Abstract

Previous studies of blood coagulation factor Xa (bovine) showed that binding of water soluble phosphatidylserine (C6PS) to factor Xa (FXa) leads to Ca2+ dependent inactive (∼1000-fold inactivation) dimer formation. We show now that human factor Xa activity is also regulated by C6PS-induced dimerization in the presence of 5 mM Ca2+ We also report that the FXa dimer is inactive:

  • despite the fact dimerization does not block the active site;

  • in part because it does block a substrate exosite;

  • the dimer interface involves lysine residues that boarder the active site and exosites, and

  • the structure of FXa in the dimer is altered relative to the monomer.

We have measured initial rates of prothrombin activation (using synthetic thrombin substrate S-2238) at varying FXa and substrate concentrations to show that the kcat/Km decreased (kcat decreased significantly and Km increased slightly) with an increase in FXa dimer formation. The observed significant decrease in kcat indicates that dimerization affects the alignment of substrate with the active site, perhaps through altering substrate binding or through altering the structure of the active site. Amidolytic activity of monomeric FXa (using synthetic substrate S-2765) decreased in response to C6PS binding, while that of the dimer increased slightly. This indicates that dimerization did not block the active site but may alter its conformation. CD and mass spectrometry showed that both Ca2+ and C6PS binding alter FXa structure and that dimerization further alters structure. Acetylation of exposed lysine residues and analysis of MS patterns obtained under conditions that favor either monomer or dimer FXa revealed that the dimerization buries lysines residues 222 and 224 (chymotrypsin numbering) that boarder the active site and are in putative exocytes. We used MS data, fluorescence energy transfer data for active site labeled FXa, to model the FXa dimer structure based on a FXa monomer model (from Gla-domainless Xa X-ray structure and Gla-EGFn with Ca2+) but the requirement that known membrane binding sites or paired FXa molecules would be located in plane was failed. Our lack of success supports our other measurements suggesting that the structure of FXa in a dimer is very different from that in a monomer.

Supported by grant from the NHBL (HL 072827 to BRL).

Author notes

Disclosure: No relevant conflicts of interest to declare.