`JAN 311991
`
`PFIZER EX. 1614
`Page 1
`
`
`
`Journal of Molecular Biology
`
`Editor-in-Chief
`P. Wright
`Department of Molecular Biology, Resear·ch 1 nstitute of Scripps Clinic·
`10666 N. Torrey Pines Road. La Jolla, CA 92037, U.S.A.
`
`Assistant Editor
`J . Ka.rn
`MRC Laboratory of Molecular Biology
`Hills R oad, Cambridge Cl32 2QH . U.K.
`
`Founding Editor
`Sir .John K endrew
`
`Consulting Editor
`Sydney Brenner
`
`Edito rs
`A. R. Fersht, University Chemical Laboratory, Cambr·idge University, Lensfield Road, Cambridge CB2 l EW, U. l<.
`Jll. Gottesman, Institute of Canuer I-tesear·ch, College of Physicians & Surgeons of Columbia Univer'Sity,
`701 W. !68th Street, New York, NY 10032, U.S.A.
`P. ron H ippe{, Institute of Molecular Biology, Univers it.y of OrPgon, Eugene, OR 9740:3- 1229, U.S.A.
`R. Huber, Max-Planck-lnstitut fiir Biochemic. 8033 Ma rtinsried bei Miinchen , Germany.
`A. Klug, MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH. U.K.
`J/ . Y1tnir. Depart.ment of Biotechnology, Pasteur Institute. 25 rue de Dr Roux , 757 24 P aris Cedex L5, France.
`
`Associate Editors
`C. R. Cantor, Human Genome Center, Donner Laborat;ory, La wrence Berkeley Laboratory. University of California.
`Berkeley. CA 94720, U.S.A.
`N.-H. Chua, The Rockt>feller University , 1230 York Avenue, New York. NY 10021, U.S.A.
`F. E. Cohen, Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, San Francisco,
`CA 94143--0446. U.S.A.
`D. J . DeRosier, R.osenstiel Basic Medical Sciences Research Center, Brandeis Univer·sity, Walt ham, MA 02254, U.S.A.
`W. A . Hendrichon, Department of Biochemistry & Molecular Biophysics. College of Physicians & Surgeons of
`Columbia University, 630 '..Vest 168th Street. New York, NY 10032 . U.S.A.
`1.8 . Holland. Jnstitut de Genetique et Microbiologie, Batiment 409, Universite de Pari>:> X T, 9 1405 Orsay Cedex 05,
`France.
`B. Honig, Department of Biochemistry & Molecular Biophysic·s, College of Physicians & Surgeons of Colombia
`University, 6:30 West !68th Street, New York, NY 10032, U.S.A.
`fl. E. fl~txley, Rosenstiel Basic Medical Sciences Research Center, Bmndeis University, Walt ham, MA 02254, U.S.A.
`V. /..;uzzali. Centre de Genetique Moleculaire, Centre National de Ia R.echerche Scientifique, 91 Gif-sur-Yvette, France.
`J. D. Mandel, Laboratoire de Cenetique Mo leculaire des Eucaryotes du CNRS, Instit ut de Chimie Biologique,
`Faculte de Medecine, II Rue Humann, 67085 Strasbourg Cedex, France.
`B. Mal/hews, Institute of Molecular Biology, University of Oregon, Eugene, OR 97403- 1229, U.S.A.
`J . H . Mille·r. Department of Microbiology, University of Califor·nia., 405 H ilgard Avenue. Los Angeles. CA 90024, U.S.A.
`M. F. Moody. chool of Pha rmacy. University of London. 29/39 B runswick Square, London WC JIIi' lAX . U . .K
`T . Richmond, Jnstit ut ftir Molekula rbiologie und Biophysik, Eidgenossische Technische Hochschule, Hiinggerberg,
`CH 8093 Zurich, Switzerland.
`R. Schleif, Biology Department, J ohns Hopkins University, Char·les & 34th Str·eets, Baltimore. M 0 212 18. U.S.A.
`N . L. Sternberg, Central Research & Development Department. E. 1. du Pont Nemours & Company, Wilmington.
`DE 19898. U.S.A.
`K . R. Yamamoto, Department of Biochemistry and Biophysics, School of Medicine, University of C'alifomia.
`Sun F ra.ncisco, CA 94143- 0448. U.S. A.
`M. Yanag£da. Department of Biophy~ics, Farult.y of Science. Kyoto University. Sakyo-Ku, Kyoto 60o. Japan.
`
`Editorial Office
`G'. flarri~. J ournal of Molecular Biology, IOd St Edwa rds Passage, Cambr·idge CB2 3 PJ, U.K.
