Abstract Background Programmed Cell death (PCD) encompasses a spectrum of genetically regulated cell death processes and plays a double-edged sword role in neoplastic progression and therapeutic resistance of Triple-Negative Breast Cancer(TNBC)through the tumor microenvironment (TME). However, the specific mechanisms by which PCD mediates microenvironmental dysregulation remain elusive. Methods Analyzing nine samples of TNBC through single-cell RNA sequencing (scRNA-seq), this study employed nonnegative matrix factorization (NMF) to assess genes associated with 13 PCD modes. Single-cell regulatory network inference and clustering (SCENIC), Monocle, CellChat, and scMetabolism were used for pseudotime analysis, intercellular communication mapping, determination of transcription factor activities (TFs), and immune infiltration of PCD-related cell clusters in TME. A robust prognostic model and drug resistance analysis were constructed using gene set enrichment analysis (GSEA), Kaplan-Meier survival analysis, and multivariable Cox regression. Finally, hub genes and critical PCD-related cell clusters were validated in the clinical breast cancer samples and the TNBC model mice. Results This investigation demonstrated that PCD significantly modulated the functional and phenotypic diversity of fibroblasts, macrophages, T cells, and B cells in the TME of TNBC. Furthermore, this study revealed that PCD-regulated CEBPB-positive cancer-associated fibroblast (CAF) populations are a key determinant of the TNBC immune Microenvironment heterogeneity and poor prognosis. Notably, CellChat analysis unveiled diverse and extensive interactions between PCD-related cell clusters and tumor immune cells, highlighting the CEBPB+ CAF subtype as a signaling ligand communicated with other immune cell clusters through the Midkine (MDK)-Nucleolin (NCL) signaling axis. Moreover, the TIDE analysis verified that CEBPB+ CAF is a predictor of poor prognosis in Immunotherapy. The ex vivo analyses of tumor specimens from both TNBC patients and syngeneic murine models were performed by quantitative reverse-transcription PCR (qRT-PCR), immunoblotting, immunohistochemical staining, and multiplexed immunofluorescence co-localization assays. They confirmed differential expression of the PCD-related prognostic genes and the presence of CEBPB+ CAFs. Conclusion In summary, our study provides a comprehensive molecular framework to understand the role of PCD-mediated TME dysregulation in TNBC pathogenesis. This study also offers new insights into the underlying mechanisms of immune therapy resistance in TNBC and identifies promising therapeutic targets for enhancing treatment efficacy and patient outcomes. Keywords: programmed cell death, triple-negative breast cancer, immune microenvironment, prognostic model, fibroblasts, drug resistance 1. Introduction Breast cancer (BC) is one of the most common malignant tumors among women. It accounts for approximately 31% of new cancer cases and is the second leading cause of cancer-related deaths among women Triple-negative breast cancer (TNBC) demonstrates significantly higher invasive and metastatic abilities than the other types of BC. The high recurrence rates of TNBC are associated with its increased propensity for vascular invasion. Therefore, patients with TNBC have the poorest prognosis among the various BC types ([31]1). Furthermore, because of the lack of ER, PR, and HER2 expression, the efficacy of conventional treatments such as endocrine and targeted therapies is limited, and advanced patients with TNBC have a median survival of less than 24 months ([32]2). In recent years, immunotherapy targeting the tumor microenvironment (TME) has shown higher clinical efficacy in patients with TNBC. The cell types within the TME of TNBC patients play contradictory roles in tumor growth and progression. Some cell types promote TNBC progression by secreting and expressing factors that enhance tumor cell proliferation and suppress anti-tumor immune responses. However, other cell types in the TME of TNBC patients suppress tumor growth by promoting adaptive immunity ([33]1, [34]3). In the phase I KEYNOTE-012 clinical trial, PD-1 inhibition via pembrolizumab monotherapy demonstrated sustained antineoplastic activity in both early-stage and advanced PD-L1-positive TNBC patients (Combined Positive Score [CPS] ≥ 1), but a subset of TNBC patients exhibited primary resistance to this immunotherapeutic intervention ([35]4). Zhang et al. stratified TNBC into two distinct immunological subtypes, namely, macrophage-enrichment subtype and neutrophil-enrichment subtype, and identified differential mechanisms of immunotherapy resistance across myeloid cell populations in TNBC ([36]5). The underlying mechanisms of immunotherapeutic resistance in TNBC are complex because of significant TME heterogeneity and intricate cellular crosstalk between the neoplastic cells and the diverse stromal cell populations within the TME. Consequently, an in-depth analysis of the TME-mediated immunotherapy resistance mechanisms is necessary to improve the