The predominant PAR4 variant in individuals of African ancestry worsens murine and human stroke outcomes

Protease-activated receptor 4 (PAR4) (gene F2RL3) harbors a functional dimorphism, rs773902 A/G (encoding Thr120/Ala120, respectively) and is associated with greater platelet aggregation. The A allele frequency is more common in Black individuals, and Black individuals have a higher incidence of ischemic stroke than White individuals. However, it is not known whether the A allele is responsible for worse stroke outcomes. To directly test the in vivo effect of this variant on stroke, we generated mice in which F2rl3 was replaced by F2RL3, thereby expressing human PAR4 (hPAR4) with either Thr120 or Ala120. Compared with hPAR4 Ala120 mice, hPAR4 Thr120 mice had worse stroke outcomes, mediated in part by enhanced platelet activation and platelet-neutrophil interactions. Analyses of 7,620 Black subjects with 487 incident ischemic strokes demonstrated the AA genotype was a risk for incident ischemic stroke and worse functional outcomes. In humanized mice, ticagrelor with or without aspirin improved stroke outcomes in hPAR4 Ala120 mice, but not in hPAR4 Thr120 mice. P selectin blockade improved stroke outcomes and reduced platelet-neutrophil interactions in hPAR4 Thr120 mice. Our results may explain some of the racial disparity in stroke and support the need for studies of nonstandard antiplatelet therapies for patients expressing PAR4 Thr120.

Trizol.(11) RNA was extracted and cDNA generated as described.(11) Brain, liver, and lungs were also isolated from hPAR4 Ala/Ala and hPAR4 Thr/Thr mice and lysed in Trizol.RNA was extracted and cDNA generated as described.(11) Primers against F2RL3 (forward -GCTGCTGCATTACTCGGAC and reverse -ACGTAGGCACCATAGAGGTTG) were used with Gapdh as a housekeeping gene (forward -AGGTCGGTGTGAACGGATTTG and reverse -TGTAGACCATGTAGTTGAGGTCA).
Restriction Enzyme Digestion.DNA was isolated as described above from hPAR4 Ala/Ala and hPAR4 Thr/Thr mice and amplified using primers spanning F2RL3 (forward -CCTTGGCCCAGTTCTTTATGAG and reverse CAGACGTATAGTACCCAGACC).After amplification, NruI (New England Biolabs, Ipswich, MA) was added for 30 minutes at 37 o C. Digestion products were visualized on an agarose gel.
Protein.Platelet lysates from hPAR4 Ala/Ala and hPAR4 Thr/Thr mice were immunoblotted using a polyclonal antibody raised against a peptide corresponding to the cytoplasmic domain of hPAR4.(12) PAR4 Knockout mice PAR4 global knockout mice were a gift from Dr. Satya Kunapuli (Temple University, Philadelphia, PA) and Dr. Steven McKenzie (Thomas Jefferson University, Philadelphia, PA).(13)

Tail-Clip Bleeding Assay
Tail clip bleeding assays were performed as described before with limited modifications.(14) Briefly, anesthesia was induced by inhalation of 5% isoflurane and maintained by inhalation of 2% isoflurane.Four millimeters of the tail end was transected and immediately placed in warm PBS.The time until bleeding stopped was measured for initial and complete (defined as ceased bleeding for >1 min) bleeding cessation.If bleeding did not cease after ten minutes, the experiment was terminated.To quantify blood loss, the number of red blood cells collected during the bleeding was counted using a Hemavet (Drew Scientific; Miami Lakes, FL).
Procoagulant platelet (Annexin V positive) formation was induced by co-stimulation with human α-thrombin and CVX and detected by APC-labelled rat anti-mouse CD41 and FITClabelled annexin V (ThermoFisher).Samples were then analysed on a Beckman Coulter Cytoflex (Beckman-Coulter, Pasadena, CA) located in the Utah Flow Cytometry Core

