Paratope and Epitope Mapping of the Antithrombotic Antibody 6B4 in Complex with Platelet Glycoprotein Ib 

Paratope and Epitope Mapping of the AntithromboticAntibody 6B4 in Complex with Platelet Glycoprotein Ib�*□S

Received for publication, March 2, 2007, and in revised form, May 30, 2007 Published, JBC Papers in Press, June 14, 2007, DOI 10.1074/jbc.M701826200

Alexandre Fontayne‡1, Bauke De Maeyer‡2, Marc De Maeyer§¶, Mayo Yamashita�, Tadashi Matsushita**,and Hans Deckmyn‡¶3

From the ‡Laboratory for Thrombosis Research, IRC, the §Laboratory of Biomolecular Modeling, Department of Chemistry, and¶BioMacS, KU Leuven Campus Kortrijk, E. Sabbelaan 53, B-8500 Kortrijk, Belgium, the �Department of Medical Technology, NagoyaUniversity School of Health Sciences, Nagoya 466-8560, Japan, and the **Department of Hematology, Nagoya UniversityGraduate School of Medicine, Nagoya 466-8560, Japan

The monoclonal antibody 6B4 has a potent antithromboticeffect in nonhuman primates by binding to the flexible loop, alsoknownas the�-switch region (amino acids 230–242), of glycopro-tein Ib� (GPIb�). This interaction blocks, in high shear stress con-ditions, the specific interaction between GPIb� and von Will-ebrand factor suppressing platelet deposition to the damagedvessel wall, a key event in the pathogenesis of arterial thrombosis.To understand the interactions between this antibody and its anti-gen at the amino acid level, we here report the identification of theparatope and epitope in 6B4 and GPIb�, respectively, by usingcomputer modeling and site-directed mutagenesis. The dockingprograms ZDOCK (rigid body docking) and HADDOCK (flexibledocking) were used to model the interaction of 6B4 with GPIb�and to delineate the respective paratope and epitope. 6B4 andGPIb�mutants were constructed and assayed for their capacity tobindGPIb� and6B4, respectively. From these data, it is found thatthe paratope of 6B4 is mainly formed by five residues: Tyr27D,Lys27E,Asp28, andGlu93 located in lightchainCDR1and-3, respec-tively, andTyr100C of theheavy chainCDR3.These residues formavalley, where the GPIb� flexible loop can bind via residues Asp235and Lys237. The experimental results were finally used to build amore accurate docking model. Taken together, this informationprovides guidelines for the design of new derivatized lead com-pounds with antithrombotic properties.

Platelets are a key factor in hemostasis (1). However, in somepathological situations, such as stroke ormyocardial infarction,shear rate increases, causing platelet activation and thrombusformation, leading to vessel occlusion. This process is depend-ent on the binding of the platelet glycoprotein Ib� (GPIb�)4 to

von Willebrand factor (VWF), which is bound to the collagenmatrix exposed to the flowing blood upon vessel damage.The structure of GPIb� consists of a globular N-terminal

region, a sialomucin core, an anionic sequence, a transmem-brane region, and a cytoplasmic tail. The N-terminal region(residues 1–282) consists of eight leucine-rich repeats (LRRs)and contains the binding sites for VWF, �-thrombin, P-selec-tin, Mac-1, high molecular weight kininogen, and coagulationfactors XI and XII (2–4). Under nonliganded conditions, plate-lets present the flexible loop (residues 230–242) within theN-terminal domain of GPIb� into a �-switch conformation,which changes upon binding of VWF into a �-hairpin confor-mation, extending the existing VWF antiparallel �-sheet (5).The GPIb� gain-of-function mutations G233V and M239Vfound in platelet-type von Willebrand disease stabilize the�-hairpin conformation and increase the affinity of GPIb� forVWF 5–6-fold (4, 6). The globular domain is presented wellabove the plasma membrane by the sialomucin core, which isconnected by a flexible hinge domain: the anionic sequence.The cytoplasmic tail of GPIb� contains binding sites for filaminA and 14-3-3 �, which play an important role in intracellularsignaling upon ligand binding (3, 7, 8).Previously, we prepared and characterized a murine mono-

clonal antibody (mAb) targeting the human GPIb�, designatedas 6B4 (9). This mAb inhibits platelet adhesion under highshear stress conditions, as was shown in flow chambers (10).Injection of 6B4-Fab fragments has a potent in vivo antithrom-botic effect in baboons (9, 11) but also on inhibiting ex vivoristocetin-induced platelet aggregation (9). Contrary to mostantithrombotic drugs, 6B4-Fab administration did not induce asignificant prolongation of the bleeding time. The epitope rec-ognized by 6B4 was mapped previously, using human/caninechimeric rGPIb�, to be within the C-terminal flanking region,between residues 201 and 268 (10), containing the flexible loop(residue 230–242) within theN-terminal domain of GPIb�. Anindication that upon binding of 6B4, this loopmight not assumethe �-hairpin conformation, as seen upon binding of VWF,comes from the finding that 6B4 no longer binds to the gain-of-function G233V and M239V (5, 10). The goal of this study wasto further determine which residues are involved in the bindingof 6B4 to GPIb�.Docking approaches using computer programs such as

ZDOCK (12), an algorithmmore appropriate as an initial stagedocking algorithm to explore vast putative binding areas in

* This work was supported by Instituut voor de Aanmoediging van Innovatiedoor Wetenschap en Technologie in Vlaanderen Grant IWT 020473 and bythe Sankyo Foundation of Life Science. The costs of publication of thisarticle were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Table S1.

