Rare genetic variants in SMAP1, B3GAT2, and RIMS1 contribute to pediatric venous thromboembolism

Recent genome-wide association studies (GWAS) have confirmed known risk mutations for venous thromboembolism (VTE) and identified a number of novel susceptibility loci in adults. Here we present a GWAS in 212 nuclear families with pediatric VTE followed by targeted next-generation sequencing (NGS) to identify causative mutations contributing to the association. Three single nucleotide polymorphisms (SNPs) exceeded the threshold for genome-wide significance as determined by permutation testing using 100 000 bootstrap permutations (P <10(-5)). These SNPs reside in a region on chromosome 6q13 comprising the genes small ARF GAP1 (SMAP1), an ARF6 guanosine triphosphatase-activating protein that functions in clathrin-dependent endocytosis, and beta-1,3-glucoronyltransferase 2 (B3GAT2), a member of the human natural killer 1 carbohydrate pathway. Rs1304029 and rs2748331 are associated with pediatric VTE with unpermuted/permuted values of P = 1.42 x 10(-6)/2.0 x 10(-6) and P = 6.11 x 10(-6)/1.8 x 10(-5), respectively. Rs2748331 was replicated (P = .00719) in an independent study sample coming from our GWAS on pediatric thromboembolic stroke (combined P=7.88 x 10(-7)). Subsequent targeted NGS in 24 discordant sibling pairs identified 17 nonsynonymous coding variants, of which 1 located in SMAP1 and 3 in RIMS1, a member of the RIM family of active zone proteins, are predicted as damaging by Protein Variation Effect Analyzer and/or sorting intolerant from tolerant scores. Three SNPs curtly missed statistical significance in the transmission-disequilibrium test in the full cohort (rs112439957: P = .08326, SMAP1; rs767118962: P = .08326, RIMS1; and rs41265501: P = .05778, RIMS1). In conjunction, our data provide compelling evidence for SMAP1, B3GAT2, and RIMS1 as novel susceptibility loci for pediatric VTE and warrant future functional studies to unravel the underlying molecular mechanisms leading to VTE.