Murine ischemic stroke model
All animal experiments complied with the regulatory standards of the University of Utah (IACUC 21-09012) and were performed following the ARRIVE guidelines (www.nc3rs.org.uk),including randomization and analysis blind to the genotype.All experiments were performed using 8-12-week-old male and female mice.
Transient middle cerebral artery occlusion stroke model.Transient middle cerebral artery occlusion (tMCAO) was performed as described previously.(15)(16)(17)(18) Briefly, occlusion of the right MCA was achieved by inserting a standardized monofilament (Doccol Corp, Sharon, MA) via the right internal carotid artery to occlude the origin of the right MCA.The occluding suture was left in situ for variable lengths of time.Induction of IS was confirmed by neurological testing of the mice while the MCA was occluded.Anesthesia was induced by inhalation of 5% isoflurane and maintained by inhalation of 2% isoflurane.Buprenorphine (University of Utah pharmacy) was administered one hour before surgery and every 12 hours as needed.Sham surgery was performed similarly, without insertion of the monofilament.The following conditions excluded mice from endpoint analyses (exclusion criteria): (1) death within 12 hours after tMCAO, (2) operation time > 10 minutes or (3) when surgical complications occurred.
Brains of dead mice were visually checked for surgical complications and stained with 2,3,5triphenyl-tetrazolium chloride (TTC, T8877; Sigma) when possible, as described below, to confirm IS-related mortality.For murine studies, mice were excluded based on pre-specific exclusion criteria, which involved a surgical bleed during the transient middle cerebral artery model.
Neurological and motor scoring.Twenty-four hours post stroke onset, mice were subjected to the modified Bederson test and the grip test to assess neurological and motor function, respectively.(15)(16)(17)(18) Determination of brain infarct size.To quantify IS brain damage, 2-mm-thick coronal brain sections were stained with 2% TTC to distinguish unaffected brain tissue from infarcted tissue, 24 hours after stroke induction.(15)(16)(17)(18) Stained slices were photographed and infarct areas (white) were measured using Image J software (National Institutes of Health; Bethesda, MD) by an operator blinded to genotype and treatment.

Platelet-neutrophil aggregate formation
Blood was collected and diluted 1:10 into M199 supplemented with 100 U/mL heparin (University of Utah pharmacy).Diluted blood was left resting or activated with different concentrations of AYPGKF for 15 minutes.For the detection of platelet-neutrophil aggregates, diluted blood was stained with APC-labelled rat anti-mouse CD41 and BV510-labeled rat antimouse Ly6G (1A8, Biolegend, San Diego, CA).(16-18)Samples were fixed with FACS lysis buffer, centrifuged at 500 x g for 10 minutes and resuspended in PBS before analysis on a Beckman Coulter Cytoflex located in the Utah Flow Cytometry Core.

Transmission Electron Microscopy
For ultrastructural analyses, platelets were adhered to Acylar coated with poly-L-lysine and fixed in 2.5% glutaraldehyde in PBS and processed as previously described (19).The samples were subsequently washed and postfixed with 2% osmium tetroxide, rewashed, dehydrated by a graded series of acetone concentrations (50%, 70%, 90%, 100%; 2 × 10 minutes each), and embedded in Epon.Thin sections were counterstained (ie, uranyl acetate and lead citrate), viewed with a JEOL JEM-1011 electron microscope, and digital images were captured with a side-mounted Advantage HR CCD camera (Advanced Microscopy Techniques).
Platelets were imaged and alpha granules counted.Alpha granules in at least 125 individual platelets were counted per independent experiment.The number of granules per independent experiment were averaged together to generate the number of granules per independent experiment.

ELISAs
Platelet Factor 4 (PF4) -Plasma samples were diluted 1:200 and PF4 was measured following the manufacturer instructions (ab202403, Abcam).Isolated platelets were lysed by freezethawing three times followed by shear through a 27-gauage needle.Platelet lysates were centrifuged at 12000xg to remove cellular debris and the analyzed.
MPO-DNA complexes -An in-house ELISA was used to quantify MPO-DNA complexes.(18,20,21) Briefly, after overnight coating with anti-MPO antibody (2 µg/ml; 0400-0002; Bio-Rad, Hercules, CA) at 4°C, a 96-well plate was blocked with 2.5% bovine serum albumin in PBS for 2 hours at room temperature.The plate was subsequently washed, before incubating for 90 minutes at room temperature with 20% human or mouse plasma in blocking buffer.The plate was washed five times, and then incubated for 90 minutes at room temperature with anti-DNA antibody (1:10; Cell Death detection ELISA, 11544675001, Sigma).After five washes, the plate was developed with TMB substrate (T0440; Sigma).