`
`,JOUR . • AL OF MOLECULAR BIOLOGY: JSSN0022- 2836. Volumes 217- 222. 1991 , published t wice a month on t he
`5th and 20th by Academic Press at 24- 28 Oval Road, London NWI 7DX, England. Annual subscription p rice including
`postage: £852 U.K. and U.S.$ 1596 overseas. Personal subscription rate: £23:3 U.K. and U.S. $:350 overseas. ubscription
`order'S should be sent to Academic Press Limited. Foots Cray, Sidcup, K ent DA 14 5HP. U.K . (Tel: 081- 300 3322). Send
`notices of changes of address to t he publisher at least 6- 8 weeks in advonce, including both old and new addresses.
`Second class postoge rate pair! at .Jamaica. NY 114:ll. U.S.A.
`Air freight and mailing in t he U.S.A. by Publications Expediting I nc., 200 Meacha m Avenue, Elmont, NY 11003, U.S.A.
`U.S.A. POSTMA TERS: send change of addresses to ,JOURNAL OF MOLECULAR BIOLOGY. c/o Publications
`Expediting, Inc., 200 Meacham Avenue, Elmont , K'Y l l00:3, U.S.A.
`Pr·i nted in U.K.
`
`PFIZER EX. 1614
`Page 2
`
`
`
`...
`
`J. Afol. B-ioi. ( 1991) 217, 133- 151
`
`Structure, Function and Properties of
`Antibody Binding Sites
`
`I. Saira Mian1t, Arthur R. BradwelJl and Arthur J. Olson1:j:
`
`'Department of Molecular Biology
`Research 1 nstitute of Scripps Clinic
`10666 North Torrey Pines Road, La Jolla, CA 92037, U.S.A.
`2Depa.rtment of Immunology
`Medical School, Birmingham University, BI5 2TJ, U.K.
`
`(Received 16 May 1990; accepted 19 J~tly 1990)
`
`Do antibody combining sites possess general properties that enable t hem to bind different
`antigens with varying affinjties and to bind novel antigens? H ere, we address this question
`by examining t he physical and chemical characteristics most favourable for residues
`i1wolved in antigen accommodation and binding. Amphipathic amino acids could readily
`tolerate the change of environment fmm hydrophilic to hy drophobic that occurs upon
`antibody-antigen complex formation. Residues that are large and can participate in a wide
`variety of van der Waals' and elect rostatic interactions would permit binding to a range of
`antigens. Amino acids with flexible side-chains couJd generate a structurally plastic region,
`i.e. a binding site possessing t he a bility to mould itself around the antigen to improve
`complementarity of the interacting surfaces. Renee, antiibodies could bind to an array of
`novel a ntigens using a limited set of residues interspersed with more unique residues to
`which greater binding specificity can be attributed. An individua l antibody molecule could
`thus be cross-reactive and have the capacity to bind structurally similar Ligands. The
`accommodation of var·iations in antigenic structure by modest combinjng site flexibility
`could make an important contribution to immune defence by allowing antibody binding to
`djstinct but closely related pathogens.
`Tyr and Trp most readily fulfi l these catholic physicochemical requirements and thus
`would be expected to be common in combining sites on theoretical grounds. E xperimental
`support.. for this comes from t hree sources, (l ) the high frequency of participation by t hese
`amino acids in t he antigen binding observed in six crystallographically determined
`antibodly- antigen complexes, (2) their freq uent occurrence in the putati ve binding regions
`of antibodies as determined from structural and sequence data and (3) t he poten tial for
`movement of their side-chains in known antibody binding sites and model systems. The six
`bound a ntigens comprise two small different haptens, non -overlapping regions of the same
`large pt·otein and a 19 amino acid residue peptide. Out o f a total of 85 complementarity
`determining region positions, only 37 locations {plus 3 framework) a re directly in volved in
`antigen interaction. Of t hese, light chain residue 9 1 is utilized by all the complexes
`examined, whilst light chain 32, light chain 96 and heavy chain 33 are emp loyed by five out
`of the six. T he binding sites in known antibody- antigen complexes as well a-s t he postulated
`combining sites in free Fab fragments show simila r characteristics with regard to the types
`of a mino acids present. The possible role of other amino a.cids is also assessed. P otential
`implications for the combining regions of class I major histocompatibility molecules and the
`rationa l design of molecules are discussed.
`
`t Present address: Sinsheimer Laboratory , Biology
`Department, University of California Santa Cru7., Santa
`Cruz, CA 95064, U .S.A.
`t Aut hor to whom reprint requests should be
`addressed.
`§Abbreviations used: CDR, complementarity
`determining region; M HC, major histocompatibility
`complex; FR, framework.
`
`0022- 2836{91{0 I 0 13 3- 19 $03.00/0
`
`l. Introduction
`Antibodies are powerful recognition a nd binding
`molecules that the immune system employs to
`eliminate foreign molecules. Antibody binding sites
`a re formed by six hypervariable loops OJ' comple(cid:173)
`mentarity determining regions (CDRs§). The CDRs,
`133
`
`© 1991 Academic Press Limited
`
`PFIZER EX. 1614
`Page 3
`
`
`
`134
`
`1. S. Mian et al.
