Visit https://ebookgate.

com to download the full version and

explore more ebooks

Genomics and Pharmacogenomics in Anticancer Drug

Development and Clinical Response 1st Edition
Natalie Stickle Msc

_____ Click the link below to download _____

https://ebookgate.com/product/genomics-and-
pharmacogenomics-in-anticancer-drug-development-and-
clinical-response-1st-edition-natalie-stickle-msc/

Explore and download more ebooks at ebookgate.com

Here are some recommended products that might interest you.
You can download now and explore!

Anticancer Drug Development 1st Edition Bruce C. Baguley

(Author)

https://ebookgate.com/product/anticancer-drug-development-1st-edition-
bruce-c-baguley-author/

ebookgate.com

Pharmacogenomics in Drug Discovery and Development From

Bench to Bedside 1st Edition Qing Yan (Auth.)

https://ebookgate.com/product/pharmacogenomics-in-drug-discovery-and-
development-from-bench-to-bedside-1st-edition-qing-yan-auth/

ebookgate.com

Novel Anticancer Drug Protocols 1st Edition John K.

Buolamwini

https://ebookgate.com/product/novel-anticancer-drug-protocols-1st-
edition-john-k-buolamwini/

ebookgate.com

Pharmacogenomics Social Ethical and Clinical Dimensions

1st Edition Mark A. Rothstein

https://ebookgate.com/product/pharmacogenomics-social-ethical-and-
clinical-dimensions-1st-edition-mark-a-rothstein/

ebookgate.com
Drug Metabolism in Drug Design and Development 1st Edition
Donglu Zhang

https://ebookgate.com/product/drug-metabolism-in-drug-design-and-
development-1st-edition-donglu-zhang/

ebookgate.com

Antibody Drug Conjugates Fundamentals Drug Development and

Clinical Outcomes to Target Cancer 1st Edition Kenneth J.
Olivier Jr.
https://ebookgate.com/product/antibody-drug-conjugates-fundamentals-
drug-development-and-clinical-outcomes-to-target-cancer-1st-edition-
kenneth-j-olivier-jr/
ebookgate.com

Bone Metastasis Experimental and Clinical Therapeutics

Cancer Drug Discovery and Development 1st Edition Gurmit
Singh
https://ebookgate.com/product/bone-metastasis-experimental-and-
clinical-therapeutics-cancer-drug-discovery-and-development-1st-
edition-gurmit-singh/
ebookgate.com

Rethink Growth and Learning through Coaching and

Organisational Development 1st Edition Natalie Cunningham

https://ebookgate.com/product/rethink-growth-and-learning-through-
coaching-and-organisational-development-1st-edition-natalie-
cunningham/
ebookgate.com

Novel Anticancer Agents Strategies for Discovery and

Clinical Testing 1st Edition Alex A. Adjei

https://ebookgate.com/product/novel-anticancer-agents-strategies-for-
discovery-and-clinical-testing-1st-edition-alex-a-adjei/

ebookgate.com
Genomics and Pharmacogenomics
in Anticancer Drug Development
and Clinical Response
 Cancer Drug Discovery
and Development™
Beverly A. Teicher, Series Editor
Genomics and Pharmacogenomics in Cytokines in the Genesis and Treatment of
Anticancer Drug Development and Clinical Cancer, edited by Michael A. Caligiuri,
Response, edited by Federico Innocenti, Michael T. Lotze, and Frances R. Balkwill,
2008 2007
Checkpoint Responses in Cancer Therapy, Regional Cancer Therapy, edited by Peter M.
edited by Wei Dai, 2008 Schlag and Ulrike Stein, 2007
Cancer Proteomics: From Bench to Bedside, Gene Therapy for Cancer, edited by Kelly K.
edited by Sayed S. Daoud, 2008 Hunt, Stephan A. Vorburger, and Stephen G.
Antiangiogenic Agents in Cancer Therapy, Swisher, 2007
Second Edition, edited by Beverly A. Teicher Deoxynucleoside Analogs in Cancer Therapy,
and Lee M. Ellis, 2007 edited by Godefridus J. Peters, 2006
Apoptosis and Senescence in Cancer Cancer Drug Resistance, edited by Beverly A.
Chemotherapy and Radiotherapy, Second Teicher, 2006
Edition, edited by David A. Gerwitz, Shawn Histone Deacetylases: Transcriptional
Edan Holtz, and Steven Grant, 2007 Regulation and Other Cellular Functions,
Molecular Targeting in Oncology, edited by edited by Eric Verdin, 2006
Howard L. Kaufman, Scott Wadler, and Immunotherapy of Cancer, edited by Mary L.
Karen Antman, 2007 Disis, 2006
In Vivo Imaging of Cancer Therapy, edited by Biomarkers in Breast Cancer: Molecular
Anthony F. Shields and Patricia Price, Diagnostics for Predicting and Monitoring
2007 Therapeutic Effect, edited by Giampietro
Transforming Growth Factor- in Cancer Gasparini and Daniel F. Hayes, 2006
Therapy, Volume II: Cancer Treatment and Protein Tyrosine Kinases: From Inhibitors to
Therapy, edited by Sonia Jakowlew, Useful Drugs, edited by Doriana Fabbro and
2008 Frank McCormick, 2005
Transforming Growth Factor- in Cancer Bone Metastasis: Experimental and Clinical
Therapy, Volume 1: Basic and Clinical Therapeutics, edited by Gurmit Singh and
Biology, edited by Sonia Jakowlew, 2008 Shafaat A. Rabbani, 2005
Microtubule Targets in Cancer Therapy, edited The Oncogenomics Handbook, edited by William
by Antonio T. Fojo, 2007 J. LaRochelle and Richard A. Shimkets, 2005
Genomics and
Pharmacogenomics
in Anticancer
Drug Development
and Clinical
Response

Edited by

Federico Innocenti,
MD, PhD
University of Chicago,
Chicago, IL, USA
Editor
Federico Innocenti, MD PhD
University of Chicago
Department of Medicine
Section of Hematology/Oncology
5841 South Maryland Avenue
Chicago IL 60637-1470
USA
[email protected]

Series Editor
Beverly A. Teicher, PhD
Genzyme Corporation
Framingham, MA

ISBN: 978-1-58829-646-7 e-ISBN: 978-1-60327-088-5

Library of Congress Control Number: 2008938265

©Humana Press, a part of Springer Science+Business Media, LLC 2008

All rights reserved. This work may not be translated or copied in whole or in part without the written permission of
the publisher (Humana Press, 999 Riverview Drive, Suite 208, Totowa, NJ 07512 USA), except for brief excerpts in
connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed
is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified
as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of going to press,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that
may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

987654321

springer.com
P REFACE
Genomics and Pharmacogenomics in Anticancer Drug Development and Clinical
Response provides the most comprehensive body of knowledge available on the role
of genetic and genomic variation in the individualization of drug therapies in cancer
patients. As a consequence of the intrinsic chromosomal and genetic instability of the
tumor genome, it is generally believed that tailoring of chemotherapy in cancer pa-
tients might be achieved by molecular analysis of patient tumor DNA. In addition,
to reduce the toxicity risk of patients, the tumor DNA information should be inte-
grated with the available data on polymorphic drug-metabolizing enzyme and trans-
porter genes mediating the exposure of patients to active drugs and/or their active
metabolites. The chapters of this book clearly show how DNA information from both
the host (germline) and the tumor should be taken into account for rational selection
of drug therapies in cancer patients, an aspect that received little attention, despite its
importance.
The availability of new molecular approaches to the selection of drug therapy is an
emerging need, because the traditional approach based on the evaluation of patient and
tumor characteristics is clearly far from optimal. Many treated patients do not experience
significant benefits from the treatment, while they often experience moderate to severe
toxicities. In addition, the development and clinical use of novel molecularly targeted
agents (alone or in combination with classical cytotoxic therapy) requires the under-
standing of the molecular features of the tumors and the identification of tumor markers
of response.
In this book, the readers will find a series of chapters addressing the role of ge-
nomic information in cancer therapy and in drug development. Several books on phar-
macogenomics are currently available, but this book represents a unique source, as it
describes experimental approaches, statistical strategies, and clinical examples of the
application of genomic medicine in oncology. Many outstanding scientists in the field
of cancer pharmacogenomics have been invited to contribute, and I am grateful to
have had the opportunity to work with them, learning a great deal from reading their
chapters.
I have approached this book from both a basic and an applied perspective. Among
three different sections, six chapters in the first section are focused on up-to-date ge-
nomic experimental approaches in oncology, including genome-wide phenotyping (mi-
croarray and proteomics) and genotyping methods, as well as novel cell-based models
used for the identification of genetic markers of drug response.
The second section shows how genetic and genomic information is currently applied
to treatment individualization and optimization: Eleven chapters describe some of the
most elegant examples of genetic and genomic markers that are predictive of the survival
and toxicity risk of cancer patients.
Finally, in the third section, readers will find four chapters that address the role of
pharmacogenomics in drug development in oncology, including an industry perspective
on this subject, as well as statistical aspects related to the discovery of pharmacogenomic
biomarkers during drug development.

v
vi Preface

Because the discovery of genetic markers of response is of high relevance in on-

cology, we perceived the need that a collection of multidisciplinary topics be gathered
together to discuss this important aspect of pharmacogenomics applied to cancer
patients.
I believe that this book represents the first reference for researchers in the field of
cancer pharmacogenomics and clinicians, from both the academia and industry.

Federico Innocenti
C ONTENTS

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Part I Genomic Experimental Approaches in Oncology

1 Toward the Realization of the Promise of Microarrays in Oncology . . . . . . . . . 3

Natalie Stickle and Neil Winegarden

2 Cell-Based Models to Identify Genetic Variants Contributing

to Anticancer Drug Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
M. Eileen Dolan and Howard McLeod

3 Proteomic Analysis in Cancer Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Yasuhiro Kuramitsu and Kazuyuki Nakamura

4 MicroRNAs and Discovery of New Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Soken Tsuchiya, Yasushi Okuno, and Gozoh Tsujimoto

5 Pharmacogenomics of the National Cancer Institute’s

60-Tumor Cell Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Anders Wallqvist, Ruili Huang, and David G. Covell

6 Use of Single-Nucleotide Polymorphism Array for Tumor

Aberrations in Gene Copy Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Kwong-Kwok Wong

Part II Pharmacogenomics of Toxicity and Response Of Chemotherapy

7 Concordance Between Tumor and Germline DNA . . . . . . . . . . . . . . . . . . . . . . . 91

Sharon Marsh

vii
viii Contents

8 Epidermal Growth Factor Receptor Mutations and Sensitivity to

Selective Kinase Inhibitors in Human Lung Cancer . . . . . . . . . . . . . . . . . . . 103
Anurag Singh, Sreenath V. Sharma, and Jeffrey Settleman

9 BCR-ABL Mutations and Imatinib Resistance in Chronic Myeloid

Leukemia Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Mark R. Litzow

10 Role of Thymidylate Synthase Gene Variations in Colorectal Cancer

Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Georg Lurje and Heinz-Josef Lenz

11 Thiopurines in the Treatment of Childhood Acute Lymphoblastic

Leukemia and Genetic Variants of the Thiopurine
S-Methyltransferase Gene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Martin Stanulla, Elke Schaeffeler and Matthias Schwab

12 Impact of Polymorphisms on the Clinical Outcomes of Monoclonal

Antibody Therapy Against Hematologic Malignancies . . . . . . . . . . . . . . . . 203
Dong Hwan Kim

13 DNA Repair and Mitotic Checkpoint Genes as Potential Predictors

of Chemotherapy Response in Non-Small-Cell Lung Cancer . . . . . . . . . . . 231
Rafael Rosell, Miquel Taron, Mariacarmela Santarpia, Fernanda
Salazar, Jose Luis Ramirez, and Miguel Angel Molina

14 Dihydropyrimidine Dehydrogenase (Dpyd) Gene

Polymorphism: Portrait of a Serial Killer . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Joseph Ciccolini, Cédric Mercier, and Gérard Milano

15 Impact of UDP-Glucuronosyltransferase 1A Haplotypes on

Irinotecan Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Kimie Sai, Hironobu Minami, Yoshiro Saito, and Jun-ichi Sawada

16 Microarray Profiling in Breast Cancer Patients . . . . . . . . . . . . . . . . . . . . . . . . . 287

Yong Qian, Xianglin Shi, Vincent Castranova, and Nancy L. Guo

17 Role of the Folate-Pathway and the Thymidylate Synthase Genes in

Pediatric Acute Lymphoblastic Leukemia Treatment Response . . . . . . . . . 299
Lea Cunningham and Richard Aplenc
Contents ix

Part III Pharmacogenomics in Clinical Drug Development in Oncology

18 Pharmacogenomics in Drug Development: A Pharmaceutical

Industry Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Tal Zaks

19 Identification of Pharmacogenomic Biomarker Classifiers in Cancer

Drug Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Richard Simon

20 Toxicogenomics Application to Oncology Drug Development . . . . . . . . . . . . 339

Luigi Calzolai and Teresa Lettieri

21 Strategies to Identify Pharmacogenomic Biomarkers:

Candidate Gene, Pathway-Based, and Genome-Wide Approaches . . . . . . 353
Xifeng Wu, Jian Gu, and Margaret R. Spitz

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
C ONTRIBUTORS
R ICHARD A PLENC , MD, MSCE • Children’s Hospital of Philadelphia, Philadelphia,
Pennsylvania, USA
L UIGI C ALZOLAI , P H D • Medway School of Pharmacy, University of Kent, Kent, UK
V INCENT C ASTRANOVA , P H D • The Pathology and Physiology Research Branch,
Health Effects Laboratory Division, National Institute of Occupational Safety and
Health, Morgantown, West Virginia, USA
J OSEPH C ICCOLINI , P HARM D, P H D • Pharmacokinetics—Medical Oncology,
Université de la Méditerranée, Marseille, France
DAVID G. C OVELL , P H D • Developmental Therapeutics Program, National Cancer
Institute, Frederick, Maryland, USA
L EA C UNNINGHAM , MD • Department of Pediatrics, Children’s Hospital of
Philadelphia, Philadelphia, Pennsylvania, USA
M. E ILEEN D OLAN , P H D • Section of Hematology/Oncology, University of Chicago,
Chicago, Illinois, USA
J IAN G U , P H D • Department of Epidemiology, The University of Texas, M.D. Anderson
Cancer Center, Houston, Texas, USA
NANCY L. G UO , P H D • Mary Babb Randolph Cancer Center, Department of
Community Medicine, West Virginia University, Morgantown, West Virginia, USA
RUILI H UANG , P H D • Developmental Therapeutics Program, National Cancer
Institute, Frederick, Maryland, USA
D ONG H WAN K IM , MD, P H D • Department of Hematology/Oncology, Samsung
Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea
F EDERICO I NNOCENTI , MD, P H D • Department of Medicine, Section of Hematology/
Oncology, University of Chicago, Chicago, Illinois, USA
YASUHIRO K URAMITSU , MD, P H D • Department of Biochemistry and Functional
Proteomics, Yamaguchi University, Graduate School of Medicine, Yamaguchi, Japan
H EINZ -J OSEF L ENZ , MD, FACP • Division of Medical Oncology, University of
Southern California, Norris Comprehensive Cancer Center, Keck School of Medicine,
Los Angeles, California, USA
T ERESA L ETTIERI , P H D • European Commission-DG Joint Research Centre, Institute
for Environment and Sustainability, Ispra (VA), Italy
M ARK R. L ITZOW, MD • Division of Hematology Research, Mayo Clinic, Rochester,
Minnesota, USA
G EORG L URJE , MD • Division of Medical Oncology, University of Southern
California, Norris Comprehensive Cancer Center, Keck School of Medicine, Los
Angeles, California, USA
xi
xii Contributors

