Time-series transcriptome analysis identified differentially expressed genes in broiler chicken infected with mixed Eimeria species

Abstract

Coccidiosis caused by the Eimeria species is a highly problematic disease in the chicken industry. Here, we used RNA sequencing to observe the time-dependent host responses of Eimeria-infected chickens to examine the genes and biological functions associated with immunity to the parasite. Transcriptome analysis was performed at three time points: 4, 7, and 21 days post-infection (dpi). Based on the changes in gene expression patterns, we defined three groups of genes that showed differential expression. This enabled us to capture evidence of endoplasmic reticulum stress at the initial stage of Eimeria infection. Furthermore, we found that innate immune responses against the parasite were activated at the first exposure; they then showed gradual normalization. Although the cytokine-cytokine receptor interaction pathway was significantly operative at 4 dpi, its downregulation led to an anti-inflammatory effect. Additionally, the construction of gene co-expression networks enabled identification of immunoregulation hub genes and critical pattern recognition receptors after Eimeria infection. Our results provide a detailed understanding of the host-pathogen interaction between chicken and Eimeria. The clusters of genes defined in this study can be utilized to improve chickens for coccidiosis control.

Keywords: Eimeria; anti-inflammation; chicken; gene co-expression network; host-pathogen interaction; innate immunity; transcriptome analysis.

FIGURE 1

Overview of the experimental design. NC, Negative control; PC, Positive control; dpi, days post-infection.

FIGURE 2

Volcano plots and Venn diagram showing significant DEGs obtained via time-series comparison between PC and NC treatments. (A) Volcano plots showing different time points, (B) Venn diagram indicating numbers of DEGs.

FIGURE 3

Clustered DEGs using two different grouping methods (A) Three types of clusters separated based on their gene expression patterns. In total, 436, 414, and 149 genes were assigned to Types 1, 2, and 3, respectively. (B) GCNs clustered using gene co-expression values. Each node is a gene colored with its expression type. Node size reflects absolute log₂FC. Edges between nodes represent significant co-expression of genes. Network names reflect the number of genes they contain.

FIGURE 4

KEGG and GO analysis results of Type 1, 2, and 3 gene clusters. (A–C) Enriched KEGG pathways of Type 1, 2, and 3 genes clustered using the k-means clustering method. (D–E) Enriched GO terms of the biological process category of Types 1 and 3 gene clusters. Representative GO terms were selected from 53 to 188 significant terms of Types 1 and 3, respectively. GO analysis of Type 2 genes did not show any significant terms. (A–E) All enrichment terms passed a significance cut-off (p < 0.05) and were sorted in the order of largest significance levels.

FIGURE 5

KEGG and GO analysis results of three GCNs. (A–C) Enriched KEGG pathways of the GCN278, GCN53, and GCN40 networks. (D–F) Enriched GO terms of the GCN278, GCN53, and GCN40 networks. GO terms were categorized as a biological process and molecular functions (colored blue and purple, respectively). Representative GO terms were selected from 26, 42, and 70 significant biological process terms and 12, 15, and 6 significant molecular functions terms for GCN278, GCN53, and GCN40, respectively. (A–F) All enrichment terms passed a significance cut-off (p < 0.05) and were sorted in the order of largest significance levels.