannose. The purified (1,000-fold) enzyme from pineapple300hy-
drolyzed the substrates in the order: p-nitrophenyl p-D-mannopy-
ranoside >P-D-Man-(GlcNAc),-Asn (a core glycopeptide) > methyl
P-D-mannopyranoside > p-D-Man-(1+4)-D-Glc > @-&Man-(1 4 ) - D -
mannitol.
A 10,000-fold purified enzyme from hen oviduct305 hydrolyzed
p-D-Man-(GlcNAc),-Asn at one quarter the rate, and P-D-Man-
(GlcNAc), at one-third the rate, of p-nitrophenyl p-D-mannopyrano-
side. The snail enzyme286hydrolyzed the core glycopeptide, as well
as a tri-D-mannose, a-~-Man-( 1+4)-/3-D-Man-( 1+4)-D-Man, but the
latter was treated in the presence of an a-D-mannosidase; here, the
action of p-D-mannosidase seems to follow that of a-D-mannosidase.
Hylin and S a ~ a i reported
~'~ that the crystalline depolymerase
from Leucaena leucocephala converts the galactomannan from the
same source into D-galactose and D-mannose. Hence, it is most likely
that the enzyme preparation contained p-D-mannosidase activity, in
addition to a-D-galactosidase and p-Dmannanase activities.
Reid and Meier202showed the presence of p-Dmannosidase in
the embryo and the endosperm of fenugreek at all stages of germina-
tion. The embryo had a low level of the enzyme, and there was no
significant increase in activity on germination. On the other hand,
the endosperm possessed low activity up to 30 h, after which it
rose steeply and levelled off at 60 h; the rise was approximately
10-fold. This increase coincided with the depletion of the reserve
galactomannan. The authors further concluded from in vivo experi-
ments (using inhibitors of protein synthesis) that production of the
enzyme was controlled by the aleurone layer.
McCleary and Mathe~on,~, showed the presence of two forms of
P-Dmannosidase in the seeds of guar, lucerne, carob, and honey
locust. Form A (low molecular weight) was located in the cotyle-
don-embryo part, whereas (the larger) form B was in the endo-
sperm. They suggested25zthat, although the enzyme activity in these
 BIOCHEMISTRY OF PLANT GALACTOMANNANS 375

seeds is low, it would be sufficient to degrade the D-manno-oligo-

saccharides produced by the action of P-D-mannanase on the endog-
enous galactomannans.
d. Oligo-~-~-mannosyl-(1-,4)-phosphorylase.-In 1970, SOmme3O3
reported that extracts of the germinated, galactomannan-containing
seeds of Trifolium repens, T . pratense, Medicago sativa, Anthyllis
vulneraria, and Lotus corniculatus contain no free D-mannose. The
activity of p-D-mannosidase could not be detected in these extracts,
either with M2 or p-nitrophenyl P-D-mannopyranoside as the sub-
strate. The author doubted303that p-D-mannosidase had a role in the
mobilization of galactomannan in these seeds. She suggested that
phosphorolysis was a possible, alternative pathway. In 1972,
Foglietti and Percheron222were successful in preparing a phosphoro-
1ytic enzyme, namely, oligo-P-D-mannosyl-(1+4)-phosphorylase,
from the germinated seeds of fenugreek. The enzyme [(1+4)-P-~-
mannan:orthophosphate P-D-mannosyl transferase] catalyzes the
following reversible reaction, in which the d.p. of the D-manno-
oligosaccharide must exceed two.

(D-Mannose), + orthophosphate *
(D-mannose),-, + P-D-mannosyl phosphate

The D-mannosyl phosphate produced during this reaction can be

further transformed into glycosyl esters of nucleoside diphos-
p h a t e ~ .With
' ~ ~ the participation of a suitable epimerase, this might
be a possible pathway for the formation of sucrose and starch in
the seed. Courtois and coworkers14shave, in fact, shown a rise in
the levels of UDP-D-glucose, UDP-D-galactose, and GDP-D-mannose
during the germination of fenugreek seeds. Further work from the
same laboratory on fenugreek indicated the presence of a nucleotide
pyrophosphorylase capable of converting the glycosyl esters of nucleo-
side diphosphates into the respective D-hexosyl phosphates, and of an
epimerase that can catalyze the conversion of UDP-D-galactose
into UDP-D-glucose.14'

IV. FUNCTION
Seed galactomannans appear to have a double physiological func-
tion. Firstly, they retain water by solvation (see Ref. 43 for details
on gel formation), and thereby prevent (in regions having high
atmospheric temperatures) complete drying of the seeds which
376 PRAKASH M. DEY

would cause protein denaturation, in particular, the denaturation of

those enzymes essential for seed germination. In this connection,
the D-galactosyl side-branches of the polymer may be regarded as
hydrophilic parts of the molecule. Secondly, the galactomannans
serve as food reserves for the germinating seeds.
Microbial galactomannans whose structures do not conform to
those of leguminous seeds have been the subject of a few special
studies. In Acer plantanoides and A . p s e u d o p l e n t h e n u ~ ,infected
~~~
with Rhytisma acerinum, a strong inhibitor of tobacco mosaic
virus was found. The active compound was water-extractable and
was identified as a galactomannan; this is of interest for future
studies on the possible inhibition of viruses. Being water-soluble,
galactomannans form highly viscous solutions which, on drying,
leave a transparent film that can, therefore, adhere to the surface
of cells, making them impermeable to viruses. These may also
anchor the viruses to the surface of the cell, or immobilize them
on its network structure. Also, a galactomannan from Lipomyces
starkyi307 was shown to have an interferon-inducing ability in cell
cultures. In addition, some microbial galactomannans have been
shown to act as serological antigen^.^^^,^^'
Some plant galactomannans are also known to interact with milk
proteins,310plant lectinsY3" and protein a n t i b o d i e ~ . ~It' ~is not yet
known whether these polysaccharides can exist in the form of com-
plexes with seed proteins. Such complex-formation often results in
increased stability of proteins towards heat inactivation, proteolytic
degradation, and other unfavorable conditions.

(306) M. Gubanski and M. Saniewski, Actu Microbiol. Pol., 13, 227-232 (1964).
(307) L. Borecky, V. Lackovic, D. Blaskovic, L. Masler, and D. Sikl, Acta Virol. Engl.
Ed., 11,264-266 (1967).
(308) I. Azuma, F. Kanetsuna, Y. Tanaka, Y. Yamamura, and L. M. Carbonell,
Mycopathol. Mycol. Appl., 54, 111-125 (1974).
(309) W. Lee and K. 0. Lloyd, Arch. Biochem. Biophys., 171, 624-630 (1975).
(310) A. S. Ambrose, U.S.Pat. 1,991,189 (1935); Chem. Abstr., 29,2255 (1935).
(311) J. P. Van Wauwe, F. G . Loontiens, and C. K. DeBruyne, Biochim. Biophys. Actu,
313,99-105 (1973).
(312) M. Heidelberger, J. Am. Chem. SOC., 77,4308-4311 (1958).
 BIBLIOGRAPHY OF CRYSTAL STRUCTURES
OF POLYSACCHARIDES
1975

BY PUDUPADI R. SUNDARARAJAN AND ROBERT H. MARCHESSAULT

Xerox Research Centre of Canada, Mississauga, Ontario L5L 119, Canada;

Department of Chemistry, University of Montreal, Montreal, Quebec H3C 3V1, Canada

I. Introduction .......................................................... 377

11. Amylose and Other wD-Glycans.. ...................................... 378
111. Cellulose and Other P-D-Glycans ....................................... 379
IV. Glycosaminoglycans (Amino Polysaccharides) ........................... 381

I. INTRODUCTION
In Volume 33 of this Series, we presented’ a review of the crystal-
line structures of polysaccharides published during the period 1967-
1974. Detailed accounts of progress in structural studies on specific
types of polysaccharides were presented in the Proceedings of the
Twenty-sixth Symposium of the Colston Research Society and were
subsequently published as a book.2 Precise methods for X-ray diffiac-
tion analysis of biopolymer structures were discussed by H ~ k i n sThe
.~
aspects of the structures of cellulose, mannan, and xylan, their organi-
zation in the cell wall, and the biosynthesis of cell-wall polysaccha-
rides were described by M a ~ k i e .Work
~ on the structures of the
connective-tissue polysaccharides, 0-acetylcellulose, and the various
forms of amylose was reviewed by at kin^,^ Chanzy,6 and Sarko,’

(1) R. H. Marchessault and P. R. Sundararajan, Ado. Carbohydr. Chem. Biochem.,

33,387-404 (1976).
(2) “Structureof Fibrous Biopolymers,” E. D. T. Atkins and A. Keller, eds., Buttenvorth,
London, 1975.
(3) D. W. L. Hukins, Ref. 2, pp. 293-305.
(4) W. Mackie, Ref. 2, pp. 391-416.
(5) E. D. T. Atkins, Ref. 2, pp. 35-45.
(6) H. D. Chanzy, Ref. 2, pp. 417-434.
(7) A. Sarko, Ref. 2, pp. 335-354.

377
378 P. R. SUNDARARAJAN AND R. H. MARCHESSAULT

respectively. An excellent review by Prestons contains an account of

X-ray diffraction and conformational studies on agar, alginic acid,
carrageenan, cellulose, mannan, pectic acid, and xylan. Two other
brief reviews on the structural features of cellulose have also been
publi~hed.~*’~
The crystal-structure data reported in 1975 on polysaccharides are
given in this article. In addition to the unit-cell dimensions, the signif-
icant features of the structures are described. Unless specified other-
wise, the chain axis is along the c direction of the unit cell. The
helical symmetry is denoted by n(&h),where n specifies the number
of repeat units per turn, and h is the projected height, in nanometers,
of the repeat unit onto the helix axis. A positive h denotes a right-
handed helix, and a negative h, a left-handed helix. As before,’ in the
title to each abstract, a common name or descriptive title for the poly-
saccharide described is given on the left, and the chemical formula on
the right.

11. AMYLOSEAND OTHERWD-GLYCANS

1. V-Amylose” POly[ ( 1-*4)-a-D-Glcp]

The dehydrated form crystallizes in an orthorhombic unit-cell with

a = 1.292 nm, b = 2.24 nm, and c = 0.795 nm, with two 6(-0.133)
chains per unit cell. The space group is P2,2,2,. Models constructed
with “residue 3” of the cyclohexaamylose structure12were best suited.
Two intrachain hydrogen-bonds, Of 2---0i+1-3of length 0.275 nm, and
0f6---0i-6-2 of length 0.284 nm, were proposed. Intermolecular hydro-
gen-bonds were ruled out. The final error function was 66%. The re-
sults were compared with those of Winter and Sarko.I3

2. Mycodextran (nigeran)l 4 Poly[(1+3)-a-~-Glcp-(1+4)-a-~-Glcp]

A method was presented for recording electron diffraction diagrams

of hydrated, crystalline biopolymers, and was applied to mycodextran.

(8)R. D. Preston, Phys. Lett. C, 21, 185-226 (1975).

(9)H. Sihtola and L. Neimo, in “Symposium on the Enzymatic Hydrolysis of Cellu-
lose,” M. Bailey, T. M. Enari, and M. Linko, eds., Biotechnical Laboratory of the
Technical Research Centre of Finland, Helsinki, Finland, 1975,pp. 9-21;Chem.
Abstr., 83,207,743d(1975).
(10)G.La1 and A. Pande, Colourage, 22,21-27 (1975).
(11)V. G.Murphy, B. Zaslow, and A. D. French, Biopolymers, 14,1487-1501 (1975).
(12)A.Hybl, R. E. Rundle, and D. G. WilliamsJ. Am. Chem. Soc., 87,2779-2788(1965).
(13)W. T.Winter and A. Sarko, Biopolymers, 13, 1447-1460 (1974).
(14)K.J. Taylor, H. Chanzy, and R. H. Marchessault,j. Mol. Biol., 92,165-167(1975).
 BIBLIOGRAPHY OF CRYSTAL STRUCTURES 379

With only the lattice water remaining, and at - lo”,the diffiactogram of

the single crystals of mycodextran contained 50 independent reflec-
tions, which were indexed with a = 1.76 nm, b = 0.735 nm, and y =
90”. On annealing in the range of 100 to 160”,a dry form was obtained,
with a = 1.76 nm, b = 0.51 nm, and y = 90”. It was suggested that the
water of hydration is of the sheet type, alternating along the b axis of
the crystal. On dehydration, the development in the lamellar crystals
of cracks in the direction normal to the b axis was observed. From the
difference in the unit-cell volume between the hydrated and dehy-
drated forms, it was deduced that there are two water molecules per
D-glUCOSe residue in the “hydrate.”

111. CELLULOSEAND OTHERP-D-GLYCANS

1. Cellulose1s Poly[( 1-*4)-p-~-Glcp]

Sodiocellulose I1 crystallizes in a hexagonal unit-cell, with

a = 1.0 nm and c = 1.51 nm. The meridional reflection on the third
layer-line led to a three-fold helical structure. The unit cell contains
two 3(-0.503) chains of sodiocellulose, six sodium ions, six hydroxyl
ions, and at least six water molecules. Conformational calculations and
packing analysis of the chains led to a structure which is similar in
several features to that of poly[(1+4)-p-~-Xylp].’~

2. Cellulose’7 Poly[(l+ 4)-p-~-Glcp]

X-Ray data on regenerated cellulose (rayon) led to a unit cell with

a = 0.801 nm, b = 0.904 nm,c = 1.036 nm, and y = 117.1”.The space
group is P21. An antiparallel arrangement of the chains, with a relative
stagger of 0.216c, was proposed. The torsion angle x(0-5, (2-5, C-6,
0-6) is different for the “up” and “down” chains. The “down” chain
has an 0-2’---0-6 intrachain hydrogen-bond. The intermolecular
hydrogen-bonds are: (i) 0-6-03, parallel to the a axis, between adja-
cent “down” chains; (ii) 0-6-0-2 between adjacent “up” chains along
the a axis, and (iii)0-2---0-2’ between the “up” chain at 1,0,0 and the
“down” chain at the center.
3. Cellulose’* Poly[(1+4)-P-~-Glcp]
(15)P. M. Whitaker, I. A. Nieduszynski, and E. D. T. Atkins,PoZymer, 15,125-127 (1974).
(16) I. A. Nieduszynski and R. H. Marchessault, Biopolymers, 11, 1335-1344 (1972).
(17) F.J. Kolpak and J. Blackwell, Macromolecules, 8,563-564 (1975).
(18)A.Koura, B.Philipp, H. Schleicher, and W. Wagenknecht, Fuserforsch. Textiltech.,
26, 514-515 (1975);Chem. Abstr., 84, 46,345d (1976).
380 P. R. SUNDARARAJAN A N D R. H. MARCHESSAULT

Aqueous solutions of guanidine cause a structural change in cellu-

lose midway between that produced by aliphatic amines and by alkali
hydroxides. The lattice change to a cellulose I1 structure, which oc-
curs with aqueous alkali hydroxide above a certain concentration,
does not occur with guanidine. This was attributed to N-H---0 bridges,
with the participation of guanidinium ions.

4. C e l l ~ l o s e ' ~ POly[( 1+4)-P-D-Clcp]

The ratio of the intensities of the 020 and 040 reflections for cellu-
lose I11 from various sources leads to two classes. Cellulose I11 from
cellulose I has prominent reflections at 0.247 nm on the first layer-line,
and 0.28 nm and 0.223 nm on the second. For cellulose I11 from cellu-
lose 11, these are at 0.257 and 0.236 nm on the first layer-line, and at
0.258 and 0.235 nm on the second. It was suggested that cellulose I11
from cellulose I and I1 be termed Cell 1111and Cell IIIn, respectively.
The cellulose IV from cell 1111gave reflections at 0.227 nm on the first
layer, and at 0.305 and 0.289 nm on the second layer. For cellulose IV
from Cell 11111and cellulose 11, the reflections are at 0.251 and 0.298 nm
on the first and second layers, respectively. It was suggested that the
crystalline modifications of cellulose be classified into two families:
Cell I family (I, 1111,and IVJ and Cell I1 (11, IIIn, and IVII).A member
of the Cell I1 family cannot be transformed into one of the Cell I family.

X-Ray diagrams of O-p-tolylsulfonylcellulose prepared from cotton

slivers showed an amorphous pattern, with a broad maximum at 20
= 21",showing that p-toluenesulfonylation of cotton is an intrafibrillar
reaction causing the breakdown of the initial cellulose I lattice. In the
initial stages of reaction of O-p-tolylsulfonylcellulose with an anhy-
drous mixture of potassium fluoride and ethylene glycol at 190", there
is a recrystallization into cellulose I1 lattice. The extent of crystalliza-
tion of cellulose improves with increasing unsaturation up to a certain
stage, beyond which the unsaturated cellulose crystallizes in a new,
highly crystalline phase.

6. Cellulose2' Poly[( 1+4)-P-D-Gkp]

(19) J. Hayashi, A. Sufoka, J. Ohkita, and S. Watanabe, J . Polym. Sci., Polym. Lett.
Ed., 13,23-27 (1975).
(20) H . C. Srivastava, A. K. Kulshreshtha, and V. K. Srivastava,]. Polym. Sci., Polym.
Lett. Ed., 13,65-70 (1975).
(21) W. Herth, A. Kuppel, and W. W. Franke,]. Ultrastruct. Res., 50,289-292 (1975).
 BIBLIOGRAPHY OF CRYSTAL STRUCTURES 38 1

The alkali-resistant material from the cyst walls of the green alga
Acetabularia mediterranea was studied by X-ray and other physical
methods. The predominant, structural polysaccharide was found to be
cellulose I. The stalk and cap walls of the alga contain poly[(l+4)-
P-D-Manp]. It is possible that the change from mannan to cellulose in
Acetabularia is paralleled by a transition from the diploid to the hap-
loid stage.

7. O-AcetylpachymanzZ POly[(I+ 3)-A~-p-~-Glcp]

The X-ray pattern from O-acetylpachyman derived from the fungus

Porio cocos can be indexed either with a hexagonal unit-cell of dimen-
sions a = b = 1.149 nm, c = 1.86 nm, or with an orthorhombic cell
having a = 1.149 nm, b = 2.013 nm, c = 1.86 nm. The presence of both
5th- and 6th-order meridional reflections led to the detection of two
polymorphs. Films stretched 25 to 50% at 125" led to a unit cell with
a = b = 1.099 nm, c = 2.238 nm. Upon further stretching to 300% at 215",
the repeat distance was diminished to 1.86 nm. Both polymorphs con-
tain six-fold helices. Conformational and packing analyses with both
hexagonal and orthorhombic unit-cells showed that there is parallel
packing of the helices, which are right-handed. The mode of transition
from one polymorph to the other was discussed.

IV. GLYCOSAMINOGLYCANS
(AMINO POLYSACCHARIDES)

1. a-Chitinz3 Poly[(1+4-)-p-~-GlcNAcp]

The structures of chitins derived from shrimp and various kinds of

crabs are all very similar. For chitin from King crab, the unit cell is
orthorhombic, with a = 0.47 nm, b (fiber axis) = 1.05 nm, and c = 1.03
nm. The molecule does not take up water, and the d spacings do not
vary on drying or soaking.

2. p-ChitinZ4 PoIY[(1+ ~ ) - ~ - D - G I c N A c ~ ]

The structure of p-chitin from the pogonophore Oligobrachia

ivanovi was refined by using X-ray data and a packing of the 2(0.519)
chains in a monoclinic unit-cell with a = 0.485 nm, b = 0.926 nm, c =

(22) T. L. Bluhm and A. Sarko, Biopolymers, 14, 2639-2643 (1975).

(23) B. L. Averbach, Report MITSG 75-17, National Technical Information Service, U. S.
Department of Commerce, 1975.
(24) K. H. Gardner and J. Blackwell, BiopoZymers, 14, 1581-1595 (1975).
382 P. R. SUNDARARAJAN AND R. H. MARCHESSAULT

1.038 nm, and y = 97.5". The space group is P2,. The analysis indicated
"parallel" chain-packing in a sheet arrangement, with an 0-3'---0-5
intramolecular hydrogen-bond (0.275 nm), an N-H---0-7 intrasheet
hydrogen-bond (276 pm), and an 0-6'---0-7 intrasheet hydrogen-bond
(0.289 nm). The intersheet 0-6-0-7 hydrogen-bond proposed pre-
vious1yz5was found to be unacceptable. The R factor is 26.7%. The
swelling properties were discussed on the basis of the present struc-
ture, and comparison was drawn between the structures of chitin and
~ellulose.'~

3. ChitosanZ3 POly[(~ + ~ ) - P - D - G ~ c N H ~ ~ ]

Crystalline, flake chitosan from King crab exhibited patterns similar

to those of chitin, but the 002 peak (b is the fiber axis) shifted from
0.962 nm for chitin to 0.857 nm for flake chitosan. The position of this
peak depends on the content of water and is lowered to 0.745 nm on
drying at 134". It was postulated that water molecules that enter the
lattice are loosely bound between the chains along the 001 direction.

4. Heparinz6 Poly[( 1+4)-a-~-GlcNSO~-p-6SO~--( 1+4)-P-D-GlcpA-

SO3--(1+ ~ ) - C Y - D - G ~ C N S O ~ - ~1+
- ~ S~O ~-L
)-Q - (-
Idop A-2SO3-]

The sodium salt of macromolecular heparin occurs with a periodic-

ity of 1.73 nm if the relative humidity (r.h.) is less than 78%. This
pattern can be indexed with an orthorhombic unit-cell. If the r.h. is
increased to 84%, the periodicity decreases, irreversibly, to 1.65 nm.
Two geometrically equivalent, triclinic unit-cells were proposed, with
a = 1 . 3 7 n m , b = 1 . 1 0 n m , c = 1 . 6 5 n m , a = 116",/3=70",y=123";and
w i t h a = 1 . 2 0 n m , b = 1 . 1 0 n m , c = 1 . 6 5 n m , a = 116",/3=9Oo,y=73",
with one tetrasaccharide chain segment per unit cell. The dimensions
are subject to variations in the r.h. The structure consists of sheets of
chains, all at the same relative translation, adjacent sheets being dis-
placed by 0.48 nm along the fiber axis. The merits of the models with
"C,(L) and l C 4 ( ~conformations
) for the L-idosyluronic acid residues
were discussed.

5. Heparin2' Poly[( 1~4)-a-D-GlcNS03-p-6S03--( 1-*4)-P-~-GlcpA-

SOa--( ~ + ~ ) - C Y - D - G ~ C N S O ~ - ~1- +
~ 4S )O- a~--~- (-
Id0p A-2S03-1

(25) N. E. Dweltz, J. R. Colvin, and A. G. McInnes,Can.J. Chem., 46,1513-1521 (1968).

(26) I. A. Nieduszynski and E. D. T. Atkins, Ref. 2, pp. 323-334.
(27) E. D. T. Atkins and I. A. Nieduszynski, Ado. E x p . Med. Biol., 52, 19-37 (1975).
