PROJECT SUMMARY
Congenital heart disease (CHD) is the most common human birth defect and occurs in ~0.8% of live
births. CHD has a known genetic etiology in approximately 45% of cases. Whole exome (WES) or genome
(WGS) sequencing of 9072 CHD probands by the Pediatric Cardiovascular Genetics Consortium (PCGC) has
identified a significant burden of damaging protein-coding variants in 269 genes (denoted CHD genes).
Although WES analyses of the full PCGC cohort (17,000 probands) is ongoing, the predicted functions of these
CHD genes indicate that CHD often arises from dysregulated expression of cardiac developmental genes.
Within the ~55% of CHD cases with unknown etiologies (denoted WES-negative), CHD genes harbor many
rare missense, splice site-associated, and noncoding variants with unknown pathogenicity. Here we test the
hypothesis that de novo and rare variants of unknown significance (VUSs) in non-coding and coding
regions are responsible for a subset of CHD that remains unexplained. Non-coding de novo VUS
(ncDNV) identified in proximity to genes that impact heart development will be studied in massively parallel
reporter assays to determine if they impact cis-regulatory element activity in cardiac lineages differentiated
from iPSCs. Candidates will be further studied by introduction into the endogenous locus and impact on gene
expression and cell transcriptional states will be assessed by single nucleus RNA sequencing (snRNAseq).
Concurrently we will study coding region VUS in two ways. Using CLINVAR assignment of pathogenicity and
AlphaMissense, a deep learning model built on protein structure that variant effects, we will prioritize rare
missense VUS for introduction into iPSCs and cardiac lineages. Using snRNAseq we will compare missense
VUS to reference and loss of function variants. Additionally, we will interrogate splice site-associated VUS
using minigene splice assays. To validate cell-based analyses we will introduce and study a selected subset of
non-coding and rare missense variants in mice. With functional data on non-coding and rare missense VUS,
we expect to define and characterize a subset that cause or contribute to CHD and improve predictive models
of pathogenicity in genes that cause CHD and other human disorder. Identifying genetic contributors in a
subset of unexplained CHD cases will improve clinical assessment of outcomes and recurrence risks and
enable genotype-phenotype analyses. These insights will also advance knowledge about the mechanisms for
normal cardiac development and CHD and identify patients that are appropriate for studies of non-genetic
etiologies. We propose to:
Aim 1. Assess the contribution of de novo noncoding VUS (ncDNV) to unexplained CHD.
Aim 2. Define the contribution of rare coding region VUS in CHD genes to unexplained CHD.