DONALD VOET
`University ofPennsylvania
`
`JUDITH G. VOET
`
`Swarthmore College
`
`BIOCHEMISTRY
`
`SECOND EDITION
`
`JOHN WILEY & SONS, INC.
`New York Chichester Brisbane
`Toronto Singapore
`
`1EX' 2029
`CFAD V NPS
`IPR2015—01093
`
`Page 1
`
`Page 1
`
`NPS EX. 2029
`Part 1
`CFAD v. NPS
`IPR2015-01093
`
`

`
`Cover Art: Two paintings of horse heart cytochrome c by
`Irving Geis in which the protein is illuminated by its sin-
`gle iron atom. On the front cover the hydrophilic side
`chains are drawn in green, and on the back cover the hy-
`drophobic side chains are drawn in orange. The paintings
`are based on an X-ray structure by Richard Dickerson.
`
`Acquisitions Editor: Nedah Rose
`Marketing Manager: Catherine Faduska
`Designer: Madelyn Lesure
`Manufacturing Manager: Susan Stetzer
`Photo Researcher: Hilary Newman
`Illustration Coordinator: Edward Starr
`Illustrators: J/B Woolsey Associates
`Production Management: Pamela Kennedy Oborski and
`Suzanne Ingrao
`Cover and Part Opening Illustrations © Irving Geis
`This book was set in 9.5/12 Times Roman by Progressive
`Information Technologies and printed and bound by
`Von Hoffman Press. The cover was printed by Lehigh
`Press Lithographers. Color separations by Lehigh Press
`Colortronics.
`
`Recognizing the importance of preserving what has been written, it is a policy of John Wiley &
`Sons, Inc. to have books of enduring value published in the United States printed on acid-free
`paper, and we exert our best efforts to that end.
`
`Copyright © 1995, by John Wiley & Sons, Inc.
`
`All rights reserved. Published simultaneously in Canada.
`
`Reproduction or translation of any part of
`this work beyond that permitted by Sections
`107 and 108 of the 1976 United States Copyright
`Act without the permission of the copyright
`owner is unlawful. Requests for permission
`or further information should be addressed to
`the Permissions Department, John Wiley & Sons, Inc.
`
`ISBN: 0-471-58651-X
`
`Printed in the United States of America
`
`10987654
`
`Page 2
`
`Page 2
`
`

`
`
`
`CHAPTER
`
`
`
`Three-Dimensional
`Structures OfProteins
`
`
`
`The properties of a protein are largely determined by its
`three-dimensional structure. One might naively suppose
`that since proteins are composed of the same 20 types of
`amino acid residues, they would be more or less alike in
`their properties. Indeed, denatured (unfolded) proteins
`have rather similar characteristics, a kind of homogeneous
`
`“average” of their randomly dangling side chains. How-
`ever, the three-dimensional structure of a native (physio-
`logically folded) protein is specified by its primary structure
`so that it has a unique set of characteristics.
`In this chapter, we shall discuss the structural features of
`proteins, the forces that hold them together, and their hier-
`archical organization to form complex structures. This will
`form the basis for understanding the structure—function
`relationships necessary to comprehend the biochemical
`roles of proteins. Detailed consideration of the dynamic
`behavior of proteins and how they fold to their native struc-
`tures is deferred until Chapter 8.
`
`1. SECONDARY STRUCTURE
`
`A po1ymer’s secondary structure (2° structure) is defined as
`the local conformation of its backbone. For proteins, this
`has come to mean the specification of regular polypeptide
`backbone folding patterns: helices, pleated sheets, and
`turns. However, before we begin our discussion of these
`basic structural motifs let us consider the geometrical prop-
`erties of the peptide group because its understanding is pre-
`requisite to that of any structure containing it.
`
`
`
`Page 3
`
`1. Secondary Structure
`A. The Peptide Group
`B. Helical Structures
`C. Beta Structures
`D. Nonrepetitive Structures
`2. Fibrous Proteins
`A.
`or Keratin—A Helix of Helices
`B. Silk Fibroin—A B Pleated Sheet
`C. Collagen — A Triple Helical Cable
`D. Elastin —— A Nonrepetitive Coil
`3. Globular Proteins
`A.