 BIBLIOGRAPHY OF CRYSTAL STRUCTURES 383

At a relative humidity of 80%, the sodium salt of pig-mucosal

heparin shows a fiber repeat of 1.65 nm, with a triclinic unit-cell
containing one chain. The sodium salt of macromolecular heparin
from rat skin, at 78% r-h., showed a repeat of 1.73 nm, and crystal-
lized in an orthorhombic unit-cell. After 24 hours at 84% r.h., the
repeat changed to 1.65 nm, the pattern corresponding to that of pig-
mucosal heparin. The stereochemistry of the chain in terms of the
chain repeat and 4, $ maps was discussed. It was suggested that the
chemical formula for heparin is poly[( 1+4)-p-D-GlcpA-( 1-4-p-D-
GlcNAcpl, and a scheme of 5-epimerization of PD-glucuronic acid to
a-L-iduronic acid was discussed.

6. Heparan sulfate2* Poly[(1+4)-a-~-GlcNAcp-GSO,--(1+4)-p-D-

GlcpA-(~ + ~ ) - ~ - D G ~ c N A c ~ - 1+4)-
~SO~--(
a-L-IdopA]

The sodium salt of heparan sulfate crystallizes in an orthorhombic

unit-cell with a = 1.18 nm, b = 1.10 nm, and c = 1.86 nm. The barium
and strontium salts exhibit the same value of c. The calcium salt of
heparan sulfate (human aorta) crystallizes in a unit cell with a = 1.70
nm, b = 1.27 nm, and c = 1.68 nm. The space group is P2,2,2,. Two
parallel-stranded, double helices of opposite polarity pass through the
unit cell. Each strand has a pitch of 3.36 nm. The chain conformations
were interpreted in terms of a tetrasaccharide having alternating a - ~
(1-4)- and fi-~-(1*4)-linkages, with all the residues in the “C,(D)
conformation.

7. HyaluronateZ8 Poly[( 1+4)-P-D-GlcpA-(1*3)-fi-~-GlcNA~p]

Sodium hyaluronate crystallizes with a = b = 1.17 nm, and c = 2.85

nm. The 3(-0.95) helix is favored. The space group is P3,21, with
antiparallel chains, and with two hexasaccharide segments per cell.
On annealing, the dimensions change to a = b = 1.87nm, and c = 2.85
nm, with six hexasaccharide segments in the unit cell. A new ortho-
rhombic form is found, with a = 3.44 nm, b = 1.17 nm, and c = 2.85 nm.
Several two-dimensional packing-arrangements were discussed.

8. HyaluronateZ8 POly[(1+4)-P-~-GlcpA-(1- 3)-p-~-GlcNAcp]

For calcium hyaluronate, at 50 to 60% r.h,, the unit-cell dimensions

are a = b = 1.54 nm, and c = 2.85 nm; at 90% r.h., the cell enlarges to
(28) J. K. Sheehan, E. D. T. Atkins, and I. A. Nieduszynski,J. Mol. B i d . , 91, 153-
163 (1975).
384 P. R. SUNDARARAJAN A N D R. H. MARCHESSAULT

a = b = 1.62 nm, and c = 2.85 nm. After annealing for three weeks
under humid conditions, this further changes to a = b = 2.04 nm, and
c = 2.85 nm. The 3(-0.95)helix is favored. It was suggested that there
is a cooperative interaction between the threefold helices and a net-
work of water molecules that are hydrogen-bonded to one another and
to the hyaluronate chains. Various two-dimensional packing-schemes
were discussed.

9. H yaluronic acidzB Poly[( 1+4)-P-~-GlcpA-(1-.3)-P-D-GlcNAcp]

Sodium hyaluronate, oriented at 40”and 90% r.h., crystallizes in a

trigonal unit-cell with a = b = 1.17 nm, and c = 2.85 nm. Some pat-
terns also showed the presence of a second phase,30having c = 3.39
nm. Two 3(-0.95) helices having an antiparallel arrangement pack in
the space group P3z21. In addition, there are about 9 sodium ions,
three chloride ions, and 31 to 35 water molecules per unit cell, and
one calcium ion for every 9 unit cells. There are two intramolecular
hydrogen-bonds, 0-3-0-5’ (0.267 nm), between the residues in
(1+3’)-linkage, and 0-4-0-5’ (0.285 nm), between the residues in
(1+4’)-linkage. Extensive, interchain hydrogen-bonds through water
molecules were found. The acetamido groups are not involved in
hydrogen bonding. Each sodium ion is located at the center of a dis-
torted, octahedral, coordination shell involving a carboxylate oxygen
atom from one helix, 0-2 (GlcpA) and 0 - 7 (GlcNAcp) from the related
helix, and three water molecules. The R factor is 23%. Comparison was
made with other polymorphs of h y a l ~ r o n a t e . ~ ~ ~ ~ ~

10. Hyaluronic acid30 Poly[(1+4)-@~-GlcpA-(1+3)-P-~-GlcNAcp]

Sodium hyaluronate crystallizes in a tetragonal unit-cell, with a =

b = 0.989 nm, and c = 3.394 nm. The space group is P432,2. Two
4(-0.85) helices pass through the unit cell, antiparallel to each other.
The cell contains 8 sodium ions, and no water molecules. Intrachain
hydrogen-bonds of the type NB-H---OA-6(0.278 nm) and OB-4---OA-5
(0.253 nm) were proposed (A = GlcpA, B = GlcNAcp). Adjacent
chains along the cell edge are related by hydrogen bonds of the type
OB-6---OB-7(0.279 nm) and OB-6---OA-2(0.237 nm). The antiparallel
chains are bridged by an OA-3---OA-6(0.259 nm) hydrogen-bond. In
addition, octahedrally coordinated sodium ions link the chains through
(29) W. T. Winter, P. J. C. Smith, and S. Amott,]. Mol. Biol., 99, 219-235 (1975).
(30) J. M. Guss, D. W. L. Hukins, P. J. C. Smith, W. T. Winter, S. Arnott, R. Moorhouse,
and D. A. Rees,]. Mol. Biol., 95,359-384 (1975).
 BIBLIOGRAPHY OF CRYSTAL STRUCTURES 385

O---Na+---0 bridges. No double-helix model, as originally proposed

for this structure, has been found to be in acceptable agreement with
the observed data. The R factor is 29%.
At higher r.h., for example, 75%, only the a dimension increases,
so that the new unit cell has a = 1.153 nm, b = 0.989 nm, and c =
3.386 nm. The space group in this case is P212121.The chain is a
2(1.693) helix, with tetrasaccharide repeat-units. The intrachain
hydrogen-bonds are 0$-4---0;4+1-5 (0.269 nm), NE1---Ok1-6(0.257 nm),
og14---Of-5(0.292 nm), and Np---0:-6 (0.295 nm). Whereas the OB-
6---OA-2interchain hydrogen-bond and those between the corner and
central chains are the same as in the tetragonal form, the other inter-
chain hydrogen-bonds and the Na---0 bonding are replaced by hy-
drogen bonds by way of water molecules, which form extensive inter-
chain bridges. The R factor is 29%.
A third crystalline form, from rooster-comb sodium hyaluronate, at
high r,h., also crystallizes with a = 1.162 nm, b = 0.984 nm, and c =
3.331 nm. Meridional reflections are on every layer line, indicating
a perturbation of the 4-fold helix symmetry.
This Page Intentionally Left Blank
 AUTHOR INDEX FOR VOLUME 35

Numbers in parentheses are reference numbers and indicate that an author’s work is
referred to although his name is not cited in the text.

A Amagaeva, A. A., 52, 54(71), 56(71),

57(73), 80(71, 73)
Abe, Y., 154
Achenbach, H., 89,92,122 Ambrose, A. S., 376
Acree, T. E., 33 Amen, K.-L., 70
Adair, W. L., 275,276(666), 322(666), Anagnostopoulos, C., 348, 363(117)
Andersen, B. R., 178
323, 324, 331(666)
Anderson, A. J,, 147
Adams, G. A., 349
Anderson, B., 253
Adams, M., 372(299),373
Anderson, E., 344,348(59)
Agrawal, B. B. L., 137, 138(134, 135),
Anderson, L., 102
140(135),151(134,135), 152(261,
Anderson, R. L., 232
262), 153, 155(262),162, 171(134,
Andoh, T., 346,367,370(265)
135),179(135),335
Andrews, A. T., 130
Agrawal, K. M. L., 148, 363, 365(233),
370(233),372(235) Andrews, E. P., 319
Andrews, P., 348(92, 93), 352(92)
Ahlgren, E., 366
Ansell, N., 279, 280(673)
Ahmad, A., 177, 272, 275(659), 276,
Anstee, D. J., 131
277(669)
Aranda, L. H., 132
Ahmed, A. I., 269
Archibald, A. R., 176
Ahmed, 2.F., 344,347
Amott, S., 384
Aida, K., 307,308(778)
Aro, H., 294, 295(711), 335(711)
Ainsworth, C. F., 151, 152(262a),
Asai, M., 96
154(262a),156, 159, 160, 161,
Ashwell, G., 131
335(262a, 324)
Aspberg, K., 137, 138(137), 304(137),
Akedo, H., 157, 160,324
306(145)
Akiya, S., 250,251(594), 253(594)
Aspinall, G. O., 343,345,346(85),
Akiyama, Y.,300,312, 324(740, 741)
347(85), 371
Alam, M., 345
Atkins, E. D. T., 377,379, 382, 383(26),
Albarsheim, P., 147
384(28)
Alberto, B. P., 172
Atwood, K. C., 244
Allan, D., 294,325
Aub, J. C., 130,317(28,29)
Allen, A. K., 136, 143(128),144(128,
207), 195, 199(213),200(213), Auger, J., 325
Aull, F., 139,306(155), 307(155),329,
203(213),204(213),205(213),211,
212(207), 213,221, 233,309(213), 330(868)
336(213),337(128, 207, 213), Averbach, B. L., 381
338(213),339(128, 207, 213) Avrameas, S., 139,317
Axelsson, K., 282,291(682)
Allen, H. J., 138, 203, 205, 337(467a)
Azavache, V., 166
Allen, L. W., 293,294(708, 714),
Azuma, I., 376
295(708),335(708, 714)
Allen, N. K., 132
Alter, G. M., 156, 158, 161(310)
B
Alving, C. R., 219 Babczinski, P., 355
Alzer, G., 299,301(735), 302(736) Bachhawat, B. K., 141, 168(193), 177,

387
388 AUTHOR INDEX, VOLUME 35

179(193),272,274(193), 275(659), Bausch, J. N., 267, 333

276,277(669) Bazhanova, E. T., 54, 55(77), 80(i'7)
Bachur, N. R., 138, 296(154),301 Beattie, G., 318
Baddiley, J., 98, 99(37), 354 Beaugiraud, S., 367,372(260)
Baenziger, J., 186, 298(432),300(432), Beck, M. L., 206,338(470)
332 Beck, S., 291
Bagchi, P., 343,346(53), 348(53) Becker, J. W., 137, 138(140), 152,
Bahl, 0. P., 148, 363, 365(233), 370(233), 153(325), 154(268,269, 275, 276),
372(235) 155(267), 156(268,275), 157(267,
Bailey, P. J., 366 268), 158(267,268, 280), 161(280),
Bailey, R. W., 172,343, 345, 356(57), 202(140), 203(140), 205(140),
363(57) 304(276), 335(267),336(140),
Bains, G., 343 337(140), 339(140)
Bajpai, K., 108 Beckman, L., 300
Baker, B. R., 58 Beevers, H., 149
Baker, C. W., 349,350(128) Behrens, N. H., 354
Baker, E. A., 363 Beitsch, D. D., 315
Balasubramanian, K. A., 177 Ben-Bassat, H., 150
Balding, P., 132,250, 252, 253, 254(599), Benz, G., 342
316, 339(599) Beppu, M., 162, 163(332),165(332)
Baldo, B. A., 282, 314, 315(800, 807,808) Bergel'son, L. D., 68
Balint, G. A,, 270 Bernadac, A., 165
Ballantyne, M. J., 348 Bernhard, W., 317
Ballas, A., 292 Bessler, W., 140, 159(174), 160, 164, 178,
Ballou, C. E., 354 179(174),228(410),246(410),
Bamberger, E. S., 363 247(410), 248(342),340(589)
Banks, J. R., 219 BBtail, G., 179, 197, 199(428),210,
Barber, B. H., 156, 158, 159(311), 160 339(428)
Barber, G. A., 353 Bettman, B., 36
Barker, B. E., 309,311 Beveridge, R. J., 347
Barker, R., 136,337(130) Bewley, J. D., 366,370(254), 371(254)
Barker, S. A., 49, 63(64),65, 172, 354 Beychok, S., 140, 160, 166(170),
Barham, D., 344,345(58) 168(319), 176, 179(170), 180(170),
Barkhan, P., 292 181(170), 184(170),338(170)
Barlow, C. B., 67, 102 Bezer, A. E., 176,270,271(644)
Baron, D., 101 Bezkorovainy, A., 281,282(679), 335(679)
Baron, J. M., 294(714),295,335(714) Bhatia, D., 343
Barondes, S. H., 139, 290(158), 306(159), Bhatia, H. M., 239,244,245,246
307(159),308,309(779,780), 325(159), Bhavanandan, V. P., 219
336(158),337(158) Bielskis, E., 341
Barr, R. M., 355 Biely, P., 178
Barrett, J. T., 131 Biemann, K., 47,48(56), 55(56), 65(56),
Bartholemew, B. A., 372,373(284,290) 67(56), 68,69(56)
Barton, L., 39 Bierry, H., 366
Basarab, O., 354 Bird, G. W. G., 129, 132(10), 134, 140(10,
Basham, T. Y., 310 104), 143, 144(206),202(68), 210,
Basu, S., 264 211(68, 206), 212(206), 213(206), 218,
Batrakov, S. G., 68 224(502), 226(508d), 227, 257,
Bauer, H., 174, 175(388) 258(201), 258(201),267, 268(632),
Bauer, s., 178 269(632), 276(104, 119, 206),
Baues, R. J., l37,262(149b) 297(632), 300(632),307, 311(132,
 AUTHOR INDEX, VOLUME 35 389

510),333(632),338(201,510,511,512, Bowie, R. A., 34,35(13),42(13), 78(13)

616, 617), 339(206, 632) Bowles, D. J., 139
Birdsell, D. C., 162, 177(330) Bowman, C. M., 32, 35(9), 36(9), 38(9),
Bishayee, S., 177 41(9), 50(9),52(9), 77(9),78(9)
Bishop, C. T., 353 Boyd, W. C., 128, 129, 131, 132(2, 5),
Bittiger, H., 150 134(62), 140, 147(2,4),202, 224, 226,
Bjork, I., 240,241(569), 243(569), 239(62), 243, 244(2, 3, 5), 245(575),
335(569) 246,248(103), 253,258,259(614),
Bjorndal, H., 282, 291(682), 349 277(13), 289(12, 13), 313, 339(2, 5,
Black, M., 360 12, 13)
Black, N. H., 342 Bragg, P. D., 349
Blackwell, J., 379, 381, 382(17) Branch, G. E. K., 36
Blake, C. C. F., 221 Branstrator, M. L., 231
Blaskovic, D., 376 Braun, C., 250,253(593), 339(593)
Blaustein, J., 137, 138(146), 142, Bretscher, M. S., 323
144(146),270(146, 194), 271(146, Bretthauer, R. K., 354
194), 272(146),273(146), Bretting, H., 315,316
274(146), 293(194), 3 17(194), Brewer, C. F., 155, 156, 157, 158,
336(146), 337(146, I%), 339(146, 161(307,308), 163, 179(308),
1%) 184(429),201(308)
Bloch, R., 137, 138(133),214(133), Brill, W. J., 148
215(133), 224(133), 313 Brilliantine, L., 132
Bluhm, T. L., 381 Brillinger, G. U., 83, 84(8)
Blumberg, M., 233,234(541), 283,284, Brimacoinbe, J. S., 48, 52(58), 54(58),
335(684) 55(60), 57(58), 74(58, 59, 60), 75(58),
Blumson, N. L., 98,99(37) 85, 89, 95(13)
Boeseken, J., 33 Brodie, G. N., 141
Bohlool, B. B., 147 Brown, H. C., 58
Boldt, D. H., 219 Brown, J. C., 329
Bomchil, G., 132 Brown, J. M., 224
Bonner, T. G., 77(117, 118, 119, 120), 78 Brown, M. C., 303, 304(762), 330(762),
Borberg, H., 297 33l(762)
Borecky, L., 376 Brown, R., 131, 134(62),239(62)
Borjeson, J., 139,292,309,310 Brown R. D., 111, 156, 157, 179, 184(429)
Borodulina-Shvetz, V. I., 38, 51(28), Brownhill, L. E., 309
54(28), 80(28) Browning, W. C., 342
Bosmann, H. B., 372(295), 373 Bruce, G. T., 172
Bottger, M., 239,240(563) Bruce, R. M., 107
Bouchard, P., 214,215(490b), 219(494), Bruce, W. R., 130, 150(33)
222(494), 223(494) Briicher, O., 297
Boundy, J. A., 168, 174(363) Brunngraber, E. G., 326
Bourne, E. J., 35, 38,42(23), 45, 50(23), Buchala, A. J., 345
51(23), 53(23,48),54(23), 55(23, 48), Bullis, C. M., 329
57(23), 58, 60(86), 72(23,48), 77(23, Buonassisi, V., 168, 301(366a)
117, 118, 119, 120, 121), 78(23), 172 BureS, L., 193, 198, 201(446)
Bourquelot, E., 370 Burger, M. M., 130, 132, 133, 136, 137,
Bourrillon, R., 139, 145,227,228(518), 138(133), 165,212(129), 214(129,
230, 231(523),298, 303, 311(160, 133), 215(129, 133,487), 216,218,
523, 731), 320(731), 322(731), 219,224(133), 313,317(30,492),
331(731),335(519), 337(220) 318(92a), 328, 329(865, 866),
Bouveng, R., 292 337(129), 339(30), 340(129)
390 AUTHOR INDEX, VOLUME 35

Burke, G. C., 296 Child, T. F., 349

Burton, H., 341(9), 342 Chin, P. S., 48, 54(61),62(61), 79(61)
Bush, D. A., 174, 175(388) Chittenden, G. J. F., 354
Butte, J. C., 122 Chiu, T. H., 355
Bywater, R., 138 Chopra, A. K., 49
Choudhury, D., 343,346(53), 348(53)
C Chowdhury, T. K., 132, 139(94)
Choy, Y. M., 344, 345(65),346(65),
Cabib, E., 354 350(91), 351(91)
Cacan, M., 234,235(546), 236, 335(546) Chrisp, D. J., 342
Cacan, R., 234,235(546), 236,335(546) Christensen, T. B., 270, 271(651),
Caccam, J. F., 355 272(651),335(651),336(651),
Callies, Q. C., 178 337(651)
Callow, J. A., 132 Christie, D. J., 156
Caprioli, R. M., 58 Christner, J. E., 224
Carbonell, L. M., 376 Chuba, J. V., 268(637),269, 291,
Carchon, H. A., 140, 195(182),197(182), 339(637)
199(182),200(182) Chudzikowski, R. J., 342
Carlson, W. A,, 341(11),342 Cifonelli, J. A., 134, 140(117),151,
Carlsson, H. E., 222 166(117),168(353,354, 355),
Carter, W. A., 141, 168(192), 178(192), 169(353, 355), 171(260,355),
179(192,418) 173(353),179(117),187(422)
Carter, W. G., 228,229(520, 521), 234, Clardy, J., 96
235(546),335(520, 521, 546) Clark, A. E., 134, 135(116), 139(116),177
Carver, J. P., 156, 158, 159(311),160 Clegg, R. M., 216
Catsimpoolas, N., 233 Clermont, S . , 360
Cawley, L. P., 148,267, 268, 269(635), Clermont-Beaugiraud, S., 366,367(255),
339(631,635) 368
Cazal, P., 132,222, 284(81), 289, 339(81) Cline, M. J., 317
Celano, M. J., 302,339(750) Closs, O., 276
Cerezo, A. S., 344,346(67,68), 348(67, Coapes, H. E., 176
68), 351(67) Coats, J. H., 119
Cermakova, M., 199 Codington, J. F., 275, 303, 304(762),330,
Chabanier, A. M., 197 331
Chandra, G. R., 360 Cohen, D., 360
Chanzy, H., 377,378 Cohen, E., 132, 150(132),307
Chapman, H. R., 341(9), 342 Colburn, P., 168,301(366a)
Charpentier, M., 367 Colvin, J. R., 382
Chase, P. S., 178 Conard, R. A., 294,335(710), 338(710)
Chassy, B. M., 123 Conchi, J., 372(293,297), 373
Chattoraj, A,, 244 Condos, R. G., 111
Chauser, E. G., 38,45(24), 54(24),72(24), Connett, S. L., 131,308(51)
80 Conrad, H. E., 349
Chen, C.-C., 130,254(32), 271(32) Cooper, A. G., 303,304(762), 330(762),
Chen, S. S. C., 371 331(762)
Cherneva, E. P., 55 Cooper, D. J., 110
Chessin, L. N., 139,309(162), 310 Cooper, F. P., 353
Cheung, G., 312 Cooper, H. L., 309
Chien, S. M., 140,251(175),252(184), Corcoran, J. W., 84, 122
253(184),337(184), 339(175, 184, Cote, M.N., 130,317(29)
601) Cottom, G. L., 156
 AUTHOR INDEX, VOLUME 35 391

Coulet, M., 131, 179, 197, 199(428),210, 350(43), 356(43), 363(43), 364(43),
339(428) 369, 375(43)
Courtois, J. E., 344, 345, 347, 348(72, Dean, B. R., 162
107), 350(72, 107), 351(72, 107), 352, Debray, H., 234, 235(546), 236, 335(546)
353(145), 356(81), 360(145), 361, De Bruyne, C. K., 134(124), 135,
363(81, 108, 117), 364(107, 145), 140(124), 145, 160, 188(181, 323),
366(259), 367(110, 112), 369(110, 189(181,323), 195(182), 197(182),
112), 372,375(145, 146) 199(182,211),200(182, 211), 216,
Cragg, R. H., 70 273(124), 274(124),339(124,211),
Creger, W. P., 202,339(462) 340(124), 376
Cruickshank, C. N. D., 354 Dechary, J. M., 132
Crumpton, M. J., 193, 294, 325(445), De Fekete, M. A. R., 358
335(709) Degand, P., 270, 271(650), 272(650),
Cuatrecasas, P., 130, 219, 250(38) 337(650)
Cunningham, B. A., 151, 152, 153(325), De Gussem, R., 140, 188(181), 189(181)
154(263,268,269, 275), 155(263, Dekker, R. F. H., 366, 367(243),
267, 270, 271,289), 156(268, 275), 368(243), 369(243)
157(267,268), 158(267, 268), de la Chapelle, A., 292, 293(703)
160(289), 161, 162, 165(336), Delmotte, F., 134, 135(115),211,
203(263), 335(271) 212(483), 213(483),215, 216(115),
Cybulska, E. B., 174 217(115), 219(115,494,497),
Cyr, M. J. S., 168, 175(364), 184(364) 222(115,494), 223(494), 339(483),
