quaternary organization of gpib-ix complex and insights into bernard-soulier syndrome revealed by...

THROMBOSIS AND HEMOSTASIS

Quaternary organization of GPIb-IX complex and insights into Bernard-Souliersyndrome revealed by the structures of GPIb� and a GPIb�/GPIX chimera*Paul A. McEwan,1 *Wenjun Yang,2 *Katherine H. Carr,1 Xi Mo,2 Xiaofeng Zheng,2 Renhao Li,2 and Jonas Emsley1

1Centre for Biomolecular Sciences, School of Pharmacy, University of Nottingham, Nottingham, United Kingdom; and 2Center for Membrane Biology,Department of Biochemistry and Molecular Biology, The University of Texas Health Science Center at Houston, Houston, TX

Platelet GPIb-IX receptor complex has3 subunits GPIb�, GPIb�, and GPIX, whichassemble with a ratio of 1:2:1. Dysfunc-tion in surface expression of the complexleads to Bernard-Soulier syndrome. Wehave crystallized the GPIb� ectodomain(GPIb�E) and determined the structure toshow a single leucine-rich repeat with N-and C-terminal disulphide-bonded cap-ping regions. The structure of a chimeraof GPIb�E and 3 loops (a,b,c) taken fromthe GPIX ectodomain sequence was also

determined. The chimera (GPIb�Eabc), butnot GPIb�E, forms a tetramer in the crys-tal, showing a quaternary interface be-tween GPIb� and GPIX. Central to thisinterface is residue Tyr106 from GPIb�,which inserts into a pocket generated by2 loops (b,c) from GPIX. Mutagenesisstudies confirmed this interface as a validrepresentation of interactions betweenGPIb� and GPIX in the full-length com-plex. Eight GPIb� missense mutationsidentified from patients with Bernard-

Soulier syndrome were examined forchanges to GPIb-IX complex surface ex-pression. Two mutations, A108P andP74R, were found to maintain normalsecretion/folding of GPIb�E but were un-able to support GPIX surface expression.The close structural proximity of thesemutations to Tyr106 and the GPIb�E inter-face with GPIX indicates they disrupt thequaternary organization of the GPIb-IXcomplex. (Blood. 2011;118(19):5292-5301)

Introduction

GPIb-IX-V complex is an abundant membrane receptor complexon the platelet surface that plays a critical role in mediating plateletadhesion to the damaged vessel wall under conditions of high shearstress.1 Platelets adhere, and integrins are subsequently activatedby interactions of GPIb-IX-V with VWF bound to the subendothe-lium. How GPIb-IX-V transmits the VWF-binding signal acrossthe membrane is not clear, partly because the structure andorganization of this complex receptor remain to be elucidated.Because GPV is only weakly associated with the receptor complexand is not essential for complex expression, assembly, VWFbinding, or signal transduction,2,3 we focus on the GPIb-IXcomplex here.

The GPIb-IX complex contains 3 subunits, GPIb�, GPIb�, andGPIX, with a 1:2:1 stoichiometry.4 Each subunit is a type Itransmembrane (TM) protein, containing an ectodomain withleucine-rich repeats (LRRs),5 a single TM helix, and a relativelyshort cytoplasmic tail. The GPIb� ectodomain contains bindingsites for a growing list of hemostatically important ligands,including VWF and thrombin.6-8 Covalent and noncovalent interac-tions are important to the quaternary stabilization of the receptor.GPIb� links to 2 GPIb� subunits through membrane-proximaldisulfide bonds to constitute the GPIb complex.4 GPIX tightlyassociates with GPIb through noncovalent interactions.9 Assemblyof these subunits into a tightly integrated complex is also supportedby genetic evidence. Bernard-Soulier syndrome (BSS) is a heredi-tary bleeding disorder that is characterized in most cases by giantplatelets, low platelet counts, and little or no expression of GPIb-IXon the platelet surface.10,11 More than 30 mutations have been

identified from patients with BSS and mapped to GPIb�, GPIb�,and GPIX,12,13 indicating that all 3 subunits are required for propersurface expression of the complex.

Consistent with genetic evidence, efficient surface expressionof GPIb-IX also requires all 3 subunits in transfected mammaliancells.14 For instance, GPIX alone cannot be expressed on thesurface of transfected Chinese hamster ovary (CHO) cells. Onlywhen coexpressed with GPIb� can it be detected on the cellsurface, indicating that GPIb� interacts with and stabilizes GPIX.15,16

Coexpression with GPIb� and GPIb� produces even higher surfaceexpression levels of GPIX.14 Thus, the surface expression level ofindividual subunits can be used as an indicator for the assembly ofGPIb-IX and, implicitly, quaternary interactions among the sub-units. Using this approach, we had previously shown that TMdomains of GPIb-IX are essential for complex assembly.17,18

Biophysical characterization of recombinant GPIb-IX–derived TMpeptides in detergent micelles indicated that they form a parallel4-helical bundle and that their association leads to formation ofmembrane-proximal disulfide bonds between GPIb� and GPIb�,further stabilizing the complex.4,19

In addition to TM domains, GPIb� and GPIX ectodomains,termed GPIb�E and GPIXE in this study, respectively, are requiredfor proper assembly and efficient surface expression of GPIb-IX,because numerous BSS-causing mutations have been mapped tothese domains. However, GPIXE is intrinsically unstable, impedingstructural and biochemical investigation. Taking advantage of thehigh sequence homology between GPIb�E and GPIXE (Figure 1A),we had earlier identified a GPIb�E/GPIXE chimera (abbreviated

Submitted May 22, 2011; accepted August 22, 2011. Prepublished online asBlood First Edition paper, September 8, 2011; DOI 10.1182/blood-2011-05-356253.

*P.A.M., W.Y., and K.H.C. contributed equally to this study.

An Inside Blood analysis of this article appears at the front of this issue.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page chargepayment. Therefore, and solely to indicate this fact, this article is herebymarked ‘‘advertisement’’ in accordance with 18 USC section 1734.

© 2011 by The American Society of Hematology

5292 BLOOD, 10 NOVEMBER 2011 � VOLUME 118, NUMBER 19

to GPIb�Eabc) that interacts with GPIb� in a manner that mimicsthe GPIX ectodomain.16 Here, we report X-ray crystal structuresof human GPIb�E and GPIb�Eabc. These structures, in combina-tion with assays to probe inter-subunit interactions in the contextof full-length subunits, give insight into GPIb-IX quaternaryassembly and provide greater understanding of the molecularmechanism of BSS.

