The Common Racially Diverse Human PAR4 Variant Thr120 Causes Worse Outcomes in a Mouse Stroke Model Via Platelet-Mediated NET Formation: Prevention By P-Selectin Blockade

Background: Human platelets express two thrombin-activated G-protein coupled receptors, PAR1 and PAR4. Because PAR1 blockade is contraindicated in treating stroke, PAR4 has been considered as a potential anti-platelet target. Human PAR4 has a common, racially divergent Ala120Thr variant; the Thr120 allele frequency is ~0.7 in blacks and ~0.2 in whites. The Thr120 variant induces platelet hyperreactivity, and a previous GWAS study suggested the PAR4 Thr120 allele conferred an increased risk of ischemic stroke. Mouse platelets express Par4 and not Par1. Because human PAR4 (hPAR4) differs substantially from mouse Par4, in vivo studies of hPAR4 and the Ala120Thr variant have not been possible.

Aims: The goals of the current study were to (1) express and characterize hPAR4 in vivo, (2) determine if the PAR4 Thr120 variant results in platelet hyperactivity and worse outcomes in a murine stroke model, and (3) to test for a pharmacogenetic interaction between the hPAR4 Ala120Thr variants and FDA-approved anti-thrombotic agents for stroke outcome.

Methods: Humanized PAR4 Ala120 (hPAR4^Ala) and PAR4 Thr120 (hPAR4^Thr) mouse lines were generated using CRISPR/Cas9 to replace mouse F2rl3 with either human variant of F2RL3 under control of mouse regulatory elements. Using these mice and a transient middle cerebral artery occlusion (tMCAO) model of stroke, we assessed stroke volume and neurologic outcomes, as well as in vivo and ex vivo measures of platelet and neutrophil function.

Results: hPAR4 mRNA and protein were expressed at similar levels in hPAR4^Ala and hPAR4^Thr lines. Compared to hPAR4^Ala platelets, hPAR4^Thr platelets stimulated with the PAR4 activation peptide AYPGKF, showed a higher dose-dependent activation of aIIbb3 and P-selectin expression (p<0.05). To directly assess the effect of the hPAR4 variant on stroke outcomes, hPAR4^Ala or hPAR4^Thr mice underwent a tMCAO model of stroke. Compared to hPAR4^Ala mice, hPAR4^Thr mice had larger infarct volume (p<0.05), more circulating platelet-neutrophil aggregates (PNAs) (p<0.05), and more neutrophil extracellular traps (NETs) in the brain (p<0.05). Since platelets activate neutrophils to induce NETs that regulate ischemic stroke, we tested if purified platelets and neutrophils from each mouse line directly altered NET formation. We observed that compared to AYPGKF-activated platelets from hPAR4^Ala mice, platelets from hPAR4^Thr mice produced more NETs in vitro (p<0.05). Similar results were observed in human platelets expressing either one or two PAR4 Thr120 alleles compared to platelets expressing two PAR4 Ala120 alleles (p<0.05).

Both hPAR4^Ala and hPAR4^Thr mice were significantly protected in the tMCAO model after administration of the specific PAR4 blocker BMS-986120, indicating PAR4 is directly involved in brain injury. We then examined if expression of the PAR4 Thr120 variant altered therapeutic benefit during ischemic stroke using clinically-approved treatments. The P2Y12 blocker ticagrelor indirectly inhibits PAR4 activation and is used in brain ischemic conditions. Single agent ticagrelor administration by gavage reduced ex vivo ADP-induced aggregation in both mouse lines (indicating an effective therapeutic dose) and reduced infarct volume, PNAs, and NET formation in hPAR4^Ala mice but not in hPAR4^Thr mice (p<0.05). Moreover, dual anti-platelet therapy (combined aspirin and ticagrelor) also failed to protect hPAR4^Thr mice from ischemic stroke brain injury. Since more PNAs and NETs were produced after stroke in hPAR4^Thr mice (see above), we tested if inhibiting platelet-neutrophil interactions with anti-P-selectin therapy was beneficial. In hPAR4^Thr mice, P-selectin blockade reduced PNAs and NET formation and improved stroke outcomes (p<0.05).

Conclusion: We have generated humanized PAR4 mouse strains that directly demonstrate the role of human PAR4 in stroke. This new model will allow for novel insights into PAR4 biology in both platelets and other tissues, and provides a system to test other antithrombotic therapies in disease settings where the hyperactive PAR4 Thr120 variant induces worse outcomes. Importantly, these findings suggest blockade of platelet-neutrophil interactions may be an effective stroke treatment, especially for subjects with the hyperactive PAR4 Thr120 variant.

No relevant conflicts of interest to declare.