Characterization of the Interaction of Activated FVII with the Endothelial Protein C Receptor (EPCR) in the Rat

Abstract Activated FVII (FVIIa) has been shown to interact with EPCR resulting in long-term extravascular tissue presence in mice, well beyond its short circulating half-life. In addition, data in hemophilia patients on short term FVIIa prophylaxis indicate that clinical improvements can persist even in the post-prophylaxis period. Taken together, these data suggest that EPCR may sequester administered FVIIa in the extravascular space that retains activity, potentially explaining the hemostatic improvements that persist long after its infusion has ceased. As such, the EPCR-FVIIa interaction may have mechanistic and translational ramifications in the treatment of bleeds in hemophilia which needs to be further investigated. Unfortunately, hemophilic mice do not model the human disorder in terms of bleeding diathesis. In contrast, the novel rat model of hemophilia A (Nielsen LN et al, 2014) exhibits spontaneous bleeds that, if untreated, can be fatal. Therefore, this model can be utilized to better understand the FVIIa-EPCR interaction in the on-demand and prophylactic action of FVIIa administration in hemophilia. As an initial step towards that goal, here we characterize for the first time the rat FVIIa-EPCR interaction in vitro. We generated rat FVIIa by introduction of a PACE/Furin intracellular cleavage site (RKRRKR) between the light and heavy chains of rat FVII, an approach we have previously shown to result in secretion of two-chain activated FVII protease of various species (human, mouse and canine) in vitro. Recombinant rat FVIIa was purified in a two-chain form using conditioned medium from a HEK-293 stable cell line. The affinity of rat FVIIa to rat EPCR was assessed by incubation of the protease on engineered CHO-K1 cells that stably expressed full length rat EPCR. We found that rat FVIIa exhibited minimal binding to rat EPCR; in contrast human FVIIa bound rat EPCR with a Kd of ~400 nM. The lack of EPCR binding of rat FVIIa, similar to what we previously observed with mouse FVIIa (Pavani G et al., 2014 and 2016), complicates studies aimed at deciphering the hemostatic effects of the FVIIa-EPCR interaction in hemophilia using the rat. Therefore, we decided to engineer a rat FVIIa molecule with EPCR binding capacity so that it can be compared to rat FVIIa (non-EPCR binder) in subsequent in vivo experiments. For this, a rat FVIIa chimeric protease was generated using the first 11 amino acids of the Gla domain of rat protein C (rat FVIIa-FVRAG, with substitutions L4F, L8V, W9R, S10A, S11G). Using CHO-K1 expressing rat EPCR, we found that purified recombinant rat FVIIa-FVRAG bound rat EPCR with a Kd of ~500 nM, comparable to human FVIIa. Considering that the Gla domain, where the modifications are located, is essential for protease function, we also determined the procoagulant activity of rat FVIIa-FVRAG vs. FVIIa. We performed a thrombin generation assay in which each protease was added at increasing concentration (0-500 nM) and we measured the lag time, peak thrombin, endogenous thrombin potential. We found that rat FVIIa, rat FVIIa-FVRAG and human FVIIa (control) behaved similarly in all measured parameters for each concentration tested. In conclusion, we have defined the lack of EPCR binding of rat FVIIa and characterized a recombinant rat FVIIa molecule that can bind rat EPCR in vitro as a gain of function, but is otherwise similar to rat FVIIa. This molecule can now be used to probe the FVIIa-EPCR interaction in the novel and relevant hemophilia A rat model. Disclosures Margaritis: Bayer Hemophilia Awards: Research Funding; Bristol Myers Squibb: Other: Salary (spouse); Novo Nordisk A/S: Research Funding.