1 An EU-Research Training Network (HPRN-CT-2002-00253) postdoctoral fellow.2 A bursary of the IWT.3 To whom correspondence should be addressed. Tel.: 32-56-246422; Fax:

32-56-246997; E-mail: [email protected] The abbreviations used are: GPIb�, glycoprotein Ib�; VWF, von Willebrand

factor; LRR, leucine-rich repeat; CDR, complementarity determiningregion; Ab, antibody; mAb, monoclonal antibody; WT, wild type.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 282, NO. 32, pp. 23517–23524, August 10, 2007© 2007 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23517

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

http://www.jbc.org/cgi/content/full/M701826200/DC1

Supplemental Material can be found at:

cases were the target binding site is unknown or less defined,and the HADDOCK1.3 method, which allows flexibility (13–15) in both ligand and target, were used to predict interactionsbetween ligand-bound or ligand-free GPIb� and variousmodels of 6B4, followed by mutagenesis experiments in aniterative approach. By this method, we identified residues inthe complementarity determining regions (CDRs) of 6B4,crucial for the binding and constituting its paratope. In par-allel, the epitope mapping on the flexible loop of GPIb� wasconfirmed and refined. An optimized interaction model wasfinally constructed combining all of the findings.

EXPERIMENTAL PROCEDURES

Materials—Protein A-Sepharose CL-4B and ECLTM werepurchased from Amersham Biosciences. The QuikChange XLsite-directed mutagenesis kit was provided by Stratagene (LaJolla, CA), and theDNA sequencingwas performed byGenomeExpress (Meylan, France). Primer sequences listed in Table 1were purchased from Eurogentec (Seraing, Belgium). Cell cultureproducts for the IgG4 expression andLipofectamine 2000TMwereprovided by Invitrogen.Centriprep-30 andCentricon-100 deviceswere provided byMillipore (Billerica,MA). Themonoclonal anti-human IgG4, Fc-specific Ab was from BD Pharmigen (San Diego,CA), the anti-human IgG horseradish peroxidase-labeled Ab waspurchased from Imtec Diagnostic (Antwerpen, Belgium), and thegoat anti-mouse horseradish peroxidase-labeled Ab and theorthophenylenediamine were from Sigma.Docking—Docking of the 6B4-Fab to GPIb� (amino acids

1–266) was performed using the software ZDOCK (12) andlater HADDOCK1.3 (14). Results were analyzed with the

BrugelTM modeling package (16). The starting structures forthe docking were four computermodels of 6B4-Fab; onemodelwas built by Algonomics (17), and three models were retrievedfrom theWebAntibodyModeling tool (18), the structure of theligand-free conformation of GPIb� (Protein Data Bank code1M0Z) (5) and the structure of the ligand-bound conformationofGPIb� (ProteinData Bank code 1SQ0) (19). TheHADDOCKoption to use structural information from all models in thesame run was selected. As a guide to the docking, a list of pos-sible interaction sites was used. HADDOCK divides the inter-acting residues into two classes: active residues, which play animportant role in binding, and passive residues, which may beindirectly involved in the binding.On themAb (the ligand) side,it is obvious that the CDRs are involved in the binding. Previousexperimental evidence indicated that in GPIb�, onemajor loop(the flexible loop at amino acids 228–242) is involved in thebinding (10). The solvent-accessible residues of the CDRs andthe GPIb� flexible loop were defined as passive residues in thedocking experiment. The preliminary ZDOCK (12) and conse-quent site-directedmutagenesis had indicated that Y27DA andE93A (Kabat numbering (20)), situated in the light chain CDR1and -3, respectively, contribute directly to the binding. There-fore, these residues were defined as being active. In the dockingexperiment, only the variable domains of the mAb were used.Since these were not covalently bonded, distance restraintswere used to keep the light and heavy chains together duringthe simulated annealing phase of the docking.Production and Purification of Monoclonal Antibodies—The

mAbs 6B4, 27A10, and 24G10 were developed in mice

TABLE 1Candidate residues for mutagenesis

MutantaRationale for mutation Primerb

Name MutationVL Y27DA Tyr3 Ala Rupture of salt bridge 5�-AGGTCTAGTAAGAGTCTCCTAGCAAAGGATGGGAAGACATACTTG-3�

