PARtitioning protease signaling

In this issue of Blood, Aisiku et al describe a novel class of protease-activated receptor-1 (PAR1) inhibitors that block proinflammatory pathways but spare cytoprotective signaling in endothelial cells.¹  These compounds, parmodulins, target the cytoplasmic face of PAR1, where they selectively interfere with Gα_q, but not Gα_12/13. This strategy of blocking specific pathways provides the ability to modulate the activity of receptors with multiple functions (such as PAR1) and may have therapeutic advantages.

Proteases initiate signaling pathways in a variety of cell types via PARs, which are a family of G-protein–coupled receptors (GPCRs) that can initiate signaling through multiple G-protein pathways.²  PAR1 was originally identified as the receptor responsible for thrombin-induced activation of platelets. As such, PAR1 has been intensely studied as an antiplatelet target. These studies culminated in March 2014 with the US Food and Drug Administration’s approval of the PAR1 antagonist vorapaxar (Zontivity). The advisory panel ruled that there was an overall benefit despite the increased risk of intracranial hemorrhage in some patient populations.³  Given these bleeding complications and PAR1’s role on other cell types in addition to platelets, there is a continued interest in developing unique PAR1 antagonists.

Vorapaxar is a classic orthosteric inhibitor that interacts at the ligand-binding site, which inhibits all signaling from PAR1 (see figure). An alternative approach is to block PAR1’s interactions with intracellular signaling molecules. This was first described using lipidated peptides based on the receptor sequence called pepducins.⁴  Depending on the sequence used, these can act as agonists or antagonists. The parmodulin compounds described in the accompanying article by Aisiku et al take advantage of PAR1’s ability to signal through multiple G-proteins and β-arrestin.⁵  Signaling through these distinct pathways is termed biased agonism, whereby receptors can be directed to signal through distinct pathways by select agonists (biased agonists) or through allosteric modulators.

The compounds described here, which build upon earlier efforts by the Flaumenhaft group,⁶  were identified by a high-throughput screen for the ability to inhibit platelet-dense granule release. Receptor chimeras mapped the binding epitope to the C-terminal tail. Interestingly, the binding of parmodulins to PAR1 does not appear to change the overall shape of the receptor because it does not alter the binding affinity of a PAR1 ligand. The authors used human platelets to demonstrate that parmodulins specifically block Gα_q, but not Gα_12/13. The effects on platelets were tested in vivo with the mouse laser injury thrombosis model and tail clip assay. A complicating factor with the in vivo studies is that mouse platelets have a different repertoire of PARs on their surface and do not express PAR1; PAR4 is the primary signaling receptor on mouse platelets. In contrast to human PAR4, mouse PAR4 has conserved features similar to human PAR1 that allow parmodulins to function in mouse platelets.^1,6  However, many GPCRs have these conserved features, and there is the chance of off-target effects. The authors have minimized this concern by performing a screen of several GPCRs and show specificity for PAR1.

Perhaps the most intriguing findings are that the proinflammatory pathways mediated by thrombin are inhibited by parmodulins, whereas the cytoprotective pathways mediated by activated protein C (APC) are unaffected. In these experiments, the authors use barrier integrity and apoptosis as cellular readouts. However, the specific pathways mediating these events were not examined. One can predict from previous studies that Gα_i and β-arrestin signaling may not be influenced by parmodulins.^5,7  These data suggest that parmodulins are primarily blocking Gα_q pathways; this needs to be formally tested.

Endogenous regulation of PAR1 activation and signaling in endothelial cells is regulated by heterodimerization,⁸  cofactors, subcellular localization,^9,10  and alternative cleavage sites, all of which have the potential to be differentially affected by orthosteric inhibitors and parmodulins. Like many GPCRs, PAR1 can heterodimerize with multiple partners, and each of these interactions can influence the cellular consequences of receptor activation. The PAR1-PAR2 heterodimer is of particular interest in the current context. Following PAR1 activation by thrombin, PAR1 is able transactivate PAR2.⁸  Because vorapaxar is directed to the ligand-binding site of PAR1, one would expect that it would not influence PAR2 transactivation. In contrast, parmodulins may influence PAR1-PAR2 interactions either directly, by disrupting heterodimers, or indirectly, through PAR1 allostery. For PAR1 activation by APC, the endothelial protein C receptor (EPCR) is a cofactor that facilitates cleavage and sequesters PAR1 into caveolae, which protects it from cleavage by thrombin. The compartmentalization of PAR1 into caveolae may protect it from interacting with parmodulins, which would allow APC-mediated cytoprotective signaling. This is consistent with parmodulins primarily blocking Gα_q pathways, because APC signals through Gα_i and β-arrestin.^5,7  However, these ideas need to be tested empirically given the substantial differences between platelets and endothelial cells.

Prolonged exposure to vorapaxar results in endothelial cell apoptosis and loss of barrier function. This is in contrast to parmodulins, where endothelial cell function was not affected. It would be intriguing if a subset of PAR1 molecules on the surface of the endothelium were protected from parmodulins and this subset was able to mediate the protective effects. An alternative possibility is that vorapaxar influences the constitutive trafficking of PAR1, which leads to disruption of endothelial cell functions.

Future studies need to examine how parmodulins influence interactions with other proteins or membrane localization and the effects that these have on cell function. The specific mechanism or pathways through which parmodulins are working will not only shed light on these compounds but also help to unravel the complex regulation of PAR1 activation on endothelial cells. In summary, the study by Aisiku and colleagues advances the possibilities of targeting PAR1 by selectively inhibiting specific pathways. This study demonstrates the value of nontraditional compounds that are directed away from the ligand-binding site. This strategy may have important implications for allosteric modulators of GPCRs in general.