The activation of coagulation factors IX and X by tissue factor(TF)/VIIa is the key step in initiating coagulation. The mechanisms by which FX binds to phospholipid and interacts with TF/VIIa remains disputed; and the reported dissociation constants between FX and FXa to well-defined phospholipid vary by orders of magnitude. In accord with other investigators we have found that at low lipid to TF ratios, adding additional lipid vesicles lacking TF, can accelerate the activation of FX. This acceleration has previously been attributed to the physically unrealistic notion that bound FX, regardless of whether it be bound to vesicles containing TF or naked vesicles lacking TF, is the preferred substrate for TF/VIIa. We have explored an alternative approach, namely that this acceleration is due to relief of Xa-mediated inhibition. Whereas it is well known that FXa can bind to TF/VIIa thus inhibiting further activation of FX, the presence of additional lipid appears to segregate FXa away from the enzymatic complex and localize it to lipid far from the reactive site. Using classical initial velocity measurements of product formation (typically 30 pM TF), we found that as we increased the lipid to TF ratio from 1.1 to 15 to 110 (x103), the level of inhibition achieved by 40 nM Phe-Phe-Arg-chloromethylketone-inactivated Xa (Xai), dropped from 67% to 40% to 10%. This concept is supported further by our new observation, evaluated by ellipsometry, that at saturation 30% PS/PC surfaces (phosphotidylserine/phosphatidylcholine) bind exactly twice as much FXa as FX; The measured dissociation constants (Kd) for FX and FXa binding to our lipid were 34.5 ± 2.9 nM and 53.9 ± 2.7 nM respectively and have been verified using independent techniques. We also consider herein the likely possibility that lipid-bound FXa exists on the surface as an opposing dimer. At very high lipid to TF ratios (>2x106) we observed reaction velocities decreasing with increasing lipid concentration. This decrease in velocity was associated with an increase in the apparent Km, which we show is due to loss of available substrate to phospholipid surfaces far from the reactive site. The interactions between TF/VIIa, FX, FXa and phospholipid surfaces, although extensively studied, have not been well understood nor well-agreed upon. The true parameters describing these interactions are needed to correctly interpret typical TF/VIIa experimental results as well as to model and understand the roles of these proteins in coagulation. Results of a computational kinetic model addressing the interplay of these reaction components will be presented.

Author notes

Corresponding author

Sign in via your Institution