Skip to main content
NCBI home page
American Journal of Human Genetics logoLink to American Journal of Human Genetics
. 2020 Jul 9;107(2):211–221. doi: 10.1016/j.ajhg.2020.06.008

A Platelet Function Modulator of Thrombin Activation Is Causally Linked to Cardiovascular Disease and Affects PAR4 Receptor Signaling

Benjamin AT Rodriguez 1,14, Arunoday Bhan 2,14, Andrew Beswick 3, Peter C Elwood 4, Teemu J Niiranen 5,6, Veikko Salomaa 5; FinnGen Study, David-Alexandre Trégouët 7, Pierre-Emmanuel Morange 8,9,10, Mete Civelek 11,12, Yoav Ben-Shlomo 13, Thorsten Schlaeger 2, Ming-Huei Chen 1, Andrew D Johnson 1,∗
PMCID: PMC7413854  PMID: 32649856

Abstract

Dual antiplatelet therapy reduces ischemic events in cardiovascular disease, but it increases bleeding risk. Thrombin receptors PAR1 and PAR4 are drug targets, but the role of thrombin in platelet aggregation remains largely unexplored in large populations. We performed a genome-wide association study (GWAS) of platelet aggregation in response to full-length thrombin, followed by clinical association analyses, Mendelian randomization, and functional characterization including iPSC-derived megakaryocyte and platelet experiments. We identified a single sentinel variant in the GRK5 locus (rs10886430-G, p = 3.0 × 10−42) associated with increased thrombin-induced platelet aggregation (β = 0.70, SE = 0.05). We show that disruption of platelet GRK5 expression by rs10886430-G is associated with enhanced platelet reactivity. The proposed mechanism of a GATA1-driven megakaryocyte enhancer is confirmed in allele-specific experiments. Utilizing further data, we demonstrate that the allelic effect is highly platelet- and thrombin-specific and not likely due to effects on thrombin levels. The variant is associated with increased risk of cardiovascular disease outcomes in UK BioBank, most strongly with pulmonary embolism. The variant associates with increased risk of stroke in the MEGASTROKE, UK BioBank, and FinnGen studies. Mendelian randomization analyses in independent samples support a causal role for rs10886430-G in increasing risk for stroke, pulmonary embolism, and venous thromboembolism through its effect on thrombin-induced platelet reactivity. We demonstrate that G protein-coupled receptor kinase 5 (GRK5) promotes platelet activation specifically via PAR4 receptor signaling. GRK5 inhibitors in development for the treatment of heart failure and cancer could have platelet off-target deleterious effects. Common variants in GRK5 may modify clinical outcomes with PAR4 inhibitors, and upregulation of GRK5 activity or signaling in platelets may have therapeutic benefits.

Keywords: platelets, thrombin, thrombosis, Cardiovascular Disease, venous thromboembolism, stroke, eQTL, GWAS, regulatory, PAR-4

Introduction

Activated platelets provide the link between inflammation, thrombosis, and atherosclerotic cardiovascular disease.1 Platelet reactivity is highly heritable,2,3 but the limited number of previously identified loci explain only a small portion of the estimated heritability.4 Despite thrombin being the principal enzyme of hemostasis and viewed as the strongest platelet agonist,5 the genetics of thrombin-induced platelet reactivity is not well understood and heretofore has not been investigated on a genome-wide scale. Many of thrombin’s cellular effects are initiated by protease-activated receptors (PARs) which are G protein-coupled receptors (GPCRs).6 PAR1 (MIM: 187930) and PAR4 (MIM: 602779) are the receptors primarily responsible for mediating the effects of thrombin in human platelets.5

Dual antiplatelet therapy (DAPT) reduces the occurrence of both stent-related and spontaneous myocardial infarction (MI) after acute coronary syndrome (ACS), but with concomitant increase in bleeding risk.7 Thus, there is a need for milder DAPT targets in order to maintain or increase efficacy while reducing bleeding, given the narrow therapeutic window of most antiplatelet treatments.8 Development of more effective strategies could potentially expand anti-platelet therapy into primary prevention where, due to bleeding risks, it generally is not recommended. Given their role in platelet biology, PAR1 and PAR4 have both been the focus of antithrombotic drug development. Targeting PAR1, the high-affinity thrombin receptor, led to vorapaxar, approved for preventing thrombotic events in patients with MI when used in combination with standard-of-care DAPT.9 Of limited clinical utility, vorapaxar is associated with increased risk of major bleeding events,9 where the mortality risk due to bleeding can be comparable to or greater than that due to MI.10 Compared to targeting PAR1, there is evidence that targeting PAR4 is associated with a lower bleeding risk and can achieve an effective antithrombotic profile, though large trials of PAR4 inhibition are still lacking.11, 12, 13

While prospective studies have demonstrated the association of platelet function with cardiovascular disease (CVD) events in patients with established CVD, there is less existing evidence that platelet function predicts CVD or CVD outcomes in the healthy population.14 Clinical trials have demonstrated a relationship between high on-treatment platelet reactivity and adverse clinical ischemic events, but tailoring therapy based on platelet reactivity remains uncertain.15 Platelet function traits remain relatively unexplored in large populations, in particular for thrombin and PAR1/PAR4 platelet activation. In order to address this broad knowledge gap—to discover genes that may mediate CVD or bleeding risk, classify treatment subpopulations or suggest new therapeutic targets—we performed the first genome-wide association study (GWAS) of thrombin-induced platelet aggregation.

Subjects and Methods

Participants and Genome-wide Analyses

We conducted and present analyses from the following methodological approaches: (1) a GWAS of platelet thrombin activation, (2) platelet and other cell and tissue expression quantitative trait loci (eQTL) analyses, including Mendelian randomization (MR), for our lead locus, and signal co-localization analyses, (3) two-sample MR for CVD outcomes from multiple consortia and biobanks, (4) integration of megakaryocyte and other epigenetic data at our lead locus, (5) site-directed mutagenesis and regulatory enhancer assays in three cell backgrounds, and (6) induced pluripotent stem cell (iPSC)-derived megakaryocyte, and platelet, small interfering RNA (siRNA) and chemical inhibitor experiments to dissect functional effects on platelets. The samples and resources utilized are described in Table S1. The demographic characteristics of the GWAS sample are shown in Table S2.

The Caerphilly Prospective Study assessed platelet aggregation induced by full-length thrombin (0.056 units/mL, Sigma Aldrich) in middle-aged males through the use of light transmission aggregometry (LTA).16 All participants provided written informed consent. Genotyping of 1,248 samples was performed with the Affymetrix UK BioBank Axiom array. Following sample and genotyping quality control, imputation was done on 1,184 samples through the use of the Haplotype Reference Consortium (HRC) 1.1 panel. We conducted a GWAS using a linear mixed model adjusting for age and medication usage. A significance threshold of p < 7 × 10−9 was adopted to account for all variants tested. Conditional analyses adjusting for the strongest peak SNP in GRK5 (MIM: 600870), rs10886430, were conducted by adding SNP dosage as a covariate to the base model. Our methods for platelet aggregation, genotyping, quality control procedures, imputation, and GWAS are further described in the Supplemental Information.