`
`three from each of the heavy a nd ljght chain vari(cid:173)
`able doma ins, a re connect ed to a relatively in var(cid:173)
`iant P-sheet fra mework (Aizari et al., 1988; Davies &
`Metzger, 1983; Capra & Edmundson, 1977; Wu &
`Ka bat , 1970). E arly a na lysis of a data bank of
`complete and pa rtial sequences of 415 light and 197
`heavy chains demonstrated that CDRs are rich in
`aromatic
`residues
`(K a bat et al., 1977). The
`combining region represen ts only a sma ll part of the
`antibody molecule, whose overall three-dimensional
`structure is highly conserved . Alt hough, the pairing
`of light and heavy chains can genera te some anti(cid:173)
`body diversity, most of it is generated by the
`somatic recombination o f variable region gene
`segments (Yancopoulos & AIL, 1986; Wysocki &
`Gefter, 1989). Such genetic mechanisms yield anti(cid:173)
`bodies exhibiting extensive diversity in hyper(cid:173)
`varia ble loop sequences. This potentia l repertoire is
`estimated to be approximately 109 in mouse (Berek
`et al., 1985). However, t he initial repertoire that
`confron ts an antigenic cha llenge is smaller t han the
`potentia l repertoire, since it is restricted to the
`ant ibody
`specificities
`expressed
`on
`existing
`immunocompetent B cells at a point
`in
`time
`(Holmberg et al., 1986). This available repertoire
`can y ield an apparently unlimited repertoire of
`antigen binding specificities and affinities.
`Although a single ant ibody has a unique three(cid:173)
`dimensiona l stn teture, bio physical a nd biochemical
`evidence indicates that it is mult ispecific or cross(cid:173)
`reactive (Richa rds et al., 1975). This capacity to
`combine both wit h its inducing antigen and with
`ant igens of similar or dispa rate structm:e augments
`t he genetically determined ant igen-binding capabili(cid:173)
`ties of antibodies. The extent of molecula r comple(cid:173)
`mentarity between determinants on the ant igen
`molecule and a mino acid residues in the combining
`site determines t he degree of antibody specificity.
`Increased cross-reactivity,
`therefore,
`is at the
`expense of specificity and affini ty.
`An improved understanding of both a nt ibody
`c;ross- reactivity and biJ1ding can be obtained by a
`study of antibody- antigen
`interactions at the
`atomic level. The role of residues in the defini tion of
`combining site structuJ'e and
`inter·action with
`ant igen can be assessed as a function of the chemical
`and structural p ropert ies o f individual a mino acids.
`First, we exa mine those characterisLics that appear
`to be of gener·a l importance in antibody- antigen
`interactions. This is followed by a detailed study of
`t he binding sites in six ant ibody- antigen complexes
`and four free Fab fragmen ts of known three-dimen(cid:173)
`sional structure, and the much la rger data base of
`ant ibody sequences. Padlan ( 1990) has performed a
`similar, t hough not ident ical, ana lysis of antibody
`combining sites in genera l, and three anti-lysozyme
`an tibody- ant igen complexes in pa rticular. On t he
`basis of their propensity to occur in the combining
`sites and their greater exposure relative to t hose in
`the framework regions, he has suggested that these
`amino acids determine specificity. Our results and
`their interp retation lead us to conclude that Tyr
`resid ues may play more genera lly irnport,ant roles in
`
`and
`bind ing
`interactions.
`
`non-specific
`
`antibody-antigen
`
`2. Physical and Chemical Properties
`of Amino Acids
`Since ant ibody binding sites are formed by six
`hypervariable
`loops
`supported on
`a
`highly
`conserved P-sheet framework, t here is likely to be a
`bias towards amino acids t hat are genera lly fo und in
`non-helical regions of proteins. Figure 1 shows t he
`norma lized fTequencies of occurrence of amin o acids
`in a-helix, P-sheet a nd reverse turns in 66 globula r
`proteins comprising 31 djfferent conformations
`(Levitt, 1978). In these structures, the occu rrence of
`Pro, Gly, Ty r, Ser, Thr, Asn, Val, Arg, lie and Trp
`in a-helices is less frequent t han random. Leu, His,
`Trp, Thr , Ty r, Phe, Ile and Val have a greater than
`random pr·obability of occurring in P-sheet s; the
`same .is t rue for Thr, Tyr, Asn, Ser, Asp, Gly and
`P ro in reverse turns. Ar·g appears to be equally
`tolerated in all the seconda ry st ructures elements
`consider·ed . J n general structural terms, Ty r a nd
`T hr seem to be the most useful non-helix form ing
`residues, since they could be positioned in either t he
`strand or turn regions of the hy pervariable loops.