Methods

Expression and purification of recombinant GPIb�E andGPIb�Eabc

To produce recombinant GPIb�E and GPIb�Eabc, a gene fragment encodinghuman GPIb� and GPIb�Eabc residues Cys1-Leu121, respectively, wascloned into a pBlueBac4.5-derived baculovirus expression vector (Invitro-gen) that appended the signal sequence of baculovirus envelope gp67(amplified from pAcGP67A vector; BD Biosciences) and a Ser-Ser-hexahistidine tag to the N-terminus and C-terminus of GPIb�E, respec-tively. The chimeric GPIb�Eabc contains 3 stretches of GPIX sequences:Ala29-Arg36, Ser49-Gln60, and Ser76-Arg87 (Figure 1A). The matureprotein with the hexahistidine tag was secreted from the infected insect cellsinto the culture media. After 40%-85% ammonium sulfate fractionation ofthe collected culture media, the protein was dissolved in 50mM Tris�HCl,20mM imidazole, 4mM EGTA, pH 7.6, at 4°C overnight; loaded onto theNi-sepharose column (QIAGEN); and subsequently eluted with 50mMTris�HCl, 300mM NaCl, 100mM imidazole, and 4mM EGTA, pH 7.6. Theeluent was concentrated and further purified by gel filtration chromatogra-phy (Superdex 75; GE Healthcare) in 20mM Tris�HCl and 100mM NaCl,pH 7.0 (supplemental Figure 1, available on the Blood Web site; see theSupplemental Materials link at the top of the online article). N-terminalsequencing analysis of the purified protein confirmed the proper removal ofthe signal sequence.

Crystallization and structural determination

Purified GPIb�E with the hexahistidine tag was dialyzed into 20mMTris�HCl, and 100mM NaCl, pH 7.0, and concentrated to 18 mg/mL forcrystallization at 19°C. Sparse matrix screens (QIAGEN) obtained initialconditions and refinement resulted in the crystallization condition of 1.6M(NH4)2SO4, 0.4M LiCl, and 0.1M MES, pH 6.5. For GPIb�Eabc the samplewas concentrated to 1.5 mg/mL, and sparse matrix screens (PACT; Molecu-lar Dimensions) identified initial conditions at 10°C from D2 (form 1) andA9 (form 2). The D2 condition is 0.1M MMT buffer pH 5 (MMT buffer is amixture of DL-malic acid, MES, and Tris base in the molar ratios 1:2:2,DL-malic acid/MES/Tris base), 25% (w/v) polyethylene glycol 1500 andA9 is 0.2M lithium chloride, 0.1M sodium acetate, pH 5, and 20%polyethylene glycol 6000. Single crystals were transferred to the samesolution containing 25% and 10% glycerol, respectively, and flash cooled inliquid nitrogen. Diffraction data were collected with beam line ID23-2,ID29-1, and ID23-2, respectively, for GPIb�E, GPIb�Eabc (form 1), andGPIb�Eabc (form 2) at the European Synchrotron Radiation Facility. TheGPIb�E structure was determined by molecular replacement with the use ofthe structure for C-terminal 133 residues of the Nogo-66 ectodomain (PDBcode, 1P8T)20 and programs MrBUMP21 and PHASER.22 Initial electrondensity maps were improved with 2-fold noncrystallographic symmetry andsolvent flattening with the use of the CCP4 program suite. Model rebuildingwas performed with COOT23 and crystallographic refinement was per-formed in REFMAC.24 Crystallographic statistics are listed in Table 1. ARamachandran plot shows 115 residues in preferred regions, 2 in allowedregions, and none in outlier regions. The GPIb�Eabc form 1 structure wasdetermined by molecular replacement with the use of the GPIb�E structure.The model was built with COOT and refined with REFMAC. A Ramachan-dran plot shows 116 of residues in favored regions and 3 in allowed regionswith none in outlier regions. The form 2 structure was determined bymolecular replacement with the use of the form 1 structure and is identicalwith the exception of side chains involved in crystal packing.

GPIb� and GPIX constructs

Vectors expressing hemagglutinin (HA)–GPIb� (full-length GPIb� withN-terminal HA epitope tag, YPYDVPDYA), HA-GPIb�E (GPIb� extracel-lular residues Cys1-Leu121 with N-terminal HA tag), HA-GPIb�Eabc-GPIXTC (HA-tagged GPIb�Eabc fused to GPIX TM and cytoplasmicdomains), HA-GPIb�Eabc (HA-tagged GPIb�Eabc), GPIX, and GPIb� hadbeen described.16,17,25 Site-directed mutagenesis was performed by theoverlap extension PCR procedure with the use of the above-mentionedvectors as the template. Each PCR fragment was digested by appropriaterestriction enzymes and subcloned into its target vector as describedearlier.17 All constructs were confirmed by DNA sequencing.

Transient transfection of CHO cells

CHO K1 cells were grown in DMEM supplemented with 10% FCS at 37°Cand 5% CO2. Transient transfection of vectors encoding GPIb�, GPIb�, andGPIX-derived constructs, in desired combinations, into CHO K1 cells wasperformed with lipofectamine 2000 (Invitrogen) as described earlier.17 Keyparameters of the transfection, such as cell density and DNA amount, werekept the same as previously described to allow proper comparison amongvarious constructs.17 After transfection, the cells were grown for anadditional 48 hours before being analyzed.

Characterization of effects of GPIb� or GPIX mutations

The secretion and folding of mutant HA-GPIb�E and HA-GPIb�Eabc

proteins were characterized as previously described.16 Briefly, the N-terminally HA-tagged protein secreted into the culture medium wascollected by coimmunoprecipitation with the use of the anti-HA Ab,resolved in 12% Bis-Tris SDS gel in the presence or absence of reducingagents, and immunoblotted by HRP-conjugated anti-HA Ab. The cellularexpression and assembly of the GPIb-IX complex was assessed by Westernblot of each subunit as previously described.17,18 The surface expressionlevels of GPIb�, HA-tagged GPIb�, and GPIX in transiently transfectedcells were measured with WM23, anti-HA, and FMC25 Abs, respectively,on a Beckman-Coulter Gallios flow cytometer as described.17,25 To assessthe mutational effect, the measured mean fluorescence value of the entirecell population (10 000 cells) is normalized with the value of CHO cellsexpressing wild-type GPIb-IX complex (GPIb�/HA-GPIb�/GPIX) being100% and that of empty vector-transfected cells 0%.17 Groups werecompared with the 2-tailed Student t test.