Ex vivo NET formation in humans and mice
For human NET assays, healthy human donors aged between 18 years and 50 years of selfidentified race from the greater Salt Lake City, Utah area were screened and genotyped for rs773902 as previously described.(22 Platelets were purified as described previously (18,20,21), resuspended to 1 x 10 8 cells/mL in Medium 199 and activated for 15 minutes.Activated platelets and neutrophils were incubated at a 100:1 ratio to induce NETs for 2.5 hours at 37°C in 5% CO2/95% air.NET levels were measured using an MPO-DNA ELISA as described above.We observed no statistical difference at any time point between hPAR4 Ala/Ala and hPAR4 Thr/Thr mice.N=4 mice per group.Significance determined by 2-way ANOVA with a Sidak's multiple comparisons test.(B) Whole blood platelets counts were measured in hPAR4 Ala/Ala and hPAR4 Thr/Thr mice before mice were injected intravenously with an anti-GPIba (2 µg/g) in PBS to deplete platelets.Subsequently, whole blood platelet counts were measured every 24 hours to measure platelet production.We observed no significant difference between hPAR4 Ala/Ala ) All donors provided informed consent based on approval from the University of Utah IRB (IRB 00095539).Donors were age and gender matched for the NET assays.Neutrophils were isolated from freshly collected whole blood of healthy adults or adult mice using the EasySep Direct Human Neutrophil Isolation kit (19666; Stemcell Technologies, Vancouver, Canada) or EasySep Mouse Neutrophil Enrichment Kit (19762; Stemcell Technologies), respectively, with greater than 95% purity.(18)Neutrophils were resuspended to a concentration of 1 x 10 6 cells/mL in Medium 199 (ThermoFisher).

Supplemental Figure 1 .Supplemental Figure 4 .
Proper insertion of F2RL3 into F2rl3 locus.Whole Genome sequencing (WGS) was performed on the founder mouse from both the (A) hPAR4 Ala and (B) hPAR4 Thr lines.Human sequencing reads were aligned against the mm10 genome in IGV.Paired-end human reads are identified by colored reads.Reads that align against the mouse genome are in grey.(A) In the hPAR4 Ala founder mouse, the full-length F2RL3 gene was present and was inserted at the targeted 5' and 3' junction sites.The F2RL3 allele inserted in this locus was different by one SNP (at the expected variant nucleotide) compared to the hPAR4 Thr mouse.WGS identified heterogeneity in the hPAR4 Ala founder mouse at the F2rl3 locus as both human F2RL3 and mouse F2rl3 were present.(B).The full-length F2RL3 coding region was present and was inserted at the targeted 5' and 3' junction sites in the hPAR4 Thr founder mouse.The hPAR4 Thr/Thr founder mouse was homogeneous for human F2RL3 as no mouse F2rl3 was present.Both lines were backcrossed eight generations onto C57BL/6J.After eight generational backcrosses, heterozygous hPAR4 mice were crossed with each other to generate homozygous hPAR4 Ala/Ala and hPAR4 Thr/Thr mice.Platelet half-life and production are comparable between between hPAR4 Ala/Ala and hPAR4 Thr/Thr mice.(A) hPAR4 Ala/Ala and hPAR4 Thr/Thr mice were inject intravenously with Dylight 488 anti-GPIbb antibody.Four hours later whole blood was drawn by tail poke and platelets were labeled with anti-CD41 antibody.The labeled platelet population was set to a 100% based on the percentage of CD41-positive, Dylight-positive platelets as analyzed by flow cytometry.This was set as time = 0.The percentage of the remaining CD41-positive, Dylight-positive platelets was measured subsequently every 24 hours and divided by the percentage of CD41-positive, Dylight-positive platelets at time = 0.