K27EA Lys3 Ala Rupture of salt bridge 5�-CTAGTAAGAGTCTCCTATATGCGGATGGGAAGACATACTTG-3�K27EE Lys3 Glu Enhanced effect 5�-CTAGTAAGAGTCTCCTATATGAGGATGGGAAGACATACTTG-3�D28A Asp3 Ala Rupture of salt bridge 5�-GTAAGAGTCTCCTATATAAGGCCGGGAAGACATACTTGAATTG-3�D28R Asp3 Arg Idem � enhanced effect 5�-GTAAGAGTCTCCTATATAAGCGCGGGAAGACATACTTGAATTGG-3�V92A Val3 Ala Negative control 5�-ATTACTGTCAACAACTTGCAGAGTATCCGCTCACG-3�E93A Glu3 Ala Rupture of salt bridge 5�-TATTACTGTCAACAACTTGTAGCTTATCCGCTCACGTTCGG-3�Y94A Tyr3 Ala Reduce contact area 5�-GTCAACAACTTGTAGAGGCCCCGCTCACGTTCGGTG 3�

VH S56A Ser3 Ala Negative control 5�-GGGAGTAATATGGACTGGTGGAGCAACAAATTATAATTCGGCTCTCATG-3�N58A Asn3 Ala Rupture of salt bridge 5�-ATGGACTGGTGGAAGCACAGCATATAATTCGGCTCTCATGTCC-3�S97T Ser3 Thr Rupture of Van der Waals 5�-CTACTGTGCCAGAGATCGAACCACCATGATTACGGCCTATG-3�I100A Ile3 Ala Reduce contact area 5�-CAGAGATCGATCTACCATGGCAACGGCCTATGCTATGGACT-3�Y100CA Tyr3 Ala Reduce contact area 5�-CTACCATGATTACGGCCGCCGCTATGGACTACTGGG-3�

a Mutated residues are located on the variable domain of the light (VL) or heavy (VH) chain of 6B4-IgG4. The name of the mutant includes the WT residue type (single-lettercode) followed by its position in Kabat numbering (number with, if needed, an insertion character) followed by the single-letter code of the mutant residue. The three-lettercode for the changes and the rationale for the mutation is also given.

b Oligonucleotide sequence of the primers used to construct light or heavy chain mutants of 6B4-IgG4 (17). The sense primer hybridizes with the noncoding strand of eitherpKaneo-CM30-Lvar or pKaneo-50-dhfr-Hleuvar, used for the preparation of 6B4 mutants. The boldface and underlined codons introduce the mutation.

TABLE 2Results from the HADDOCK docking of the 6B4 model to the resting conformation of GPIb�Results are grouped (cluster) according to the root mean square deviation. Only the three energetically best models for each cluster are reported.

Cluster 1 Cluster 2 Cluster 3Modelnumber

Interfacearea

Nonbondedenergy

Modelnumber

Interfacearea

Nonbondedenergy

Modelnumber

Interfacearea

Nonbondedenergy

Å2 kcal/mol Å2 kcal/mol Å2 kcal/mol89 771.1 �71 81 859.1 �71 25 972.8 �7623 851.5 �63 50 814.5 �50 55 872.4 �6721 917.2 �86 88 813.8 �49

Paratope Identification of the Anti-GPIb� Antibody 6B4

23518 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

against purified human GPIb� (9), purified by affinity chro-matography with protein A-Sepharose CL-4B, and dialyzedovernight at 4 °C against PBS. Antibody purity was checkedby SDS-PAGE under nonreducing conditions, followed byCoomassie Brilliant Blue staining. Concentrations wereevaluated by optical density at 280 nm, and antibodies werekept at �20 °C before use.Construction and Expression of 6B4 WT and Mutants—

Recombinant 6B4 WT and mutants were prepared as chi-meric human/murine IgG4 as previously described (17).Construction of 6B4 mutants was performed with theQuikChange XL site-directed mutagenesis kit according tothe manufacturer’s instructions using pKaneo-CM30-Lvarand pKaneo-50-dhfr-Hleuvar vectors, coding for the respec-tive chimeric light and heavy chain of 6B4, and the appropri-ate primer couple (Table 1). After DpnI digestion and bacte-ria transformation, clones positive for the presence ofplasmid DNA were selected, and their purified DNA wassequenced. All 6B4 antibodies were expressed in a transientexpression system using human embryonic kidney cell line

293T/17 and Lipofectamine 2000TM as described before(17).Purification andCharacterization of 6B4WTandMutants—