Results

Crystal structure of GPIb�E

To explore the architecture of the GPIb-IX receptor complex wecrystallized recombinant GPIb�E and solved the crystal structure to1.25-Å resolution (Table 1). The topology of GPIb�E spanningresidues 1-118 is shown in Figure 1B, showing the first structurewith only a single LRR repeat. A typical example of the refinedelectron density is shown in stereo in Figure 1C and supplementalVideo 1. As seen in other LRR proteins, GPIb�E assumes aright-handed coiled structure with a parallel �-sheet on one side(the concave face) and connecting loops containing �-turns andshort 310 helices on the opposite (convex) face. The central LRR iscovered on both ends by N- and C-terminal capping regions thatcontain several short �- and 310-helices. Four disulfide bonds areobserved in the GPIb�E structure: 2 (Cys1-Cys7 and Cys5-Cys14)are located in the N-terminal cap and the other 2 (Cys68-Cys93 andCys70-Cys116) in the C-terminal cap, which are topologicallyequivalent to those in the Nogo-66 and SLIT receptor.20,26 Withonly a single LRR it assumes a compact rectangular shape with arelatively flat, rather than a curved concave, face commonlyobserved in multi-LRR structures. Moreover, the single LRR

GLYCOPROTEIN GPIb�-GPIX STRUCTURE 5293BLOOD, 10 NOVEMBER 2011 � VOLUME 118, NUMBER 19

accommodates a unique feature in GPIb�E that has not beenobserved in previously reported LRR structures; interactions ofside chains bridging N- and C-terminal capping regions. Extendingover the convex face, the aromatic ring in Trp21 is locked betweenthe amino group of Pro46 and the guanidinium group of Arg71 bycation-� interactions (Figure 1B). These interactions exemplifynumerous interloop interactions on the convex face, which prob-ably add stability to the structure in a manner similar to the buriedGPIb� Asn residues, Asn40 and Asn64, on the concave face.27

Different pathways in the pathogenesis of GPIb� BSSmutations

Eight missense mutations in GPIb�E (C5Y,28 R17C,29 P29L,30

N64T,31 P74R,32 Y88C,33,34 P96S,35 and A108P34) have beenidentified in patients with BSS. We have examined the context ofthese 8 mutations with the use of the GPIb�E structure. Residuesaffected are shown in a ribbon diagram of the structure in Figure2A and also in supplemental Video 2. Information on surfacelocalization and conservation of each affected residue is summa-rized in supplemental Table 1. Cys5 and Cys14 form a disulfidebond. Substitution of Cys5 with a tyrosine would result in loss ofthe disulfide bond and leave an unpaired Cys14. Asn64 is fullyburied in the structure and bridges the C-terminal cap and the LRRwith 2 hydrogen bonds, both of which would be lost if substitutedby a smaller Thr residue. The other GPb�E residues (Arg17, Pro29,Pro74, Tyr88, Pro96, and Ala108) are present on the proteinsurface, although Arg17, Tyr88, and Pro96 are partially buried bysurrounding side chains (Figure 2B). Both R17C and Y88C wouldintroduce an additional Cys residue to the domain, which willprobably interfere with formation of the 4 native disulfide bonds.The other 4 mutations involve either removal or addition of a Proresidue, which can affect local conformation or global stability.

To elucidate the molecular pathogenesis of these BSS muta-tions, we have characterized systematically the mutational effectson GPIb-IX expression and assembly in transiently transfected

CHO cells. Key parameters of transient transfection were keptconstant to ensure proper comparison of protein expression levelsamong different experiments.17 CHO cells largely recapitulated thereported clinical observations that all 8 mutations led to asignificant decrease in surface expression of the GPIb-IX complex,albeit to various degrees (supplemental Figure 2). To test whetherthese BSS mutations are detrimental to the structural integrity ofGPIb�E, each mutation was introduced to the HA-tagged GPIb�E

(with the native signal sequence), and the resulting gene wastransfected transiently into CHO cells. Western blot analysis of thecell lysate indicated that translation of the HA-GPIb�E gene wasnot affected by any of the mutations (Figure 2C). However, C5Y,P29L, and P96S mutant proteins failed to secrete from the cellsbecause no HA-tagged protein was detected in the culture media.The secretion of R17C was significantly decreased. In N64T-expressing cells, only the ectodomain with a higher molecular masswas detected in the culture media. The other mutants, P74R, Y88C,and A108P, were secreted similar to the wild type. Of the mutantsthat were secreted, R17C, N64T, and Y88C exhibited significantformation of intermolecular disulfide bonds as detected in SDS-PAGE under nonreducing conditions. Because GPIb�E contains4 intramolecular disulfide bonds and no intermolecular ones(Figure 1B), the existence of the latter indicated misfolding of thesemutant proteins. In contrast, P74R and A108P GPIb�E mutantproteins, like the wild-type, contained only intramolecular disulfidebonds and should therefore be well folded.

The effects of mutations on the interaction between GPIb� andGPIX were analyzed next. GPIX does not express on the cellsurface in isolation but becomes detectable when coexpressed withGPIb�.14 GPIb� is thought to interact with and stabilize GPIX,which involves the ectodomain and TM domain of both pro-teins.15,16 As shown in Figure 2D, when HA-tagged GPIb�(HA-GPIb�) and GPIX were cotransfected into CHO cells, bothHA-GPIb� and GPIX were detected on the cell surface. Mutationsthat disrupt secretion or folding of GPIb�E, such as P29L, resulted

Table 1. Crystallographic data collection and refinement statistics

Sample GPIb�E GPIb�Eabc (form 1) GPIb�Eabc (form 2)

Data collection

Space group P2(1) P3(1)21 C2(1)

Cell dimensions

a, b, c, Å 61.60, 34.83, 71.77 72.01, 72.01, 171.73 124.27, 72.72, 72.57

�, �, �, degree 90, 90.31, 90 90, 90, 120 90, 107.06, 90

Resolution, Å* 31.54-1.25 (1.28-1.25) 42.17-2.35 (2.48-2.35) 36.36-3.2 (3.82-3.20)

Rsym*† 0.145 (0.400) 0.131 (0.517) 0.136 (0.630)

I/sigI* 9.8 (3.8) 14.4 (5.0) 7.6 (1.9)

Completeness, %* 93.4 (92.2) 100.0 (100.0) 95.2 (95.0)

Redundancy* 4.6 (4.0) 10.9 (11.1) 2.4 (2.4)

Refinement

No. of reflections 78 922 19 565 9796

Rwork/Rfree ‡ 0.203/0.223 0.211/0.259 0.246/0.284

B-factors, Å2

Protein 12.5 8.8 25.2

Ligands 21.0 39.3 61.3

Solvent 28.1 13.1 18.7

Rms deviations

Bond lengths, Å 0.0188 0.189 0.008

Bond angles, degree 1.889 1.301 1.435

*Values in parentheses are for highest-resolution shell.†Rsym � Sum(h) �Sum(j) �I(hj) � �Ih/Sum(hj) �Ih where I is the observed intensity and � Ih is the average intensity of multiple observations from symmetry-related

reflections calculated with SCALA.‡Rwork � Sum(h) ��Fo�h � �Fc�h�/Sum(h)�Fo�h, where Fo and Fc are the observed and calculated structure factors, respectively. Rfree was computed as in Rwork, but only for

(5%) randomly selected reflections, which were omitted in refinement, calculated with REFMAC.