Czech, M. P., 130, 150(39) 340(115)
Czonka, F. A., 155 de Maelenaere, H . J. H., 130, 150(41),
232
Demain, A. L., 104, 106, 108(50)
D
Den, H., 131
Dahlgren, K., 297 Denborough, M. A., 177
Dahlhoff, W. V., 36,39,40(17), 41(33a, De Neve, R., 191, 198(442b)
41,42), 43(39), 46(33a), 47(33a), Desai, N. N., 144, 195(213), 199(213),
48(33a), 52, 53(41), 58(17), 66(40), 200(213), 203(213), 204(213),
70(17,33a, 40,41,42), 71(17, 33a, 205(213), 309(213),336(213),
40,41,42), 74(33a), 75(33a), 76(33a), 337(213), 338(213),339(213)
77(42), 78(40,41) Desai, P. R., 131,227,278(674, 676),
Dahr, W., 259,303,304(761), 305(761), 279, 280(66,671, 674, 676), 281(66,
306(761), 338(617, 618) 671, 677), 282(679, 680), 284(674),
Dale, J., 33,35 303(678), 304(678),318(678),
Daniel, T. M., 179, 184(424) 338(674,676)
Daniels, P. J. L., 110, 111 Deuel, H., 64,348
Dankert, M., 355 de Waard, A., 131
Danon, D., 143, 258(202), 259(202), Dey, P. M., 344, 345(58), 350,351(139),
261(202), 335(202), 338(202) 362, 363(233),365(139, 223),
Daoud, K. M., 348 366(223,224), 367(139), 368(139,
Davey, M. W., 141, 168(192), 178(192), 140), 369(139, 140)
179(192,418) Diamond, M. A., 246
Davidson, E. A., 219 Di Ferrante, N., 168
Davies, C., 356(204),357, 366(204), Di Girolamo, A., 98
370(204), 371(204) Di Girolamo, M., 98
Davies, D. R., 358 Dillner, M.-L., 243
Dazzo, F., 147, 148 Dixon, M., 148
Dea, I. C. M., 343,345, 347(43), 349, Dixon, T., 270
392 AUTHOR INDEX, VOLUME 35

Dodd, R. Y.,315 Ekstedt, R. D., 176

Dolhun, J. J., 38,68(29) Elbein, A. D., 353
Dong, T., 366(259),367 El Khadem, H. S., 343,346(54)
Donnelly, A. J., 291 Ellestad, G. A., 89,91(21)
Donnelly, E. H., 173 Elting, K. A., 342
Dorset, M., 129,291 Ely, K. R., 151, 152(264)
Doty, D. M., 362,370(220) Emi, S., 366,367(251)
Douglas, S. D., 139, 309(162) Emoto, S., 74(111),77
Doyle, R. J., 156, 159, 160(297),161, Engvall, E., 178, 301(413),330(413)
162, 165, 168, 177(330),179(366) Ensgraher, A., 147, 148(228)
Drabik, J. S., 258, 259(614) Entlicher, G., 134(122),135, 137,
Drewes, L. R., 354 138(141, 142), 141, 155, 190(142),
Duke, J., 173, 179, 182(381),187(422) 191(142,442), 192(442),193(142),
Durand, R., l49,297(243e) 194(440,442), 196(141),197(141,
Durso, D. F., 347 440), 198(454,455), 199(122,440),
Dutcher, J. D., 81 201(122,446), 321(185),322(185),
Dweltz, N. E., 382 325(844),326, 335(141, 142,442),
Dyckes, D. F., 190, 191(441),192(441), 337(141,454,455), 388(142),
193,335(441) 339(122, 141,440)
Dymova, S. F., 38,42(26),44(26), 70(26), Eoff, W. E., 363,370(220)
72(26), 73(26),78(26) Eriksson, K. E., 366
Eriksson-Quensel, I.-B., 151, 152(257,
258), 335(258)
E Ermishkina, S. A., 38, 53(30), 54(30),
80(30)
Eagles, J., 48 Erskine, A. J., 346
Ehisu, S., 206, 208, 224 Ersson, B., 137, 306
Ebisu, T., 262, 266(621) Eshimoto, N., 108
Eckhardt, A. E., 266,295(629) Estola, E., 131
Edelev, M. G., 44,45(45), 70(45) Eto, E., 108
Edelman, G. M., 137, 138(140),139, 151, Etzler, M. E., 134, 137, 138(108, 146),
152, 153(325),154(263,268,269, 140(108),142(108), 144(108, 146),
275, 276), 155(263,267, 270,271, 149, 162,215, 223(335),227(108),
280,289), 158(267,),160(283,289), 228,229(520, 521), 231, 249,
161(280),162, 163(333),165(333, 270(146),271(146),272(146),
336), 202(140), 203(140,263), 273(146), 274(146), 335(108, 520,
205(140),224(156),296(156), 521), 336(146), 337(146, 335,495),
304(276),306, 307(769),335(267, 338(108), 339(146)
271, 769), 336(140), 337(140), Evans, P. J., 355
339(140) Evans, P. M., 165
Edmundson, A. B., 151, 152(264) Evans, V. J., 317
Edwards, J. O., 32,42(6), 49(6) Evdokimova, G. S., 54, 55(77),80(77)
Egami, F., 372 Everhart, D. L., 147,288
Eginitis-Rigas, C., 293, 335(706) Eylar, E. H., 355
Egorin, M. J., 296
Eguchi, Y., 86, 87(16),88(16)
Ehrlich, P., 128, 129
Eichmann, K., 314, 315(808)
F
Einstein, J. R., 255,256, 257(610), Fahn,A., 356
334(610),335(610) Fairbanks, G., 323
Eisenherg, F., 65 Fanger, H., 309
 AUTHOR INDEX, VOLUME 35 393

Farkas, E. H., 341 Frerman, F. E., 354,355

FarkaS, J., 75(112, 113),76(114), 77 Fridman, C., 232, 235(539)
Farmer, P. B., 123, 124(86) Fried, W., 104
Farnes, P., 309, 311 Friedman, B. A., 263
Fath, J., 342 Frost, R. G., 290, 336(691), 337(691),
Feeney, R. F., 268(637),269, 339(637) 339(691)
Feizi, T., 253, 285, 288(686) Fmmin, A. M., 202, 203, 205(468,469),
Feldman, J. D., 267,268(633) 339(468,469)
Felsted, R. L., 138, 296(154), 301 Fuhr, B. J., 156, 158, 159(311)
Fennessey, P., 355 Fujinaga, D. M., 258,259(614)
Feniindez-Moran, H., 306 Fujita, T., 189
Ferrier, R. J., 32, 37,38(3), 46(22), 47(3), Fujita, Y.,131,307(52),308(52, 778)
48(3), 50(55), 51, 52(3), 53(3, 22, 51, Fukuda, M., 178, 187,297(417),
55, 69), 54(22, 51, 69), 55(22, 52), 300(417), 301(417),304(417), 319,
57(3,22, 55), 58(69), 62, 73(69, 70), 323,324,374
74(3, 51,55), 75(3, 51, 55), 76(22, 52, Fukumoto, J., 366,367(251), 372(298),
69, 115, 116), 77, 79(3, 55) 373,374(298)
Ferris, B., 325 Funabashi, M., 85, 95(14)
Ferris, C., 178, 179(416), 186(416), 187, Funatsu, G., 270, 271(642), 272(642),
312(416) 339(641, 642)
FialovB, D., 193, 194(447) Funatsu, M., 270,271(642), 272(642),
Filippova, T. M., 44,45(45), 70(45) 339(641, 642)
Finch, A., 34, 35(11) Furihata, K., 124
Findlay, J. B. C., 323,324 Furthmayr, H., 323,324(845)
Finstad, C. L., 306,307(771), 309(771),
335(771)
Fishel, C. W., 168, 179(366)
G
Fliegerovi, O., 192,335(444)
Foglietti, M. J., 360, 362, 375 Gabriel, O., 89, 99(17), lOO(17)
Folling, I., 276 Gachelin, G., 156
Font, J., 139, 227, 228(518), 311(160), Gaillard, B. D. E., 345
335(519) Galbraith, W., 138, 142, 144(199),
Fontand, J., 227 164(199), 235(151),246(151, 199),
Ford, W. W., 131 247(151, 199), 248(151, 199),
Fordom, M. T., 366 249(199), 295(151),335(151, 199),
Foriers, A., 191, 198(442a,442b, 442c), 339(151, 199)
235(442c) Galun, E., 147
Fornstedt, N., 311 Gander, J. E., 354
Foster, A. B., 33,45, 50, 51(67), 53(67), Gantt, R. R., 317
54(67), 56(67), 70, 72(49), 77(67), Gardell, S., 292
78(67) Gardner, D. A., 264
Foster, D. M., 132 Gardner, K. H., 381
Fountain, D. W., 239 Gardner, P. J., 34,35(11)
Franke, W. W., 380 Garegg, P. J., 62, 71
Franks, D., 218, 339(501) Garrido, J., 317
Fraser, A. R., 162, 163(333), 165(333) Gatt, S., 363
Fraser-Reid, B., 96 Gaugler, R. W., 89, 99(17), 100(17)
Frkchet, J. M. J., 53 Gebb, C., 101
Freedman, S. O., 330 Gellhorn, E., 139
French, A. D., 378 Genaud, L., 179, 199(428),339(428)
French, D., 363 Genin, A., 370
394 AUTHOR INDEX, VOLUME 35

Gepner, I. A., 165 365),182(169,209, 240, 365,381),

Gerrard, W., 32 183(197,240,365), 184(120,168,
Geserick, G., 131 169,197,209,215,240,423,430),
Gesner, B., 297 185(168,197,204,240,365,430),
Giaja, J., 366 186(209,240,431), 187(197,209,
Gibson, W. A., 317 422,430),188(180,365),189(113,
Gielen, W., 218,239 180,436),190(204),200(169,197,
Gifford, H., 202,339(462) 240, 365),206(125),207(125),208,
Gilboa-Garber, N., 131 212(204),215(131),216,217(498),
Gilham, P. T., 63 218(498),219(498),222(498),224,
Gilliam, E. B., 327,328(862,863) 228(410),235(151),246(151,199,
Gillies, D. G., 78 410),247(151,199,240,410),
Ginsburg, V., 100 248(151,199,342),249(199,587,
Glaser, C., 139 591),262(125),263(132,622),
Glaser, L., 89,99(18),100(18) 264(126,131,625), 265(125),
Glaudemans, C. P. J., 317 266(191,621),295(157,628,629),
Glew, R. H., 156,160(297),163 30l(413),322(209),330(413),
Glicksman, M., 341(10),342 335(125,131,151,199,261,262),
Goddard, V. R., 292 338(125,131,168,169,204,215,
Goffart, P. R., 342 365,470,622,625,626), 339(151,
Gold, E. R., 132,250,252, 253,254(599), 199,498), 340(589,626)
315,316,339(599) Goldstein, L., 156
Gold, P., 330 Goldwasser, S.M., 132,313(85)
Goldberg, A. R., 130,214,218,219, Gombos, G., 326,327(857)
317(30),339(30) Gonatas, N. K., 317
Goldberg, M. L., 296 Gonzales, J. I., 342
Goldman, D. S.,355 GonzPlez, D . I., 178
Goldstein, A. W., 118,119(77) Gonzalez-Porque, P., 95
Goldstein, I. J., 129,134(121),135(112, Good, R. A., 306,307(771),309(771),
113,120),136,137(128),138(125, 335(771)
134,155),140(113,120,121,135), Gordon, J. A., 133,140,233,234,238(99)
141,142(125,131,168,197), Gorin, P. A. J., 71,143,168(209),
143(131,168,169),144(131,168, 179(209),182(209),184(209),
169,199),146(131),150(121,168), 186(209),187(209),322(209),343,
151(120,121,134,135,168), 354(41)
152(261,262),153(172),154, Goring, H., 360
155(173,262), 156, 157(180), Gottlieb, D., 104,107(54),108(54)
159(173,174),160(173,174),161, Gould, N.R., 246,247(586),335(586,
162(326),163(329),164(173,199, 588)
339),165(173,204,290),166(121, Goussault, Y., 311
320),167(120,121,168,197,204, Graham, V. A., 151
359),168(120,204,209, 215,240, GralBn, N., 151,152(257,258),335(258)
320,359),169(320),170(359), Grasbeck, R., 292,293(703,704),
171(120,121,134,135,359,369), 294(716),310(703),335(716,718)
172(320,362),173(121,320,362, Gray, D., 163
369),174(120,204,365), 175(204, Gray, G. R., 137,262(149b)
215,320),178(240),179(112,113, Gray, R. D., 156,160(297)
120,121,126,135,168,169,173, Green, D. M., 258,259(614)
179,197,204,209, 215,240), Green, J. W., 33
180(168,169,197,204,215,365), Greenaway, P. J., 219
181(118,169,197,204,215,240, Greer, J., 152,154(274a)
 AUTHOR INDEX, VOLUME 35 395

Gregory, W. T., 187, 298, 301(733), Halmer, P., 366, 370(254), 371(254)
320(434),331(434) Halpern, B., 307
Griffiths, D., 344, 345(58) Halprin, K. M., 372(296), 373
Grimaldi, J. J., 163, 164 Hamblin, J., 147
Grimmett, M. R., 347 Hammarstrom, S., 131, 134(63), 138(63),
Grisebach, H., 82,83(4, 5, 6), 84,89, 140, 141, 144(63),154, 165(290), 178,
90(24), 91(24), 92, 95(28), 98(4, 6), 216, 217(498),218(498), 219(498),
99, 100(5), 101, 102, 109(40), 122 222(498), 239(63), 240(63, 561,562),
Grollman, A. P., 155, 158, 161(307,308), 241(63, 561, 562, 564), 242(63, 178,
163, 179(308),201(308) 562, 563), 243(63, 189, 569), 249,
Gross, R., 299,307(735) 282, 291(682),301(413), 330(1),
Grubb, R., 131,277(65), 338(65) 335(562, 569, 570), 338(63, 189, 562,
Grundbacher, F. J., 208,288 563), 339(498),340(570)
Griiss, J., 370 Hannaford, A. J., 47, 50, 53(55, 68),
Gubanski, M., 376 57(53), 74(53), 75(53), 79(53)
Giirtler, L. G., 270,271(648), 272, Hannig, K., 297
336(648), 337(648) Hansch, C., 189
Guilbert, B., 139 Hara, K., 271
Guillot, J,, 131, 197 Hara, M., 366
Gul, B., 254 Haratz, A., 312
Gumpf, D., 314 Harboe, M., 276
Gunja Smith, Z. H., 166, 167(359), Hardman, K. D., 151, 152(262a),
168(359), 170(359), 171(359) 154(262a, 274), 156, 159, 160,
Gunther, G. R., 152, 154(268), 156(268), 161(274),335(262a, 324)
157(268), 158(268),162, 165(336) Harper, A. A., 271,272(658)
Gupta, D. S., 347 Harris, H., 134, 177(106)
Gupta, P. C., 343, 344,345(45) Hartigay, B., 35
Curd, J. W., 326 Hartman, F. C., 255, 334(607), 335(607)
Guseva, A. S., 38, 39,42(26), 44(26), Hanvood, S. E., 77(119, 120), 78
45(45), 51(36), 53(36), 54(46), 56(46), Hashem, N., 202
70(26,45), 72(26,46), 73(26), 80(36) Hashimoto, Y.,366, 367,372(298), 373,
Guss, J. M., 384 374(298)
Gussin, A. E. S., 214, 215(491) Haikovec, C., 325,326(851)
Guthrie, R. D., 67, 110 Hassid, W. Z., 353,355
Gutz, C. G., 156 Hassing, G. S., 140, 155(173), 159(173),
Gyaw, M. O., 353 160(173), 161, 162(326), 163(329),
164(173), 165(173),179(173)
Hatano, M., 366
Hatt, B. W., 49,63(64), 65(64)
H Hatton, L. R., 51, 73(70), 76(115, 116), 77
Haak, W. J., 104, 107(53), 115(53), Hawkins, D. C., 315
116(53), 117(53), 119 Hay, A. J., 372(297),373
Haddock, J. W., 354 Hayashi, J., 380
Haglid, K. G., 326 Hayashi, K., 270, 271(642), 272(642),
Hague, D. R., 157 339(642)
Haines, A. H., 50, 51(67). 53(67), 54(67), Hayes, C. E., 136, 141, 142(131),
56(67), 77(50), 78(50) 143(131), 144(131), 146(131),
Hakomori, S., 216, 217(499), 315 148(131),215(131), 263(131),
Hall, J. L., 313 264(131,625), 266(191), 295(629),
Hall, L. D., 70 335(131), 338(131, 625, 626),
Hall, S . , 347 340(626)
396 AUTHOR INDEX, VOLUME 35

Hayman, M. J., 193, 325(445) Holland, P., 300

Head, C., 293,335(705, 706) Hollerman, C. E., 134(121),135,
Healey, P. L., 356 140(121), 142(168),143(168),
Heath, E. C., 354,355 150(121, 168),151(121, 168),
Heath, M. F., 211 166(121), 167(121, 168), 168(121),
Heding, H., 108 169(121), 171(121),173(121),
Hedrick, J. L., 97 179(121, 168), 180(168),181(168),
Heggen, M., 267,268(632), 269(632), 184(168), 185(168),338(168)
297(632), 300(632),333(632), Holmen, H., 137, 138(137), 304(137)
339(632) Holt, P. D., 131
Hehre, E. J., 150, 151(253), 168(253), Homer, B. R., 162
171(253),173(371), 186(253,371, Horecker, B. L., 355
372) HoiejHi, V., 135, 136, 137, 289, 290,
Heidelberger, M., 270, 271(644),376 291(690), 333, 336(690), 337(690),
Heinrikson, R., 267,294(630), 295(630, 339(690)
712), 335(630,712) Horii, S., 120, 122(81)
Hellin, H., 254, 259(604) Horisberger, M., 138, 174, 175(388),
Hellstrom, U., 243 262(150c), 264
Helman, J. R., 177 Homer, V. V., 342
Hemperly, J. J., 162, 163(333), 165(333) Hornick, C. L., 141
Hems, R., 45, 70, 72(49) Horowitz, P., 163
Henderson, M. E., 352 Horstmann, H. J.,270,271(648), 272,
Henley, R. R., 129,291 336(648), 337(648)
HBrissey, H., 370 Horton, C. B., 148, 149(243)
Hermetet, J. C., 326 Horvei, K. F., 345,348(86)
Herth, W., 380 Hossaini, A. A., 132
Heyne, E., 348 Hotta, K., 302,339(757)
Hicklin, B. L., 206, 338(470) Hough, L., 346,348(92, 93), 349, 352(92)
Hickman, S., 131 Houston, C. W., 372(302),373
Higashi, Y.,355 Howard, I. K., 134(123),135, 137,
Higuchi, T., 355 138(138), 141, 142(123), 144(123),
Hildesheim, J., 283,284(684), 335(684) 148, 149(243),190(138), 191(441),
Hill, R. L., 136, 337(130) 192(441), 193(138,441), 194(123,
Hirschhom, R., 297 138, 186), 195(125),202(138),
Hirschmann, W. D., 299,301(735), 335(138,441),338(125, 138)
302(736) Howe, M. L., 131
Hirst, E. L., 343, 347, 348(100), 351(100), Howell, S. F., 134, 136(102), 140(102),
352(100) 150, 151, 155, 166(102), 168(102),
Hitchcock, C., 347 169(102), 173(102), 177(102),
Hitz, W. D., 58 179(102), 338(102,246, 247)
Hiyama, K., 367 Howes, F. N., 356
Hoglund, S., 297 Hiebabeckf, H., 75(112),77
Hoeksema, H., 119(80), 120 Hrgovcic, R., 168
Hof, H. I., 326 Hsu, R., 267, 294(630),295(630),
Hoffman, J., 349 335(630)
Hoffman, M. K., 58 Huang, A. H. C., 149
Hoffmann-Ostenhof, O., 104 Huang, J. W., 141, 168(192), 178(192),
Hofheinz, W., 84, 122 179(192)
Hofnung, D., 156 Hubbard, A., 325
Hogenkamp, H. P. C., 123, 124(86) Hubbell, D., 147, 148
Holland, N. H., 300 Hubbell, W., 318
 AUTHOR INDEX, VOLUME 35 397

Huber, G., 348 Iwasa, J., 189

Hubert, A. J., 35 Iwen, M. H., 347
Hudgin, R. L., 131 Iyer, R. N., 134, 138(112,113),140(113),
Hudson, C. S., 372(299),373 179(112,113),181,184(430),
Huet, C., 165,317 185(430),189(113)
Huet, M.,152,165
Hui, P. A., 347,348,350(109,120),363
Hukins, D. W. L., 377,384
Hunedy, F., 48,52(58),54(58),57(58), 1
74(58,59),75(58) Jack, A., 156,161(296)
Hungerford, D.A., 291 Jackson, J. J., 355
Hunt, R. C., 329 Jackson, R. L., 319,320,321, 324(843)
Huprikar, S. V., 145,196,197(448), Jacoby, M., 272
303(219),339(448) Jaffe, M., 166
Husain, A., 48,52(58),54(58),57(58), Jaff6, W. G., 130,132(44),150,159,166,
74(58,59),75(58) 178,297
Hutson, D. H., 172 Jamieson, G. A., 300
Hybl, A., 378 Janata, J., 150
Hylin, J. W., 361,370(219),
374 Jansons, V. K., 328,329(865,866)
Janzen, D. H., 149
I Jaumatte, J., 226
Javaid, J. I., 326
Ichihara, N., 108 Jayne-Williams, D.J., 130,341(9),342
Ichiki, N.,137,138(144),254(144), Jeanes, A., 354
270(144),271(144) Jeanloz, R. W., 143,186(210),187(210),
Ikuma, H., 371 195(210),196(210),275,303,
Il’ina, E. F., 68 304(762),322(210),330(762),
Imahori, K., 131,307(52),308(52) 331(762)
Inamine, E., 104, lOS(50) Jenkins, J., 313
Inbar, M.,130,150(31),179(250),288, Jentoft, N. H., 354
317(31) Jermyn, M. A., 134,135(116),139(116)
Inch, T. D., 50,51(67),53(67),54(67), Jindal, V. K., 343
77(67),78(67) Jirgensons, B., 215,223(495a),230,
Innawi, S., 366 231(523),272,281,282(680),
Inuzuka, T., 366 31l(523)
Irimura, T., 140,238,251,252(183),254, Johansson, B. G., 178,301(413),331(413)
258,259(616),261(616),273(183), Johnson, E. A., 292,293(697, 698),
275,297,300(183),304(183),305, 294(697,698),335(697,698,707)
318(183),331(183),335(616), Johnson, E. A. Z., 138,203,205,
339(183) 337(467a)
Isaacs, R., 330 Johnson, L.N., 221
Ischiguro, S., 271 Jones, B . M.,165
Ishiguro, M., 270,271(642),272(642), Jones, D. A., 147
339(642) Jones, D. B., 155
Ishii, S.-I., 154 Jones, J. K. N., 179,346,347,348(92,93,
Ishiyama, I.,239,240(559, 560),241(560, loo),351(100),352(92)
567),242(567),243(559,567) Jones, J. M., 148,267,268(633),
Isono, K., 126 269(635),339(631,635)
Ivanova, E. A., 39,49,51(36),53(36),63, Jones, Q., 344,348(61)
64,80(36) Jones, R. D., 348,350(121)
Iwabuchi, M., 154 Jones, R. L., 371
398 AUTHOR INDEX, VOLUME 35

Jones, R. T., 292, 293(698),294(698), Kargin, V. A., 55

335(698) Karl, W., 89, 90(24), 91(24), 92
Jonsson, B., 131,277(64),338(64) Karlsson, B., 326
Joseph, J. P., 58 Karlstrom, B., 156, 193(301)
Jourdain, G. W., 224 Karush, F., 141
Judd, W. J., 206,263,338(470) Katar, M., 312
Juergens, W. G., 176 Katchalski, E., 138,232,233(538),
Jung, P., 355 234(538), 235(537, 538, 539), 238,
Jurgelsky, W., 244 335(538)
Juster, H. B., 149 Kates, M., 363
Katsuno, A., 177
Katzen, H. M., 130, 150(40)
K Kaufman, H. W., l52,154(274a)
Kabarity, A,, 202 Kaufman, S. J., 257, 335(612)
Kabat, E. A,, 131, 133, 134(63), 138(63, Kaul, R. K., 344
108), 140(108),142(108), 144(63, 108, Kauss, H., 139,355
198, 200), 154, 156(170), 176, 178, Kawaguchi, H., 102
179(170),180(170), 181(170), Kawaguchi, T., 140, 238(183), 251(183),
184(170),208,227( 108),228,229,231, 252(183), 254(183), 261(183),
236(552), 237, 238(552),239(63), 273(183),275(183),297(183),
240(63), 241(63), 242(63), 243(63), 300(183), 304(183), 313,331(183),
253,259(203), 260(198), 261(203), 339(183)
270, 271(644),284, 285, 286, 287, Kawamura, T., 108
288(200, 526), 313, 314(803),316, Kawasaki, T., 131
335(108,200,570), 338(63, 108, 180, Kawauchi, H., 315
552), 340(570) Kay, C. M., 153, 154, 160(285), 163(285a)
Kafka, J. A., 308,309(779) Keen, J. L., 342
Kahane, I., 319,320(843), 321(843), 323, Kehoe, J. M., 306
324(843) Keilich, G., 366
Kahlem, G., l49,297(243e) Keith, C., 215
Kahn, A. H., 254 Kelkar, P. S., 347
Kaifu, R., 143, 186(210),187(210), Keller, J., 186,298(432),300(432)
195(210),196(210),302,322(210) Keller, L., 270,271(646, 647)
Kaiser, R., 361 Keller-Schierlein, W., 81, 122(2)
Kalb, A. J., 138, 152, 153, 154(274a), Kent, S. P., 147
155(278),156, 161(296),179(278), Keough, A. H., 32, 38(8), 50(8), 51(8),
283(153),285(683), 335(265, 683), 77(8), 78(8)
340(683) Khalap, S., 315, 316
Kalvoda, L., 75(114), 77 Khanna, S. N., 343,345(45)
Kamata, T., 346 Kieda, C., 211, 212(483),213(483),
Kameda, Y., 120, 122(81) 339(483)
Kan, T.-J., 168, 177 Kiehs, K., 189
Kanarek, L., 191, 198(442a,442b) Kiernan, J. A., 317
Kandler, O., 352 Kikuchi, T., 173
Kanellakes, T. M., 264 Kim, Y. C., 244,245(575)
Kanetsuna, F., 376 Kim, Y. S., 330
Kaplan, M. J., 214, 215(488),216(488) Kim, Z., 239,242(556)
Kaplan, N. O., 294(717),295 Kimata, K., 86,87(16), 88(16)
Kaplan, R., 306 Kingdon, H . S., 294, 295(712), 335(712)
Kapoor, V. P., 344, 346(66), 347,348, Kiyohara, H., 109
351(124) Klages, F., 343
 AUTHOR INDEX, VOLUME 35 399

Kniep, B., 109 276(666), 297(216),298(216, 432),

Knight, J. C., 119(80), 120 299(216, 730, 732), 300(432, 732),
Knoboch, I., 239, 240(563) 301(216,733),304(216),308,311(731),
Knoboch, W., 239,240(563) 318(216,435, 747), 319(216,435,
Knop, F. B., 342 730, 747), 320(216,434, 731),
Knox, R. B., 134, 135(116), 139(116) 321(216,435, 730, 747), 322(435,
Kobayashi, M., 160, 173 666, 730, 731, 747), 323, 324(216,
Koch, G. L. E., 372 730, 747), 331(216, 434, 435, 666,
Kocourek, J., 134(122), 135, 136, 137, 730, 731, 747), 332
138(141), 141, 145, 155, 190, Koshiyama, H., 102
191(442), 192(442), 193, 194(440, KoStif, J. V., 134(122), 135, 137, 138(141,
442,447), 196(141),197(141,440), 142), 155, 190(142), 191(142),
198(454,455), 199(122), 201(122), 193(142), 196(141), 197(141),
289, 290, 291(690),313, 314, 199(122),201(122), 335(141, 142),
321(185), 322(185), 325(844), 337(141), 338(142), 339(122, 141)
326(851),333,335(141,142,442,444), Koulumies, R., 208,338(472,473)
336(690),337(142,454,455, 690), Koura, A., 379
338(142), 339(122, 141,440, 690) Kovacs, P., 341
Kohler, W., 133, 145, 227, 239(loo), Kozak, L. P., 354
240(558), 241(100), 242(100), KritkL, Z., 178
243(100, 558), 338(100) Kreger, D. R., 343, 347(46)
Koenig, S. H., 156, 157, 179, 184(429) Krishnaswami, N. R., 345
Kossel, H., 65 Kristenko, L. V., 52, 54(71), 56(71),
Koster, R., 36, 39,40,40(17), 41(33a, 41, 80(71)
42), 43(39), 46(33a), 47(33a), 48(33a), Kristiansen, T., 129,304
52, 53(41), 58(17), 66(40), 70(17,33a, Kriipe, M., 132, 133(77), 139(77), 140(77),
39, 40,41,42), 71(17, 33a, 40,41,42) 141, 142, 146, 147,208,210,212(77,
74(33a), 75(33a), 76(33a), 78(40,41, 480), 224,226,244, 250, 252(195),
42) 253(593), 263(195), 264(195),
Kottgen, E., 132 265(195), 282(77), 284(77), 289(77),
Koh, C., 256, 257(610), 334(610), 290, 302, 303(754),304(77, 755),
335(610) 305(77),313(77),338(77,195), 339(77,
Kolecki, B. J., 278(674, 676), 279, 480, 593,754,755, 756)
280(674,676), 338(674, 676) Kubknek, J., 321,325
Kolodkina, I. I., 37, 38, 39,49, 51(27,28, Kiihnemund, O., 227, 239, 240(588),
36), 53(27,36),54(27, 28, 35), 55(77), 243(558)
63, 64,80(27, 28,35, 36, 75, 76, Kuhn, R., 277, 283, 284(672), 338(672)
77) Kuhns, W. J., 268(637), 269, 291,
Kolpak, F. J.,379,382(17) 339(637)
Kooiman, P. J., 343, 347(46), 366(257), Kuivila, H. G., 32,38(8), 50(8), 51(8),
367 7 7 W 78(8)
Kornfeld, R., 145, 178, 179(416), Kulshreshtha, A. K.,380
186(416), 197, 297(216), 298(216, Kunitz, M., 270,271(645)
432), 299(216, 730, 732), 300(432, Kunstmann, M. P., 89, 91(21)
732), 301(216, 733), 304(216), Kuppel, A., 380
311(416),318(216, 730, 747), Kurokawa, T., 137, 138(144),202,
319(216,730, 747), 320(216), 254(144), 270(144), 271(144), 312,
321(216, 730, 747), 322(730, 747), 340(796)
324(216, 730, 747), 331(216, 730, Kustanovich, I. M., 52, 54(71), 56(71),
747) 80(71)
Kornfeld, S., 131, 145, 186, 187, 275, Kuzuhara, H., 74(111), 77)
400 AUTHOR INDEX, VOLUME 35

Letsinger, R. L., 35
L
Leunberger, R., 348
Lackovic, V., 376 Le Vine, D., 214,215(488), 216,219
Lal, B. M., 356, 361(200) Levine, P., 302,339(750)
Lal, G., 378 Levine, V., 372(296),373
Lalaurie, M., 132,224,284(81),289, Levitzki, A., 153, 155(278), 179(278)
339(81) Levvy, G. A., 372(297),373
Lamblin, G., 270,271(650), 272(650), Levy, A., 150, 166, 178
337(650) Lewis, B. A., 151, 168, 171(260),
Lampen, J. O., 174 175(364), 184(364)
Lamport, D. T. A,, 211 Lewis, D., 77(117, 118, 119, 120), 78
Lancaster, J. E., 89, 91(21) Lewis, S. D., 163, 164(339)
Landsteiner, K., 129, 139, 190, 291 Lhermitte, M., 270,271(650), 272(650),
Lang, A,, 371 337(650)
Langridge, R., 215 Li, J. G., 129, 292
Lankester, A., 130 Li, S. C., 363
Lantz, R. S., 327 Li, S. S.-L., 271, 306
Lappert, M. F., 32 Li, Y.-T., 363,372(300), 373,374(300)
L a m , O., 349 Liao, J., 316
Larson, E. B., 345, 346(84) Liener, I. E., 130, 132(45),134, 137,
Latimer, S. L., 372(302),373 138(136), 149, 151, 152(136, 264),
Lau, P. Y., 96 154, 171(136),231,232(529), 233,
Leavitt, R. D., 138,296(154),301 234(534), 235(532),296, 301,
Lechevallier, D., 365 335(136)
Le Dizet, P., 344, 347,348(72, 107), Liese, W., 366
350(72, 107), 351(72, 107), 354, Lightbody, J. J., 224
361(112),363,364(107), 367(110, Lin, C. M., 110
112), 369(110, 112), 372 Lin, H., 130, 150(33)
Lee, B. K., 110, 111 Lin, J.-Y., 130, 254(32),255, 271(32)
Lee, J. K. N., 203, 205(469),339(469) Lin, K., 271
Lee, J. W., 372(301),373 Lin, L.-T., 130,254(32), 271(32)
Lee, S. R., 364, 367,370(236), 372(236), Lindahl-Kiessling, K., 297
374(236) Lindberg, A. A., 140, 241(178), 242(178)
Lee, W., 376 Lindberg, B., 174,349
Lee, Y. C., 138,258(150d), 333, 372(300), Lindberg, Bengt, 46, 53(53), 57(53), 62
373, 374(300) Lindberg, Boje, 62
Lees, E. M., 38,42(23), 50(23), 51(23), Linden, J. C., 355
53(23), 54(23), 55(23), 57(23), 58, Lindstrom, K., 71
60(86), 72(23), 77(23), 78(23) Ling, A. R., 370
Lemanski, T., 253,339(601) Ling, N. R., 130,291(36)
Lemonnier, M., 311 Linsley, K. B., 275
Lennarz, W. J., 355 Lipscomb, W. N., 152
Leo, A. J., 341 Lis, H., 129, 130(18,26), 132(18, 26,
Leon, M. A., 134(123),135, 138, 141, 37b), 133, 134(97),136(18), 137(18,
142(123), 144(123), 150, 178(109), 132), 144, 150(72),164, 208(18),
179(249),190, l91(441), 192(441), 209(18), 214,215(491), 232(97),
193(441),194(109, 123), 195(123), 233(538), 234(538,541), 235(537,
224,338(109, 123) 538, 539), 236(212),238(99,212),
Leschziner, C., 344, 346(68), 348(68) 239(343,343a), 296, 317(354),
Leseney, A. M., 227, 228(518), 298, 318(18), 335(538),338(212)
311(73l),320(731),322(73l), Liske, R., 218,339(501)
33l(731) Lisowska, E.,302,303(759), 339(753)
 AUTHOR INDEX, VOLUME 35 401

Litman, G. W., 306,307(771), 309(771), McElhinney, R. S., 58

335(771) McGinnis, G. D., 48, 54(61), 62(61),
Little, L. L., 341(8), 342 79(61)
Livingston, D. C., 317 McInnes, A. G., 382
Lloyd, K. O., 140, 166(170), 172, 175, McKenzie, G. H., 152, 153, 154(279),
176, 179(170), 180(170), 181(170), 160(266), 164(279),165
184(170),338(170),354,376 Mackie, W., 377
Lockhart, J. C., 34,70 McKinley, I. R., 35,42,43, 45, 53(43,
Lohmar, R. L., 344, 348(61) 48), 55(43,48),66, 67(lo), 72(48),
Lonchampt, M., 165 77(43),78(43)
Longcor, F., 39 MacLennan, A. P., 315
Lonngren, J., 135, 138, 179(126), 224, McMaster, M., 134, 147, 244(103),
262, 263(622), 264(126, 622), 245(103), 248(103)
338(622) McNamara, P. M., 34, 35(11),
Loontiens, F. G., 134(124), 135, 140(124), McPherson, A., 257, 335(612)
144, 160, 188(181,323), 189(181, Makela, O., 129, 132, 133(78, 79),
323), 195(182),197(182), 199(182, 134(78), 140(78),141, 142(20),
211), 200(182,211), 216, 273(124), 144(20), 146, 147, 148(20), 199(20),
274(124),339(124,211),340(124),376 202, 208, 209,(20, 78), 210, 226, 244,
Lorand, J. P., 32,42(6), 49(6) 250(78), 252(195), 253(78), 262,
Lotan, R., 136, 137(132), 143, 147, 164, 263(195, 620), 264(195), 265(195),
214, 215(491),218, 219(504), 223, 282(78), 284(78), 290, 304(78),
234, 235(544),236,239(343, 343a), 305(78), 306(78), 313(78),338(20,78,
258, 259(202,203), 261(202,203), . 95), 339(78)
262(619), 318,335(202, 544, 546), Makela, P., 142,252(195),262, 263(195,
338(202, 619), 340(544) 620), 264(195),265(195), 338(195)
Lucas, J. J., 355 Magnuson, J. A., 156, 158, 161(310)
Luecker, P., 341(17), 342 Mahler, H. R., 326
Luisada, A., 202 Mahmood, S., 85(13), 95(13)
Lundeen, D. E., 164,248(342) Maier, S., 102, 109
Lundstrplm, H., 349 Mair, G. A., 221
Lustig, A., 152,335(265) Maisonrouge-McAuliffe,F., 178
Lynn, W. S., 130, 150(39) Maiti, B. C., 56
Lyr, H., 366 Majems, P. W., 141
Majumdar, M. K., 117, 118
Majumdar, S. K., 117, 118
M Malcolm, E. W., 33
McAlpine, T. S., 84 Maley, F., 372(292),373
McCleary, B. V., 347,356,(205), 357, Malik, J. M., 104, 107(53), 115(53),
364(205), 365,366, 369,370(205), 116(53), 117(53)
372(252), 373(252), 374 Malinzak, D. A,, 131
McClendon, J. H., 32,49(5) Mann, T., 372(293),373
McCormick, J. E., 58 Manners, D. J., 166, 168(357, 358), 169
McCredie, R. J., 348 Mannschreck, A,, 96
McCubbin, W. D., 153, 154, 160(285), Marchalonis, J. J.,306, 307(769),
163(285a) 335(769)
McDaniel, L., 110 Marchesi, V. T., 214, 317(493), 319,
McDannel, M. L., 177 320(843), 321(843), 323, 324(843,
McDermed, J. D., 292, 293(698), 845)
294(698),335(698) Marchessault, R. H., 377, 378(l),379
McDonald, M. R., 270, Marcus, D. M., 155, 158, 161(307,308),
271(645) 163, 179(308),201(308)
402 AUTHOR INDEX, VOLUME 35

Marcusson-Bequn, H., 210 167(121), 168(121), 169(121),

Maiik, T., 197, 198(454),337(454) 171(121),173(121),179(121)
Marini, M., 161, 162(326) Meyer, E. W., 233
Marinkovich, V. A., 210,211(482), Meyer, H . W., 317
339(482) Meyers, F. L., 366(258),367
Markowitz, H., 140, 177(177) Mialonier, G., 149, 215, 219(494),
Marquardt, M. D., 140 222(494),223(494),297(243e)
Marsh, C. A., 372 Michelson, A. M., 94
Marsh, W. L., 253 Miescher, P., 297
Marshall, R. D., 211 Miki, T., 248
Marshall, W. H., 300 Miller, A. L., 290, 336(691),337(691),
Martin, C. R., 344, 348(62) 339(691)
Martin, J. R., 118, 119(77),317 Miller, F., 178, 186, 187(433)
Martin, L. L., 118, 119(77) Miller, J. B., 267,294(630),295(630,712),
Martin, T., 132 335(630, 712)
Masler, L., 376 Miller, J. T., 246, 253
Massaro, E. J., 307 Miller, S. E., 347
Matern, H., 83, 84(8),92, 93, 95(28) Minshall, J., 89
Matern, U., 99, 101, 102, 109(40) Mirelman, D., 147,218
Matheson, N. K., 347,356(205), 357, Misaki, A., 143, 148, 168(209,240), 172,
364(205),365,366, 369(116), 173, 174(240),178(240),179(209,
370(205), 372(252), 373(252),374 240), 181(240),182(209,240, 381),
Mathews, K. P., 264 183(240),184(209,240,423),
Matsubara, M., 366 185(240),186(209,240), 187(209),
Matsubara, S., 245, 246 200(240),247(240),322(209)
Matsuda, K., 173 Misawa, Y.,366
Matsumoto, I., 137, 138(149), 142, 143, Mitchell, E. D., 372(302),373
144(208),145(208),208(196,208), Mitscher, L. A., 89, 91(21), 118, 119(77)
209(149, 196), 210(196), 212, Mizoguchi, T., 109
225(149, 196,208, 226), 226, Moe, 0. E., 347
262(149),282(476), 289(149, 196, Moller, G., 130, 154, 165(290),291(37)
226), 290(196), 291(196), 296(149), Mogel, L. G., 45, 52, 54(47),56(72),
305(196),313,335( 149,476), 72(47)
336(226),337(196,208,226, 509), Moldow, C. F., 300
338(476),339(196, 509, 692) Monsigny, M., 134, 135(115),149, 211,
Matsumoto, T., 156, 193(300) 212(483),213(483),214,215(490b),
Matsuzaki, M., 102 216(115),217(115),219(115,494,
Mayer, M. M., 133 497), 222(115,494), 223(494),
Meier, H., 345,346,352, 353(143),355, 297(243e),306, 307(770),335(770),
356(204),357, 360,361(201), 339(483),340(115)
364(202), 365, 366(204),370(201), Montgomery, R., 134, 140(117),151,
371(201, 202,204), 372(202),374 166(117),168(354,356), 169(117,
Meir, H., 344, 345(70),348(70) 356), 171(260),179(117),343
Mekinnon, A. A., 349 Moore, L., 189
Melo, A., 89,99(18), lOO(18) Moorhouse, R., 384
Meloche, H. P., 123 Mordman, C. T., 292,293(703), 310(703)
Melton, L. L., 342 Morel], A. G., 131
Merdel, L. B., 231,292 Moreno, C., 178
Merlis, N. M., 55 Morgan, I. G., 326, 327(857)
Merrick, J. M.,134(121), 135, 140(121), Morgan, W. T. J., 129, 131(21, 22),
150(121), 151(121),166(121), 140(21,22), 141, 142(22),144(22),
 AUTHOR INDEX, VOLUME 35 403

207,208(22,471),209(471,475), Nachman, R. L., 325

210(471), 226(476), 244, 248(22), 250, Nadelmann, H., 358
253(22), 277(21), 282(22), 283, Nagai, Y., 302, 303(758),304(758),
284(22), 305(471), 338(21, 22,471, 339(752, 758)
475,477), 339(22) Nagaoka, K., 106
Morgenstem, M., 177 Nagaoka, T., 372
Morgenthaler, W. W., 342 Nagata, Y.,136,212(129), 214(129),
Mori, Y., 157, l60,324(306b) 215(129,487), 216,337(129),
Morinioto, J. Y., 344 340(129)
Morita, K., l57,324(306b) Nainawatee, H. S., 356, 361(200)
Morley, R. G., 346 Nakamura, S., 134, 157, 177(107),
Moroux, Y., 214,215(490b) 300(107)
Morrison, A,, 343,345,347(43), 349, Nakano, K., 312
350(43), 356(43), 363(43), 364(43), Nanji, D. R., 370
369, 375(43) Nanno, S., 271
Morrison, J. D., 35 Napier, P. W., 288
Morse, J. H., 134, 177(105), 178(105), Napoli, C., 147, 148
300(105), 301 Naspitz, C. K., 130, 291(35)
Morse, S. I., 176 Natarajan, C., 343
Morton, G., 89, 91(21) Nathenson, S. G., 176
Moscona, A. A., 317 Neely, W. B., 172
Moskal, J. R., 264 Neimo, L., 378
Mountfield, B. A., 32 Neri, G., 272,327, 328(861, 862, 863)
Mueller, G. W., 292 Neter, E., 131
Mueller, H., 341(16, 17), 342 Neuberger, A., 136, 143(128), 144(128),
Mukaida, M., 293, 240(559, 560), 195(213), 199(213),200(213),
241(561), 243(560) 203(213), 204(213),205(213), 211,
Mukherjee, A. K., 343,346(52), 348(52) 212(207), 213,221, 233,309(213),
Mukhejee, S., 343,347, 348, 351(124) 336(213), 337(128,207, 213),
Munro, M. H. G., 104, 107(54), lOB(54) 338(213), 339(128,207, 213)
Murachi, T., 333 Neukom, H., 348,350(120)
Murakami, M., 54, 79(78) Neupert, G., 317
Murakawa, S., 134, 177(107), 300(107)’ Newman, A. D., 156
Muramatsu, T., 372(294), 373 Nichol, L. W., 152, 160(266), 165
Murase, K., 54,79(78) Nicholson, S. K., 160
Murawski, A., 111 Nicolson, G. L., 129, 130(25), 132(25),
Murphy, J. C., 155 133(25), 137, 138(146), 144(146), 142,
Murphy, L. A., 248,249(591), 262, 263, 150(25),270(146, 194), 271(146, 194),
264(625), 266(621), 295(628),338(625) 272(146), 273( 146), 274(146),
Murphy, V. G., 378 293(194), 317(194),318(25),
Musgrave, 0. C., 34 35(13), 42(13), 56, 336(146), 337(146, 194),339(146,
78(13) 194), 340(146)
Mustier, J., 131 Niederhuber, J. E., 135, 179(126),
Mustier, M., 197 264(126)
Muzurek, M., 71 Nieduszynski, I. A., 379,382,383(26),
384(28)
N Nigrelli, R. F., 268(637),269, 291,
339(637)
Nabors, M. W., 371 Nijenhuis, L. E., 302
Nachbar, M. S., 139,291, 306(155), Nikaido, H., 355
307(155), 329,330(868) Nikaido, K., 355
404 AUTHOR INDEX, VOLUME 35

Nimi, O., 109 150, 639, 649, 651), 272(639, 651,

Nisizawa, K., 367 658), 273,274( 147), 276,334(147,
Noguchi, H., 307 150), 335(147, 150, 651), 336(147,
Nomi, R., 109 150, 649, 651), 337(147, 150, 651),
Noonan, K. D., 132, 165,318 338(147, 649), 339(147), 340(147)
Nordbn, A., 292 Olson, M. 0. J., 137, 138(136), 151,
Nordgren, R.,342 152(136), 154, 171(136), 335(136)
Nordman, C. T., 292,293(704), 294, Onodera, K., 197, 199(452),200(452),
295(71l), 335(711) 339(452)
Norins, L. C., 300 Onozaki, K., 137, 138(144),202,254(144),
North, A. C. T., 221 270(144), 271(144)
Northcote, D. H., 211, 366(258), 367 Opie, J. W., 342
Northrup, R. L., 301 Oppenheim, J. D., 139, 291, 306(155),
Novak, E., 366 307(155), 329, 330(868)
Novogrodsky, A., 164, 238, 239(343, Orenstein, N. S., 100
343a), 261, 262(619),338(619) Ortmann, R.,99
Nowak, H., 341(17),342 Osawa, T., 129, 132(19), 137, 138(139,
Nowak, T. P., 139,306(159), 307(159), 143, 144, 149), 140(19), 142(143),
325(159) 143, 144(143,208), 162, 163(332),
Nowakova, N., 145 165(332), 178, 186(210), 187(210),
Nowell, P. C., 130, 291, 291(34) 190(143), 191(143),192(143),
Noyes, C., 294,295(712), 335(712) 194(143), 195(143,210), 196(210),
Nuernberg, E., 341(12, 13, 14, 16, 17), 202(139), 205(139),207(19), 208(195,
342 208), 209(149, 195),210(195), 212,
Nystrom, R. F., 104, 107(53), 115(53), 225(149, 195, 208,226), 226,
116(53), 117(53) 238(183), 250,251( 183, 594),
252(183), 253(594),254(144, 183),
258,259(616), 261(183, 616),
0 262(149), 270(144), 271(144),
275(183,289), 282(476),289(149,
Oblin, A,, 214 226, 290,291), 290(196), 291(196),
Obrenovitch, A., 214 296(149), 297(183,417), 300(183,
O’Brien, J. S., 290, 336(691),337(691), 417), 301(417), 302,304(183,417),
339(691) 305(19, 96),310,312, 313, 318(183),
Oerkermann, H., 299, 301(735), 302(736) 319, 321(210), 323,324(740, 741),
Oh, Y. H., 294,335(710), 338(710) 331(183), 335(143, 149,476, 616),
Ohashi, S., 102 336(226), 337(196,208, 226, 509),
Ohata, Y., 109 338(19, 143,476), 339(183, 196, 509,
Ohkawara, A., 372(296),373 692), 372
Ohkita, J., 380 Osborn, M. J., 355
Ohmari, T., 102 Osborne, T. B., 231
Oikawa, K., 153, 154, 160(285),163(285a) Osgood, E. E., 129,292, 339(696)
