Single-Cell and Spatial Transcriptomics Explore Purine Metabolism–Related Prognostic Risk Model and Tumor Immune Microenvironment Modulation in Ovarian Cancer

Background: Ovarian cancer (OC) ranks as the second leading cause of gynecological cancer–related deaths in women globally. Single-cell and spatial transcriptomics could precisely describe the heterogeneity of OC that affect the clinical treatment.

Methods: Single-cell sequencing and spatial transcriptomics information were from different public datasets. A pseudotime analysis of cellular developmental pathways, score single-cell gene sets, and cell activity ratings in each metabolic pathway were performed. A prognostic model was created using univariate regression analysis, LASSO, and multivariate regression analysis. Finally, the immune microenvironment and immunotherapeutic effects were analyzed for their association with purine metabolism activity. Finally, RT-qPCR was used to estimate the mRNA level of OC in cell lines.

Results: We observed a higher purine metabolism score by a signature of 12 purine metabolism–related gene in tumor cells. When compared with fibroblasts, epithelial cells with high scores displayed more intense TGF-β signaling pathway activity. Forty-four differentially expressed purine metabolism–related genes were identified to be substantially expressed in the tumor’s core region and were closely linked to purine and pyrimidine metabolic activities. Low-risk population had higher immune infiltration level and immunotherapy results. The NME6+ epithelial cell high-expression group had a greater prognosis and showed a negative connection with the tumor immune dysfunction and exclusion score and cancer-associated fibroblast cell concentration.

Conclusion: Purine metabolism was a predictor for OC patients’ prognosis. The presence of positive NME6 expression in epithelial cells emerges as a protective factor for OC patients, presenting a possible therapeutic target for personalized treatment.