Identification of Oligomerization Motifs in the β3 Transmembrane Domain.

The integrin heterodimer αIIbβ3 resides on the platelet surface in an equilibrium between inactive (low affinity) and active (high affinity) conformations. We have reported that the transmembrane (TM) domains of its αIIb and β3 subunits engage in specific heteromeric and homomeric interactions that define the αIIbβ3 activation state. Further, we have proposed a “push-pull” hypothesis to explain how αIIbβ3 activity is regulated. Thus, processes that disrupt the TM domain heterodimer stabilize the active αIIbβ3 conformation and “push” the low affinity/high affinity equilibrium toward the activated state. Conversely, interactions that either require separation of the TM domains or are more favorable when the TM domains are apart “pull” the equilibrium in the same direction. Moreover, as would be predicted by this hypothesis, we found that adding a soluble peptide corresponding to the β3 TM domain to suspensions of gel-filtered platelets induced signaling-independent, αIIbβ3-mediated platelet aggregation, confirming that disruption of the αIIb/β3 TM domain heterodimer activates αIIbβ3. To identify the motifs responsible for oligomerization of the β3 TM domain, we used the TOXCAT assay. In TOXCAT, a chimeric protein consisting of an N-terminal ToxR’ DNA binding domain, a C-terminal maltose-binding protein domain, and an interposed β3 TM domain was expressed in the inner membrane of E. coli. TM domain-mediated dimerization of the chimeric protein then drove the transcriptional activation of a reporter gene chloramphenicol acetyl transferase (CAT). We replaced each residue in the β3 TM domain with either Leu, Ala, Val or Ile and measured the effect on CAT synthesis. The results were then used to calculate a perturbation index that reflects the mean fold-change in observed CAT activity for each of the mutants. When analyzed in this way, TOXCAT revealed that β3 helices associate homomerically with a 4 residue periodicity and that the interactive side of the helix lies along residues 700, 704, 708, and 712. Next, based on the TOXCAT results, we introduced selected mutations into the β3 TM domain of full-length αIIbβ3 and measured their effect on FITC-fibrinogen binding to αIIbβ3 expressed in CHO cells. We found that the disruptive mutations M701L, I704L, L705L, L712A, and L713A, located on the interactive side of the β3 helix, induced constitutive fibrinogen binding to αIIbβ3, whereas mutation of the intervening residues did not. Lastly, the TOXCAT and fibrinogen binding results were applied to a model of the β3 TM helix, revealing that the residues involved in oligomerization of the β3 TM domain were present as a “sticky” stripe along one face of the helix. Our results indicate that the face of the β3 helix that mediates its homomeric interaction also mediates its interaction with the αIIb TM domain; that mutations at this interface disrupt both homomeric and heteromeric interactions, and that these mutations strongly induce fibrinogen binding to αIIbβ3. Thus, our findings indicate that disruption of the α/β TM domain interface alone is sufficient to activate αIIbβ3 and that homomeric β3 interactions by themselves are not required to induce fibrinogen binding to αIIbβ3.