Oishi, K., 131,307(52),308(52, 778) Osman, H. G., 277,283,284(672),
Okada, M., 160 338(672)
O’Kane, D. J., 166, 168(352) Osuga, D. T., 268(637),269, 339(637)
Okanishi, M., 102 Ottensooser, F., 148,302,339(750)
Okazaki, R., 94 Overend, W. G., 47, SO(SS), 53(55),
Okazaki, T., 94 57(55),58,62,74(55), 75(55), 76(115),
Okuda, S., 85,86,87(16), 88(16) 77, 79(55)
Okumura, T., 372 Overton, J. D., 341(11), 342
Olsnes, S., 137, 138, 254(147, 150), 255, Oyen, R.,297
256, 257,270(147, 150), 271(147, Ozanne, B., 214,318(489)
 AUTHOR INDEX, VOLUME 35 405

Perrodon, Y., 307

P
Perricone, A. C., 342
Pacak, F., 145, 313, 314 Perry, A. L., 372, 373(284)
Pachtman, E. A., 202,205(469), 339(469) Perry, M. B., 179
Painter, T. J., 208, 209(475), 226(475), Petek, F., 348,350, 363(117), 366(259),
338(475, 476), 349, 352 367, 368(140, 253), 369(140, 253)
Paley, L. G., 360 Petek, P., 347, 363(108)
Pallansch, M. J., 231, 232(529), 233, Petryniak, J., 313, 314
234(534) Pettitt, D. J., 342
Palmer, K. J., 348 Pezzanite, J. O., 96
Palozzo, A., 159, 166, 297 Pfeiffer, V. F., 344, 348(62)
Panchenko, S. I., 64 Pflumm, M. N., 153,154, 155(289),
Pande, A., 378 160(283,289), 168(319)
Pandolfino, E. R., 156 Pfuderer, P., 255, 256, 257(610), 334(607,
Pape, H., 83,84(8) 610), 335(607, 610)
Pappenheimer, A. M., 271, 272(658) Phelps, C. F., 316
Pardoe, G . I., 131, 133, 143, 144(206), Philipp, B., 379
211,212(206), 213(206), 218, Phillips, D. C., 221
224(502), 258(201), 259(201), 267, Pichuzhkina, E. I., 64
268(632, 636), 269, 276(206), 299, Pihl, A., 137, 138, 254(137, 150),
300(632), 301(735), 307, 333(632), 255(137), 256(137), 257(137),
338(201), 339(206, 632, 636) 270(137, 155), 271(137, 155, 639,
Park, R. B., 363 649, 651), 272(137, 639, 651),
Park, W. M., 371 273(137), 274(137), 334(137, 150),
Partridge, J., 314 335(137, 150, 651), 336(137, 150,
Paton, F. J., 370 649, 651), 337(137, 150, 651),
Paulovi, M., 134(122), 135, 190, 338(137, 649), 339(137), 340(137)
191(442),192(442), 194(442), 197, Pittner, F., 104
199(122),201(122), 335(442), Pittz, E. P., 159
339(122) Plow, E. F., 178, 182(419)
Paulsen, H., 89,99 Plummer, T. H., Jr., 372(292), 373,
Pavlista, A. D., 371 374(305)
Pecht, I., 155 Podder, S. K., 141, 168(193), 177,
Pelly, R., 163 179(193), 274(193)
Pemberton, R. T., 239(566), 240 Poleflca, T. G., 219
Percheron, F., 352,353(145), 360(145), Pollitzer, W., 302, 339(750)
362,363, 364(145), 366, 367(255), Porath, J., 137, 138(137),297, 304(137),
368(255), 372(260), 375(145) 306(145), 311
Percival, E. G. V., 343 Poretz, R. D., 140, 151, 157(180), 166,
Perdomo, J. M., 330 168, 172(362),173(362), 174(365),
Pere, M., 228,230, 231(523), 311(523), 180(365), 181(365),182(365),
335(519) 183(365), 185(365), 188(180, 365),
Pereira, M. E. A., 142, 144(198, 200), 189(180,436), 190(365),200(365),
231, 236(552), 237, 238, 259, 250, 251(175, 184, 598), 252(184),
260(198), 261(203), 284, 285, 286, 253(184), 267, 312, 333, 337(184,
287, 288(200, 526, 686), 313, 598), 338(365), 339(175, 184, 601)
314(803), 316,335(200), 338(552) Portsmouth, D., 48, 55(60), 74(60)
Perera, C. B., 203, 205(468), 339(468) Pospelova, T. A., 54, 56(74), 80(74)
Perila, O., 353 PospiSilovii, J., 141, 321, 322(185),325,
Perkins, H. R., 353 326(851)
Perlmann, H., 243 Powell, A. E., 150,179(249)
Perlmann, P., 243 Powers, D. A., 151, 152(264)
406 AUTHOR INDEX, VOLUME 35

Prasad, D., 38,46(22),51, 53(22, 51, 69), Ragheb, H . S., 107

54(22, 51,69), 55(22, 52), 57(22), Rahman, M. A., 254
58(50, 69), 73(69), 74(51), 75(51), Ramachandra, G., 365
76(22, 50, 52, 69) Ramachandramurthy, P., 296
Pratt, R. M., Jr., 317 Ramirez, G., 326
Preobrazhenskaya, M. E., 168, 171 Randall, M. H., 50,51(67), 53(67), 54(67),
Preobrazhenskii, N. A., 38,45(24) 51(27, 56(67), 77(67), 78(67)
28), 52, 53(27), 54(24, 25, 27,28,71), Rao, C. V. N., 343,346(53), 348(53)
55(77), 56(71), 57(73), 72(24, 25), Rao, M. V. L., 345
80(27, 28, 71, 73, 75, 76, 77) Rapin, A. M. C., l32,318(92a)
Presant, C. A., 187,308, 318(435), Raubitshek, H., 190,291
319(435),320, 321(435), 322(435), Ravid, A., 261, 262(619),338(619)
331(435) Rebers, P., 179
Preston, R. D., 378 Reckin, E., 360
Pricer, W. E., Jr., 131 Redlich, H., 89
Pridham, J. B., 344,345(58), 358,362, Redwood, W. R., 219
363(223),365(223),366(223, 224) Reeber, A., 326
Prigent, M. J., 145,303,337(220) Reeder, W. J., 176
Pringsheim, H., 370 Reeke, G. N. Jr., 137, 138(140), 152,
Prior, A. M., 110 153(325), 154(268,269, 275,276),
Privat, J.-P., 134, 135(115), 149,214, 155(267), 156(268,275), 157(267,
215(490b),216,217,219(115,494, 268,280), 158(267,268,280),
497), 222(494),223(494), 297(243e), 161(280),202(140), 203(140),
340(115) 205(140), 304(276),335(267),
Prokop, O., 129, 131, 133, 145(14a),227, 336(140), 337(140),339(140)
239(58, 60, 61, loo), 241(60, loo), Rees, D. A., 341,349,384
242(61, 100, 556), 243(100), 338(60, Reese, E. T., 362, 369,372(221), 374
61, 100) Refsnes, K., 255, 270,271(651), 272(651),
Pryce, N. G., 349 335(651), 336(651),337(651)
Punin, W., 147 Rege, V. P., 208,209(475), 226(475),
Punnett, H. H., 292 338(475,477)
Punnett, T., 292 Reguera, R. M., 128, 147(4),243(4)
Pusztai, A., 297,301 Reichert, C. M., 143, 148, 168(209,240),
174(240), 178(240),179(209, 240),
Q 181(240), 182(209),183(240),
184(209,240), 185(240), 186(209,
Quicke, G. V., 232 240), 187(209),200(240), 247(240),
Quiocho, F. A., 152 322(209)
Quirt, A., 158, 159(311) Reid, J. S. G., 344,345(70), 346(199),
348(70), 352, 353(143),355, 356(204),
357, 359(199),360,361(201),
R 364(202), 365, 366(204),370(199,
Race, R. R., 129, 133(14),226(14), 201), 371(199, 201,202, 204),
289(14), 339(14) 372(202), 374
Rackis, J. J., 232 Reifenberg, U., 299
Rackwitz, A., 131, 239(58, 60), 241(60), Reiner, R., 102
338(60) Reinhold, V. N., 47,48(56), 55(56),
Rashen, V., 295, 335(718) 65(56), 67(56),68(56), 69(56)
Rafestin, M. E., 214 Reisfeld, R. A., 139, 309(162),
Rdferty, G. A., 58,62 310
Rafique, C. M., 346,347(90) Reitherman, R. W., 139, 290(158),
 AUTHOR INDEX, VOLUME 35 407

336(158, 691), 337(158, 691), Romanowska, E., 303

339(691) Ronalds, J. A,, 372(301),373
Rekunova, V. N., 39, 54(37), 80(37) Rose, A. W., 307
Renkonen, K. O., 128, 132(6),208(6), Rose, J. E., 709
282, 283,304(6), 338(6) Rosen, M., 342
Renwrantz, L., 316 Rosen, S. D., 139,290(158), 308,
Resheff, G., 315 309(779,780), 336(158), 337(158)
Resnick, H., 178, 182(419) Rosenau, W., 296
Rettig, E., 341(12, 13, 16), 342 Rosenfel’d, E. L., 168, 171
Revel, J.-P., 317 Rosenfeld, L., l38,258(150d)
Rice, R. H., 162, 215, 223(335), 337(335, Rosenthal, A. S., 317
495) Rosenwasser, A., 164,239(343)
Richards, E. L., 347 Ross, T. T., 266
Richards, G. N., 345,366, 367(243), Rosset, J., 264
368(243), 369(243) Roth, J., 313,317
Richards, J. B., 355 Roth, K. L., 202
Richards, R. L., 219 Rothfield, L., 355
Richter, M., 130, 291(35) Rougier, M., 356
Richtmyer, N. K., 372(299), 373 Roussel, P., 270,271(650), 272(650),
Rick, P. D., 354 337(650)
Rigas, D. A., 292, 293(697,698), 294(697, Rovis, L., 285, 288(686)
698), 335(697,698, 705, 706, 707), Rowlands, D. T., Jr., 313
339(696) Roy, J., 256,257(611), 334(611), 335(611)
Rinehart, K. L., Jr., 102, 104, 107(53,54), Roychoudhury, R., 65
108(54), 115(53), 116(53), 117(53), Rozenberg, M., 307
119 Rozynov, B. V., 68
Riss, H., 140, 251(184),252(184), Ruckel, E. R., 172
253(184), 337(184),339(184) Rudakova, I. P., 38, 39,44, 51(28), 52,
Rist, C. E., 172 54(28, 37,46, 71), 56(46, 71, 74),
Rivera, A., 292 57(73), 72(46),80(28,37, 71, 73, 74)
Rizk, A. M., 344 Ruddon, R. W., 164,248(342)
Rizvi, S. A. I., 344 Rudloff, E., 65
Robbins, P., 355 Rudowski, A., 38,46(22), 51, 53(22, 69),
Robertson, E. S., 140, 241(178), 54(22, 69), 55(22), 57(22), 58(50, 69),
248(178) 76(22, 50, 69)
Robinson, D. S., 48 Riidiger, H., 305
Robinson, R., 174 Ruelius, H. W., 279, 280(673)
Robos, V. N., 44,45(45), 70(45) Ruff, B. A., 104, 107(53), 115(53),
Robson, E. B., 134, 177(106) 116(53), 117(53), 119
Roche, A.-C., 306,307(770), 335(770) Rule, A. H., 140
Rochmilevitz, T., 356 Rundle, R. E., 378
Ronnback, L., 326 Rydlund, P. H., 342
Rogers, H. J., 353 Rzedowski, W., 366
Rogers, J., 187,320(434),331(434)
Rogers, T. O., 104, 107(54), 108(54)
Roguet, R., 231 S
Roholt, 0. A., 161, 162 Saarnio, J., 344
Roland, J.-C., 356 Sachs, L., 130, 144, 150(31),236(212),
Rollins, A. J., 85(13), 95 238(212), 288, 296, 317(31,250, 554),
Rolls, J. P., 104, 107(53), 115(53), 338(212),339
116(53), 117(53), 119 Sadaksharaswami, M., 365
408 AUTHOR INDEX, VOLUME 35

Sadovskaya, V. L., 68 Scher, M., 355

Sage, H. J., 131, 134(123), 135, 137, Schertz, K. F., 244
138(138), 141, 142(123),144(123), Schiffer, M., 152
145, 148, 149(243), 190(138), Schilling, E. D., 341,342
191(441),192(441), 193(138,441), Schleicher, H., 379
194(123, 138, 186), 195(123), Schlesinger, D., 131, 239(58, 60),
202(138),308(51, 221), 335(138,441), 241(60), 242(556),338(60)
338(123, 138),340(186) Schmid, R., 82,83(5), 89, 90(24), 91(24),
Saint-Paul, M., 147 lOO(5)
Saito, N., 46, 76(54) Schmidt, E. L., 147
Sakakibara, F., 315 Schmidt, P., 65
Sakakibara, K., 172 Schnebli, H. P., 150
Sakamoto, C. K., 328, 329(866) Schnitzler, S., 239, 240(563)
Sakurai, Y., 137, 138(139),202(139), Schott, H., 64,65
205(139) Schuerch, C., 172
Salfner, B., 299, 302(736) Schulze, R. E., 342
Sallach, H. J., 97 Schumacher, K., 299,301(735), 302(736)
Sallam, M. A. E., 343, 346(54) Schuppner, H. R., 341
Saltmarsh-Andrew, M., 355 Schussler, W., 40, 41(41), 52(41), 53(41),
Salton, M. J. R., 139,306(155),307(155) 70(41), 71(41), 78(41)
Saltvedt, E., 137,254(147),255(147), Scocca, J. R., 333
256(147), 257(147), 270(147), Segrest, J. P., 319, 320(843), 321(843),
271(147), 272,273( 147), 274(147), 824(843)
276,334( 147),335(147), 336(147), Sehgal, K., 356,361(200)
337(147),338(147), 339(147), Seib, P. A., 58
340(147) Seidl, D. S., 166
SalvetovL, A., 192,335(444) Seiler, A., 356(203), 357, 365, 372(203)
Sambrook, J., 214,318(489) Sela, B.-A., 139, 144,224(156),236(212),
Sandermann, H., 355 238(212), 239, 296(156), 317(554),
Sanderson, A. R., 176 338(212)
Sandoz, D., 356 Self, R., 48
Sanford, B. H., 130,317(29) Sen, A., 256,257(611), 334(611), 335(611)
Sanger, R., 129, 133(14), 226(14), Seshadri, T. R., 345
289(14), 339(14) Seto, H., I24
Sangster, I., 38,46(22), 53(22), 54(22), Seymour, E., 53
55(22),57(22), 76(22) Shafer, J. A., 140, 159(174), 160, 163,
Saniewski, M., 376 164(339), 179(174), 185
Sarko, A., 377, 378, 381 Shafizadeh, F., 48, 54(61), 62(61), 79(61)
Sarma, V. R., 221 Shankar Iyer, P. N., 135, 137(125),
Sasada, Y., 46, 76(54) 138(125), 142(125), 146(125),
Sasame, H. A., 232 206(125), 207(125), 262(125),
Sastry, P. S., 363 265(125), 335(125), 338(125, 470)
Sawai, K., 361, 370(219),374(219) Shannon, L., 314
Sawyer, W. H., 152, 153, 154(279), Shaper, J. H., 136, 337(130)
160(266), 164(279),165(290) Shapleigh, E., 128, 129, 131, 132(2), 134,
Schaffer, J. W., 172 147(2),224,243(2), 244(2, 3, 103),
Schaffner, C., 110 245(103), 248(103), 289(12, 13),
Schaub, R. E., 58 339(2, 12, 13)
Schechter, B., l64,239(343a) Sharma, B. R., 345
Scheinberg, I. H., 131 Sharon, N., 129, 130(18,26,27), 132(18,
Scheinberg, S. L., 244,246, 247(586), 26, 37b), 133, 134(97), 136(18),
335(586, 588) 137(18, 132), 142, 143(128), 144(128),
 AUTHOR INDEX, VOLUME 35 409

147, 150(72), 164, 191, 198(442c), Skutelsky, E., 143, 258(202), 259(202),
208(18), 209(18),214, 215(491), 216, 261(202), 335(202), 338(202)
218, 219(504),222(504), 223, 232(97), Slechta, L., 119
233(538), 234(538, 541), 235(442c, Slessor, K. N., 46, 53(53),57(53)
537, 538, 539, 544, 546), 236(212, Slodki, M. E., 168, 174, 354
552), 237, 238(99, 212, 552), 239(343, Sly, D. A., 151, 152(264)
343a),258(202), 259(202, 203), Small, D. M., 347,369(116)
260(198), 261(202, 203). 262(619), Small, P. A., 310
291(37a), 296, 317(554), 318(18), Smith, A. K., 232
335(202, 538, 544, 546), 337(128), Smith, B. C., 47, 50(55), 53(55), 57(55),
338(202, 212, 552, 619), 339(128), 74(55), 75(55), 79(55)
340(544) Smith, C. W., 89
Shaw, Y.-S., 255,271 Smith, D. F., 327, 328(861, 862, 863)
Sheehan, J. K., 383, 384(28) Smith, E. E., 140, 142(168), 143(168,
Sheichenko, V. I., 56 197), 144(168),150(168), 151(168),
Sherman, W. R., 69,70 166, 167(168, 197, 359), 168(359),
Sherry, A. D., 156 170(359),171(359), 173, 179(168,
Shibata, Y.,362, 369, 372(221), 374 197), 180(168, 197), 181(168, 197),
Shier, W. T., 134,206(114), 216,222 183(197), 184(168, 197,430),
Shimahara, H., 346,367, 370(265) 185(168, 197,430), 187(197,430),
Shimanouchi, H., 46, 76(54) 200(197), 338(168)
Shimazaki. K., 272 Smith, F., 134, 140(117), 151, 166(117),
Shinkai, K., 157, 324(306b) 168(353,354,355),169(353),
Shinohara, T., 197, 199(449,452), 171(260), 173(353), 175(364),
200(452), 339(449,452) 179(117), 184(364),343, 345, 346(84),
Shiroya, T., 360,365(210) 347(90), 348(99),350(99)
Shishido, K., 173 Smith, P. J. C., 384
Shmyrev, I. K., 44,45(45), 70(45) Smith, S. B., 317
Shoham, J., 150 So, L. L., 134, 135(120), 140(120),
Shoham, M., 155 143(169), 144(169), 151(120),
Siddiqui, I. R., 38,43,45(32), 53(32), 65, 153(172), 160, 162, 165(204),
67(32, 44, 99), 70(32, 44), 72(32, 44), 166(320), 167(120,204), 168(120,
73(32,44) 169, 204, 215, 320), 169(320), 170,
Sidebotham, R. L., 171 171(120,369), 172(320,362),
Siegelman, H. W., 234,235(544), 173(320,362,369), 174(120, 204),
335(544),340(544) 175(204,215,320), 178, 179(120,
Sigarlakie, E., 174, 175(388) 169, 172,215), 180(169,204,215),
Sihtola, H., 378 181(169,204,215), 182(169),
Sikl, D., 376 184(120, 169, 215, 430), 185(204,
Silber, R., 297,300 430), 187(430),196(204),200(169),
Silberschmidt, K., 302 212(204), 338(169, 204, 215),
Silman, I., 315 340(172)
Simpson, D. L., 308, 309(779,780) Soboczenski, E. J., 32, 38(8), 50(8), 51(8),
Singer, S. J., 317 77(8), 7863)
Singh, G., 210 Soderman, D. D., 130, 150(40)
Singla, S., 140, 251(175), 339(175) Somme, R., 344, 345(76), 372(303, 304),
Sinha, M. P., 344 373,375
Sinnwell, V., 99 Sohonie, K., 196, 197(448),339(448)
Sioufi, A., 352, 353(145), 360(145), Solms, J., 32, 38(7), 43(7), 44(7), 50(7),
364(145),375(145), 52(7), 64,72(7), 73(7)
Skorvaga, M., 109 Som, S., 256, 257(611),334(611),
Skoyles, D., 56 335(611)
410 AUTHOR INDEX, VOLUME 35

Somers, G. F., 32,49(5) Subrahamnyan, V., 343

Somers, P. J., 49,63(64),65(64) Suescun, E., 202
$om, F., 75(114),77 Sufoka, A., 380
Southworth, D., l49,229(243a) Sugahara, K., 372,373(286),
374(286)
Speckart, S. F., 219 Sugihara, J. M.,32,35(9),
36(9),38(9),
Spencer, J. F. T., 343,354(41) 41(9),50(9),52(9),77(9),78(9)
Spiro, R. G., 300 Sugimori, T., 123
Sprenger, I.,227,267, 268(632,636), Sugino, Y., 312,340(796)
269(632),297(632),300(632), Sugishita, S., 277,338(670)
333(632),339(632,636) Sugiyama, N., 346,367,370(265)
Springer, G . F., 131,134,140,144(167), Suhadolnik, R. J., 122,123,124(86),126
145,277,278(167,676), 279(118, Sukeno, T., 372(292),373,374(305)
167),280(66,167,671,673,676), Sulkowski, E., 141,168(192),178(192),
281(66,671,677),282(679),284,285, 179(192,418)
302,303(219,678,758),304(678, Sumner, J. B., 134,136(102),140(102),
758),318(678),335(679),338(167, 151,152(257,258),155,166(102),
674,676),339(752,757,758) 168(102,352),169(102),173(102),
Srivastava, H.C., 380 177(102),179(102),335(258),
Srivastava, V. K., 380 338(102,246,247)
Stacey, B. E., 39,45(33), 70(33),72(33), Sundberg, L., 304
73(33) Sundblad, G., 178,216, 217(498),218(498),
Stacey, M., 48,74(59),172 219(498),222(498),301(413),
Stadler, P., 99 33l(413),339(498)
Staub, A. M., 179 Surolia, A., 141,168(193),177,179(193),
Stead, R. H., 232 272,274(193),275(659),276,
Steck, T. L., 323 277(669)
Steigerwald, J. C., 372,373(290) Susz, J. P., 326
Stein, J. Z., 347 Sutoh, K., I38,258(150d)
Stein, M. D., 141,190,191(441), Suzuki, H., 171,186(372),363,367
192(441),193(441),194(186), Suzuki, K.,131,307(52),308(52)
335(441),340(186) Suzuki, N., 85,86,87(16), 88(16),95
Steinberg, M. S., 165 Suzuki, S., 85,86,87(16),88(16)
Steiner, E. A., 263 Suzuno, R., 157
Steinhausen, G., 282,314, 315(806) Svenson, R. H., 293,294(708, 714),
Stepanenko, B. N., 343 295(708),335(708,713,714)
Sternlicht, H., 155,158,161(307,308), Svensson, S., 62,174,178,282,291(682),
163,179(308),201(308) 301(413),331(413),349
Stillmark, H., 128,254,258(603), 270, Sweeley, C. C., 355
338(603),339(603) Swenson, H. A., 33
Stobo, J. D., 317 Sykes, B. D., 163,164
Stocked, R. J., 131
Stoddart, R. W., 317
Stone, K.J., 355
Stoudt, T. H., 104, 107(54),108(54) T
Strauchen, J. A., 300 Takagi, M., 86, 87(16),88(16)
Streips, U. N.,177 Takahashi, N.,333
Strominger, J. L., 94,176,355 Takahashi, T., 134(123),135,142(123),
Strosberg, A. D., 191,198(442a,44213, 144(123),194(123),195(123),270,
442c),235(442c) 271(642),272(642),278(676),279,
Stroshane, R. M., 102,104,107(53), 280(676),296,338(123, 676),
115(53),116(53), 117,119 339(641,642)
Subbarao, P. V., 345 Takatsu, A., 239,240(560),241(560)
 AUTHOR INDEX, VOLUME 35 411

Takayama, K., 355 Tisdale, V. V., 292,293(698), 294(698),

Takayanagi, G., 315 335(698)
Takemoto, M., 346,367, 370(265) Tiwari, R. D., 344
Takenishi, S., 367 Tixier, R., 214,215(490b)
Talbot, C. F., 149,229 Tkacz, J. S., 174
Tanaka, K., 134, 177(107),300(107) Tobiska, J., 132,313(80)
Tanaka, Y., 376 Todd, L. S., 179
Tanigaki, Y., l57,324(306b) Tokuyama, H., 310,311(788)
Taniguchi, M., 104, 107(53,54), 108(54), Tominaga, S., 177
115(53),116(53), 117(53),119 Tomita, M., 137, 138(139, 144), 202(139),
Tanner, W., 352,355 205( 139), 254(144), 270(144),
