Lecture 5 – Protein Structure II

• Protein Structure (cont.)

• Super-secondary structure
• Turns
• Motifs

• Tertiary Structure
• Protein Domains
• 3o Structure Representation Methods:
• visual methods
• non-visual methods of representation:
• contact plots

• Quaternary structure
The Structure of Globular
Proteins
Fibrous structural proteins are long helices.
 e.g., collagen (C), α-, β-keratin.

Water-soluble and membrane proteins more globular.

 therefore, adopt more complex folds…
 that vary greatly from protein to protein.
 the 3o structure describes the global conformation.

However, different proteins often locally similar:

 will generally have regions of similar 2o structure.
 may also contain similar groups of helices:
 super-secondary structure.
 there are numerous common groups, called motifs.

3o structure can also often be segregated into domains:

 independently folding units, with discrete function.
 thus, we first discuss these 2 intermediate levels of
structure.
Super-secondary Structure
and The Need for
Turns
Super-secondary Structure:
 a structural motif formed by the association of 2 or
more helices.
 many common examples exist.
 we’ve discussed several types of helices,
 but something is missing.

Simplest example: a β-hairpin…

 a simple, antiparallel β-sheet, formed
by a single chain.
 this structure requires a tight, 180o turn...
 between the two β-strands.

Turns required for folding,

 …and are non-helical.
 Type I and Type II β-turns
There are 2 typical β-turns
observed in proteins:
 distinguished by the orientation of the
peptide bond connecting residues 2 and
3.
 Type I – keto-O of res. 2 pointed away
from R2, R3 (shown in peptide
bond 3).
 no steric problems.
 the turn in our previous slide was type I.
 Type II – keto-O of residue 2 pointed
towards R2, R3.
 steric problems b/w the keto-O and the C
βof R3.
 thus, a Glycine often observed at R3 (no
Cβ).
Some Extended β-structures
Several common motifs are repetitions of a β-hairpin.
 The β-meander:
 formed by repeated antiparallel β-strands.
 accommodates 1 H-bonding face of each strand:
 all H-bond donors/acceptors of the face paired.
 still have 2 unpaired faces to associate w/ H20.
 intrastrand H-bonding disfavored in Aq. solution.

 The β-barrel:
 enough strands are associated to allow closure.
 H-bonding faces of all strands accommodated.
 H20 interaction eliminated.

 The Greek Key:

 more complicated arrangement of 4 β-strands;
 each pair of neighbors anti-parallel.
The β−α−β Motif
A 2-stranded parallel β-sheet:
 must incorporate a longer turn.
 at least the length of 1 β-strand.
 …often adopts an α-helix.
 to accommodate H-bond donors
and acceptors in the stretch.
 eliminates interaction with H20.
 the β−α−β motif.
 The 2 β-strands are H-bonded…
 in spite of the intervening helix.
Motifs Involving only α-
helices
Finally, there are motifs involving only α-
helices:
 e.g.: The α−α unit, or ‘α-helix hairpin’.
The Domainal Picture of
Protein Structure
Many globular proteins contain independently
folding elements:
 each of which is independently stable, with specific
function.
 longer than super-2o structures (30-300 residues).
 such elements called Domains.
 Many proteins have separate domains:
 for binding of substrates;
 for binding of effectors;
 for enzymatic activity.

