Summary

Although discovered as potent inhibitors of αIIbβ3-mediated platelet aggregation, snake venom disintegrins are now known to bind to other integrins according to different degrees of potency and specificity. More recently, homologues of the disintegrinlike loop have been found as a discrete domain in the ADAM family, yet the potency and specificity of each of these domains in terms of integrin binding is relatively unknown. In this present study, we have selected the disintegrin-like domain (dd) of ADAM-15 (designated as ddADAM-15), the only RGD containing domain in the ADAM family, for a structure/function study. Experimentally, the ddADAM-15 and a number of mutants in which the RGD-containing loop was substituted by cognate regions from ADAM-2, -12 and -19 were tested in terms of integrin- binding activity. For comparison with ADAM-15, an addi- tional mutant (dddenADAM-15) was designed based upon the RGD-containing loop of snake venom dendroaspin, a disintegrin- like integrin antagonist. The results showed that ddADAM- 15 is an inhibitor of platelet aggregation, though with less potency than dddenADAM-15. None of the other mutants exhibited significant inhibition of platelet aggregation. ddADAM-15 was found to have higher binding ability for α2β1 and α9β1 than the ADAM-2 derived mutant which appeared to be more selective for αVβ3 and α4β1 than either ddADAM-15 or its ADAM-19 based mutant. The integrin-binding properties of ddADAM-15 were completely abolished by point mutation within the RGD motif (R64GD→A64GD). These results suggest a more subtle contribution of this loop sequence to defining the functionality of the ADAMs compared to dendroaspin.