Whole-exome sequencing uncovers the genetic complexity of bicuspid aortic valve in families with early-onset complications

Abstract

Bicuspid aortic valve (BAV) is the most common congenital heart lesion with an estimated population prevalence of 1%. We hypothesize that specific gene variants predispose to early-onset complications of BAV (EBAV). We analyzed whole-exome sequences (WESs) to identify rare coding variants that contribute to BAV disease in 215 EBAV-affected families. Predicted damaging variants in candidate genes with moderate or strong supportive evidence to cause developmental cardiac phenotypes were present in 107 EBAV-affected families (50% of total), including genes that cause BAV (9%) or heritable thoracic aortic disease (HTAD, 19%). After appropriate filtration, we also identified 129 variants in 54 candidate genes that are associated with autosomal-dominant congenital heart phenotypes, including recurrent deleterious variation of FBN2, MYH6, channelopathy genes, and type 1 and 5 collagen genes. These findings confirm our hypothesis that unique rare genetic variants drive early-onset presentations of BAV disease.

Keywords: bicuspid aortic valve; cardiovascular genetics; congenital heart disease; thoracic aortic aneurysm; whole-exome sequencing.

Graphical abstract

Figure 1

Filtration workflow to select candidate EBAV variants The initial seqr project variant search generated 4,790 candidates. Missense variants that were not rated as deleterious by at least three functional predictors, had normalized CADD scores < 20, and had the same base position in two different families were removed. We prioritized genes that are known to cause dominant cardiac phenotypes and had pLOF or pDMI variants in three or more families. We excluded genes that are known to cause recessive phenotypes.

Figure 2

Enrichment of EBAV candidate gene variants The number of pDMI and pLOF variants, selected according to the criteria defined in the subjects, material, and methods, were compared to the number of variants with equivalent functional classifications from non-Finnish European ancestry populations in gnomAD v.4.1 (see supplemental methods). If pDMI or pLOF variants were present in GenTAC BAV samples, the prevalence ratio was also reported (see Table 1 for details). Black circles indicate enrichment vs. gnomAD (0–100×); red circles indicate enrichment vs. GenTAC (0–100×); horizontal lines indicate 95% confidence intervals; orange boxes indicate segregates with BAV in more than one EBAV-affected family; red box indicates de novo variant.

Figure 3

Gene ontology enrichment analysis of EBAV candidate genes with filtered pDMI and pLOF variants Plot was generated using over-representation analysis of gene ontologies in WebGestalt 2019. The −log10 of false discovery rates (FDR) were plotted against the log2 of enrichment ratios.