S HARON M ARSH , P H D • Washington University, Division of Molecular Oncology,

St. Louis, Missouri, USA
H OWARD L. M C L EOD , P H D • University of North Carolina, Chapel Hill, Chapel Hill,
North Carolina, USA
C EDRIC M ERCIER , MD, P H D • Medical Oncology, Université de la Méditerranée,
Marseille, France
G ERARD M ILANO , P H D • Laboratoire d’Oncopharmacologie, Centre Antoine
Lacassagne, Nice, France
H IRONOBU M INAMI , MD • Division of Oncology/Hematology, National Cancer
Center Hospital East, Kashiwa, Japan
M IGUEL A NGEL M OLINA , P H D • Catalan Institute of Oncology, Hospital Germans
Trias i Pujol, Badalona (Barcelona), Spain
K AZUYUKI NAKAMURA , P H D • Department of Biochemistry and Functional
Proteomics, Yamaguchi University, Graduate School of Medicine, Yamaguchi, Japan
YASUSHI O KUNO , P HARM D • Department of PharmacoInformatics, Graduate School
of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
YONG Q IAN , P H D • The Pathology and Physiology Research Branch, Health
Effects Laboratory Division, National Institute of Occupational Safety and Health,
Morgantown, West Virginia, USA
J OSE L UIS R AMIREZ , P H D • Catalan Institute of Oncology, Hospital Germans Trias i
Pujol, Badalona (Barcelona), Spain
R AFAEL ROSELL , MD • Catalan Institute of Oncology, Hospital Germans Trias i
Pujol, Badalona (Barcelona), Spain
K IMIE S AI , P H D • Division of Biosignaling, National Institute of Health Sciences,
Tokyo, Japan
YOSHIRO S AITO , P H D • Division of Biochemistry and Immunochemistry, National
Institute of Health Sciences, Tokyo, Japan
F ERNANDA S ALAZAR , P H D • Catalan Institute of Oncology, Hospital Germans Trias
i Pujol, Badalona (Barcelona), Spain
M ARIACARMELA S ANTARPIA , MD • Department of Clinical Oncology and Innovative
Therapy, University of Messina, School of Medicine, Messina, Italy
J UN - ICHI S AWADA , P H D • Division of Biochemistry and Immunochemistry, National
Institute of Health Sciences, Tokyo, Japan
E LKE S CHAEFFELER , P H D • Margarete-Fischer-Bosch, Institute of Clinical
Pharmacology, Stuttgart, Germany
M ATTHIAS S CHWAB , MD • Department of Clinical Pharmacology, University
Hospital Tuebingen, Tubingen, Germany
Contributors xiii

J EFFREY S ETTLEMAN , P H D • Massachusetts General Hospital Cancer Center,

Charlestown, Massachusetts, USA

S REENATH V. S HARMA , P H D • Massachusetts General Hospital Cancer Center,

Charlestown, Massachusetts, USA

X IANGLIN S HI , P H D • The Pathology and Physiology Research Branch, Health

Effects Laboratory Division, National Institute of Occupational Safety and Health,
Morgantown, West Virginia, USA

R ICHARD S IMON , DS C • Biometric Research Branch, National Cancer Institute,

Bethesda, Maryland, USA

A NURAG S INGH , P H D • Massachusetts General Hospital Cancer Center, Charlestown,

Massachusetts, USA

M ARGARET R. S PITZ , MD, MPH • Department of Epidemiology, The University of

Texas, M.D. Anderson Cancer Center, Houston, Texas, USA

M ARTIN S TANULLA , MD, MS C • Pediatric Hematology Oncology, Hannover Medical

School, Hannover, Germany

NATALIE S TICKLE , MS C • UHN Microarray Centre, University Health Network, T.

Robert Beamish Family Convergence Centre of Medical Discovery,
Toronto, Ontario, Canada

M IQUEL TARON , P H D • Catalan Institute of Oncology, Hospital Germans Trias i

Pujol, Badalona (Barcelona), Spain

S OKEN T SUCHIYA , P HARM D • Department of Genomic Drug Discovery Science,

Department of PharmacoInformatics, Graduate School of Pharmaceutical Sciences,
Kyoto University, Kyoto, Japan

G OZOH T SUJIMOTO , MD, P H D • Department of Genomic Drug Discovery Science,

Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan

A NDERS WALLQVIST, P H D • Laboratory of Computational Technologies, SAIC-

Frederick, Inc., Frederick, Maryland, USA

N EIL W INEGARDEN , MS C • UHN Microarray Centre, University Health Network,

T. Robert Beamish Family Convergence Centre of Medical Discovery,
Toronto, Ontario, Canada

K WONG -K WOK W ONG , P H D • Department of Gynecologic Oncology, The University

of Texas, M.D. Anderson Cancer Center, Houston, Texas, USA
xiv Contributors

X IFENG W U , MD, P H D • Department of Epidemiology, The University of Texas,

M.D. Anderson Cancer Center, Houston, Texas, USA
TAL Z AKS , MD, P H D • Translational Medicine-Oncology, GlaxoSmithKline,
Collegeville, Pennsylvania, USA
I G ENOMIC E XPERIMENTAL
A PPROACHES IN O NCOLOGY
 1 Toward the Realization
of the Promise of Microarrays
in Oncology

Natalie Stickle, MSc, and Neil

Winegarden, MSc
CONTENTS
M ICROARRAYS FOR C ANCER R ESEARCH :
H ISTORICAL P ERSPECTIVES
T UMOR H ETEROGENEITY
S TEM C ELL T HEORY OF C ANCER
P ROFILING S MALL S AMPLES WITH M ICROARRAYS :
A MPLIFICATION
P ROSPECTIVE V ERSUS R ETROSPECTIVE S TUDIES
D IAGNOSTICS AND P ROGNOSTICS
C ONCLUSIONS
R EFERENCES

S UMMARY
Microarrays have long been promised to be a tool that might one day revolutionize
oncology research and drug development. Despite the tremendous potential, however,
there have been few breakthroughs that can be directly attributed to microarray-based
profiling. While many researchers now say that microarrays have been over-hyped, it
is more likely that early experiments were simply conducted in a naı̈ve manner. Many
believe that as technology and our understanding of experimental design improves, so

From: Cancer Drug Discovery and Development: Genomics and Pharmacogenomics

in Anticancer Drug Development and Clinical Response
Edited by: F. Innocenti, DOI: 10.1007/978-1-60327-088-5 1, 
c Humana Press, Totowa, NJ

3
4 Part I / Genomic Experimental Approaches in Oncology

too will the end results. We believe that with new approaches, particularly the use of
pure cell populations potentially coupled with new and improved RNA amplification
methodologies, the promise of microarrays for oncology finally will be realized.

Key Words: Microarrays; gene expression; RNA amplification; laser capture

microdissection; cancer stem cell theory; pure cell populations; diagnostics;
prognostics

1. MICROARRAYS FOR CANCER RESEARCH:

HISTORICAL PERSPECTIVES
Gene expression profiling in cancer has become routine in recent years and to date en-
compasses the largest category of research using DNA microarrays. Microarray analysis
has been used to assess transcriptome-level (expression analysis) (1) and genome-level
(single-nucleotide polymorphism (2), amplification/deletion (3)) differences in cancer-
ous versus normal cell samples. Furthermore, microarray analysis has been used to iden-
tify putative diagnostic and prognostic markers that will hopefully translate into clinical
tests that can be used to stratify patients and determine treatment regimens (4).
Comparison of treatment response at both the RNA and DNA levels has been con-
ducted in an effort to further direct the most optimal therapy for an individual patient.
Classically, many large-scale studies have revealed groups of genes differentially ex-
pressed between cancerous and normal cells, including genes known to be important
for neoplastic transformation. However, despite the vast amounts of data generated to
date, there has been limited translation from the research bench to the clinic. There are
few drugs currently in the pipelines of the pharmaceutical industry that were developed
because of a microarray-identified target (microarrays have been used to validate and
test many potential pharamcophores, but few targets have actually been identified by
microarrays).
One of the greatest early promises of microarrays was the ability to identify signatures
of biomarkers (panels of genes), which would be able to diagnose or prognose complex
diseases such as cancer. Despite some notable exceptions, however, this potential benefit
has largely gone unmet. As researchers have looked to capitalize on their investments in
these high-throughput technologies, there has been some shift recently toward genotyp-
ing rather than transcript analysis in cancer biology. While these techniques are promis-
ing, there is still little sign at this early stage that suggests that this shift in focus will
lead to revolutionary changes in the type of data obtained.
All of these developments have led to the discouragement of many researchers who
have begun to re-evaluate their desire to use high-throughput technologies for oncology
research. Despite this, it is our belief that the technology in-and-of itself is not inher-
ently flawed, and rather it is the way that microarrays have typically been applied to
cancer research that has led to a decrease in the potential power of the analyses. Recent
advances in microarray design, labeling, and amplification technologies—and even our
theory of the origin of cancerous lesions—are leading to changes in the way microarray
experiments are conducted. These developments are likely to finally fulfil the realization
of the promise of microarrays in oncology.
Chapter 1 / Microarrays and Oncology 5

2. TUMOR HETEROGENEITY
Many of the anti-neoplastic drugs currently available on the market specifically target
the cancerous tumor and do not account for the overall heterogeneity of the disease. It is
reasonable to assume that due to the complex make-up of a tumor, combined therapies
may be necessary to treat the different cell types involved. One major hurdle to under-
standing tumor cells at the gene expression level is our limited ability to isolate pure
populations of diseased cells.
Tumor cells can be extracted from blood with relative ease, but with solid tumors,
such as those found in prostate or breast cancer, tumor cells cannot be isolated so eas-
ily. Indeed, these cell populations often show very large degrees of heterogeneity. Solid
tumors expand by infiltrating within the normal tissue scaffold, making it difficult but
necessary to dissect the malignant cells away from the normal tissue structures (5). For
those samples that are dissectible in theory, some cells outside the area of interest may
remain attached, resulting in a sub-pure sample that confounds analysis ( 5). Even in
cases where the tumor can be distinguished easily from the surrounding normal tissue,
the tumors themselves remain heterogeneous.
Tumor heterogeneity refers to the existence of distinct subpopulations of tumor cells
with specific characteristics within a single neoplasm. Breast cancer is a classic example
of a heterogeneous disease. First, the term breast cancer does not itself refer to a single
disease. Breast cancers include many different diseases including (but not limited to)
adenomas, papillomas, invasive ductal carcinoma, ductal carcinoma in situ (DCIS), and
lobular carcinoma in situ (LCIS) (6).
Breast tissue itself is highly heterogeneous, containing stroma, epithelial, ductal, lym-
phocytic, fibroblastic and other structures which further complicate issues. Even in mor-
phologically pure populations, there remains a great deal of molecular heterogeneity.
This is true also of breast cancer, which includes alterations in ER, Her2 BRACA1, and
BRCA2 status among others (4,7,8,9).
In addition to the well-recognized clonal mutations that afford a survival advantage to
cells within the tumor microenvironment (i.e., areas of hypoxia, acid pH, and poor nutri-
ent supply) (10,11), there is evidence of many randomly distributed unselected mutations
that contribute to the heterogeneity of cells in a tumor (12,13). Overall, heterogeneity
is likely the major hurdle in the usefulness of microarray technology in tumor biology.
Important cell-specific signatures, particularly in rare cell populations, would be lost in
a microarray analysis as to date, many microarray analyses used heterogeneous tumor
samples for profiling. More recently, researchers have begun to turn their attention to
microdissected, relatively pure cell populations for microarray analysis.

2.1. Tumor Microdissection

Microdissection of a tumor was initially carried out using a standard syringe ( 5).
The tumor was viewed under a microscope so that it could be separated from the sur-
rounding normal tissues. For this method to be effective, tumors needed to be easily
defined under the microscope, which limited the number of samples that were compati-
ble. Microdissection techniques have further evolved to include techniques such as laser
microbeam microdissection (LMM) and laser capture microdissection (LCM). These
6 Part I / Genomic Experimental Approaches in Oncology

techniques afford laser precision and the possibility to isolate single cells ( 14). As a
result, these methods have gained importance as tools to obtain purified cell populations
from tissue (15).
LCM has been used as a tool for the purification and stratification of tumor samples
that are to be used in subsequent microarray analyses (16,17,18,19,20,21). Separation of
the tumor from the surrounding tissue provides cleaner populations of cells that can be
used for the subsequent analysis of gene expression profiles or chromosomal aberrations.
Furthermore, by purifying individual cell types, distinct signatures for each cell type can
also be obtained using microarray profiling ( 22, 23). Such analyses demonstrate that
the cancers are not homogenous at the molecular level. Yang et al. (23) demonstrated
that the signature obtained from bulk tumors can include many common genes found
in more pure populations of epithelial cells, but that both the purified epithelia and the
bulk tumor have characteristic sets of genes that are unique as well. In analysing the
results of a gene expression profiling on purified epithelia from ER-␣ positive and ER-␣
negative cells, a total of 146 ER-related genes were identified. When the authors then
compared the expression profiles to those of the bulk tumors from which the purified
cell populations were obtained, 61 of these genes were identified as being in-common.
Thus, 85 of the genes identified (58%) could only be identified by using the pure cell
population (23).
As LCM becomes more commonly used upstream of microarray analysis, the quality
of data as well as the usefulness of the signatures obtained will likely increase. With
a greater understanding of the molecular heterogeneity at the transcriptome level for
example, it will be possible to develop combinatorial therapies that use multiple different
drugs to target the several different cell types involved in the disease. Very few other
technologies hold the promise of microarrays when it comes to understanding cancers at
this level of detail, and thus it is likely that these new and more focused approaches will
lead to new therapeutic strategies.

3. STEM CELL THEORY OF CANCER

While natural heterogeneity of cancer specimens is well documented, and the ability
to isolate hundreds to thousands of cells with techniques such as LCM have allowed
microarray analysis of more pure populations, the recent re-introduction of the cancer
stem cell theory (24,25) has presented an extreme example of the need to isolate rare
populations of cells. Cancer stem cells appear to exist at a frequency on the order of 1
in 1000 cells (26). While obtaining completely pure populations of stem cells remains
difficult, highly enriched populations of these cells can be isolated using flow cytometry
separating on the basis of the presence of certain specific cell markers.
One hypothesis is that cancer stem cells are relatively immune to current therapeutic
methodologies, due in large part to the fact that these cells divide only infrequently and
are generally quiescent. As these cells also have the potential to cause the formation of
a new cancer, therapeutics targeted to these specific cells would be of immense benefit.
The fact that cancer stem cells occur in such low frequency dictates that what may ar-
guably be the most important signature to identify within the tumor is completely lost in
most microarray analyses.
Chapter 1 / Microarrays and Oncology 7

While laser capture microdissection and other related techniques may allow for the
generation of relatively homogenous cell populations, most such purification steps will
largely eliminate any potential cancer stem cells, because these techniques necessitate
the identification of differentiated cell types for isolation. As such, cancer stem cells
would frequently be eliminated from analysis, and any cancer stem cells that may be
accidentally included in the analysis would be in such low occurrence that their signature
would most certainly be missed.
Assuming that the cancer stem cell theory is in fact correct, then we can see how
it is possible that the reason we have not yet learned a great deal about cancer using
microarrays is not due to the technology itself, but more due to its application and the
sample that is being used to profile the cancer. As techniques are introduced that will
allow the profiling of single cells, it may be possible to obtain specific gene expression
signatures for these cancer stem cells. Drugs developed specifically to target these genes
will revolutionize the means by which the disease is treated. Elimination of these tumor
progenitors would not only help eradicate the initial disease, but would also essentially
prevent recurrence as well. The problem, however, is that it is very difficult to profile
individual cells, and techniques to do so have only recently been introduced.

4. PROFILING SMALL SAMPLES WITH MICROARRAYS:

AMPLIFICATION
With advances in tumor microdissection and flow sorting, the current trend in as-
sessment of malignancies is toward evaluation of ever-smaller samples, in an attempt
to separate the effects of tumor heterogeneity from the tumor-specific genetic signature.
Furthermore, as diagnostic methodologies improve, the overall tumor, when biopsied,
is becoming smaller and smaller. In many cases, much of the tumor material obtained
during a biopsy is required for pathology, leaving very small amounts of tissue for the
molecular researcher.
There is also a trend toward performing minimally invasive biopsy techniques such as
needle/core biopsies and fine needle aspirates (FNAs). The amount of material available
from such techniques, even if the bulk (non-microdissected) sample were to be used,
can be very small. Traditional DNA microarray analysis requires a fairly substantial
amount of material for analysis; some 5–10 ␮g of total RNA is usually required for
analysis in the absence of amplification. In order to obtain this quantity of RNA from
typical epithelial cells, as many as 5 ×105 or 1 ×106 cells are required. Clearly, this total
RNA requirement poses a challenge when studying microdissected samples, but it also
presents a challenge when using small, unique clinical samples, such as FNA biopsies.
Assersohn et al. (27) reported that the median recovery of breast FNAs was 202,500
cells, which translated to between 0.81 and 1.42 ␮g of total RNA, well below the typi-
cal requirement for a microarray experiment. Consequently, many researchers routinely
incorporate amplification methods into their gene expression screens. There is currently
no consensus in the literature with regard to the best amplification strategy to utilize.
Indeed, even the fidelity of the various amplification methods available is the subject
of dispute. Some researchers report that amplified RNA can substitute for total RNA
in gene expression experiments (28,29,30) while others have shown that amplification
of RNA introduces bias to the results ( 31, 32). It is quite possible, however, that this
8 Part I / Genomic Experimental Approaches in Oncology

apparent disagreement is caused by a difference in what the researchers consider to

be an acceptable level of fidelity. It is well accepted that any amplification technique
will introduce some degree of bias ( 33). What is more up for debate is to what level
this bias can be introduced before it causes a detrimental effect on the analysis. Some
would argue that no level of bias is acceptable; however, it is likely that as long as the
amplification methodology is highly reproducible, statistical methods can be used to
model and remove or correct for the bias.

