Development of a novel prognostic prediction model using mitochondrial-related genes and single-cell sequencing data...

Colorectal carcinoma (CRC) is a prevalent malignancy worldwide. Due to suboptimal screening practices, CRC is frequently diagnosed at an advanced stage. The role of mitochondrial abnormalities in the advancement of CRC is significant, but the prognostic effect of mitochondrial genes remains unclear. Recent research emphasizes mitochondrial dysfunction's key role in oncogenesis and cancer progression. Mitochondria are vital for cellular metabolism, energy production, and regulating processes like apoptosis, redox homeostasis, and signal transduction. Their dysfunction causes metabolic reprogramming, heightened oxidative stress, and cell death resistance. Though mitochondrial abnormalities link to CRC's aggressive phenotypes, the prognostic value of a comprehensive set of mitochondrial genes remains unclear, leaving a critical knowledge gap. The primary objective of this study was to systematically investigate the prognostic potential of mitochondrial genes in CRC and to develop a novel, reliable risk-scoring model. We aimed to identify key mitochondrial genes associated with patient survival, construct a predictive signature, and validate its efficacy in independently prognosticating overall survival (OS).

This study used single-cell RNA sequencing (scRNA-seq) data of CRC tissues from The Cancer Genome Atlas (TCGA), and a comprehensive set of 1,650 mitochondrial genes from MitoCarta 3.0. A differential gene expression analysis, gene set enrichment analysis (GSEA), pathway analysis, and Cox proportional hazards regression analysis were conducted.

The Cox regression analysis identified five mitochondrial genes (i.e., CPT2, ACSL6, MOCS1, TERT, and PTRH1). These five genes were used to establish a reliable risk-scoring system. Patients with elevated risk scores had more severe clinical manifestations and worse survival outcomes. These results were corroborated in an external validation set (GSE17536 cohort). A predictive model was developed based on these genes that had robust predictive capabilities with areas under the curve (AUCs) of 0.75, 0.77, and 0.78 in the prediction of the 1-, 3-, and 5-year OS of CRC patients, respectively. Additionally, the correlation between the risk score and immune microenvironment characteristics (e.g., immune infiltration patterns, stromal/immune scores) further substantiated the predictive power of the model.

This study established a novel prognostic model for CRC based on mitochondrial genes, thereby extending the understanding of the disease's progression. Subsequent studies should seek to validate these findings within a broader cohort of patients and explore the potential therapeutic roles of the identified mitochondrial genes in the management of CRC.