Targeted myocardial gene expression in failing hearts by RNA sequencing.

BACKGROUND: Myocardial recovery with left ventricular assist device (LVAD) therapy is highly variable and difficult to predict. Next generation ribonucleic acid (RNA) sequencing is an innovative, rapid, and quantitative approach to gene expression profiling in small amounts of tissue. Our primary goal was to identify baseline transcriptional profiles in non-ischemic cardiomyopathies that predict myocardial recovery in response to LVAD therapy. We also sought to verify transcriptional differences between failing and non-failing human hearts.

METHODS: RNA was isolated from failing (n = 16) and non-failing (n = 8) human hearts. RNA from each patient was reverse transcribed and quantitatively sequenced on the personal genome machine (PGM) sequencer (Ion torrent) for 95 heart failure candidate genes. Coverage analysis as well as mapping the reads and alignment was done using the Ion Torrent Browser Suite™. Differential expression analyses were conducted by empirical analysis of digital gene expression data in R (edgeR) to identify differential expressed genes between failing and non-failing groups, and between responder and non-responder groups respectively. Targeted cardiac gene messenger RNA (mRNA) expression was analyzed in proportion to the total number of reads. Gene expression profiles from the PGM sequencer were validated by performing RNA sequencing (RNAseq) with the Illumina Hiseq2500 sequencing system.

RESULTS: The failing sample population was 75% male with an average age of 50 and a left ventricular ejection fraction (LVEF) of 16%. Myosin light chain kinase (MYLK) and interleukin (IL)-6 genes expression were significantly higher in LVAD responders compared to non-responders. Thirty-six cardiac genes were expressed differentially between failing and non-failing hearts (23 decreased, 13 elevated). MYLK, Beta-1 adrenergic receptor (ADRB1) and myosin heavy chain (MYH)-6 expression were among those significantly decreased in failing hearts compared to non-failing hearts. Natriuretic peptide B (NPPB) and IL-6 were significantly elevated. Targeted gene expression profiles obtained from the Ion torrent PGM sequencer were consistent with those obtained from Illumina HiSeq2500 sequencing system.

CONCLUSIONS: Heart failure is associated with a network of transcriptional changes involving contractile proteins, metabolism, adrenergic receptors, protein phosphorylation, and signaling factors. Myocardial MYLK and IL-6 expression are positively correlated with ejection fraction (EF) response to LVAD placement. Targeted RNA sequencing of myocardial gene expression can be utilized to predict responders to LVAD therapy and to better characterize transcriptional changes in human heart failure.