Abstract 1638 Bioinformatic Prioritization of Cardiovascular Disease-Linked Variants Modulating TBX5 Binding

Cardiovascular diseases (CVDs) are the leading cause of death worldwide, responsible for approximately 17.9 million lives each year. Genome-wide association studies (GWAS) have linked multiple types of CVDs with mutations, specifically within non-coding regions of the genome (e.g., promoters and enhancers). CVD-associated mutations, such as single nucleotide polymorphisms (SNPs), can alter protein-DNA interactions of regulatory proteins, like the cardiac transcription factor (TF) TBX5. TBX5 plays a crucial role in regulating cardiac genes needed for proper heart development. Previous research has associated TBX mutants with detrimental cardiac defects and congenital diseases. However, multiple CVD-associated SNPs occur within cardiac TF binding sites, altering TF-DNA interactions needed to regulate gene expression. Due to the increasing number of CVD-associated mutations, prioritizing for biochemical and genetic analysis remains challenging. To address this, we develop a bioinformatic pipeline to prioritize potential CVD-causal variants modulating TBX5 binding that are in expression quantitative trait loci (eQTL) in cardiac tissue. Our pipeline uses TBX5-DNA binding profiles (ChIP-seq data) from human cardiomyocytes to predict TBX5 binding sites. Through electrophoretic mobility shift assay (EMSA), we were able to validate our computational predictions on DNA sequences of different TF-DNA binding affinities (different scores). For in vitro validation, we expressed the TBX5 DNA-binding domain (T-box domain) and purified it through Ni-NTA affinity chromatography. The in vitro TBX5-DNA binding affinities agreed with the computational predictions and in proportion to the binding scores. We then implemented the model to predict the impact of CVD-associated SNPs on TBX5-DNA binding. We identified 1,535 CVD-associated SNPs that occur within TF footprints and putative enhancers in the human heart and 14,218 variants in linkage disequilibrium (LD) with eQTL in cardiac tissue. To test the model, we chose two variants (rs7612445 and rs7790964) predicted to alter TBX5-DNA binding and with differential gene expression profiles in cardiac tissue. Through EMSA, we observed that rs7612445 alternate allele increased TBX5-DNA binding, whereas rs7790964 decrease TBX5-DNA binding. These changes in TBX5-DNA binding affinities have been associated with increased gene expression of GNB4 (rs7612445) and decreased expression of TBX20 (rs7790964) in the heart atrial appendage, both genes crucial for cardiovascular traits, such as heart rate stabilization and heart development, respectively. In summary, we validated the computational predictions through a biochemical approach, supporting the applicability of this computational approach to prioritize potential CVD-causal mutations.