Causal Analysis and Multi-trait Colocalization for GRK5 Locus

Testing for causal association with eQTL was conducted with the use of the platelet RNA and expression 1 (PRAX1) platelet eQTL dataset.17 We investigated whether thrombin-induced aggregation and platelet cell traits share a common association signal at the GRK5 locus by performing a co-localization analysis with quantitative traits derived from multiple blood cell lineages: platelets (mean platelet volume [MPV], platelet count [PLT], and platelet distribution width [PDW]), red blood cell count (RBC), and white blood cell count (WBC).18 Furthermore, we conducted similar genetic colocalization analyses for the GRK5 locus through the use of genome-wide study data for thrombin generation potential;19 platelet aggregation to ADP, collagen, and epinephrine;4,20 eQTLs from 44 cells and tissues from the Genotype-Tissue Expression (GTEx) Project version 7; five white blood cell types and platelets from the CEDAR project;21 and aortic endothelial cells.22 Two-sample MR analysis was conducted using the rs10886430-G allele as the genetic instrument and thrombin-induced platelet aggregation as exposure in separate analyses for nine pulmonary, stroke, or heart disease outcomes from the UK BioBank,23 four stroke outcomes from the MEGASTROKE consortium,24 and CVD codes in the FinnGen Biobank (version 4).25 Only this SNP was used in MR analyses because in conditional analyses of chromosome 10, this was the only independently significant SNP associated with thrombin reactivity or platelet GRK5 expression levels. Details regarding these analyses are given in the Supplemental Information.

Regulatory Function

We integrated epigenetic regulatory maps of chromatin accessibility, enhancer RNA (eRNA), histone marks, enhancer elements, and DNA-binding proteins assessed in megakaryocyte-erythroid lineage models (ENCODE)26,27 and cultured primary megakaryocytes (BLUEPRINT) to annotate potential functional impacts of the intronic GRK5 rs10886430 variant.28,29 Protein network analysis of transcriptional regulators that bind the rs10886430 variant was performed with STRING 10.5. We used site-directed mutagenesis to investigate the impact of the rs10886430 variant on enhancer activity in vitro in GATA1- (MIM: 305371) and GATA2- (MIM: 137295) overexpressing HEK293 cells, as well as HUVEC and K562 cells. Details regarding the data integration, network analysis, enhancer reporter luciferase assays, and conditional overexpression of GATA1 and GATA2 are provided in the Supplemental Information.

Platelet Function siRNA and Inhibitor Experiments

The role of G protein-coupled receptor kinase 5 (GRK5) in platelet function was assessed through the use of siRNA as well as pharmacologic inhibition of G protein-coupled receptor kinase (GRK) activity. Immortalized megakaryocyte progenitor cell lines (imMKCLs) were generated from human induced pluripotent stem cells and maintained as previously described.30 For siRNA experiments, imMKCLs were transfected with GRK5 or control EGFP siRNA for 48 h, then analyzed using qRT-PCR and in vitro flow cytometric analysis of platelet function (via P-selectin and PAC-1) following stimulation with either 20 μM ADP plus 20 μM thrombin receptor-activating peptide 6 (TRAP-6 that activates platelets via PAR1) or control. Pharmacologic inhibition of GRK activity was investigated with the pan-GRK inhibitor CCG215022 (MedChemExpress) which exhibits nanomolar IC50 values against both GRK2 as well as GRK5 and selectivity against PKA.31 Platelet-rich plasma (PRP) samples obtained from healthy donors (n = 3) were treated with 0.78 μM CCG215022 or DMSO vehicle for 45 min prior to stimulation with increasing concentrations of one of three platelet agonists: PAR4 Activating Peptide (PAR4-AP, 1, 20, 50 μM), TRAP-6 (1, 10, 20 μM), ADP (1, 10, 20 μM) or vehicle control followed by flow cytometric analysis of platelet function (via P-selectin and PAC-1). Details regarding the siRNA experiments in imMKCL, GRK inhibition in PRP samples, and platelet function assessment by flow cytometry are provided in the Supplemental Information.

Results

We performed a GWAS of thrombin-induced platelet aggregation in the Caerphilly Prospective Study including >7.75 million common and low-frequency (minor allele frequency [MAF] > 0.01) SNPs imputed via the HRC panel. There was no evidence for inflation of test statistics (λ = 1.005) (Figure S1A). We observed 17 variants that surpassed the genome-wide significance threshold (p < 7 × 10−9) for association with thrombin-induced aggregation, all localized to 10q26.11 (Figure 1A). Conditional analysis identified no additional signals independent of the sentinel variant located in the first intron of the GRK5 locus (GRK5, rs10886430, p = 3.0 × 10−42) ∼43-kb downstream of the transcription start site (TSS) within consensus intron 1 (Figure S1B, Table S3). The minor G allele (MAF 0.136) of the GRK5 SNP was associated with increased platelet reactivity to thrombin (β = 0.70, SE = 0.05; with other covariates fixed, this is ∼3.9% per allele increase in thrombin reactivity) (Figure 1B). Variance component analysis indicates that this single variant explained 18.3% of variation in the thrombin phenotype. There was no significant population structure in Caerphilly, which consists overwhelmingly of participants of European ancestry, based on principal components clustering with multi-ethnic samples in the UK BioBank (Figure S2). Likewise, we found little relatedness in Caerphilly: only n = 303 had any first-, second-, or third-degree relations. Inclusion of the genetic relatedness matrix in the GWAS accounted for this. Nonetheless, we conducted a sensitivity analysis removing the n = 303 individuals and analyzing chromosome 10, and rs10886430 remained highly significant (p = 2.01 × 10−31), indicating that the results are population associations rather than strong family effects.

Figure 1.

Figure 1

Open in a new tab

Regional and SNP Associations of GRK5 with Thrombin-Induced Platelet Aggregation

(A) Locus Zoom plot of the lead SNP (rs10886430) from genome-wide association analysis of thrombin-induced platelet aggregation in Caerphilly Prospective study (n = 1,184). The plot depicts the 600-kb window flanking the rs10886430 (purple) variant which is located in the first intron of GRK5.

(B) Boxplot of the effect of rs10886430-G variant dosage on thrombin-induced platelet aggregation (0.056 U/mL). Data points are plotted as red circles. The bold horizontal line represents the median. The upper and lower hinges indicate the 25th and 75th percentiles, respectively. The whiskers extending from the hinges represent the values no further than 1.5 ∗ interquartile range.

We next asked whether thrombin-induced aggregation and platelet, red cell, and white cell count traits share a common association signal at the GRK5 locus. We performed a Bayesian test for co-localization between the Caerphilly thrombin GWAS and multiple blood cell lineage traits from the UK BioBank/INTERVAL study meta-analysis,18 interrogating shared variants in an ∼1.8 Mb independent linkage disequilibrium block containing the GRK5 lead SNP. We observed strong evidence for co-localization (posterior probability >0.99) between thrombin reactivity and platelet cell traits (MPV, PLT, and PDW) but not WBC or RBC; this result supports the hypothesis that a single variant affects these traits in platelets (Figure 2A). The GRK5 lead SNP (rs10886430) was identified as the shared, potentially causal variant in each case (Table S4). With the association limited to platelets, we next asked whether the rs10886430 variant affects platelet reactivity mediated by other agonists. We conducted a GWAS on aggregation to ADP and collagen in Caerphilly participants, and we performed colocalization analyses with the thrombin GWAS. No evidence of colocalization was observed between thrombin reactivity and either agonist (Table S5). Further, no colocalization was observed with aggregation to ADP, collagen, or epinephrine in the largest such GWAS published to date in independent cohorts4 (Table S6). We next asked whether the effect on thrombin reactivity was limited to platelet activation or was also observed in thrombin generation potential traits, as this could indicate an effect mediated via thrombin levels. No colocalization was observed between platelet reactivity to thrombin and three phenotypic markers of thrombin generation19 (Table S7).

Figure 2.

Figure 2

Open in a new tab

GRK5 rs10886430-G Is Linked to Platelet Traits and Causally Effects Platelet GRK5 Expression and Multiple CVD Outcomes

The rs10886430 variant regulates platelet cell traits and GRK5 platelet gene expression, as well as both cardiovascular and cerebrovascular disease pathologies.