`Interpretation of Protein X—Ray and NMR Structures
`B. Tertiary Structure
`
`4. Protein Stability
`Electrostatic Forces
`
`.m.U.O.==.>
`
`Hydrogen Bonding Forces
`Hydrophobic Forces
`Disulfide Bonds
`Protein Denaturation
`
`5. Quaternary Structure
`A. Subunit Interactions
`
`B. Symmetry in Proteins
`C. Determination of Subunit Composition
`
`Appendix: Viewing Stereo Pictures
`
`141
`
`Page 3
`
`

`
`
`142 Chapter 7. Three-Dimensional Structures OfProteins
`
`
`
`
`
`A. The Peptide Group
`
`In the 1930s and 1940s, Linus Pauling and Robert Corey
`determined the X-ray structures of several amino acids and
`dipeptides in an effort to elucidate the structural constraints
`on the conformations of a polypeptide chain. These studies
`indicated that the peptide group has a rigid, planar structure
`(Fig. 7-1) which, Pauling pointed out, is a consequence of
`resonance interactions that give the peptide bond an ~40%
`double-bond character:
`
`FIGURE 7-1. The standard dimensions (in angstroms, A, and
`degrees,‘’) of the planar trans-peptide group derived by
`averaging the results of X-ray crystal structure determinations of
`amino acids and peptides. [After Marsh, R.E. and Donohue,
`J ., Adv. Protein Chem. 22, 249 (1967).]
`
`This explanation is supported by the observations that a
`peptide’s C—N bond is 0.13 A shorter than its N—C,,
`single bond and that its C=O bond is 0.02 A longer than
`that of aldehydes and ketones. The peptide bond’s reso-
`nance energy has its maximum value, ~85 kJ - mol”‘, when
`the peptide group is planar because its 7:-bonding overlap is
`maximized in this conformation. This overlap, and thus the
`resonance energy, falls to zero as the peptide bond is twisted
`to 90° out of planarity, thereby accounting for the planar
`peptide group’s rigidity.
`Peptide groups, with few exceptions, assume the trans
`conformation: that in which successive C“ atoms are on
`opposite sides of the peptide bond joining them (Fig. 7-1).
`This is partly a result of steric interference which causes the
`cis conformation (Fig. 7-2) to be ~8 kJ -mol“ less stable
`than the trans conformation (this energy difference is
`somewhat less in peptide bonds followed by a Pro residue;
`indeed, ~10% of the Pro residues in proteins follow a cis
`peptide bond, whereas cis peptides are otherwise extremely
`rare).
`
`Polypeptide Backbone Conformations May Be Described
`by Their Torsion Angles
`The above considerations are important because they in-
`dicate that the backbone ofa protein is a linked sequence of
`rigid planar peptide groups (Fig. 7-3). We can therefore
`specify a polypeptide’s backbone conformation by the tor-
`sion angles (rotation angles or dihedral angles) about the
`
`Main chain ,
`
`
`
`
`
`cis-Peptide group
`
`FIGURE 7-2. The cis-peptide group.
`
`J-.-~.-_
`
`
`
`|1l_I!:Il.I_II!fi-;.,m-.._.._.
`
`Ca—N bond (<75) and the C,,—C bond (w) of each of its
`amino acid residues. These angles, (1) and I//, are both de-
`fined as 180° when the polypeptide chain is in its planar,
`fully extended (all-trans) conformation and increase for a
`
`FIGURE 7-3. A polypeptide chain in its fully extended conformation showing the planarity of
`each of its peptide groups. [Figure copyrighted © by Irving Geis.]
`
`Page 4
`
`

`
`
`
`
`Section 7-]. Secondary Structure
`
`143
`
`i
`
`Amide plane
`
`
`H
`
`H
`
`H
`
`H
`
`H
`
`H
`
`H H
`
`H
`H
`
`.
`
`H
`H
`
`(a) Staggered
`
`(b)
`
`Eclipsed
`
`
`
`FIGURE 7-5. Newman projections indicating the (a)
`staggered conformation and (b) eclipsed conformation of ethane.
`
`
`
`
`
`rangement because opposing hydrogen atoms are maxi-
`mally separated in this position. In contrast, the eclipsed
`conformation (Fig. 7-5b), which is characterized by a tor-
`sion angle of 0°, is least stable because in this arrangement
`
`the hydrogen atoms are closest to each other. The energy
`
`difference between the staggered and eclipsed conforma-
`
`tions in ethane is ~12 kJ - mol‘1, a quantity that represents
`
`an energy barrier to free rotation about the C—C single
`
`bond. Substituents other than hydrogen exhibit greater
`steric interference; that is, they increase the size of this en-
`ergy barrier due to their greater bulk. Indeed, with large
`Substituents, some conformations may be sterically for-
`bidden.
