Diagnostic and prognostic values of differentially expressed genes in canine mammary carcinoma: An integrated bioinformatics analysis

Abstract

Background Canine mammary carcinoma (CMC) is a common tumor in unspayed dogs and poses a significant health concern for animals. This study aimed to identify differentially expressed genes (DEGs) between CMC and adjacent healthy mammary tissue through an integrated RNASeq bioinformatics analysis that combined an independently generated dataset from the present study, CPA-UN (CPA-UN), composed of RNA-seq data obtained from CMC and matched normal mammary tissue samples, with publicly available GEO datasets generated using next-generation sequencing (NGS). Candidate genes associated with diagnostic and prognostic potential were subsequently explored through integrative downstream analyses. Methods and findings DEGs were identified using DESeq2 and further analyzed for functional enrichment (ClusterProfiler, Pathview, and GSEA), co-expression gene networks, and tumor immune infiltrate deconvolution (CIBERSORTx). The GSE119810 dataset was used for exploratory overall survival (OS) analysis. Transcriptomic data from 88 CMC cases and adjacent healthy mammary tissue samples identified eight DEGs common across all four datasets. Among these, ACAN , COL11A1 , EDIL3 , NDUFA4L2 , and IGFBP5 showed higher transcript levels in CMC tissues than in healthy mammary tissue, whereas TNNC1 , PCK1 , and METTL24 showed lower transcript levels in CMC tissues than in healthy mammary tissue. Functional enrichment analyses indicated that DEGs with higher transcript levels in CMC were predominantly associated with extracellular matrix remodeling, cell adhesion, tumor microenvironment interactions, immune and inflammatory responses, and signaling pathways involved in tumor progression and metastasis. In contrast, DEGs with lower transcript levels were mainly enriched for cytoskeletal organization and tissue structural integrity, suggesting a loss of normal mammary gland architecture and myoepithelial-associated functions during tumor progression. GSEA further demonstrated coordinated enrichment of hallmark gene sets related to cell-cycle dysregulation, proliferation, epithelial–mesenchymal transition, metabolic adaptation, inflammatory signaling, and stromal remodeling, supporting the presence of integrated transcriptional programs that drive tumor progression and TME remodeling in CMC. Co-expression gene network analysis revealed a highly modular organization, with densely interconnected clusters of co-expressed genes that may reflect coordinated biological processes and regulatory programs. Exploratory CIBERSORTx analysis found no significant differences in the relative proportions of the 22 infiltrating immune cell types between CMCs and paired healthy controls after multiple-comparison corrections. Nevertheless, dataset-specific trends were observed in regulatory T cells, M1 macrophages, activated CD4 memory T cells, plasma cells, dendritic cells, and mast cells, though these findings should be interpreted cautiously. An exploratory survival analysis of 1,759 DEGs from the GSE119810 dataset identified 53 genes nominally associated with overall survival in CMC. However, none remained statistically significant after multiple-testing correction, underscoring the exploratory nature of these findings and the need for validation in larger cohorts. Conclusion This study provides a preliminary transcriptomic framework for CMC, identifying candidate genes and pathways associated with tumor-related processes and supporting future functional validation to clarify their roles in tumorigenesis, progression, and tumor aggressiveness. However, these findings remain exploratory and require validation in larger cohorts to confirm their diagnostic and prognostic relevance, given the limitations of secondary data analyses and potential variability in tissue collection and processing across studies.