`The free energy of interaction between an anti(cid:173)
`body a nd its a ntigen is a function of both enthalpy
`and ent ropy. Non-bonded forces between t he inter(cid:173)
`acting molecules include hydrophobic. hydrogen
`bond, ''an der Waals' and electrostatic interactions
`(for a review, see Fersht, 1985). I n general terms,
`a ntibody combining site residues need to be as
`multifaceted as possible to accommodate t he varied
`stereochemical and electronic
`features of
`t he
`ant igen . Hence, a mino acids with non-pola r (for
`exa mple Leu, Tie and Val) and charged (for example
`Asp, Glu , Lys a nd Arg) side-chains would be of more
`limited usefulness than, fo r example, His, which is
`known to be capable of cross-linking sequentially
`distant but spat ially close regions of proteins (Ba ker
`& H ubbard , 1984; J.S.M. & A.J .O., unpu blished
`results). Similarly , the amides Asn and Gin would be
`generally more preferable t ha n Asp and Glu, since
`the former pair are both hydrogen bond donors a nd
`acceptors whereas t heir charged counterparts are
`only acceptors.
`If a posit ive cha rge is required in the antibody
`combining site, Arg would be more suitable than
`Lys because of its greater fu notional versatility; for
`example, Arg can form a larger nu mber of hydrogen
`bonds t han Lys. As a consequence of its planar
`nature and n-electron system, the terminal gua nidi(cid:173)
`nium g rou p of Arg often exhibits pseudo-aromatic
`behaviour by participating in most of the inter(cid:173)
`actions
`previously
`catalogued
`for
`true
`a romatic- a romatic interactions (l .S.M. & A.J .0. ,
`unpublished results). These interactions occur at t he
`intersubunit interfaces of a number of oligomeric
`proteins ,
`including viml coat proteins and a
`membmne protein; the photosynt hetic reaction
`centre of Rhodopseudonwnas viridi8. The ab i)jty to
`fo rm hydrogen bonds, hydr·ophobic interactio ns and
`
`PFIZER EX. 1614
`Page 4
`
`
`
`Antibody Binding Sites
`
`135
`
`'".-----.------.----,.----,.----,-----.
`
`2 ·0 , - - - - - --,.-- - - - - - , - - - - - - - , - - - - -- - . . - - - - - - .
`p
`
`1·4
`
`0·8
`
`G
`
`w
`
`H
`
`0
`
`N
`
`0 c
`
`0
`
`N
`
`c
`
`1 5
`
`0·5
`
`G
`
`M
`
`' v
`
`w
`
`H
`
`~· ~--~~--~----~----~----~----~
`o;;
`0·4
`08
`1·0
`1·2
`16
`14
`Probllbilityo-he",.
`(o l
`
`2·0 , - - - - - - . - - - - , - - - - - , - - - - - , - - - - - - . - - - - - .
`
`G
`
`~ Lo
`f
`
`0-5
`
`0
`
`N
`
`A
`
`w
`
`c
`
`H
`
`v
`
`M
`
`OO L-----L---~~--~----~----~~--~
`06
`0·&
`1·0
`1·2
`1.
`1·8
`Probabl!lltya-neltx
`
`••
`
`(b)
`
`00 ~-----L------~----~------~----~
`08
`10
`1· 2
`00
`16
`Pfottab!My f'.Sheel
`
`"
`
`lei
`
`F igure 1. Scatter diagrams showing the norma lized
`frequencies of occurrence of amino acids m a-helix,
`fJ-sheet and reverse
`tums
`in 66 globular· proteins
`comprising 31 differ-ent conformations (Levitt, 1978).
`versus p .. sheet.
`(a) Probability of forming a -helix
`(b) Probability of forming a-helix versus reverse turn.
`(c) Probability of forming P-sheet versus reverse turn. 'rhe
`values represent t he ratio of the fraction of residues of
`each amino acid t hat occm·red in t he secondary str·ucture
`element to this fraction for all residues. To elimi•nate a
`bias towards structures that were determined more than
`once, the values were each weighted by a fac'tor of
`!/ (number of related proteins wit h same conformation).
`l'<orma lized frequencies of I indicate random occurrence,
`whilst > I
`indicate more frequent occurrence
`than
`random. The act ua l point is marked by the bottom left of
`the I letter amino acid code: A, Ala; R, Arg; N", Asn:
`D. Asp ; C, Cys; Q, Gin; E, Olu; 0, Gly; H, His; I , £1e:
`L. Leu; K , Lys; M, Met; F. Phe; S. Ser; T , Thr; \V, Trp;
`Y. Tyr; and V. Va l.
`
`attractive e lectrost atic inter·actions between pos i(cid:173)
`t ive ly cha rged gro ups a nd a ro matic rings permits
`T y r a nd Trp to in teract wit h structura lly diverse
`antigens. Anot.he r functiona l advantage in locating
`Tyr and Trp in antibody co mbining sites is that,
`unlike a mino a cids having shor ter side-cha ins, s uc h
`a.s Asn a nd Se r, t hey lack t he capacity to interact
`easily wit h othe r groups on t he antibody surface but
`a re ideally sui t,ed to interact with a nother molecule.