`
`
`
`
`
`Amide plane
`
`
`
`
`-u—¢--L».-——IJ4:
`
`
`
`FIGURE 7-4. A portion of a polypeptide chain indicating the
`torsional degrees of freedom of each peptide unit. The only
`reasonably free movements are rotations about the Ca—N bond
`((1)) and the C,,—C bond (1//). The torsion angles are both 180'‘
`in the conformation shown and increase, as is indicated, in a
`clockwise manner when viewed from C0,. [Figure copyrighted ©
`by Irving Geis.]
`
`Allowed Conformations of Polypeptides Are Indicated by
`the Ramachandran Diagram
`The sterically allowed values of d) and a// can be deter-
`mined by calculating the distances between the atoms of a
`tripeptide at all values of <1) and l// for the central peptide
`unit. Sterically forbidden conformations, such as the one
`shown in Fig. 7-6, are those in which any nonbonding inter-
`atomic distance is less than its corresponding van der Waals
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`clockwise rotation when viewed from C“ (Fig. 7-4). (In ear-
`lier literature, this fully extended conformation was defined
`as having gb = y/ = 0°. These values can be made to corre-
`spond to the present convention by subtracting 180° from
`both angles.)
`There are several steric constraints on the torsion angles,
`(I) and 1//, of a polypeptide backbone that limit its conforrna-
`tional range. The electronic structure of a single (0) bond,
`such as a C—C bond, is cylindrically symmetrical about its
`bond axis so that we might expect such a bond to exhibit
`free rotation. Ifthis were the case, then in ethane, for exam-
`ple, all torsion angles about the C-C bond would be
`equally likely. Yet, certain conformations in ethane are fa-
`vored because of repulsions between electrons in the C—H
`bonds. The staggered conformation (Fig. 7-5a), which
`occurs at a 180° torsion angle, is ethane’s most stable ar-
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`FIGURE 7-6. Steric interference between the carbonyl oxygen
`and the amide hydrogen on adjacent residues prevents the
`occurrence of the conformation (15 = —60°, l// = 30°. [Figure
`copyrighted © by Irving Geis.]
`
`
`
`Page 5
`
`

`
`
`
`144 Chapter 7. Three-Dimensional Structures OfProteins
`
`
`
`180
`
`90
`
`3
`
`‘I
`
`n
`
`1
`-1
`
`-I
`_l
`
`7
`
`-1
`
` 1'
`
`_I
`
`-90 -
`
`‘I
`
`-1
`
`I
`
`T
`
`'
`
`'
`
`1
`-1
`
`-1
`-1
`
`"
`
`-
`
`_
`
`480
`-180
`
`-90
`
`l_|__L
`o
`¢(deE)
`
`.n.
`90
`
`180
`
`FIGURE 7-7. A Ramachandran diagram (named in honor of
`its inventor, G.N. Ramachandran) shows the sterically allowed
`4» and VI angles for poly-L-alanine. The diagram was calculated
`using the van der Waals distances in Table 7-1. Regions of
`“normally allowed” (1) and VI angles are shaded in blue, whereas
`green-shaded regions correspond to conformations having
`“outer limit” van der Waals distances. The conformation angles,
`(I) and I//, of several secondary structures are indicated below:
`
`Secondary Structure
`
`4) (deg) w (deg)
`
`Right-handed oz helix (at)
`Parallel B pleated sheet (TT)
`Antiparallel B pleated sheet (T l)
`Right-handed 3,0 helix (3)
`Right-handed 1t helix (7t)
`2.27 ribbon (2)
`Left-handed polyglycine II and poly-L-proline
`II helices (II)
`Collagen (C)
`Left-handed oz helix (a,_)
`
`— 57
`- 1 19
`- 139
`— 49
`— 57
`—78
`-79
`
`— 5 1
`57
`
`-47
`1 13
`135
`— 26
`- 70
`59
`150
`
`153
`47
`
`[After Flory, P.J., Statistical Mechanics ofChain Molecules, p. 253,
`Interscience (1969); and IUPAC-IUB Commission on Biochemical
`Nomenclature, Biochemistry 9, 3475 (1970).]
`
`distance. Such information is summarized in a conforma-
`tion map or Ramachandran diagram (Fig. 7-7).