(A) Heatmap of posterior probabilities from Bayesian colocalization analyses of thrombin reactivity and five blood cell traits in the 1.8 Mb LD block containing the rs10886430 variant (10q26.11). Shared SNP—probability of one shared SNP associated with both traits; Independent SNPs—probability of two independent SNPs associated with each trait; Trait 2 Only—probability of association with the blood cell trait and not with thrombin-induced aggregation; Thrombin Only—probability of association with thrombin-induced aggregation and not with the blood cell trait; No Association—probability of no association with either trait.

(B) Manhattan plot depicting summary data-based Mendelian Randomization (SMR) analysis of association between platelet gene expression and platelet reactivity to thrombin (0.056 U/mL).

(C) Mendelian Randomization analysis of thrombin reactivity (rs10886430G instrument) and cardiopulmonary phenotypes in UK BioBank (Outcomes). Plotted are the OR of the causal estimates (circles) and associated 95% confidence intervals (error bars), color of circle indicates -log transformed p value of estimate. Further description of medical outcomes codes and statistics is found in the Supplemental Information and Table S12.

To determine potential regulatory impacts of the rs10886430 variant, we first examined its association with functional expression changes cataloged in the platelet-specific PRAX1 study eQTL dataset.17,32 We utilized summary-data-based Mendelian Randomization (SMR) analysis with the PRAX1 eQTL dataset to test for association between platelet gene expression and platelet reactivity to thrombin. We identified GRK5 as the only gene at a genome-wide significance level (βSMR = −1.54, SESMR = 0.20, PSMR = 3.67 × 10−14) (Figure 2B, Table S8). Finding no evidence to suggest that the SMR association could be due to genetic linkage (PHEIDI = 0.16), we concluded that expression of GRK5 is associated with platelet reactivity driven by the rs10886430 variant. As predicted by our SMR analysis, the minor rs10886430-G allele is a strong cis-eQTL for decreased platelet GRK5 expression (βeQTL = −0.456, PeQTL = 8.27 × 10−20) in the PRAX1 study.17 We further replicated this strong GRK5 eQTL in the independent CEDAR platelet dataset21 (βeQTL = −0.429, PeQTL = 1.11 × 10−20). To investigate the tissue specificity of the genetic effect of rs10886430-G on GRK5 expression, we conducted co-localization analyses between the Caerphilly thrombin GWAS in the GRK5 locus and 51 tissue or cell types. We observed no evidence of co-localization between thrombin reactivity and GRK5 expression among 44 tissues profiled by the GTEx Project (Table S9), five other white blood cell types21 (Table S10), or vascular endothelial cells22 (Table S11).

Having observed the strong association of rs10886430-G with lower GRK5 expression exclusively in platelets, we asked whether the genetic effect of the variant on platelet reactivity was further associated with relevant cardiovascular or cerebrovascular disease pathology. We utilized GWAS statistics for nine pulmonary, stroke, or heart disease diagnoses in the UK BioBank cohort. There was evidence for an association of rs10886430-G with several diagnoses, most strongly with pulmonary embolism (PE) (odds ratio [OR] = 1.25, SE = 1.03, p = 8.1 × 10−13) (Table S12). We performed two-sample MR to test for a causal relationship between thrombin reactivity (exposure) and disease outcome for the nine UK BioBank diagnoses. We observed strong, positive association between thrombin-induced platelet reactivity at the GRK5 locus and multiple disease outcomes, most significantly with diagnosis of PE (ORMR = 1.38 [1.26 – 1.52], PMR = 2.40 × 10−11), also with diagnoses of cerebral infarction (ORMR = 1.19 [1.08 – 1.32], PMR = 7.43 × 10−4) and acute MI (ORMR = 1.14 [1.07 – 1.21], PMR = 6.67 × 10−5), but not with heart failure (ORMR = 1.04 [0.96 – 1.12], PMR = 0.39) (Figure 2C, Table S12). Next, we investigated the rs10886430-G variant in stroke subtypes from MEGASTROKE,24 the cohorts of which are independent of UK BioBank. There was evidence for a stronger association in cardioembolic stroke (OR = 18.42, SE = 2.96, p = 6.16 × 10−4) than ischemic stroke (OR = 1.61, SE = 1.15, p = 2.29 × 10−4), and no association for large-artery stroke (OR = 4.32, SE = 5.09, p = 0.39) (Table S13). In MR analyses, we observed strong, positive association between thrombin-induced platelet reactivity at the rs10886430-G variant and cardioembolic stroke (ORMR = 64.36 [3.09 – 1340.44], PMR = 7.18 × 10−3), all stroke (ORMR = 1.79 [1.19 – 1.27], PMR = 7.69 × 10−4), and ischemic stroke (ORMR = 1.97 [1.22 – 1.33], PMR = 7.17 × 10−4), but not large artery stroke (ORMR = 10.24 [0.08 – 773.88], PMR = 0.37) (Table S13). Finally, the G allele was again associated with increased risk of multiple CVD outcomes in the FinnGen study—deep venous thrombosis (DVT) of lower extremities (ORMR = 1.25 [1.11– 1.41], PMR = 2.28 × 10−4), ischemic stroke (ORMR = 1.15 [1.06 – 1.25], PMR = 5.75 × 10−4), portal vein thrombosis (ORMR = 2.51 [1.30-4.85], PMR = 6.31 × 10−3), right bundle branch block (ORMR = 1.68 [1.14 – 2.48], PMR = 8.51 × 10−3), MI (ORMR = 1.12 [1.03 – 1.22], PMR = 1.09 × 10−2), atrioventricular block (ORMR = 1.24 [1.04– 1.47], PMR = 1.64 × 10−2)—and with reduced risk for hypertension (ORMR = 0.94 [0.89 – 0.98], PMR = 7.49 × 10−3) and cardiomyopathies (ORMR = 0.79 [0.68 – 0.91], PMR = 1.85 × 10−3) (Table S14).

A GRK5 gain-of-function coding variant (p.Gln41Leu) leading to enhanced β-adrenergic receptor (βAR) desensitization of excessive catecholamine signaling has been proposed to provide a “genetic β-blockade” that improves survival in African Americans with heart failure.33 Functional studies have also shown β2-ARs can inhibit platelet aggregation and adhesion.34 In our study, this variant was not associated with thrombin reactivity (p = 0.51, data not shown). Having observed associations with multiple pathologic cardiopulmonary traits, we asked whether the GRK5 variant’s effect on platelet reactivity was modified by pharmacological β-blockade. A small subset of Caerphilly participants were taking cardiac-specific, β1-AR selective (5.1%) or non-selective drugs (3.1%) at the time platelet reactivity was measured. Fitting linear models with specific and non-specific beta-blockers as additional covariates, we observed a negative interaction effect on platelet reactivity between the GRK5 variant and β1-AR selective drugs (β = −0.60, SE = 0.24, p = 0.01) (Figure S3).

To study the regulatory function of the GRK5 variant, we next integrated cell-type-specific epigenome maps derived from primary megakaryocyte cells.28,29 The variant localizes to a region of open chromatin (DNase hypersensitivity peak site) in a predicted enhancer region characterized by broad enrichment of active marks H3K27Ac and H3K4me1 (Figure 3A). Active enhancers are often characterized by short, unstable bi-directional transcripts termed “enhancer RNAs (eRNAs).” Integrating nascent transcription maps in K562 cells, we observed that the GRK5 variant localizes to the predicted TSS region of an eRNA27 (Figure S4). Having observed multiple lines of evidence that the variant is in an enhancer element, we asked whether the position was occupied by transcriptional regulators in vivo, as this could provide a clear testable mechanism of action in a non-coding DNA context. To this end, we scanned transcription factor (TF) binding datasets from mega-erythroid cell models generated by the ENCODE consortium.26 The variant position was bound by 27 factors, including the master hematopoietic TF GATA1 in peripheral blood erythroblasts and the histone acetyltransferase p300 (MIM: 602700) in K562 cells (Table S15).

