# Identification of prognostic genes associated with mitochondria and macrophage polarization in prostate adenocarcinoma based on transcriptome and Mendelian randomization analysis

## TL;DR

This study identifies seven genes linked to mitochondria and immune cells in prostate cancer, offering new insights for prognosis and treatment.

## Contribution

A novel risk prediction model using mitochondrial and macrophage-related genes for prostate adenocarcinoma prognosis.

## Key findings

- Seven prognostic genes (ABHD11, PTRH2, CAT, NTHL1, SLC25A39, OXR1, GSTZ1) were identified for prostate adenocarcinoma.
- A risk prediction model using these genes showed high accuracy in outcome prediction.
- Drug sensitivity analysis revealed higher IC50 values for bexarotene and CCT018159 in high-risk patients.

## Abstract

Prostate adenocarcinoma (PRAD) is a common malignancy in the male genitourinary system, with growing evidence linking its progression to mitochondrial function and macrophage polarization. This study identifies prognostic genes associated with these factors in PRAD through integrated transcriptomic data analysis and Mendelian randomization (MR).

This study utilized transcriptome datasets from The Cancer Genome Atlas prostate adenocarcinoma (TCGA-PRAD). Candidate genes were selected by integrating mitochondrial-related genes (MRGs), macrophage polarization-related genes (MPRGs), and differentially expressed genes (DEGs). Prognostic genes were subsequently identified through MR and regression analyses, enabling the construction and validation of a risk prediction model. The model underwent independent prognostic assessment and nomogram validation, followed by comprehensive analyses including functional enrichment, immune profiling, and drug sensitivity evaluation comparing high- and low-risk cohorts.

From the overlap of 6,734 DEGs, 5,940 key module genes, and 1,136 MRGs, 103 CGs were identified. MR and regression analyses revealed seven prognostic genes (ABHD11, PTRH2, CAT, NTHL1, SLC25A39, OXR1, GSTZ1), which formed a robust risk prediction model. The model confirmed risk score, prostate-specific antigen, and Gleason score as independent prognostic factors for PRAD. A validated nomogram demonstrated high accuracy in outcome prediction. Functional enrichment analysis highlighted differential E2F target activity between risk groups, while immune profiling identified nine distinct cell populations, including immature dendritic cells. Finally, drug sensitivity analysis showed elevated IC50 values for bexarotene and CCT018159 in high-risk patients.

This study identified seven prognostic genes and provided a new theoretical basis for exploring immune defense mechanisms and targeted therapeutic drugs in PRAD.

The online version contains supplementary material available at 10.1007/s12672-025-03858-5.