`Figure 7-7 indicates that most areas of the Ramachan-
`dran diagram (most combinations of (I) and VI) are confor-
`mationally inaccessible to a polypeptide chain. The partic-
`ular regions of the Ramachandran diagram that represent
`allowed conformations depend on the van der Waals radii
`chosen to calculate it. But with any realistic set of values,
`such as that in Table 7-1, only three small regions of the
`conformational map are physically accessible to a polypep-
`tide chain. Nevertheless, as we shall see, all ofthe common
`types of regular secondary structures found in proteins fall
`within allowed regions of the Ramachandran diagram. In-
`
`180hlL_Il__Ll_l
`-180
`-90
`o
`90
`lllldegl
`FIGURE 7-8. The conformation angle distribution of all
`residues but Gly and Pro in 12 precisely determined high-
`resolution X-ray structures with a superimposed Ramachandran
`diagram. [After Richardson, J.S. and Richardson, D.C., in
`Fasman, G.D. (Ed.), Prediction ofProtein Structure and the
`Principles ofProtein Conformation, p. 6, Plenum Press (1989).]
`
`180
`
`TABLE 7-1. VAN DER WAALS DISTANCES FOR '
`INTERATOMIC CONTACTS
`9
`
`Contact Type
`
`Normally Allowed (A)
`
`Outer Limit (A)
`
`-
`
`1.9
`2.0
`- H
`H -
`2.2
`2.4
`- O
`H -
`2.2
`2.4
`- N
`H -
`2.2
`2.4
`- C
`H -
`2.6
`2.7
`- O
`-
`O -
`2.6
`2.7
`- N
`-
`0 -
`2.7
`2.8
`- C
`-
`O -
`2.6
`2.7
`- N
`-
`N -
`2.8
`2.9
`N -- - C
`2.9
`3.0
`C -
`-
`- C
`3.0
`3.2
`C -
`-
`- CH2
`
`
`3.2CH, -- - CH2 3.0
`Source: Ramachandra, G.N. and Sasisekharan, V., Adv. Protein Chem.23,
`326 (1968).
`
`deed, the observed conformational angles of most non-Gly
`residues in proteins whose X-ray structures have been de-
`termined lie in these allowed regions (Fig. 7-8).
`Most points that fall in forbidden regions of Fig. 7-8 lie
`between its two fully allowed areas near VI = 0. However,
`these “forbidden” conformations, which arise from the col-
`lision of successive amide groups, are allowed if twists of
`only a few degrees about the peptide bond are permitted.
`This is not unreasonable since the peptide bond offers little
`resistance to small deformations from planarity.
`
`Page 6
`
`Page 6
`
`

`
`Gly, the only residue without a C, atom, is much less
`sterically hindered than the other amino acid residues. This
`is clearly apparent in comparing the Ramachandran dia-
`
`Section 7-1. Secondary Structure
`
`145
`
`gram for Gly in a polypeptide chain (Fig. 7-9) with that of
`other residues (Fig. 7-7). In fact, Gly often occupies posi-
`tions where a polypeptide backbone makes a sharp turn
`which, with any other residue, would be subject to steric
`interference.
`
`Figure 7-7 was calculated for three consecutive Ala resi-
`dues. Similar plots for larger residues that are unbranched
`
`at C5, such as Phe, are nearly identical. In Ramachandran
`diagrams of residues that are branched at C5, such as Thr,
`the allowed regions are somewhat smaller than for Ala. The
`cyclic side chain of Pro limits its (1) to the range -60“ i
`25 °, making it, not surprisingly, the most conformationally
`restricted amino acid residue. The conformations of resi-
`
`dues in chains longer than tripeptides are even more re-
`stricted than the Ramachandran diagram indicates be-
`cause, for instance, a polypeptide chain cannot assume a
`conformation in which it passes through itself. We shall see,
`however, that despite the great restrictions that peptide
`bond planarity and side chain bulk place on the conforma-
`tions of a polypeptide chain, every unique primary struc-
`ture has a correspondingly unique three-dimensional struc-
`ture.
`
`-180
`-180
`
`-90
`
`O
`
`90
`
`180
`
`B. Helical Structures
`
`¢.'(de8)
`FIGURE 7-9. The Ramachandran diagram of Gly residues in
`a polypeptide chain. “Normally allowed” regions are shaded
`in blue, whereas green-shaded regions correspond to “outer
`limit” atomic distances. Gly residues have far greater
`conformational freedom than do other (bulkier) amino acid
`residues as the comparison of this figure with Fig. 7-7 indicates.
`[After Ramachandran, G.N. and Sasisekharan, V., Adv. Protein
`Chem. 23, 332 (I968).]
