Prognostic model construction and drug prediction in colorectal cancer using mitochondrial programmed cell death-related genes
摘要
Colorectal cancer (CRC) is a highly aggressive and heterogeneous malignancy with a poor prognosis. Mitochondria and programmed cell death (PCD) play crucial roles in CRC tumorigenesis and cancer progression. However, the prognostic significance of mitochondrial programmed cell death (MPCD) remains unclear. This study aims to investigate prognostic value of MPCD-related genes.
MethodsCRC patient data from The Cancer Genome Atlas (TCGA) and MPCD gene sets from previous studies were collected. A prognostic model based on MPCD-related genes was developed using regression analysis and validated using an external CRC cohort from Gene Expression Omnibus (GEO). Immune microenvironment of the two risk cohorts was evaluated using ssGSEA and CIBERSORT algorithms. Potential anti-tumor drugs targeting model genes were predicted using pRRophetic R package. In addition, CRC-related cell lines were collected, and functional tests were conducted on the key risk gene SRPX.
ResultsA prognostic model comprising 13 MPCD genes (ACSL6, ELANE, G6PC2, ITGB3, MTM1, NAT1, PLAU, SCG2, SLC7A11, SMPD1, SRPX, TPM1, TRAP1) was successfully constructed, which demonstrated stable prognostic performance across different CRC cohorts. Immune infiltration analysis revealed significant enrichment of Tregs and Tfh cells in the high-MPCD cohort, suggesting that immune tolerance may contribute to the poor prognosis in this group. Three candidate drugs—CGP-60474, WH-4-023, and Trametinib—were identified as having significant negative correlations with model genes SRPX, SCG2, and SMPD1. Cell experiments demonstrated that overexpression of SRPX significantly inhibited the migration and invasion abilities of HCT116 cells.
ConclusionThis study developed an independent and reliable MPCD-related prognostic model. The model provides new targets for personalized risk assessment and immunotherapeutic interventions in CRC.