Tarentino, A. L., 372(292),373, 374(305) 271( 144)
Taylor, J. R., 372(296),373 Toms, G. C., 132
Taylor, K. J., 378 Tonornura, A., 137, 138(143),142(143),
Tegtmeyer, H., 145,302, 303(219, 758), 144(143),190(143),191(143),
304(758),339(758) 192(143), 194(143), 195(143),312,
Teichberg, V. I., 315 335(143), 338(143)
Tell, G. P. E., 130, 150(38) Tookey, H. L., 344,348(61, 62)
Terao, T., 140, 162, 163(332), 165(332), Torii, M., 172, 176
238(183), 250,251(183), 252(183), Torssell, K., 32,48, 49(5)
254(183),258, 259(616),261(183, Toyoshima, S., 137, 138(143),142(143),
616), 273(183),275(183), 297(183), 144(143), 178, 190(143),191(143),
300(183),304(183),310, 318(183), 192(143), 194(143), 195(143),
33 1(183), 335(616), 339(183) 2 97(417), 300(417), 301(417),
Teresa, G. W., 148,267, 268, 269(635), 304(417),312,335(143), 338(143),
339(631, 635) 372
Testa, R. T., 110, 112, 113(71) Trejo, G., 354
Thambi-Dorai, D., 177 Treska-Ciesielski, J., 326
Thierne, T. R., 354 Trowbridge, I. S., 197, 198(453),
Thimann, K. V., 371 199(453),337(453),340(453)
Thomas, D. B., 319,321 Troy, F. A., 354,355
Thomas, M. W., 215,223(495a) Tmitt, S. G., 104, 107(53), 115(53),
Thomasson, D. L., 156, 160(297),165 116(53), 117(53)
Thompson, T. E., 315 Tschirch, A., 345, 356(82)
Thorpe, T. A., 366,370(254), 371(254) Tserng, K.-Y., 130,254(32),271(32)
Thoss, K., 317 Tsuda, M., 312, 340(796)
Thunell, S., 292 Tsujisaka, Y., 367
Tichi, M., 134(122),135, 136, 137, Tully, R. E., 149
138(142),190(142),191(142,442), Tung, T.-C., 130, 254(32), 255,271(32)
192(442),193(142),194(440,442, Tunis, M., 245, 247(579)
447), 196, 197(440),199(121,440), Turner, R. H., 130
201(121,446), 335(142,442,444), Tyminski, A., 353
338(142), 339(122,440)
Tichf, M., 192
Tierney, B., 39,45(33), 70(33), 72(33),
73(33)
U
Tieslau, C., 130,317(28) Uchida, T., 156, 193(300)
Tilley, B. C., 112, 113(71) Uda, F., 86,87(16), 88(16)
Timberlake, J. W., 140, 251(184), Uernatsu, T., 123, 124(86)
252( 184),253(184), 337(184), Uhlenbruck, G., 129, 131, 133, 143,
339(184) 144(206),145(14a),211(206),
Timell, T. E., ,345,353,367(150) 212(206),213(206),218,224(502),
412 AUTHOR INDEX, VOLUME 35

227, 239(61, loo), 240, 241(100, 567), Vesterberg, O., 366

242(61, 100, 556, 567), 243(100, 567), Vicari, G., 253
258(201),259(201), 267, 268(632, Vieweg, H. G., 358
636), 269(632),276(206), 282, 299, Villafranca, J. J., 157, 158, 161(309),
300(132),301(735), 302(736), 165(309), 167(309),179(309)
303(754),304(755, 761), 305(761), Villarroya, E., 363
306(761),307, 314,315(806, 807, Villarroya, H., 350, 366, 367, 368(140,
808), 333(632), 338(61, 100, 201, 617, 253), 369(140, 253)
618), 339(206, 632, 636, 754, 755, Villiers, T. A., 358
756) Viola, R. E., 157, 158, 161(309), 165(309),
Ukita, T., 137, 138(139, 144), 202(139), 167(309), 179(309)
205(139), 254(144), 270(144), Vlodavska, I., 288
271(144) VO@, W.-E., 239, 240(563)
Ulevitch, R. J., 267, 268(633) Vretblad, P., 137
Ullrich, J., 96
Umemoto, J., 219
W
Umezawa, S., 81, 102(1)
Unrau, A. M., 344,345(65), 346(65), Wada, S., 235
347(90), 361(69) Waechter, C. J., 355
Unrau, I. C. J., 344 Wagenknecht, W., 379
Unzelman, J. M., 356 Wagman, G. H., 110, 111
Uy, R., 138 Wagner, M., 317
Wahl, H. P., 99, 101, 109(40)
Walborg, E. F., Jr., 215, 223(495a), 272,
V
327,328(861, 862,863)
Valdovinos, J. G., 371 Waleroft, M. J., 130, 150(33)
Valueva, S. P., 55 Waldschmidt-Leitz, E., 270 271(646,647)
Van Cleve, J. W., 172 Walker, D. L., 96
Vani.urovi, D., 196 Walker, J. B., 98, 104(36), 106, 108(36),
Vandenheede, J. R., 268(637),269, 109(36)
339(637) Walker, M. S., 109
Van Driessche, E., 191, 198(442a,442b) Walker, R. E., 342
Van Landschoot, A,, 216 Wall, H. M., 58,59(84), 62
Van Wageningen, T., 302 Wallach, D. F. H., 323
Van Wauwe, J. P., 134(124),135, Wallenfels, K., 363
140(124),144, 160, 188(181, 323), Walter, M. W., 360
189(181,323), 195, 197(182), Wang, J. L., 137, 138(140), 139, 151, 152,
199(182,211),200(211),273,274(124), 153, 154(263,268,269, 275),
339(124, 211), 340(124),376 155(263, 267,270, 271, 289),
Vargha, L., 31 156(268,275), 157(267,268),
Varner, J. E., 360 158(267,268), 160(283,289), 162,
Varsharskaya, L. S., 38, 39, 51(27,28, 163(333), 165(333,336), 202(140),
36), 53(27,36), 54(27, 28, 35), 80(27, 203(140,263), 205(140), 224(156),
28, 35, 36, 76) 296(156), 335(267, 271), 336(140),
Vartia, K. O., 131 337(140), 339(140)
Vasquez, J. J., 145,308(221) Wgngstrbm, B., 349
Verenikina, S. G., 37,38,42(26), 44(26), Ward, R. M., 168, 174(363)
45(24), 54(28,25,46), 56(46), 70(26), Waszczenko-Zacharczenko, E., 132,258,
72(28,25,26,46), 73(26), 78(26), 80 259(614), 313(85)
Verma, S. D., 210 Watanabe, K., 216,217(499), 315
Vesel?, P., 201 Watanabe, S., 380
 AUTHOR INDEX, VOLUME 35 413

Watkins, W. M., 129, 131(21, 22), 140(21, 206(125), 207(125), 262( 125),
22), 141, 142(22),207, 208(22, 24, 265( 125),335(125), 338(125)
471), 209(471,475), 210(471), Williams, D. G., 378
225(24), 226(475), 244, 248(22), 250, Williams, J., 350, 368(140), 369(140)
253(22), 277(21), 282(22), 283, Williams, J. H., 58
284(22),305(471), 338(21, 471, 475), Williams, N. R., 58, 62
339(22) Williams, P., 140, 144(167), 278(167),
Watson, P. R., 354 279( 167), 280(167), 284( 167),
Watson, R. R., 100 285( 167), 338(167)
Watt, W. B., 297 Williams, T. J., 185
Waxdal, M. J., 152, 155(267, 270, 271), Williamson, J. R., 343
157(267), 158(267), 310, 335(267, Williamson, P., 144
271) Willoughby, E., 355
Webb, E. C., 148 Wilson, K., 283, 284(684), 335(684)
Webber, J. M., 45, 50, 51(67), 53(67), Wingham, J., l32,224,226(508d),257
54(67), 56(67), 72(49), 77(67), 78(67) Winter, W. T., 378,384
Weber, T. H., 292, 293, 294(715, 716), Wintzer, G., 299 301(735), 302(736)
295, 335(715, 716) Winzler, R. J., 319,321
Wecksler, M., 150 Wirtz-Peitz, F., 47,48(56), 55(56), 65(56),
Wei, C. H., 255, 256, 257, 271, 334(160, 67(56), 68(56), 69(56)
607), 335(607, 610) Wisnieski, J. J., 179, 184(424)
Weigel, H., 35, 38,42(23), 43, 45, 50(23), Wissler, F. C., 307
51(23), 53(23,43,48), 54(23), 55(23, Wold, F., 138
43, 48), 57(23), 58,60(86), 66, 67(10), Wolfrom, M. L., 32, 38(7), 43(7), 44(7),
72(23, 48), 77(23,43), 78(23,43), 172 50(7), 52(7), 72(7), 73(7)
Weinbaum, G., 123 Wolpert, J. S., 147
Weiner, H., 107 Wood, C., 208
Weiner, I. M., 355 Wood, G., 96
Weinstein, B., 341(7), 342 Wood, M. K., 152
Weinzierl, J., 156, 161(296) Wood, P. J., 38,43,45(32), 53(32), 65,
Weisenthal, L. M., 164, 248(342) 67(32,44, 99),70(32, 441, 72(32,44),
Weiss, A. K., 132, 139(94) 73(32,44)
Weisz, J., 65 Woodbury, R. R., 49,63(64), 65(64)
Weith, H. L., 63 Woodruff, J., 297
Wellum, G. R., 34,35(11) Woodside, E. E., 159, 168, 179(366)
Welsch, P. D., 139,309(162) Worth, H. G. J., 360
Wester, D. W., 39 Wray, V. P., 327
Western, A., 132 Wright, A., 166, 168(357, 358), 169,355
Westholm, F. A., 151, 152(264) Wright, C. S., 215, 216(496a)
Westoo, A., 240, 241(569), 243(569), Wright, D. C., 58
335(569) Wu, T. T., 154
Whistler, R. L., 343,344,347, 348, 349, Wuilmart, C., 191, 198(442c), 235(442c)
350( 128), 362, 370(220)
Whitaker, P. M., 379 Y
Whyte, J. N. C., 345,346(85), 347(85)
Wickstrbm, A., 345, 348(86) Yabroff, D. L., 36
Wiebers, J. L., 38, 63, 68(29) Yachnin, S., 134, 267,294(630, 714),
Wiecko, J., 69, 70 295(630, 712), 301(111), 335(630,
Wiener, A. S., 145 712, 713,714)
Wilkinson, K. D., 135, 137(125), Yahara, I., 152, 154(269), 162, 165(336)
138(125), 142(125), 146(125), Yamamoto, T., 366,367(251)
4 14 AUTHOR INDEX, VOLUME 35

Yamamura, Y., 376 194(109, 123), 195(123),335(441),

Yamano, T., 120, 122(81) 338(109, 123)
Yamashina, I., 372, 373(286),374(286) Yudis, M. D., 110
Yamazaki, S., 85, 95(14) Yueh, M. H., 341,342
Yancik, J. J., 342 Yurkevich, A. M., 37, 38,39, 42(26),
Yang, C. K., 246 44(26), 45(24), 46,49, 51(27,28, 36),
Yang, W.-K., 239, 255,334(607), 335(607) 52, 53(27,30, 36), 54(24,25, 27, 28,
Yang, Y., 168, 172(362),173(362),178 30, 35,37,46,47, 71), 55(77), 56(46,
Yano, O., 310 71, 72, 74), 57(73), 63, 64,70(26),
Yariv, J., 138, 153, 155(278), 179(278), 72(24, 25,26,46,47), 73(26), 78(26),
283, 284(684),335(684) 80(27, 28,30, 35, 36, 37, 71, 73, 74,
Yasuda, Y., 333 75, 76, 77)
Yesner, I., 297
Yokoyama, K., 310
Yomaguichi, T., 239
Yomo, H., 360 2
Yonehara, H., 124 Zand, R., 153,263
Yoshimura, J., 85, 95(14) Zanetta, J.-P., 326, 327(857)
Yosizawa, Z., 248,250,253(596) Zaslow, B., 378
Youle, R. J., 149 Zeissig, A., 150, 338(246)
Young, F. E., 177 Ziaya, P. R., 229
Young, N. M., 134(123),135, 142(123), Ziegenfuss, E. M., 341(11), 342
144(123), 162, 178(log), 190, Ziska, P., 205
191(441), 192(441),193(441), Zweifel, G., 58
 SUBJECT INDEX FOR VOLUME 35

A interaction with lectins, 180

paper chromatography, phenylboronic
Aaptos papillata lectins, isolation and acid in, 60
properties, 316 Aldohexose
Abrin 2,3,6-trideoxy-46-(2-hydroxyacetyl)-
carbohydrate-binding specificity, 255 L-threo-, 96
immunization to toxic, 129 Algae, polysaccharides of' marine, 8, 9
isolation and purification, affinity Alginic acid, structure, 7, 8, 10
chromatography, 138 Allopyranoside
purification and properties, 254-257 methyl 6-deoxy-p-~-,2,4-phenyl-
Abrus precatorius, seed-extracts, boronate, oxidation, 57
hemagglutinating and toxic activi- preparation, 48
ties, 254-257 Amino acids
Abrus precatorius lectin, see Abrin of asparagus-pea lectin, 284
Acetylation, of carbohydrate boronates, of Bauhinia purpurea alba lectin, 305
53 of castor-bean lectin, 272
Actinamine, biosynthesis, 119 of concanavalin A, 162
Actinospectinoic acid, structure, 119 of discoidin, 309
Adenine of eel-serum lectin, 281
9-P-D-arabinofuranosyl-,biosynthesis, of horse-shoe-crab lectin, 306
123, 124 of lectins, 335,337
arabinosyl-, biosynthesis, 124 of lentil lectin, 191-193
Adenosine of lima-bean lectin, 247
complex formation with phenylboronic of Osage-orange lectin, 267
acids, 49 of pea lectins, 197
5'-phosphate, preparation, 54 of peanut lectin, 258
-, 3'-amino-3'-deoxy-, biosynthesis, 123 of potato lectin, 211
-, 5'-0-trityl-, preparation, 55 of red kidney-bean lectin, 292,293
Agaricus bisporus lectin, see Mushroom of ricin, 271
lectin of snail lectin, 240
Agglutinins, see also Lectins of soybean lectin, 234
plant, history, 128 of sun-hemp lectin, 306
Aldgamycin C , periodate oxidation, 90 of Ulex europeus lectin, 289
Aldgamycin E, components, and degra- of wheat-germ lectin, 215
dation, 89-91 Amino acid sequence, of concanavalin A,
Aldgarose, D, biosynthesis, 89-91 152
Alditols Amylopectin
borinate-boronates, preparation, 40 protozoal, 11
boronates, acetates and benzoates, 53 reaction with concanavalin A, 169-171
preparation and structure of, 42,43 Amylose, V-, crystal structure bibliogra-
properties, of, 77,78 phy, 378
column chromatography, boronic acids Anguilla anguilla serum lectin, see Eel-
in, 63 serum lectin
electrophoresis, sulfonylated Anhydro sugars
phenylboronic acids in, 62 boronates, properties, 79
gas-liquid chromatography, boronic paper chromatography, phenylboronic
acids in, 65 acid in, 60

415
416 SUBJECT INDEX, VOLUME 35

Antibiotics Barley lectin, isolation and properties,

aminocyclitol, biosynthesis, 81, 314
102-122 Barry degradation, of carbohydrates, 9
anthracycline, 91-96 Bauhinia purpurea alba lectin
biosynthesis of sugar components, carbohydrate-binding specificity, 305
81-126 composition and purification, 305
macrolide, glycosidic, 81 isolation, 138
nucleoside, biosynthesis, 122-126 Beans, lectins and toxic properties, 130
Antibodies, determination, 129 Benzoylation, of carbohydrate boronates,
Antibody activity, of lectins, 147 53
Antigens, lectins and blood-group, 129 Bibliography
Apiose, D-, biosynthesis, 100, 101, 125 of crystal structures of polysaccharides,
Arabinitol, 1,5-dideoxy-~-, 377-385
phenylboronate, structure, 43 of Edmund L. Hirst and colleagues
Arabinogalactans, interaction with publications, 17-29
concanavalin A, 179 Biochemistry, of plant galactomannans,
Arabinomannans, interaction with 34 1-376
concanavalin A, 179 Biosynthesis
Arabinopyranose, 1,2-O-isopropylidene- of antibiotic sugar components,
p-L-,3,4-phenylboronate, 81-126
preparation, 39 of plant galactomannans, 352-356
Arabinose, L-, Black-locust lectin, see Robinia
butylboronates, hydrolysis, 51 pseudoaccacia lectin
preparation, 43 Blasticidin S, biosynthesis, 124, 125
phenylboronates, hydrolysis, 50, 51 Bluensidine, biosynthesis, 105, 107
preparation, 43 Bluensomycin, structure, 102, 103
Arachis hypogaea lectin, see Peanut Boranediyl, nomenclature, 36
lectin Boric acid, reaction with D-glucose, 31
Ascorbic acid, history, 5, 6 Borinates
Asparagus-pea lectin boronates from, 39-41
carbohydrate-binding specificity, carbohydrate, preparation, 70
285-288 Boronates
composition, 284 in aqueous solutions, 48-52
isolation, 138 carbohydrate, 5, 31-80
purification, 283 esterification, 53
Axinella polypoides lectins, isolation and etherification, 55
properties, 316 hydrolysis, 50-52
mass spectrometry, 65-70
nomenclature, 36
B
nuclear magnetic resonance
Bandeiraea simplicifolia lectin, isolation spectroscopy, 70
of, 138 nucleophilic displacement reactions,
Bandeiraea simplicifolia I lectin 55-57
carbohydrate-binding specificity, oxidation, 57
264-266 preparation and structure, 43-45
purification and properties, 262 properties, 72-80
Bandeiraea simplicifolia I1 lectin removal of boronate group, 52
carbohydrate-binding specificity, 206 in separation of carbohydrates, 57,
isolation, properties, and structure, 58
206-208 stability, 35, 53-55
Bandeiraea simplicifolia seeds, lectin structure, 33,41-48
isolation, 137 synthesis, 37-41
 SUBJECT INDEX, VOLUME 35 417

Boronic acids -, Op-tolylsulfonyl-, crystal structure

in biochemistry, 32 bibliography, 380
in column chromatography, 63-65 Cellulose I, green algae, crystal structure
in gas-liquid chromatography, 65 bibliography, 381
interactions with carbohydrates in Cellulose 11, crystal structure bibliogra-
aqueous solutions, 48-52 phy, 379,380
Borylene, nomenclature, 36 Cellulose 111, crystal structure bibliogra-
phy, 380
C a-Chitin, crystal structure bibliography,
38 1
Calcium hyaluronate, crystal structure P-Chitin, crystal structure bibliography,
bibliography, 383 38 1
Canavalia ensifomis, see Jack bean Chitosan, crystal structure bibliography,
Canavalia ensifomis lectin, see 382
Concanavalin A Chloroacetylation, of carbohydrate
Cancer therapy, lectins in, 130 boronates, 54
Carugana uborescens lectins, purifica- Chromatography
tion, composition, and properties of, affinity, of concanavalin A, 157
3 13 in lectin isolation and purification,
Carbamic acid, N-phenyl-, esterification 137
of carbohydrate boronates by, 54 column, boronic acids in, 63-65
Carbohydrate-binding specificity, of gas-liquid, boronic acids in, 65
lectins, 139-145,331-333, see also history, 6
specific lectins paper, phenylboronic acid in, 58-62
Carbohydrates Cladinose, L-, biosynthesis of, 82
boronates, see Boronates Cobalt compounds, carbohydrate, prepa-
interaction with boronic acids in ration, 56
aqueous solutions, 48-52 Coffee beans, a-D-galactosidase from, 363
phosphates, mass spectrometry of Concanavalin A
boronates, 69 acetylation and succinylation, 162
separation, by use of boronates, 57, 58 amino acids, 161, 162
Carob galactomannan, structure, 349 amino acid sequence, 152
Carob gum, structure, 349 carboh ydrate-binding specificity, 142,
Castor-bean extracts, agglutinating 143,157-164,179-190,204
action, 128 complex formation, 163, 168, 178
Castor-bean lectin, see also Ricin crystal structure, 152
carbohydrate-binding specificity, hemagglutinating activity, 201
273-276 interaction with amylopectin, 169-171
interaction with cellular structures, with cellular structures, 317
317 with dextrans, 166, 171-173
isolation, purification, and properties with D-fnictans, 175
of, 137,270,272 with glycogen, 169-171
Cellobiose, and cellulose structure, 5 with glycoproteins, 177-179
Cells, see also Neuronal cells; Tumor with mannans, 173-175
cells with polysaccharides, 166-169, 179
interaction with lectins, 317-333 with teichoic acids, 175-177
Cellulose isolation, 136, 138
regenerated, crystal structure bibliog- of jack bean, physical and chemical
raphy, 379 characterization, 150-157
structure, 5 mono-, di-, and tetra-valent, prepara-
-, sodio-, crystal structure bibliography, tion and biological activities, 164,
379 165
418 SUBJECT INDEX, VOLUME 35

preparation and properties, 150, 151 diphenyl-, formation, 67

structure, 153-155 Disaccharides
Cordycepin, biosynthesis, 123 interaction with concanavalin A, 184
Cosmetics, plant galactomannans, 342 with lectins, 142
Cotyledon, galactomannans, 346, structure, 5
356-361 Discoidin, preparation, composition, and
Crotalaria juncea lectin, isolation, puri- properties of, 309
fication, and properties of, 306 Dolichos bijlorus lectin
Crystal structure carbohydrate-binding specificity, 229
of concanavalin A, 152 circular dichroic spectrum, 230
of polysaccharides, bibliography, hemagglutinating activity, 226
377-385 isolation, 138
Cucumber, wild, lectin, isolation, 138 purification and properties of, 227-229
Cyclitol, amino-, antibiotics, uses, 231
biosynthesis, 81, 102-122 Drilling, well, plant galactomannans in,
1,3-cis-Cyclohexanediol,cyclic 342
boronates, preparation, 36
1,4-Cyclohexanediol, cyclic boronates,
preparation, 36 E
Cycloheximide, inhibitor of
galactomannan degradation, 360 Echinocystis lobata lectin, isolation, 138
Cytidine, 5'-phosphate, preparation, 54 Eel, electric, electrolectin from, 315
Cytisus sessilifolius lectin Eel-serum lectin
