Evaluating the Impact of Non‐coding Variants in the Cardiac Transcription Factor NKX2‐5 Binding Sites

Abstract

Transcription Factors (TFs) are DNA‐specific binding proteins that regulate gene expression and other biological processes like cell differentiation. For example, the cardiac TF NKX2‐5 has been proven to be a master regulator in the transcriptional network needed for heart development. Mutations in NKX2‐5 have been shown to cause congenital heart diseases (CHDs), the most common birth defect that result in structural abnormalities in the heart. However, ~95% of disease‐associated mutations occur within the non‐coding genome, such as promoters and enhancers. The molecular mechanism of these genetic variations and their role in many diseases has remained unexplored. We hypothesize that non‐coding variants within NKX2‐5 binding sites will affect DNA recognition and alter TF‐DNA binding. First, using ~84 million Single Nucleotide Polymorphisms (SNPs) reported in the 1000 Genomes Project, I identified 8,475 that are predicted to affect NKX2‐5 binding. After filtering these variants, I found 901 SNPs localized in cardiac enhancers, and 30 disease‐associated SNPs from the GWAS catalog predicted to affect NKX2‐5 binding sites. Using position weight matrix (PWM) DNA‐specificity models, binding scores for the identified mutations were calculated. The variants rs7350789 and rs7719885 were predicted to have the greatest impact in NKX2‐5 binding affinity with ΔPWM scores of 258 and ‐212 respectively. These variants were prioritized for in vitro validation through Electrophoretic Mobility Shift Assay (EMSA) using purified full‐length NKX2‐5 to evaluate TF‐DNA complex formation. After performing an EMSA, both variants resulted in an increase in TF‐DNA binding even though rs7719885 had a negative ΔPWM score which predicts a decrease in binding affinity. Although not as expected, two of the variants predicted to impact TF‐DNA binding were successfully validated.