`
`Helices are the most striking elements of protein 2° struc-
`ture. If a polypeptide chain is twisted by the same amount
`about each of its C, atoms, it assumes a helical conforma-
`tion. As an alternative to specifying its ()5 and z// angles, a
`helix may be characterized by the number, n, of peptide
`units per helical turn, and its pitch, p, the distance the helix
`rises along its axis per turn. Several examples of helices are
`diagrammed in Fig. 7-10. Note that a helix has chirality;
`
`n=5
`
`n=4
`
`n.=3
`
`n=2
`
`n=-3
`
`\
`
`\ ©
`
`llzvwa
`65i5
`
`FIGURE 7-10. Examples of helices indicating the definitions
`of the helical pitch, p, the number of repeating units per turn,
`n, and the helical rise per repeating unit, d = p/n. Right- and
`left-handed helices are defined, respectively, as having positive
`
`and negative values of n. For n = 2, the helix degenerates to a
`nonchiral ribbon. For 1) = 0, the helix degenerates to a closed
`ring. [Figure copyrighted © by Irving Geis.]
`
`Page 7
`
`

`
`l l2 44 t
`
`i
`£-
`:5.
`
`3.
`
`A polypeptide helix must, of course, have conformation
`angles that fall within the allowed regions of the Rama-
`chandran diagram. As we have seen, this greatly limits the
`possibilities. Furthermore, ifa particular conformation is to
`have more than a transient existence, it must be more than
`
`just al1owed—it must be stabilized. The “glue” that holds
`polypeptide helices and other 2° structures together is, in
`part, hydrogen bonds.
`
`The at Helix
`
`Only one helical polypeptide conformation has simulta-
`neously allowed conformation angles and a favorable hy-
`drogen bonding pattern: the C! helix (Fig. 7-1 I), a particu-
`larly rigid arrangement of the polypeptide chain.
`Its
`discovery through model building, by Pauling in 1951,
`ranks as one of the landmarks of structural biochemistry.
`For a polypeptide made from L-a-amino acid residues,
`the oz helix is right handed with torsion angles (1) = - 57°
`and 1/1 = -47 °, n = 3.6 residues per turn, and a pitch of 5.4
`A. (An oz helix of D-oz-amino acid residues is the mirror
`image of that made from L-amino acid residues: It is left
`handed with conformation angles (12 = + 57° and 1/1 =
`+47° but with the same values of n and p.)
`Figure 7-11 indicates that the hydrogen bonds of an a
`helix are arranged such that the peptide C=O bond of the
`nth residue points along the helix towards the peptide
`N—H group of the (n + 4)th residue. This results in a
`strong hydrogen bond that has the nearly optimum
`N -
`-
`- 0 distance of 2.8 A. In addition, the core of the a
`helix is tightly packed; that is, its atoms are in van der Waals
`contact across the helix, thereby maximizing their associa-
`tion energies (Section 7-4A). The R groups, whose posi-
`tions, as we saw, are not fully dealt with by the Ramachan-
`dran diagram, all project backward (downward in Fig. 7-1 1)
`and outward from the helix so as to avoid steric interference
`
`with the polypeptide backbone and with each other. Such
`an arrangement can also be seen in Fig. 7-12. Indeed, a
`major reason why the left-handed ct helix has never been
`observed (its helical parameters are but mildly forbidden;
`Fig. 7-7) is that its side chains contact its polypeptide back-
`bone too closely. Note, however, that 1 to 2% ofthe individ-
`ual non-Gly residues in proteins assume this conformation
`(Fig. 7-8).
`The at helix is a common secondary structural element of
`both fibrous and globular proteins. In globular proteins, at
`helices have an average span of ~ 12 residues, which corre-
`sponds to over three helical turns and a length of 18 A.
`However, oz helices with as many as 53 residues have been
`found.
`
`Other Polypeptide Helices
`Figure 7-13 indicates how hydrogen bonded polypeptide
`helices may be constructed. The first two, the 2.27 ribbon
`and the 31., helix, are described by the notation n,,, where n,
`as before, is the number of residues per helical turn, and m
`is the number of atoms, including H, in the ring that is
`
`Page 8
`
`146 Chapter 7. Three-Dimensional Structures OfProteins
`
`that is, it may be either right handed or left handed (a right-
`handed helix turns in the direction that the fingers of a right
`hand curl when its thumb points along the helix axis in the
`direction that the helix rises). In proteins, moreover, n need
`not be an integer and, in fact, rarely is.
`
`©
`IRVING
`6215
`
`FIGURE 7-11. The right-handed oz helix. Hydrogen bonds
`between the N—H groups and the C=O groups that are four
`residues back along the polypeptide chain are indicated by
`dashed lines. [Figure copyrighted © by Irving Geis.]
