We have recently shown that FVIIa binds to EPCR on the endothelial cell surface [13
] and this binding could result in EPCR-FVIIa activation of PAR1-mediated cell signaling [16
] and internalization of FVIIa bound to EPCR [17
]. However, the physiological significance of this interaction in vivo
is not precisely known. In the present study, by employing EPCR-deficient and EPCR-over-expressing mice, along with wild-type, we provide convincing evidence that rFVIIa administered to mice readily binds to EPCR on the vascular endothelium and thereby reducing the circulating rFVIIa concentration immediately after rFVIIa administration. More importantly, our data show that FVIIa binding to EPCR facilitates the entry of FVIIa to extravascular tissues and the FVIIa redistributed to these compartments remained functionally active and stayed much longer there than in circulation. These data may have significant relevance to hemostasis as well as clinical treatment as rFVIIa is widely used to treat bleeding disorders associated with hemophiliacs with inhibitors as well as other bleeding complications.
The data presented in the manuscript clearly show that exogenously administered rFVIIa associates with EPCR on the endothelium even in the presence of endogenous circulating protein C, the primary ligand for EPCR. The extent of rFVIIa association with endothelium soon after its administration strongly correlates with the degree of EPCR expression on the endothelium. Endothelial and perivascular association of rFVIIa with the vasculature remained more perceptible for somewhat longer periods of time in EPCR-over-expressing mice compared with wild-type mice; nonetheless, an association of FVIIa with the endothelium decreased time-wise in all mice and was negligible 1 day after rFVIIa administration. Negligible staining of FVIIa on the vessels of EPCR-deficient mice suggest that if FVIIa specifically binds the endothelium at additional sites other than EPCR, such binding may be of lower affinity and constitutes a relatively minor fraction. The observations that an EPCR-dependent association of FVIIa with the endothelium is most pronounced soon after its administration and subsequently declines time-wise, along with the retention of FVIIa in tissue homogenates for at least 7 days, suggest that FVIIa bound to the endothelium effectively enters tissues in an EPCR-dependent manner and retained there for longer time periods. It may be pertinent to note here that in unrelated experiments we found no perceptible differences in endothelial permeability at basal conditions among these mice, which rules out the possibility of disparate extravasation as the result of variation in vascular integrity.
An EPCR-dependent transport mechanism of circulating, procoagulant FVIIa from the bloodstream to extravascular tissues may have implications in both physiological and pathophysiological circumstances. In this study, the use of pharmacological concentrations of rFVIIa and EPCR-over expressing mice has provided robustness in detecting the FVIIa transport mechanism. Nonetheless, it is likely that EPCR-dependent transcytosis of FVII/FVIIa operates in normal physiology where low but physiologically meaningful amounts of FVII or FVIIa are transported from the circulation to extravascular compartments. This mechanism would permit the formation of basal TF–FVIIa complexes that are potential primers of the coagulant and signaling responses, which could contribute to hemostasis and vascular integrity. It has been previously shown that TF–FVIIa complexes may form continuously in the adventitia of blood vessels [26
]. More recently, FVII–TF complexes were demonstrated around dermal vessels in the absence of injury [28
]. It is possible that EPCR-mediated FVII transcytosis could be responsible for the above-reported observations of the presence of TF–FVIIa complexes in adventitia. Although EPCR appears to play a critical role in FVIIa transcytosis and accumulation of FVIIa in tissues, a small amount of FVIIa may be distributed to tissues independent of EPCR as we found a small and gradual increase in FVIIa even in EPCR-deficient mice. Further, our earlier studies indicate that a small fraction of FVIIa may associate with the endothelium independent of EPCR [13
]. Although this binding appears to be specific, at present no receptor other than EPCR has been definitively identified for FVIIa on endothelial cells.
The importance of EPCR-dependent redistribution and the accumulation of extravascular FVIIa assume immediate clinical relevance in terms of therapeutic and prophylactic use in the treatment of bleeding disorders. First, it opens the possibility that the plasma concentration of rFVIIa measured after its administration may not be fully indicative of its bioavailability and subsequent hemostatic efficacy. It is possible that a slow release of rFVIIa bound to EPCR on the endothelium or from tissues back into the circulation, although not readily detectable in blood, could prolong the hemostatic effect of rFVIIa. It is also possible that rFVIIa–TF complexes formed on extravascular cell surfaces after EPCR-mediated transport could contribute to hemostasis in hemophilic patients even in the absence of rFVIIa in the circulation. Second, it could provide rationale for the prophylactic use of rFVIIa. In spite of having a circulatory half-life in the blood of approximately 2–3 h in humans, clinical evidence obtained from hemophilic patients show that once-daily administration of rFVIIa is capable of preventing bleeding episodes during the treatment period [9
]. Interestingly, the hemostatic effect of rFVIIa not only significantly surpassed the short circulatory half-life of rFVIIa during the treatment period, but can also prevent hemorrhagic events during a 3-month post-treatment period [12
]. FVIIa binding to EPCR on the endothelium may result in a long-term benefit by protecting the integrity of the vascular endothelium directly via an EPCR-FVIIa-dependent signaling mechanism [16
] or indirectly via facilitating a FVIIa interaction with extravascular TF, which in turn would lead to TF-FVIIa signaling and/or thrombin generation. The present observation that FVIIa, at least pharmacologically administered rFVIIa, binds to EPCR thereby promoting rFVIIa entry into the extravascular tissue may also have important implications on the development of second-generation rFVIIa derivatives for therapeutic purposes. For example, if one considers that the interaction of FVIIa with EPCR plays a role in hemostasis either by direct or indirect mechanisms [29
], then rFVIIa derivatives that fail to interact with EPCR, such as PEGylated rFVIIa [32
], would not be a good therapeutic option even if they were engineered to have prolonged circulatory half-lives [33
]. Conversely, if one believes that the hemostatic effect of rFVIIa solely stems from circulating rFVIIa, then engineering a rFVIIa derivate that does not bind to EPCR would prevent sequestration of rFVIIa on the endothelium and thus increase the plasma concentration of rFVIIa by about 30%.
The transcytosis of coagulation factors may have far-reaching implications in hemostasis and vascular biology as the extravascular reservoirs of coagulation factors may contribute to hemostasis and cell signaling. However, at present there is little information on potential mechanisms by which these factors enter the extravasculature space in the absence of injury to blood vessel. The present study puts forward an EPCR-dependent mechanism by which circulating FVIIa is sequestered to the vascular endothelium, redistributed to extravascular tissues and functionally retained for significant periods of time. In terms of pharmacological application of rFVIIa, the present study demonstrates that the plasma concentration may not be accurately indicative of the bioavailability of i.v. injected rFVIIa and functionally active rFVIIa may remain in tissue for a longer period. This possibility in the human system may at least to some degree explain the prolonged hemostatic efficacy of rFVIIa in prophylaxis, as well as place importance on the functional relevance of extravascularly compartmentalized FVIIa. Further studies are needed to fully evaluate the physiological significance of the present finding.