Chemotherapy resistance in triple-negative breast cancer (TNBC) remains a critical clinical challenge, with a substantial proportion of patients failing to achieve pathological complete response following neoadjuvant chemotherapy (NAC). Using an integrative single-cell RNA sequencing (scRNA-seq), bulk transcriptomic, and spatial proteomic framework, we aimed to identify the malignant epithelial subset driving this resistance and the intercellular signaling axes through which it reprograms the tumor microenvironment (TME). scRNA-seq analysis of NAC-treated breast tumors revealed a Fatty Acid-EMT co-expressing epithelial subset (FA-EMT) that is selectively enriched in the chemotherapy-resistant residuum. Critically, FA-EMT co-expression-rather than either program individually-most powerfully predicted chemotherapy resistance and reduced overall survival across two independent bulk transcriptomic cohorts comprising 277 TNBC patients (p < 0.001). CellChat ligand-receptor analysis established FA-EMT cells as the dominant TME signaling hub, deploying MDK-NCL and MIF-CD74-CXCR4 axes to simultaneously suppress adaptive and innate anti-tumor immunity via T-cell exhaustion, Treg activation, and the expansion of myeloid-derived suppressor cells. Spatial CyCIF validation in a published paclitaxel-resistant TNBC mouse model (n = 69 cores) confirmed significant Metabolic-EMT enrichment in resistant tumor cores (p = 0.0085) with physical co-localization with immunosuppressive MDSC and Treg populations. These findings establish the FA-EMT subset as a key cellular driver of treatment failure in TNBC and nominate MDK-NCL and MIF-CD74-CXCR4 as mechanistically grounded therapeutic targets with the potential to dismantle the FA-EMT-driven immunosuppressive niche and sensitize chemotherapy-resistant TNBC to cytotoxic treatment.

Triple-negative breast cancer (TNBC) is characterized by marked immune microenvironment heterogeneity and variable chemotherapy response, yet the epithelial transcriptional programs governing cytotoxic immune activation remain poorly understood. We performed an exploratory, integrative analysis using single-cell RNA sequencing of 31,962 cells from eight TNBC patients operationally stratified into Good and Bad Prognosis groups based on pathological lymphoid infiltration, a discovery grouping subsequently validated against pathological complete response (pCR) in three independent bulk RNA-seq cohorts. This analysis identified four epithelial transcriptional states. The G5 DNA damage subpopulation—predominantly restricted to Good Prognosis tumors (29.2% vs. 0%)—and the G4 Metabolism subpopulation—2.4-fold enriched in Bad Prognosis—were the primary prognostic signatures. Machine learning validation using nested leave-one-cohort-out (LOCO) cross-validation across 614 samples demonstrated that G4 + G5 raw genes with random forest yielded the largest observed mean AUC of 0.653, though these results are exploratory and do not establish a validated clinical classifier. CellChat ligand–receptor interaction analysis revealed that G5 DNA-damage epithelial cells are the dominant immune activators in Good Prognosis TNBC, predominantly engaging CD8 cytotoxic T cells through MHC-I antigen presentation via HLA-A/B/C/E/F → CD8A/CD8B interactions, the highest-probability signaling pathway identified. Spatial transcriptomics independently validated significantly higher DNA damage and CD8 T-cell scores in Good Prognosis tissue. Together, these exploratory findings suggest a framework in which tumor-intrinsic DNA damage signaling is associated with MHC-I antigen presentation upregulation and CD8 cytotoxic T-cell engagement, supporting further investigation of this axis and its potential implications for combining DNA-damaging chemotherapy with immune checkpoint blockade in TNBC.