`
`Page 8
`
`

`
`Section 7-]. Secondary Structure
`
`147
`
`FIGURE 7-12. A stereo, space-filling representation of an oz
`helical segment of sperm whale myoglobin (its E helix) as
`determined by X-ray crystal structure analysis. In the main
`chain, carbon atoms are green, nitrogen atoms are blue, oxygen
`
`atoms are red, and hydrogen atoms are white. The side chains
`are yellow. Instructions for viewing stereo diagrams are given in
`the appendix to this chapter.
`
` v
`
`6535
`
`l
`t
`_ _ _ _ _ _ __ /_ ——____.____/ ______._......_../
`/310 helix
`I on helix
`2.27 ribbon
`
`1
`//1: helix
`
`FIGURE 7-13. The hydrogen bonding pattern of several polypeptide helices. In the cases shown,
`the polypeptide chain curls around such that the C=O group on residue n forms a hydrogen
`bond with the N—H groups on residues (n + 2),
`(n + 3), (n + 4), or (n + 5).
`[Figure copyrighted © by Irving Geis.]
`
`
`
`Page 9
`
`Page 9
`
`

`
`
`
`148 Chapter 7. Three-Dimensional Structures OfProteins
`
`\_.
`
`on helix
`310 helix
`FIGURE 7-14. Two polypeptide helices that occasionally
`occur in proteins compared with the commonly occurring oz
`helix. (a) The 3,0 helix is characterized by 3.0 peptide units per
`turn and a pitch of 6.0 A, which makes it thinner and more
`elongated than the at helix. (b) The oz helix has 3.6 peptide units
`
`1: helix ¢
`
`per turn and a pitch of 5.4 A (also see Fig. 7-11). (c) The n helix,
`with 4.4 peptide units per turn and a pitch of 5.2 A, is wider
`and shorter than the oz helix. The peptide planes are indicated.
`[Figure copyrighted © by Irving Geis.]
`
`closed by the hydrogen bond. With this notation, an an helix
`is a 3.613 helix.
`The right-handed 3 ,0 helix (Fig. 7- 14a), which has a pitch
`of 6.0 A, is thinner and rises more steeply than does the an
`helix (Fig. 7-14b). Its torsion angles place it in a mildly
`forbidden zone ofthe Ramachandran diagram that is rather
`near the position of the oz helix (Fig. 7-7) and its R groups
`experience some steric interference. This explains why the
`3 ,0 helix is only occasionally observed in proteins, and then
`mostly in short segments that are frequently distorted from
`the ideal 3,0 conformation. The 3,0 helix most often occurs
`as a single-tum transition between one end ofan oz helix and
`the adjoining portion of a polypeptide chain.
`The 1: helix (4.416 helix), which also has a mildly forbid-
`den conformation (Fig. 7-7), has only rarely been observed
`and then only as segments of longer helices. This is proba-
`bly because its comparatively wide and flat conformation
`(Fig. 7- 14c) results in an axial hole that is too small to admit
`water molecules but yet too wide to allow van der Waals
`associations across the helix axis; this greatly reduces its
`
`stability relative to more closely packed conformations.
`The 2.27 ribbon, which as Fig. 7-7 indicates, has strongly
`forbidden conformation angles, has never been observed.
`Certain synthetic homopolypeptides assume conforma-
`tions that are models for helices in particular proteins. Po-
`lyproline is unable to assume any common secondary
`structure due to the conformational constraints imposed by
`its cyclic pyrrolidine side chains. Furthermore, the lack ofa
`hydrogen substituent on its backbone nitrogen precludes
`any polyproline conformation from being knit together by
`hydrogen bonding. Nevertheless, under the proper condi-
`tions, polyproline precipitates from solution as a left-
`handed helix of all-trans peptides that has 3.0 residues per
`helical turn and a pitch of 9.4 A (Fig. 7-15). This rather
`extended conformation permits the Pro side chains to avoid
`each other. Curiously, polyglycine, the least conformation-
`ally constrained polypeptide, precipitates from solution as a
`helix whose parameters are essentially identical to those of
`polyproline, the most conformationally constrained poly-
`peptide (although the polyglycine helix may be either right
`
`Page 10
`
`Page 10
`
`

`
`(a) Antiparallel
`
`Section 7- 1. “Secondary Structure
`
`149
`
`
`
`Parallel
`
`(13)
`
`FIGURE 7-15. The polyproline II
`helix. Polyglycine forms a nearly
`identical helix (polyglycine II). [Figure
`copyrighted © by Irving Geis.]