4.1. T7-(“Eberwine”)–Type Amplification

The most popular methods of RNA amplification for microarray-based transcription
profiling are modified from the classic Eberwine protocol ( 34). The basic premise of
this approach is that mRNA is converted into cDNA using an oligo-dT primer that has
an additional sequence encoding the T7-promoter appended to it. After performing a
second strand cDNA synthesis, an artificial gene under the control of a T7-promoter is
created. This gene can then be transcribed by T7-polymerase creating multiple RNA
copies of each gene.
This basic technique is utilized by many researchers and is a standard part of the
Affymetrix, Agilent, and Illumina labeling protocols. Proponents of this technique point
to the fact that it is linear in nature, while providing great increases in sensitivity. This
protocol can be performed in tandem (multiple rounds of amplification) to obtain even
greater amounts of material; however, each successive round of amplification does in-
troduce more bias (35). With multiple rounds of amplification, as little as 10 ng of total
RNA can be used to create labeled material for microarray analysis ( 35, 36). While
this thousand-fold increase in sensitivity is impressive, the requirement for a minimum
of 1000 cells can still be problematic for many researchers using LCM techniques for
tumor profiling. Another drawback of this method is the time requirement for amplifica-
tion: a single round of amplification adds between 1.5 and 2 days of time to the labeling
procedure, whereas two rounds of amplification can add 3 days (37).

4.2. Isothermal Amplification Strategies

While T7-mediated amplification has shown a great deal of promise and has been
employed in numerous studies (including nearly all microarray experiments to date us-
ing LCM isolated cells), other methodologies are being sought that may provide both
greater sensitivity and fidelity while reducing the time requirement. One such method,
Ribo-SPIA from NuGen Technologies, allows for amplification from as little as 5 ng of
total RNA in as little as 5 hr (37,38). Ribo-SPIA utilizes a chimeric RNA/DNA primer
that links a known RNA sequence tag to the 5 -end of an oligo-dT cDNA primer. After
creating double-stranded cDNA from the mRNA, RNaseH is used to digest only the
short RNA portion of the primer, which enables another chimeric primer to bind, and a
new cDNA strand then displaces the old one. This process is repeated over many cycles
creating thousands of single-stranded cDNA all in one tube (37,38). The major advantage
of this technique over T7-based amplification is that it can be completed in a matter of
hours, rather than days.
A novel signal amplification method based on the rolling circle amplification (RCA)
technology has also been reported. RCA involves many rounds of isothermal enzymatic
Chapter 1 / Microarrays and Oncology 9

synthesis where DNA polymerase extends a hybridized primer (consisting of a DNA

minicircle) by continuously progressing around the circular DNA probe to replicate the
sequence over and over again, producing a single amplified product that remains linked
to the DNA primer.
Nallur et al. (39) demonstrated that RCA is a sensitive method for detecting nucleic
acid signals of microarrays. They reported that for DNA targets there was at least an
8000-fold increase in detection sensitivity over standard hybridization under the same
conditions. Although this indicates a great potential to amplify signal from very small
amounts of hybridized material, it is necessary to alter the hybridization protocol to
account for the corresponding hybridization kinetics.

4.3. Signal Amplification Post-Hybridization

Another approach employed to increase sensitivity of microarrays is signal amplifica-
tion at the post-hybridization stage. One such technique is tyramide signal amplification
(TSA) which requires 20–100 times less RNA than direct cDNA labeling ( 40). This
method was originally used in immunohistochemistry and has been an important tool
for immunofluorescence microscopy (41,42).
The TSA technique involves the incorporation of a biotin- or fluorescein-labeled nu-
cleotide and its subsequent detection with conjugate reporter molecules after hybridiza-
tion. The enzyme portion of the conjugate is horseradish peroxidase (HRP), which cat-
alyzes the breakdown of tyramide that results in the deposition of numerous Cy3 labels
adjacent to the immobilized HRP. Karsten et al. (40) evaluated the use of the TSA method
with archived samples on microarray gene expression analysis. Although they report
that the TSA method worked well to amplify the signal from frozen tissue with good-
quality RNA, there were still issues of poor reproducibility and reliability especially
with formalin-fixed tissue samples. RNA from fixed tissue is generally of poor quality.
Partial degradation and modification by proteins or chemicals usually results in disrupted
cDNA synthesis with a high rate of misincorporations and short product length. These
factors influence the specificity of hybridization, and the TSA method did not reduce
their effects (40).
Alternatively, dendrimers (highly branched molecules) can be used to increase the
amount of label per nucleotide, and subsequently per labeled cDNA/cRNA molecule
(www.genisphere.com). This technology involves dendrimers, each carrying hundreds
of fluorescent tags, binding to each hybridized molecule. Yu et al. (43) evaluated both
the TSA and Genisphere 3DNA dendrimer labeling systems and reported that the 3DNA
system was less time consuming and produced superior and more consistent results.
However, hybridizations with less than 2 ␮g of RNA produced low and variable spot
intensities (43).

4.4. Exponential Amplification

Each of the aforementioned amplification strategies is of potential use in studying
smaller, purer cell populations; however, few if any of these techniques are amenable to
the study of single cells, which may be necessary if analyzing cancer stem cells. In order
to study individual cells, it is necessary to have a technique that is sensitive down to the
10 Part I / Genomic Experimental Approaches in Oncology

low picogram range (<10 pg of total RNA). In order to obtain this level of sensitivity,
exponential amplification methods are generally necessary.
The initial opinion about the use of exponential amplification strategies for microarray
research was sceptical, with many researchers concerned that exponential amplification
would introduce too much bias into the analysis (44,45,46,47,48,49,50). There is, how-
ever, a growing body of evidence that exponential amplification is indeed a viable option
for microarray profiling (33,51,52,53,54,55,56). Exponential amplification strategies can
amplify as little as 10 pg of total RNA (roughly equivalent to one cell) while preserving
most of the original abundance relationships of the mRNA species (29).
The method presented by Iscove et al. (29) employs two important strategies to help
maximize fidelity of the amplified products with the original mRNA pool: generation of
short, uniform length amplicons (generally around 300 bp) and minimizing the number
of amplification cycles to avoid saturation. Extensions of this basic method have also
been presented in which fidelity has been even further improved, indicating that expres-
sion profiling of single cells has become a reality (52,57,58).

5. PROSPECTIVE VERSUS RETROSPECTIVE STUDIES

Microarray technology generally requires the isolation of high-quality, intact RNA.
The labeling technologies used in microarray experiments utilize oligonucleotide
primers comprised of poly-dT in order to create cDNA in a process mediated by the
polyA tail of the mRNA ( 59). With high-quality intact RNA it is not unusual to be
able to obtain cDNA products in excess of 1 kb in length. The probes for microarrays,
although typically 3 -biased, may be present at varying lengths from the polyA tail of the
mRNA, and in the case of multiple probe designs (such as those implied by Affymetrix),
some of these probes may extend several hundred bases away from the polyA tail.
In order to ensure that highly intact RNA is isolated, researchers have generally relied
on either tissue culture or, in the case of clinical samples, fresh or frozen specimens.
This has presented another level of challenge to the study of oncology because most
cancer centers around the world have relied on the process of archiving biopsies through
a process of formalin fixation and paraffin embedding (FFPE). It is exceedingly difficult
to isolate good-quality intact RNA from FFPE archived samples. Often, RNA yields
are low, and the RNA is highly fragmented when extracting from an FFPE sample.
In addition, due to chemical modification of the mRNA during the fixation process,
it has been shown that as little as 3% of the total RNA is actually accessible to cDNA
synthesis strategies (60). Traditional microarray techniques and designs have therefore
limited the ability to utilize such archival samples and as such have greatly reduced the
power of analysis. The cDNAs produced from the highly fragmented RNA isolated from
FFPE archived samples tend to be short and heterogeneous in length compared to those
produced from intact RNA.
Due to the issues surrounding the analysis of archived samples, most microarray-
based oncology studies have relied on prospective analyses—running samples as they
become available, rather than retrospective analyses—running archived samples. While
the prospective analysis tends to provide excellent-quality data, the associated clinical
data from archived samples tends to be of greater value. In particular, in obtaining
archived samples, it is possible to concentrate on those patients for which outcome
information is known (e.g., survival rates, recurrence, treatment success, and so on).
Chapter 1 / Microarrays and Oncology 11

Because much of this information may not be known for several years after the samples
are first obtained, the true value of a prospective analysis may not be realized for some
time to come. The other issue is that for rare cancers, the use of prospective analyses
can severely limit the overall sample size in terms of number of biopsies obtained. The
ability to delve into archived samples can significantly increase the number of samples
profiled, dramatically increasing the power of the analysis.
A recent analysis of the effect of FFPE archiving on the gene expression patterns
obtained using traditional Affymetrix arrays has been published by Scicchitano et al.
( 61). Although the authors make the case that there are a great number of genes that
overlap when comparing frozen to FFPE archived samples these represent only 27%
(218) of the total number (799) of differentially expressed genes found in the frozen
tissue samples. What is perhaps of more concern is that an additional 138 genes were
found to be differentially regulated only in the FFPE samples. If one concludes that
these genes are indeed false-positives, then the false positive rate is 39% for these FFPE
samples. The authors conclude, however, that for certain types of analysis, in which one
is attempting to implicate pathways that may be involved in a particular process, this rate
of success may be sufficient.
Many newer technologies from labeling techniques to array designs are helping to
combat the issues of dealing with FFPE samples, opening the door to larger and more
robust retrospective analyses.

5.1. New Array Designs

As mentioned, the labeled products (cDNA or cRNA) generated from FFPE-derived
mRNA tend to be smaller in size than those prepared from fresh or frozen samples. Most
arrays that have been designed and used to date have attempted to bias probe design to
the 3 end of the mRNA; however, the degree of this bias has been based on the idea of
cDNA products that are up to 1 kb or more in length. Newer designs using more rigorous
algorithms that further bias the probe design to within only a few hundred bases of the
polyA tail of the mRNA have been made available.
Affymetrix, for example, has released a product termed X3P (for eXtreme 3-Prime).
The X3P arrays have been designed such that the majority of probe-sets fall within
300 bases of the 3 end of the mRNA, compared to traditional Affymetrix de-
signs which may use probe-sets spanning as much as the first 600 bases (see
http://www.affymetrix.com/products/arrays/specific/x3p.affx). Despite the introduction
of these new arrays, publications using FFPE-archived samples remain few in number
(62,63). Indeed, to date, no comprehensive comparison, such as the one carried out by
Scicchitano et al., has been carried out using the 3XP arrays, although such an analysis
would be of great value.

5.2. New Approaches

Rather than attempting to adapt current microarray technologies to FFPE samples,
others have decided to develop new approaches to the study of archived samples via
array-based analysis. Illumina has developed a new approach termed DASL (for cDNA-
mediated annealing, selection, extension, and ligation) which has been demonstrated to
be effective for some FFPE archived samples (64).
12 Part I / Genomic Experimental Approaches in Oncology

The first technical difference between the DASL assay and traditional microarray
assays is that a random priming technique is utilized, and as such an intact polyA tail
is not required for efficient labeling. Other differences include a mechanism by which
two oligonucleotide probes must recognize the RNA of interest in order to be extended,
ligated, and later amplified via PCR. This technique leads to both high specificity and
high sensitivity (due to the PCR reaction).
In practical tests it was shown that the DASL technique, when assayed for a to-
tal of 231 genes, showed improved overlap of identified genes (between 33% and
44%) compared to the previous study by Scicchitano et al., but that the false positive
rate was similar (26%–39%) ( 65). While the DASL technique provides an interest-
ing alternative for studying FFPE-archived samples, the current technology is limited
by the ability to multiplex the PCR reaction. As such, users are generally limited
to using between 512 and 1536 different genes in the assay, which is far from the
whole genome (see http://www.illumina.com/products/geneexpression/dasl assay.ilmn).
However, Illumina has developed a product that targets specific cancer-related genes
with the intention of allowing analysis out of FFPE samples (see http://www.illumina.
com/products/geneexpression/dasl human ca panel.ilmn).
Amplified labeling techniques are generally employed when dealing with FFPE
samples. Arcturus has generated a specific kit designed to work with FFPE-archived
samples. The general procedure involves deparaffinization of the sample, isolation
of RNA, and subsequent amplification of the mRNA by a T7-mediated technique
(www.arctur.com/research portal/products/paradise main.htm). Also, although it has
not yet been demonstrated, it is possible that an exponential technique such as that from
Iscove et al. (29) may also be beneficial because such techniques are highly amenable to
the shorter mRNA fragments that are typically isolated from FFPE-archived samples.
As we move forward we are witnessing two trends that will dramatically improve
our ability to extract information from high-throughput genomic studies in cancer: (1)
techniques are being introduced that enable analysis of FFPE-archived samples, and
(2) clinical institutes are increasingly choosing to freeze at least some of the biopsy
material obtained from a patient. The impact of the first trend will be more immediate
as we are able to correlate gene expression profiles and genomic aberrations with the
vast amounts of clinical data that has already been obtained. In the near future, we will
also see the benefit of the switch to using freezing for archiving samples because these
samples are much more amenable to microarray analysis. As the clinical information
for these samples expands, the amount of information that we are able to extract from a
microarray analysis will also greatly increase.

6. DIAGNOSTICS AND PROGNOSTICS

There has been a great deal of discussion in the literature about the use of microarrays
to develop panels of biomarkers that can be used for diagnostics, prognostics, and thera-
nostics (66,67,68,69). The ultimate goal of microarray-based diagnostics/prognostics and
other techniques is to identify a panel of genes that can be used to provide information
that is of higher diagnostic/prognostic value compared to traditional methodologies. To
date, most reports of diagnostic and prognostic panels require use of primary tumor
material. As such, there is little that has been done to provide for earlier diagnosis of
Chapter 1 / Microarrays and Oncology 13

disease using surrogate markers. Consequently, the greatest progress has really been in
the area of prognostics. Many studies have been designed to compare various aspects of
the tumor to clinical outcome.
Prognosis also leads to the concepts of pharmacogenomics and theranostics. The goal
of pharmacogenomics is to tailor therapies to individual patients, choosing the regimen
that is most likely to provide a positive treatment outcome (69,70,71,72). Theranostics,
on the other hand, aims to monitor treatment as it is being administered, adjusting the
therapy as necessary to keep up with the progression of the disease (73,74,75). There
has been some promising data obtained on pharmacogenomics for cancer (76); however,
although there has been some focus on theranostics for infectious diseases, there has
been little progress in this area related to cancer to date.

6.1. Prognostics
It is not uncommon for a pathologist to examine tumor samples from two patients
and to classify them as being the same type, stage, and grade, only to have the two
patients respond very differently to treatment and to progress through their diseases at
very different rates ( 66). This underscores the fact that current pathology tests, while
often well defined, are still insufficient for the efficient and effective prognosis of disease.
Pathologists are employing more and more molecular techniques for prognosis rather
than relying simply on histology and morphology; however, with a disease as complex
as cancer, simplex molecular analyses may not suffice.
It is the goal of array-based prognostics to utilize a panel of markers that are more
accurately able to stratify patients into prognostic groupings. There are several examples
in the literature of biomarker panels identified by microarray analysis that may have
some relation to prognostic outcome (4,77,78,79,80). However, the issue remains that
many of these biomarker panels do not afford sufficient specificity or sensitivity to war-
rant clinical application. Microarrays offer the potential to profile tumors at a finer level
of detail than what can be obtained by the pathologist; however, it is the reliance on
classification from traditional pathology that has confounded many studies.
Often a microarray study will attempt to profile a series of tumors and look for the
data to cluster on the basis of some pathological classification such as stage, grade, or
size. In truth, this approach is likely more applicable to understanding the molecular un-
derpinnings of the transition from one stage to another, but is less informative in terms of
prognostics. One of the reasons is that due to the use of prospective studies, researchers
have relied on secondary indicators of outcome to develop prognostic markers. However,
with this approach, realistically, the best that a prognostic panel of markers can hope to
achieve is the same success rate as what is seen when using the generally easier-to-
obtain pathology-based classification data. In several instances however, there have been
demonstrations that microarray-based profiling of tumors can identify sub-classifications
that could not be determined by classical pathology.