5294 McEWAN et al BLOOD, 10 NOVEMBER 2011 � VOLUME 118, NUMBER 19

in little cell surface expression of HA-GPIb� and, as a conse-quence, GPIX as detected by flow cytometry. By contrast, adistinctive feature of P74R and A108P was that HA-GPIb� waspresent on the cell surface but GPIX was not. Overall, these resultsshowed that BSS missense mutations disrupt assembly and surfaceexpression of the GPIb-IX complex via different mechanisms.

Crystal structure of GPIb�Eabc and conformational change inthe Tyr88 loop

Despite high sequence similarity between GPIb�E and GPIXE,attempts to express recombinant GPIXE have not been successful.16

However, sequence analysis did allow the engineering of GPIb�Eabc,a chimera of GPIb�E and GPIXE, as a stable protein that is readilysecreted.16 Built on the GPIb�E scaffold, GPIb�Eabc incorporates3 discontinuous stretches of GPIXE that correspond to the 3 convexloops, including the �1 helix (termed loops a, b, c; spanning GPIXresidues Ala29-Arg36, Ser49-Gln60, and Ser76-Arg87 as shown inFigure 1A). GPIb�Eabc was purified from insect cells and crystal-lized, and the structure was solved to 2.35-Å resolution, using theGPIb�E crystal structure for a molecular replacement calculationidentifying 4 molecules in the asymmetric unit (Figure 3A; Table1). To avoid confusion, residues of GPIb�Eabc are labeled in thisstudy by their residue numbers in respective source domains (eg,GPIX-Asp56 in loop b) rather than by GPIb�Eabc’s own.

The GPIb�Eabc structure shares the same topology as GPIb�E,but it has 3 changes in the structure (Figure 3B). The first, asexpected, is in the convex loops. Because of the substantialsequence difference, the 310 helix in loop a and the cation-�interloop interaction between Trp21 and Arg71 in GPIb�E are notpresent in the GPIb�Eabc structure (Figure 3A). Second, in helix �1of loop c GPIX-Tyr79 replaces GPIb�-Pro74, which surprisinglydoes not affect the kinked main chain conformation but insteadforms a substantial new interaction with loop b (Figure 3A). Here,the GPIX-Tyr79 side chain is buried under the main chain ofGPIX-Gly53 and the Tyr OH group forms a hydrogen bond to themain chain nitrogen of adjacent GPIX-Phe55 (Figure 3A). Theextensive interactions observed in these convex loops substantiateour earlier observation that only when all 3 convex loops aregrafted together onto the GPIb�E scaffold did the GPIb�Eabc

chimeric protein become stable and well folded.16 A third differ-ence occurs in the C-terminal cap, despite GPIb�E and GPIb�Eabc

sharing the same amino acid sequence in this region. Residues86-89 that form the 310 helix in loop c of GPIb�E unravel, andresidues 80-86 coil up to form a novel turn of helix in GPIb�Eabc

with GPIb�-Glu84 forming a new salt bridge to GPIb�-Arg57. Inaddition, the C-terminal helix �2 makes a rigid body movement of3 Å away from helix �1. Compared with GPIb�E, the anglebetween helices �1 and �2 in GPIb�Eabc is widened (Figure 3C).

Figure 1. The crystal structure of GPIb� ectodomain(GPIb�E). (A) Sequence alignment of GPIb�E (blue),

GPIXE (purple), and GPIb�Eabc (blue and purple) withrespective residue numbers for GPIb�E and GPIX markedon the top and bottom. Elements of secondary structurein GPIb�E are shown on top and colored as in panel B.Residues affected by BSS missense mutations are inred. Three stretches of the GPIXE sequence that areincluded in GPIb�Eabc are shown in purple. (B) Twoorientations are shown of a ribbon diagram of the GPIb�E

structure viewed from the concave face (left), with�-strands labeled in blue, �-helices in red, 310 helices inpurple, and loop regions in gray. Asn residues 40, 41, and64 are shown as stick, and a single residue from theN-linked oligosaccharide attached to Asn41 is also shownin green. The diagram on the right is rotated 180 degreeswith side chains in the inter-LRR cap cation-� interactionshown as stick. (C) Refined 1.25-Å electron density,calculated with 2Fo-Fc coefficients and contoured at1.5 rms.


The re-coiling conformational change in the loop containingresidue Tyr88 is in a topologically equivalent region of the LRRfold to the �-switch loop in GPIb� known to alter conformationwhen binding to VWF-A1 domain or peptide inhibitor.6,36 InGPIb� the conformational change involves unraveling of a short�-helix to form an extended �-hairpin, whereas in GPIb� the�-helix unravels and a second stretch is formed, displaced towardthe C-terminus in a twisting helical motion (Figure 3D).

Tetrameric structure of GPIb�Eabc shows a GPIb�-GPIXinterface

Previous studies with the GPIb�Eabc construct have shown that the3 loops (a, b, c) from GPIX form ectodomain interactions withGPIb� and that this is required for surface expression of theGPIb-IX complex.16 The side chains from GPIb� involved in thisinteraction with GPIX were unknown. In the crystal structure weobserve 4 GPIb�Eabc molecules in the asymmetric unit form atetrameric ring structure. Here, the C-terminal cap region of onesubunit packs against the convex loops (b, c) of another placing the�-helices toward the center of the ring where they are partiallyburied (Figure 4A). At this interface the side chain of GPIb�-Tyr106 from one subunit lies at the center of a shallow pocketcreated between loop b and helix �1 (loop c) from a second subunitthat is rotated by 90 degrees. Each interface buries a surface area of� 1100 Å,2 which is significantly greater than values typical forcrystal contacts37 or the contacts found in the GPIb�E structure. Asecond monoclinic crystal form (space group C21) of GPIb�Eabc

grown under different conditions has an identical structure andtetrameric arrangement (labeled form 2 in Table 1).

Figure 4B shows hydrophobic contacts contribute to the GPIXpocket, including GPIX-Tyr79, GPIX-Leu82, and GPIX-Trp83,

and more peripheral contacts to the interface come from GPIb�-Ala108 and GPIX-His57. The OH group of GPIb�-Tyr106 forms ahydrogen bond to the guanidinium of GPIX-Arg87. GPIb�-Tyr106is further surrounded by 3 salt bridges between GPIb� residuesArg102, Asp90, and Arg89 and GPIX residues Asp56, Arg87, andAsp86, respectively. GPIb�-Arg92 and GPIX-Arg87 side chainsform hydrogen bonds to the main chain carbonyls GPIX-Asp56and GPIb�-Ala86, respectively.

