Activated protein C (APC) exerts direct cytoprotective activities on cells that are required for APC's ability to reduce mortality in murine sepsis models and that likely have potential contributions to mortality reduction in patients with severe sepsis. Although multiple receptors have been implicated to promote APC direct cytoprotective activities on various cell types, the endothelial protein C receptor (EPCR) and PAR1 play pivotal roles in our current understanding of the molecular mechanisms for APC's direct effects on cells. According to the current generally accepted model, binding of APC to EPCR on the endothelial cell membrane facilitates activation of PAR1 and induction of signaling that ultimately results in cytoprotective effects. However, this model raises the dilemma of how thrombin and APC can often have seemingly opposing effects when activating PAR1. To gain insights into mechanisms for the contrasting effects of APC vs. thrombin signaling, PAR1 activation by APC was studied in greater detail. We hypothesized that APC might cleave PAR1 not only at Arg41 but also at Arg46 with distinctly different consequences than caused by Arg41 cleavage and that this alternative cleavage could distinguish APC's from thrombin's effects. Previously we found that, in the presence of EDTA, APC cleaved a synthetic PAR1 N-terminal peptide (TR33-66) at Arg41. Now we have found that, in buffers containing CaCl2, APC cleaved the PAR1 TR33-66 peptide at Arg41 and also at an additional site distal from Arg41. Proteolysis of the TR33-66 peptide with APC resulted in fragments corresponding to TR33-41 and TR42-66 similar to thrombin. But in contrast to thrombin, a third fragment was generated by APC and the TR42-66 fragment disappeared over time with the concomitant accumulation of a novel peptide. Furthermore, incubation of thrombin-cleaved TR33-66 with APC resulted in additional proteolysis of TR42-66 and generation of the same novel fragment, indicating the possibility of a second APC cleavage site in PAR1 that was distinct and distal from Arg41. Isolation of the novel proteolytic fragments and their MALDI-TOF analysis identified Arg46 as the second APC cleavage site in the TR33-66 peptide. When cells containing wild-type EPCR were transfected with SEAP-PAR1 wild type and mutant constructs, both thrombin and APC cleaved wt-PAR1. As anticipated, efficient cleavage by thrombin was observed for R46Q-PAR1 but not R41Q-PAR1 or R41Q/R46Q-PAR1. In contrast, APC readily cleaved both R41Q-PAR1 and R46Q-PAR1 whereas cleavage of R41Q/R46Q-PAR1 by APC was negligible. APC mediated cleavage of R41Q-PAR1 and R46Q-PAR1 required the presence of functional EPCR and was not supported by the APC-binding-defective E86A-EPCR mutant. These results indicate that on cells Arg46 in PAR1 can serve as a second cleavage site for APC. Since the new PAR1 N-terminus after proteolysis acts as a tethered ligand for receptor activation, cleavage at Arg41 vs. Arg46 could potentially create structurally distinct agonists, which might help explain the divergent patterns for PAR1-mediated cytoprotective APC signaling vs. proinflammatory IIa signaling. In studies to see if the APC-induced new PAR1 N-terminus starting at Asn47 could promote signaling, we found that a synthetic peptide with the PAR1 47–66 N-terminal sequence (NPND-peptide) increased Akt phosphorylation at Ser473 in endothelial cells over 30 min whereas neither a control scrambled sequence (47-66)-peptide nor a TRAP peptide had a similar remarkable effect on Akt phosphorylation. Moreover, the NPND-peptide, but neither the scrambled sequence-related peptide nor a TRAP peptide, inhibited staurosporine-induced endothelial cell apoptosis. Thus, it appears that the new N-terminus generated by APC's cleavage at Arg46 in PAR1 generates a novel tethered ligand that can induce cytoprotective APC-like but not thrombin-like signaling characteristics. In summary, APC is capable of a unique, functionally significant cleavage of PAR1 at Arg46 that can initiate distinctive signaling compared to canonical cleavage of PAR1 at Arg41. This implies activated PAR1 is a GPCR with multiple active conformations capable of multiple, agonist selective activity profiles which may help explain the divergent patterns for cytoprotective APC signaling vs. proinflammatory thrombin signaling.
Mosnier:Scripps Research institute: Employment, Patents & Royalties. Griffin:ZZBiotech LLC: Consultancy, Membership on an entity's Board of Directors or advisory committees; Scripps Research Institute: Employment, Patents & Royalties.