The different expressed antibodies were purified on a proteinA-Sepharose CL-4B column and dialyzed against phosphate-buffered saline overnight at 4 °C. Quality control was per-formed by SDS-PAGE and Western blot analysis using amonoclonal anti-human IgG4-Fc-specific Ab, followed by goatanti-mouse horseradish peroxidase-labeled Ab before revela-tion using ECLTM. Antibody concentration was estimated, incomparison with an IgG4 reference, by sandwich ELISA usinganti-human IgG4-Fc-specific Ab for capture and an anti-hu-man IgGhorseradish peroxidase-labeledAb for detection.mAbconcentrations were adjusted to 1 �g/ml and kept at �20 °Cbefore use.Production and Characterization of rGPIb� Mutants—WT

and mutant rGPIb� were produced in a transient expressionsystem using 293T/17 cells and Lipofectamine 2000TM. After48 h, rGPIb� secreted in the medium was concentrated usingCentriprep-30 and Centricon-100 devices. The concentration

of eachmutant was determined by atwo-step ELISA as described before(21).Binding of 6B4 WT and Mutants

to Human Platelets—The capacityof WT and mutant 6B4 to bind toplatelets was tested in an ELISA sys-tem where the antibody was added,in a dilution series of 1:2, into wellsprecoated with fixed intact humanplatelets (17). Revelation was doneby incubatingwithamonoclonalanti-human-IgG4 antibody (1:4000), fol-lowedby a goat anti-mouse horserad-ish peroxidase-labeled Ab (1:5000),before the addition of H2O2 andorthophenylenediamine, stop withH2SO4, and optical density determi-nation (490–630 nm) on a micro-plate reader. Binding of 6B4 WT atsaturation was set as 100%.Binding of rGPIb� WT and

Mutants toMonoclonal Anti-GPIb�Antibodies—Recombinant GPIb�WT and mutants were tested for

FIGURE 1. HADDOCK computer docking model of 6B4 to GPIb�. A, ribbon representation of the HADDOCKmodel number 21 complexing GPIb� (green) and its �-switch region in the “ligand-free” conformation (red)with 6B4 light chain (magenta) and 6B4 heavy chain (blue). Amino acid side chains predicted to form hydrogenor salt bridges are represented as sticks. B, focused ribbon representation showing important residues consti-tuting the 6B4 GPIb� interface. Hydrogen bonds and salt bridges are represented as dashed lines. Y100C is notshown, since it does not form hydrogen bonds or salt bridges as mentioned in Table 3. Amino acid residues thatlater were confirmed to be part of the paratope are lettered in red. All images of the three-dimensional modelswere generated with PyMOL (available on the World Wide Web).

TABLE 3Hydrogen bonds and salt bridges between 6B4 and GPIb� in model 21Heavy chain Y100C is not in the table, since it does not form hydrogen bonds.

Number6B4 GPIb�

Hydrogen bond energyAmino acid Positiona Atom Amino acid Position Atom

kcal/mol1 Ser 67 HG Glu 14 OE1 �2.52 Asp 28 OD1 Arg 64 HH22 �1.73 Lys 27E HZ3 Asp 83 OD2 �2.14 Asp 28 OD1 His 86 HE2 �3.65 Lys 27E HZ2 Asp 106 OD2 �2.26 Thr 53 O Lys 231 HZ1 �2.57 Tyr 27D HH Asp 235 OD1 �2.58 Glu 93 OE1 Lys 237 HZ3 �2.5

a Positions of the residues in 6B4 are noted in Kabat numbering.

Paratope Identification of the Anti-GPIb� Antibody 6B4

AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23519

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

their capacity to bind to coated monoclonal 6B4, 27A10, and24G10 in an ELISA set up as described previously (21).Statistical Analysis—The binding capacity of 6B4 and its

mutants to GPIb� as well as the binding of GPIb� and itsmutants to monoclonal anti GPIb� antibodies were comparedby Student’s t test. The differences were considered statisticallysignificant when p was �0.05.

RESULTS

First Model Using ZDOCK—The 6B4 paratope was tenta-tively determined by constructing computermodels of the vari-able regions of 6B4 bound to different crystal structures ofGPIb�. In a first approach, a 6B4 computer model (17) wasdocked to the ligand-free conformation of GPIb� (1M0Z.pdb)using the ZDOCKprogram (12). The resultingmodel was char-acterized by one major binding site in which the mAb binds tothe flexible loop of GPIb�. The model suggested that both thelight and heavy chain of 6B4 contribute to the binding toGPIb�.