6.2. Diagnostics
Early diagnostics is perhaps the least well represented among microarray-derived
clinical applications. Analysis of the primary tumor requires a priori that one already
14 Part I / Genomic Experimental Approaches in Oncology

be aware of the existence of the disease. As such, profiles and markers that may be
identified at this late stage may not be useful in early stage diagnosis. In addition, many
of the markers identified in such analyses may be specific to the tumor tissue itself, and
may not exist outside of the tumor niche. Arguably, a much more useful tool would be
one that allowed for early diagnosis of a disease using peripheral blood samples or cells
sedimented from urine, for example. There are a couple of different strategies that might
be employed to allow for this; however, neither strategy has been fully implemented in
the literature as of yet.
The first approach would be to assay peripheral blood of cancer patients looking for
signatures that are predictive of the disease. The major issue with this approach, of
course, is that the cells of interest in non-blood-borne cancers are likely to be in very
low concentration in the blood, and obtaining a signature is therefore unlikely.
The other approach is based largely on informatics. In such an approach, tumors
would be profiled in contrast to normal tissue. Tumor-specific markers would be iden-
tified via microarrays, as has been done in many publications already. From here, the
list of tumor-specific markers would be analyzed to determine if any of these markers
represented proteins which were likely to be secreted out of the cell and which may be
detected in the peripheral blood stream. Preferably multiple markers would be identified
that could be tested using multiplex ELISA assays (antibody arrays). Such work will
take time, however, because once the potential markers are identified, antibodies must
be generated, validated, and tested for effectiveness as an early diagnostic tool. Such
work is being done, but little has been published so far.

7. CONCLUSIONS
Many reports, reviews, opinions, and letters have been published that extol the virtues
of microarray technology, suggesting that microarrays will eventually become the new
standard for drug discovery, diagnostics, and prognostics. However, progress in these
areas to date has been disappointing.
In a recent report, the U.S. federal Government Accountability Office (GAO) criti-
cized genomic tools such as microarrays, indicating that not only did they increase the
cost of R&D, but that they also actually slowed research progress (81). Such criticisms
are also being heard in the scientific sphere as well (82). Despite this, many researchers
remain confident that the promise of microarray technology will be fully realized in
the not-too-distant future. It is likely that microarrays have been seen to be a mature
technology when, in fact, the hardware itself may have been mature, but the approaches
are still in development.
Despite the fact that oncology has been the primary area of focus for most microarray
research, issues in the way that such analyses have been carried out may have severely
limited the value of the data obtained. Cancer is a complex and highly heterogeneous
disease that metes out its effects over years rather than days. The recent introduction of
technologies that allow for dissection of the disease to a more homogenous level, and
the ability to perform retrospective rather than prospective analyses, should allow for the
evolution of the technology to a point where real-world impact is seen.
Development of drugs that target the progenitors of disease and multi-drug therapies
that combat the different types of cancerous cells will be made possible by the ability
Chapter 1 / Microarrays and Oncology 15

to profile homogenous, yet minute, cell samples. In addition, identification of specific

markers from the progenitor cells may allow for early detection before the tumor can be
detected by conventional means.
Microarray technology has perhaps not yet lived up to the hype, but it is coming of
age, and its promise is soon to be realized at last.

REFERENCES
1. DeRisi J, Penland L, Brown PO et al. Use of a cDNA microarray to analyze gene expression patterns
in human cancer. Nat Genet 1996;14:457–460.
2. Primdahl H, Wikman FP, von der Maase H et al. Allelic imbalances in human bladder cancer:
genome-wide detection with high-density single-nucleotide polymorphism arrays. J Natl Cancer Inst
2002;94:216–223.
3. Pollack JR, Perou CM, Alizadeh AA et al. Genome-wide analysis of DNA copy-number changes
using cDNA microarrays. Nat Genet 1999;23:41–46.
4. van’t Veer LJ, Dai H, van de Vijver MJ et al. Gene expression profiling predicts clinical outcome of
breast cancer. Nature 2002;415(6871):530–536.
5. Rubin MA. Tech.Sight. Understanding disease cell by cell. Science 2002;296:1329–1330.
6. Mallon E, Osin P, Nasiri N et al. The basic pathology of human breast cancer. J Mammary Gland Biol
Neoplasia 2000;5:139–163.
7. Namba R, Maglione JE, Davis RR et al. Heterogeneity of mammary lesions represents molecular
differences. BMC Cancer 2006;6:275.
8. Perou CM, Sorlie T, Eisen MB et al. Molecular portraits of human breast tumors. Nature
2000;406:747–752.
9. Sorlie T, Perou CM, Tibshirani R et al. Gene expression patterns of breast carcinomas distinguish
tumor subclasses with clinical implications. Proc Natl Acad Sci USA 2001;98:10869–10874.
10. Dachs GU, Chaplin DJ. Microenvironmental control of gene expression: implications for tumor an-
giogenesis, progression, and metastasis. Semin Radiat Oncol 1998;8:208–216.
11. Schornack PA, Gillies RJ. Contributions of cell metabolism and H+ diffusion to the acidic pH of
tumors. Neoplasia 2003;5:135–145.
12. Bielas JH, Loeb KR, Rubin BP et al. Human cancers express a mutator phenotype. Proc Natl Acad
Sci USA 2006;103:18238–18242.
13. Loeb LA, Loeb KR, Anderson JP. Multiple mutations and cancer. Proc Natl Acad Sci USA
2003;100:776–781.
14. Suarez-Quian CA, Goldstein SR, Pohida T et al. Laser capture microdissection of single cells from
complex tissues. Biotechniques 1999;26:328–335.
15. Fend F, Raffeld M. Laser capture microdissection in pathology. J Clin Pathol 2000;53:666–672.
16. Alevizos I, Mahadevappa M, Zhang X et al. Oral cancer in vivo gene expression profiling assisted by
laser capture microdissection and microarray analysis. Oncogene 2001;20:6196–6204.
17. Hergenhahn M, Kenzelmann M, Grone HJ. Laser-controlled microdissection of tissues opens a win-
dow of new opportunities. Pathol Res Pract 2003;199:419–423.
18. Mohr S, Bottin MC, Lannes B et al. Microdissection, mRNA amplification, and microarray: a study
of pleural mesothelial and malignant mesothelioma cells. Biochimie 2004;86:13–19.
19. Ohyama H, Zhang X, Kohno Y et al. Laser capture microdissection–generated target sample for high-
density oligonucleotide array hybridization. Biotechniques 2000;29:530–536.
20. Wu MS, Lin YS, Chang YT et al. Gene expression profiling of gastric cancer by microarray combined
with laser capture microdissection. World J Gastroenterol 2005;11:7405–7412.
21. Yim SH, Ward JM, Dragan Y et al. Microarray analysis using amplified mRNA from laser capture
microdissection of microscopic hepatocellular precancerous lesions and frozen hepatocellular carci-
nomas reveals unique and consistent gene expression profiles. Toxicol Pathol 2003;31:295–303.
16 Part I / Genomic Experimental Approaches in Oncology

22. Coco S, Defferrari R, Scaruffi P et al. Genome analysis and gene expression profiling of neuroblastoma
and ganglioneuroblastoma reveal differences between neuroblastic and Schwannian stromal cells. J
Pathol 2005;207:346–357.
23. Yang F, Foekens JA, Yu J et al. Laser microdissection and microarray analysis of breast tumors reveal
ER-alpha–related genes and pathways. Oncogene 2006;25:1413–1419.
24. Clarke MF, Dick JE, Dirks PB et al. Cancer stem cells: perspectives on current status and future
directions: AACR Workshop on Cancer Stem Cells. Cancer Res 2006;66:9339–9344.
25. Wang JC, Dick JE. Cancer stem cells: lessons from leukemia. Trends Cell Biol 2005;15:494–501.
26. Becker AJ, McCulloch CE, Till JE. Cytological demonstration of the clonal nature of spleen colonies
derived from transplanted mouse marrow cells. Nature 1963;197:452–454.
27. Assersohn L, Gangi L, Zhao Y et al. The feasibility of using fine needle aspiration from primary breast
cancers for cDNA microarray analyses. Clin Cancer Res 2002;8:794–801.
28. Gomes LI, Silva RL, Stolf BS et al. Comparative analysis of amplified and nonamplified RNA for
hybridization in cDNA microarray. Anal Biochem 2003;321:244–251.
29. Iscove NN, Barbara M, Gu M et al. Representation is faithfully preserved in global cDNA amplified
exponentially from sub-picogram quantities of mRNA. Nat Biotechnol 2002;20:940–943.
30. Li Y, Ali S, Philip PA et al. Direct comparison of microarray gene expression profiles between non-
amplification and a modified cDNA amplification procedure applicable for needle biopsy tissues.
Cancer Detect Prev 2003;27:405–411.
31. Li J, Adams L, Schwartz SM, Bumgarner RE. RNA amplification, fidelity, and reproducibility of
expression profiling. C R Biol 2003;326:1021–1030.
32. Nygaard V, Loland A, Holden M et al. Effects of mRNA amplification on gene expression ratios in
cDNA experiments estimated by analysis of variance. BMC Genomics 2003;4:11.
33. Subkhankulova T, Livesey FJ. Comparative evaluation of linear and exponential amplification tech-
niques for expression profiling at the single-cell level. Genome Biol 2006;7:R18.
34. Van Gelder RN, von Zastrow ME, Yool A et al. Amplified RNA synthesized from limited quantities
of heterogeneous cDNA. Proc Natl Acad Sci USA 1990;87:1663–1667.
35. Wang E, Miller LD, Ohnmacht GA et al. High–fidelity mRNA amplification for gene profiling. Nat
Biotechnol 2000;18:457–459.
36. Baugh LR, Hill AA, Brown EL et al. Quantitative analysis of mRNA amplification by in vitro tran-
scription. Nucleic Acids Res 2001;29:E29.
37. Barker CS, Griffin C, Dolganov GM et al. Increased DNA microarray hybridization specificity using
sscDNA targets. BMC Genomics 2005;6:57.
38. Dafforn A, Chen P, Deng G et al. Linear mRNA amplification from as little as 5 ng total RNA for
global gene expression analysis. Biotechniques 2004;37:854–857.
39. Nallur G, Luo C, Fang L et al. Signal amplification by rolling circle amplification on DNA microar-
rays. Nucleic Acids Res 2001;29:E118.
40. Karsten SL, Van Deerlin VM, Sabatti C et al. An evaluation of tyramide signal amplification
and archived fixed and frozen tissue in microarray gene expression analysis. Nucleic Acids Res
2002;30:E4.
41. Wiedorn KH, Kuhl H, Galle J et al. Comparison of in situ hybridization, direct and indirect in situ
PCR as well as tyramide signal amplification for the detection of HPV. Histochem Cell Biol 1999;111:
89–95.
42. Qian X, Bauer RA, Xu HS et al. In situ hybridization detection of calcitonin mRNA in routinely fixed,
paraffin-embedded tissue sections: a comparison of different types of probes combined with tyramide
signal amplification. Appl Immunohistochem Mol Morphol 2001;9:61–69.
43. Yu J, Othman MI, Farjo R et al. Evaluation and optimization of procedures for target labeling and
hybridization of cDNA microarrays. Mol Vis 2002;8:130–137.
44. Gallardo TD, Hammer RE, Garry DJ. RNA amplification and transcriptional profiling for analysis of
stem cell populations. Genesis 2003;37:57–63.
45. Kenzelmann M, Klaren R, Hergenhahn M et al. High-accuracy amplification of nanogram total RNA
amounts for gene profiling. Genomics 2004;83:550–558.
Chapter 1 / Microarrays and Oncology 17

46. Liu CL, Schreiber SL, Bernstein BE. Development and validation of a T7-based linear amplification
for genomic DNA. BMC Genomics 2003;4:19.
47. Park PJ, Cao YA, Lee SY et al. Current issues for DNA microarrays: platform comparison, double
linear amplification, and universal RNA reference. J Biotechnol 2004;112:225–245.
48. Schlingemann J, Thuerigen O, Ittrich C et al. Effective transcriptome amplification for expression
profiling on sense-oriented oligonucleotide microarrays. Nucleic Acids Res 2005;33:e29.
49. Shearstone JR, Allaire NE, Campos-Rivera J et al. Accurate and precise transcriptional profiles from
50 pg of total RNA or 100 flow-sorted primary lymphocytes. Genomics 2006;88:111–121.
50. Xu T, Shu CT, Purdom E et al. Microarray analysis reveals differences in gene expression of circulat-
ing CD8(+) T cells in melanoma patients and healthy donors. Cancer Res 2004;64:3661–3667.
51. Brandt SP. Microgenomics: gene expression analysis at the tissue-specific and single-cell levels. J Exp
Bot 2005;56:495–505.
52. Kurimoto K, Yabuta Y, Ohinata Y et al. An improved single–cell cDNA amplification method for
efficient high-density oligonucleotide microarray analysis. Nucleic Acids Res 2006;34:e42.
53. Montrose MH. The future of GI and liver research: editorial perspectives: I. Visions of epithelial
research. Am J Physiol Gastrointest Liver Physiol 2003;284:G547–550.
54. Nygaard V, Hovig E. Options available for profiling small samples: a review of sample amplification
technology when combined with microarray profiling. Nucleic Acids Res 2006;34:996–1014.
55. Osawa M, Egawa G, Mak SS et al. Molecular characterization of melanocyte stem cells in their niche.
Development 2005;132:5589–5599.
56. Ramos CA, Bowman TA, Boles NC et al. Evidence for diversity in transcriptional profiles of single
hematopoietic stem cells. PLoS Genet 2006;2:e159.
57. Tietjen I, Rihel J, Dulac CG. Single-cell transcriptional profiles and spatial patterning of the mam-
malian olfactory epithelium. Int J Dev Biol 2005;49:201–207.
58. Tietjen I, Rihel JM, Cao Y et al. Single-cell transcriptional analysis of neuronal progenitors. Neuron
2003;38:161–175.
59. Winegarden N. Microarrays. In: Webster J (ed.). Encyclopedia of Medical Devices and Instrumenta-
tion. 2 nd ed. Hoboken, NJ: John Wiley & Sons; 2006:361–371.
60. Godfrey TE, Kim SH, Chavira M et al. Quantitative mRNA expression analysis from formalin-fixed,
paraffin-embedded tissues using 5’ nuclease quantitative reverse transcription–polymerase chain re-
action. J Mol Diagn 2000;2:84–91.
61. Scicchitano MS, Dalmas DA, Bertiaux MA et al. Preliminary comparison of quantity, quality, and mi-
croarray performance of RNA extracted from formalin-fixed, paraffin-embedded, and unfixed frozen
tissue samples. J Histochem Cytochem 2006;54:1229–1237.
62. Kim M, Lee S, Yang SK et al. Differential expression in histologically normal crypts of ulcerative
colitis suggests primary crypt disorder. Oncol Rep 2006;16:663–670.
63. Chung CH, Parker JS, Ely K et al. Gene expression profiles identify epithelial-to-mesenchymal tran-
sition and activation of nuclear factor–␬b signaling as characteristics of a high-risk head and neck
squamous cell carcinoma. Cancer Res 2006;66:8210–8218.
64. Fan JB, Yeakley JM, Bibikova M et al. A versatile assay for high-throughput gene expression profiling
on universal array matrices. Genome Res 2004;14:878–885.
65. Bibikova M, Talantov D, Chudin E et al. Quantitative gene expression profiling in formalin-fixed,
paraffin-embedded tissues using universal bead arrays. Am J Pathol 2004;165:1799–1807.
66. Dietel M, Sers C. Personalized medicine and development of targeted therapies: The upcoming chal-
lenge for diagnostic molecular pathology: a review. Virchows Arch 2006;448:744–755.
67. Pusztai L, Mazouni C, Anderson K et al. Molecular classification of breast cancer: limitations and
potential. Oncologist 2006;11:868–877.
68. Nielsen TO. Microarray analysis of sarcomas. Adv Anat Pathol 2006;13:166–173.
69. Gerhold DL, Jensen RV, Gullans SR. Better therapeutics through microarrays. Nat Genet 2002;32
Suppl:547–551.
70. Jain KK. Role of nanobiotechnology in developing personalized medicine for cancer. Technol Cancer
Res Treat 2005;4:645–650.
18 Part I / Genomic Experimental Approaches in Oncology

