Multipopulation GWAS for venous thromboembolism identifies novel loci followed by experimental validation in zebrafish

Abstract

Venous thromboembolisms (VTEs) are a leading cause of morbidity and mortality. Although many genetic risk factors have been identified, a substantial portion of the heritability remains unexplained. In this study, we employed a genome-wide association study (GWAS) for VTE across 9 international cohorts of the Global Biobank Meta-Analysis Initiative to address this question, along with in vivo functional validation. In this multipopulation GWAS (VTE cases, 27 987; controls, 1 035 290), 38 genome-wide significant loci were identified, 4 of which were potentially novel. For each autosomal locus, we performed gene prioritization using 7 independent, yet converging, lines of evidence. Through prioritization, we identified genes associated with VTE through GWAS and/or functional studies (eg, F5, F11, VWF, STAB2, PLCG2, TC2N), functionally validated those that did not have evidence other than GWAS (TC2N, TSPAN15), and discovered 1 not previously associated with coagulation (RASIP1). We evaluated the function of 6 prioritized genes with strong genetic evidence, including F7 as a positive control, using laser-mediated endothelial injury to induce thrombosis in zebrafish after CRISPR/Cas9 knockdown. From this assay, we have supportive evidence for the role of RASIP1 and TC2N in the modification of human VTE and suggestive evidence for STAB2 and TSPAN15. This study expands on the currently identified genomic architecture of VTE through biobank-based, multipopulation GWASs, in silico candidate gene predictions, and in vivo functional follow-up of candidate genes.

Graphical abstract

Figure 1.

Schematic view of genes. Of the 38 genome-wide significant loci, 4 are potentially novel, and 34 are known from previous GWASs. Of the potentially novel genes, 2 have supportive evidence from recent GWAS meta-analyses.^, Six of the previously known genes were functionally validated in this study using a zebrafish model of blood clotting with 3 genes showing supportive evidence (significant validation) as the causal gene for modification of VTE in humans.

Figure 2.

Integrative gene prioritization. Autosomal genome-wide significant loci labeled by prioritized gene (x-axis) with shading for each line of evidence used in the bioinformatics-driven prioritization scheme (y-axis). Genes in bold were on the gold standard list (supplemental Table 7). The 7 lines of evidence evaluated were chosen to cover different mechanisms through which genetic variants contribute to disease risk, for example, regulatory changes vs protein perturbations. For VPS13D;DHRS3, F7;F10, and LINC02375;LINC02411, the genes had equal numbers of supporting lines of evidence. LINC00656 is also known as RP4-737E23.2. rs536995174 had 1 line of evidence each for SERPING1, SLC43A3, SLC43A1, F2, OR5AK4P, and LRRC55 and was excluded from this visualization. Genes with asterisks were selected for follow-up in a functional assay in zebrafish (Figure 3).

Figure 3.

Functional evidence for causal genes in genetically modified zebrafish.P values from Wilcoxon rank sum tests are listed at the top. The y-axis represents the experimental TTO for control and sgRNA-injected zebrafish embryos with the x-axis showing the genes targeted through CRISPR. Injections made without sgRNA served as a negative control. Factor 7 (F7) served as a positive control.