The interaction of lipoprotein(a) [Lp(a)] with platelets is not well defined, particularly with regards to the individual contribution of the protein components of Lp(a), the apo B-100 and the apolipoprotein apo(a). This study investigated the binding of different recombinant apo(a) [r-apo(a)] isoforms, to human platelets and its effect on platelet aggregation. Scatchard analysis of saturation binding experiments demonstrated that human platelets display a single class of high affinity r-apo(a) binding sites (71 +/- 46 molec./platelet, Kd = 5.6 +/- 2.0 nmol/L). Platelet activation with strong agonists (thrombin, arachidonic acid) increased 2- to 10-fold the r-apo(a) binding, without affecting the affinity. Competition assays showed that the binding sites are highly specific for r-apo(a) and Lp(a). At high concentration t-PA could also bind to the r-apo(a) binding sites. By contrast, neither fibrinogen nor plasminogen inhibited to the r-apo(a) binding. The lysine analogue EACA inhibits the binding of r-apo(a) to platelets, thus suggesting the involvement of lysine residues in that interaction. Moreover, the r-apo(a) binding to platelets is unlikely mediated by GPIIb/IIIa-attached fibrin since it is not affected by platelet treatment with either LJ-CP8, a monoclonal antibody that specifically blocks fibrinogen binding to GPIIb/IIIa, nor GPRP, an inhibitor of fibrin polymerisation. Finally, we show that the distinct recombinant apo(a) proteins, as well as native Lp(a), promote an aggregation response of platelets to otherwise subaggregant doses of arachidonic acid. This proaggregant effect of r-apo(a) is dependent on its binding to platelets since it requires a minimum incubation time, and it is prevented by EACA at concentration inhibiting the r-apo(a)-platelet interaction. These results suggest that the prothrombotic action of Lp(a) may be in part mediated by modulating the platelet function through the interaction of its apo(a) subunit with a specific receptor at the platelet surface.