Single-cell transcriptomic and chromatin accessibility analyses of dairy cattle peripheral blood mononuclear cells and their responses to lipopolysaccharide

Abstract

Background: Gram-negative bacteria are important pathogens in cattle, causing severe infectious diseases, including mastitis. Lipopolysaccharides (LPS) are components of the outer membrane of Gram-negative bacteria and crucial mediators of chronic inflammation in cattle. LPS modulations of bovine immune responses have been studied before. However, the single-cell transcriptomic and chromatin accessibility analyses of bovine peripheral blood mononuclear cells (PBMCs) and their responses to LPS stimulation were never reported.

Results: We performed single-cell RNA sequencing (scRNA-seq) and single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq) in bovine PBMCs before and after LPS treatment and demonstrated that seven major cell types, which included CD4 T cells, CD8 T cells, and B cells, monocytes, natural killer cells, innate lymphoid cells, and dendritic cells. Bioinformatic analyses indicated that LPS could increase PBMC cell cycle progression, cellular differentiation, and chromatin accessibility. Gene analyses further showed significant changes in differential expression, transcription factor binding site, gene ontology, and regulatory interactions during the PBMC responses to LPS. Consistent with the findings of previous studies, LPS induced activation of monocytes and dendritic cells, likely through their upregulated TLR4 receptor. NF-κB was observed to be activated by LPS and an increased transcription of an array of pro-inflammatory cytokines, in agreement that NF-κB is an LPS-responsive regulator of innate immune responses. In addition, by integrating LPS-induced differentially expressed genes (DEGs) with large-scale GWAS of 45 complex traits in Holstein, we detected trait-relevant cell types. We found that selected DEGs were significantly associated with immune-relevant health, milk production, and body conformation traits.

Conclusion: This study provided the first scRNAseq and scATAC-seq data for cattle PBMCs and their responses to the LPS stimulation to the best of our knowledge. These results should also serve as valuable resources for the future study of the bovine immune system and open the door for discoveries about immune cell roles in complex traits like mastitis at single-cell resolution.

Keywords: Cattle; Lipopolysaccharide; Peripheral blood mononuclear cell; Single-cell ATAC-seq; Single-cell RNA-seq.

Fig. 1

Cluster analysis of single-cell transcriptomes using four cattle PBMC samples. A UMAP projection plot showing seven major clusters of the 26,141 individual cell transcriptomes from all four PBMC samples. B The cell types were annotated using Azimuth (https://satijalab.org/azimuth/), based on their similarity to the human PBMC reference. C Plots and relative proportions of seven clusters/cell types across four PBMC samples, as annotated in B. The percentages in the table represent the relative proportions of cell types in four samples

Fig. 2

Cell-cycle, SCENIC, and Pseudotime analyses. A Cell-cycle analysis. Heatmap showing expression levels of cell-cycle-related genes in each cell type. Cells were ordered according to the average expression level of cell-cycle-related genes. The color key from white to red indicated expression levels from low to high. The cell-cycle index of each cell type is shown at the right. B SCENIC results. SCENIC binary regulon activity matrix showing all correlated regulons that were active in at least 1% of all regulons. Each column represents a single cell, and each row represents one regulon. The “regulon” refers to the regulatory network of TFs and their target genes. “On” indicates active regulons; “Off” indicates inactive regulons. Cluster labels correspond to those used in the UMAP plot. Representative transcription factors are highlighted. All cells (C) or individual cell type (D) pseudotime analysis using Monocle 2 for cell transcriptomes. Solid black lines indicate the main diameter path of the minimum spanning tree (MST) and provide the backbone of Monocle’s pseudotime ordering of the cells

Fig. 3

Co-expression analyses. A Dendrogram showing the gene co-expression network constructed using WGCNA. The color bar labeled as “Module colors” beneath the dendrogram represents the module assignment of each gene. B The relationship between modules and cell type. The upper numbers within each grid are the correlation between each module and cell type. The numbers in brackets represent the p values. C Selected significantly enriched GO terms based on genes within each module

Fig. 4

Specific gene expression responses of innate immunity induced by lipopolysaccharide in cattle PBMC. Gene expressions of CXCL2 (A), IRF9 (B), and CCL2 (C) in seven cell types, four PBMC samples of different treatment time points, or across their combinations. On their right, the changes of chromatin accessibility peak profiles near these three gene promoters over the treatment time course were derived from scATAC-seq. D Heatmap showing scaled expression levels of three gene modules (core antiviral, peaked inflammatory, and sustained inflammatory) in monocytes

Fig. 5

Associations of cell clusters with complex traits based on GWAS signal enrichment analyses using DEGs/marker genes among cell types (A) and among cattle PBMC LPS-treatment samples (top 5%) (B). “*” denotes FDR < 0.05