Figure 3.

Figure 3

Open in a new tab

GRK5 rs10886430 Overlaps an Active Meta-erythroid Lineage Enhancer Bound by Interconnected Factors

The rs10886430 variant resides in a megakaryocyte enhancer element.

(A) Epigenetic regulatory maps of primary-derived MK and K562 cells.

(B) STRING Protein Network Analysis of DNA binding factors recruited to the rs10886430 variant in ENCODE mega-erythroid cell models. Proteins are represented as circles. Colors indicate network cluster membership. Solid lines indicate interactions within a network cluster. Dotted lines indicate interactions between proteins in different clusters. Line colors indicate type of evidence: cyan—known interaction from curated database; magenta—known interaction experimentally determined; yellow-green—text mining; black—co-expression.

We verified that binding of the GRK5 variant by GATA1 is also observed in primary megakaryocyte (MK) cells35 and in hematopoietic stem and/or progenitor-derived erythroid precursors.36 The GATA1 binding motif is highly enriched at active eRNA in K562 cells.37 We next considered whether there was functional connectivity or association between the factors binding the GRK5 variant locus. Using the STRING algorithm, we constructed a highly connected network model incorporating 24 of 26 mappable transcriptional regulators (protein-protein interaction enrichment p value: < 1.0 × 10−16) (Figure 3B). Clustering of the network revealed several properties: SWI/SNF chromatin remodeling complex clustering with GATA1 (red cluster), transcriptional repressors (including one cluster of polycomb-related proteins) (green cluster), cAMP responsive factors (blue cluster), p300 and lineage factor IKZF1 (MIM: 603023) (khaki cluster), and NF-E2 (MIM: 601490) complex (light green cluster). These included p45 and MAFK (MIM: 600197), subunits of the heterodimeric NF-E2 complex required for megakaryocyte maturation and platelet production in vivo.38 Collectively, these results suggest that the GRK5 non-coding variant affects platelet reactivity by modulating a functional megakaryocyte lineage enhancer, leading to platelet populations with altered GRK5 expression and function.

Using luciferase reporter assay, we confirmed that the locus drives enhancer activity in vitro through binding of GATA1 and not through binding of GATA2 (Figure S5A–S5C). Enhancer activity in endogenous GATA1-expressing cells is largely abolished upon deletion of the GATA1 core binding motif (GATA1del; Figure S5B). Given its close proximity to the GATA1 motif, we hypothesized that the rs10886430-G could reduce the regulatory potential of the enhancer by interfering with binding of GATA1. Thus, we investigated the impact of the GRK5 variant on enhancer activity, both by targeted four-base deletion as well as by single-base substitution of the minor “G” allele (Figure 4). In K562 cells, introduction of the GRK5 variant effect allele (AtoGmut) repressed enhancer activity 1.5-fold (p < 0.001). Targeted deletion of the base position (AGTGdel) produced nearly identical results. Endothelial HUVEC cells, which express lower levels of GATA1 (confirmed using qRT-PCR, data not shown), exhibited weak enhancer activity and a diminished capacity for the GRK5 variant effect allele to repress enhancer activity (1.3-fold, p < 0.05) (Figure 4).

Figure 4.

Figure 4

Open in a new tab

Mutagenesis Causing Deletion or rs10886430 A to G Transition Disrupts Enhancer Activity in Multiple Cell Backgrounds

The rs10886430 variant modulates megakaryocyte enhancer activity. Top, diagram of GRK5 enhancer constructs: row 1, WT allele; row 2, four-base deletion including allele position; row 3, substitution with “G” effect allele. Bottom, normalized luciferase activity in mega-erythroid K562 cells and endothelial HUVEC. Results are the mean of three independent experiments performed in quadruplicate (n total = 12), error bars represent SEM. ∗p < 1 × 10−4, ∗∗p < 5 × 10−5.

Having established a potential mechanism by which the DNA variant regulates GRK5 expression, we next investigated the role of GRK5 in platelet physiology. To this end, we utilized iPSC-derived imMKCLs which produce functional platelets expressing cell markers CD42b (MIM: 606672) and VWF (MIM: 613160).30 We first assessed the impact of reducing GRK5 expression in imMKCL through the use of siRNA testing, achieving a knockdown efficiency of ∼75% (Figure S6A). We performed platelet function testing of GRK5-depleted and control platelet progeny cells stimulated with agonists ADP/TRAP-6 via flow cytometry, measuring platelet surface activated GPIIb-IIIa (PAC1 antibody) and platelet surface P-Selectin (CD62P). Transient knockdown of GRK5 increased the percentages of platelets that were positive for P-Selectin 1.6-fold (p < 0.05) and activated GPIIb-IIIa 1.2-fold (p < 0.05) (Figure 5A). The amounts of P-Selectin and activated GPIIb-IIIa exposed on the surface of each platelet, as judged by the geometric mean fluorescence intensity (MFI), were largely unchanged (Figure 5A).

Figure 5.

Figure 5

Open in a new tab

Disrupting Platelet GRK5 via siRNA or Chemical Inhibitor Causes Increased Thrombin Activation via a PAR4 Receptor Signaling Process

GRK5 perturbation promotes platelet activation.

(A) Change in platelet activation markers P-Selectin and activated GPIIb-IIIa (PAC1) in response to 20 μM ADP/TRAP6 stimulation (compared to untreated) following siRNA knockdown of GRK5 expression (or Green Fluorescent Protein [GFP] siRNA negative control) in imMKCL cells. Data represent mean ± SEM of four independent experiments. Differences assessed via Student’s T-Test. ∗p < 0.05, ∗∗p < 0.01.

(B) Change in P-Selectin in response to specific activation of either ADP, PAR1 (TRAP-6), or PAR4 (PAR4-AP) signaling (compared to untreated) following pharmacologic inhibition of GRK activity with 0.78 μM CCG215022 (GRKi) or vehicle control (Con) in PRP samples from healthy donors. Plotted data represent mean ± SEM (n = 3) difference in percentages of marker-positive platelets following treatment with the indicated concentrations of agonist. Differences between PRP treated with GRK inhibitor versus control assessed via Student’s T-Test. ∗p < 0.05, ∗∗p < 0.01.

We next investigated the effect of repressing GRK activity on platelet function in the specific context of PAR1- or PAR4-activated signaling ex vivo via CCG215022, a pan-specific small-molecule inhibitor exhibiting nanomolar IC50 values against both GRK2 (MIM: 109635) and GRK5. To this end, we performed platelet function testing on treated and control PRP stimulated with either TRAP-6 (PAR1 activator) or PAR4-AP via flow cytometry. Treatment with the GRK inhibitor increased the percentages of platelets positive for P-Selectin up to 2.3-fold in the presence of PAR4-AP but not in the presence of TRAP-6 (Figure S6B). Given that we observed activation in siRNA experiments upon ADP/TRAP-6 co-stimulation, the absence of an effect by PAR1 signaling alone in healthy donor PRP treated with GRK inhibitor led us to consider whether stimulation with ADP alone would have an effect. We performed additional platelet function testing with a range of concentrations of TRAP-6, PAR4-AP, or ADP following GRK inhibition (Figure 5B, Figure S7). Significant increases in P-Selectin-positive cells were again observed for PAR4-AP (20uM: 2.7-fold, p < 0.01) and to a lesser extent for ADP (1 uM: 1.2-fold, p < 0.05) but not TRAP-6 (Figure 5B, Figure S7). Because TRAP-6 activation may have been saturated at doses between 1 uM to 20 uM, we ran independent experiments at lower doses of TRAP-6 in donor PRP (0.001 uM, 0.01 uM, and 0.10 uM), and again we found no effect of GRK inhibition on platelet activation (Figure S8). Together, these results suggest that inhibition of platelet GRK5 promotes PAR4-mediated platelet activation and to a lesser extent ADP-mediated activation, but not PAR1-mediated platelet activation.

