Abstract
Introduction
Triple Negative Breast Cancer (TNBC) is an aggressive cancer with fewer targeted therapies. Chemotherapy combined with anti–PD-1/PD-L1 immunotherapy is currently the standard of care for TNBC patients; however, fewer than 50% of patients achieve a durable response, pointing to additional immunosuppressive mechanisms within the tumor microenvironment (TME). TNBC progression is shaped by reciprocal signaling among malignant, immune, stromal, and vascular cells, which can establish protumorigenic and immunosuppressive circuits. We hypothesize that TNBC tumors harbor non-cell-autonomous intercellular communication networks that facilitate immune evasion through specific ligand-receptor signaling pathways.
Methods
We conducted an exploratory bioinformatics analysis using publicly available scRNA-seq and spatial transcriptomics data of 10 TNBC patients. Quality control filtered out cells with 20 features or >15% mitochondrial reads. Single-cell data were processed in Seurat to annotate major tumor, stromal, and immune lineages using canonical marker expression. CellChat was used to infer cell-cell communication networks and identify dominant ligand-receptor interactions. Spatial transcriptomics data were analyzed in R to map gene expression and evaluate the spatial co-localization of predicted interacting cell populations and ligand-receptor pairs within tumor tissues.
Results
Single-cell analysis identified diverse cell populations, including cancer epithelial cells, cancer-associated fibroblasts (CAFs), myeloid cells, and lymphocytes, within the TNBC TME. CellChat analysis revealed multi-lineage signaling networks, with CAFs showing the strongest outgoing signaling intensity and myeloid cells the greatest incoming communication. Across stromal-epithelial and immune-epithelial communication axes, the PPIA–BSG emerged as a prominent, recurrent ligand–receptor signaling bridge linking CAFs and myeloid cells to cancer epithelial cells. Comparative CellChat analysis revealed key tumor-specific signaling shifts: cancer cells gained myeloid-derived EGFR signaling and a broad CAF/myeloid receptor switch toward Syndecan-1 (SDC1) matrix interactions, alongside the loss of CAF-derived IGF signaling. Spatial transcriptomic analysis confirmed the co-enrichment of cancer epithelial cells, CAFs, and myeloid populations along with the spatial co-localization of PPIA and BSG transcripts at dense tumor–stroma interfaces
Conclusion
TNBC intercellular communication is coordinated across a complex multi-lineage network rather than driven by a single cell type. The altered communication in TNBC involves both tumor-specific gains (EGFR activation and SDC1 receptor rewiring) and conserved stroma–tumor drivers. The pervasive recurrence and spatial colocalization of the PPIA–BSG matrix-remodeling axis represents a high-priority candidate for functional validation and novel therapeutic targeting.