Alteration of the Factor X Zymogen to Protease Transition Provides Evidence for Allosteric Linkage between the S1 and FVa Binding Sites.

The zymogen to protease transition in the chymotrypsin-like serine protease family follows a well described mechanism in which bond cleavage at a highly conserved site (Arg¹⁵-Ile¹⁶; chymotrypsin numbering system) results in the unmasking of a new N-terminus that acts as an intramolecular ligand for Asp¹⁹⁴. This new salt-bridge drives a conformational change in the so-called “activation domain”, surface loops consisting of the S1 specificity pocket, oxyanion hole, autolysis loop, and sodium biding site. It is well documented in the trypsin system that Ile¹⁶-Asp¹⁹⁴ internal salt-bridge formation is allosterically linked to the S1 specificity site; that is changes at one site influence the other and vice versa. Blood coagulation factor Xa (FXa) reversibly associates with its cofactor factor Va (FVa) on a membrane surface in the presence of Ca²⁺ ions with high affinity; an interaction which is not mimicked by the zymogen FX. To determine whether the FX zymogen to protease transition contributes to the expression of a high affinity FVa binding site, we constructed a series of FXa variants which are shifted along this transition pathway. To generate these “zymogen-like” proteins, we made several substitutions at position 16 or 17, with the intent of destabilizing the intramolecular salt bridge to varying degrees. Following a series of preliminary experiments, three mutants were chosen for expression, purification, and activation with RVV-X: I16L, I16G, and V17A. Kinetic studies using peptidyl substrates and active site directed probes revealed that I16L and V17A have an impaired ability to bind these probes (15 to 25-fold increase in the K_m or K_i) while the rate of catalysis (k_cat) was reduced by 3-fold compared to wild-type FXa (wtFXa; plasma-derived and recombinant). The I16G variant was not inhibited by any of the probes examined and its chromogenic activity was severely impaired (>500 to 1000-fold), precluding calculation of kinetic parameters. These data are consistent with the idea that destabilization of internal salt-bridge formation (Ile¹⁶-Asp¹⁹⁴) influences binding at the S1 specificity site. In contrast to these results, assembly of I16L and V17A into prothrombinase almost completely restored the K_m for peptidyl substrates while the k_cat was still 3-fold reduced, indicating that FVa binding can rescue binding at the active site. Surprisingly, even the K_m value for I16G was almost completely restored (3-fold increased compared to wtFXa) when assembled in prothrombinase; however a 60-fold reduction in the k_cat was found. Consistent with these data, kinetic studies using prothrombin or prethrombin-1 revealed that each of the FXa variants had a normal K_m value when assembled in prothrombinase; while the k_cat values where reduced to a similar extent as for the chromogenic substrates. Overall our data indicate that direct binding of these FXa variants to FVa rescues binding at S1 site, suggesting allosteric linkage exists between these sites. Thus the FX zymogen to protease transition not only influences the formation of the S1 pocket, but also contributes in a substantial way to the formation of a FVa binding site.