Discussion

In a GWAS of thrombin-induced platelet aggregation, we identify a GRK5 non-coding variant (rs10886430-G) strongly associated with increased reactivity to thrombin. We observed that thrombin-induced aggregation shares a common association signal at the GRK5 locus with three platelet cell indices (MPV, PLT, and PDW), the rs10886430 SNP being the shared causal variant in each case. Interestingly, the SNP had previously been identified as a sentinel variant positively associated with two indices: MPV and PDW,18 which are suggested by some as partial surrogates for platelet activation. Our study suggests a direct role for GRK5 in platelet activation, with an overall mechanism outlined in Figure 6.

Figure 6.

Figure 6

Open in a new tab

Suggested Mechanism of GRK5 Regulatory Variant in Influencing PAR4 Platelet Activation and Thrombosis Risk

Schematic depicting the platelet-specific effects of GRK5 variant rs10886430 via a GATA1 enhancer and modified suppression of platelet PAR4 signaling on thrombin activation. PAR4 has been studied as a drug target (BMS-986120). The model is supported by thrombin platelet reactivity association in PRP, a strong platelet eQTL in two independent samples, a lack of eQTL co-localization in other cells and tissues, mutagenesis and enhancer assays, siRNA and inhibitor (CCG215022) experiments in iPSC-derived megakaryocytes and platelets, and multiple population genetic studies for CVD outcomes (UK BioBank, MEGASTROKE, FinnGen, INVENT, MVP).

To determine the regulatory potential of the rs10886430 variant, we first applied SMR to test for association between platelet gene expression and reactivity to thrombin. The analysis supported a model whereby decreased GRK5 expression is associated with increased platelet reactivity through disruption by the rs10886430-G variant. The strong rs10886430-G eQTL for GRK5 replicated in two independent platelet datasets17,21 but none of the other 44 tissues in GTEx, five white blood cell subtypes, or aortic endothelial cells; these results indicate significant platelet specificity. Utilizing megakaryocyte and erythroid epigenetic datasets, we found evidence for an active cell lineage enhancer at the SNP site bound by master hematopoietic TF GATA1. GATA1 plays a critical role in megakaryocyte maturation and platelet formation in vivo.39 We supported the disruption of GRK5 expression through multiple experiments. We further investigated the role of GRK5 in platelet function through the use of an iPSC-derived megakaryocyte and platelet production model. Knockdown of GRK5 by siRNA increased functional markers of platelet activation in agonist-stimulated cells, including α-granule release (surface P-Selectin) and inside-out activation of the platelet integrin complex GPIIb-IIIa. These results indicate that GRK5 plays an important functional role in negatively regulating thrombin-induced platelet reactivity.

Thrombin is the most potent endogenous platelet activator, and it plays an important role in clot promotion and inhibition, and cell signaling, as well as additional processes that influence fibrinolysis and inflammation. The role of GRK5 could be mediated through several mechanisms: first, through canonical GRK GPCR desensitization of signal initiators PAR1 or PAR4,6 and second, GPCR desensitization of numerous downstream effectors of PAR signaling such as β2-AR (MIM: 107941),34,40 Akt (MIM: 164730), or the SRC family kinases.41 To further determine whether GRK5 affects PAR1 or PAR4 signaling, we conducted experiments that showed GRK5 inhibitory effects are mediated via inhibition of PAR4-driven platelet activation.

PAR4 is involved in sustained platelet activation, and it invokes sustained intracellular calcium response in platelets, phosphyltidylserine exposure, thrombin generation, and fibrin deposition.13 Thus, PAR4 has been suggested as a novel anti-platelet therapeutic target, with primate models and other studies indicating that PAR4 inhibition could provide superior inhibition with reduced bleeding diatheses.11,12 Our results suggest that a significant genetically determinable fraction of the population could potentially receive greater benefit from PAR4 inhibition. Interestingly, a missense variant, p.Ala120Thr, in PAR4 has been described as affecting activation.42 However, this variant was not associated with full thrombin reactivity in our study (p = 0.07, data not shown), and responses to a monoclonal antibody directed at PAR4 did not vary by genotype in another study.43 Taken together, this suggests that GRK5 variant rs10886430 may have stronger implications for thrombin-based platelet activation than does the previously reported PAR4 coding variant. Our study does have some limitations. Given the modest sample size, false negatives are expected. Because there are no other reported thrombin, PAR1, or PAR4 platelet reactivity GWASs, we cannot yet assess replication of those findings.

Given the importance of thrombin in clot formation, we used two-sample MR to determine whether the genetic effect of the variant on platelet reactivity may play a putatively causal role in cardiovascular or cerebrovascular disease pathology. Whereas the role of platelets in arterial thrombosis is well established, our MR analysis utilizing the strong GRK5 instrumental variable also suggests that thrombin-driven platelet reactivity contributes to the trajectory of venous thromboembolism (VTE), both in DVT and in PE. Recently, the variant was independently tied to increased VTE risk in the INVENT Consortium and this was replicated in the Million Veteran’s Project.44 Critical for DVT propagation in vivo, platelets are recruited to developing venous thrombi where they support leukocyte accumulation and promote formation of procoagulant neutrophil extracellular traps.45 Markers of platelet activation are elevated in acute PE, correlate with the severity of right ventricular dysfunction, and can persist for several months.46 The contributing role of platelets in VTE is further supported by the observation that aspirin therapy reduces the risk of DVT and PE in patients undergoing orthopedic surgery.47 Thrombosis and excess platelet activation are common pathological features of pulmonary arterial hypertension,48 another cause of pulmonary heart disease. While the etiologic heterogeneity characterizing ischemic stroke makes it difficult to assign a causative role for platelet reactivity to any given subtype, the platelet content of embolized thrombi is twice that of in situ thrombi,49 suggesting that the platelet aggregate increases propensity for embolization. The importance of platelets in pathogenesis of acute MI is supported by both clinical and in vivo animal studies which show that the initial thrombus (following endothelial injury) is primarily composed of activated platelets.50

In our investigation of the rs10886430-G variant among stroke subtypes available in MEGASTROKE, we observed a markedly strong effect in the prevalence of cardioembolic stroke, though not of large vessel or small vessel disease. Compared to cardioembolic stroke, the large artery atherosclerosis and small vessel occlusion subtypes have vastly different etiologies.24 MR analysis supported a causal role for thrombin-driven platelet reactivity specifically in the cardioembolic subtype. Causal associations with PE and cardioembolic stroke suggest that the rs10886430-G variant is particularly enriched in emboli forming distally to the site of vascular occlusion. More broadly, our results underline the importance of thrombin-driven platelet reactivity in both venous and arterial disease. It remains to be seen whether rs10886430-G is an important variant to segment populations at risk relative to treatment for either venous or arterial disease.