most proteins contain overlapping regions.
Mosaic proteins:
 consist exclusively of a chain of non-overlapping
Example: Calmodulin
Calmodulin is an intercellular protein:
 regulates the behavior of many other proteins.
 is a transducer: Ca+ sensor.
 Ca+ binding converts it to an active form.
 has a clear, domainal structure:
Calmodulin composed of 2
Ca+- binding domains.
 2 separate binding domains.
 each domain = a pair of ‘EF-hands’.
 domains connected by a long α-helix.
 helix shortens on Ca+ binding…
 allowing folding into active form.
Protein Tertiary Structure
Tertiary Structure (3o):
 the overall 3-D conformation of the polypeptide
chain.
 may be determined, for instance, by X-ray
crystallography.
How do we represent this overall structure?
 many methods, with different degrees of detail…
 atomic coordinate models.
 various visual inspection models:
 Stick model.
 van der Waals surface (CPK) model.
 Ribbon model.
 Solvent accessible surface (SAS) area model.
 Simple caricature (exaggerated cartoon).
Representation 1 - Atomic
Coordinates
The most detailed representation is a list:
 atomic coordinates for each atom in the polypeptide.
 each list element specifies:
 atom name;
 type, number of the residue;
 spatial coordinates, (x,y,z).
 conceptually simple, but…
 focus is not conformation.
 not very intuitive (not a picture):
 does not indicate helices, or
super-2o structures...
Visual models focus on conformation.
 easily produced (by computer) from atomic
coordinates.
Visual Model 1 – Stick Model
Each bond represented as a stick in 3-D.
 i.e., a line connecting each bonded pair of atoms in
3-space.
 correctly represents atomic positions, relative bond
lengths.
 gives a notion of overall shape.
 clearly shows residue adjacencies.
Atomic sizes not included.
 no indication of steric interactions.
Helical structure not indicated.
 helical structure not shown.
 does not show super-2o structure.
 does not indicate domains.
Visual Model 2–
van der Waals Surface Model
Developed by Corey, Pauling, and Kultun:
 also called ‘CPK’ models.
 includes the size of each atom:
 represents each atom as a vdW sphere.
 radius = van der Waals radius.
 atom types distinguished by color.
Various Disadvantages:
 protein interior obscured.
 primarily a surface plot.
 backbone not distinguished from
side-chains.
 no indication of helical structure.
Visual Model 3 – Ribbon
Model
The Ribbon Model:
 removes the vdW spheres to reveal interior atoms.
 backbone traced by a ribbon.
 aids in visualization of helices.
 still may be too much information.
 2o structure obscured.
 often, only critical residues retained.
 such as those at an active site.
 much more informative.
Numerous variations on the
ribbon model.
 residues may be completely omitted.
 …choice depends on purpose.
Visual Model 4 -
Solvent Accessible Surface
Model
In a Solvent-Accessible Surface (SAS) model:
 new focus: interaction of the macromolecule with
H20.
 backbone may still be represented by a ribbon:
 to indicate the helical structure.

Residues are not explicitly modeled.

 instead, modeled implicitly.
 residues exclude solvent differently,
based on hydrophobicity:
 Limits the solvent’s accessibility.

Resulting water boundary depicted

as a transparent surface…
 the solvent accessible surface (SAS).

Visual Model 5 – Caricature
Finally, we may represent the protein using a
caricature (cartoon) model:
 here, our focus is again on conformation.
 atoms in the backbone replaced by simple symbols.
 side-chains may be retained in stick form.
Each symbol represents a 2o structure.
 each helix represented by a cylinder.
 each β-strand represented by an arrow.
 points N to C.
 greatly simplifies inspection:
 location of 2o structures.
 types of super-2o structures.
 domains clearly shown.
The best model depends on focus.
Non-visual Methods
Conformation may also be represented non-
visually.
 alternative: define helical structures analytically…
 in terms of the mean values of (φ,ψ) required:
 for a given residue to adopt each type of helix structure.
Simple method for a protein of interest:
 calculate the (φ,ψ) value for each residue in the chain.
 from the list of atomic coordinates.
 assign a ‘structure’ to each residue.
 i.e., α-helix (H), β-sheet (E), or random coil (c).
 Example: N-ccHHHHcccEEEEEcc-C
Problem: mapping not unique…
 actual distributions of each type of structure overlap.
 H-bonding patterns may also be considered…
 Kabsch and Sanders (1983).
o
Contact Plots
Alternative method of representing protein
structure:
 is by a plot of residue adjacency 　
 called a contact, or diagonal plot.
Method:
 Calculate the distance (dij) b/w each pair of residues
(i,j).
 distance b/w Cα carbons.
 Construct a sequence vs. sequence grid:
 assign residue contacts at each grid point, (i,j) based on
dij:
 dij = 0 nm (white)
 dij = 0.4-0.5 nm (gray)
 dij = 0.5-0.6 nm (black).
Definition of Close Contact
Definition of ‘close contact’: dij = 0.4-0.6 nm.
 excludes successive residues (i,i+1)…
 In particular:
 α-helix helical rise, h = 0.15 nm;
 β-strand helical rise, h = 0.34 nm;
 Trans-conformation helical rise = 0.36 nm (max helical
rise).
 includes residues that bond to form helices and
sheets.
Residue self-contact, so that dii = 0 nm…
 produces a set of white points along the main
diagonal, i = j.
 Note that contacts will be mirror symmetric about
i=j.
Contact Signature of an α-
helix
An α-helix repeats each 3.6
residues.
 As a result, residue i is within:
 0.4-0.5 nm of residue j = i+3 (gray dot).
 0.5-0.6 nm of residue j = i+4 (black
dot).
 for example, residue 1 is in close
contact with residues 4 (gray) and 5
(black).
An α-helix will therefore generate:
 two linear sets of points:
 on each side of the main diagonal.
 one gray set, and one black set.
 each set parallels the main diagonal.
 each displaced an average of 3.5
Contact Signature of an
Antiparallel β-sheet
An anti-parallel β-sheet:
 has a local hairpin structure.
 places residues i and j in close
contact,
 for i,j symmetric about the central
residue of the hairpin.
An anti-parallel β-sheet will
therefore generate:
 2 linear sets of points;
 perpendicular to the main diagonal;
 centered about the residue which
makes the hairpin turn.
Shown at right, for a 12 residue
sheet:
 centered at residue i=7, the central
residue of the hairpin.
Contact Signature of a
Parallel β-sheet
Consider a parallel β-sheet:
 of length L residue pairs, with
a spacing chain of length N,
 for convenience, let K = L+N;
 total chain length = K + L =
2L+N
 residues i will be in close
contact with residue j =
(L+N)+ i = K+ i.
A parallel β-sheet will
therefore generate:
 2 linear sets of points…
 each parallel to the main
diagonal;
 And offset K residues to either
Contact Plots: Example
Example: Protein with a pair of
adjacent α-helices,
 each helix produces a
characteristic signature.
 details of 2o structure.
 helix-to-helix contact also
indicated:
 by points at the corners of the
‘box’ formed by the 2 helices.
 thus, reveals details of 3o
structure.
Particularly useful when:
 applied to data from of a
technique that measures
distance, directly:
Protein Quaternary Structure

