Transcriptome and Weighted Gene Co-Expression Network Analysis for Feather Follicle Density in a Chinese Indigenous Breed

Abstract

Feather follicle density plays an important role in appealing to consumers' first impressions when making purchasing decisions. However, the molecular network that contributes to this trait remains largely unknown. The aim of this study was to perform transcriptome and weighted gene co-expression network analyses to determine the candidate genes relating to feather follicle density in Wannan male chickens. In total, five hundred one-day-old Wannan male chickens were kept in a conventional cage system. Feather follicle density was recorded for each bird at 12 weeks of age. At 12 weeks, fifteen skin tissue samples were selected for weighted gene co-expression network analysis, of which six skin tissue samples (three birds in the H group and three birds in the L group) were selected for transcriptome analysis. The results showed that, in total, 95 DEGs were identified, and 56 genes were upregulated and 39 genes were downregulated in the high-feather-follicle-density group when compared with the low-feather-follicle-density group. Thirteen co-expression gene modules were identified. The red module was highly significantly negatively correlated with feather follicle density (p < 0.01), with a significant negative correlation coefficient of -0.72. In total, 103 hub genes from the red module were screened. Upon comparing the 103 hub genes with differentially expressed genes (DEGs), it was observed that 13 genes were common to both sets, including MELK, GTSE1, CDK1, HMMR, and CENPE. From the red module, FOXM1, GTSE1, MELK, CDK1, ECT2, and NEK2 were selected as the most important genes. These genes were enriched in the DNA binding pathway, the heterocyclic compound binding pathway, the cell cycle pathway, and the oocyte meiosis pathway. This study suggests that FOXM1, GTSE1, MELK, CDK1, ECT2, and NEK2 may be involved in regulating the development of feather follicle density in Wannan male chickens. The results of this study reveal the genetic structure and molecular regulatory network of feather follicle density in Wannan male chickens, and provide a basis for further elucidating the genetic regulatory mechanism and identifying molecular markers with breeding value.

Keywords: Wannan male chickens; feather follicle density; mRNA expression profiles; weighted gene co-expression network analysis.

Figure 1

Identification of DEGs in the H and L groups. (a) Volcano plots of DEGs in the H and L groups; red denotes significantly downregulated and blue denotes significantly upregulated genes in the H group when compared with the L group. (b) Heatmap of DEGs in the H and L groups. H, high feather follicle density; L, low feather follicle density. (c) Enriched KEGG and GO pathways for differentially expressed genes (DEGs) identified between H and L groups. H, high feather follicle density; L, low feather follicle density.

Figure 2

WGCNA of gene expression profile with feather follicle density in Wannan male chicken skin tissue at 12 weeks. (a) Heat map of sample clustering and traits; J, feather follicle density; The color red indicates the correlation between phenotypic data and gene expression. The darker the shade of red, the stronger the correlation. (b) Scale-free topology model fit, and gene mean connectivity under different soft threshold powers.

Figure 3

WGCNA of gene expression profiles with feather follicle density in Wannan male chicken skin tissue at 12 weeks. (a) Analysis of the correlation between different module genes and phenotypes; (b) GS and MM analysis for density; (c) enriched KEGG and GO pathways for hub genes in the red modules; (d) interaction network of the hub genes in the red modules. J, feather follicle density.

Figure 4

(a) Comparisons between qRT-PCR and RNA-seq measurements of the expression abundance of 10 random differentially expressed genes; (b) correlation analysis of the values between RNA-seq and qPCR results.