The interface between GPIb�Eabc molecules is composed of theGPIb�E-derived C-terminal region and GPIXE-derived convexloops. To test whether this interface is a valid representation offull-length GPIb� and GPIX in the whole complex, 3 residues wereselected for mutagenesis. GPIb�-Tyr106 lies at the center of theinterface, and GPIX-Leu82 and GPIX-Asp86 are examples ofresidues that contribute hydrophobic and electrostatic contacts tothe interface, respectively (shown as underlined in Figure 4B). Inthe GPIb�E crystal structure, the side chain of Tyr106 is exposed tothe solvent and flanked by Arg85 and Arg89. Mutating Tyr106 toPhe, Asn, Asp, Ala, or Val largely preserved proper secretion andfolding of HA-GPIb�E that were expressed transiently from CHOcells (Figure 5A). However, when expressed as a full-lengthsubunit on the surface of cells, all Tyr106 mutations resulted in theloss of the ability of HA-GPIb� to enhance surface expression ofGPIX in transfected CHO cells (Figure 5B). Thus, as predicted bythe GPIb�Eabc crystal structure, these results pinpoint GPIb�-Tyr106 as a critical part of a GPIXE binding site.

Similarly, HA-GPIb�Eabc protein containing a GPIX-L82A orGPIX-D86E mutation was able to secrete to the culture media andfold well (Figure 6A). However, both mutations failed to retain theGPIb�-binding ability of GPIb�Eabc in the context of the full-lengthsubunit (Figure 6B). Because GPIXE alone cannot express as a

Figure 2. BSS mutations in GPIb�E disrupt expres-sion and assembly of GPIb-IX complex by differentmechanisms. (A) Two views of the GPIb�E structure areshown as a ribbon diagram related by a 90-degreerotation. Residues affected by BSS missense mutationsare highlighted as stick and colored green for beingsolvent accessible and red for buried. (B) Space-fillingrepresentation of the GPIb�E structure, showing theconcave (left) and convex (right) faces. Main chain atomsare colored white, and residues affected by BSS muta-tions are colored green. (C) SDS gels showing differentialeffects of BSS mutations on expression (top), secretion(middle), and folding (bottom) of GPIb�E expressed fromtransfected CHO cells. Each mutation is identified by theresidue number. The HA epitope tag was appended tothe N-terminal end of GPIb�E for easy detection. Immuno-precipitation was performed with anti-HA monoclonal Aband immunoblotting with HRP-conjugated anti-HA mono-clonal Ab (HRP-HA). Molecular weight markers aremarked on the left of each gel. (D) Sample flow cytomet-ric histograms showing the effects of selected BSSmutations on surface expression levels of HA-taggedfull-length GPIb� (HA-GPIb�) and GPIX (GPIX) that arecoexpressed transiently in CHO cells. (E) Relative sur-face expression levels of HA-GPIb� (gray column) andGPIX (white column) in transfected CHO cells. Thesurface expression levels were measured by flow cytom-etry and quantified as mean fluorescence intensity, whichwere normalized with expression levels in cells trans-fected with wild-type GPIb-IX (GPIb�/HA-GPIb�/GPIX)being 100% and those in cells transfected with shamvectors 0%.17 The data are presented as mean � SD(n � 3). *P � .001.


well-folded form,16 it would be difficult to assess the effect of eithermutation on its folding and secretion. Nonetheless, full-lengthGPIX bearing either mutation could not be expressed on the cellsurface even in the presence of GPIb� (Figure 6C), which isconsistent with the GPIb�Eabc crystal structure that both GPIX-Leu82 and GPIX-Asp86 participate directly in the interactionbetween GPIb�E and GPIXE.

Figure 3. The crystal structure of GPIb�Eabc and its comparison with GPIb�E.(A) A ribbon diagram of the GPIb�Eabc structure, showing the grafted GPIX convexloops (magenta) in front. Residues derived from GPIb�E are colored in blue. Sidechains of several residues are shown in stick and labeled. GPIX-Ala29 andGPIX-Ser76 correspond to GPIb�-Trp21 and GPIb�-Arg71, respectively. (B) Super-position of GPIb�E (green) and GPIb�Eabc (blue/magenta) structures. (C) A close-upview of the conformational difference in the C-terminal cap region between GPIb�E

(green) and GPIb�Eabc (blue) structures. The locations of several side chains in bothstructures are marked. (D) Topology diagrams of ligand-free (green) and ligand-bound structures in orange for GPIb� and blue for GPIb�. The observed conforma-tional change occurs in a topologically equivalent loop (boxed) and in each caseinvolves unwinding of a helix for GPIb� (R-loop) and GPIb� (loop containing residueTrp88) on engaging ligand; VWF-A1 or GPIX, respectively.

Figure 4. The tetramer structure of GPIb�Eabc shows a binding interfacebetween GPIb�E-derived C-terminal cap region and GPIXE-derived convexloops. (A) Cartoon diagram showing 2 views of the GPIb�Eabc tetramer structurerelated by a 90-degree rotation. The N-termini (N) of the molecules are at theperipheral of the tetramer, and the C-termini (C) are all located in proximity at thebottom of the lower panel. Sequences derived from GPIb�E are colored in blue andthose derived from GPIXE in magenta. The side chain of GPIb�-Tyr106 at eachinterface is shown in green as stick. (B) A close-up view of the interface between2 GPIb�Eabc molecules (GPIb�E-derived residues in blue and GPIXE-derived resi-dues in magenta). Interacting side chains are labeled and colored accordingly, andunderlined residues were subject to mutagenesis.


Discussion

One of the fundamental but unanswered questions about theGPIb-IX complex is how the subunits organize and assemble.Earlier studies have shown that TM helices of the GPIb-IXcomplex interact with one another to form a parallel tetrameric�-helical bundle.4,18,19 The GPIX TM helix bridges the 2 GPIb�TM helices, and GPIb� performs a similar function providingnoncovalent and covalent quaternary interactions and stability byformation of the 2 interchain disulfide bonds bridging 2 GPIb�s.Previously, it has also been shown that the LRR-containingectodomains form an interface involving the convex loops of GPIXinteracting with GPIb�.16 Overall protein-protein interactionsmediated by LRR domains use a diverse range of scaffolds withdifferent numbers of LRRs (ranging from 1 to 15) mediatinghomotypic and heterotypic interactions.5 These interactions areprincipally mediated by the LRR concave face that typically has a

curved or horseshoe shape as observed in GPIb�.6 GPIb� andGPIX are unusual in that they have only a single LRR. The GPIb�E

structure is the first description of this fold and shows a compactshape with little LRR curvature.

