Abstract 3259

Integrin αIIbβ3 undergoes allosteric conformational changes in its extracellular domains, resulting in integrin activation that allows high affinity binding with soluble ligands. The crystal structure of the integrin β subunit revealed an interaction of the β-tail domain (βTD) with the βI domain containing ligand-binding sites, suggesting that βTD may be involved in allosteric mechanism for integrin activation. However, previous studies have shown conflicting results on the functional role of βTD in integrin activation. In this study, we conducted site-directed mutagenesis in the βTD domain and tested ligand binding to αIIbβ3 mutants. We produced αIIbβ3 mutants in which the β3TD loop residues (DSSG) were substituted with the corresponding β1 (NGNN) or β2TD residues (DGMD). The αIIbβ3 mutants were expressed on the surface of CHO cells by cotransfection of mutant β3 and wild-type αIIb cDNAs, and were tested for binding of PAC1, a ligand-mimetic anti-αIIbβ3 antibody. The NGNN, but not DGMD mutant bound significant PAC1 binding without any stimulation, indicating a constitutively active state. To identify the residue(s) responsible for αIIbβ3 activation in the βTD, we produced αIIbβ3 mutants in which the individual residues in the β3TD loop were substituted with the corresponding β1TD residues. Among them, only G675N bound significant PAC1 binding without any stimulation. Since G675N mutation creates a sequence known to be a consensus sequence for glycosylation (Asn-X-Ser/Thr), it is possible that the insertion of glycans into the βTD loop induces conformational changes in αIIbβ3 which allow ligand binding. To test this hypothesis, we added substitution of S677 with Thr, Ala or Asp to the G675N mutation. The resultant G675N/S677T double mutant, in which the N-glycosylation site was preserved, was constitutively active. In contrast, G675N/S677A and G675N/S677D, in which the N-glycosylation site was disrupted, were in an inactive state. These results suggest that an artificial glycan wedge between βTD and βI domains activates αIIbβ3. However, our study does not provide evidence that the βTD domain constrains wild type αIIbβ3 inactive although the separation of βTD and βI domains may be able to activate integrins.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.