Beta-blockers are a common preventative therapy following MI and a mainstay for the management of heart failure.33 Functional studies have shown β2-ARs to inhibit platelet aggregation and adhesion through activation of platelet nitric oxide synthase.34 Given the role of GRK5 in desensitizing β-AR signaling, we investigated whether the GRK5 variant’s effect on platelet reactivity was modified by beta-blocker therapy in a subset of Caerphilly participants. We observed a negative interaction effect on platelet reactivity between the GRK5 variant and β1-AR selective drugs, and we saw no effect with non-selective drugs. The absence of an appreciable effect with the latter may in part be explained by low sample sizes in the model (selective n = 56, non-selective n = 34) as well as by differences in chemistry which may affect platelet uptake.51 Future work could potentially examine the genotypic effect of the GRK5 variant on multiple drug classes in larger samples including direct thrombin inhibitors, beta-adrenergic blockers, and other anticoagulant and anti-platelet therapies. Overexpression of cardiac GRK5 leads to early heart failure after pressure overload in mouse models.52 Also upregulated in human heart failure, GRK5 is being investigated as a therapeutic target with selective small-molecule inhibitors under development.53 Our work highlights the potential for significant platelet-driven off-target effects with this or other strategies seeking to inhibit GRK5. Notably, in platelet RNA-sequencing data, GRK5 is by far the most expressed member of the GRK family. The next highest-expressed GRK family members are GRK6 (MIM: 600869) (∼18% expression level of GRK5) and GRK4 (MIM: 137026) (∼2% expression level of GRK5), suggesting that GRK5 is likely to be the critical protein family member active in platelets.54 Given the role of GRK5 in controlling PAR4-mediated platelet activation and the association of the rs10886430-G genetic effect on platelet reactivity with cardiovascular and cerebrovascular embolic events, we suggest that finding a mechanism to maintain GRK5 activity in platelets could prove beneficial in preventing venous and arterial CVD.

Data and Code Availability

Caerphilly genotypes and phenotypes are available by request to the study steward at the University of Bristol: https://www.bristol.ac.uk/population-health-sciences/projects/caerphilly/. Other code and data are either freely available at the websites listed in the Web Resources or in the Supplemental Information, or by request to the corresponding author.

Declarations

The authors declare no competing interests.

Acknowledgments

This work was supported by National Heart, Lung, and Blood Institute (NHLBI) Intramural Research Program funding (to B.A.T., M-H.C., and A.D.J.). Additional support came from the National Blood Foundation/American Association of Blood Banks (FP01021164), the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK; U54DK110805) and the National Research Service Award (NRSA)’s Joint Program in Transfusion Medicine (T32 4T32HL066987-15 to Ar.B.). D.A.T. was supported by the «EPIDEMIOM-VTE» Senior Chair from the Initiative of Excellence of the University of Bordeaux. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the NHLBI, the National Institutes of Health, or the U.S. Department of Health and Human Services.

Published: July 9, 2020

Footnotes

Supplemental Data can be found online at https://doi.org/10.1016/j.ajhg.2020.06.008.

Web Resources

Supplemental Data

Document S1. Figures S1–S8, Tables S1–S16, FinnGen Members List, Supplementary Methods, and Supplementary References
mmc1.pdf (1.4MB, pdf)
Document S2. Article plus Supplemental Information
mmc2.pdf (3.2MB, pdf)