KPC cell lines show differential growth response to HGF treatment

ATAC-seq and RNA-seq reveal altered TGFβ response with PIN1 knockdown

PIN1 loss impacts KPC pancreas weight and fibroblast ⍺SMA expression

RNA-seq on inhibitor samples, FPKMs all protein coding genes

PIN1 inhibition slows PSC growth and prevents TGFβ-driven activation

NABA_MATRISOME overlap with DEGs (FPKMs)

The prolyl isomerase PIN1 is overexpressed in cancer and contributes to cancer cell-intrinsic phenotypes, including proliferation and migration. However, PIN1 may also function in stromal cells within the tumor microenvironment. In this study, we showed that PIN1 is a critical regulator of pancreatic stellate cell (PSC) state at baseline and in response to the myofibroblast-activating factor TGFβ. Loss or inhibition of PIN1 altered the epigenetic and transcriptional responses of PSCs to TGFβ, preventing PSC differentiation to a myofibroblast state and altering expression of secreted matrix proteins and signaling molecules. Consistent with inhibition of the TGFβ response, low fibroblast PIN1 expression in mouse and human pancreatic ductal adenocarcinoma correlated with low expression of αSMA, a marker of myofibroblast activation. Decreased PIN1 expression at baseline also altered paracrine hepatocyte growth factor (HGF) signaling from fibroblasts to tumor cells. PSCs with low PIN1 expression displayed reduced expression and secretion of HGF, resulting in an attenuation of c-MET receptor phosphorylation and signaling in nearby cancer cells. In allograft models, host PIN1 was critical for normal growth of a subset of pancreatic cancer cell lines that are responsive to HGF signaling. Through the identification of changes to fibroblast activation state and cross-talk following PIN1 loss or inhibition, these data suggest that systemic targeting of PIN1 will suppress the protumorigenic pancreatic ductal adenocarcinoma microenvironment and may differentially affect heterogeneous patient populations. SIGNIFICANCE:PIN1 plays a critical role in the response of pancreatic stellate cells to TGFβ and can be targeted to attenuate myofibroblast activation and protumor cross-talk to suppress pancreatic cancer progression.

FPKMs of 2698 DEGs across inhibitor samples

Loss of PIN1 in host cells slows tumor growth and alters trichrome staining in orthotopic allografts of two KPC cell lines

DESeq2 for shSCR PSCs comparing TGF-β treated vs untreated

PIN1 knockdown or overexpression impacts PSC phenotypes at baseline and in response to TGFβ

PIN1 knockdown in PSCs impacts signaling to tumor cells via altered HGF secretion

FPKMs of 1220 DEGs across knockdown samples

BACKGROUND/OBJECTIVES:Triple-negative breast cancer (TNBC) exhibits high immune infiltration yet remains clinically aggressive. Although immune checkpoint blockade benefits a subset of patients, the molecular programs enabling concurrent immune activation and immune evasion in TNBC are not fully defined. This study aimed to identify TNBC-specific tumor-intrinsic and tumor-extrinsic molecular features that may explain this paradox. METHODS:Publicly available single-cell RNA-sequencing data from primary breast tumors were analyzed to characterize subtype-specific transcriptional programs across epithelial and stromal compartments. Tumor-intrinsic findings were independently validated using bulk transcriptomic and clinical data from the METABRIC cohort. Tumor microenvironment remodeling was evaluated using multiplexed tissue imaging of TNBC tumors. Functional analyses were done included Gene Ontology enrichment, Hallmark gene set enrichment analysis, and SERPINB3-centered protein-protein interaction network analysis using STRING. RESULTS:Single-cell analysis identified SERPINB3 as a TNBC-enriched epithelial gene relative to ER+ and HER2+ tumors. This subtype-restricted pattern was validated in the METABRIC cohort and associated with pathways related to epithelial-mesenchymal transition, interferon signaling, and antigen presentation. TNBC tumors also displayed a humoral immune signature characterized by B-cell and plasmablast enrichment, as well as ectopic immunoglobulin gene expression in cancer-associated fibroblasts, endothelial cells, and myeloid populations. Multiplex imaging revealed coordinated associations between immune suppression, stromal activation, and tumor proliferation. Network analysis placed SERPINB3 within interconnected immune-regulatory and stromal signaling modules. CONCLUSIONS:Together, these data indicate that TNBC exhibits co-existing immune activation and immune-suppressive features. The identified epithelial and stromal signatures represent candidate biomarkers that may inform future studies of immune regulation and therapeutic stratification in TNBC.

CAF marker FPKMs in inhibitor samples

Secreted Signals overlap with DEGs (FPKMs)

DESeq2 comparing DMSO untreated vs TGFβ treated

PIN1 knockdown or inhibition impacts the transcriptional response to TGFβ