71. Jain KK. Personalized medicine. Curr Opin Mol Ther 2002;4:548–558.
72. Nawrocki S, Skacel T, Brodowicz T. From microarrays to new therapeutic approaches in bladder
cancer. Pharmacogenomics 2003;4:179–189.
73. Bissonnette L, Bergeron MG. Next revolution in the molecular theranostics of infectious diseases:
microfabricated systems for personalized medicine. Expert Rev Mol Diagn 2006;6:433–450.
74. Picard FJ, Bergeron MG. Rapid molecular theranostics in infectious diseases. Drug Discov Today
2002;7:1092–1101.
75. Sablon E, Shapiro F, Zoulim F. Early detection of hepatitis B drug resistance: implications for patient
management. Expert Rev Mol Diagn 2003;3:535–547.
76. Deeken JF, Figg WD, Bates SE et al. Toward individualized treatment: prediction of anticancer drug
disposition and toxicity with pharmacogenetics. Anticancer Drugs 2007;18:111–126.
77. Golub TR, Slonim DK, Tamayo P et al. Molecular classification of cancer: class discovery and class
prediction by gene expression monitoring. Science 1999;286:531–537.
78. Alizadeh AA, Eisen MB, Davis RE et al. Distinct types of diffuse large B-cell lymphoma identified
by gene expression profiling. Nature 2000;403:503–511.
79. van’t Veer LJ, Dai H, van de Vijver MJ et al. Expression profiling predicts outcome in breast cancer.
Breast Cancer Res 2003;5:57–58.
80. van de Vijver MJ, He YD, van’t Veer LJ et al. A gene-expression signature as a predictor of survival
in breast cancer. N Engl J Med 2002;347:1999–2009.
81. Aronovitz LG, Redican-Bigott G, Hormozi S et al. New drug development: science, business, regula-
tory, and intellectual property issues cited as hampering drug development efforts. In: Office USGA,
2006;1–47.
82. Russo E. Chip critics countered. The Scientist 2003;17:30.
 2 Cell-Based Models to Identify
Genetic Variants Contributing
to Anticancer Drug Response

M. Eileen Dolan, PhD, and

Howard McLeod, PharmD
CONTENTS
Introduction
Us e of Lymphoblas toid Cell Lines
for Pharmacogenomic Dis covery
Relevance to Drug Toxicities in Humans
Acknowledgements
References

S UMMARY
A goal of pharmacogenetics is the design of individualized therapy based on
a patient’s genomic sequence to maximize response and minimize toxicity. Cur-
rent chemotherapy is associated with serious, at times life-threatening, toxicity.
Identifying heritable genetic variants responsible for chemotherapeutic toxicities is
challenging because of its multigenic nature and the difficulty of studying chemother-
apeutic toxicity using traditional genetic screening strategies, such as unaffected
families.
Cell-based models using lymphoblastoid cell lines from individuals comprising
large pedigrees are a means to overcome these limitations. Several investigators
within the NIH Pharmacogenetics Research Network have used cell-based models,
genome-wide mapping strategies including linkage, and association studies to search
for genes responsible for toxicities and for responses associated with chemotherapy.

From: Cancer Drug Discovery and Development: Genomics and Pharmacogenomics

in Anticancer Drug Development and Clinical Response
Edited by: F. Innocenti, DOI: 10.1007/978-1-60327-088-5 2, 
c Humana Press, Totowa, NJ

19
20 Part I / Genomic Experimental Approaches in Oncology

By comparing the variations in chemotherapy-induced cytotoxicity in cell lines from

within and among families, the extent to which genetics contributes to human variation
in susceptibility to such cytotoxicity can be evaluated.
Centre d’ Etude du Polymorphisme Humain (CEPH) pedigrees have genetic infor-
mation in the form of SNPs and microsatellites in the public domain, which allows
for linkage analysis to identify genetic regions harboring genes contributing to cyto-
toxicity. Another cell-based model utilizes International HapMap cell lines to identify
genetic variants and gene expression associated with chemotherapeutic agent–induced
cytotoxicity. With extensive genotypic information and gene expression levels avail-
able for these cell lines, associations can be made between genotype and cytotoxicity,
and between genotype and gene expression; and correlations can be made between
gene expression and cytotoxicity. These studies enable the determination of genetic
variants contributing to sensitivity to cytotoxicity through the modulation of gene ex-
pression.
There are a number of advantages to the use of CEPH cell lines for evaluating toxici-
ties associated with chemotherapy. Use of cell lines avoids giving toxic chemotherapy
to healthy volunteers; cells can be grown under identical conditions; and extensive
genotype data for CEPH cell lines are publicly available, allowing investigators to
utilize classical mapping techniques (e.g., linkage analysis, association) to study the
genetic influence of human variation in drug sensitivity.
Despite the advantages of this cell-based system, limitations include that LCLs rep-
resent only one specific tissue type from unaffected individuals, which may or may
not be appropriate when assessing a phenotypic effect that occurs in a different tis-
sue (e.g., peripheral neuropathy, diarrhea). Another limitation is that most cell lines
do not express cytochrome P450 genes and thus the contribution of metabolism and
pharmacokinetics on drug affect is not taken into account. In spite of these limitations,
these cell lines provide an opportunity for initial identification of important genes and
variants that can be validated in further studies. This comprehensive approach can
be broadly applied to other phenotypes that are measurable in cell lines, and for the
first time allows the application of familial genetic analysis in the search for genes
important in chemotherapeutic response and toxicity.

Key Words: Pharmacogenetics; cellular phenotype; CEPH; genome scan; familial

genetics

1. INTRODUCTION
Most chemotherapeutic drugs exhibit serious toxicity; hence elucidating the genetic
variants that alter their pharmacodynamic effects is an important but challenging project.
Challenges include our inability to do family studies evaluating the effects of chemother-
apy on individuals without cancer and the multigenic nature of drug response. Therefore,
several cell-based models to identify genetic variants important in chemotherapeutic-
induced toxicity have been developed. The models employ EBV-transformed lym-
phoblastoid cell lines from related healthy Caucasians of European descent (CEPH) and
HapMap trios to evaluate chemotherapeutic-induced cytotoxicity and/or apoptosis.
Chapter 2 / Pharmacogenetic Models of Drug Cytotoxicity 21

1.1. Interpatient Variability in Response to Drugs and Toxicity

The main prognostic factors in cancer are age, tumor size, stage of disease, histo-
logic type of the tumor, pathological grade, and in some cases hormone-receptor status.
There are suggestions that the same genes that are implicated in cancer risk may also
be involved in a person’s propensity to carcinogenic exposure and/or to modulation of
therapeutic outcome ( 1). Therefore, constructing genetic profiles that can be used to
individualize therapy may also increase our understanding of cancer risk genes and may
be applied to cancer development and prediction of outcome.
A large number of genes are likely to influence the response of a tumor to an individ-
ual chemotherapeutic agent. However, studying the genetic contribution to chemosen-
sitivity in the clinic is challenging because drug responses reflect not only properties
intrinsic to the target cell but also host metabolic properties. Candidate gene approaches
have had some success in identifying genes important in response to chemotherapy used
to treat cancer; however, there has been limited success in identifying genetic factors
that predict patients at risk for chemotherapeutic induced toxicities (2). The focus of this
chapter is to describe the use of classical genetic approaches in the context of identifying
genes important in response/toxicities associated with chemotherapy.

1.2. Challenge of Studying Anticancer Agents

There are a number of limitations specific to the study of the pharmacogenetics of an-
ticancer agents. Anticancer agents affect tumor and non-tumor cells. They have a narrow
therapeutic index and a wide toxicity profile with the extreme end of the spectrum result-
ing in death. Thus, it is particularly important to identify genetic polymorphisms asso-
ciated with toxicities related to these medications. Identifying heritable polymorphisms
associated with chemotherapy cannot be performed in related, healthy individuals, thus
forcing researchers to consider alternatives to current approaches.
Until recently, only a few studies have utilized human lymphocytes or Epstein-Barr
virus (EBV)–transformed lymphoblastoid cell lines (LCLs) derived from the CEPH to
identify heritable traits related to drug sensitivity or other drug-induced pharmacody-
namic parameters ( 3, 4). However, these studies are providing evidence for the broad
applicability of this approach to pharmacogenomic research.

1.3. Advantages and Limitations of LCLs for Evaluating

the Pharmacogenomics of Anticancer Agents
The greatest advantages of the use of CEPH cell lines for studying human variation
in response to drugs is that they provide a system to perform phenotypic studies of anti-
cancer agents that would be considered unsafe and unethical in healthy volunteers. The
confounders of in vivo studies (metabolism, pharmacokinetics, clearance) are ignored,
which can be considered both an advantage and disadvantage. CEPH cells can be grown
under identical conditions; thus the phenotype to be tested is in a well-controlled, iso-
lated system without the confounders present in vivo.
In contrast, it is questionable how well the results apply to a drug that will be ad-
ministered to a patient and thus will undergo metabolism. If one is identifying pharma-
codynamic markers of drug sensitivity, the model still holds much promise. A major
22 Part I / Genomic Experimental Approaches in Oncology

advantage of the use of CEPH cell lines is the extensive genotypic data that is pub-
licly available (www.cephb.fr/cephdb/), allowing the use of classical mapping techniques
such as linkage analysis to define regions of the genome that co-segregate with the phe-
notype without requiring extensive genotyping. In particular, the International HapMap
Project used cell lines from these same families with more than 6 million SNPs typed in
3–6 members of CEPH families for association studies that can be used to determine if
SNPs or haplotypes containing multiple SNPs explain the variation in drug sensitivity
(www.hapmap.org/abouthapmap.html).
One limitation of LCLs or other human-derived cell lines is that they represent only
one specific tissue type, which may or may not be appropriate when assessing a toxic
effect that occurs in a different tissue (e.g., peripheral neuropathy, diarrhea). Another
limitation is that EBV transformation can introduce phenotypic changes or expression
changes that may affect the phenotype of interest. One must consider different ethnic
populations when evaluating phenotype–genotype relationships, and the large pedigrees
currently available are Caucasian. Despite these limitations, the cell model provides an
opportunity for initial identification of important genes and variants that can be validated
in further studies.

2. USE OF LYMPHOBLASTOID CELL LINES

FOR PHARMACOGENOMIC DISCOVERY
The CEPH genotype consists of 61 families with large sibships that have served as a
reference panel for the construction of human genetic maps by the international scientific
community (5). Until recently, very few pharmacogenetic studies utilized these cell lines;
however, researchers are beginning to appreciate the value of these cell lines in studying
genetic determinants of chemotherapy, radiation, and gene expression.

2.1. Generation of EBV-Transformed Lymphoblastoid Cell Lines

Infection by EBV in vitro easily transforms resting B-cells from human peripheral
blood into actively proliferating lymphoblastoid cell lines (6). The percentage of B-cells
susceptible to transformation has been estimated to be between 10% and 100% (7,8).
EBV proteins, such as EBV nuclear antigens and latent membrane proteins, participate in
the transformation of resting B-cells to actively proliferating cells, which is the first step
in immortalization and tumorigenesis (6). Recent studies have shown that LCLs can be
in a pre- or post-immortal state, and that cellular events and host genetic factors probably
cooperate with EBV during transition from the pre-immortal to the post-immortal state
(6). LCLs have been used for various purposes–as sources of DNA and cells to study var-
ious genetic disorders; for studies of immunology and biology; and as described below
for studies of pharmacogenetics.

2.2. Use of Lymphoblastoid Cells from Targeted Populations

LCLs derived from targeted populations have been utilized to evaluate genetic sus-
ceptibility to radiation sensitivity (9), sensitivity to mitomycin C in patients with Wilms’
tumor (10), and camptothecin-induced apoptosis (11). These cell lines have been useful
for studying bleomycin-induced chromatid breaks and oxidant-induced apoptosis (12)
Chapter 2 / Pharmacogenetic Models of Drug Cytotoxicity 23

the genetic contribution to variation in expression that influences response to ionizing

radiation (IR) (13) variation in sodium–lithium countertransport (14); and variation in
global gene expression (15,16).

2.3. Measures of Anticancer Agent Response

Until recently, CEPH LCLs have represented an untapped resource for identifying
heritable traits related to anticancer drug sensitivity. Cytotoxicity can be measured as
cell growth inhibition using a short-term assay using a colorimetric such as alamarBlue,
propidium iodide or MTT, to name a few (Fig. 1) (17,18,19). These assays focus on the
ability of mitochondrial reduction of dye, rather than a direct measure of cytotoxicity.
However, the dye-reduction approaches are very amenable to high-throughput assays.
In addition, apoptosis can be measured using ELISA, flow cytometry, or similar ap-
proaches. Other phenotypes might include phosphorylation of enzymes, transport of
known substrates, expression levels of genes, enzyme activity, protein levels or drug-
induced DNA breaks (13,14,17,19,20,21).

2.4. Genomic Information on HapMap Cell Lines

The most apparent genetic difference between individuals is variation in their
DNA sequence. This includes insertions and deletions, as well as repetitive ele-
ments of variable copy number whose span may range from a few bases to many
kilobases. Perhaps most importantly, there are about 10 million single nucleotide poly-
morphisms. The CEPH pedigrees have served as a reference panel for the construc-
tion of human genetic maps; thus information has accumulated in these families on
thousands of genetic markers in the form of restriction fragment length polymor-
phisms, minisatellites, microsatellites, SNPs (www.cephb.fr/cephdb/) and haplotype
data (www.hapmap.org/abouthapmap.html), much of which isavailable in the public do-
main (22). The genetic maps can be combined with the phenotype data (e.g., cytotoxicity
data) described above for genotype–phenotype studies. The large pedigrees allow for
linkage analysis, and the HapMap trios can be used for association studies.

2.4.1. H ERITABILITY S TUDIES

Our laboratories were the first to use CEPH family cell lines to demonstrate that a
significant genetic component contributed to susceptibility to the cytotoxic effects of
cisplatin, 5-FU and docetaxel (17,18). Dolan et al. estimated the heritability for suscep-
tibility to cisplatin-induced cytotoxicity to be approximately 0.47; therefore sensitivity to
the cytotoxic effects of cisplatin is under appreciable genetic influence. Linkage analysis
was performed and the strongest signal (lod 2.16, empirical p-value 0.0005) was found
on chromosome 1 at 44 cM (18).
Watters et al. found that narrow sense heritability estimates for docetaxel, varying by
dose, ranged from 0.21 to 0.70, while the heritability estimates for 5-fluorouracil ranged
from 0.26 to 0.65 (17). Thus the heritability of cytotoxicity observed in this system is
similar to or greater than that of several common phenotypes. The statistically significant
effects of docetaxel were mapped to chromosomes 5q11-21 and 9q13-q22 (Fig. 2) (17).
The effects of 5-fluorouracil were mapped to chromosome 9q13-q22 (17).
24
Fig. 1. Docetaxel concentration–viability curves for sensitive (A) and resistant (B) CEPH cell lines.
25
Fig. 2. Quantitative trait locus on chromosome 9 for docetaxel. Data derived from Watters et al. (17).
26 Part I / Genomic Experimental Approaches in Oncology

Susceptibility to cytotoxicity of chemotherapy is likely due to multiple loci with

low locus-specific heritability contributing to the trait. These data show the power of
using large pedigrees that have been extensively genotyped for evaluating the genetic
contribution to sensitivity to cell growth inhibition by anticancer agents. Correa et al.
(21) determined that heritable variation contributed to radiation response by evaluating
a group of well-characterized genes that induced at least a two-fold change following
radiation in cell lines derived from unrelated CEPH individuals compared to those from
monozygotic (MZ) twins. The most correlated genes were shown to be target genes of
p53, suggesting co-regulation.