therapeutic outcomes and enhance the prognostic indicators for TNBC patients. Immunological heterogeneity within the tumor microenvironment is associated with differential outcomes of treatment modalities targeting the programmed cell death (PCD) mechanisms in the cancer cells ([37]6). PCD encompasses a spectrum of genetically regulated cell death processes that are orchestrated by distinct molecular cascades and signal transduction pathways. PCD plays a pivotal role in neoplastic progression, therapeutic resistance, and immunological escape mechanisms. It is currently known that the PCD spectrum comprises 13 distinct modalities, including macroautophagy, type I PCD (apoptosis), RIPK1/RIPK3-mediated necroptosis, inflammatory caspase-dependent pyroptosis, copper-induced cuproptosis, and iron-dependent ferroptosis ([38]7, [39]8). Neoplastic cells undergoing multimodal PCD modulate the immune system by secreting a wide array of cytokines, which facilitate chemotactic recruitment of diverse immune cell populations and phenotypic transformation of tumor-infiltrating lymphocytes and resident immune cells into immunosuppressive phenotypes. This facilitates TME remodeling and promotes immune evasion and therapeutic refractoriness ([40]9). Conversely, PCD mechanisms facilitate conventional dendritic cell (cDC) trafficking to the tumor-draining lymph nodes (tdLNs) and potentiate adaptive immune responses by releasing damage-associated molecular patterns and proinflammatory mediators, thereby suppressing tumor growth and enhancing immunotherapeutic efficacy ([41]10). This functional dichotomy has established PCD as a critical homeostatic regulator within the TME. Therefore, PCD has emerged as a central focus in immuno-oncological research. Previous studies have reported that PCD mechanisms are dysregulated in TNBC ([42]11). Novel pharmacological agents targeting the PCD pathways have demonstrated significant therapeutic potential in TNBC. Chen et al. reported that phloretin (Ph), a dihydrochalcone flavonoid derivative, suppressed the proliferation of TNBC cells by downregulating the mTOR/ULK1 signaling pathway and suppressing autophagy ([43]12). Spirooxindole 6e induced PCD in the MDA-MB-231 cells by modulating the intrinsic apoptotic pathway through downregulation of Bcl-2, upregulation of Bax, and activation of caspase-3 ([44]13). The current research paradigms focus on individual PCD mechanisms in TNBC. However, sustained modulation of individual PCD pathways may induce therapeutic resistance. Furthermore, the bidirectional effects of PCD represent significant challenges in developing precision-targeted therapeutics for TNBC ([45]8). The exponential expansion of public genomic repositories and revolutionary advances in single-cell analytical methods have facilitated multi-omics approaches to identify novel tumor pathogenetic mechanisms and therapeutic targets. This study comprehensively analyzed data derived from Next Generation Sequencing to unravel novel cellular phenotypes and molecular signatures mediated by PCD-related prognostic genes in TME of TNBC. Applying nonnegative matrix factorization (NMF), single-cell regulatory network inference and clustering (SCENIC), gene set enrichment analysis (GSEA) and CellChat, this study aimed to explore signaling pathways, functional enrichments, intricate interplay, transcriptional features, immune characteristics, various developmental roles and prognostic implications within these distinct subgroups of CAFs, macrophages, T cells, and B cells in the TME, providing a robust framework for developing highly effective targeted therapeutics for TNBC. In summary, this study sheds light on the potential role of PCD-related prognostic genes in the dysregulated immune microenvironment in Triple-Negative Breast Cancer, and provides a robust framework for developing highly effective targeted therapeutics for TNBC. 2. Materials and methods 2.1. Single-cell data processing We downloaded the single-cell RNA sequencing dataset [46]GSE176078 ([47]14)from the Gene Expression Omnibus (GEO) database ([48]www.ncbi.nlm.nih.gov/geo) and extracted nine single-cell samples of triple-negative breast cancer (TNBC). We calculated and filtered cells that expressed more than 300 genes using the PercentageFeatureSet function, with mitochondrial gene expression below 15%, and red blood cell gene proportion less than 1%. The merged ScRNA-seq data were normalized, and the top 2000 highly variable genes were identified using the FindVariableFeatures function. We then dimensionality reduction on the selected top 2000 highly variable genes using the ScaleData function and the RunPCA function. Batch correction was performed using the Harmony algorithm. Cells were clustered using the ‘FindNeighbors’ and ‘FindClusters’ functions (resolution=0.3) to identify cell clusters, and further dimensionality reduction was carried out using the UMAP method. Finally, the FindAllMarkers function was used to screen for marker genes in six subgroups, which were annotated and visualized using references and the CellMarker 2.0