A crystal structure for the SLIT receptor LRR ectodomaindomain 4 (pdb code 2WFH) shows a homodimer with a concaveface-to-face interaction of the LRRs. We initially speculated thatthe 2 copies of GPIb�E may also interact in a similar way to form adimer; however, in the crystal structure we only observe amonomer. By contrast the structure of a chimeric GPIb�Eabc, where3 loops from GPIX are added, shows a heterodimeric GPIb�-GPIXinterface. Because of the positioning of the 2 subunits in the dimerat right angles, this structure is then able to cyclize with a seconddimer to form a ringlike tetramer where the same interface isobserved 4 times. Rather than using the LRR concave face, thisstructure shows that interactions occur through the C-terminal capof GPIb�E and convex face loops of GPIXE. Here, a central sidechain Tyr106 from the GPIb�E cap inserts into a pocket created by2 loops from GPIX on the convex face. This is more familiar to theinterfaces observed in Toll-like receptors in which heterodimericinteractions form between the C-terminal cap and one side ofthe LRRs.38

An important spatial constraint for the GPIb-IX complex lies inthe uniform topology of its subunits. Because GPIb�, GPIb�, and

Figure 6. GPIXE convex loops participate in direct interaction with GPIb�E in thefull-length complex. (A) SDS gels showing the lack of disruptive effects of the L82Aor D86E mutation on expression, secretion, and folding of HA-GPIb�Eabc. Residuenumbers of both mutations are in the context of GPIX. The other annotations followthose of Figure 2C. (B) Relative surface expression levels of HA-GPIb�Eabc-GPIXTC,either wild-type or containing the indicated mutation, in the absence (gray column) orpresence (white column) of coexpressing GPIb� in transfected CHO cells. HA-GPIb�Eabc-GPIXTC is a protein that contains N-terminally HA-tagged GPIb�Eabc andGPIX TM and cytoplasmic domains.16 The expression level was measured by flowcytometry with the use of anti-HA mAb and quantified as mean fluorescence intensity,which was normalized with that in cells transfected with wild-type GPIb-IX (GPIb�/GPIb�/HA-GPIX) being 100% and those in cells transfected with sham vectors 0%.16

Note that the enhancement of HA-GPIb�Eabc-GPIXTC surface expression by GPIb� issignificantly diminished by both mutations. The data are presented as mean � SD(n � 4). *P � .01. (C) Relative surface expression levels of GPIX, either wild-type orcontaining the indicated mutation, in the absence (gray column) or presence (whitecolumn) of coexpressing GPIb� in transfected CHO cells. GPIX surface expressionlevel was measured by flow cytometry with the use of anti-GPIX mAb FMC25 andanalyzed as described above (n � 3).

Figure 5. Tyr106 mutations do not disrupt secretion and folding of GPIb�E butdisrupt its interaction with GPIXE. (A) SDS gels showing the lack of disruptiveeffects of Tyr106 mutations on expression, secretion, and folding of GPIb�E

expressed from transfected CHO cells. The identity of each Tyr106 mutation ismarked on top of the gels. The other annotations follow those of Figure 2C.(B) Relative surface expression levels of HA-GPIb� (gray column) and GPIX (whitecolumn) in transfected CHO cells measured by flow cytometry. The annotations followthose of Figure 2E. Note that none of the Tyr106 mutations retain the ability ofwild-type HA-GPIb� to enhance surface expression of GPIX. The data are presentedas mean � SD (n � 3). *P � .001.


GPIX are all type I TM proteins, the C-termini of all 4 ectodomainsin the GPIb-IX complex should remain in close proximity to oneanother to enable formation of the adjacent parallel TM helixbundle. We noticed that in the GPIb�Eabc crystal structure, theC-termini of all 4 GPIb�Eabc molecules are located on the same sideof the tetramer (Figure 4A; supplemental Video 3). If the GPIb�1-GPIX-GPIb�2 trimer has a similar cyclic arrangement positionedabove the TM helices, then a schematic model of this trimer isshown in Figure 7A with the GPIb� TM helix added (shown ingreen) extending toward the N-terminal ligand binding domain. A

second observation from the GPIb�Eabc structure is that GPIX canonly bind one GPIb�E through its b,c loop pocket; thus, furtherstudies will be required to define what the context and conforma-tion of a second GPIb�E (GPIb�E2) is. Because we observe nointeraction between 2 GPIb�E molecules in the crystal structure,the model in Figure 7A shows GPIb�E2 on the opposite side ofGPIX to GPIb�E1.

Different pathways to BSS

The genetic basis for BSS was established 3 decades ago.11 Morethan 30 mutations have since been identified from patients withBSS, and the number will probably grow.39 As we showed in thisstudy, several novel mutations could produce BSS-like phenotypesin transfected cells (Figures 5-6). Some BSS mutations are locatedin the promoter region and are presumed to decrease genetranscription.40 Some are frameshifting or nonsense mutations,resulting in a typically truncated and nonfunctional protein prod-uct.41-43 In this study, we have for the first time systematicallycharacterized all the reported missense BSS mutations in GPIb�and observed distinct mechanisms that underlie the pathogenesis ofthe disease (Figure 2). Six of the 8 missense mutations aredetrimental to proper tertiary folding or secretion of GPIb�E.However, the other 2 mutations A108P and P74R maintain thenative disulphide bonds and probably fold normally.

The clinical data on the A108P mutation describe a patient whois compound heterozygous with mutation Y88C.34 In these plateletsGPIb-IX receptor complex is present on the surface at a reducedlevel and does bind VWF. Antibody SZ1 (binds GPIb�/GPIXsubunits in complex but not in isolation) recognizes the GPIb-IXpresent. As a compound heterozygote of GPIb�, the mutationscould form different combinations of the chains, that is, A108P/A108P, A108P/Y88C, and Y88C/Y88C. A separate family that ishomozygous for Y88C has classic BSS giant platelets with noreceptor present at the platelet surface.33 As shown in Figure 4B,Ala108 is adjacent to Tyr106 which is central to the GPIb�Eabc

interface. We showed mutating Tyr106 abolishes GPIX expressionin the same way as A108P. Mutating Ala108 to Pro at the peripheryof the interface may produce a less-stable subunit association that,in turn, gives rise to the reduction in surface expression of thewhole complex, which is what we observe in CHO cells. The P74Rmutation shows classic BBS platelets in homozygous patients withno receptor complex detected at the surface.32 This mutation ismore severe than A108P in our cell-based assays and does notsupport any GPIX expression at the cell surface even in thepresence of GPIb� (supplemental Figure 2). Pro74 does notcontribute directly to the Tyr106 interface but is located close by inhelix �1, which does contribute through side chains from residuesAla86, Arg89, and Asp90 (Figure 4B). Proline residues are commonlyfound at the N-termini of helices where the special main chain propertiesinfluence the secondary structure formation.Achange here could disruptthe orientation of the helix and thus its contributions to the interface.