`The a ccommod atio n of charged a reas on
`t he
`a nt igen need not necessita t.e a n a ntibody co mbining
`site possessing a mino acids o f complementa•·y
`cha rge. Ana lysis of Arg, Lys, Glu a nd Asp s ide(cid:173)
`cha ins buried at the intermolecula r interfaces of
`oligomeric systems ind icates t ha t orien ted dipoles
`a re usua lly prefe rred over countercha rges in stabi(cid:173)
`lizing t hese buried residues (I.S.M. & A.J.O., unpub(cid:173)
`lished results). Thus , t he peptide backbone and
`polar side-cha ins of hypervaria ble
`loop residues
`co uld be d eployed to st a bilize both negatively a nd
`posit ively cha rged regions. l n some i11stances, this
`
`may be as effective a s e mploy men t of fo,·ma lly
`cha rged a mino acids: in ca ses of cha rge- c ha rge
`interactio n, t he steric effects of neig hbow·ing reg ions
`may preven t t he format ion of geome trically optima l
`ion-pa irs s uch t hat t he poten t ia lly availa ble e ne rgy
`is not fully realized.
`The non-cova lent associatio n between a nt.ibody
`a nd a ntigen requires t he re moval of water fro m
`surfaces buried by
`t he
`interacting mole:cules.
`Antibody regions involved in t his process should be
`capa ble of tolera ting both t he pola r a nd non -po la r
`en vironment.s t hat exist before complex fo rmation
`a nd upo n a n t igen binding, respectively. Individua l
`residues exposed on t he surface of t he free a nt,ibody
`can become completely or pa r tially buried in t he
`complex. In a ddition to residue a mphipa thicity,
`residue s ize might be a factor. There is a good
`corre la tion between t he surface area of amino acids
`and their free energies o f tra nsfe r fro m water to an
`o rga nic phase (Chothia , 1974, 1975; Cellles &
`Kla pper , 1978). A value o f I A2 ( I A= O·l nm) of
`
`PFIZER EX. 1614
`Page 5
`
`
`
`136
`
`I . S . Jllli an et al.
`
`AntigenA ~
`
`0
`
`M
`
`Antibody A -
`
`Antigenr\ ~
`
`w
`
`AntibodyA (cid:173)
`
`(o l
`
`( b)
`
`Antigen 8
`
`Antibody 8
`
`Antigen B'
`
`Antibody 8
`
`-10
`
`a
`E H
`
`N
`
`0
`
`~ L_--~--------~------~------~·~--~
`2!10
`200
`100
`70
`
`Surface area tA7l
`
`Figure 2. Compat·ison of the si7.e of a mino acids and t he
`affinit ies of their side-chains for water. The surface area
`(R ose et al ., 1985) comprises the mean solvent access·
`ibility for amino acid X in an ensem ble of Gly-X-Giy
`tripeptides. The hydration potent ial (Wolfenden ef, al.,
`198 1) is t he effective free energy of tra11sfer fi'Om t he
`vapour phase to dilute aqueous solut ion buffer at pH 7 of
`molecules having t he structure R-H, where R is the side(cid:173)
`chain of ea ch amino acid; for P (P ro), only t he surface
`ind icated , since no hydration potent ial was
`area is
`evaluated. ~ ide-chains wet·e modelled by the following
`compounds: A, methane; R, methy lguanidine; N , aceta(cid:173)
`mide; D, acet.ic acid: C. owtha nethiol; Q. propionamide:
`E , propionic acid; 0 , H 2 ; H, 4-methylirnidazole; l. isobu(cid:173)
`tane; L, butane; K, n-butylamine; M, ethy lrnethyl
`toluene; S, methanol; T , et.hanol; W.
`sulp hide: F ,
`3-met hylindole; Y, p-cresol: and V. propane. As a result of
`technical difficulties (Wolfenden el al., 1981), methyl(cid:173)
`guanid ine (shorter than t he side-chain of Arg by 2 methy(cid:173)
`lene groups) was employed to estimate t he value for
`propylguanid ine; t his probably leads to the hyd rophilic
`and hydrophobic. nature of Arg being over· and under(cid:173)
`estimated , respecti,,ely.
`
`surface a rea gives a hydrophobic energy of
`25 calfmol (I cal = 4·1 84 J : Chothia , 1974). Whilst
`these van der Waals' energies may be small
`compared to a hydrogen bond , when summed over
`the entire combining site they may be important in
`stabilizing the complex. Figure 2 compM·es the
`affinities of amino acid side-chains
`for water
`(Wolfenden et al., 198 1) with t he surface area of t he
`entire a mino acid (Rose et al., 1985). The classical
`groupings into small, large, hydrophobic and hydro(cid:173)
`philic amino acids are evident . With respect to
`amphipathicity, Ser, Thr, T yr and Trp seem desir(cid:173)
`able residues to locate in ant ibody binding sites,
`since their side-chains are in the midrange of hydro(cid:173)
`gen potent ial values. The aromatic residues Tyr and
`Trp a re also two of t he la rgest and are capable of
`contt·i bnting significantly to the total interaction
`energy (Fig. 2).
