The improvement of the prediction of prostate cancer (PCa) is a major challenge in disease management. This study analysed a total of 147,856 cells and identified 15 distinct cell types using single-cell RNA-sequencing (scRNA-seq) and bulk RNA-seq data from TCGA and GEO databases. Of these cells, 31,256 exhibited a high telomere-related gene score and were predominantly composed of myeloid dendritic cells (mDCs). Simultaneously, pseudo-temporal analysis indicated that mDCs are in the later stages of the differentiation trajectory, suggesting the significant role of mDCs as telomere-active cells in the development of PCa. Analysis of cell-cell communication revealed significant differences, particularly an increase in communication between mDCs and CTLs, alongside a decrease in communication between mDCs and B cells. These variations may represent critical nodes influencing the development of PCa. Additionally, two hub genes were utilized to create risk models, with ROC curves confirming their predictive efficacy for 3-, 5-, and 10-year survival rates in patients. Functional analysis of these genes was conducted, and NPY siRNA transfection notably inhibited proliferation in LNCaP and DU145 cells. Furthermore, the models demonstrated that high-risk patients had poorer overall survival, greater immune infiltration, and reduced sensitivity to chemotherapeutic drugs.
Keywords: Prostate cancer; Risk mode; Single-cell RNA-sequencing; Telomere-related gene; Tumour immune environment.
© 2025. The Author(s).