carbohydrate-binding specificity of, agglutinating activity, 277
209 carbohydrate-binding specificity,
hemagglutinating activity, 208 277-279
Cytosine, 1-(P-D-glUCOpyranOSylUrOniC purification, 281,282
acid)-, formation, 124 Ehrlich cells, lectin-reactive, 329,330
isolation, 124 Electrolectin, isolation and properties,
315
Electrophoresis
D
affinity, for lectin detection, 135
Deboronation, of boronates, 52 sulfonylated phenylboronic acid in, 62
Decoyinine, biosynthesis, 123 Electrophorus electricus, electrolectin,
Degradation, see also Barry degradation; properties, 315
Smith degradation Endo-p-D-mannanase, in galactomannan
biochemical, of galactomannans, degradation during germination of
356-375 seeds, 361
Desosamine, D-,biosynthesis, 84, 122 Endosperm, galactomannans in, 345,
Dextrans 355,356-361
interaction with concanavalin A, 166, Enzymes
171-173 in biosynthesis of tylosin, 84
with lentil lectins, 194 for galactomannan degradation,
Dictyostelium discoideum lectin, 361-375
isolation and properties, 308 Enzymic analysis, of carbohydrates, 11
Dinucleoside phosphates, synthesis, 54 Erythritol, 4-deoxy-~-,phenylboronate,
1,3,2-Dioxaborinane, nomenclature, 36 structure, 43
-, 5-hydroxy-2-phenyl-, nomenclature, Erythrocytes
36 lectin-reactive glycoproteins, 318-325
1,3,2-Dioxaborolane, nomenclature, 36 lentil lectin in hemagglutination, 193
-, 2-methyl-, preparation, 39 Erythromycin A, formation, 85
1,3,5,2,4-Dioxazadiborepane,2,4- Erythromycin C, biosynthesis, 84
 SUBJECT INDEX, VOLUME 35 419

Esterification of carbohydrate boronates, -, tri-0-methyl-D-, synthesis, 11

53 L-Fucose-binding lectins, 277-291
l,e-Ethanediol, reaction with Fucoxylomannan, isolation, 282
trimethylborane, 39 Furze seed, see Uler europeus I1
Etherification, of carbohydrate
boronates, 55
Euonymus europeus lectin
G
carbohydrate-binding specificity, 145
purification and properties, 313 Galactan, e-, from larch, 10
Explosives, plant galactomannans in, 342 Galactitol
1,6-bis(diethylborinate) 2,3:4,5-bis-
(ethylboronate), preparation, 40
2,3:4,5-his(ethylboronate),prepara-
F
tion, 41
Fava-bean lectin 1,6:2,3:4,5-tris(ethylboronate),prepa-
agglutinating activity, 201,202 ration, 40, 43
carbohydrate-binding specificity, tris(phenylboronate), preparation, 4 1
203-205 Galactomannan, peptido-0-phosphono-,
isolation, 138 biosynthesis, 354
purification and properties, 202 Galactomannan depolymerase, in seed
Favism, fava-bean, 202 germination, 361
Fenugreek seeds Galactomannans
components of immature, 352,355 biochemistry of plant, 341-376
galactomannan biosynthesis in, 352, hiosynthesis, 352-356
355 function, 375,376
galactomannan location in, 345 isolation, 345
a-D-galactosidase in, 365 location in oiuo, 345
germination, galactomannan degrada- metabolism during seed germination,
tion during, 356-361 356-361
P-”mannanase from, 367 in micro-organisms, 354
P-D-mannosidase in, 374 occurrence, 343-345
oligo-P-Dmannos yl-( 1+4)- structure, 347-351
phosphorylase from, 375 uses, 341,342
Fetuin, separation, 277 Galactopyranoside, methyl 6-deoxy-a-~-,
Fire-fighting, plant galactomannans in, 3,4-phenylboronate, preparation, 48
342 -, methyl 2,3-di-O-benzoyl-a-~-,prepa-
Food, plant galactomannans in, 341 ration, 53
Frog-egg lectin, isolation and proper- Galactose,
ties, 315 D-,
Fructans in endosperm during germination,
interaction with concanavalin A, 175 357
structure, 7 lectins, 254-277
Fructopyranose, P-D-, 2,3:4,5-bis- -, 2-acetamido-2-deoxy-D-, lectins,
(phenylboronate), preparation, 45 226-254
Fructosans, review, 8 -, 6-deoxy-a-~-,1,2:3,4-bis-
Fructose, (phenylboronate), preparation of, 45
D-, complex formation with -, 6-0-methyl-D-, isolation of, 7
phenylboronic acid, 48 a-DGalactosidase
polymers, 7 in galactomannan degradation during
-, 3,4- and 4,5-dia-methyl-~-,synthe- germination of seeds, 362-366
sis, 11 multimolecular forms, 365
-, 4-0-methyl-D-, synthesis, 11 specificity, 364
420 SUBJECT INDEX, VOLUME 35

a-D-Galactoside galactohydrolase, in -, 2-deoxy-2-(rnethylamino)-~-,

galactomannan degradation, 362 biosynthesis, 107-109
Carosamine, L-, biosynthesis, 112 -, 6-O-(N,”-dimethylglycyl)-~-, synthe-
Centarnicins, structure, 110-115 sis, 54
Germination -, 1,2-O-isopropylidene-a-D-
galactomannan degradation during, of 5,6-diphenylcyclodiboronate, prepara-
seeds, 356-361 tion, 45
lectins during, 149 3,5-phenylboronate, preparation, 45
Glebomycin, biosynthesis, 102 -, 6-O-methyl-D-, preparation, 55
p-D-Glucan, structure, 7, 8, 10 Glucuronic acid, D-, methyl ethers, syn-
Glucofuranose, 3-deoxy-3-fluoro-1,2-0- thesis, 11
isopropylidene-a-D-, 5,6-phenyl- Glutarimide, 3-[2-(3,5-dimethyl-2-
boronate, preparation, 4 5 oxocyclohexyl)-2-hydroxyethyl]-,
-, 1,2-0-isopropylidene-a-~- inhibitor of galactomannan degrada-
3,5-borate, preparation, 31 tion, 361
3- and 6-chloroacetates, preparation, a-D-Glycans, crystal structure bibliogra-
54 phy, 378
3,5-phenylboronate, preparation and P-D-Glycans, crystal structure bibliogra-
use, 32 phy, 379-381
Glucomannan, biosynthesis, 353 Glycerol
Glucono-l,4-lactone, 6-O-(N,N-dimethyl- ethylboronate, structure, 4 3
glycyl)-D-, synthesis, 54 phenylboronate, nomenclature, 36
Glucopyranose, 1,6-anhydro-P-~-,2,4- properties and structure, 42
phenylboronate, preparation, 48 -, 3-O-glycopyranosyl-, mass spectrom-
Glucopyranoside, methyl a - ~ - etry of phenylboronates, 68
4,6-phenylboronate 2,3-(diphenyl- -, 1,3-O-(phenylboranediyl)-,
cyclodiboronate), hydrolysis, 52 nomenclature, 36
preparation, 47 -, 1,3-O-(phenylborylene)-,
phenylboronate, preparation, 38,46, nomenclature, 36
52 Glycine mox lectin, see Soybean lectin
-, methyl 6-deoxy-a-, 2,4-phenylboronate, Glycogen
preparation, 48 end-group assays, 8
-, methyl 2,3-di-O-benzoyl-a-~-,prepa- reaction with condanavalin A, 169-171
ration, 53 separation from guaran, 277
Glucopyranosides, interaction with structure, 5
concanavalin A, 188-190 Glycolipids, Dolichos biflorus lectin in
Glucose study of, 231
D-, Glycopeptides
1,2:3,5-bis(phenylboronate),prepa- concanavalin A-reactive, from calf
ration, 45 thymocytes, 325
complex formation with phenyl- interaction with concanavalin A, 187
boronic acid, 49 with lectins, 140
reaction with boric acid, 31 lectin-reactive, from human
-, 2-acetarnido-2-deoxy-~-, lectins, erythrocytes, 319
206-226 from tumor cells, 327-333
-, 1,8anhydro-n-, 2,4-phenylboronate, from neuronal cells, lectins in isolation
reaction with rnethacrylic anhydride, of, 326
55 pea lectin-reactive, structure, 321
-, 6-bromo-6-deoxy-D-, preparation from Glycoproteins, see also Lectins
boronate, 56 cell-surface lectin-reactive, 205,
-, 6-chloro-6-deoxy-D-,preparation from 317-333
boronate, 56 distribution and mobility, 129
 SUBJECT INDEX, VOLUME 35 42 I

Dolichos biflorus lectin in study of, Heparan sulfate, crystal structure hibli-
23 1 ography, 383
interaction with concanavalin A, Heparin, crystal structure bibliography,
177-179 382
with lectins, 140, 141 1,6-Hexanediol, cyclic phenylhoronate,
with lentil lectins, 184 preparation, 35
lectin-reactive, of erythrocyte Hexopyranose, 2,6-dideoxy4C-acetyI-~-
membrane, 318-325 xylo-, component of quinocycline B
from tumor cells, 328-333 and isoquinocycline B, 92
lentil lectin-reactive, from pig -, 2,6-dideoxy4C-(l-hydroxyethyl)-~-
lymphocyte, 325 xylo-, component of quinocycline A
from neuronal cells, lectins in and isoquinocycline A, 91
isolation, 326 Hexose, 2,3,6-trideoxy-46-(2-hydroxy-
from platelet membrane, 325 acety1)-L-threo-, see Pillarose
separation of partially sialated, 277 Hirst, Edmund Langley, obituary, 1-29
Glycopyranosides, ethylhoronates, prep- Hoinarus americanus lectins, isolation,
aration, 41 3 13
Glycosaminoglycans, crystal structure Horse gram, see Dolichos biflorus
bibliography, 38 1-385 Horseshoe-crab lectin, isolation, purifi-
Gl ycosides cation, and composition, 306
horonates, acetates and henzoates, 53 Hyaluronates, crystal structure hihliogra-
preparation, 45-48 phy, 383-385
properties, 72-76 Hyaluronic acid, and salts, crytal struc-
reaction with lectins, 140 tures of, 384,385
with phenylboronic acid, 32 Hydrolysis, of carbohydrate horonates,
Glycosid-3-uloses, preparation, 46 50-52
Glycosylation, of phenylboronates, 55
Gorse-seed extract, see Ulex europeus I1 I
Guanosine, 5’-phosphate, preparation,
54 Immunoglobulins, myeloma, interactions
Guaran with polysaccharides, 317
separation from glycogen, 276 Immunology, lectins role in, 128
structure, 349 Infrared spectroscopy, and carbohydrate
Guar gdactoniannan, structure, 349 boronate structure, 42
Guar seeds Inositols, paper chromatography,
enzymes in galactomannan degrada- phenylhoronic acid in, 60
tion during germination, 361 Inulin, structure, 7
a-D-galactosidase from, 364 Isolectins
P-D-mananase from, 367 of Bandeiraea simplicifolia I, 266
Guluronic acid, L-, of alginic acid, 8, 10 from haricot kidney-bean, 297
Gums lentil, properties, 190,192
plant, constitution, 6 pea, 197
review, 8 ofPhaseolus vulgaris, 294
sbucture, 9 Isoquinocyclines A and B, components
of, 91

H J
Helix pomatiu lectin, see Snail lectin Jack-bean lectin, see Concanavalin A
Hemagglutination tests, for lectins, 133, Jack beans, lectin isolation from, 137
140 Japanese pagoda tree, see Sophora
Hemagglutinins, see Lectins japonicu
Hemicelluloses, problems, 8, 10 Jequirity bean, see Abrus precatoriiis
 422 SUBJECT INDEX, VOLUME 35

K lectin isolation from, 137, 138

Kidney-bean extracts, effect on lympho- Lentil lectin
cyte division, 130 carbohydrate-binding specificity, 194,
204,205
L hemagglutinating activity, 190,201
interaction with cellular structures,
Laburnum alpinum lectin, 208 317
carbohydrate-binding specificity of, 305 with erythrocyte glycopeptide, 320
Landsteiner hapten-inhibition tech- isolation and properties, 138, 190-196
nique, for lectins, 139 structure, 190-192
Lectins, 127-340, see also Isolectins and Lettuce seeds, p-D-mannanase and ger-
specific lectins mination, 371
2-acetamido-2-deoxy-~-galactose- Leucoagglutinin, purification and
binding, 226-254 composition of, from Phaseolus
2-acetamido-2-deoxy-~-glucose-binding, vulgaris, 295
206-226 Levans
antibody activity, 147 interaction with concanavalin A, 175
applications in serological laboratories, plant, structure, 7
129 Lima-bean extracts, agglutinating action,
blood-group and carbohydrate-binding 128
specificities of purified, 338,339 Lima-bean lectin
carbohydrate association, 340 agglutinating activity, 243,244,247
carbohydrate-binding specificity, carbohydrate-binding specificity, 249
139-145,331-333 composition, 247
classification, 133, 141, 146 isolation, 138
definition, 128, 131 purification, 245,246
detection, 133-136 Limulus polyphemus lectin, see
L-fucose-binding, 277-291 Horseshoe-crab lectin
functions, 146-149 Lipopolysaccharides, interaction with
D-galactose-binding, 254-277 concanavalin A, 179
interaction with cellular structure, Lotus tetragonolobus lectin, see
317-333 Asparagus-pea lectin
with glycosides, 140 Lymphocytes, lectin-reactive
with polysaccharides, glycoproteins, glycoproteins from, 325
and glycolipids, 140
isolation and purification, 136-139
in life cycle of plant, 148
D-mannose(D-glucose)-binding, M
150-205 Maackia amurensis lectins, isolation,
nomenclature, 145, 146 purification, and properties, 313
physical and chemical properties, 133, Maclura pomifera lectin, see Osage-
334-337 orange lectin
reviews, 132 Macrolide antibiotics, glycosidic, 81
sources, 139 Maleylation, of lentil lectin, 196
toxicity, 149 P-D-Mannanase
uses, 129 action on galactomannans, 350
Lens culinaris, see Lentil activity, occurrence, and properties,
Lens esculenta, lentil lectin from, 193 366-371
Lentil Mannans
large-seed and small-seed, lectins interaction with concanavalin A,
from, 193 173-175
 SUBJECT INDEX, VOLUME 35 423

structure, from ivory nuts, 7 Mining, plant galactomannans in, 342

Mannitol Mitogenesis, lectin-induced, 291
D-, Mitogenic activity
3,4-ethylboronate, preparation, 41 of Phytolacca lectins, 309,311
1,2,5,6-tetrakis(diethylborinate)3,4- of Wistaria floribunda lectins, 311,
ethylboronate, preparation, 40 312
1,2:3,4:5,6-triboronate,preparation, Molecular weight
40 of asparagus-pea lectin, 284
tris(phenylboronate), hydrolysis, 50 of castor-bean lectin, 272
Mannofuranose, 2,3-O-isopropylidene-D-, of concanavalin A, 152
5,6-phenylboronate, preparation, 39 of fava-bean lectin, 203
Mannopyranoside, methyl a-D-, 2,3- and of lectins, 334, 336
4,6-phenylboronates and 2,3:4,6-bis- of Osage-orange lectin, 268
(phenylboronate), preparation, 48 of ricin, 271
methyl 6-deoxy-a-~-,2,3-phenyl- of Ulex europeus lectin, 289
boronate, preparation, 48 Monosaccharides
Mannopyranosides, alkyl and aryl, interaction with concanavalin A,
interaction with concanavalin A, 188 181-184
Mannose, lectin-reactive, classification, 141
D-, complex formation with phenyl- structure, 5,7
boronic acid, 49 Mucilages
structure, history, 5 constitution, 6
P-D-Mannosidase structure of plant, 9
in galactomannan degradation during Mucoprotein, from Phaseolus vulgaris,
germination of seeds, 362, 292
366-371 Mung-bean seedlings, enzyme from, 353
occurrence, purification, and proper- Mushroom lectin
ties, 372-375 carbohydrate-binding specificity, 145
Mannoside, methyl a-D-,phenylboronate, interaction with cellular glycopeptides,
preparation, 38 318
Mannuronic acid, with erythrocyte glycopeptide, 320
D-, of alginic acid, 10 isolation, purification, and structure,
methyl ethers, synthesis, 11 308
Mass spectrometry, of carbohydrate Mycaminose, D-, biosynthesis, 122
boronates, 41,65-70 Mycarose, L-, biosynthesis, 82,83, 88
Meadow-mushroom lectin, see Mycodextran, crystal structure bibliogra-
Mushroom lectin phy, 378
Melibiase, in galactomannan degradation
during germination of seeds,
N
362-366
Methacrylic anhydride, reaction with Neomycins, biosynthesis, 115-118
1,6-anhydro-D-glucose 2,4-phenyl- Neuronal cells, glycoproteins and
boronate, 55 glycopeptides from, isolation and
Methylation properties of, 326
of carbohydrate boronates, 42, 55 Nigeran, crystal structure bibliography,
carbon, of sugars, 83 378
in structural analysis, 5 Nomenclature
Microanalysis, of sugars, 5 of carbohydrate boronates, 36
Micro-organisms furanose and pyranose, 5
galactomannans in, 354, 376 of lectins, 145, 146
p-D-mannanase in, 366 Noviose, L-, biosynthesis, 82
424 SUBJECT INDEX, VOLUME 35

Nuclear magnetic resonance Paints, plant galactomannans in, 342

spectroscopy, of carbohydrate Pangamic acid, synthesis, 54
boronates, 41, 70 Pangamolactone, synthesis, 54
Nucleoside antibiotics, see Antibiotics Paper products, plant galactomannans in,
Nucleosides 342
arabinosyl, biosynthesis, 123 Paromamine, formation, 113, 114
boronates, acetates and benzoates of, Pea lectin, carbohydrate-binding
53 specificity, 199, 204
preparation, 38,45-48 hemagglutinating activity, 197, 201
properties, 80 interaction with erythrocyte
cobalt-containing, preparation, 56 glycopeptides, 321
column chromatography, boronic acids with glycopeptides, 141
in, 63,64 isolation, 138, 196
complex formation with boronic acids, purification, 196
49 Peanut lectin, 257-262
cytosine, isolation, 124 biological activity, 261
phenylboronates, hydrolysis, 51 carbohydrate-binding specificity,
mass spectrometry, 68 259-261
Nucleotides isolation, 138
column chromatography, boronic acids purification and properties, 258
in, 64 Pea-tree lectin, composition, purifica-
complex formation with boronic acids, tion, and properties, 313
49 Pectic substances, constitution, 6
phenylboronates, mass spectrometry, 1,5-Pentanediol, cyclic phenylboronate,
68 preparation, 35
0 Pentopyranosid3-ulose, methyl (Y-D-
Obituary, Edmund Langley Hirst, 1-29 and P-D-erythro-, preparation by
Oligo-P-D-mannosyl-(1+4)-phos- oxidation of boronate, 57
phorylase, in galactomannan degra- P e p t i d o e -phosphonogalactomannan,
dation during germination of seeds, biosynthesis, 354
362,375 Periodate oxidation
Oligosaccharides of carbohydrate phenylboronates,
interaction with asparagus-pea lectin, 57
286-288 in structural analysis of polysac-
with concanavalin A, 181-186 charides, 8
with lectins, 142 Pharmaceuticals, plant galactomannans
with wheat-germ lectin, 219 in, 341
structure, 5 Phaseolus coccineus lectin,
Optical rotatory dispersion, of sugars, 5 carbohydrate-binding specificity, 145
Osage-orange lectin Phaseolus lunatus lectin, see Lima-bean
carbohydrate-binding specificity, lectin
268-270 Phaseolus uulgaris lectin, see Red
isolation, purification, and properties, kidney-bean lectin
267 Phenol, ep-dinitro-, inhibitor of
1,3,2-Oxazaborolane, 2-phenyl-, galactomannan degradation, 361
formation, 67 Phenylboronates, stability to hydrolysis,
Oxidation, of carbohydrate boronates, 57 50
Phenylboronic acid
P complex formation with carbohydrates,
48,49
Pachyman, 0-acetyl-, crystal structure effect in paper chromatography, 59-62
bibliography, 381 reaction with glycosides, 32
 SUBJECT IND,EX, VOLUME 35 425

sulfonylated, in electrophoresis, 62 R
Phosphates, carbohydrate, mass spectro-
Raffinose, structure, 5
metry of boronates of, 69
Phosphorylation, of nucleoside
Rana catesbiana lectin, isolation and
boronates, 54 properties, 315
Red kidney -)bean, phytohemagglutinin,
Phytohemagglutinin, *seealso Lectins
from red kidney-beans, 291, 293 291,293
Phytolacca americanum, see Poke-weed Red kidney-bean lectin
carbohydrate-binding specificity, 145,
Phytolacca esculenta lectin, isolation
297-302
and properties, 310
Pillaromycin A, structure, 96 composition, 292
