The platelet integrin αIIbβ3 is the prototypic example of regulated integrin function. Thus, αIIbβ3 is present in a resting conformation on unstimulated platelets, but switches to an active conformation following platelet stimulation. Recent experiments suggest that disrupting a heteromeric interaction between the αIIb and β3 transmembrane (TM) and cytoplasmic domains shifts αIIbβ3 from its resting to its active conformation. However, structural information about the heteromeric interaction is sparse. Thus far, the structure of the TM heterodimer has only been studied by molecular modeling. Interactions between soluble cytosolic tail peptides have been studied by NMR spectroscopy, but these studies may not reflect native contacts because they fail to account for constraining TM domain interactions. To obtain an NMR structure for the αIIbβ3 cytosolic tail heterodimer that reflects its native structure, we expressed 13C- and 15N-labeled peptides corresponding to αIIb residues 988–1008 and β3 residues 713–762 in E. coli. Residues 987 in αIIb and 712 in β3 were replaced with cysteines, based on modeling that predicts the resultant disulfide bond will fix the peptides in their native orientation. Crosslinked heterodimers were dissolved in dodecylphosphocholine micelles at pH 6.5 and analyzed at 37°C on a 750 MHz NMR spectrometer. Previously, we presented a preliminary analysis of this construct indicating that when constrained by the proximal disulfide bond, the αIIb and β3 cytoplasmic tails interact and the cytosolic tail of β3 consists of three helices. We have now solved the final structure which defines the β3 interface that interacts with the αIIb cytoplasmic tail. The αIIb-β3 heterodimer interface is dynamic, but can be localized to β3 residues 716 and 719 because they have different chemical shifts in the crosslinked heterodimer than they do in the component monomers. This positions β3 residue 723 at the αIIb-β3 interface, consistent with the putative Arg995-Asp723 salt bridge. Interestingly, the αIIb tail is natively unstructured so a static interface for αIIb could not be identified. Additionally, the completed structure defines the relative orientations of the three β3 helices. The β3 cytoplasmic tail contains a sharp kink at residue 724 that fixes the membrane embedded helix (residues 713–723) and the first cytoplasmic helix (residues 725–736) at a right angle. The kink was defined by multiple NMR parameters including NOE distance restraints between residues 721 and 727. The distal cytoplasmic helix (residues 746–757) is related to the rest of the molecule by a flexible loop (residues 737– 745). N15 NOESY-HSQC crosspeak intensities provide evidence that the flexible loop and distal helix undergo increased motion relative to the first two helices, and the final structure reflects this motion because there is no preferred orientation for the distal helix relative to the first two helices. Lastly, the distal helix and flexible loop are joined by β3’s canonical NPXY motif which forms an N-terminal cap for the distal helix. In conclusion, we have solved the NMR structure of a disulfide-crosslinked αIIb/β3 cytoplasmic tail heterodimer. Our analysis indicates that, when constrained by a disulfide bond, the αIIb and β3 cytoplasmic tails interact, providing one mechanism for maintaining αIIbβ3 in a resting state.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution