Abstract

Phosphatidylethanolamine (PE) and glucosylceramide (GlyCer) are cell surface lipids that preferentially enhance anticoagulant, rather than procoagulant pathways. In particular, both PE and GlyCer enhance the anticoagulant activity of activated protein C (APC). Previous studies have indicated that specific APC Gla domain residues may mediate APC interaction with PE and GlyCer. To investigate whether specific APC residues mediate PE and GlyCer enhanced APC anticoagulant activity, we expressed a series of APC variants in which APC Gla domain residues not shared with the human prothrombin Gla domain were substituted with their prothrombin amino acid equivalent. The anticoagulant activity of each APC Gla domain variant was assessed in a tissue factor-initiated thrombin generation assay containing phospholipid vesicles of differing composition (80% PC/20% PS); or PC/PS/PE (60%/20%/20%); or PC/PS/GlyCer (60%/20%/20%). For each of these lipid mixtures, thrombin generation (endogenous thrombin potential, ETP) was not significantly different in the absence of APC. In the presence of PC/PS vesicles, APC reduced thrombin generation by 63±3% at the highest APC concentration tested (6nM). However, APC impairment of thrombin generation was enhanced 3-fold in the presence of PC/PS/PE compared with vesicles containing PC/PS alone, and in the presence of PC/PS/GlyCer was enhanced 4.3-fold. Enhancement of anticoagulant function by PE and GlyCer was similar for the majority of the APC Gla domain variants tested. Interestingly, one APC variant (APC-I18V) exhibited similar anticoagulant activity to that of wild type APC with PC/PS vesicles, but was not enhanced by the presence of PE- or GlyCer-containing vesicles. Phospholipid vesicles containing PE or GlyCer have been previously described to enhance protein S cofactor enhancement of APC. Therefore, to further characterize APC-I18V, we assessed the ability of wild type APC and APC-I18V to be enhanced by protein S in the presence of PC/PS, PC/PS/PE or PC/PS/GlyCer using a protein S-sensitive thrombin generation assay. In the presence of PC/PS, increasing protein S concentration in protein S-deficient plasma resulted in an APC-mediated slow decrease in thrombin generation, irrespective of whether wild type or APC-I18V was used (IC50 for protein S-mediated APC inhibition of thrombin generation with PC/PS, 130nM). However, in the presence of PC/PS/PE or PC/PS/GlyCer, thrombin generation was impaired by wild type APC at 3–4-fold lower protein S concentration than that observed when PC/PS vesicles alone were used (IC50, PC/PS/PE=31.5nM and PC/PS/GlyCer=37.5nM). APC-I18V, however, did not exhibit a similarly increased sensitivity to protein S in the presence of PE or GlyCer, as the anticoagulant activity of this variant was the same as when only PC/PS was included. To investigate whether the loss of specific neutral lipid enhancement in APC-I18V affected its ability to initiate cytoprotective signaling via EPCR-PAR-1 on endothelial cells, the capacity of APC-I18V to inhibit thrombin-induced endothelial cell barrier permeability was assessed. When cells were pre-treated with either wild type APC or APC-I18V, there was a significant enhancement in barrier integrity and attenuation of thrombin-induced permeability (P<0.05), demonstrating that loss of PE/GlyCer enhancement of APC anticoagulant activity does not adversely affect EPCR binding and EPCR/PAR-1 cytoprotective signaling. Collectively, these results suggest PE and GlyCer enhancement of APC anticoagulant activity is mediated by increased sensitivity to protein S, and that Ile-18 in the APC Gla domain is critical for mediating APC-specific functional enhancement by PE/GlyCer.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author