`
`Figure 3. A diagram illustrating specificity and cross(cid:173)
`reactivity for a given ant i body. The specific binding of t he
`ant ibodies A and B to a111t igens A and B is a function of
`the high degt·ee of complementa.rity between
`their
`molecular surfaces in terms of shape, size and functiona(cid:173)
`lity. (a) Cross-reactivity may arise a~ a result of structural
`similarity of epitopes between a ntigens A and A'
`(Richa rds et tll.. 1975). A poor fit in one region may be
`compensated for by a good fi t elsewhere. This could result
`in a sufficient number of short-range interactions to
`p roduce a stable a nt ibody- ant igen comp lex. Another
`the antibody
`cross-reacting antigen , A", may fit
`combining site in a slight ly different way. (b) The anti(cid:173)
`body B may accommodate the related ant igens B and B'
`if it is able to vary the stereochemical features of the
`combining site. i.e. if it is int rinsically pliable.
`
`In an ant ibody- ant igen complex, the stabiliza(cid:173)
`tion energy gained from the va rious intermolecular
`forces must mor·e tha n offset losses due to conforma(cid:173)
`tional entropy and conformational strain. The free
`energy of complex formation could therefore be
`maximized by minimizing the loss of conformationa l
`en tropy upon associat ion. I t is known that a single
`antibody is able to combine with a spectrum of
`different antigens (Richards et al., 1975). Alt hough
`such cross-reaction may occur either because the
`ant igens share epitopes, or because the epitopes are
`sufficiently similar in shape to bind t he same ant i(cid:173)
`body (Richards et al., 1975), it couJd arise a lso if the
`topogmphy of the combining site could be modu(cid:173)
`lated (Fig. 3). Thus, antibodies could utilize amino
`acids whose side-chains were sufficiently structur·
`ally and fun ctionally flexible to permit t hem to alter
`the stereochemical features of t he combining site
`wit h minima l loss of entropy. The potentia l impor(cid:173)
`tance of side-chain motion ha.s been further hig h·
`lighted by a recent comparative study of known
`antibody structures and sequences (Chothia el al.,
`1989). It has been suggested that the number of
`main-chain conformations of at least five of t he six
`loops appears to be limited . T he adoption of a
`specific backbone conformation is believed to be a
`reflection of only a few key conserved residues in the
`loop or fmm ework of the a ntibody (Chothia et al.,
`1989). T his sm all repertoire of canonical structures
`would represent a reduction in the s pectrum of
`specificity and affini ty potentially available to t be
`antibody binding site were the n umber of conforma(cid:173)
`tions proport ional to the number of sequences that
`could be produced genetically.
`
`PFIZER EX. 1614
`Page 6
`
`
`
`Antibody Binding Sites
`
`137
`
`Table I
`Preferred conformatio'M of arnino acid side-chai'M as described by their torsion pararneters (Cody, 1985)
`
`Residue
`
`A
`R
`
`N
`D
`c
`
`Q
`
`E
`
`G
`H
`
`1.
`
`K
`M
`
`~~
`
`J>
`~
`T
`
`w
`
`y
`
`v
`
`"'
`
`5; - 19
`- 15
`
`II
`-7; ±35
`-30
`
`- 20
`
`0; ±3.5
`
`(I; 25
`-25
`
`- 15; - 45
`
`-
`
`IS: ±36
`
`-20
`±20
`
`±20
`
`± 10
`5
`-25
`
`- 10
`
`±20
`
`-
`
`II ; - 35
`
`x"
`
`130
`- 60
`60
`- 60
`- 60
`60
`60
`- 60
`180
`-60
`60
`
`- 60
`60
`60
`ISO
`- 60
`180
`-60
`60
`-60
`ISO
`- 60
`60
`180
`-60
`130
`± 35
`60
`-60
`60
`- 60
`60
`-60
`180
`60
`- 60
`
`xll
`
`Torsion parameters (")
`x22
`x21
`
`- 120
`- 120
`
`180
`180
`ISO
`
`180
`180
`0
`±5
`
`180
`180
`180
`180
`ISO
`
`60
`60
`180
`ISO
`180
`-60
`-60
`60
`180
`ISO
`180
`180
`90
`90
`90
`±35
`
`90
`90
`90
`90
`90
`
`ISO
`- 60
`180
`
`180
`- 60
`
`180
`
`xs
`
`x•
`
`±10
`±10
`
`180
`
`x'
`
`180
`180
`
`- 15
`25
`± 15
`± 15
`± 15
`
`180
`- 60
`180
`180
`
`±25
`
`± 20
`
`± 10
`
`These were derived from crystallographic studies of amino acids and t heir derivatives. All atoms are numbered using t he Greek
`the co "' = (0--C--C"-N), X II = (N-C"--c'--c' 1
`) , x21
`), x' 2
`= (N-CO-C'--G'2
`= (CO-C'-C7-C'1
`letter de~nations start ing with
`),
`), x'= (C'- C'- G" - C'). x•= (07- C'-C'-(:(). To account for t he ring pucker in Pro. x•= {C7-C"- N- Cu) and
`7.22 =(CO-
`- C'-C'2
`7.5 = (C'-~-C'-C'). The most frequently observed values are given first.
