A Comprehensive Prognostic Model for Colon Adenocarcinoma Depending on Nuclear-Mitochondrial-Related Genes

Abstract

Background: Colon adenocarcinoma (COAD) has increasing incidence and is one of the most common malignant tumors. The mitochondria involved in cell energy metabolism, oxygen free radical generation, and cell apoptosis play important roles in tumorigenesis and progression. The relationship between mitochondrial genes and COAD remains largely unknown. Methods: COAD data including 512 samples were set out from the UCSC Xena database. The nuclear mitochondrial-related genes (NMRGs)-related risk prognostic model and prognostic nomogram were constructed, and NMRGs-related gene mutation and the immune environment were analyzed using bioinformatics methods. Then, a liver metastasis model of colorectal cancer was constructed and protein expression was detected using Western blot assay. Results: A prognostic model for COAD was constructed. Comparing the prognostic model dataset and the validation dataset showed considerable correlation in both risk grouping and prognosis. Based on the risk score (RS) model, the samples of the prognostic dataset were divided into high risk group and low risk group. Moreover, pathologic N and T stage and tumor recurrence in the two risk groups were significantly different. The four prognostic factors, including age and pathologic T stage in the nomogram survival model also showed excellent predictive performance. An optimal combination of nine differentially expressed NMRGs was finally obtained, including LARS2, PARS2, ETHE1, LRPPRC, TMEM70, AARS2, ACAD9, VARS2, and ATP8A2. The high-RS group had more inflamed immune features, including T and CD4⁺ memory cell activation. Besides, mitochondria-associated LRPPRC and LARS2 expression levels were increased in vivo xenograft construction and liver metastases assays. Conclusion: This study established a comprehensive prognostic model for COAD, incorporating nine genes associated with nuclear-mitochondrial functions. This model demonstrates superior predictive performance across four prognostic factors: age, pathological T stage, tumor recurrence, and overall prognosis. It is anticipated to be an effective model for enhancing the prognosis and treatment of COAD.

Keywords: colon adenocarcinoma; nuclear mitochondrial-related genes; risk prognostic model.

Figure 1.

Identification and enrichment analysis of differentially expressed-nuclear mitochondrial-related genes (DE-NMRGs). (A) Volcano plot of differentially expressed genes. (B) A Venn-diagram of the DEGs and nuclear mitochondrial-related genes to obtain DE-NMRGs. The top 10 Gene Ontology (GO) function terms (C) and Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathways (D) enriched by the overlapping differentially expressed-nuclear mitochondrial-related genes (DE-NMRGs). The horizontal axis represents the number of genes, and the vertical axis represents the term. The color indicates significance, and the size of the dot is proportional to the number of genes.

Figure 2.

(A) The process of building the signature containing 19 DE-NMRGs most correlated with prognosis. The hazard ratios (HRs), 95% confidence intervals (CIs) calculated by univariate Cox regression. (B) LASSO logistic regression analysis for selection of optimal DE-NMRGs from DE-NMRGs. (C) The Kaplan–Meier (KM) survival curve of the 10 optimized DE-NMRGs combination. The blue and red curves represent the low- and high-expression groups, respectively.

Figure 3.

The Kaplan–Meier (KM) curve, risk score (RS) value distribution, and receiver operating characteristics (ROC) curve of the training dataset (A–C) and the validation dataset (D–F): (A and D) the blue and red curves represent the low-risk and high-risk sample groups, respectively; (B and E) survival time status; (C and F) the numbers in parentheses represent the specificity and sensitivity value for the corresponding ROC curves.

Figure 4.

The prognostic forest plots (A), nomogram survival rate model (B), and the survival nomograms predicting the 1-, 3-, and 5-year survival rate for independent prognostic clinical factors. (C) The horizontal axis represents the predicted survival, and the vertical axis represents the actual survival.

Figure 5.

The mutation information of 10 differentially expressed-nuclear mitochondrial-related genes (DE-NMRGs). (A) Statistical presentation of various mutation types of methylation regulators. (B) Mutation waterfall diagram of DE-NMRGs in samples. (C) Co-mutation map of NMRGs.

Figure 6.

The distribution of various immune cells (A), and correlation analysis among immune cells, ESTIMATE score, and nine differentially expressed-nuclear mitochondrial-related genes (DE-NMRGs) (B) in two risk groups.

Figure 7.

Representative protein expression of the nine hub genes.

Figure 8.

The expression levels of mitochondria-associated proteins in liver metastasis model. (A) Western blot analysis of related targets, including LRPPRC (150 kDa), LARS2 (70 kDa), ETHE1 (25 kDa) and GAPDH (37 kDa). (B) changes in the protein expression of LRPPRC, LARS2 and ETHE1 were visualized by histograms. Relative protein expression was adjusted by GAPDH. Statistical evaluations were carried out by ANOVA. All data are represented as the mean ± SD of experimental values in triplicate. *p < 0.05, **p < 0.01, ***p < 0.001.