Inverse Effects of EDTA on αIIbβ3-Mediated Adhesion to Immobilized Fibrinogen Versus the D98 Plasmin Fragment of Fibrinogen

αIIbβ3 is known to mediate adhesion of platelets to immobilized fibrinogen through its interaction with the C-terminal γ chain dodecapeptide (γ12) and EDTA inhibits the adhesion by binding divalent cations required for ligand interaction with the β3 metal ion-dependent adhesion site (MIDAS) cation. Studies by several groups, however, suggest that αIIβb3 can also interact with other sites on fibrin(ogen). To identify potential additional sites of interaction between fibrinogen and αIIbβ3, we studied the adhesion of HEK293 cells expressing αIIbβ3 (αIIbβ3-HEK) to the D98 plasmin fragment of fibrinogen, which lacks the γ12 peptide. The D98 fragment did not contain the γ12 peptide as judged by both immunoblotting with mAb 7E9 (anti-γ12) and mass spectroscopy. αIIbβ3-HEK did not bind to immobilized D98 (10 µg/ml coating concentration) in the presence of 2 mM Ca²⁺/1 mM Mg²⁺, but they did bind to immobilized intact fibrinogen (10 µg/ml coating concentration) and the adhesion was inhibited by mAbs 10E5 (anti-αIIbβ3), 7E3 (anti-αIIbβ3 + αVβ3), and 7E9. Adhesion of αIIbβ3-HEK to fibrinogen was nearly eliminated by 10 mM EDTA [13,007 ± 3,676 vs 304 ± 331 arbitrary fluorescence intensity units (AFU); n=9; p<0.001]. Unexpectedly, and in dramatic contrast, 10 mM EDTA increased adhesion of αIIbβ3-HEK to D98 nearly 25-fold, from 458 ± 601 to 10,718 ± 3,106 AFU (n=9; p=0.001). The adhesion to D98 in the presence of EDTA was not inhibited by mAb 7E9 or mAb LM609 (anti-αVβ3), and was inhibited by mAb 7E3 by less than 10%.  EDTA-dependent adhesion was, however, inhibited by mAb 10E5, which binds to the αIIb cap domain and inhibits fibrinogen binding to αIIbβ3, by 85% ± 4% (n=7; p=0.003). Dose-response experiments demonstrated that 3 mM EDTA was sufficient to block adhesion to fibrinogen and 3-4 mM EDTA was required to enhance adhesion to D98. Adding the β3 D119 mutation to αIIbβ3-HEK (αIIbβ3-D119-HEK), which disrupts the MIDAS, eliminated adhesion to fibrinogen (αIIbβ3-HEK: 17,342 ± 6,148 vs. αIIbβ3-D119-HEK: 0 ± 241 AFU; n=3; p=0.006), but had little or no effect on the binding to D98 in the presence of EDTA (αIIbβ3-HEK: 11,363 ± 3,700 vs. αIIbβ3-D119-HEK: 9,026 ± 3,252 AFU; n=3; p=0.054). However, unlike EDTA-dependent adhesion of αIIbβ3-HEK to D98, the adhesion of αIIbβ3-D119-HEK was inhibited by mAb 10E5 by only 20% ± 19% (n=3; p=0.247).

We conclude that EDTA exposes one or more sites on αIIbβ3 that bind(s) to a site(s) on immobilized D98 that is either not accessible or not expressed on intact fibrinogen. These data are consistent with the known effect of EDTA in altering the conformation of αIIbβ3 as judged by its exposing “ligand-induced” binding sites recognized by mAbs such as AP5 and PMI-1, even in the absence of ligand, and the ability of αIIbβ3 to mediate interactions with fibrin to support clot retraction even in the presence of EDTA. EDTA-treated αIIbβ3 may, therefore, provide insights into potential ancillary sites of interaction between αIIbβ3 and fibrin(ogen).