Integrated transcriptomic and regulatory RNA profiling reflects complex pathophysiology and uncovers a conserved gene...

Heart failure (HF) is a complex syndrome. Despite availability of multiple treatment options, the mortality remains high and the quality of life poor. Better understanding of the underlying pathophysiological processes can lead to development of novel therapies. Multiple comparative transcriptomics studies, which revealed gene level changes in the key pathophysiological pathways in failing hearts, point towards heterogeneity from interplay of disease stage, etiologies and ethnicity. Transcriptomic characterization of HF in patients from different ethnicities can potentially help in understanding the heterogeneity imparted by various factors and the core elements in heart failure.

An integrated analysis of bulk transcriptome and microRNA sequencing from the cardiac tissues of 30 South Asian (SA) patients having HF with reduced ejection fraction (HFrEF) and 19 control subjects was conducted. Plasma miRNAs from a subset of HFrEF and control patients were also sequenced to understand their biomarker potential. The altered transcriptome from the myocardium of SA HFrEF patients reflected cardiac muscle contraction, cellular energetics, altered immune signaling and extracellular matrix remodelling as predominant pathophysiological mechanisms. The SA HFrEF patients also showed dysregulation of multiple microRNAs in cardiac tissue like miR-216, miR-217, miR-184 and miR-9983. Many of these miRNAs, such as miR184 and few others, showed altered levels in both the plasma and cardiac tissue of HFrEF patients suggesting their biomarker potential. The diversity in the HFrEF transcriptomes from published studies led us to examine the core HF genes in our cohort. A gene signature generated using machine learning (ML) from the top dysregulated genes in SA HFrEF cohort stratified HF from controls in other cohorts. The sensitivity of the HF gene signature was further improved when union of two cohorts was used as a training set. Our ML analyses developed a core HF gene signature consisting of 21 genes that can stratify HF patients from controls with 98 % sensitivity in all the tested cohorts.

This study reveals molecular changes underlying the pathophysiology as reflected by coding and regulatory non-coding components of transcriptome from South Asian patients and uncovers a conserved gene signature for HF.