References

  • 1.Fuentes Q E., Fuentes Q F., Andrés V., Pello O.M., Font de Mora J., Palomo G I. Role of platelets as mediators that link inflammation and thrombosis in atherosclerosis. Platelets. 2013;24:255–262. doi: 10.3109/09537104.2012.690113. [DOI] [PubMed] [Google Scholar]
  • 2.O’Donnell C.J., Larson M.G., Feng D., Sutherland P.A., Lindpaintner K., Myers R.H., D’Agostino R.A., Levy D., Tofler G.H., Framingham Heart Study Genetic and environmental contributions to platelet aggregation: the Framingham heart study. Circulation. 2001;103:3051–3056. doi: 10.1161/01.cir.103.25.3051. [DOI] [PubMed] [Google Scholar]
  • 3.Faraday N., Yanek L.R., Mathias R., Herrera-Galeano J.E., Vaidya D., Moy T.F., Fallin M.D., Wilson A.F., Bray P.F., Becker L.C., Becker D.M. Heritability of platelet responsiveness to aspirin in activation pathways directly and indirectly related to cyclooxygenase-1. Circulation. 2007;115:2490–2496. doi: 10.1161/CIRCULATIONAHA.106.667584. [DOI] [PubMed] [Google Scholar]
  • 4.Johnson A.D., Yanek L.R., Chen M.-H., Faraday N., Larson M.G., Tofler G., Lin S.J., Kraja A.T., Province M.A., Yang Q. Genome-wide meta-analyses identifies seven loci associated with platelet aggregation in response to agonists. Nat. Genet. 2010;42:608–613. doi: 10.1038/ng.604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Crawley J.T.B., Zanardelli S., Chion C.K.N.K., Lane D.A. The central role of thrombin in hemostasis. J. Thromb. Haemost. 2007;5(Suppl 1):95–101. doi: 10.1111/j.1538-7836.2007.02500.x. [DOI] [PubMed] [Google Scholar]
  • 6.Tiruppathi C., Yan W., Sandoval R., Naqvi T., Pronin A.N., Benovic J.L., Malik A.B. G protein-coupled receptor kinase-5 regulates thrombin-activated signaling in endothelial cells. Proc. Natl. Acad. Sci. USA. 2000;97:7440–7445. doi: 10.1073/pnas.97.13.7440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Wallentin L., Becker R.C., Budaj A., Cannon C.P., Emanuelsson H., Held C., Horrow J., Husted S., James S., Katus H., PLATO Investigators Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N. Engl. J. Med. 2009;361:1045–1057. doi: 10.1056/NEJMoa0904327. [DOI] [PubMed] [Google Scholar]
  • 8.Scaglione F. New oral anticoagulants: comparative pharmacology with vitamin K antagonists. Clin. Pharmacokinet. 2013;52:69–82. doi: 10.1007/s40262-012-0030-9. [DOI] [PubMed] [Google Scholar]
  • 9.Tricoci P., Huang Z., Held C., Moliterno D.J., Armstrong P.W., Van de Werf F., White H.D., Aylward P.E., Wallentin L., Chen E., TRACER Investigators Thrombin-receptor antagonist vorapaxar in acute coronary syndromes. N. Engl. J. Med. 2012;366:20–33. doi: 10.1056/NEJMoa1109719. [DOI] [PubMed] [Google Scholar]
  • 10.Valgimigli M., Costa F., Lokhnygina Y., Clare R.M., Wallentin L., Moliterno D.J., Armstrong P.W., White H.D., Held C., Aylward P.E. Trade-off of myocardial infarction vs. bleeding types on mortality after acute coronary syndrome: lessons from the Thrombin Receptor Antagonist for Clinical Event Reduction in Acute Coronary Syndrome (TRACER) randomized trial. Eur. Heart J. 2017;38:804–810. doi: 10.1093/eurheartj/ehw525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Wong P.C., Seiffert D., Bird J.E., Watson C.A., Bostwick J.S., Giancarli M., Allegretto N., Hua J., Harden D., Guay J. Blockade of protease-activated receptor-4 (PAR4) provides robust antithrombotic activity with low bleeding. Science Translational Medicine. 2017;9:eaaf5294. doi: 10.1126/scitranslmed.aaf5294. [DOI] [PubMed] [Google Scholar]
  • 12.Wilson S.J., Ismat F.A., Wang Z., Cerra M., Narayan H., Raftis J., Gray T.J., Connell S., Garonzik S., Ma X. PAR4 (Protease-Activated Receptor 4) Antagonism With BMS-986120 Inhibits Human Ex Vivo Thrombus Formation. Arterioscler. Thromb. Vasc. Biol. 2018;38:448–456. doi: 10.1161/ATVBAHA.117.310104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.French S.L., Arthur J.F., Lee H., Nesbitt W.S., Andrews R.K., Gardiner E.E., Hamilton J.R. Inhibition of protease-activated receptor 4 impairs platelet procoagulant activity during thrombus formation in human blood. J. Thromb. Haemost. 2016;14:1642–1654. doi: 10.1111/jth.13293. [DOI] [PubMed] [Google Scholar]
  • 14.Puurunen M.K., Hwang S.J., Larson M.G., Vasan R.S., O’Donnell C.J., Tofler G., Johnson A.D. ADP Platelet Hyperreactivity Predicts Cardiovascular Disease in the FHS (Framingham Heart Study) J. Am. Heart Assoc. 2018;7:e008522. doi: 10.1161/JAHA.118.008522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sibbing D., Aradi D., Alexopoulos D., Ten Berg J., Bhatt D.L., Bonello L., Collet J.P., Cuisset T., Franchi F., Gross L. Updated Expert Consensus Statement on Platelet Function and Genetic Testing for Guiding P2Y12 Receptor Inhibitor Treatment in Percutaneous Coronary Intervention. JACC Cardiovasc. Interv. 2019;12:1521–1537. doi: 10.1016/j.jcin.2019.03.034. [DOI] [PubMed] [Google Scholar]
  • 16.Elwood P.C., Renaud S., Sharp D.S., Beswick A.D., O’Brien J.R., Yarnell J.W. Ischemic heart disease and platelet aggregation. The Caerphilly Collaborative Heart Disease Study. Circulation. 1991;83:38–44. doi: 10.1161/01.cir.83.1.38. [DOI] [PubMed] [Google Scholar]
  • 17.Simon L.M., Chen E.S., Edelstein L.C., Kong X., Bhatlekar S., Rigoutsos I., Bray P.F., Shaw C.A. Integrative Multi-omic Analysis of Human Platelet eQTLs Reveals Alternative Start Site in Mitofusin 2. Am. J. Hum. Genet. 2016;98:883–897. doi: 10.1016/j.ajhg.2016.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Astle W.J., Elding H., Jiang T., Allen D., Ruklisa D., Mann A.L., Mead D., Bouman H., Riveros-Mckay F., Kostadima M.A. The Allelic Landscape of Human Blood Cell Trait Variation and Links to Common Complex Disease. Cell. 2016;167:1415–1429.e19. doi: 10.1016/j.cell.2016.10.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Rocanin-Arjo A., Cohen W., Carcaillon L., Frère C., Saut N., Letenneur L., Alhenc-Gelas M., Dupuy A.-M., Bertrand M., Alessi M.-C., CardioGenics Consortium A meta-analysis of genome-wide association studies identifies ORM1 as a novel gene controlling thrombin generation potential. Blood. 2014;123:777–785. doi: 10.1182/blood-2013-10-529628. [DOI] [PubMed] [Google Scholar]
  • 20.Eicher J.D., Xue L., Ben-Shlomo Y., Beswick A.D., Johnson A.D. Replication and hematological characterization of human platelet reactivity genetic associations in men from the Caerphilly Prospective Study (CaPS) J. Thromb. Thrombolysis. 2016;41:343–350. doi: 10.1007/s11239-015-1290-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Momozawa Y., Dmitrieva J., Théâtre E., Deffontaine V., Rahmouni S., Charloteaux B., Crins F., Docampo E., Elansary M., Gori A.-S., International IBD Genetics Consortium IBD risk loci are enriched in multigenic regulatory modules encompassing putative causative genes. Nat. Commun. 2018;9:2427. doi: 10.1038/s41467-018-04365-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Erbilgin A., Civelek M., Romanoski C.E., Pan C., Hagopian R., Berliner J.A., Lusis A.J. Identification of CAD candidate genes in GWAS loci and their expression in vascular cells. J. Lipid Res. 2013;54:1894–1905. doi: 10.1194/jlr.M037085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Canela-Xandri O., Rawlik K., Tenesa A. An atlas of genetic associations in UK Biobank. Nat. Genet. 2018;50:1593–1599. doi: 10.1038/s41588-018-0248-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Malik R., Chauhan G., Traylor M., Sargurupremraj M., Okada Y., Mishra A., Rutten-Jacobs L., Giese A.-K., van der Laan S.W., Gretarsdottir S., AFGen Consortium. Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium. International Genomics of Blood Pressure (iGEN-BP) Consortium. INVENT Consortium. STARNET. BioBank Japan Cooperative Hospital Group. COMPASS Consortium. EPIC-CVD Consortium. EPIC-InterAct Consortium. International Stroke Genetics Consortium (ISGC) METASTROKE Consortium. Neurology Working Group of the CHARGE Consortium. NINDS Stroke Genetics Network (SiGN) UK Young Lacunar DNA Study. MEGASTROKE Consortium Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat. Genet. 2018;50:524–537. doi: 10.1038/s41588-018-0058-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Mars N., Koskela J.T., Ripatti P., Kiiskinen T.T.J., Havulinna A.S., Lindbohm J.V., Ahola-Olli A., Kurki M., Karjalainen J., Palta P., FinnGen Polygenic and clinical risk scores and their impact on age at onset and prediction of cardiometabolic diseases and common cancers. Nat. Med. 2020;26:549–557. doi: 10.1038/s41591-020-0800-0. [DOI] [PubMed] [Google Scholar]
