Document Type : Research Paper
Extended Abstract
Introduction
Mastitis is one of the major health challenges in dairy herds, typically arising from the invasion of a wide range of microorganisms into the mammary gland, which leads to intramammary infections (IMI) and inflammatory responses. Beyond its detrimental effects on animal welfare, mastitis causes substantial economic losses to dairy producers due to decreased milk yield and altered milk composition, increased veterinary and treatment costs, and the premature culling of chronically affected cows. Over the past two decades, the prevalence of Prototheca spp. microalgae have increased steadily, establishing them as emerging mastitis pathogens in dairy herds. Owing to their intrinsic resistance to antimicrobial treatment, the difficult to diagnose Prototheca-associated infections are frequently persistent, and often overlooked. In this context, RNA sequencing (RNA-seq) and transcriptomic profiling represent powerful tools for elucidating complex host-pathogen interactions. However, to date, comprehensive studies integrating co-expression module analysis and the reconstruction of multi-partite regulatory networks to elucidate gene regulatory mechanisms associated with Prototheca spp. infection in milk somatic cells have not been conducted. Therefore, the objective of this study was to systematically characterize the transcriptomic response of mammary tissue to Prototheca spp. infection in Holstein dairy cows. Using an integrative, multi-partite analytical framework, including comparative RNA-seq analysis, weighted gene co-expression network analysis (WGCNA), and bipartite miRNA–mRNA regulatory network reconstruction, we aimed to identify key hub co-expression modules, regulatory transcripts, and associated functional pathways. Ultimately, this study provides deeper genetic insight into bovine mastitis caused by Prototheca spp. and supports the development of informed breeding strategies to select cows with enhanced resistance as parents of the next generations.
Materials and Methods
This study analyzed RNA-seq data derived from milk somatic cell samples of Holstein dairy cows, comprising two groups: mastitis-affected cows infected with Prototheca spp. and healthy control cows. The dataset, corresponding to project number PRJNA911953, is publicly available in the GEO database from NCBI and includes 15 biological replicates (eight clinical mastitis and seven healthy samples). The quality of raw sequencing reads was initially assessed using FastQC, followed by adapter trimming and removal of low-quality reads using Trimmomatic. Cleaned reads were then aligned to the bovine reference genome (Bos taurus) using HISAT2, and gene-level read counts were generated with featureCounts. Differential expression analysis of transcripts, including both mRNAs and miRNAs, was performed using the DESeq2 package. To stabilize variance and facilitate downstream co-expression network analyses, Variance Stabilizing Transformation (VST) normalization was applied using the vst(blind = TRUE) function. Weighted gene co-expression network analysis (WGCNA) was conducted in the R software environment. The optimal soft-thresholding power was determined using the pickSoftThreshold function with powers ranging from 1 to 20 and networkType = "signed". A soft-threshold power of β = 19 was selected based on achieving a scale-free topology fit index (R²) was > 0.8 in conjunction with an appropriate mean connectivity profile. Gene Ontology (GO) and pathway enrichment analyses were performed to identify significantly enriched biological processes and metabolic-signaling pathways using the DAVID online tool and the STRING database. Statistical significance was determined using a false discovery rate (FDR) threshold of < 0.05, adjusted by the Benjamini–Hochberg method. To further elucidate the regulatory mechanisms underlying bovine mastitis caused by Prototheca spp., an integrated mRNA–ncRNA regulatory network framework was constructed. This framework incorporated protein–protein interaction data and miRNA–target gene interactions to investigate molecular relationships and regulatory patterns influencing gene expression in bovine mastitis. Network visualization and analysis were performed using Cytoscape software.