Based on this model, seven 6B4 residues were selected formutation toAla: four on the light chain (Y27D,K27E, Val92, andGlu93) and three on the heavy chain (Ser56, Asn58, and Ile100).Mutations of Val92 and Ser56 to Ala were included as negativecontrol, since no major effect was expected. Only Y27DA andE93A were found to affect the binding of 6B4 to GPIb� (Fig. 2,A and B).SecondModel UsingHADDOCK—Since the rigid body dock-

ing method ZDOCK only allow the prediction of two interact-ing residues, we, in a second approach, used the results from thefirst round as input to construct a new docking model usingHADDOCK1.3 that allows for flexible docking in both 6B4 andGPIb�. We used the four available structures of the mAb plusthe ligand-free and ligand-bound structures of GPIb� simulta-neously in the docking experiment. The docking of 6B4 to theligand-bound conformation of GPIb� (1SQ0.pdb) did not pro-duce any acceptable model, since hardly any hydrogen bridgesbetween the two proteins were found that furthermore mainlyoccurred between main chain elements (data not shown).The docking results with the ligand-free GPIb� structure

were grouped into clusters, which are defined as an ensemble ofat least two conformations displaying a backbone root meansquare deviation at the interface smaller than 1.0 Å (14). Of theenergetically best models in each of the three clusters thusobtained (Table 2), docking model 21 in cluster 1 was selected,because both Y27D and Glu93 are predicted to bind GPIb�,which is in agreementwith our previous experimental data (Fig.1). Based on this docking model, light chain K27E, Asp28, andTyr94 and heavy chain Ser97 and Y100C were selected formutagenesis (Table 1).In thismodel, Asp28 forms four interactions, twoofwhich are

salt bridges, with the side chains of Arg64 and His86 located intheGPIb�LRR2 andLRR3, respectively. To disrupt these inter-actions, Asp28 was not only mutated to Ala but also to the pos-itively charged Arg (Table 1). Furthermore, also in this model(and in most generated models) K27E is predicted to play animportant role in binding, since it forms two ionic interactionswith Asp83 and Asp106 of GPIb� (Table 3, lines 3 and 5). In thefirst round, however, somewhat unexpectedly, mutation of

K27E to Ala did not inhibit the binding of 6B4 to GPIb� (Fig.2A), so also here we made a second mutant in which the nega-tively charged Asp is introduced instead. All 6B4 mutants,except for Y94A, which only was expressed in very low quanti-ties, were tested for their capacity to bind to immobilizedhuman platelets (Fig. 2A). Of the five new mutants tested, onlyS97T still bound normally to GPIb� on platelets. Fig. 2B clearlyshows that six mutants (Y27DA and E93A from round 1 and

FIGURE 2. Paratope resolution of 6B4. Comparison of 6B4 WT (purple) andmutants for the binding to GPIb� on human platelets (A and B). Differentconcentrations (A) of purified 6B4 WT or mutant were added to wells pre-coated with human platelets. Bound mAb were detected with anti-humanIgG4 Ab followed by a goat anti-mouse horseradish peroxidase-labeled Ab, asdescribed under “Experimental Procedures.” 6B4 WT (�) binding was com-pared with light chain mutants (Kabat numbering) Y27DA (E), K27EA (Œ),K27EE (F), D28A (�), D28R (ƒ), V92A (�), and E93A (�) and with heavy chainmutants S56A (‚), N58A (�), S97T (�), I100A (f), and Y100CA (�). B, percent-age binding of 0.25 (open bar) or 0.5 �g/ml (black bar) purified 6B4 WT ormutant to GPIb�. Binding of 6B4 to WT was set as 100%. Data in A and B arethe mean � S.E. from three independent experiments; most of the error barsare within the size of the symbols. *, statistically different with p � 0.05. Shownis a surface representation of the variable domains of 6B4 (C) with the residuesthat are critical for binding (red), the residues where a mutation did not affectthe binding (green), and Tyr94 (blue), which was suggested but which couldnot be tested because of low expression levels. The image of the three-di-mensional model was generated with PyMOL (available on the World WideWeb).

Paratope Identification of the Anti-GPIb� Antibody 6B4

23520 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

K27EE,D28A,D28R, andY100CA from round 2) at 0.25 and 0.5�g/ml maximally allowed 25% binding of 6B4 to GPIb�,whereas the other mutants (K27EA, V92A, S56A, N58A, andI100A from round 1 and S97T from round 2) show nearly nor-mal binding. Based on the mutagenesis data, we can concludethat residues Tyr27D, Lys27E, and Asp28 of CDR L1, Glu93 ofCDR L3, and Tyr100C of CDRH3 are part of the paratope of 6B4(Fig. 2C). The three other CDRs do not seem to be involved inthe binding.Epitope Determination on GPIb�—Based on the docking

model of 6B4 bound to GPIb�, a number of residues were pre-dicted to be part of the epitope. All residues present at theantigen surface and forming hydrogen bonds or ionic interac-tions with residues of 6B4 could be putative residues of theepitope (Table 3). Several of these residues are located in theLRR: Arg64, Asp83, His86, and Asp106. Some other residuesinvolved (Lys231, Asp235, and Lys237) are part of the �-switchregion, the key element that changes its conformation duringligand binding (5).To validate our model, a set of 38 single to triple GPIb�

mutants (21) containing 62 charged residues mutated to Ala(Fig. 3) was tested for the binding to wild type 6B4. As an addi-tional control, two other inhibitory anti-GPIb� mAbs weretested, namely 24G10,which competeswith 6B4 for the bindingto human platelets (10), and 27A10 (22), which does not com-pete (Fig. 4A).Amarked impairment of the binding ofmAb 6B4was seen to

