Sequence Determinants of Mouse Protein C Interaction with Mouse Endothelial Protein C Receptor

The Endothelial Protein C Receptor (EPCR) is an important component of the Protein C anticoagulant pathway. The interaction of Protein C (PC) through its Gla domain with EPCR enhances PC activation, thus down-regulating thrombin production. EPCR can also bind human activated Factor VII (FVIIa) and modulate its activity and localization. The residues involved in receptor recognition in both PC and FVIIa are Phe4 and Leu8, located in the first portion of the Gla domain. The importance of Phe8 is indicated by the lack of EPCR binding of a PC variant that contains a Val8 from human prothrombin. Remarkably, the similarity between PC and FVIIa for EPCR binding is lacking in the mouse. Others and we have shown that mouse FVIIa (mFVIIa, which contains a Leu4 and a Leu8) interacts with mouse EPCR (mEPCR) poorly, thus failing to model the spectrum of known human FVIIa properties. In previous work, we generated mFVIIa chimeras that contain parts of the mouse PC (mPC) Gla domain and determined that three residues in the mPC Gla domain can confer mEPCR binding to mFVIIa. Specifically, molecule mFVIIa-FMR that contained the Leu4->Phe, Leu8->Met and Trp9->Arg from mPC was functionally similar to mFVIIa and could bind mEPCR as a true gain-of-function. However, little is known on the contribution of any/all of these positions in mPC binding to mEPCR. Here, we wanted to understand the sequence determinants that dictate this interaction. For this, we generated single amino acid mutants of mPC at position 4, 8 or 9 from the corresponding residues of mFVIIa. Using conditioned medium from transiently transfected cells, we tested the ability of each mPC mutant to bind to mEPCR expressed on the surface of CHO-K1 cells. A single substitution of Phe4 with Leu abolished mEPCR binding of mPC, in contrast to modifications at position 8 (Met to Leu) or 9 (Arg to Trp). The importance of Phe4 for the mPC-mEPCR interaction was confirmed in a reverse experiment modifying mFVIIa (that has poor mEPCR affinity) individually at position 4 (Leu to Phe), 8 (Leu to Met) or 9 (Trp to Arg) according to the mPC sequence. We found that Leu4->Phe was the sole modification that could confer mEPCR binding to mFVIIa. We have previously shown that the interaction of mFVIIa-FMR with mEPCR enhances its hemostatic function (vs. mFVIIa) after administration in hemophilic mice that have undergone injury (Pavani G et al, Blood 2014). To further explore the contribution of position 4 (Leu->Phe [L4F]) in these effects, recombinant mFVIIa-L4F was purified. Titration of mFVIIa-L4F on CHO-K1 cells expressing mEPCR showed a specific and dose-dependent receptor binding, in contrast to mFVIIa, confirming our previous data (see above). Moreover, mFVIIa-L4F showed no difference in clotting activity compared to mFVIIa in a prothrombin time-based assay. In order to compare mFVIIa-L4F to mFVIIa in its ability to generate mouse thrombin, we used a thrombin generation assay using hemophilia B plasma spiked with either procoagulant. We found that addition of either mFVIIa or mFVIIa-L4F generated similar amounts of thrombin at all concentrations tested (3.1 - 25 nM). Therefore, mFVIIa-L4F exhibited similar coagulant activity to mFVIIa but gained mEPCR binding capacity, a feature shared with mFVIIa-FMR. Further experiments are underway to determine whether the single substitution in mFVIIa-L4F is sufficient to recapitulate the enhanced hemostatic properties observed in vivo with mFVIIa-FMR. In conclusion, our findings identify a single amino acid residue (Phe4) in the Gla domain of mouse PC that plays a critical role in the binding to its natural receptor. This property can also be transplanted into mFVIIa, without affecting its coagulant activity. These observations reveal another difference between human and mouse systems and may have implications for EPCR-dependent functions or properties of other vitamin K-dependent proteins.