2.4.2. L INKAGE A NALYSIS

In trying to link genetic variation to disease susceptibility at the level of the whole
genome, investigators generally use the approach of linkage analysis. This method uses
families that have more than one member who has the disease of interest, and defines
regions of the genome that segregate with disease through the different generations of
the pedigree. Because cell lines are derived from individuals within large CEPH pedi-
grees, the same approach can be used to evaluate susceptibility to drug-induced effects
such as cytotoxicity or apoptosis, or a biochemical effect measured within the cell lines.
Linkage analysis provides broad regions of the genome that harbor disease resistance or
susceptibility alleles which may contain up to hundreds of different genes. Thus, a major
challenge in the use of genetic linkage analysis is the relatively poor resolution of link-
age mapping. The selection of candidate genes is usually based on reported functional
information that may be related to the phenotype. The function of a vast proportion of
human genes remains elusive and assumptions are made about which genes are the most
important.

2.4.3. W HOLE G ENOME A SSOCIATION

Association studies using biologically plausible candidate genes have shown variable
success. They are comparatively more powerful than linkage tests for strong association,
weak penetrance models. There are a number of whole genome association studies that
can be performed using anticancer agent cytotoxicity as a phenotype, each study with
advantages and disadvantages. This approach is well supported by the recent data show-
ing expression of a considerable number of genes is directly controlled by cis-acting
elements in their genotype (23,24,25). Association studies with dense SNP maps, such as
those within HapMap samples, can identify susceptibility loci or other determinants for
some complex traits (e.g., expression levels, drug sensitivity). If significant association
results are found that overlap with regions that had evidence of linkage, this provides
further evidence that the SNP is associated with the phenotype.
Genetic association studies may be either family based, or population based (26,27).
Family based studies can use transmission disequilibrium tests (TDT), a test that de-
tects association in the presence of linkage. Population-based tests use analysis of vari-
ance (ANOVA) among multiple subgroups. The TDT was first proposed by Spielman
and Ewens ( 28). The basic premise of the TDT is that certain alleles at a locus are
disproportionately transmitted from parents to an affected offspring; therefore the ba-
sic sampling unit for the TDT is a nuclear family with at least one affected offspring
Chapter 2 / Pharmacogenetic Models of Drug Cytotoxicity 27

(i.e., father, mother, andaffected offspring). For the association studies of the quantita-
tive traits in complex structured families such as the three-generation CEPH pedigrees,
mixed models with variance–covariance structures are implemented in the statistic tools
of QTDT, FBAT (family-based association test), and SOLAR (sequential oligogenic
linkage analysis).
With the completion of the Human Genome Project, there is a tremendous amount
of information readily available about the occurrence and frequency of polymorphisms
throughout the human genome. Ideally, SNPs that have a high degree of polymorphism
(i.e., minor allele frequency > 10%) in the population and are exonic resulting in non-
synonymous coding changes, or are located in known regulatory regions, are considered
to be the “smoking gun.” Most scientists would argue that synonymous SNPs that do not
produce altered coding sequences, and therefore are not expected to change the function
of the protein in which they occur, would not be worth follow-up studies. However,
recent results demonstrate that a synonymous SNP in P-glycoprotein with similar mRNA
and protein levels resulted in altered conformation of the protein ( 29). This SNP was
associated with altered drug and inhibitor function, illustrating the importance of an
unbiased approach.

2.4.4. I NFERENCE S TUDIES

A powerful means to identify genetic variants associated with any phenotype includ-
ing cytotoxicity is to perform association studies with denser markers. The HapMap
trios (parents and grandparents of some large pedigrees) have millions of SNPs, but the
children within those large pedigrees have microsatellites and some SNPs. One way
to obtain denser markers in the offspring of the HapMap trios is to use a genotype
inference method. This method combines sparse marker data from the children with
high-resolution SNP genotypes from the parents to infer genotypes for the offspring.
Burdick et al. inferred over 53 million SNP genotypes for 78 children in the CEPH fam-
ilies (30). This ultimately leads to high-density genotypes in large families and can be
used in mapping studies with cytotoxicity or apoptosis as the phenotype.

2.5. Expression Studies

Gene expression, a determinant of a cell’s characteristics, is another phenotype that
can be studied using lymphoblastoid cell lines. Studies have shown that gene expression
levels in humans differ not only among cell types within an individual, but also among
individuals (16,31). As a result, there have been several recent studies that have identified
polymorphic genetic variants that influence gene expression levels (15,25,32).
Expression data on 233 CEPH cell lines using the Human Genome Focus R
Ar-
ray representing over 8,500 verified human sequences from the NCBI RefSeq database
is publicly available (www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE1485) and can
be used for correlation and regression analysis with cytotoxicity and apoptosis. One
use of this data to obtain information on gene expression that is important in drug
sensitivity/resistance is to use vitro phenotype (e.g., cytotoxicity, apoptosis) as a
continuous variable and use correlation and/or regression analyses to determine the
effect of gene expression on phenotype for each probe set (4,276 probe sets after fil-
tering). One limitation is that the Focus array does not represent the entire genome
28 Part I / Genomic Experimental Approaches in Oncology

(www.affymetrix.com/products/arrays/specific/focus.affx) and there is a 3 bias in probe

design. There has also been data generated on 90 CEPH and 90 Yoruban cell lines
at the baseline using the Human Exon 1.0 ST GeneChip R
array. Containing roughly
1.4 million probe sets designed to represent all known exon regions within the human
genome (Build 34), this exon chip overcomes the two limitations of the earlier Focus R

Array (8,500 probesets and the 3 bias in probe design). This array allows exon-level
profiling and the ability to interrogate each exon across the genome on a single chip.
Gene expression data for the 180 HapMap lines can be correlated to phenotypes (IC50
for cytotoxicity and apoptosis).
Differences among individuals has led to several studies demonstrating expression
differences among populations. Recently, ethnic differences in gene expression as a
complex quantitative phenotype and its regulation by polymorphic genetic variants have
not been investigated comprehensively. Spielman et al. utilized a subset of human genes
(∼4,200 expressed in LCLs) with samples derived from unrelated individuals from the
CEU (CEPH from Utah, USA) and the CHB/JPT (Han Chinese in Beijing and Japanese
in Tokyo) samples to demonstrate that cis-acting regulators may account for some of the
differences in gene expression between the populations (33). Using the same platform,
Storey et al. showed that 17% of genes are differentially expressed between CEU and
YRI (Yoruban from Ibadan, Nigeria) in a limited set of 16 samples (34).

2.6. Integrating Different Approaches

Thus, the use of complementary approaches including heritability analysis, linkage
analysis, expression studies, and association studies can be used to identify and charac-
terize novel genes important in sensitivity and resistance to chemotherapy. Ultimately,
one must consider validating the genes identified in an appropriate cell or tissue (Fig. 3).
Among the advantages of using linkage analysis and association in cell-based models
to identify genetic variants important in sensitivity to drugs are: (1) The use of cell lines
from pedigrees allows genetic strategies such as linkage analysis to be used to identify
regions of the genome important in sensitivity to chemotherapy without any assumptions

Fig. 3. Identifying genes/genetic variants contributing to susceptibility to anticancer agent

cytotoxicity.
Chapter 2 / Pharmacogenetic Models of Drug Cytotoxicity 29

about the nature of the genes likely to be involved; (2) if heritability is low or linkage
does not result in high lod scores, global association and expression studies can be per-
formed on the CEPH and Yoruban HapMap samples; (3) chemotherapy does not have to
be given to non-affected family members for a genetic study; (4) the phenotypic effects
(e.g., cytotoxicity, apoptosis) are protected from confounding variables that exist in vivo;
(5) the possibility of identifying genes that were previously unknown or unrecognized is
realized.

3. RELEVANCE TO DRUG TOXICITIES IN HUMANS

While the use of immortalized cell lines has the obvious benefits of convenience for
high-throughput analysis, ability to replicate, and avoiding exposure of toxic molecules
to patients or volunteers, cell lines are not human patients. The cells only represent one
of many cellular constituents of a human (B-lymphocytes), and one that is not a major
concern for life-threatening toxicity. This cell lineage may not adequately represent GI
tract, nerves, and the complete bone marrow milieu. The cell lines do not have a liver to
evaluate pharmacokinetics nor an immune system or other dynamic processes that are
important for the in vivo effects of many drugs. Indeed, most CYPs and transporters are
down-regulated in CEPH cell lines and therefore will not be evaluable in the toxicity
screens. However, the cell lines do offer a cell autonomous model of drug effect, in that
any mechanism that is common to multiple cell lineage is likely to be represented in the
CEPH cells.
The argument against any given cell lineage can be made for hepatocytes and/or hu-
man cancer cell lines, both of which are a major part of the drug development process.
There is also an emphasis on the pharmacodynamic mechanisms of the drug effect, be-
cause of the absence of CYP/transport members. Because a drug’s pharmacokinetics are
usually adequately characterized as part of the preclinical and early clinical develop-
ment, pharmacodynamics remains an area of needed research.Specific steps have been
taken to try to understand how relevant the CEPH cell lines are to other cancer cell line
systems.
The observed CEPH population mean IC50 for both docetaxel and 5-fluorouracil
was similar to IC50 values observed across the NCI60 panel of human tumor cell
lines (http://dtp.nci.nih.gov) ( 17). In addition, docetaxel- and 5-fluorouracil-induced
cell death is mediated by caspase-3 cleavage, similar to that observed in tumor cells
( 17). These data are encouraging for the use of CEPH pedigrees as a discovery tool.
However, the ultimate proof of the value of the cell-based models will be the hu-
man validation of markers derived from this process. These studies lie ahead and
will help position cell-based models in their correct place in the drug development
process.

ACKNOWLEDGEMENTS
The authors are supported by the Pharmacogenetics of Anticancer Agents Re-
search (PAAR) and the Comprehensive Research on Expressed Alleles in Therapeu-
tic Evaluation (CREATE) groups within the NIH Pharmacogenetics Research Network
(GM63340, GM61393).
30 Part I / Genomic Experimental Approaches in Oncology

REFERENCES
1. Spitz MR, Wu X, Mills G. Integrative epidemiology: from risk assessment to outcome prediction.
J Clin Oncol 2005;23:267–275.
2. Walgren RA, Meucci MA, McLeod HL. Pharmacogenomic discovery approaches: will the real genes
please stand up? J Clin Oncol 2005;23:7342–7349.
3. Shukla SJ, Dolan ME. Use of CEPH and non-CEPH lymphoblast cell lines in pharmacogenetic stud-
ies. Pharmacogenomics 2005;6:303–310.
4. Meucci MA, Marsh S, Watters JW et al. CEPH individuals are representative of the European Ameri-
can population: implications for pharmacogenetics. Pharmacogenomics 2005;6:59–63.
5. Dausset J, Cann H, Cohen D et al. Centre d’Etude du Polymorphisme Humain (CEPH): collaborative
genetic mapping of the human genome. Genomics 1990;6:575–577.
6. Sugimoto M, Tahara H, Ide T et al. Steps involved in immortalization and tumorigenesis in human
B-lymphoblastoid cell lines transformed by Epstein–Barr virus. Cancer Res 2004;64:3361–3364.
7. Miller G, Lipman M. Release of infectious Epstein–Barr virus by transformed marmoset leukocytes.
Proc Natl Acad Sci USA 1973;70:190–194.
8. Henle W, Diehl V, Kohn G et al. Herpes-type virus and chromosome marker in normal leukocytes
after growth with irradiated Burkitt cells. Science 1967;157:1064–1065.
9. Gipps EM, Kidson C. Cellular radiosensitivity: expression of an MS susceptibility gene? Neurology
1984;34:808–811.
10. Imray FP, Smith PJ, Relf W et al. Wilms’ tumour: association with cellular sensitivity to mitomycin
C in patients and first-degree relatives. Lancet 1984;1:1148–1151.
11. Poot M, Gollahon KA, Rabinovitch PS. Werner syndrome lymphoblastoid cells are sensitive to
camptothecin-induced apoptosis in S-phase. Hum Genet 1999;104:10–14.
12. Cloos J, Reid CB, van der Sterre ML et al. A comparison of bleomycin-induced damage in lympho-
cytes and primary oral fibroblasts and keratinocytes in 30 subjects. Mutagenesis 1999;14:87–93.
13. Jen KY, Cheung VG. Transcriptional response of lymphoblastoid cells to ionizing radiation. Genome
Res 2003;13:2092–2100.
14. Schork NJ, Gardner JP, Zhang L et al. Genomic association/linkage of sodium lithium counter-
transport in CEPH pedigrees. Hypertension 2002;40:619–628.
15. Morley M, Molony CM, Weber TM et al. Genetic analysis of genome-wide variation in human gene
expression. Nature 2004;430:743–747.
16. Cheung VG, Conlin LK, Weber TM et al. Natural variation in human gene expression assessed in
lymphoblastoid cells. Nat Genet 2003;33:422–425.
17. Watters JW, Kraja A, Meucci MA et al. Genome-wide discovery of loci influencing chemotherapy
cytotoxicity. Proc Natl Acad Sci USA 2004;101:11809–11814.
18. Dolan ME, Newbold KG, Nagasubramanian R et al. Heritability and linkage analysis of sensitivity to
cisplatin-induced cytotoxicity. Cancer Res 2004;64:4353–4356.
19. Huang RS, Kistner EO, Bleibel WK et al. Effect of population and gender on chemotherapeutic agent-
induced cytotoxicity. Mol Cancer Ther 2007;6:31–36.
20. Cheung VG, Spielman RS, Ewens KG et al. Mapping determinants of human gene expression by
regional and genome-wide association. Nature 2005;437:1365–1369.
21. Correa CR, Cheung VG. Genetic variation in radiation-induced expression phenotypes. Am J Hum
Genet 2004;75:885–890.
22. The International HapMap Consortium. A haplotype map of the human genome. Nature
2005;437:1299–1320.
23. Brem RB, Yvert G, Clinton R et al. Genetic dissection of transcriptional regulation in budding yeast.
Science 2002;296:752–755.
24. Lo HS, Wang Z, Hu Y et al. Allelic variation in gene expression is common in the human genome.
Genome Res 2003;13:1855–1862.
25. Cheung VG. Polymorphic landscape of the human genome. Eur J Hum Genet 2005;13:133–135.
26. Lewis CM. Genetic association studies: design, analysis and interpretation. Brief Bioinform
2002;3:146–153.
Chapter 2 / Pharmacogenetic Models of Drug Cytotoxicity 31

27. Cardon LR, Bell JI. Association study designs for complex diseases. Nat Rev Genet 2001;2:91–99.
28. Spielman RS, Ewens WJ. The TDT and other family-based tests for linkage disequilibrium and asso-
ciation. Am J Hum Genet 1996;59:983–989.
29. Kimchi-Sarfaty C, Oh JM, Kim IW et al. A “silent” polymorphism in the MDR1 gene changes sub-
strate specificity. Science 2007;315:525–528.
30. Burdick JT, Chen WM, Abecasis GR et al. In silico method for inferring genotypes in pedigrees. Nat
Genet 2006;38:1002–1004.
31. Schadt EE, Monks SA, Drake TA et al. Genetics of gene expression surveyed in maize, mouse and
man. Nature 2003;422:297–302.
32. Stranger BE, Forrest MS, Clark AG et al. Genome–wide associations of gene expression variation in
humans. PLoS Genet 2005;1:e78.
33. Spielman RS, Bastone LA, Burdick JT et al. Common genetic variants account for differences in gene
expression among ethnic groups. Nat Genet 2007;39:226–231.
34. Storey JD, Madeoy J, Strout JL et al. Gene-expression variation within and among human populations.
Am J Hum Genet 2007;80:502–509.
 3 Proteomic Analysis in Cancer
Patients

Yasuhiro Kuramitsu, MD, PhD, and

Kazuyuki Nakamura, MD, PhD
CONTENTS
Introduction
Materials
Methods
Res ults from Our Study
Res ults from the Literature
Potential Application in Oncology
References

S UMMARY
The term proteome means the total protein complement of a genome, and pro-
teomics means the analysis for proteome. The combination of two-dimensional gel
electrophoresis (2-DE) and mass spectrometry (MS) is a proteomic method of high-
throughput analysis of protein expression. By using this 2-DE and MS, proteomic
studies have identified many proteins that may be involved in the pathogenic mecha-
nism of cancers. These studies analyzed cancer cell lines, as well as cancer tissues or
serum from patients.
In the present study, we analyzed proteome in hepatocellular carcinoma (HCC),
esophageal cancer, and pancreatic cancer tissues. We identified many proteins whose
expression in cancer tissues was different from corresponding non-cancerous tissues
by using 2-DE and MS. Furthermore, we identified some auto-antibodies reacting
to proteins in HCC cancer tissues. In this chapter, we will describe the method, our
experimental result, and reports from other researchers about proteomic analysis in
cancer patients.
From: Cancer Drug Discovery and Development: Genomics and Pharmacogenomics
in Anticancer Drug Development and Clinical Response
Edited by: F. Innocenti, DOI: 10.1007/978-1-60327-088-5 3, 
c Humana Press, Totowa, NJ

33
Random documents with unrelated
content Scribd suggests to you:
my first impression of Crystal Grey was that she was something
between a proud goddess and a sweet angel: the former aspect
slumbering in her coal-black eyes and wavy black hair, the latter
wide awake upon her lovely face and perfect form, clad, as all angels
are, in white. The mysterious eyes of deep night, and the hair of
deeper night contrasted strangely with the innocent wistfulness of
the rest of the face. If the eyes were those of some severe sage,
made young again by a draught of his wondrous elixir, the sweet
girlish lips looked as if they had kissed the early morning dew from a
ripe peach and carried away the freshness of it. I rose from my
wicker chair and stood facing her, with the hammock between us. I
was too dazzled by this sudden apparition of girlish beauty, beyond
my power to describe, to stammer out a single word; and, while I
was trying to begin an apology for my rough appearance in her
garden sanctum, she spoke first.
“Are you the stranger that brought the good news?” she asked, as
she let go the branch and advanced a step towards me.
“I am,” I replied; “but that hardly excuses my trespassing here
perhaps.”
 “SHE SEATED HERSELF SIDEWAYS ON THE HAMMOCK, WHILE I
RESUMED MY WICKER CHAIR, AND TOLD AGAIN THE STORY
WHICH I HAD NARRATED TO HER FATHER.”