`
`The number of degrees of freedom of each amino
`acid side-cha in can be approximated by examining
`the range and distribution of t he observed confor(cid:173)
`mations (Table I). Residues with short side-chains
`su ch as Ser and Thr lose little entropy when fixed
`upon antigen association, since they have only one
`and two variable side-chain torsion angles, respect(cid:173)
`ively. However, they project only a short d istance
`from the surface of t he antibody and so could not
`effect substantial changes in binding site topo(cid:173)
`g raphy. Large residues can elicit great changes in
`tJ1e surface contours of the combining site, since
`t hey can sweep out large volumes of space. Residues
`wit h the largest surface area (Fig. 2) are Trp, Arg,
`Tyr and Phe and of these, Arg has twice t he number
`of variable
`torsion angles of the other three.
`Additionally, Arg is less suita ble than the other
`a romatic residues because it is charged and there-
`
`fore requires a more restrictive interaction at the
`interface.
`the many
`Tyr represents a balance between
`different, though sometimes conflicting, desirable
`aspects we believe to be of general importance in
`antibody- antigen interactions. Thus, it would be
`expected to be the most common residue in anti(cid:173)
`body combining sites. Experimental evidence from
`antibody sequences and structures appea rs to verify
`these assumptions.
`
`3. Combining Sites in Known
`Antibody- Antigen Complexes
`T able 2 lists t he antibody residues that bind
`antigen in six complexes whose structures have been
`by X-ray
`crystallogra phy. The
`determined
`numbering scheme used follows the convent ion of
`
`PFIZER EX. 1614
`Page 7
`
`
`
`138
`
`/ . S . Mian et al.
`
`Table 2
`Antigen binding residues observed in six crystallographically determined antibody- antigen complexes
`
`Residue position
`
`McPC603
`
`HyHEL-5 RyHEL-10
`
`Fab .01.3
`
`Fab 4-4-20
`
`8 1312
`
`Percentage comPosit ion
`
`Antibody- ant igen complex
`
`FH.2
`C.DR.2
`
`COH.3
`
`A. Light chain
`CDR. I
`27.D
`2S
`30t
`31t
`32t
`34
`49
`50t
`53
`91t
`92t
`93t
`94t
`95t
`96t
`13. H ea.~>y chain
`FRI
`30t
`CDR1
`31t
`32t
`33
`35
`47
`50
`51
`52
`52At
`53t
`54t
`55t
`56
`57
`58
`95
`96
`97
`98
`99
`100
`IOOA
`100B
`IOOI
`
`FR2
`CDR2
`
`CDR.3
`
`0 (30)
`N (31)
`N (32)
`
`y (50)
`Q (53)
`s (91)
`N (92)
`
`y (96)
`
`T (30)
`s (31)
`D (32)
`y (33)
`
`y (50)
`s (52)
`
`y (53)
`s (54)
`s (56)
`
`y (58)
`W (98)
`
`N (30)
`y (31)
`y (33)
`
`D (49)
`w (90)
`G (9 1)
`H. (92)
`
`p (94t)
`
`w (33)
`E (35)
`w (47)
`E (50)
`
`s (55)
`s (57)
`T (58)
`N (59)
`0 (99)
`
`y ( 10 1)
`
`H
`
`y
`
`y
`y
`
`F
`w
`s
`
`T
`G
`y
`
`w
`
`0
`D
`
`H.
