Single-Cell Analysis of Host Responses in Bovine Milk Somatic Cells (bMSCs) Following HPAIV Bovine H5N1 Influenza Exposure

Abstract

The 2024 outbreak of highly pathogenic avian influenza virus (HPAIV) H5N1 in U.S. dairy cattle presented an unprecedented scenario where the virus infected bovine mammary glands and was detected in milk, raising serious concerns for public health and the dairy industry. Unlike previously described subclinical influenza A virus (IAV) infections in cattle, H5N1 infection induced severe clinical symptoms, including respiratory distress, mastitis, and abnormal milk production. To understand the host immune responses and changes, particularly in the mammary gland, we performed single-cell RNA sequencing analysis on bovine milk somatic cells (bMSCs) in vitro exposed to an H5N1 isolate from an infected dairy farm. We identified ten distinct cell clusters and observed a shift toward type-2 immune responses, characterized by T cells expressing IL13 and GATA3, and three different subtypes of epithelial cells based on the expression of genes associated with milk production. Our study revealed temporal dynamics in cytokine expression, with a rapid decline in luminal epithelial cells and an increase in macrophages and dendritic cells, suggesting a role in increased antigen presentation. While viral RNA was detected in bulk-exposed bMSC samples via qRT-PCR, no viral reads were observed in the scRNA-seq data, indicating that the immune responses captured may be due to exposure to viral components rather than productive infection. This research fills a critical gap in understanding the immune responses of bovine mammary glands to H5N1 exposure and highlights the need for further investigation into therapeutic strategies for managing such outbreaks.

Keywords: bovine H5N1; bovine milk somatic cells; single-cell RNA-seq.

Figure 1

Single-cell RNA-seq analysis of bovine milk somatic cells (bMSCs). (A) t-SNE embeddings of merged scRNA-seq profiles from in vitro-unexposed and bovine H5N1-exposed bMSCs collected from three healthy cattle. Distinct cell populations are represented by different colors and assigned labels based on expression of selected gene markers. (B) Heatmap showing z-scored average expression of curated cell marker genes that had a fold change >1 and p < 0.05 for at least one cluster. Genes are ordered by the cluster in which they have the highest expression. (C) Heatmap displaying z-scored average expression of top 10 genes in each cluster that had a fold change >1 expressed and p < 0.05 for at least one cluster. Genes are organized by cluster in which they have highest expression.

Figure 2

Comparison of cell populations and gene expression between control and bovine H5N1-exposed bMSCs. (A) t-SNE embeddings of bMSCs either exposed to bovine H5N1 or unexposed control cells. Distinct cell populations are represented by different colors and assigned labels based on expression of selected gene markers, same as in Figure 1. (B–D) Bar graphs showing percentage (x-axis) of innate immune (B), T (C), and epithelial (D) cells in total cells, corresponding to different cell types (y-axis) in control (blue) and bovine H5N1-exposed (brown) bMSCs. (E) Heatmap displaying z-scored average expression of top 50 upregulated genes after bovine-H5N1 infection in each cluster that had a fold change >1 expressed and p < 0.05 for at least one cluster. Genes are organized by cluster in which they have highest expression and analyzed for Gene Ontology (GO)-enriched processes.

Figure 3

Viral genome detection by RT-qPCR. The graph shows the results of IAV M gene-based RT-qPCR in MDCK cells (positive control) and bMSCs infected with bovine H5N1. The x-axis represents the mean Ct values, and y-axis represents the cell type.