Analyzing the Muscle Transcriptome and the Splicing Variants of Pigs with Divergent Lipid Phenotypes Through RNA-Seq

We have used RNA-Seq to examine the gluteus medius (GM) muscle transcriptome of commercial Duroc pigs with divergent phenotypes (HIGH and LOW) for 13 fatness traits. The whole transcriptome of 52 GM samples from HIGH (n=26) and LOW (n=26) pigs were sequenced with a HiSeq2000 platform. An average of 70-million sequence reads were obtained from each sample and mapped to the pig reference genome. In all samples, 80% of reads were categorized as mapped. Differential expression analysis was performed at the gene and transcript (splicing variants) levels by comparing the LOW and HIGH groups. Results detected 96 genes and 30 splicing variants differentially expressed between pigs with divergent fatness traits (p<0.01 and Fold-change>1.5). Pathway analysis evidenced that most of differentially expressed genes belong to the glycolysis and gluconeogenesis, oxidative phosphorylation, insulin signaling and fatty acid biosynthesis metabolic routes. In order to detect genes acting as key regulators, an in-depth analysis was performed focusing on the differentially expressed splicing variants. This approach allowed us to identify several splicing variants corresponding to genes associated to lipid metabolism (i.e. HP, RXRG, SCD and TIMP1). Stearoyl-CoA desaturase (SCD) and Tissue Inhibitor Metallopeptidase1 (TIMP1) genes present one out of four splicing variants that appear to be differentially expressed. Also, the Haptoglobin (HP) gene shows differential expression between HIGH and LOW pigs in two out of seven splicing variants (p<0.01 and Fold-change>2). Besides, SNP discovery and eQTL analysis allowed us to identify several co-localization signals of muscle cis/trans-eQTL and genomic regions displaying associations with fatness traits.