Finally, not all mutations characterized in GPIb� result in lossor reduction of receptor complex at the platelet surface. Thepolymorphism G15E is associated with the human platelet-specificalloantigen Iya and does not adversely affect the expression level ofGPIb-IX.44 Gly15 is exposed on the protein surface and located inthe N-terminal cap. Figure 7B shows the locality of these mutationsin the same model as Figure 7A and shows how 2 copies of GPIb�in this arrangement present mutations in different contexts (supple-mental Video 4). Overall these studies provide valuable insights onthe assembly of GPIb-IX complex and will prove a scaffold forfurther investigation of this important platelet receptor.

Figure 7. A schematic model for the membrane-proximal portion of the GPIb-IXcomplex. (A) Cartoon diagrams related by a 90-degree rotation showing the GPIb� chainTM domain (in green) extending toward the N-terminus (top), GPIb� ectodomain and TMdomain (in sky blue and dark blue), and GPIX ectodomain and TM domain (in purple).Tyr106 from GPIb�1 is shown as stick along with interchain disulphides. (B) Same view aspanelA, except without TM domains. GPIb� is colored white, showing residues affected byBSS and human platelet-specific alloantigen (HPA) mutations. Side chains of GPIb�residues Ala108 (blue), Pro74 (orange), and Gly15 (red) are shown in space-filling mode;Tyr106 is shown as stick (gray).


Acknowledgments

The authors thank the European Synchrotron Radiation Facility forprovision of synchrotron radiation source and thank David Flot andGordon Leonard for assistance in using the beamline ID23-2.

This work was supported in part by the National Institutes ofHealth (grant HL082808, R.L.) and Cancer Research UK (grantC21418/A8573), British Heart Foundation (grant RG/07/002/23 132, J.E.).

Coordinates and structure factors deposited with the ProteinData Bank are 3RFE for GPIb�E and 3REZ for GPIb�Eabc.

Authorship

Contribution: P.A.M. and K.H.C. crystallized and determined thestructures of GPIb�E and GPIb�Eabc, respectively; W.Y. produced

recombinant proteins for crystallization, characterized and ana-lyzed mutational effects on GPIb-IX expression and assembly, andwrote the paper; X.M. and X.Z. cloned many constructs andestablished the baculovirus protein expression system; R.L. initi-ated and designed research, analyzed results, and wrote the paper;and J.E. analyzed results and wrote the paper.

Conflict-of-interest disclosure: The authors declare no compet-ing financial interests.

The current affiliation for R.L. is Division of Hematology,Oncology, and Bone Marrow Transplant, Department of Pediatrics,Emory University School of Medicine, Atlanta, GA.

Correspondence: Jonas Emsley, Centre for Biomolecular Sci-ences, School of Pharmacy, University of Nottingham, Notting-ham, NG7 2RD, United Kingdom; e-mail: [email protected]; and Renhao Li, Division of Hem/Onc/BMT,Department of Pediatrics, Emory University School of Medicine,2015 Uppergate Dr, Rm 464, Atlanta, GA 30322; e-mail: [email protected].

References

1. Andrews RK, Lopez JA, Berndt MC. Molecularmechanisms of platelet adhesion and activation.Int J Biochem Cell Biol. 1997;29(1):91-105.

2. Modderman PW, Admiraal LG, Sonnenberg A,von dem Borne AE. Glycoproteins V and Ib-IXform a noncovalent complex in the platelet mem-brane. J Biol Chem. 1992;267(1):364-369.

3. Kahn ML, Diacovo TG, Bainton DF, Lanza F,Trejo J, Coughlin SR. Glycoprotein V-deficientplatelets have undiminished thrombin responsive-ness and do not exhibit a Bernard-Soulier pheno-type. Blood. 1999;94(12):4112-4121.

4. Luo S-Z, Mo X, Afshar-Kharghan V, Srinivasan S,Lopez JA, Li R. Glycoprotein Ibalpha forms disul-fide bonds with 2 glycoprotein Ibbeta subunits inthe resting platelet. Blood. 2007;109(3):603-609.

5. Kobe B, Kajava AV. The leucine-rich repeat as aprotein recognition motif. Curr Opin Struct Biol.2001;11(6):725-732.

6. Huizinga EG, Tsuji S, Romijn RA, et al. Structuresof glycoprotein Ibalpha and its complex with vonWillebrand factor A1 domain. Science. 2002;297(5584):1176-1179.

7. Dumas JJ, Kumar R, Seehra J, Somers WS,Mosyak L. Crystal structure of the GpIb�-throm-bin complex essential for platelet aggregation.Science. 2003;301(5630):222-226.

8. Celikel R, McClintock RA, Roberts JR, et al.Modulation of alpha-thrombin function by distinctinteractions with platelet glycoprotein Ibalpha.Science. 2003;301(5630):218-221.

9. Du X, Beutler L, Ruan C, Castaldi PA, Berndt MC.Glycoprotein Ib and glycoprotein IX are fully com-plexed in the intact platelet membrane. Blood.1987;69(5):1524-1527.

10. Nurden AT, Caen JP. Specific roles for plateletsurface glycoproteins in platelet function. Nature.1975;255(5511):720-722.

11. Nurden AT, Dupuis D, Kunicki TJ, Caen JP. Analy-sis of the glycoprotein and protein composition ofBernard-Soulier platelets by single and two-di-mensional sodium dodecyl sulfate-polyacryl-amide gel electrophoresis. J Clin Invest. 1981;67(5):1431-1440.

12. Lopez JA, Andrews RK, Afshar-Kharghan V,Berndt MC. Bernard-Soulier syndrome. Blood.1998;91(12):4397-4418.

13. Berndt MC, Andrews RK. Bernard-Soulier syn-drome. Haematologica. 2011;96(3):355-359.

14. Lopez JA, Leung B, Reynolds CC, Li CQ, FoxJEB. Efficient plasma membrane expression of afunctional platelet glycoprotein Ib-IX complex re-quires the presence of its three subunits. J BiolChem. 1992;267(18):12851-12859.

15. Kenny D, Morateck PA, Gill JC, Montgomery RR.The critical interaction of glycoprotein (GP) Ibbetawith GPIX–a genetic cause of Bernard-Souliersyndrome. Blood. 1999;93(9):2968-2975.

16. Mo X, Nguyen NX, McEwan PA, et al. Binding ofplatelet glycoprotein Ibbeta through the convexsurface of leucine-rich repeats domain of glyco-protein IX. J Thromb Haemost. 2009;7(9):1533-1540.

17. Mo X, Lu N, Padilla A, Lopez JA, Li R. The trans-membrane domain of glycoprotein Ibbeta is criti-cal to efficient expression of glycoprotein Ib-IXcomplex in the plasma membrane. J Biol Chem.2006;281:23050-23059.