She extended her fair, white hand, and, as I took it in my rough

brown one, looking into her eyes the while, a combination of feelings
took possession of me. I can only liken it to the laying of the
foundation-stone of a love which would mount upwards for ever and
ever, like a crystal staircase leading to the far-off heaven of her soul.
Her sweet lips moved, then trembled, but no words came. Only
her eyes spoke unfathomable things, as they burned with feelings,
tender and mysterious; only a sigh escaped her as she turned her
head away.
“What is it affects you so in the news I bring?” I asked.
 “My mother,” she replied; “is she still living? Do you think we can
find her? Oh! tell me the story of your adventures again; perhaps
there was something my father left out.”
She seated herself sideways on the hammock, while I resumed
my wicker chair, and told again the story which I had narrated to her
father the night before.
When I had finished, she said, “May Heaven reward you for all you
went through.”
“I am rewarded already,” I replied.
“You mean that Heaven has rewarded you in advance by giving
you a disposition that gets its happiness from making other people
happy.”
I was ashamed of my poor attempt at a compliment and said, “Yes,
that is what I meant”—though it was nothing of the kind—“only I am
not clever at expressing myself.”
“Do you think that last shot of yours killed Ngaraki?” she asked,
with a gentle concern in her voice, for I had told her more about the
chief than I had to Grey the previous evening.
“I hope not,” I replied; “he was a noble fellow, and had a fine
hatred of the Vile Tohungas, because, as he said, they were the
Destroyers of Women.”
A slight change of expression passed over her face, and with a
quick intuition she said:
“There is something you have kept back. Almost my only memory
besides that of my mother’s face is what I think must have been an
image carved out of marble, all white and beautiful. Did you see any
such thing?”
“No, I did not see what you speak of, but I saw Ngaraki hold up his
arms and gaze—like the Twelfth Tohunga I have told you about—up
through the darkness to something which he knew was far above,
and I judged from the chief’s manner that the object he was
addressing was sublime and beautiful.”
There was a silence, during which Crystal was evidently engaged
in trying to recall more of this earliest memory, while I was
considering whether, if I spoke of Kahikatea’s experience, I should
be breaking my promise to Te Makawawa. At length, proceeding on
the argument that his discovery was independent of the old chief’s
revelations, I concluded that I was on safe ground.
I said, “Now that you have mentioned the matter, I might tell you
that I have a friend who says he has seen that beautiful image in a
high cave in the mountain. How he got there I hardly know, and how
he got out again he does not know himself, but he says he saw the
marble statue of a lovely woman, young—almost a girl. She was
standing near the mouth of the cave with her arms outstretched, as if
to some vision in the western sky, and on her face was stamped a
divine and radiant beauty, while her form, still and cold, yet full of
motion, seemed ready to spring to life at a touch. The prayers of all
women who lift their eyes unto the hills were upon her lips. The
sightless eyes derived their love-light from the longing expressed by
the whole figure yearning forward to some glorious future of our race
when——”
I paused, for while I had been speaking Crystal’s hands had
clasped themselves together in her lap, a rapt look had come to her
eyes, and my thoughts wandered from the statue. However beautiful,
however dazzling it might be, it could not be more so than this girl
before me. Therefore, as I said, my thoughts wandered from the
statue; I paused and she, with a start, turned her eyes upon me with
looks of serious wonder.
“What a symbol of the ideal woman!” she said. “All white—
standing far above the world—waiting, with a prayer upon her lips,
for the dawn of a brighter day. She is the higher self of all women
who wait and pray, and try to be white. What did your friend think?”
“He thought just what you think. Indeed, he even went so far as to
fall madly in love with that ideal woman in his own strange, poetical
style, and he now swears he will find his way to that cave to look
upon her face again. Pygmalion and Galatea make a very pretty
story between them, but don’t you think it’s rather a wild kind of
poetry for a man of the nineteenth century to love a stone?”
She smiled a sweet, sad smile at one of the little leaves overhead,
as it opened and shut its tiny door against the blue. “Surely it is not
the stone your poetical friend has fallen in love with,” she said
presently; “it is the beauty and meaning depicted on that stone—
how, and by whom, is a mystery.”
 The breakfast bell rang vigorously from the verandah, and covered
the silence with which I greeted her last remark. It was not because I
saw any reason for secrecy that I kept this part of Te Makawawa’s
secret, but simply because it had been tacitly understood between
him and me that it was not a matter for repetition.
Crystal rose from the hammock, and saying that her father would
be waiting for us, led the way towards the house.
As she hastened before me among the trees of the garden, and,
later, when she stood and waited for me on the verandah steps,
looking down between the clustering vines, I thought that any man,
no matter how poetical, was a fool to fall in love with the beauty
depicted on a stone, when the world of living things contained such
loveliness in the flesh. Truly it was as Kahikatea had said: the
woman who had conceived the image of Hinauri and reproduced it
upon the stone could not have borne an unlovely child. Yet to say
that Crystal Grey was “not unlovely” would be a very inadequate
description. More positive statements than that would have fallen
from lips more matter-of-fact than mine. If eyes were made for
seeing, then Crystal Grey had her own excuse for being, as
someone somewhere sings, but if words were made for description,
the subtle charm of this child of dreams could find no vehicle but
music.
 CHAPTER XIV.
THE CHIEF OF THE VILE TOHUNGAS.

Having made up his mind to accompany me with his daughter on

our search expedition, Dreamer Grey began setting his sheep-run
and his household matters in order, in view of an absence which
might prove prolonged. It was necessary to engage a competent
manager to look after things, and this meant a delay of at least a
week, which, however, would afford ample time to prepare for the
journey. During the first two or three days of this week I saw much of
Grey, helping him with his sheep and various other things that had to
be seen to. As a consequence we got to know and trust each other
well.
Tiki, who worshipped ‘the little maiden’ as if she were a divine
being, and, when she spoke to him in his own tongue, replied
invariably with a mixture of acquired politeness and native poetry,
half comical, half grand to listen to, had made himself and me
thoroughly uneasy about that taepo he had seen in Cazotl’s boat.
With the wisdom of a savage, who, long accustomed to intertribal
wars, knows almost intuitively when he is being tracked, if not why,
Tiki had it firmly fastened in his mind that the people on the yacht
ought to be watched. As I had not informed Tiki of my suspicion that
we had been tracked along the whole course of our journey, I
regarded his independent view more seriously than if he had known
and exaggerated my own weird feeling in regard to that wizened
negro.
“Very well, Tiki,” I said, the day after our arrival, when he spoke
about it, “if you think ‘the little maiden’ is in danger from those people
you might keep your eye on them.”
He needed no second permission. From that time I saw very little
of him for several days, but I knew he was keeping a strict eye on
the movements of Cazotl and his crew. It was not until the evening of
the fourth day after our arrival that my suspicions received
verification, and his watchfulness nearly cost him his life.
 In the evening of that day, when Grey was busy with some
correspondence in his library, I strolled down into the garden, where I
knew I should find Crystal, for I had seen her go out some time
before with her sketching book. I had dreams of a heaven on earth—
indeed, I should have been less than human if more than three days
had passed over my heart without bringing my ‘love at first sight’ to a
stage in which I felt that the garden where Crystal moved and had
her being was a sacred place. Sweetness lingered in the air. The
dreamy trees, as they rolled in the summer zephyrs, made music
which could not be written down; the rustic retreat beneath the
hazels was full of an influence which I can only describe as the
presence of angels lingering in an atmosphere which has been
purified for them. Sitting here alone late at night, I had been able to
cast aside the littleness of my life and feel that by right of an
ennobling love I might remain there awhile on sufferance. I was
aware that a great change had taken place in me. A new world had
sprung into being, and the splendour of its sun, moon, and stars was
centred in Crystal.
It was with a feeling that all this must soon come to an avowal of
love, as surely as water boils at a given temperature, that I sought
her that evening in the garden; and, I reflected, it would in all
probability reach a sudden end just as surely as the same water
under different conditions freezes at a given degree, for in all sober
reason, who or what was I to deserve the love of such a girl? But I
went to find her all the same, and making my way to the retreat
beneath the nut-trees, held aside the leaves about the entrance and
looked in.
Crystal was sitting in the wicker chair with an open book on her
knees. Her hat was laid aside, and a wisp of her raven hair, fanned
loose from the good-natured mass, half screened her cheek.
“May I come in?” I asked.
“Yes, of course you may,” she laughed, looking up and brushing
the wayward wisp back into its place again. “Come and read me
some of Kawana Kerei’s legends in the original Maori.” She held up
the book as she spoke.
“In the original Maori?” I said, seating myself on the rough bench
against the hazel stems. “That reminds me. I saw a picture in an out-
of-the-way corner of the drawing-room to-day with something in
Maori written beneath it: ‘Degrade the pure one! Whose is the task?
Mine! Mine!’ Did you paint that picture?”
“Yes,” she replied simply; “I painted it quite lately.”
Seeing by her manner that she was a little confused, I asked: “Is it
founded on some Maori legend?”
“No; it is—it is——” She hesitated, with her eyes cast down; then,
after a pause, looked up with a shy smile and asked: “Have you ever
had a very vivid dream, of which you can remember every detail so
accurately that it seems like a real experience?”
“No,” I said; “my imagination is not strong enough for that; but if
you are going to say that picture was painted from a dream of your
own I shall believe you.”
She leaned forward in her chair and half whispered: “Yes, it was; I
saw it as plainly as I see you now, Wanaki,”—she had caught my
Maori name from Tiki—“even more distinctly, if that is possible.”
“Tell me your dream,” I said. “I could not see the faces of the men
in the picture, as the light was not good, but I judge from the legend
that they were evil. It is strange that you should see unlovely things
in your dreams. Will you tell it me?”
“It is rather a terrible subject,” she said, “and I don’t think I quite
understood it even when I had painted the picture. Perhaps you will
tell me what it means. I dreamed that I went away over the sea for
thousands of miles. The silver-tipped waves shone beneath in the
bright moonlight, and the little islands, fringed with palms and belted
with coral, were studded everywhere on the ocean. At last I came to
a vast country, where, in the interior, there were great hills and
mountain lakes, impassable swamps and deep wildernesses. I saw
ancient ruins and long lines of what looked like giant cactus——”
“Mexico,” I said, thinking aloud.
“It may have been; it was vast and it was tropical. In my dream I
found myself standing among the ruined pillars of what must once
have been a colossal temple. Now, it usually happens in a dream
that one sees things vaguely, but in my dream it was different. I saw
every detail. The scale-like feathers on the huge stone snakes that
were coiled up the pillars, the glittering eyes of the vampire bats that
clung about them, the huge green lizard that basked in a patch of
moonlight on the stone floor—all these were clear and distinct, and
on the heavy, broken stonework overhead, supported by the pillars,
were shadowy masses of creeping plants, with here and there a
glistening aloe or clump of white flowers catching the moonlight
through the crevices.
“As I was looking at these things in my dream a murmur of voices
came from within. I advanced between the treble row of pillars and
saw a large inner space where there were a number of figures
moving about a tall column. They were men of different nationalities,
and they chanted a strange song while they looked up at the full
moon which poured its rays down into the open space. These men
had strong, evil faces, with eyes that flashed red in the moonlight; I
can remember each one perfectly, and have drawn them as I saw
them.”
She paused as if she were recalling the vivid scene, and, in the
few moments’ silence, my mind flew back to the Vile Tohungas of the
Pit gazing up at the full moon, nursing their stomachs and curling
their granite lips disdainfully as they worshipped. Ngaraki, no doubt,
would have read in this dream a word from his Great Tohungas of
the Earth to the effect that the Vile Brotherhood of Huo still existed,
striving to work out the age-long degradation of Woman, and, above
all, to destroy his ancient goddess, when, as the Daughter of the
Dawn, she should return. Just as the sacred fire of Hinauri had been
nursed in the breasts of her guardians through pre-Maori races up to
the present, so the baleful red fire of the Vile Tohungas, taken into
the north by their servant fleeing from the wrath of Zun, may have
been kept burning through pre-Toltec civilisations even unto this day.
In spite of myself, this idea was growing upon my mind, when Crystal
continued.
“While I watched, their chant to the moon ended; and, as the last
notes fell, I fancied I could hear them rolling back into the distance
like the close of a song sung by a great multitude in the open air.
Then a large black mirror was brought out of the darkness and fixed
in position so that the moonray was reflected high up on to a dark
part of the smooth stone wall of the ruin. They began a wild orgy
round the pillar. It came to a sudden silence, and all stood still,
gazing at the moonlight on the wall. I looked also and saw, not on the
rock, but in the distance through the rock, what looked like the
central thoroughfare of some great city. By the glare of many lamps,
high and low, I saw carriages crossing and re-crossing, while
omnibuses for ever stopped and moved on again. I saw people
moving to and fro upon the broad pavements; all about were women
—many of them proud-looking and beautiful—who appeared to be
waiting for someone. I did not understand what they were doing
there, but when the men in the open space of the temple cried ‘It is
well! It is well!’ I knew that the vision had shown them the working
out of some great wrong.
“The picture vanished, and they returned to their orgy, which grew
more terrible and furious, then stopped suddenly as before, while
they remained gazing fixedly at the moonlight on the wall. A second
time there came a scene—not the same place, though the people
acted in just the same way—and this also was greeted with the cry,
‘It is well! It is well!’
“It vanished, and a third time the wild orgy was carried on. It
reached a pitch of fury which horrified me, and, when it stopped
suddenly, and they stood gazing at the wall, the vision came again.
But this time it was the white figure of a woman standing among the
trees of a garden far away. The place was bathed in peaceful
sunlight. It was a sun-picture reflected by the moon from a distant
spot. I could not see the features of the woman, but her arms were
raised to the sky, and she seemed to be praying. Then, as if in
answer to her prayer, there came, out of the blue, beings that
seemed more like gods and goddesses than men and women. They
came thronging down towards the world—men with noble looks and
perfect forms, and women with serene, heavenly faces full of all the
tender goodness that should belong to a woman. They appeared to
separate to the four quarters of space, and I thought that here was a
race of more perfect beings coming to people this earth in answer to
the cry of the woman.
“At a sound of murmuring and confusion, I turned to the other
watchers in the open space, and, as I did so, one among them, who
seemed to be chief, stood out from the rest and held up a
threatening hand towards the far-off vision. He laughed, and his
voice was more animal than human. Then he roared out the words
you saw beneath my picture: ‘Degrade the pure one! Whose is the
task? Mine! Mine!’ and with these words ringing in my ears I woke.
That was the dream; was it not strange?”
“Very strange,” I said; “but your father called me as I was glancing
at the picture, and I had not time to examine it very clearly. I should
like to have a good look at it if I may.”
“Yes, I’ll run and get it now.”
“Let me go,” I volunteered. But she was before me, and ran up to
the house. While she was gone I cast my memory over the
extraordinary dream she had related. Matter of fact as I was, I could
not but see that if ever there was a meaning in a dream there was a
meaning in this. The Destroyers of Woman exulting at the slow
undermining of mankind in the mass, their threat hurled at Woman
as the Mother of a nobler and more godlike race, their resolve to
degrade her as such, so that this world should be peopled with dull,
coarse forms, informed by vile minds, such as their own evil faces
portrayed—all this, I reflected with astonishment, was indeed the tale
of Ngaraki the savage, retold from the heart of an innocent girl.
My reflections were cut short by the reappearance of Crystal with
the picture.
“See!” she said, holding it up to the full light; “that is the man who
roared out the words. Why—what is—oh! Mr. Warnock, what is the
matter?”
I had turned from the picture with a gasp, and had sat down on the
wooden bench with my face buried in my hands. I looked up as she
repeated her words and saw mingled bewilderment and concern on
her lovely face.
“It is the face of a fiend,” I said, “not of a man.”
“But why are you so strange? And why do you clench your teeth
like that? You’re as pale as death—surely, Wanaki, the face of a
fiend on canvas cannot be so terrible to look at!”
“It is the face of a fiend,” I repeated fiercely, half beside myself with
maddening fears, “not of a man.”
It was the face of Cazotl. And, in those evil features, I traced a
resemblance which had eluded me on my first sight of the Mexican
in the flesh—a resemblance to the bold granite features of the chief
of the Vile Tohungas of the Pit, so fiercely cursed by Ngaraki.
 Very weird to me was this strange, circumstantial suggestion that
the legendary chief of the Vile Tohungas had returned to be the
actual head of a brotherhood whose aims and objects were identical
with those of ancient time. Yet forcible and valid seemed the
conception that as the guardians of Hinauri had brought their
protecting curses down from the remotest past, so her enemies,
who, although driven far into the north as the aged chief had said,
had preserved their continuity as a Vile Brotherhood through the
ages, had handed on even into the present their hatred of the Pure
One, always with the aim of destroying her or causing her to forget
her Sign of Power. For awhile my shrewd, practical scepticism
struggled against a strong unity of evidence derived independently
from different sources. The aged chief’s belief that the Vile Ones
would return, Crystal’s dream picture of the Destroyers of Women,
the undeniable resemblance of their chief to Cazotl and to the
granite image in the abyss—these things pointed my mind to strange
conclusions; but when I reviewed the conflicting purposes of the
Good and the Vile of ancient time and identified them with the
conflicting purposes of Ngaraki and Cazotl, I was for a moment
almost tempted to throw my common-sense to the winds and say
that the ancient giants of the two priesthoods had returned again and
again to the earth to continue the fierce struggle begun at the very
foundations of the world. But if Cazotl were the arch enemy of
Ngaraki and the would-be destroyer of Hinauri, why should he have
appeared to Crystal Grey in a dream? And why, again, should he
seem to be in pursuit of her? A vague apprehensive shuddering
within me was the only answer to these questions.
 CHAPTER XV.
THE DARKNESS PUTS FORTH A TENTACLE.