GPIb� mutants D83A/H86A (Fig. 4A, lines 3 and 4), D106A(line 5), K149A/E151A/K152A (line 8), K288A/R290A (line 9),and D235A/K237A (lines 6 and 7), which contain all of thepredicted residues except for Arg64 (line 2) and Lys231 (line 10),mutation of which did not affect binding. However, since thebinding of all three antibodies (and others)5 to the first fourmutants was similarly decreased, it is possible that these inducea conformational change in GPIb�, as previously hypothesized(21) and hence might have an indirect effect on the antibodybinding. These residues, therefore, are being treated withcaution in the description of the epitope, which on the other

hand clearly involves Asp235 and Lys237 of the �-switchregion (Fig. 4B).RefinedDockingModel—All of the results from themutagen-

esis experiments on both 6B4 and GPIb� were combined andused for another docking round with HADDOCK1.3, usingagain the four 6B4 models and the ligand-free and ligand-bound GPIb� conformation. Results were ranked by interfacearea and energy contribution to the complex formation andvisually analyzed. Also, here the only acceptable docking mod-els could be made with the ligand-free GPIb� conformation, ofwhich the best model, number 61 (Fig. 5 and supplementalmaterial), is very close to the previous model 21. In model 61,6B4 is rotated and translated into the N-terminal direction,resulting in an increased interaction area from 917.2 to 1164.8Å2. On the GPIb� side, only the �-switch region has a differentconformation, leading to a difference in rootmean square devi-ation of 3.4 Å between the mAbs and the �-switch region. Theroot mean square difference between the two docking models,21 and 61, is 8.2 Å when we consider the CDRs but only 7.7 Åwhen we restrain to the residues that are common in theinterface.A close comparison of the two docking models reveals that

some of the amino acids of 6B4 that were involved in the inter-action area of the docking model 21 no longer are (i.e. L27C).On the other hand, new residues are now situated in the contactsurface (i.e. Met51 and Phe71). In model 61, CDR L2 with resi-dues Met51, Ser52, Thr53, and Arg54 is now part of the interact-ing surface fromwhich it was absent inmodel 21. In both dock-ingmodels, the light chain has a dominant role in the binding toGPIb�, whereas CDRH1 is not contributing. The interaction ofGPIb� inmodel 61 does not differ from the one inmodel 21; theN-terminal flanking region, the five first LRRs, and the�-switchregion are all involved in the binding to 6B4.

DISCUSSION

We have developed a monoclonal antibody, 6B4, targetingthe humanGPIb�, which is responsible for the binding of plate-lets to the exposed collagen in damaged vessels via von Will-ebrand factor under high shear stresses. After promising resultsin animal models of arterial thrombosis, we have developed a5 M. Yamashita and T. Matsushita, unpublished results.

FIGURE 3. Amino acid sequence of human GPIb� targeted for charged to alanine scanning mutagenesis. The amino acid sequence shown includesresidues 1–302 of human GPIb�. The functional elements of GPIb� are indicated below the sequence. Charged residues His, Arg, Lys, Glu, and Asp (boldfacetype) were targeted for the mutagenesis; the boxes show clusters of mutations. For convenience, the mutant proteins were named according to the mutatedamino acids in GPIb� (e.g. mutant D18A/K19A/R20A contains three residues, Asp18, Lys19, and Arg20, mutated to Ala). As a result, 38 mutants were constructed,including 19 single mutations and 19 clustered mutations.

Paratope Identification of the Anti-GPIb� Antibody 6B4

AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23521

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

FIGURE 4. Epitope of 6B4, 24G10, and 27A10 on GPIb�. Red, involved in 6B4 binding; green, not involved; blue, inconclusive due to likely misfolding of themutant; orange, gain-of-function mutations that disrupt 6B4 binding. A, binding of monoclonal antibodies to mutant GPIb�. After expression and purification,each GPIb� was tested for the binding to coated mAb 6B4, 24G10, or 27A10 as described under “Experimental Procedures.” Each bar represents the mean withS.E. value obtained for at least two independent duplicate assays. *, statistically different, with p � 0.05. Results with line numbers in boldface type are discussedunder “Results,” and 1–7 corresponds to the numbering in Table 3. B, surface representation of the N-terminal part of GPIb� with predicted residues involvedor not in 6B4 binding. Gly233 and Met239, for which the gain-of-function mutation to Val disrupts the binding of 6B4, are in orange. The images of thethree-dimensional models with the surface representation were generated with PyMOL (available on the Word Wide Web).

