Abstract
BACKGROUND: Low-density lipoprotein (LDL) is a critical regulator of lipid metabolism and has been implicated in the development and progression of various malignancies. However, its specific roles and mechanisms in the ovarian cancer tumor microenvironment (TME) remain unclear. This study aimed to comprehensively elucidate the distribution, functional pathways, and prognostic value of LDL in ovarian cancer using single-cell transcriptome analysis. METHODS: Single-cell transcriptome data from ovarian cancer patients were analyzed. The AUCell algorithm was used to score LDL-related gene expression in different cell subsets, dividing cells into high and low LDL score groups. Functional pathway enrichment (Gene Set Variation Analysis [GSVA]) and cell-cell communication (CellChat) analyses were performed. Differentially expressed genes (DEGs) identified between the two groups were combined with bulk RNA-seq data from eight cohorts to construct the LDL-related ovarian cancer prognostic signature (LDLOCPS) using machine learning. Prognostic performance and immune landscape differences were evaluated between high and low LDLOCPS groups. RESULTS: LDL was predominantly highly expressed in myeloid cells (macrophages and monocytes) and stromal cells (fibroblasts, smooth muscle cells, and endothelial cells) within the ovarian cancer TME. GSVA revealed that the high LDL score group was significantly enriched for pathways including epithelial-mesenchymal transition (EMT), inflammatory response, coagulation, and angiogenesis. CellChat analysis demonstrated enhanced cell-cell communication involving IL6, CSF, and tenascin in the high LDL score group, with SPP1+ macrophages and monocytes showing stronger incoming and outgoing signals. The LDLOCPS model, constructed from bulk transcriptomic data and validated across eight cohorts, effectively stratified patients by risk; the high LDLOCPS group exhibited significantly worse overall survival. Receiver operating characteristic (ROC) and principal component analysis (PCA) analyses confirmed the robust predictive performance of LDLOCPS. Moreover, patients in the high LDLOCPS group showed reduced immune cell infiltration and lower expression of immune-related genes, suggesting an immunosuppressive microenvironment. CONCLUSION: This study systematically reveals the spatial distribution of LDL within the ovarian cancer microenvironment and uncovers its regulatory roles in tumor progression through multiple signaling pathways. The LDLOCPS model provides a valuable tool for risk stratification and prognosis prediction in ovarian cancer. LDL-mediated microenvironmental and immunosuppressive effects may offer novel insights for developing targeted and immunomodulatory therapies in ovarian cancer.