`D
`y
`R
`
`0 (97)
`
`y ( 100)
`
`L ( 102)
`
`y (33)
`E (35)
`
`R (52)
`
`N (101 )
`
`w (107)
`
`H
`
`y
`R
`
`s
`
`w
`
`w
`
`y
`y
`
`D
`D
`
`Y
`
`G
`
`V
`
`P
`
`R
`
`A
`
`I
`s
`s
`0
`s
`y
`
`F
`y
`
`p
`F
`
`S24, Ei20, ¥ 16, 0 14
`0 17, N16, S16, Y9
`825, N23, V14, L10
`S33. N2S, T IS, H5
`Y68, N6, S4, W3
`A30, N24, H20, S9
`Y83, G9, F3, 82
`0 19, 015, K l1, RIO
`S38, K 18, Tl4, Sl I
`W27, Y22, 8 16, Gl 2
`N20, Yl8, 815, 012
`836. E 14. HIO. Y9
`817, N l6, VI2, [,12
`P63, H10, LS, S6
`Wl9, YI S, LIS, R IO
`
`848. T42, K5, G I
`S39, 034, N7, R6
`Y60, F 17, 86, T6
`Y40, G24, Wl6, A5
`H26, N21, 8 17, E15
`W94. Y3, L < l , F < I
`Yl9, E l2, A l2, Rl 2
`184, 89, Fl , Rl , VI
`N27, Rl6, 8 13, 012
`P54, N l9, 89, L4
`N29, 0 21, 8 11, AIO
`827, N26, G23, 016
`0 56, Y19, S l5, D3
`825, T24, 016, YIO
`T74, 112, K3. 82
`N20, Y l 9, Ll 15, Et:l
`D2S, G IS. S IO, YIO
`Y26, 0 12, R9, D6
`Y32, 0 14, 06, L6
`Y32, 0 19, V7, 86
`G26, Y21, Sl 6, E5
`826, Yl2. F12, R7
`823, Fl2, GIO, Y8
`Y29, F l2, O t t , 8 11
`W35, Y34, AIO, M6
`
`McPC603 (8atow et al., 1986), HyHEL-5 (Sheriff eJ al .. 1987), Hy HEL-10 (Padlan et al .. 1989), F ab 0 1.3 (A mit eJ al., 1986). Fl'b 4-4-
`20 (Herron e1 al .. 19S9) and Bl312 (Stanfield et al., 1990). R-esidue positions, framework (FR) and complementarity determining regions
`(CDR) are from Kabat el at. ( 1987). The sequential residue num bers of struct ures in the Protein Data Bank (Bernstein el al., 1977):
`McPC603 (File 2MCP ), HyHE L-5 (I HFL) and HyREL-lO (3HJ<'M), are indicated in parentheses. At each of the position.s known to bind
`antigen, the column headed Percentage composit ion indicat-es the 4 most common amino acids and their frequencies as calculated from
`the summary tables in K abat et al. (1987). For example, one of the fluorescein binding residues in Fab 4-4-20 is H is L27D (located in
`CDR I). Examination of the sequence database at t his loca.tion indicates that Ser, His, Tyr and Gly are found in 74% of all sequences
`(24%, 20 %. 16% and 14 % . respectively) wit h the remaining 26 % comprising t he ot her 16 a.mino acids (eacb is < 14% of the total).
`Asn, Asp, Gin and Glu are underestimat-ed becaus-e positions given as Asx or Glx by Kabat et al. ( 1987) (i.e. where t here are uncertaint.ie.•
`in sequence data), whilst being included in the total number of sequences, are not incorporated into totals for these 4 amino acids.
`Alt hough positions L95 and L96 are listed separately, exa,mination of the crystal structures of McPC603, Hy HEJ,-5 and H yHEL-10
`indicates t hat Leu L96 (102), Pro L95 (94) and 1'yr L96 (96) are structurally equivalent (see also Fig. 5). In HyHEL-lO, Ser L93 (93)
`and Trp L94 (94) are listed as being p artially buried by the antigen lysozy me (Padlan eJ-al. , 1989). l n Bl312, Leu L270, Oys H32. Trp
`H47, Oly H50. Ser H 96. Ser H97 and Asp H98 a re buried by, but not in van der Waals' contact wit h, t he antigen (Stanfield et al., 1990).
`Also marked are residues in the hypervariable and framework regions t hat a comparati ve study of known antibody structures and
`sequences {Chothia. et al., 1989) has suggested as being important (tl and mainly responsible (t) for generating t he observed main-chllin
`conformations of 5 of the 6 hypervariable loops (predictions for CDR.3 of t he hea vy chain were not made). Amongst all the antigen
`binding residues in t hese complexes, only Pro L95 (94) of HyliEL-5 belongs to the group of key residues. Ot her posit ions that are
`mainly responsible for producing t he canonical structures occur at t he following posit.ions. Light chain: 2 (1/4, i.e. out of a total of 4
`classes in which this is important for the observed conformation, only in canonical st ruct ure number I is this a residue that is a primary
`determinant of the loop main chain conformation), 25 ( 1/4), 29 (l/4). 33 {1/4), 48 (1/ 1 ), 64 (1/ 1 ), 7 1 (1/4), 90 ([I , 2, 3]/3), 95 ([1. 3]/3).
`Heavy chain: 26 (l/ 2), 27 ( 1./2), 29 (1/2), 34 (1/2), 52a (2/4), 55 ([1, 4]/4), 71 ([2.3, 4]/4).
`
`PFIZER EX. 1614
`Page 8
`
`
`
`Antibody Binding Sites
`
`139
`
`Kabat et aJ,. (1987) with L (light) and H (heavy)
`prefix