Paratope Identification of the Anti-GPIb� Antibody 6B4

23522 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

recombinant and humanized Fab fragment of 6B4. Impor-tantly, at effective antithrombotic doses, 6B4-Fab does not pro-long the bleeding time; nor does it induce thrombocytopenia (9,11). To take full advantage of the in vivo effects of blockingGPIb�, a compound fit for oral administration is a prerequisitefor a broad andprophylactic use. As a first step toward that goal,we here identified the paratope of 6B4 that confers the inhibi-tory properties of the molecule. In this study, we mapped boththe paratope and the epitope of 6B4 by combining computerdocking models with mutagenesis studies on both 6B4 andGPIb�.Our first ZDOCK-based docking approach using 6B4 and

GPIb� in its ligand-free conformation (1M0Z.pdb) identifiedonly light chain residues Y27D and Glu93, of the seven selectedresidues, to be critical for the binding. This relatively poorresult might be due to the fact that ZDOCK is an algorithmdeveloped for rigid body docking. GPIb�, however, is a mole-cule with several known conformations, and also antibodiesrearrange their residue side and/or main chains to improve theaffinity for their targets. One way to overcome this problemwould be to performmolecular dynamics simulations followedby calculation of relative free binding energies with the molec-ular mechanics Poisson-Boltzmann surface area as, for exam-ple, Wu et al. (23) did to resolve the interaction between thescorpion toxin ScyTx and the small conductance calcium-acti-vated potassium channel Rsk2. An alternative strategy, whichwe followed, is using a docking strategy method that allowsflexibility in both ligand (6B4) and target (GPIb�), such asHADDOCK1.3. Docking tasks were submitted, including boththe ligand-free and ligand-bound conformation of GPIb� incombination with 6B4 models constructed with theWeb Anti-body Modeling program, next to the model that we previouslyused to prepare a humanized 6B4-Fab fragment (17). Afterranking the hits and visual inspection, no good candidate mod-els were identified with the ligand-bound conformation of

GPIb�. This finding is in total agreement with our previousresult, where the binding of 6B4 was some 6-fold lower to thegain of function (G233V andM239V)GPIb�, as comparedwiththe wild type (10). These gains of function are found in platelet-type vonWillebrand disease and enhance the affinity of GPIb�for VWF (6, 24). Furthermore, the structure of GPIb� carryingeither one of these mutations is similar to its conformation incomplex with the VWF A1 domain (19).Next, 6B4 residues involved in the binding to ligand-free

GPIb�, as deduced from the docking model 21, were expressedas single mutants. The chimeric human/mouse 6B4-IgG4 waschosen because it has the same characteristics (17) as the paren-tal IgG and is easy to manipulate and to produce. Indeed, pro-duction of the mutants in quantities sufficient for the bindingstudieswas possible for all 13mutants except for Y94A. Bindingexperiments of 6B4 WT and its mutants were performed onwhole fixed platelets of healthy volunteers, thereby presentingGPIb� in the GPIb�IX�V complex. All together we positivelyidentified three paratope residues in CDR L1 and one in CDRL3 and CDR H3 each, which all together are in close spatialproximity on the antibody surface.To further validate docking model 21, we next explored the

role of every charged residue in GPIb� by using the Ala scanlibrary previously described (21). Mutation of most of the pre-dicted interacting residues caused deficient binding. However,since a number of these mutations are suspected to induce aconformational change in GPIb� (21), which also here resultedin a decreased binding of two other anti-GPIb� antibodies withdifferent epitopes (10), we cannot make a definitive statementon these, in contrast to residuesAsp235 andLys237 located in the�-switch region, which specifically affected binding of 6B4.These results are furthermore in perfect agreement with thestudy of Cauwenberghs et al. (10), where we used human/ca-nine chimeric GPIb� to map the epitope.Finally, we performed a new docking experiment, including

all known experimental data, which yielded a final model intotal agreement with all of the mutagenesis results.In conclusion, by identifying the crucial residues involved in

the paratope-epitope interaction of 6B4 with GPIb�, a detailedmodel of the complex at the atomic level is proposed. Thisinformation allows for a better understanding of the antithrom-botic action of 6B4 andwill be helpful in the design of improvedmolecules.Furthermore, and in a broader perspective, the iterative

method we used here to identify the interacting amino acidresidues by alternating docking and mutagenesis experimentsallows for a more rational identification of relevant residuesthan what a classical laborious mutagenesis scan can provideand can be readily applied to other protein-protein interactionsfor which sufficient structural information is available.