hemagglutinating and mitogenic activ-
Pillarose, structure, 96
ity, 291, 292
Pisum satious lectin, see Pea lectin
Plant life-cycle, lectins in, 148 interaction with cellular glycopep-
Plasters, plant galactomannans in, 342 tides, 318-320
isolation, 138,296,297
Platelet membrane, glycoproteins from,
325 purification, 292
structure, 267
Poke-weed extract, mitogenic activity,
Ribitol, 1,5-dideoxy-, phenylboronate,
309,310
structure, 43
Poke-weed lectin, isolation and proper-
Ribofuranose, p-D-, 1,5:2,3-
ties, 309
Polyagglutinability, in serology, 257 bis(phenylboronate), preparation and
structure, 44
Polyoxins, biosynthesis, 125, 126
Ribonucleosides, reaction with isobutyl
Pol ysaccharides
diphenylborinate, 39
amino, crystal structure bibliography,
381-385 Ribopyrdnosylamine, N-( p-bromo-
pheny1)-a-D-, 2,4-phenylboronate,
conformational analysis, 8
preparation and structure, 46
crystal structure bibliography, 377-385
Ribose, CY-D-,2,4-phenylboronate, prepara-
interaction with castor-bean lectin, 274
tion and structure, 44
with concanavalin A, 166-169, 179
Ricin
with lectins, 140
carbohydrate-binding specificity, 274
with lentil lectins, 1%
immunization, 129
with myeloma immunoglobulins, 317
isolation, 270
plant, structure, 10
structure, 5, 7 purification, 138, 271
toxicity, 271
Polysphondylium pallidum lectin,
Ricinus communis lectin, see Castor-bean
isolation and properties, 309
lectin; Ricin
Potato lectin
amino acids, 211 Robinia pseudoaccacia lectin
robin, interaction with erythrocyte
carbohydrate-binding specificity, 212
glycopeptide, 320
hemagglutinating activity, 210
isolation and properties, 311
isolation, 138
purification, 210
Prolectin, isolation, 149, 229 S
Proteins, carbohydrate-binding, of plants Scarlet runner-bean lectin, carbohydrate-
and animals, see Lectins binding specificity, 145
Psicofuranine, biosynthesis, 123 Serology
Dolichos bijlorus lectin in, 226
Helix pomatia lectin in, 226, 239, 241
Q lectins in, 129
Quinocyclines A and B, biosynthesis, polyagglutinability in, 257
91-96 Ulex europeus seed extract in, 224,289
426 SUBJECT INDEX, VOLUME 35

Sheep’s rumen, carbohydrates, 11 end-group assays, 8

Sialoglycopeptide, from tumor cells, 327 structure, 10
Sialoglycoprotein, properties of, of eryth- Stereochemistry, of sugars, 5
rocyte membrane, 318, 319 Streptarnine, biosynthesis, 117
Smith degradation, of carbohydrates, 9 -, deoxy-, biosynthesis, 104, 117
Snail lectin Streptidine
carbohydrate-binding specificity, 242 biosynthesis, 102-107, 109
composition, 240 6-phosphate, biosynthesis, 103-106,
hemagglutinating activity, 239 109
isolation, 138 Streptomutin A, production and
purification and properties, 240 structure, 106
Sodiocellulose 11, crystal structure bibli- Streptomyces 2755 lectin, isolation, puri-
ography, 379 fication, and composition, 307
Sodium hyaluronate, crystal structure Streptomycin
bibliography, 383-385 biosynthesis and structure, 102-110
Solanum tuberosum lectin, see Potato 6-phosphate, biosynthesis, 109
lectin -, dihydro-, biosynthesis, 102
Sophora japonica lectin Streptose, L-, biosynthesis, 98-102
agglutinating activity, 250-254 -, dihydro-L-, biosynthesis, 98-102
carbohydrate-binding specificity, 251 Sucrose
isolation, 138 in endosperm and cotyledon during
purification, 250 germination, 357
Sophorose structure, 5
interaction with concanavalin A, Sugars
184- 186 C-acetyl-branched, 92,94, 95
structure, 186 anhydro, boronates, properties, 79
Soybean lectin paper chromatography of,
amino acids, 234 phenylboronic acid in, 60
biological activity, 238 biosynthesis of, of antibiotic sub-
biophysical characteristics, 233,234 stances, 81-126
carbohydrate-binding specificity, boronates, see Boronates
236-238 branched-chain, antibiotic compo-
hemagglutinating activity, 23 1 nents, 82-102
interaction with cellular structures, 317 column chromatography, boronic acids
isolation, 138 in, 64
purification, 232,233 electrophoresis of, sulfonylated
toxicity, 232 phenylboronic acids in, 62
Spectinomycin, biosynthesis and formyl- or hydroxymethyl-branched, 98
structure, 118-121 gas-liquid chromatography of, boronic
-, dihydro-, formation, 119, 121 acids in, 65
Spectrophotometric assay, of lectins, 134 C-methyl-branched, biosynthesis, 97
Spindle-tree lectin, carbohydrate- paper chromatography of,
binding specificity, 145 phenylboronic acid in, 59, 61
Sponge Iectins, isolation and properties, structure, 5
315,316 Sunn-hemp lectin, isolation, purification,
Stachyose and properties, 306
biosynthesis in fenugreek seeds, 352
in endosperm during germination, 357
Starch
T
in endosperm and cotyledon during Teichoic acids, interaction with
germination, 356-361 concanavalin A, 175-177
 SUBJECT INDEX, VOLUME 35 427

Tetritol, 3-deoxy-D~-glycero- carbohydrate-binding specificity, 145,

ethylboronate, structure, 43 303
phenylboronate, struchire, 43 isolation, 302
cis-3,4-Thiolanediol 1-oxide, purification and composition of, 303
phenylboronates, separation of Vinelose, L-, biosynthesis, 85-89
stereoisomers, 58
p-Toluenesulfonylation, of carbohydrate W
boronates, 53
Wax-bean lectin, purification and
p-Tolylboronates, stability to hydrolysis,
composition, 296
50
Well-drilling, plant galactomannans in,
Tridncna marima lectin, purification and
342
properties, 314
Wheat-germ lectin
Tridacnin, purification and properties,
agglutinating activity, 214
314
carbohydrate-binding sites, 216, 222
1,3,5,2,4-Trioxadiborepane,
carbohydrate-binding specificity, 220
nomenclature, 36
Trisaccharides interaction with cellular structures,
interaction with lectins, 142 317
structure, 5 isolation, 137, 214
mitogenic activity, 224
Triticum vulgaris lectin, see Wheat-germ
lectin precipitation, 216, 224
Tritylation, of nucleoside boronates, 55 purification, and properties, 214
Tumor cells structure, 215
lectin-binding, 205 Wistaria Joribunda lectin
lectin-reactive, 327-333 isolation, 138,312
Tylosin, biosynthesis, 83 purification and properties, 312

U X

Ulex europeus I lectin X-ray crystallography, of sugars, 5

carbohydrate-binding specificity, 290 X-ray diffraction, and carbohydrate boro-
hemagglutinating activity, 289 nate structure, 41
isolation and purification, 289, 290 Xylan, structure, 7
Ulex europeua 11, extract, carbohydrate- Xylitol
binding specificity, 224-226 2-diethylborinate 1,3:4,5-bis(ethyl-
Ultraviolet absorption spectra, of sugars, boronate), hydrolysis, 52
5 -, 1,5-dideoxy-, phenylboronate,
Uridine, 5’-phosphate, preparation, 54 structure, 43
Xylofuranose, l,Z-O-isopropylidene-a-D-,
V 3,5-phenylboronate, preparation, 38
Xylofuranoside
Validamycins, biosynthesis and methyl, isolation of anomeric
structure, 120-122 phenylboronates, 58
Validoxylamine, structure, 120-122 -, methyl p-D, 3,5-phenylboronate,
Verbascose, in endospem during gernii- hydrolysis, 51, 52
nation, 357 Xylopyranoside, methyl CY-D-
Vicia crucca lectin, isolation and proper- 2,4-phenylboronate, oxidation, 57
ties, 138,304 preparation, 46
Viciu eruilia lectin, isolation, separation from isomers, 58
composition, and properties, 311 -, methyl p-D-
Vicia faho lectin, see Faba-bean lectin 2,4-phenylboronate, oxidation, 57
Vicia graminen lectin preparation, 37
428 SUBJECT INDEX, VOLUME 35

Xylopyranoside-5-W, methyl (Y-D-and (Y-D-, 1,2:3,5bis(phenylboronate)and

p-D-, preparation, 58 1,2:3,5-bis(butylboronate),prepara-
Xylose, tion, 43
D- -, 3-0-~-D-glucopyranosy~-D-, synthesis,
boronates, hydrolysis, 51 55
structure, 4 -, 3-0-a- and -P-D-xylopyranosyl-D-,
sulfur-containing,8 synthesis, 55
 CUMULATIVE AUTHOR INDEX FOR VOLS. 31-35*
B G
BALLOU,CLINTONE., and BARKER, GELPI,MARIAE., and CADENAS, RAUL A.,
HORACEA., [Obituary ofl WilliamZev The Reaction of Ammonia with Acyl
Hassid, 32, 1-14 Esters of Carbohydrates, 31,81-134
BARKER,HORACEA. See Ballou, Clinton E. GLAUDEMANS, CORNELISP. J., [Obituary
BARNETT,JAMESA., The Utilization of ofl Hewitt Grenville Fletcher, Jr.,
Sugars by Yeasts, 32, 125-234 31, 1-7
BUSHWAY, ALFREDA. See Whistler, Roy L. GLAUDEMANS, CORNELISP. J., The
Interaction of Homogeneous, Murine
Myeloma Immunoglobulins with
C Polysaccharide Antigens, 31,313346
CADENAS, RAUL A. See Gelpi, Maria E. GOLDSTEIN,IRWINJ., and HAYES,
CERNY,MILOSLAV, and S T A N ~ K
JAN,
, COLLEENE., The Lectins: Carbo-
JR., 1,6-Anhydro Derivatives of Aldo- hydrate-binding Proteins of Plants
hexoses, 34,23-177 and Animals, 35, 127-340
CHEN, MINSHEN,and WHISTLER,ROY L., GRISEBACH, HANS,Biosynthesis of Sugar
Metabolism of D-Fructose, 34, Components of Antibiotic Sub-
285-343 stances, 35,81-126

H
D
HAINES,ALANH., Relative Reactivities
DAX,KARL, and WEIDMANN, HANS,Reac- of Hydroxyl Groups in Carbohydrates,
tions of ~-Glucofuranurono-6,3- 33,ll-109
Iactone, 33,189-234 HANESSIAN,STEPHEN,and PERNET,
DEA, IAINC. M., and MORRISON, ANDRE G., Synthesis of Naturally
ANTHONY,Chemistry and Interac- Occurring C-Nucleosides, Their
tions of Seed Galactomannans, 31, Analogs, and Functionalized C-GIyco-
241312 syl Precursors, 33, 111-188
DE BELDER,ANTHONY N., Cyclic Acetals HAYES,COLLEENE. See Goldstein,
of the Aldoses and Aldosides: High- Irwin J.
lights of the Literature Since 1964, HORTON,DEREK.See Wander, Joseph D.
and a Supplement to the Tables, 34,
179-241 I
DEKKER,ROBERTF. H., and RICHARDS,
GEOFFREYN., Hemicellulases: Their IGARASHIYKrKvO, The Koenigs-Knorr
Occurrence, Purification, Properties, Reaction, 34,243-283
and Mode of Action, 32,277452
DEY, P%UCASHM., Biochemistry of Plant J
Galactomannans, 35, 3 4 1 3 7 6 JEFFREY,GEORGEA., and SUNDARA-
LINGAM, MUTTAIYA,Bibliography of
Crystal Structures of' Carbohydrates,
F Nucleosides, and Nucleotides (1973),
FERRIER, ROBERTJ., Carbohydrate Boro- 31,347371; (1974), 32,353384;
nates, 3 5 , 3 1 4 0 (1975),3 4 , 3 4 5 3 7 8

* Starting with Volume 30, a Cumulative Author Index covering the previous 5
volumes will be published in every 5th volume. That listing the authors of chapters
in Volumes 1-29 may b e found in Volume 29.

429
430 CUMULATIVE AUTHOR INDEX FOR VOLS. 31-35

K R
KHAN,RIAZ,The Chemistry of Sucrose, REXOVA-BENKOVA, ~ U B O M ~and
R A MARK-
,
33,235-294 V I ~ OSKAR,
, Pectic Enzymes, 33,
323-385
RICHARDS, GEOFFREYN. See Dekker,
L Robert F. H.
LAM, OLLE, and LINDBERG, BENGT,The
Pneumococcal Polysaccharides: A S
Re-examination, 33,295-322 SINGH,PREMP. See Whistler, Roy L.
LINDBERG,BENGT,LONNGREN, JORGEN, STACEY,MAURICE,and MANNERS,
and SVENSSON, SIGFFUD,Specific DAVIDJ., [Obituary ofl Edmund
Degradation of Polysaccharides, 31, Langley Hirst, 35,"1-29
185-240 STANEK,JAN,JR. See Cernf, Miloslav.
LINDBERG,BENGT.See also, Larm, Olle. SUNDARALINGAM, MUTTAIYA. See Jeffrey,
LONNGREN, JORGEN.See Lindberg, George A.
Bengt. SUNDARARAJAN, PUDUPADIR., and
MARCHESSAULT, ROBERTH., Bib-
M liography of Crystal Structures of
Polysaccharides (1975), 35,377-385
MANNERS,DAVIDJOHN.See Stacey, SUNDARARAJAN, PUDUPADIR. See also,
Maurice. Marchessault, Robert H.
MARCHESSAULT, ROBERTH., and SUN- SVENSSON, SIGFRID. See Lindberg, Bengt.
D A R A R A JPUDUPADI
AN, R., Bibli-
ography of Crystal Structures of Poly- T
saccharides (1967-1974), 33,387404
MARCHESSAULT, ROBERTH. See also, TOKUZEN,REIKO. See Whistler, Roy L.
Sundararajan, Pudupadi R.
MARKOVIC, OSKAR.See Rexovi-Benkova, W
iubomira. WANDER,JOSEPHD., and HORTON,
MOHNSON,ANTHONY. See Dea, Iain DEREK,Dithioacetals of Sugars, 32,
C. M. 15-123
WATSON,RONALDR., and ORENSTEIN,
NEIL S., Chemistry and Biochemistry
N
of Apiose, 31, 135-184
NAKAHARA, WARO.See Whistler, Roy L. WEIDMA", HANS.See Dax, Karl.
NEUBERGER, ALBERT,[Obituary ofl WEIGEL,HELMUT,[Obituary ofl Edward
Alfred Gottschalk, 33, 1-9 John Bourne, 34,l-22
WHISTLER,ROYL., BUSHWAY, ALFRED
A., SINGH,PREM P., NAKAHARA,
0
WARO,and TOKUZEN,REIKO,Non-
ORENSTEIN,NEIL S. See Watson, cytotoxic, Antitumor Polysaccharides,
Ronald R. 32,235-275
WHISTLER,ROYL. See also, Chen,
Minshen.
P WILLIAMS,J. MICHAEL,Deamination of
PERNET,AND& G. See Hanessian, Carbohydrate Amines and Related
Stephen. Compounds, 31,9-79
 CUMULATIVE SUBJECT INDEX FOR VOLS. 31-35*
A Biosynthesis,
of sugar components of antibiotic sub-
Acyl esters,
stances, 35,81-126
of carbohydrates, reaction of, with
Boronates,
ammonia, 31,81-134
of carbohydrates, 35,3140
Aldohexoses,
Bourne, Edward John,
1,6-anhydro derivatives of, 34,23-177
obituary of, 34,l-22
Aldoses and aldosides,
cyclic acetals of, 34, 179-241
Ammonia, C
the reaction of, with acyl esters of carbo-
Carbohydrate-binding proteins
hydrates, 31, 81-134
(the lectins),
1,6-Anhydro derivatives,
of plants and animals, 35, 127-340
of aldohexoses, 34,23-177
Carbohydrates. See also,
Animals,
Polysaccharides, Sugars.
carbohydrate-binding proteins of, 35,
acyl esters of, reaction with ammonia,
127-340
31,81-134
Antibiotic substances,
amines of, and related compounds,
biosynthesis of sugar components of,
deamination of, 31,9-79
35,81-126
bibliography of crystal structures of,
Antigens,
(1973),31,347471
polysaccharide, interaction of, with
(1974),32,353484
homogeneous, murine myeloma
(1975),34,345378
immunoglobulins, 31,313-346
boronates of, 35,3140
Antitumor polysaccharides,
relative reactivities of hydroxyl groups
noncytotoxic, 32,235-275
in, 33, 11-109
Apiose,
Chemistry,
chemistry and biochemistry of, 31,
of apiose, 31, 135-184
135-184
of seed galactomannans, 31,241-312
of sucrose, 33,235-294
B Crystal structures,
Bibliography, bibliography of,
of crystal structures of carbohydrates, of carbohydrates, nucleosides, and
nucleosides, and nucleotides, nucleotides,
e
(1973),31,347-371 (1973), 31,347471
(1974),32,353-384 (1974),32,353-384
(1975),34,345-378 (1975).34,345478
of crystal structures of polysaccharides, of polysaccharides,
(1967-1974), 33,387404 (1967-1974), 33,387404
(1975), 35,377485 (1975),35,377-385
Biochemistry, Cyclic acetals,
of apiose, 31, 135-184 of the aldoses and aldosides, 34,
of plant galactomannans, 35,341476 179-241

* Starting with Volume 30, a Cumulative Subject Index covering the previous 5
volumes will be published in every 5th volume. That listing the chapters in Volumes
1-29 may be found in Volume 29.

431
432 CUMULATIVE SUBJECT INDEX FOR VOLS. 31-35

D Interactions,
of homogeneous, murine myeloma
Deamination immunoglobulins, with poly-
of carbohydrate amines and related saccharide antigens, 31,313346
compounds, 31,9-79 of seed galactomannans, 31,241-312
Degradation,
specific, of polysaccharides, 31,
K
185-240
Dithioacetals, Koenigs-Knorr reaction, the,
of sugars, 32, 15-123 34,243-283

E 1
Enzymes. See also, Hemicellulases. Lectins, 35,127-340
pectic, 33,323385
M
F Metabolism,
Fletcher, Hewitt Grenville, Jr., Of D-fructose, 34,285343
obituary of, 31,l-7 Mode of action,
D-Fructose, of hemicellulases, 32,277-352
metabolism of, 34,285443
N
G C-Nucleosides,
naturally occurring, and their analogs,
Galactomannans, and functionalized C-glycosyl pre-
of plants, biochemistry of, 35,341-376 cursors, synthesis of, 33, 111-188
of seeds, chemistry and interactions of, Nucleosides and nucleotides,
31,241-312 bibliography of crystal structures of,
D-Glucofuranurono-6,3-lactone, (1973),31,347371
reactions of, 33, 189-234 (1974),32,353384
Gottschalk, Alfred, (1975),34,345-378
obituary of, 33, 1-9
0
H
Obituary,
Hassid, William Zev, of Edward John Bourne, 34, 1-22
obituary of, 32, 1-14 of Hewitt Grenville Fletcher, Jr., 31,
Hdnicellulases, 1-7
mode of action, occurrence, properties, of Alfred Gottschalk, 33, 1-9
and purification, 32,277-352 of William Zev Hassid, 32, 1-14
Hirst, Edmund Langley, of Edmund Langley Hirst, 35, 1-29
obituary of, 35, 1-29 Occurrence,
Hydroxyl groups, of hemicellulases, 32,277352
relative reactivities of, in carbohydrates,
33,ll-109 P
Pectic enzymes, 33,323485
I
Plants,
Immunoglobulins, carbohydrate-binding proteins of, 35,
homogeneous, murine myeloma, the 127340
interaction of, with polysaccharide galactomannans of, biochemistry of, 35,
antigens, 31,313346 341-376
 CUMULATIVE SUBJECT INDEX FOR VOLS. 31-35 433

Pneumococcal polysaccharides, S
a re-examination, 33,295-322
Striictures, crystal,
Polysaccharides. See also,
of carbohydrates, nucleosides, and
Carbohydrates, Galactomannans.
nucleotides,
antigens, interaction of, with horno-
(1973), 3 1 , 3 4 7 4 7 1
geneous, murine myeloma immu-
(1974), 32,353484
noglobulins, 31,313-346
(1975), 34,345-378
bibliography of crystal structures of,
of pol ysaccharides,
(1967-1974), 3 3 , 3 8 7 4 0 4
(1967-1974), 3 3 , 3 8 7 4 0 4
(1975), 35,377-385
(1975), 35,377485
noncytotoxic, antitumor, 32,235-275
Sucrose,
the pneumococcal, a re-examination of,
the chemistry of, 33,235-294
33,295-322 Sugar components,
specific degradation of, 31, 185-240
of antibiotic substances, biosynthesis of,
Properties
35,81-126
of hemicellulases, 3 2 , 2 7 7 4 5 2
Sugars,
Proteins,
dithioacetals of, 32, 15-123
carbohydrate-binding, of plants and
the utilization of, by yeasts, 32,
animals, 35, 127-340
125-234
Purification,
Synthesis,
of hemicellulases, 32, 277-352
of naturally occurring C-nucleosides,
their analogs, and functionalized
R C-glycosyl precursors, 33, 111-188
Reaction,
of ammonia with acyl esters of carbo-
hydrates, 31,81-134
U
the Koenigs-Knorr, 34, 243-283 Utilization,
Reactions, of sugars by yeasts, 32, 125-234
of D-g~ucofuranurono-6,3-~actone, 33,
189-234
Reactivities,
Y
relative, of hydroxyl groups in carbo- Yeasts,
hydrates, 33, 11-109 utilization of sugars by, 32, 125-234
 ERRATUM
VOLUME 34
Page 86. The second line of‘ the second paragraph should read: “1,6-anhydrohexo-
pyranoses renders replacement of equatorial sul-”

A
B
c a
0 9
E O
F 1
6 2
H 3
1 4
434 J 5