The integrin tilbfl 3 contains distinct and interacting binding sites for snake - venom

We have previously demonstrated [Lu, Williams, Deadman, Salmon, Kakkar, Wilkinson, Baruch, Authi and Rahman (1994) Biochem. J. 304, 929-936] the preferential antagonism of the interactions of the integrin aIIbflb on activated platelets with three immobilized glycoprotein ligands (fibrinogen, fibronectin and von Willebrand factor) by a selected panel of snake-venom RGD (Arg-Gly-Asp)-containing proteins including the disintegrins kistrin and elegantin, and the neurotoxin variant dendroaspin. Kistrin and dendroaspin, although structurally unrelated, contain similar amino acids flanking the tripeptide RGD and behaved as identical antagonists preferentially inhibiting platelet adhesion to immobilized fibrinogen as opposed to fibronectin. In contrast, elegantin, which shares extensive sequence similarity with kistrin but has different amino acids around the tripeptide RGD, preferentially inhibited platelet adhesion to immobilized fibronectin as opposed to fibrinogen. To develop further insights into the mechanisms underlying the preferential antagonism shown by the venom proteins in the adhesion studies, we, in the present study, sought to determine the binding properties of kistrin, elegantin and dendroaspin to the zIIjbfl complex by radioligand kinetic and competition studies. In direct binding experiments, both kistrin and dendroaspin were observed to bind to a single class of binding site on ADP-activated platelets with apparent equilibrium dissociation constant (Kdapp) values of42 + 2 nM and 21+6 nM respectively. In competition studies, dendroaspin blocked the binding of 1251-labelled kistrin to ADP-activated platelets in a simple competitive manner, with an apparent equilibrium inhibition constant (K,app) of 143+14 nM, from which an indirect Kdapp = 22 nM for dendroaspin was determined. This result suggests that kistrin and dendroaspin bind to the same site on the integrin acllbfl3 consistent with their similar inhibitory properties. In contrast, elegantin recognized two classes of binding sites on the aIIb/3 complex with Kdapp values of 10.5 +0.8 nM and 175+10 nM, and, unlike dendroaspin, did not inhibit the binding of '251-labelled kistrin to ADP-activated platelets. However, in reciprocal experiments both kistrin and dendroaspin inhibited the binding of 125I-elegantin to ADPactivated platelets in a non-competitive manner, with Kiapp values of 34+ 3 nM and 21 + 2 nM respectively. Thus elegantin appears to interact with distinct but interacting sites on the a IIb/ complex from the binding site of kistrin and dendroaspin, consistent with its distinctive inhibitory preferences as shown in platelet adhesion studies. Taken in conjunction with the structural similarities and differences between these three RGD-containing venom proteins, these studies provide indirect evidence that the amino acid environment flanking the residues RGD plays a significant role in determining the binding characteristics and thus functional properties of these ligands.