That night, long after the house was quiet, I remained leaning on the
sill of my bedroom window, looking down on the peaceful garden
below and turning matters over in my mind. The night wind sighed
and died away in faint puffs upon the trees. A midnight hush was
falling upon everything—a midnight hush and something more: great
black clouds were banking up seaward, and the roses round my
window were sending out heavy odours, such as flowers do before a
thunderstorm. The air became sultry as the inky clouds banked
higher and higher. Then the land wind fell altogether and dead
silence ensued, in which I could hear the titoki-berries opening on
their little hinges, and a strange sound of a going in the high tops of
the native trees in the plantation, while always the leaves of the
aspens tossed and turned in sad unrest.
It may have been the oppressiveness of the air that weighed me
down with a vague presentiment of evil, though now I look back upon
it I am inclined to think my feelings were owing to a strong antipathy
to an evil thing. This antipathy must have been aroused and
strengthened by the discovery recorded in the last chapter. Crystal’s
dream had filled me with feelings even keener, I thought, than those
which had taken possession of Ngaraki, for he, so I reasoned in my
ignorance, had to do merely with inert stones, one sacred, others
cursed; whereas I had to do with flesh and blood. I had little doubt
that Crystal’s dream was one of those strange instances of second
sight which sometimes come to people who live pure lives in quiet
places, where they are in close touch with the nature they can see,
and in closer touch with the nature which they cannot see. The
likeness of the face in the picture to the face of Cazotl was no mere
fancied resemblance. It was striking. It was real. The details of the
picture, too, were true to life, and such as no amount of study from
books could produce. This I coupled with the knowledge that Crystal
had never been away from home except for seven successive years
spent at school in Dunedin. I was driven to the conclusion that there
was something in this dream, and, if something, why not everything?
As I leaned over the window-sill I pondered many things deeply.
Whatever might have been the reason of tracking us all the way from
the Table Land the Mexican’s presence in the Sound appeared to me
to be the speedy carrying out of the threat he had delivered in the
dream. I could well understand that Crystal, with her high ideals and
living energy, was of those women whose very existence is a nail in
the coffin of the fiend in human shape whose glance first strikes the
lily from your hand, and then the truth from beneath your feet.
Consequently, on the one side deepened my love for this perfect
woman with the eyes of night, and on the other blazed a terrible hate
for her would-be destroyer.
With these feelings I entered into the spirit of the brooding
thunderstorm, and, knowing that sleep was impossible, I resolved to
go out of the house, and take my thunder and lightning in the
garden. I had always been fond of a thunderstorm—for in a land
where there are few isolated trees and many bold mountain tops, the
danger from lightning is very small—but on this occasion I welcomed
it with a kind of vivid pleasure, as it was in strict accordance with my
mood.
Going downstairs, I found a mackintosh on the hat-stand in the
hall and put it on. Then, making my way quietly out of the house, I
went round the verandah to see if Tiki was asleep. I was not
surprised to find his mattress of straw unoccupied. He was on the
war track. Probably he had slept by day, and was not watching the
yacht in the interests of ‘the little maiden.’
As I found my way on to the lawn I heard the first rumble of the
thunder over the hills in the distance. The fan-like branches of the
cedars were moving restlessly, as if the terrified air did not know
which way to turn. I could just see their vague outlines against the
blacker sky.
While I stood listening to the ominous whispers of the cedar-
branches, a blinding flash lighted up the place, throwing the wall of
pines above the plantation into clear relief. Then, some miles away,
the thunder crashed and rattled among the hills. In the silence
between the lightning and the thunder, however, I heard what I took
to be a dog or a cat running softly on its four feet across the lawn
from the plantation. My mood of dark hate blinded my usual
wariness, and it never occurred to me that it might be something
else. After the thunder came silence, and then another flash
scribbled down the indigo sky into the hills, and, while it lighted my
surroundings as clear as noonday, my glance happened to fall upon
some gnarled, twisted, and charred remains of a patch of scrub
which had lately been burnt, about twenty yards distant, and just
midway between the plantation and the trees beneath which I stood.
One of the grotesque fragments, a trifle thicker than the others, was
twisted in such a peculiar way that its weirdness caught my attention,
and when the flash had passed I sauntered carelessly towards it and
waited. The peal of thunder was scarcely over when the vivid
lightning streamed down again, and when I looked for the weird
effect of the charred patch, it seemed to me that the grotesque-
looking twist was gone. At the same instant something struck my hat
behind—something which I mistook for the first large drop of the
thunder-shower—and, dismissing the apparent change in the burnt-
out patch of scrub with the passing explanation that it was owing to
my change of position, I sauntered on towards the path that led out
beneath the wall of trees into the fields of the valley. As I went I
certainly thought it strange that one drop of rain should fall alone,
and wondered vaguely what it was that had struck my hat during the
vivid flash.
Passing through the plantation and the wall of pines, whose leaves
threw out a resinous odour in the sultry air, I turned and walked back
along the outside of the plantation, intending to re-enter the
enclosure by a small gap which led directly on to the lawn. As I drew
near this, and flash after flash lighted up the place, I saw from time to
time something, which at first I took for a post, standing in an open
space some thirty paces away from the plantation. When I came
nearer to it, however, the lightning’s glare brought out the object in
bold relief, and it looked more like a man standing bolt upright in the
open field. The thunder now followed sharp on the heels of the
lightning with a deafening crash right overhead, and the heavy rain
came down without warning. Buttoning the mackintosh close up
under my chin, I struck out into the field towards the spot where I had
seen the object that had aroused my curiosity.
When I calculated that I was fairly near it, I stood still and waited
for the flash, for in the darkness I could see nothing. The flash came,
and there, a few steps before me, with the rain dancing from his hair
and glistening shoulders, stood Tiki like a statue, gazing fixedly at
that part of the plantation where the gap led through on to the lawn.
In the brief interval between the lightning and the thunder I called
his name:
“Tiki!”
The words left my lips as the darkness clapped down like the door
of a vault, and in the two seconds that ensued I listened and called
again, but there was only the ready reply of the thunder breaking like
an avalanche overhead.
The next moment I reached the Maori’s side in the darkness,
touched him, shook him, called him, but he made no answer. I could
hear the rain pattering on his bare shoulders; I could hear my own
voice against the final echo of the thunder; then, as the rain held up
a moment and a weird shuddering afterthought of the elements
ricochetted across the sky, I stood still, wondering what strange state
the Maori had fallen into that he stood there like a dead tree-trunk in
the field.
The next flash startled me. It showed Tiki with his teeth set and his
eyes fixed. He appeared like one in that strange cataleptic state in
which the mind and senses are more or less alive, but all volition is
gone. As my eyes rested upon him I detected on his shoulder a
slight stain of blood, which slowly trickled from a wound in which a
small reed dart of two or three inches in length was still sticking. All
this was imprinted upon my eye while the light lasted, but it was not
until darkness supervened that the picture was developed. I found
the dart and pulled it out. Then, as the heavy tread of Tawhaki again
shook the rafters on the House of Tane overhead, I came to the
horrible conclusion that this was the work of that wizard negro—that
the thing which had struck my hat by the cedars was a poisoned dart
of the same kind—that the gnarled and twisted fragment was the
negro himself, that——
 A shudder ended my train of reasoning. The door of the house
was unbarred!
That wizard devil must have been on his way to the house when
he discharged that dart at me!
With terrible thoughts surging through my brain, with the phantom
cry of Cazotl, “Degrade the Pure One!” ringing in my inner ears, and
the passing conjecture that he was now waiting with a boat on the
beach for the return of his wizard minion with Crystal, bereft of all
volition like Tiki, I dashed across the space that separated me from
the gap which led towards the house. No helping flash favoured me
on the way, and when I reached the trees I had to grope about for
the opening. At last I found it and proceeded to make my way
through, but, just as I reached the centre of the plantation, the
lightning forked down right on to the lawn and ran along the ground.
For quite five seconds a dazzling light revealed the way on to the
lawn, and in that brief space of time things happened which five
seconds will not suffice to tell.
Straight before me on a narrow path between two pine trunks, was
the lithe figure of the hideous negro in the act of groping his way
through from the house as the lightning fell. In one hand he held a
reed tube several feet long, and with the other he was feeling for the
tree trunk on his right. Behind him I had a dim idea of a white-robed
figure; but I did not shift my eyes from the negro, for he saw me as
soon as I saw him, and the tube was moving towards his wizened
lips. With a spring I was on to him, and, catching the tube with one
hand just as he set it to his lips, I turned it aside, gave him a violent
thrust in the mouth with it, wrenched it away, and flung it on the
ground. Then I gripped him by the throat, and it was just as we rolled
back together into the bushes that the bright light went out, and our
brief struggle went on in the darkness.
It was brief, for I defy any man to hold a creature of that kind
unless his hand, like Kahikatea’s, could meet right round his neck.
He twisted and turned like an eely fiend, wrenched his throat out of
my grasp, and wriggled away, leaving me snatching at air and tree
trunks.
The thunder rolled off in an angry growl. As it ceased the same
wild laugh that I had heard before came from somewhere far away.
Mistrusting that laugh, and thinking that the negro was in hiding near
by, waiting to make a dash to snatch up his deadly weapon, I quickly
scrambled towards the place where I had thrown it, and soon found it
among the leaves in the darkness.
Then I remembered the figure in white that I had seen following
the negro, and stood peering before me, listening and waiting for the
next flash. I would have called, “Who’s there?” but I knew it was best
to preserve perfect silence with that wizard thing, for there was no
telling what he might do with his infernal poisoned darts, even
without a tube. However, I could not resist throwing out a gentle hint
that I was prepared for him, and that his safest plan was to beat a
retreat. Taking my revolver from my hip pocket, where I always
carried one, I fired a shot up into the trees. It was answered by the
hideous laugh from far away down the Sound, but it followed so
quickly on the report that I knew the author of that laugh, now a
confessed ventriloquist, was near at hand. He was evidently waiting
for the next flash to recover his tube which I held in my hand.
The flash came, and the sight it revealed I shall never forget.
There stood Crystal in the path before me, draped in her night
garments soaked through and through. Her long black hair, in which
flashed countless diamonds of rain, fell loose about her like a veil.
Her mysterious eyes, now like polished obsidian, were fixed in a
glassy stare. Her face was set and pale, like a piece of beautiful
marble. She was in the same state as Tiki, conscious of much that
was passing, as I learned afterwards, but obedient only to
impressions that had been set upon her by the will of another, who
had taken control of her own. On her shoulder, showing through a rift
of her hair, was a stain of blood upon the white linen, but the dart
had been withdrawn.
No sooner had the flash of light passed than that controlling will
was expressed by a voice, harsh and hollow, coming from a little
distance outside the plantation, and pronouncing a strange word in a
 “THE FLASH CAME, AND THE SIGHT IT REVEALED I SHALL NEVER
FORGET. THERE STOOD CRYSTAL IN THE PATH BEFORE ME,
DRAPED IN HER NIGHT GARMENTS.”

language unknown to me. At the sound Crystal attempted to move

past me in the darkness, evidently impelled by the suggestion that
she must follow. But with one arm I caught her round the waist and
held her back. She struggled violently with all her strength to follow
the voice as it repeated the strange word from further in the field,
and it dawned upon me, from what little I knew of this old and new
world black magic of control by suggestion, that if I restrained her by
force the result might be some strange twist of the brain or
aberration of the nervous centres. So I let her move on, retaining one
of her hands and walking by her side for a short distance into the
field. A flash revealed a figure gliding ahead of us, and in order to
make him glide a little quicker, I fired four revolver shots in
succession after him. Then, acting upon an idea which had occurred
to me, I tried to imitate the voice and the strange word he had used.
After two or three attempts beneath my breath, I made the peculiar
sound, coming to a halt at the same time.
It was effective: Crystal stopped also and turned towards me.
Repeating the word I drew her gently back towards the plantation,
and she followed obediently. It was with the idea that her sense of
sight might contradict her sense of hearing that I pressed her eyelids
down and bound my handkerchief over her eyes, lest, when the
lightning flashed, she should see me and become aware of this
deception within a deception.
Thus reiterating the guiding sound which, by the bond of
suggestion placed upon her by the infernal negro wizard,
represented his will, I wrapped the mackintosh about her and led her
through the plantation, over the lawn, and into the house. There,
obedient to my instructions, given to her in the harsh voice of the
negro, she remained in the care of Grey and the servants, with
whom I succeeded in placing her in touch, while I, having hidden the
reed tube in a safe place, hurried out to look for Tiki.
The storm had passed over, and was grumbling itself out in the
distance. A bright star shone down through a break in the clouds, but
it was still too dark to see clearly, and it was with difficulty I made my
way to the place where Tiki had been standing.
After searching about for a long time and finding nothing, I was
favoured by the moon in its last quarter rising over the hills inland
and showing through the heavy cloud drift. By this pale light I
corrected my position and searched again. But there was no sign of
Tiki. Had he recovered and gone after the negro to kill him, or had he
followed obediently under the influence of the poison and the voice?
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookgate.com