18. Luo S-Z, Mo X, Lopez JA, Li R. Role of the trans-membrane domain of glycoprotein IX in assemblyof the glycoprotein Ib-IX complex. J Thromb Hae-most. 2007;5(12):2494-2502.

19. Luo SZ, Li R. Specific heteromeric association offour transmembrane peptides derived from plate-let glycoprotein Ib-IX complex. J Mol Biol. 2008;382(2):448-457.

20. Barton WA, Liu BP, Tzvetkova D, et al. Structureand axon outgrowth inhibitor binding of theNogo-66 receptor and related proteins. EMBO J.2003;22(13):3291-3302.

21. Keegan RM, Winn MD. Automated search-modeldiscovery and preparation for structure solutionby molecular replacement. Acta Crystallogr D BiolCrystallogr. 2007;63(Pt 4):447-457.

22. McCoy AJ, Grosse-Kunstleve RW, Adams PD,Winn MD, Storoni LC, Read RJ. Phaser crystallo-graphic software. J Appl Crystallogr. 2007;40(Pt4):658-674.

23. Emsley P, Cowtan K. Coot: model-building toolsfor molecular graphics. Acta Crystallogr D BiolCrystallogr. 2004;60(Pt 12 Pt 1):2126-2132.

24. Murshudov GN, Vagin AA, Dodson EJ. Refine-ment of macromolecular structures by the maxi-mum-likelihood method. Acta Crystallogr D BiolCrystallogr. 1997;53(Pt 3):240-255.

25. Mo X, Luo S-Z, Lopez JA, Li R. Juxtamembranebasic residues in glycoprotein Ibbeta cytoplasmicdomain are required for assembly and surfaceexpression of glycoprotein Ib-IX complex. FEBSLett. 2008;582:3270-3274.

26. Howitt JA, Clout NJ, Hohenester E. Binding sitefor Robo receptors revealed by dissection of theleucine-rich repeat region of Slit. EMBO J. 2004;23(22):4406-4412.

27. Uff S, Clemetson JM, Harrison T, Clemetson KJ,Emsley J. Crystal structure of the platelet glyco-protein Ib� N-terminal domain reveals an un-

masking mechanism for receptor activation. J BiolChem. 2002;277(38):35657-35663.

28. Gonzalez-Manchon C, Butta N, Iruin G, Alonso S,Ayuso MS, Parrilla R. Disruption of the Cys5-Cys7 disulfide bridge in the platelet glycoproteinIbbeta prevents the normal maturation and sur-face exposure of GPIb-IX complexes. ThrombHaemost. 2003;90(3):456-464.

29. Kunishima S, Naoe T, Kamiya T, Saito H. Novelheterozygous missense mutation in the plateletglycoprotein Ibbeta gene associated with isolatedgiant platelet disorder. Am J Hematol. 2001;68(4):249-255.

30. Hillmann A, Nurden A, Nurden P, et al. A novelhemizygous Bernard-Soulier Syndrome (BSS)mutation in the amino terminal domain of glyco-protein (GP)Ibbeta–platelet characterization andtransfection studies. Thromb Haemost. 2002;88(6):1026-1032.

31. Strassel C, Pasquet JM, Alessi MC, et al. A novelmissense mutation shows that GPIbbeta has adual role in controlling the processing and stabilityof the platelet GPIb-IX adhesion receptor. Bio-chemistry. 2003;42(15):4452-4462.

32. Kunishima S, Tomiyama Y, Honda S, et al. Ho-mozygous Pro74–Arg mutation in the plateletglycoprotein Ibbeta gene associated with Ber-nard-Soulier syndrome. Thromb Haemost. 2000;84(1):112-117.

33. Kurokawa Y, Ishida F, Kamijo T, et al. A missensemutation (Tyr88 to Cys) in the platelet membraneglycoprotein Ibbeta gene affects GPIb/IX complexexpression–Bernard-Soulier syndrome in thehomozygous form and giant platelets in theheterozygous form. Thromb Haemost. 2001;86(5):1249-1256.

34. Kunishima S, Lopez JA, Kobayashi S, et al. Mis-sense mutations of the glycoprotein (GP) Ibbetagene impairing the GPIb alpha/beta disulfide link-age in a family with giant platelet disorder. Blood.1997;89(7):2404-2412.

35. Tang J, Stern-Nezer S, Liu PC, et al. Mutation inthe leucine-rich repeat C-flanking region of plate-let glycoprotein Ibbeta impairs assembly of vonWillebrand factor receptor. Thromb Haemost.2004;92(1):75-88.

36. McEwan PA, Andrews RK, Emsley J. Glycopro-tein Ib� inhibitor complex structure reveals acombined steric and allosteric mechanism of vonWillebrand factor antagonism. Blood. 2009;114(23):4883-4885.

37. Ponstingl H, Henrick K, Thornton JM. Discriminat-ing between homodimeric and monomeric pro-teins in the crystalline state. Proteins. 2000;41(1):47-57.


38. Kang JY, Nan X, Jin MS, et al. Recognition of li-popeptide patterns by Toll-like receptor 2-Toll-likereceptor 6 heterodimer. Immunity. 2009;31(6):873-884.

39. Savoia A, Pastore A, De Rocco D, et al. Clinicaland genetic aspects of Bernard-Soulier syn-drome: searching for genotype/phenotype corre-lations. Haematologica. 2011;96(3):417-423.

40. Ludlow LB, Schick BP, Budarf ML, et al. Identifi-cation of a mutation in a GATA binding site of theplatelet glycoprotein Ibbeta promoter resulting in

the Bernard-Soulier syndrome. J Biol Chem.1996;271(36):22076-22080.

41. Moran N, Morateck PA, Deering A, et al. Surfaceexpression of glycoprotein Ib� is dependent onglycoprotein Ibbeta: evidence from a novel muta-tion causing Bernard-Soulier syndrome. Blood.2000;96(2):532-539.

42. Kunishima S, Matsushita T, Ito T, Kamiya T, SaitoH. Novel nonsense mutation in the platelet glyco-protein Ibb gene associated with Bernard-Souliersyndrome. Am J Hematol. 2002;71(4):279-284.

43. Strassel C, David T, Eckly A, et al. Synthesis ofGPIbbeta with novel transmembrane and cyto-plasmic sequences in a Bernard-Soulier patientresulting in GPIb-defective signaling in CHOcells. J Thromb Haemost. 2006;4(1):217-228.

44. Sachs UJ, Kiefel V, Bohringer M, Afshar-Kharghan V, Kroll H, Santoso S. Single aminoacid substitution in human platelet glycoproteinIbbeta is responsible for the formation of theplatelet-specific alloantigen Iy(a). Blood. 2000;95(5):1849-1855.


quaternary organization of gpib-ix complex and insights into bernard-soulier syndrome revealed by...

Documents