A Ca²⁺ Switch Triggers Formation of Inactive Factor X_a Dimers on Phosphatidylserine-Containing Membranes.

Factor X_a has a prominent role in amplifying both inflammation and coagulation cascades. In the coagulation cascade, its main role is catalyzing the proteolytic activation of prothrombin to thrombin. Efficient proteolysis is well known to require phosphatidylserine (PS)-containing membranes that are provided by platelets in vivo. PS, in the presence of Ca²⁺, triggers tight association of factor X_a with its cofactor, factor V_a. An interesting complication is that PS also triggers tight association of factor X_a with factor X_a, at least in solution (Majumder R, Wang JF, and Lentz BR. Biophys. J. 2003, 84:1238–1251), to form an inactive factor X_a dimer (Sen S, and Lentz BR, unpublished). In this work, we ask whether Ca²⁺ and PS also trigger formation of an inactive factor X_a dimer on a membrane and explore the possible physiological significance of this.

We have determined the proteolytic activity of human factor X_a towards human prethrombin2 as a substrate both at fixed membrane concentration (increasing factor X_a concentration) and at fixed factor X_a concentration (increasing membrane concentration). Neither of these experiments showed the expected behavior of an increase in activity as factor X_a bound to membranes. The total factor X_a activity actually decreased as low concentrations of PS-containing membranes were added, and increased at higher membrane concentrations. At fixed membrane concentrations, the total factor X_a activity did not increase proportionally with factor X_a concentration. Both observations suggested the existence of membrane-bound and inactive multimeric forms of factor X_a. Because we have observed factor X_a to form dimers in solution (Majumder R, Wang JF, and Lentz BR. Biophys. J. 2003, 84:1238–1251), we tried to fit globally four such data sets to a model that takes into account dimerization of factor X_a after binding to a membrane. This dimer model successfully described all our data, with the parameters of best fit being k_cat/K_M^dimer = 0 M⁻¹s⁻¹, k_cat/K_M^monomer = 430 M⁻¹s⁻¹, k_cat/K_M^solution = 38 M⁻¹s⁻¹, and K_d,surface^dimer = 4·10⁻¹² mol/(dm)². This surface dimerization constant corresponds to a solution-phase K_d^dimer = 10 nM at 100 μM lipid concentration, nearly what we observed for formation of bovine factor X_a dimer in the presence of short-chain PS (20 nM; Majumder R, Wang JF, and Lentz BR. Biophys. J. 2003, 84:1238–1251). Also consistent with the dimer hypothesis, we observed that the binding of factor X_a to PS-containing membranes appears to be tighter at low than at high membrane concentration. As we observed for soluble-PS-induced dimer formation in solution, dimer formation on a membrane was Ca²⁺ dependent. Unlike in solution, factor X_a was activated by membrane binding below 1.5 mM Ca²⁺, but inactivated above this Ca²⁺ concentration. This suggests that factor X_a activity may be regulated by Ca²⁺ concentrations close to plasma Ca²⁺ levels.

We conclude that:

1. factor X_a dimerizes on PS-containing membranes;

2. , factor X_a dimer is inactive; and

3. , the transition from monomer to dimer state depends critically on Ca²⁺ concentration.