`
`FIGURE 7-16. The hydrogen bonding associations in B pleated sheets. Side chains are
`omitted for clarity. (a) The antiparallel /3 pleated sheet. (b) The parallel fl pleated sheet.
`[Figure copyrighted © by Irving Geis.]
`
`, C4——N\‘.
`
`I
`
`or left handed because Gly is nonchiral). The structures of
`the polyglycine and polyproline helices are of biological
`significance because they form the basic structural motif of
`collagen, a structural protein that contains a remarkably
`high proportion of both Gly and Pro (Section 7-2C).
`
`bonding capacity of the polypeptide backbone. In /Ipleated
`sheets, however, hydrogen bonding occurs between neigh-
`boring polypeptide chains rather than within one as in oz
`helices.
`
`/3 Pleated sheets come in two varieties:
`
`C. Beta Structures
`
`In 1951, the year that they proposed the at helix, Pauling
`and Corey also postulated the existence of a different poly-
`peptide secondary structure, the ,8 pleated sheet. As with the
`a helix, the B pleated sheet’s conformation has repeating ti)
`and y/ angles that fall in the allowed region of the Rama-
`chandran diagram (Fig. 7-7) and utilizes the full hydrogen
`
`
`
`1. The antiparallel ,6 pleated sheet, in which neighboring
`hydrogen bonded polypeptide chains run in opposite
`directions (Fig. 7- 1 6a).
`
`2. The parallel fl pleated sheet, in which the hydrogen
`bonded chains extend in the same direction (Fig. 7- 16b).
`
`The conformations in which these [3 structures are opti-
`mally hydrogen bonded vary somewhat from that of a fully
`extended polypeptide (cl) = my = i 180°) as indicated in
`Fig. 7-7. They therefore have a rippled or pleated edge-on
`
`Page 11
`
`Page 11
`
`

`
`150 Chapter 7. Three-Dimensional Structures 0fProteins
`
`appearance (Fig. 7-17), which accounts for the appellation
`“pleated sheet.” In this conformation, successive side
`chains of a polypeptide chain extend to opposite sides ofthe
`pleated sheet with a two-residue repeat distance of 7.0 A.
`/3 Sheets are common structural motifs in proteins. In
`globular proteins, they consist of from 2 to as many as 15
`polypeptide strands, the average being 6 strands, which
`have an aggregate width of ~25 A. The polypeptide chains
`in a /3 sheet are known to be up to 15 residues long, with the
`average being 6 residues that have a length of ~21 A. A
`6-stranded antiparallel /3 sheet, for example, occurs in the
`jack bean protein concanavalin A (Fig. 7-18).
`Parallel ,8 sheets of less than five strands are rare. This
`observation suggests that parallel fl sheets are less stable
`than antiparallel /3 sheets, possibly because the hydrogen
`bonds ofparallel sheets are distorted in comparison to those
`of the antiparallel sheets (Fig. 7-16). Mixed parallel-
`antiparallel ,8 sheets are common but occur with only ~40%
`of the frequency that would be expected for the random
`mixing of strand directions.
`The B pleated sheets in globular proteins invariably ex-
`hibit a pronounced right-handed twist when viewed along
`
`FIGURE 7-17. A two-stranded B antiparallel pleated sheet
`drawn so as to emphasize its pleated appearance. Dashed lines
`indicate hydrogen bonds. Note that the R groups (purple balls)
`on each polypeptide chain alternately extend to opposite sides
`of the sheet and that they are in register on adjacent chains.
`
`[Figure copyrighted © by Irving Geis.]
`to this chapter.
`
`FIGURE 7-18. A stereo, space-filling
`representation of the six-stranded
`antiparallel fl pleated sheet in jack
`bean concanavalin A as determined
`by X-ray crystal structure analysis.
`In the main chain, carbon atoms are
`green, nitrogen atoms are blue,
`oxygen atoms are red, and hydrogen
`atoms are white. R groups are
`represented by large purple balls.
`Instructions for viewing stereo
`drawings are given in the appendix
`
`Page 12
`
`Page 12
`
`

`
`
`, S
`
`ection 7-]. Secondary Structure
`
`151
`
`FIGURE 7-19. Polypeptide chain folding in proteins
`illustrating the right-handed twist of B sheets. Here the
`polypeptide backbones are represented by ribbons with or helices
`shown as coils and the strands of B sheets indicated by arrows
`pointing towards the C-terminus. Side chains are not shown.