REFERENCES1. Ruggeri, Z. M. (1997) Thromb. Haemostasis 78, 611–6162. Andrews, R. K., Lopez, J. A., and Berndt, M. C. (1997) Int. J. Biochem. Cell

Biol. 29, 91–1053. Berndt, M. C., Shen, Y., Dopheide, S. M., Gardiner, E. E., and Andrews,

R. K. (2001) Thromb. Haemost. 86, 178–1884. Yip, J., Shen, Y., Berndt, M. C., and Andrews, R. K. (2005) IUBMB Life 57,

103–108

FIGURE 5. Superposition of GPIb�/6B4 model 21 and optimized model61. The “ligand-free” conformation of GPIb� (green, ribbon) is complexedwith 6B4 from the optimized model 61 (light chain in yellow and heavy chainin orange) and from the initial model 21 (light chain in magenta and heavychain in blue). The region of GPIb� in model 61 with the most notable changein conformation compared with model 21 is shown in red (�-switch region),and the remainder of GPIb� is white. The image of the three-dimensionalmodel was generated with PyMOL (available on the World Wide Web).

Paratope Identification of the Anti-GPIb� Antibody 6B4

AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23523

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from

5. Huizinga, E. G., Tsuji, S., Romijn, R. A., Schiphorst, M. E., de Groot, P. G.,Sixma, J. J., and Gros, P. (2002) Science 297, 1176–1179

6. Andrews, R. K., Gardiner, E. E., Shen, Y., Whisstock, J. C., and Berndt,M. C. (2003) Int. J Biochem. Cell Biol. 35, 1170–1174

7. Feng, S., Resendiz, J. C., Lu, X., and Kroll, M. H. (2003) Blood 102,2122–2129

8. Nakamura, F., Pudas, R., Heikkinen, O., Permi, P., Kilpelainen, I., Munday,A. D., Hartwig, J. H., Stossel, T. P., and Ylanne, J. (2006) Blood 107,1925–1932

9. Cauwenberghs, N., Meiring, M., Vauterin, S., van Wyk V., Lamprecht, S.,Roodt, J. P., Novak, L., Harsfalvi, J., Deckmyn, H., and Kotze, H. F. (2000)Arterioscler. Thromb. Vasc. Biol. 20, 1347–1353

10. Cauwenberghs, N., Vanhoorelbeke, K., Vauterin, S., Westra, D. F., Romo,G., Huizinga, E. G., Lopez, J. A., Berndt, M. C., Harsfalvi, J., and Deckmyn,H. (2001) Blood 98, 652–660

11. Wu, D., Meiring, M., Kotze, H. F., Deckmyn, H., and Cauwenberghs, N.(2002) Arterioscler. Thromb. Vasc. Biol. 22, 323–328

12. Chen, R., and Weng, Z. (2002) Proteins 47, 281–29413. Russell, R. B., Alber, F., Aloy, P., Davis, F. P., Korkin, D., Pichaud,M., Topf,

M., and Sali, A. (2004) Curr. Opin. Struct. Biol. 14, 313–32414. Dominguez, C., Boelens, R., and Bonvin, A. M. (2003) J Am. Chem. Soc.

125, 1731–173715. Bonvin, A. M. (2006) Curr. Opin. Struct. Biol. 16, 194–20016. Delhaise, P., Bardiaux,M., DeMaeyer, M., Prevost, M., Vanbelle, D., Don-

neux, J., Lasters, I., Vancustem, E., Alard, P., andWodak, S. J. (1988) J.Mol.Graphics 6, 219

17. Fontayne, A., Vanhoorelbeke, K., Pareyn, I., Van, R. I., Meiring, M., Lam-precht, S., Roodt, J., Desmet, J., andDeckmyn,H. (2006)ThrombHaemost.96, 671–684

18. Whitelegg, N. R., and Rees, A. R. (2000) Protein. Eng. 13, 819–82419. Dumas, J. J., Kumar, R., McDonagh, T., Sullivan, F., Stahl, M. L., Somers,

W. S., and Mosyak, L. (2004) J. Biol. Chem. 279, 23327–2333420. Kabat, E. A.,Wu, T. T., Reid-Miller,M., Perry, H.M., andGottesman, K. S.

(1987) Sequences of Proteins of Immunological Interest, 4th Ed., pp.VII–XLIV, National Institutes of Health, Bethesda, MD

21. Shimizu, A.,Matsushita, T., Kondo, T., Inden, Y., Kojima, T., Saito, H., andHirai, M. (2004) J. Biol. Chem. 279, 16285–16294

22. Cauwenberghs, N., Ajzenberg, N., Vauterin, S., Hoylaerts,M. F., Declerck,P. J., Baruch, D., and Deckmyn, H. (2000) Haemostasis 30, 139–148

23. Wu, Y., Cao, Z., Yi, H., Jiang, D., Mao, X., Liu, H., and Li, W. (2004)Biophys. J 87, 105–112

24. Miller, J. L. (1996) Thromb. Haemost. 75, 865–869

Paratope Identification of the Anti-GPIb� Antibody 6B4

23524 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007

at KU

Leuven - Biom

edical Library on Septem

ber 11, 2007 w

ww

.jbc.orgD

ownloaded from