In silico single-cell RNA-seq analysis reveals distinct host response programs in bovine milk somatic cells during H5N1-associated and bacterial mastitis

Abstract

ABSTRACT In silico single-cell RNA-seq analysis reveals distinct host response programs in bovine milk somatic cells during H5N1-associated and bacterial mastitis Highly pathogenic avian influenza (HPAI) H5N1 has recently been recognized as a cause of severe mastitis in dairy cattle, with viral replication in the mammary gland and shedding into milk. This emergence raises concern for animal health, dairy production and potential zoonotic transmission. In contrast, bovine mastitis is usually bacterial and has been studied mainly with bulk-level approaches that obscure cell-type–specific host responses. In this thesis, publicly available single-cell RNA sequencing datasets from H5N1-exposed, bacterial mastitis and healthy control bovine milk somatic cells were re-analyzed using an in-silico bioinformatics pipeline to compare viral and bacterial mastitis at single-cell resolution. After read alignment and rigorous quality control, an integrated bovine milk cell atlas was generated and major immune and non-immune cell populations were annotated, including neutrophils, monocyte/macrophage subsets, dendritic cells, lymphocytes and mammary epithelial cells. Condition-level and cell-type–level differential expression and functional enrichment analyses identified distinct pathogen-specific transcriptional programs. H5N1 mastitis was characterized by strong induction of interferon-stimulated genes, antiviral restriction factors, and chemokine modules across immune compartments, whereas bacterial mastitis was dominated by Toll-like receptor and NF-κB signalling, inflammasome-associated genes, and inflammatory pathways. These signatures were reflected in altered cellular composition, with focused expansion of interferon-high myeloid subsets in H5N1 and broader recruitment of neutrophil and macrophage populations in bacterial mastitis. By intersecting condition-specific gene sets with lineage markers, two concise 12-gene biomarker panels were derived: a viral panel composed mainly of interferon-stimulated antiviral genes and a bacterial panel centred on innate immune and inflammasome regulators. When projected onto neutrophil trajectories using diffusion pseudotime, these panels showed distinct activation profiles consistent with pathogen class. Through this analysis, understanding of pathogen-specific mastitis immunopathology at single-cell resolution is advanced, host-response gene panels for molecular differentiation of H5N1 from bacterial mastitis using milk-derived cells are supported, and priorities for future experimental validation and diagnostic assay development are highlighted. Keywords: H5N1, Bacterial mastitis, Single-cell RNA sequencing, Bovine milk somatic cells, Host response