Supplemental Figure 10 .Supplemental Figure 12 .
and hPAR4 Thr/Thr mice at any time point measured.N=4-5 mice per group.Significance determined by 2-way ANOVA with a Sidak's multiple comparisons test.Gating strategy to measure procoagulant platelet formation.Representative flow plots depicting PS exposure based on Annexin V binding on washed platelets at baseline or after thrombin and convulxin stimulation in hPAR4 Thr/Thr .A minimum of 10,000 total CD41 positive events were counted to determine PS expression.Platelets from hPAR4 mice activate in response to mouse thrombin.Platelets were isolated from hPAR4 Ala/Ala and hPAR4 Thr/Thr mice and activated with the indicated concentrations of mouse thrombin for 15 minutes.(A) integrin activation (measured by JON/A binding) and (B) P-selectin expression was measured on CD41+ platelets by flow cytometry.N=6 per group.Significance determined by Mixed effect ANOVA with a Sidak's multiple comparisons test (A) and by a 2-way ANOVA with a Sidak's multiple comparisons test (B).Supplemental Figure 19.BMS-986120 does not alter collagen-induced aggregation in humanized PAR4 mice.BMS-986120 (10 mg/kg) was administered i.v. to hPAR4 Ala/Ala and hPAR4 Thr/Thr mice 1 hour before blood was obtained.Washed platelets were resuspended in Tyrode's buffer and collagen-induced aggregation was measured.N=5-6 per group.Normality was determined by Shapiro-Wilk test while significance was determined by Kruskal-Wallis test with a Dunn's Multiple Comparisons test.Supplemental Figure 25.Cerebral hemorrhaging associated with anti-platelet agents in hPAR4 Thr/Thr mice.Representative images of TTC brain stains of mice treated with ticagrelor or DAPT after tMCAO.White areas indicate necrotic brain tissue, arrows indicate hemorrhages.Please note, that in parallel to bleeding complications, mice regularly develop large ischemic strokes.Images are representative of N=6-11 per treatment group.

Fig. 26 .
Fig. 26.Heterozygous hPAR4 Ala/Thr mice have immediate phenotype between hPAR4 Ala/Ala and hPAR4 Thr/Thr mice depending on stroke severity and anti-platelet therapy.(A-B) Mice were subjected 40 minutes of tMCAO followed by 23 hours of reperfusion (moderate).Brains were analyzed for ischemic stroke brain damage by TTC staining, 24 hours after stroke onset.Upon TTC staining, live brain tissue will stain red, while dead brain tissue will remain white (outlined with black dotted line).(A) Images are representative of N=6-9 per group.(B) Quantification of brain infarct volumes 24 hours after stroke.Significance was determined with one-way ANOVA with a Tukey's Multiple Comparison test.N=6-9 per group.(C-I) Ticagrelor (200 mg/kg) or vehicle was administered by gavage to hPAR4 Ala/Thr mice 2 hours before tMCAO.Mice were then subjected 60 minutes of tMCAO followed by 23 hours

Table 1 .
Blood cell counts in hPAR4 Ala/Ala and hPAR4 Thr/Thr mice.

Table 2 .
Baseline characteristics of REGARDS study participants.
1Ischemic stroke cases occurred on or before September 30, 2020.Supplemental

Table 3 .
Baseline aspirin and P2Y12 inhibitor use by genotype in REGARDS.Ischemic stroke cases occurred on or before September 30, 2020.

Table 4 .
Genotype of ischemic stroke cases by TOAST classification in REGARDS Black participants.

Table 5 .
Relationship of rs773902 with incident ischemic stroke in Black participants of REGARDS (n=487 cases, 7133 noncases) using an additive and dominant model.
1Hazard ratio for each additional copy of A allele 2 Adjusted 1: adjusted for age + sex + top 5 ancestry PCs

Table 6 .
Relationship of rs773902 with functional outcomes (modified Rankin scores, mRS) immediate post-ischemic stroke among Black participants of REGARDS (N=269) using an additive and dominant model.
1Hazard ratio for each additional copy of A allele 2 Adjusted 1: adjusted for age + sex + top 5 ancestry PCs