Integrative analysis of molecular mechanisms in prostate cancer via single-cell RNA sequencing and weighted gene co-expression network analysis.

Despite extensive prior research on prostate cancer (PCa) transcriptomics, the molecular mechanisms underlying the disease's progression, particularly in the castration-resistant or metastatic stages, remain incompletely understood. The majority of recent research has concentrated on bulk RNA sequencing, which could mask the variation found in tumor microenvironments. This study aims to address this gap by integrating single-cell RNA sequencing (scRNA-seq) and bulk RNA sequencing with weighted gene co-expression network analysis (WGCNA) to investigate the molecular mechanisms of PCa at a higher resolution. In order to further individualized treatment plans for PCa, we aim to discover important genes and signaling pathways that could be used as therapeutic targets. We first preprocessed expression profile data from prostate cancer tissue samples, selecting 9,809 high-quality cells from a dataset. Following batch correction with Harmony and dimensionality reduction with principal component analysis (PCA), we used the Louvain clustering algorithm to divide the cells into discrete subtypes. The clusters were then visualized using t-SNE. This resulted in 16 cellular subtypes categorized into five major cell types: epithelial cells, monocytes, endothelial cells, CD8 + T-cells, and fibroblasts. Analysis of receptor-ligand pairs uncovered significant interactions between monocytes and both tumor cells and endothelial cells. Applying the high-dimensional WGCNA (hdWGCNA) method to construct a gene co-expression network, we detected seven gene modules, four of which were highly expressed in tumor cell subtypes and contained 380 key genes. Combining pathway analysis, we ultimately screened six key genes: CNPY2, CPE, DPP4, IDH1, NIPSNAP3A, and WNK4. We used Cox univariate regression and least absolute shrinkage and selection operator (lasso) regression techniques to build a prognostic prediction model that included these six important genes based on clinical data gathered from PCa patients. The prognostic prediction model constructed in this study demonstrated excellent predictive performance in both the training set and an external validation set, with the high-risk group showing significantly lower overall survival (OS) than the low-risk group. Furthermore, there was a substantial correlation found between risk scores and several immune-related gene sets, chemotherapeutic drug sensitivity, and tumor immune infiltration. High- and low-risk groups exhibited significant differences in immune cell content, immune factor levels, and immune dysfunction. Further analysis revealed significant correlations between the expression levels of model genes and multiple disease-related genes. Through Gene Set Variation Analysis (GSVA) and Gene Set Enrichment Analysis (GSEA), we uncovered perturbations in multiple signaling pathways in high- and low-risk groups, potentially impacting the prognosis of PCa patients. This study uncovers key genes and signaling pathways in the prostate cancer tumor microenvironment, particularly genes such as CNPY2, CPE, DPP4, IDH1, NIPSNAP3A and WNK4, which have potential as therapeutic targets. Our findings provide new insights into personalized treatment strategies for PCa and warrant further clinical validation in the future.