RNA-Seq based selection signature analysis for identifying genomic footprints associated with the fat-tail phenotype in sheep

Abstract

Understanding the genetic background behind fat-tail development in sheep can be useful to develop breeding programs for genetic improvement, while the genetic basis of fat-tail formation is still not well understood. Here, to identify genomic regions influencing fat-tail size in sheep, a comprehensive selection signature identification analysis was performed through comparison of fat- and thin-tailed sheep breeds. Furthermore, to gain the first insights into the potential use of RNA-Seq for selection signature identification analysis, SNP calling was performed using RNA-Seq datasets. In total, 45 RNA-Seq samples from seven cohort studies were analyzed, and the F_ST method was used to detect selection signatures. Our findings indicated that RNA-Seq could be of potential utility for selection signature identification analysis. In total, 877 SNPs related to 103 genes were found to be under selection in 92 genomic regions. Functional annotation analysis reinforced the hypothesis that genes involved in fatty acid oxidation May modulate fat accumulation in the tail of sheep and highlighted the potential regulatory role of angiogenesis process in the fat deposition. In agreement with most previous studies, our results re-emphasize that the BMP2 gene is targeted by selection during sheep evolution. Further gene annotation analysis of the regions targeted by the sheep evolution process revealed that a large number of genes included in these regions are directly associated with fat metabolism, including those previously reported as candidates involved in sheep fat-tail morphology, such as NID2, IKBKG, RGMA, IGFBP7, UBR5, VEGFD and WLS. Moreover, a number of genes, including BDH2, ECHS1, AUH, ERBIN and CYP4V2 were of particular interest because they are well-known fat metabolism-associated genes and are considered novel candidates involved in fat-tail size. Consistent with the selection signature identification analysis, principal component analysis clustered the samples into two completely separate groups according to fat- and thin-tailed breeds. Our results provide novel insights into the genomic basis of phenotypic diversity related to the fat-tail of sheep breeds and can be used to determine directions for improving breeding strategies in the future.

Keywords: RNA-Seq datasets; SNP calling; fat deposition; selection signatures; thin- and fat-tailed sheep.

Figure 1

The pipeline used to perform SNP calling from RNA-Seq datasets and selection signature identification analysis.

Figure 2

Circos plot indicating the number of clean reads (outer ring), mapped reads (middle ring), and identified SNPs (inner ring) per sample of the seven studies. “Stu” represents the number of studies (Stu1, Study 1), “Fat” and “Thin” represent fat- and thin-tailed breeds, respectively, and “S” represents the number of samples per study. There are 45 samples in total.

Figure 3

Density plot of the identified SNPs after imputation across chromosomes. Different colors represent the quantity of the SNPs based on the legend.

Figure 4

Principal component analysis (PCA) based on all the identified SNPs in thin- and fat-tailed breeds. Individuals are colored based on their breed, and the red and purple colors represent the fat- and thin-tailed sheep breeds, respectively. “Stu” represents the number of studies (Stu1, Study 1), “Fat” and “Thin” represent fat- and thin-tailed breeds, respectively, and “S” represents the number of samples per study.

Figure 5

Manhattan plot of the genome-wide selection signature distribution of windowed FST in fat- and thin-tailed sheep breeds. The outermost ring shows chromosome numbers, and the middle red ring marks the 0.1% percentile threshold for F_ST > 0.15.

Figure 6

Circos plot related to genome-wide distribution of the putative selection signatures. The outermost ring shows the chromosome numbers. SNPs were divided into two groups based on their heterozygosity levels, as the SNPs that were more heterozygous in thin-tailed sheep breeds are shown in the first (red points) inner gray ring, while the more heterozygous SNPs in fat-tailed sheep breeds are presented in the second (blue points) inner gray ring. The first layer of the heatmap displays the FST values, which ranged from 0.15 (purple) to 0.28 (yellow). The second layer of the heatmap displays the Theta values, which ranged from 0.23 (purple) to 0.42 (yellow). Some of the most important genes associated with fat metabolism are displayed.

Figure 7

Principal component analysis (PCA) based on the SNPs located in the genomic regions identified to be under positive selection in thin- and fat-tailed breeds. Individuals are colored based on their group (fat- or thin-tailed), and the red and purple colors represent the fat- and thin-tailed sheep breeds, respectively. “Stu” represents the number of studies (Stu1, Study 1), “Fat” and “Thin” represent fat- and thin-tailed breeds, respectively, and “S” represents the number of samples per study.

Figure 8

Sankey plot of the functional enrichment analysis related to the genes identified in the selection signature regions. Count represents the number of genes associated with the terms.