Integrating Single-Cell RNA-Seq and ATAC-Seq Analysis Reveals Uterine Cell Heterogeneity and Regulatory Networks Linked to Pimpled Eggs in Chickens

Abstract

Pimpled eggs have defective shells, which severely impacts hatching rates and transportation safety. In this study, we constructed single-cell resolution transcriptomic and chromatin accessibility maps from uterine tissues of chickens using single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC-seq). We identified 11 major cell types and characterized their marker genes, along with specific transcription factors (TFs) that determine cell fate. CellChat analysis showed that fibroblasts had the most extensive intercellular communication network and that the chickens laying pimpled eggs had amplified immune-related signaling pathways. Differential expression and enrichment analyses indicated that inflammation in pimpled egg-laying chickens may lead to disruptions in their circadian rhythm and changes in the expression of ion transport-related genes, which negatively impacts eggshell quality. We then integrated TF analysis to construct a regulatory network involving TF-target gene-Gene Ontology associations related to pimpled eggs. We found that the transcription factors ATF3, ATF4, JUN, and FOS regulate uterine activities upstream, while the downregulation of ion pumps and genes associated with metal ion binding directly promotes the formation of pimpled eggs. Finally, by integrating the results of scRNA-seq and scATAC-seq, we identified a rare cell type-ionocytes. Our study constructed single-cell resolution transcriptomic and chromatin accessibility maps of chicken uterine tissue and explored the molecular regulatory mechanisms underlying pimpled egg formation. Our findings provide deeper insights into the structure and function of the chicken uterus, as well as the molecular mechanisms of eggshell formation.

Keywords: chicken; pimpled egg; scATAC-seq; scRNA-seq; uterus.

Figure 1

Single-cell transcriptome analysis and clustering identification of chicken uterine cells. (A) Unsupervised clustering revealed 21 distinct transcriptional cell clusters, which were visualized using a UMAP plot. Each point represents an individual cell, with colors indicating cluster assignments. (B) A UMAP plot was utilized to visualize 11 uterine cell types. Each point represents an individual cell and is color-coded according to its cell type. (C) The dot plot illustrates the distinct expression patterns of canonical marker genes across 11 cell populations. (D) Bar plot showing the percentage of different cell types within each of the 6 samples.

Figure 2

Regulatory network of DE transcription factors (TFs) and their DE target genes in chicken uterine tissue, along with GO terms. In the diagram, inverted triangles represent transcription factors (TFs), circles denote target genes, and rectangles indicate GO terms. Red highlights TFs and target genes that are upregulated in the NE group, while green highlights those upregulated in the PE group. Blue represents GO terms. The size of the shapes corresponds to the degree of involvement, with larger shapes indicating greater participation in regulatory networks.

Figure 3

scRNA-seq reveals significant changes in cell communication between the NE and PE groups. (A) Number of interactions between all cells in the NE group. Thicker links indicate a higher number of interactions. (B) Number of interactions between all cells in the PE group. Thicker links indicate a higher number of interactions. (C) Cell–cell communication chord diagram for the COLLAGEN signaling pathway in the NE group, with thicker links indicating stronger interactions between cells. (D) Cell–cell communication chord diagram for the COLLAGEN signaling pathway in the PE group, with thicker links indicating stronger interactions between cells. (E) Relative signaling pathway diagram showing the pathways identified in the NE and PE groups. Larger pathways in the NE group are depicted in cyan, while those in the PE group are depicted in red. Black indicates pathways with no significant difference.

Figure 4

Single-Cell Chromatin Accessibility Analysis of Chicken Uterine Tissue. (A) Annotation and statistics of the distribution of peaks in different genomic functional regions (such as promoters, 5′UTR, 3′UTR, exons, introns, downstream regions, and intergenic regions) for the N1 sample (top) and P1 sample (bottom). (B) UMAP plot of the scATAC-seq dataset for sample P1, with cell type assignments based on scRNA-seq data. (C) Histogram of cell frequency distributions for scRNA and scATAC data in the P1 sample. (D) Feature plot of inferred marker gene activities, including epithelial cells (KRT7), ionocytes (PDE1C), luminal epithelial cells (ATP8A2), macrophages (ENSGALG00010005330), T cells (ENSGALG00010005352), fibroblasts (COL5A1), B cells (ENSGALG00010003777), and endothelial cells (APOLD1).

Figure 5

Identification of Ionocyte through marker genes and enrichment analysis. (A) UMAP plot showing the expression patterns of ionocyte marker genes (from the literature) in scRNA-seq data. (B) Enriched GO terms for genes upregulated in Ionocytes, as identified in scRNA-seq data. (C) UMAP visualization of PDE1C, a marker gene for ionocytes identified in the literature, showing inferred gene activity within ionocyte clusters from scATAC-seq data.