  • 26.Wang J., Zhuang J., Iyer S., Lin X., Whitfield T.W., Greven M.C., Pierce B.G., Dong X., Kundaje A., Cheng Y. Sequence features and chromatin structure around the genomic regions bound by 119 human transcription factors. Genome Res. 2012;22:1798–1812. doi: 10.1101/gr.139105.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Core L.J., Martins A.L., Danko C.G., Waters C.T., Siepel A., Lis J.T. Analysis of nascent RNA identifies a unified architecture of initiation regions at mammalian promoters and enhancers. Nat. Genet. 2014;46:1311–1320. doi: 10.1038/ng.3142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Stunnenberg H.G., Hirst M., International Human Epigenome Consortium The International Human Epigenome Consortium: A Blueprint for Scientific Collaboration and Discovery. Cell. 2016;167:1145–1149. doi: 10.1016/j.cell.2016.11.007. [DOI] [PubMed] [Google Scholar]
  • 29.Petersen R., Lambourne J.J., Javierre B.M., Grassi L., Kreuzhuber R., Ruklisa D., Rosa I.M., Tomé A.R., Elding H., van Geffen J.P. Platelet function is modified by common sequence variation in megakaryocyte super enhancers. Nat. Commun. 2017;8:16058. doi: 10.1038/ncomms16058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Nakamura S., Takayama N., Hirata S., Seo H., Endo H., Ochi K., Fujita K., Koike T., Harimoto K., Dohda T. Expandable megakaryocyte cell lines enable clinically applicable generation of platelets from human induced pluripotent stem cells. Cell Stem Cell. 2014;14:535–548. doi: 10.1016/j.stem.2014.01.011. [DOI] [PubMed] [Google Scholar]
  • 31.Homan K.T., Waldschmidt H.V., Glukhova A., Cannavo A., Song J., Cheung J.Y., Koch W.J., Larsen S.D., Tesmer J.J.G. Crystal Structure of G Protein-coupled Receptor Kinase 5 in Complex with a Rationally Designed Inhibitor. J. Biol. Chem. 2015;290:20649–20659. doi: 10.1074/jbc.M115.647370. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Simon L.M., Edelstein L.C., Nagalla S., Woodley A.B., Chen E.S., Kong X., Ma L., Fortina P., Kunapuli S., Holinstat M. Human platelet microRNA-mRNA networks associated with age and gender revealed by integrated plateletomics. Blood. 2014;123:e37–e45. doi: 10.1182/blood-2013-12-544692. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liggett S.B., Cresci S., Kelly R.J., Syed F.M., Matkovich S.J., Hahn H.S., Diwan A., Martini J.S., Sparks L., Parekh R.R. A GRK5 polymorphism that inhibits β-adrenergic receptor signaling is protective in heart failure. Nat. Med. 2008;14:510–517. doi: 10.1038/nm1750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Queen L.R., Xu B., Horinouchi K., Fisher I., Ferro A. beta(2)-adrenoceptors activate nitric oxide synthase in human platelets. Circ. Res. 2000;87:39–44. doi: 10.1161/01.res.87.1.39. [DOI] [PubMed] [Google Scholar]
  • 35.Tijssen M.R., Cvejic A., Joshi A., Hannah R.L., Ferreira R., Forrai A., Bellissimo D.C., Oram S.H., Smethurst P.A., Wilson N.K. Genome-wide analysis of simultaneous GATA1/2, RUNX1, FLI1, and SCL binding in megakaryocytes identifies hematopoietic regulators. Dev. Cell. 2011;20:597–609. doi: 10.1016/j.devcel.2011.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Pinello L., Xu J., Orkin S.H., Yuan G.-C. Analysis of chromatin-state plasticity identifies cell-type-specific regulators of H3K27me3 patterns. Proc. Natl. Acad. Sci. USA. 2014;111:E344–E353. doi: 10.1073/pnas.1322570111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Azofeifa J.G., Allen M.A., Hendrix J.R., Read T., Rubin J.D., Dowell R.D. Enhancer RNA profiling predicts transcription factor activity. Genome Res. 2018;28:334–344. doi: 10.1101/gr.225755.117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Shivdasani R.A., Rosenblatt M.F., Zucker-Franklin D., Jackson C.W., Hunt P., Saris C.J.M., Orkin S.H. Transcription factor NF-E2 is required for platelet formation independent of the actions of thrombopoietin/MGDF in megakaryocyte development. Cell. 1995;81:695–704. doi: 10.1016/0092-8674(95)90531-6. [DOI] [PubMed] [Google Scholar]
  • 39.Vyas P., Ault K., Jackson C.W., Orkin S.H., Shivdasani R.A. Consequences of GATA-1 deficiency in megakaryocytes and platelets. Blood. 1999;93:2867–2875. [PubMed] [Google Scholar]
  • 40.Tran T.M., Jorgensen R., Clark R.B. Phosphorylation of the β2-adrenergic receptor in plasma membranes by intrinsic GRK5. Biochemistry. 2007;46:14438–14449. doi: 10.1021/bi700922h. [DOI] [PubMed] [Google Scholar]
  • 41.Kim S., Jin J., Kunapuli S.P. Relative contribution of G-protein-coupled pathways to protease-activated receptor-mediated Akt phosphorylation in platelets. Blood. 2006;107:947–954. doi: 10.1182/blood-2005-07-3040. [DOI] [PubMed] [Google Scholar]
  • 42.Edelstein L.C., Simon L.M., Lindsay C.R., Kong X., Teruel-Montoya R., Tourdot B.E., Chen E.S., Ma L., Coughlin S., Nieman M. Common variants in the human platelet PAR4 thrombin receptor alter platelet function and differ by race. Blood. 2014;124:3450–3458. doi: 10.1182/blood-2014-04-572479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.French S.L., Thalmann C., Bray P.F., Macdonald L.E., Murphy A.J., Sleeman M.W., Hamilton J.R. A function-blocking PAR4 antibody is markedly antithrombotic in the face of a hyperreactive PAR4 variant. Blood Adv. 2018;2:1283–1293. doi: 10.1182/bloodadvances.2017015552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Lindstrom S., Wang L., Smith E.N., Gordon W., van Hylckama Vlieg A., de Andrade M., Brody J.A., Pattee J.W., Haessler J., Brumpton B.M. Genomic and Transcriptomic Association Studies Identify 16 Novel Susceptibility Loci for Venous Thromboembolism. Blood. 2019 doi: 10.1182/blood.2019000435. blood.2019000435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.von Brühl M.-L., Stark K., Steinhart A., Chandraratne S., Konrad I., Lorenz M., Khandoga A., Tirniceriu A., Coletti R., Köllnberger M. Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in mice in vivo. J. Exp. Med. 2012;209:819–835. doi: 10.1084/jem.20112322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Chung T., Connor D., Joseph J., Emmett L., Mansberg R., Peters M., Ma D., Kritharides L. Platelet activation in acute pulmonary embolism. J. Thromb. Haemost. 2007;5:918–924. doi: 10.1111/j.1538-7836.2007.02461.x. [DOI] [PubMed] [Google Scholar]
  • 47.2000). Prevention of pulmonary embolism and deep vein thrombosis with low dose aspirin: Pulmonary Embolism Prevention (PEP) trial. Lancet 355, 1295–1302. [PubMed]
  • 48.Lannan K.L., Phipps R.P., White R.J. Thrombosis, platelets, microparticles and PAH: more than a clot. Drug Discov. Today. 2014;19:1230–1235. doi: 10.1016/j.drudis.2014.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Wysokinski W.E., Owen W.G., Fass D.N., Patrzalek D.D., Murphy L., McBane R.D., 2nd Atrial fibrillation and thrombosis: immunohistochemical differences between in situ and embolized thrombi. J. Thromb. Haemost. 2004;2:1637–1644. doi: 10.1111/j.1538-7836.2004.00899.x. [DOI] [PubMed] [Google Scholar]
  • 50.Silvain J., Collet J.-P., Nagaswami C., Beygui F., Edmondson K.E., Bellemain-Appaix A., Cayla G., Pena A., Brugier D., Barthelemy O. Composition of coronary thrombus in acute myocardial infarction. J. Am. Coll. Cardiol. 2011;57:1359–1367. doi: 10.1016/j.jacc.2010.09.077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Walle T., Webb J.G., Walle U.K., Bagwell E.E. Stereoselective accumulation of the β-receptor blocking drug atenolol by human platelets. Chirality. 1991;3:451–453. doi: 10.1002/chir.530030607. [DOI] [PubMed] [Google Scholar]
  • 52.Martini J.S., Raake P., Vinge L.E., DeGeorge B.R., Jr., Chuprun J.K., Harris D.M., Gao E., Eckhart A.D., Pitcher J.A., Koch W.J. Uncovering G protein-coupled receptor kinase-5 as a histone deacetylase kinase in the nucleus of cardiomyocytes. Proc. Natl. Acad. Sci. USA. 2008;105:12457–12462. doi: 10.1073/pnas.0803153105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Traynham C.J., Hullmann J., Koch W.J. “Canonical and non-canonical actions of GRK5 in the heart”. J. Mol. Cell. Cardiol. 2016;92:196–202. doi: 10.1016/j.yjmcc.2016.01.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Eicher J.D., Wakabayashi Y., Vitseva O., Esa N., Yang Y., Zhu J., Freedman J.E., McManus D.D., Johnson A.D. Characterization of the platelet transcriptome by RNA sequencing in patients with acute myocardial infarction. Platelets. 2016;27:230–239. doi: 10.3109/09537104.2015.1083543. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S8, Tables S1–S16, FinnGen Members List, Supplementary Methods, and Supplementary References
mmc1.pdf (1.4MB, pdf)
Document S2. Article plus Supplemental Information
mmc2.pdf (3.2MB, pdf)

Data Availability Statement

Caerphilly genotypes and phenotypes are available by request to the study steward at the University of Bristol: https://www.bristol.ac.uk/population-health-sciences/projects/caerphilly/. Other code and data are either freely available at the websites listed in the Web Resources or in the Supplemental Information, or by request to the corresponding author.

Articles from American Journal of Human Genetics are provided here courtesy of American Society of Human Genetics

RESOURCES