`((1) Bovine carboxypeptidase A, a 307-residue protein, contains
`an eight-stranded mixed B sheet that forms a saddle-shaped
`curved surface with a right-handed twist. (b) Triose phosphate
`isomerase, a 247-residue chicken muscle enzyme, contains an
`eight-stranded parallel B sheet that forms a cylindrical structure
`known as a B barrel [here viewed from the top (left) and from
`the side (right)]. Note that the crossover connections between
`successive strands of the B barrel, which each consist
`predominantly of an a helix, are outside the B barrel and have a
`right-handed helical sense. [After drawings by Jane Richardson,
`Duke University.]
`
`(a)
`
`
`
`
`
`
`
`their polypeptide strands (e.g., Fig. 7-19). Such twisted B
`sheets are important architectural features of globular pro-
`teins since B sheets often form their central cores (Fig. 7- 19).
`Conformational energy calculations indicate that a B
`sheet’s right-handed twist is a consequence of nonbonded
`interactions between the chiral L—amino acid residues in the
`
`sheet’s extended polypeptide chains. These interactions
`tend to give the polypeptide chains a slight right-handed
`helical twist (Fig. 7-19) which distorts and hence weakens
`the B sheet’s interchain hydrogen bonds. A particular B
`sheet’s geometry is thus the result of a compromise between
`optimizing the conformational energies of its polypeptide
`chains and preserving its hydrogen bonds.
`The topology (connectivity) of the polypeptide strands in
`a B sheet can be quite complex; the connecting links ofthese
`assemblies often consist of long runs of polypeptide chain
`which frequently contain helices (e.g., Fig. 7-19). The link
`connecting two consecutive antiparallel strands is topologi-
`cally equivalent to a simple hairpin turn (Fig. 7-20a). How-
`ever, tandem parallel strands must be linked by a cross-over
`connection that is out of the plane of the B sheet. Such
`crossover connections almost always have a right-handed
`
`
`FIGURE 7-20. The connections between adjacent
`polypeptide strands in B pleated sheets: (a) The hairpin
`connection between antiparallel strands is topologically in the
`plane of the sheet. (b) A right-handed crossover connection
`between successive strands of a parallel B sheet. Nearly all such
`crossover connections in proteins have this chirality (see, e.g.,
`Fig. 7-1911). (C) A left-handed cross-over connection between
`parallel B sheet strands. Connections with this chirality are rare.
`[After Richardson, J .S., Adv. Protein Chem. 34, 290, 295 (1981).]
`
`
`
`Page 13
`
`
`Page 13
`
`

`
`152 Chapter 7. Three-Dimensional Structures OfProteins
`
` FIGURE 7-21. A possible folding scheme illustrating how
`
`right-handed polypeptide chain twisting favors the formation of
`right-handed cross-over connections between successive strands
`of a parallel B sheet.
`
`helical sense (Fig. 7-20b), which is thought to better fit the B
`sheets’ inherent right-handed twist (Fig. 7-21).
`
`D. Nonrepetitive Structures
`
`Regular secondary structures—/telices and B sheets-
`comprise around half of the average globular protein. The
`protein’s remaining polypeptide segments are said to have a
`coil or loop conformation. That is not to say, however, that
`these nonrepetitive secondary structures are any less or-
`dered than are helices or B sheets; they are simply irregular
`and hence more difficult to describe. You should therefore
`not confuse the term coil conformation with the term ran-
`
`dom coil, which refers to the totally disordered and rapidly
`fluctuating set of conformations assumed by denatured
`proteins and other polymers in solution.
`Globular proteins consist
`largely of approximately
`straight runs of secondary structure joined by stretches of
`
`polypeptide that abruptly change direction. Such reverse
`turns or B bends (so named because they often connect
`successive strands of antiparallel B sheets) almost always
`occur at protein surfaces; indeed, they partially define these
`surfaces. Most reverse turns involve four successive amino
`
`acid residues more or less arranged in one oftwo ways, Type
`I and Type II, that differ by a 180° flip of the peptide unit
`linking residues 2 and 3 (Fig. 7-22). Both types of B bends
`are stabilized by a hydrogen bond although deviations from
`these ideal conformations often disrupt this hydrogen
`bond. Type I B bends may be considered to be distorted
`sections of 3,0 helix. In Type II B bends, the oxygen atom of
`residue 2 crowds the C,, atom ofresidue 3, which is therefore
`usually Gly. Residue 2 of either type ofB bend is often Pro
`since it can facilely assume the required conformation.
`Almost all proteins of >60 residues contain one or more
`loops of 6 to 16 residues that are not components of hel