Analysis of Transcriptome Complexity via RNA-Seq in Normal and Failing Murine Hearts

Regulation of gene expression has a critical role in normal cardiac function and pathogenesis of heart failure. A global change in cardiac transcriptome from normal to one with characteristics of “fetal-like” profile is a major part of the pathological remodeling in failing hearts1-3. Although much insight has been learnt from transcriptome profiling studies using microarray technologies, limitations in coverage and sensitivity still leave a significant part of the cardiac transcriptome landscape un-explored, especially concerning expression and variation at single exon resolution. Recent advances in high-throughput sequencing technologies are enabling a new way to study transcriptomes: massively parallel sequencing of short reads derived from mRNAs (RNA-Seq)4, 5. Compared to microarray technologies, RNA-Seq was shown to enable more accurate quantification of gene expression levels6, 7. More importantly, RNA-Seq does not require a priori annotation of gene and transcript structures. It allows not only in-depth studies of expression changes in known genes and alternative isoforms, but also unbiased characterization of novel exons and novel transcript clusters. It also enables investigation of long non-coding RNA (lncRNA) genes, which are not usually targeted by alternative transcriptome profiling methods, such as microarrays. Thus, RNA-Seq opens the way to de novo transcriptome reconstruction and discovery of novel transcripts of any mammalian cell. Indeed, recent reports using RNA-Seq to profile transcriptome in mouse heart have revealed interesting new insights in cardiac transcriptional and signaling networks in genetic models of heart failure7-10. In this study, we developed bioinformatic methods to identify transcript structures and analyze transcriptome complexities with a particular emphasis on quantification of RNA splicing variants at single exon resolution using RNA-Seq data of normal and failing murine hearts. The methods take full advantage of the strength of RNA-Seq. We show that they allowed in-depth profiling and quantification of alternative mRNA structures, novel exons, novel transcript clusters (NTCs) and long non-coding RNA genes. The results open ways to direct experimental investigation of these novel transcriptome features and highlight the power of RNA-Seq to provide a comprehensive bioinformatic delineation of disease-specific transcriptomes.