Results and Discussion
Following quality control confirmation, differential expression analysis was conducted to identify transcripts exhibiting significant expression changes between cows with clinical mastitis and healthy cows. In total, 6,726 genes showed significant differential expression (P < 0.05). Based on the defined thresholds (|log₂ fold change| ≥ 2 and FDR < 0.05), 119 mRNAs, five miRNAs, and four lncRNAs were significantly upregulated, while 181 mRNAs, nine miRNAs, and 10 lncRNAs were significantly downregulated in cows with clinical mastitis compared with healthy controls. Analysis of chromosomal distribution revealed non-uniform transcript abundance patterns, indicating that certain chromosomes—particularly chromosomes 2 and 5—acted as hub chromosomes, each harboring 28 differentially expressed transcripts involved in regulating molecular responses to mastitis-associated IMIs. To satisfy the criterion of approximate scale-free topology in co-expression network construction, the soft-thresholding power (β) was set to 19, achieving a scale-free topology fit index (R²) > 0.80. Using clinical mastitis samples as the reference set, hierarchical clustering combined with the Dynamic Tree Cut algorithm identified 19 distinct co-expression modules. Integration of comparative transcriptome analysis with co-expression network results revealed 104 and 135 overlapping genes between the lists of upregulated and downregulated genes and their corresponding co-expressed modules, respectively. Based on experimentally supported and predicted miRNA–mRNA interactions, bipartite miRNA–mRNA regulatory networks were subsequently reconstructed for both upregulated and downregulated transcript sets. Within the regulatory co-expression network associated with upregulated genes, 13 hub mRNAs (POU2AF1, FCRL5, JCHAIN, CD19, MS4A1, CD79A, IL17A, IFNG, PAX5, CXCR5, TNFRSF13C, CD79B, and CD22) and two hub miRNAs (bta-miR-2404 and bta-miR-2887) were identified. These hub transcripts were predominantly involved in host defense responses, immune system activation through the regulation of inflammatory processes, B- and T-cell activity, cytokine–cytokine receptor interactions, and the IL-17 signaling pathway. Collectively, they play a critical role in adaptive immunity by enhancing immune preparedness and facilitating effective responses against pathogenic invasion. Conversely, the regulatory co-expression network associated with downregulated genes revealed six hub mRNAs (CSN1S1, CSN1S2, CSN2, BTN1A1, LALBA, and PLIN2) and six hub miRNAs (bta-miR-2462, bta-miR-34a, bta-miR-1434, bta-miR-12001, bta-miR-11981, and bta-miR-1835). These downregulated transcripts were primarily associated with the regulation of milk synthesis and its components, maintenance of mammary tissue function, and hormonal signaling pathways—particularly those involving progesterone and estradiol—that govern mammary epithelial cell growth and differentiation. Additionally, these transcripts were linked to the glucagon signaling pathway, which is known to suppress milk protein synthesis and disrupt glucose metabolism during mastitis.
Conclusion
The findings of this study demonstrate that an integrative bioinformatics approach is effective in identifying key hub co-expression modules, hub transcripts (mRNAs and miRNAs), and functional pathways associated with bovine mastitis caused by Prototheca spp., thereby providing novel insights into the genetic architecture and gene regulatory mechanisms underlying this disease. These results enhance our understanding of host responses to Prototheca-induced mastitis and highlight potential molecular targets associated with resistance or susceptibility. However, to translate these findings into practical breeding strategies aimed at improving herd resistance and identifying mastitis-susceptible dairy cows, further investigations involving larger populations, independent datasets, and experimental validation of the identified hub transcripts are required.
Author Contributions
Conceptualization: Farzad Ghafouri, Mostafa Sadeghi, Seyed Reza Miraei-Ashtiani, Herman Wildrik Barkema and Masoud Shirali; methodology: Farzad Ghafouri; formal analysis: Farzad Ghafouri and Masoud Shirali; writing—original draft preparation, Farzad Ghafouri; writing—review and editing: Mostafa Sadeghi, Seyed Reza Miraei-Ashtiani, Herman Wildrik Barkema and Masoud Shirali; supervision: Mostafa Sadeghi, Masoud Shirali, Seyed Reza Miraei-Ashtiani and Herman Wildrik Barkema. All authors have read and agreed to the published version of the manuscript.
Data Availability Statement
Datasets used in this study are publicly available and can be accessed from the National Center for Biotechnology Information (NCBI) with project number PRJNA911953. Further details on accessing the data are on the NCBI website at https://www.ncbi.nlm.nih.gov/.
Acknowledgements
The authors sincerely thank the researchers of the project with accession number PRJNA911953 for generating and sharing their transcriptomic data through the NCBI public repository. Open access to these data enabled the complementary and innovative analyses of the present study, including weighted gene co-expression network analysis (WGCNA) and reconstruction of bipartite mRNA-miRNA regulatory networks. We hereby appreciate the valuable efforts of these researchers in advancing the knowledge of bovine mastitis.
Ethical Considerations
The study did not involve human or animal subjects and therefore did not require ethical approval. The authors confirm that no data fabrication, falsification, plagiarism, or misconduct occurred.
Conflict of interest
The authors declare no conflicts of interest.
