Phospholipid Components of the Activated Platelet Surface Determine Multimerin 1 and Factor V/Va Binding, and Modulate Thrombin Generation in Plasma.

Abstract

Multimerin 1 (MMRN1) is a factor V (FV) binding protein sequestered in endothelium and platelets, where it is costored with FV in α-granules for regulated release. Both MMRN1 and FV/Va bind to the outer surface of activated platelets and to phosphatidylserine (PS), with overlap in FV/Va binding sites for PS and MMRN1. Inhibitory effects of MMRN1 on FVa-dependent thrombin generation, and the activation dependent nature of MMRN1 binding to platelets led us to evaluate and compare MMRN1 and FV/Va binding to membranes of varying compositions. Binding was evaluated by established phospholipid binding ELISA, and by surface plasmon resonance (SPR) studies of liposome binding to immobilized MMRN1. Mixtures of PS: phosphatidylcholine (PC): phosphatidylethanolamine (PE): sphingomyelin (SM) and cholesterol representative of resting and activated platelet outer surfaces were investigated, including those typical of stimulation by thrombin, or thrombin plus collagen. Membrane compositions optimal for MMRN1 binding were compared to those that enhanced FV/Va binding and thrombin generation. Calibrated automated thrombograms were modified to compare functional effects of lipid mixtures on coagulation at low but saturating FV concentrations. ELISA results indicated that MMRN1 lipid binding was linear with PS content of PS:PC membranes from 0 to 100 mole% PS (R2=0.9). At 5% PS (where MMRN1 binding was ~18% of pure PS), addition of PE increased MMRN1 binding in a linear, dose-dependent manner (R2=0.92) up to 35% PE (the highest PE content tested), and this increased the MMRN1 binding relative to pure PS. A parallel trend was observed for FVa binding to these mixtures. While FV binding to PS:PC 5:95 was enhanced by the addition of increasing PE content (R2=0.98) up to 20 mole% PE (where the effect of PE appeared to saturate), significant binding (greater than 50% of binding to pure PS) occurred without PE, and to pure PC. Significant increases (p values <0.005) in MMRN1 and FV/Va binding were observed when a 25% PE was added to mixtures containing 5% PS. Incorporation of 33% cholesterol in the membrane had no significant effect on the binding of MMRN1 and FV; however, 68% cholesterol enhanced their binding. SPR measurements confirmed the relative binding affinities of MMRN1 for different membranes. There was no significant MMRN1 binding to membranes typical of the resting platelet surface (PS:PC:PE:SM 2:39:9:50), but MMRN1 binding was increasingly enhanced by membranes typical of thrombin stimulated platelets (5:40:25:30), randomized total platelet lipid contents (10:44:27:19) and thrombin plus collagen stimulated platelets (11:35:36:20), in that order. Lipid mixtures that were optimal for MMRN1, FV and FVa binding also enhanced coagulation, by increasing both the rates and total amounts of thrombin generated by plasma. Increases in thrombin generation were evident with increased membrane content of PS, PE or cholesterol. The results indicate that membrane lipid compositions characteristic of activated platelets enhance coagulation, and have parallel enhancing effects on MMRN1, FV and FVa binding. The data suggests that activation-induced differences in surface exposure of PS and PE, relative to PC and sphingomyelin, and local concentrations of cholesterol in the platelet membrane, alter the probability of MMRN1 and FVa binding, possibly by changing the outer membrane physical properties. These changes may influence the ability of MMRN1 to modulate FV/Va binding and function and possibly MMRN1 adhesive functions.