Transcriptome Analysis of Adipose Tissue Indicates That the cAMP Signaling Pathway Affects the Feed Efficiency of Pigs

Abstract

Feed efficiency (FE) is one of the main factors that determine the production costs in the pig industry. In this study, RNA Sequencing (RNA-seq) was applied to identify genes and long intergenic non-coding RNAs (lincRNAs) that are differentially expressed (DE) in the adipose tissues of Yorkshire pigs with extremely high and low FE. In total, 147 annotated genes and 18 lincRNAs were identified as DE between high- and low-FE pigs. Seventeen DE lincRNAs were significantly correlated with 112 DE annotated genes at the transcriptional level. Gene ontology (GO) analysis revealed that DE genes were significantly associated with cyclic adenosine monophosphate (cAMP) metabolic process and Ca²⁺ binding. cAMP, a second messenger has an important role in lipolysis, and its expression is influenced by Ca²⁺ levels. In high-FE pigs, nine DE genes with Ca²⁺ binding function, were down-regulated, whereas S100G, which encodes calbindin D9K that serve as a Ca²⁺ bumper, was up-regulated. Furthermore, ATP2B2, ATP1A4, and VIPR2, which participate in the cAMP signaling pathway, were down-regulated in the upstream of lipolysis pathways. In high-FE pigs, the key genes involved in the lipid biosynthetic process (ELOVL7 and B4GALT6), fatty acid oxidation (ABCD2 and NR4A3), and lipid homeostasis (C1QTNF3 and ABCB4) were down-regulated. These results suggested that cAMP was involved in the regulation on FE of pigs by affecting lipid metabolism in adipose tissues.

Keywords: adipose; cAMP; feed efficiency; lipid metabolism; pig.

Figure 1

Annotation and alignment of RNA-seq read pairs from porcine adipose tissue with different genome version. (A) Distribution of average mapped read pairs of six samples on the S scrofa 10.2 (left) and S. scrofa 11.1 (right) genomes. In the pie charts, the percentages represent the mean of all six RNA-seq data. On average, over 66% unique read pairs were aligned to annotated genes. (B) Distribution of mapped read pairs in different samples. Samples that belong to the same genome group show similar distribution. The color of bar standing for the same meaning as pie charts. (C) Boxplots showing the expression patterns (scaled log2FPKM) of annotated genes and long intergenic non-coding RNAs (lincRNAs) in high Feed Efficiency (high-FE) (left) and low-FE (right) groups, respectively.

Figure 2

Differential expression analysis of annotated genes and lincRNAs between high- and low-FE pigs. (A) The distribution of fold changes in gene expression. Genes (lincRNAs) with absolute log2FoldChange > 1 are indicated in orange (red) and those with log2Foldchange < −1 are indicated in green (blue). (B) Quantitative-PCR (qPCR) analysis results of six selected DE genes and lincRNAs. Left: Scatter diagram showing the log2FC correlation of RNA-seq and qPCR. Right: Relative expression of selected DE genes and lincRNAs. Statistically significant differences between high- and low-FE pigs are indicated by * (p value < 0.05).

Figure 3

Correlation analysis of DE genes and lincRNAs in the adipose tissues of high- and low-FE pigs. Weighted correlation network analysis (WGCNA) was applied to identify modules of highly correlated factors, including genes and lincRNAs. (A) A total of 25 modules were identified on the basis of expression patterns, which are represented by the dendrogram and correlation heat map. (B) Correlations between differentially expressed (DE) genes and lincRNAs were identified through WGCNA. Dark color indicates higher correlation. Vertical and horizontal axes in the heat map represent lincRNAs and genes, respectively. (C) Correlations between DE genes and lincRNAs were estimated on the basis of the Pearson correlation coefficient. Lattices in red are highly positive, and those in blue are highly negative. (D) Venn diagram depicting the proportion of correlated lincRNA-gene pairs detected on the basis of WGCNA and Pearson correlation. (E) Bar plot showing the Pearson correlation coefficient of overlapping lincRNA-gene pairs. Red and green bar plots represent positively and negatively correlated pairs, respectively.

Figure 4

Results of functional enrichment analysis of DE genes between high- and low-FE pigs in genes ontology (GO) molecular function, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, and GO biological process.

Figure 5

Potential pathways of the annotated DE genes and lincRNAs in the adipose tissues of high- and low-FE pigs. Red and pink colors indicate up-regulation in high-FE pigs (red, log2FC > 1; pink, 0 < log2FC < 1), while neon green and light green indicate down-regulation in high-FE pigs (neon green, log2FC < −1; light green, −1 < log2FC < 0).