Protein Quaternary Structure (4o) describes:

 formation of a multimeric complex from 2 or more
polypeptides (= units);
 by non-covalent association.
4o structure defined by the type and number
of each unit.
 For instance, there are 2 broad types of dimers:
 homodimer - complex formed by 2 identical units.
 heterodimer - complex formed by 2 different units.

3o and 4o structures generally play different

roles:
 3o structure provides basic functionality;
 4o structure provides organized contact;
 o
The Significance of 4
Structure
Example: Myoglobin (Mb) vs. Hemoglobin (Hb).
 Very similar:
 each binds Oxygen…
 1o, 2o, and 3o structures nearly identical.
 However, each has a different, specific function:
 Mb: O2 storage in muscle cells.
 Hb: transports O2 from the lungs.

Difference: 4o structure.
 Mb is a monomer; Hb is a tetramer:
 2 α units; 2 β units.
 Hb does not function as 4 Mb units:
 each unit has lower O2 affinity.
 O2 binding of units cooperative.
 Cooperativity arises from the contact;
 contact allows communication.
 o
Symmetry in Hb 4 Structure
Complexes of identical subunits typically
symmetric.
 description of 4o structure includes subunit symmetry.
Symmetry of Hb:
 1 true C2 symmetry axis
 (perpendicular to plane of slide).
 2 pseudo-C2 axes (x and y)
 relate α, β dimers, by 180o rotation.
O2 binding breaks the symmetry:
 Fe pulled out of the porphyrin plane.
 tugging on the distal histidine.
 Large conformational change in Mb.
 In Hb, this change resisted by the
other subunits.
 Hb’s tendency to maintain symmetry:
Higher Symmetries in Protein
o
4 Structure
Higher symmetries in 4o structure also
possible…
 when proteins exist in large clusters.
 there are many examples:
 Hemocyanin: D5.
 Virus Coat Proteins: Icosahedral.
Example: Aspartate Transcarbomylase
(ATCase)
 Catalyzes 1st step in pyrimidine synthesis;
 structure (X-ray crystallography)…
 symmetric arrangement (12 subunits):
 6 C subunits (catalytic).
 6 R subunits (regulatory).
 D3 symmetry:
 One C3 axis (perp. to the plane).
 also 3 C axes (in the plane).
Why Symmetric Clusters?
Reasonable Question:
 why do proteins tend to aggregate in symmetric
clusters?
Thermodynamic reason:
 aggregation is entropically unfavorable (∆So < 0).
 Thus, a very negative interaction Enthalpy, ∆Ho
required.
 symmetric aggregation maximizes the interaction
number:
 thereby maximizing -∆Ho.
Example: Homodimer Association
 each monomer: up to 2 interactions.
 Symmetric conformation:
 2 subunit-subunit interactions.
 Nonsymmetric conformation:
Conclusion
In this Lecture, we have discussed:
 The intermediate level structure of Proteins:
 Super-2o and Domainal structure of Proteins.
 Methods of Visualization of the overall 3-D structures
of Proteins.
 Protein 3o structure:
 contact plots.
 Protein 4o structure.

In Lecture 5, we will turn to polynucleotide

structure.