RESUMO
Conventional antithrombotic drug discovery requires testing of large numbers of drug candidates. We used computer-aided macromolecular interaction assessment (MIAX) to select antithrombotic molecules that mimic and therefore block platelet GPIb's binding to von Willebrand factor (vWf), an early step in thrombus formation. We screened a random array of 15-mer D-amino acid peptides for binding vWf. Structures of 4 candidate peptides were inferred by comparison to sequences in protein databases, conversion from the L to D conformations and molecular dynamics (MD) determinations of those most energetically stable. By MIAX, we deduced the amino acids and intermolecular hydrogen bonds contributing to the GPIb-vWf interaction interface. We docked the peptides onto vWf in silico to localize their binding sites and consequent potential for preventing GPIb-vWf binding. In vitro inhibition of ristocetin-initiated platelet agglutination confirmed peptide function and suitability for antithrombotic development, thereby validating this novel approach to drug discovery.