Cardiac ion channel expression in the equine model - In-silico prediction utilising RNA sequencing data from mixed tissue samples.

Understanding cardiomyocyte ion channel expression is crucial to understanding normal cardiac electrophysiology and underlying mechanisms of cardiac pathologies particularly arrhythmias. Hitherto, equine cardiac ion channel expression has rarely been investigated. Therefore, we aim to predict equine cardiac ion channel gene expression. Raw RNAseq data from normal horses from 9 datasets was retrieved from ArrayExpress and European Nucleotide Archive and reanalysed. The normalised (FPKM) read counts for a gene in a mix of tissue were hypothesised to be the average of the expected expression in each tissue weighted by the proportion of the tissue in the mix. The cardiac-specific expression was predicted by estimating the mean expression in each other tissues. To evaluate the performance of the model, predicted gene expression values were compared to the human cardiac gene expression. Cardiac-specific expression could be predicted for 91 ion channels including most expressed Na + channels, K + channels and Ca 2+ -handling proteins. These revealed interesting differences from what would be expected based on human studies. These differences included predominance of Na V 1.4 rather than Na V 1.5 channel, and RYR1, SERCA1 and CASQ1 rather than RYR2, SERCA2, CASQ2 Ca 2+ -handling proteins. Differences in channel expression not only implicate potentially different regulatory mechanisms but also pathological mechanisms of arrhythmogenesis.