Comparative analysis of the microRNA transcriptome between yak and cattle provides insight into high-altitude adaptation

Abstract

Extensive and in-depth investigations of high-altitude adaptation have been carried out at the level of morphology, anatomy, physiology and genomics, but few investigations focused on the roles of microRNA (miRNA) in high-altitude adaptation. We examined the differences in the miRNA transcriptomes of two representative hypoxia-sensitive tissues (heart and lung) between yak and cattle, two closely related species that live in high and low altitudes, respectively. In this study, we identified a total of 808 mature miRNAs, which corresponded to 715 pre-miRNAs in the two species. The further analysis revealed that both tissues showed relatively high correlation coefficient between yak and cattle, but a greater differentiation was present in lung than heart between the two species. In addition, miRNAs with significantly differentiated patterns of expression in two tissues exhibited co-operation effect in high altitude adaptation based on miRNA family and cluster. Functional analysis revealed that differentially expressed miRNAs were enriched in hypoxia-related pathways, such as the HIF-1α signaling pathway, the insulin signaling pathway, the PI3K-Akt signaling pathway, nucleotide excision repair, cell cycle, apoptosis and fatty acid metabolism, which indicated the important roles of miRNAs in high altitude adaptation. These results suggested the diverse degrees of miRNA transcriptome variation in different tissues between yak and cattle, and suggested extensive roles of miRNAs in high altitude adaptation.

Keywords: Cattle; High altitude adaptation; Yak; microRNA.

Figure 1. The overview of small RNA sequencing.

(A) Mature miRNAs and miRNA precursors identified in this study. (B) Venn charts indicate expression patterns of mature miRNAs among each sample.

Figure 2. Highly conserved heart and lung tissues showed varying degrees of divergence.

(A) Hierarchical clustering and heat map matrix of pairwise Spearman’s correlations of the counts of 718 co-expressed miRNAs between 11 miRNA libraries. (B) Principal component analysis (PCA) plot of each sample. The fraction of the variance explained is 78% for eigenvector 1 and 10% for eigenvector 2. (C) Top 10 unique miRNAs with the highest expression levels in heart and lung tissues of yak and cattle.

Figure 3. Expression pattern of differentially expressed miRNAs belong to the miRNA family in sequence and miRNA cluster in location.

(A–B) Volcano plot showing differentially expressed miRNAs in heart (A) and lung (B) between yak and cattle. (C) Box plot of Pearson’s correlation coefficient for miRNA families, clusters and background. (D) Expression pattern of miR-200 family belonging to miR-200 cluster. (E) Expression pattern of miR-34 family belonging to miR-34 cluster and miR-449 cluster.

Figure 4. Functional analysis of DE miRNAs in heart and lung.

(A-B) Gene Ontology (GO) categories and pathways enriched for target genes of DE miRNAs in heart (A) and lung (B). The p value indicating the significance of the enrichment, was corrected by Fisher’s exact test. miRNA number was the number of DE miRNAs involved in each GO term. Gene number was the number of target genes for DE miRNAs involved in each GO term. (C) DE miRNAs involved in p53 feedback circuit. (D) DE miRNAs and corresponding target genes are involved in HIF-1 pathway. miRNAs in red indicated upregulated in yak, miRNAs in black indicated downregulated in yak.