BACKGROUND

Prostate cancer (PCa) is a multifactorial and heterogeneous disease, ranking among the most prevalent malignancies in men. In 2020, there were 1,414,259 new cases of PCa worldwide, accounting for 7.3% of all malignant tumors. The incidence rate of PCa ranks third, following breast cancer and lung cancer. Patients diagnosed with high-grade PCa frequently present with existing or developing metastases, complicating their treatment and resulting in poorer prognoses, particularly for those with bone metastases. Utilizing single-cell RNA sequencing (scRNA-seq), we identified specific malignant cell subtypes that are closely linked to high-grade PCa. By investigating the mechanisms that govern interactions within the tumor microenvironment (TME), we aim to offer new theoretical insights that can enhance the prevention, diagnosis, and treatment of PCa, ultimately striving to improve patient outcomes and quality of life.

METHODS

Data on scRNA-seq was obtained from the GEO database. The gene ontology and gene set enrichment analysis were employed to analyze differential expression genes. Using inferCNV analysis to identify malignant epithelial cells. We subsequently employed Monocle, Cytotrace, and Slingshot packages to infer subtype differentiation trajectories. The cellular communication between malignant cell subtypes and other cells was predicted using the CellChat package. Furthermore, we employed pySCENIC to analyze and identify the regulatory networks of transcription factors (TFs) in malignant cell subtypes. The MDA PCa 2b and VCap cell lines were employed to validate the analysis results through cellular functional experiments. In addition, a risk scoring model was developed to assess the variation in clinical characteristics, prognosis, immune infiltration, immune checkpoint, and drug sensitivity.

RESULTS

A malignant cell subtype in PCa with high expression of was identified through scRNA-seq analysis. This subtype was situated at the differentiation terminal, exhibited a higher level of malignancy, and exhibited characteristics that were more prone to advanced tumor lesions. In addition, our research underscored the intricate interactions that exist within the TME, particularly the interaction between PTN secreted by this subtype and fibroblasts via the NCL receptor. This interaction may be closely associated with cancer-associated fibroblasts and tumor progression. Subsequently, we determined that the + malignant cell subtype was significantly correlated with the TF IRX4. This TF is linked to a worse prognosis in PCa and may affect disease progression by regulating gene transcription. Our conclusions were additionally verified through cellular experiments. Furthermore, the prognostic model we developed demonstrated satisfactory predictive performance, with gene sets from the high NmRS group facilitating tumor progression and deterioration. The analysis of immune infiltration was instrumental in the development of clinical intervention strategies and patient prognosis.

CONCLUSION

By examining the cellular heterogeneity of a unique malignant cell subtype within the PCa microenvironment, we were able to disclose their reciprocal interaction with disease progression. This offers a novel viewpoint on the diagnosis and treatment of PCa.