The alternative splicing landscape of infarcted mouse heart identifies isoform level therapeutic targets

Abstract

Alternative splicing is an important process that contributes to highly diverse transcripts and protein products, which can affect the development of disease in various organisms. Cardiovascular disease (CVD) represents one of the greatest global threats to humans, particularly acute myocardial infarction (MI) and subsequent ischemic reperfusion (IR) injury, which involve complex transcriptomic changes in heart tissues associated with metabolic reshaping and immunological response. In this study, we used a newly developed ONT full-length transcriptomic approach and performed transcript-resolved differential expression profiling in murine models of MI and IR. We built an analytical pipeline to reliably identify and quantify alternative splicing products (isoforms), expanding on the currently available catalog of isoforms described in mice. The updated alternative splicing landscape included transcripts, genes, and pathways that were differentially regulated during IR and MI. Our study establishes a pipeline to profile highly diverse isoforms using state-of-the-art long-read sequencing, builds a landscape of alternative splicing in the mouse heart during MI and IR.

Fig. 1

Overview of our studies and sequencing results. (A) The overall framework of our pipeline can be roughly divided into four parts, in which the identification of novel transcripts can be further subdivided in (B), and the operations and software involved in these steps are described in (C). (D) Percentage of sequences identified to VPNs and SSPs (Primers found) and chimeric sequences recovered by the software (Rescue) as well as removed sequences (Unusable) in the Oxford nanopore data. (E) Percentage of novel genes versus known genes resulting from the identification of novel transcripts. (F) Number of extended reference genomes after fusion of novel transcripts with known transcripts in GENCODE.

Fig. 2

Identification of novel transcripts and extension of reference annotations. (A) Schematic representation of the differences between different types of transcripts. (B) Sequence length distribution of different types of transcripts. (C) Percentage of each transcript type in different sequence length intervals. (D) Percentage of protein-coding and non-coding transcripts in different types. (E) Percentage of transcripts with only one exon (Mono-Exon) and multiple-exon transcripts (Multi-Exon). (F) Distribution of the number of transcripts of different lengths. The number of all transcripts (G), novel transcripts (H), and protein-coding novel transcripts (I) detected in MI/IR and IR-Sham/MI-Sham. (J) Number of transcripts detected in different groups of samples (Mann-Whitney test, asterisks indicate p < 0.05). (K) The number of different categories of novel transcripts. (L) The number of different categories of novel protein-coding transcripts (Mann-Whitney test, asterisks indicate p < 0.05).

Fig. 3

Gene level expression differences and functional enrichment. (A) Principal Component Analysis (PCA) of samples based on top500 genes. (B) Distance matrix of samples. (C) The number of genes identified versus NGS reads data, it can be seen that IR/MI can identify the same number of genes with less number of reads relative to two Sham groups. (D) Genes up regulated by MI relative to MI-Sham and their GO enrichment to terms. (E) Volcano plot of differential genes in the MI relative to IR-Sham. (F) Results of GO enrichment analysis of differential genes in MI relative to MI-Sham. (G) Genes up regulated by IR relative to IR-Sham and their GO enrichment to terms. (H) Volcano plot of differential genes in the IR relative to IR-Sham. (I) Results of GO enrichment analysis of differential genes in IR relative to IR-Sham.

Fig. 4

Differential expression analysis at the transcript level with a focus on MI- and IR-specific transcripts. (A) MA plot of differentially expressed transcripts of MI relative to MI-Sham, triangles represent novel transcripts. (B) GO enrichment results for genes corresponding to differentially expressed transcripts in MI relative to MI-Sham. (C) MA plot of differentially expressed transcripts of IR relative to IR-Sham, triangles represent novel transcripts. (D) GO enrichment results for genes corresponding to differentially expressed transcripts in IR relative to IR-Sham. (E) MA plot of differentially expressed novel transcripts of MI relative to MI-Sham, triangles represent novel protein-coding transcripts. (F) Changes in the proportion of types of differentially expressed transcripts in protein-coding novel transcripts and novel transcripts. (G) MA plot of differentially expressed novel transcripts of IR relative to IR-Sham, triangles represent novel protein-coding transcripts. (H) Type distribution of differentially expressed protein-coding novel transcripts with fold change. (I) IR and MI group upregulation of differentially expressed novel protein-coding transcripts.