Cryo-Electron Microscopy Structure of the aIIbβ3-Abciximab Complex

Abstract

DN, YZ, AS, & JL contributed equally, as did MF, TW, & BSC The αIIbβ3 antagonist antiplatelet drug abciximab, approved in 1994, is the chimeric antigen-binding fragment (Fab) comprising the variable regions of murine mAb 7E3 and human IgG1 and light chain κ constant domains. In studies involving thousands of patients undergoing percutaneous coronary interventions, abciximab decreased mortality and the risk of recurrent myocardial infarction. Mutagenesis studies conducted by us and others (Puzon-McLaughlin, JBC 2000; Takagi, Biochem 2002; Artoni, PNAS 2004) suggested that abciximab binds to the β3 C177-C184 specificity-determining loop (SDL) and Trp129 on the adjacent β1-α1 helix, and our negative-stain electron microscopy (EM) studies of the complex of mAb 7E3 with αIIbβ3 in nanodiscs (Choi, Blood 2013) supported its binding to the αIIbβ3 head domain. None of these studies, however, had the resolution to assess whether 7E3 or abciximab prevents fibrinogen binding by steric interference, disruption of the αIIbβ3-binding pocket for fibrinogen, or both. To address this knowledge gap, we used cryo-EM to produce a density map at 2.8-Å resolution, which allowed us to build an atomic model of the αIIbβ3-abciximab complex. The interacting surface of abciximab is comprised of residues from all three complementarity determining regions of both the light and heavy chains, with high representation of aromatic residues (Figure). Abciximab buries a total of 1,273 Å2 of solvent-exposed surface on αIIbβ3, of which 1,040 Å2 is with β3 and 218 Å2 is with αIIb. The binding of abciximab does not result in disruption of the ADMIDAS, MIDAS, or SyMBS metal ion regions, but it does produce an ~3.4 Å compression of the SDL. Binding is primarily to the β3 SDL and neighboring residues, the β1-α1 helix, and β3 Ser211-Val212 and Met335. The latter residue interacts with the ADMIDAS metal ion in the unliganded receptor; ligand binding leads to the loss of the interaction and a dramatic swing-out motion of the β3 subunit that produces a high-affinity ligand-binding conformation. Surprisingly, the structure also indicated several abciximab interactions with αIIb. To assess the contribution of individual interactions between αIIbβ3 integrin and abciximab to the stability of the αIIbβ3-abciximab complex, we carried out 4 independent molecular dynamics simulations of the cryo-EM structure as well as the X-ray crystal structure of ligand-free αIIbβ3 integrin in its closed conformation (PDB: 3FCS). These revealed that the dynamic behavior of the RGD peptide-binding pocket was similar between the cryo-EM structure of the αIIbβ3-abciximab complex and the X-ray structure of unbound αIIbβ3 during the total simulation time of 2 microseconds per system. Abciximab-protein interaction analysis of the simulations demonstrated that: 1. The αIIb subunit participated marginally in the interaction with abciximab, with only the Asp159(αIIb)-Arg7(light chain) making contact for >0.5 fraction of the simulation time. 2. The β3 residues with >0.7 contact fraction values were Lys125, Asp126, Trp129, and Gln132 on the β1-α1 helix; Glu171, Asn175, Tyr178, Lys181, Thr182, and Thr183 on the SDL; and Val212, Met335, and Asp336. We calculated the effect of every possible mutation at each residue involved in significant inter-molecular interaction with abciximab in terms of changes in free energy of binding, and the resulting relative values were compared to experimental mutagenesis data. Thus, we made αIIb Asp159Ala and β3 Met335Asp mutations, the latter producing the analogous murine residue and the mutation predicted to be most disruptive to abciximab binding. We found no effect of either mutation on the binding of either mAb 7E3 or abciximab as judged by flow cytometry. Our data demonstrate unexpected interactions of abciximab with several αIIb residues and β3 Met335. Most importantly, abciximab binding did not alter the atomic structure or dynamics of the RGD-binding pocket in the timescale of the simulation, and so unless it induces allosteric modulation over a longer time scale, it does not appear to disrupt the RGD-binding pocket. Abciximab did, however, compress the SDL, which is not a component of the RGD-binding pocket but contributes to ligand binding by a still undefined mechanism. Thus, our data are most consistent with abciximab preventing ligand binding by steric interference, with a potential contribution via alteration of the SDL. Figure Disclosures Coller: Centocor/Janssen: Patents & Royalties: abxicimab; Accumetrics/Instrumentation Laboratory: Patents & Royalties: VerifyNow assay; Scholar Rock: Consultancy, Equity Ownership; CeleCor: Consultancy, Equity Ownership, Research Funding.