pubmed.ncbi.nlm.nih.gov

Characterization of RNA Processing Genes in Colon Cancer for Predicting Clinical Outcomes - PubMed

  • ️Mon Jan 01 2024

Characterization of RNA Processing Genes in Colon Cancer for Predicting Clinical Outcomes

Jianwen Hu et al. Biomark Insights. 2024.

Abstract

Objective: Colon cancer is associated with multiple levels of molecular heterogeneity. RNA processing converts primary transcriptional RNA to mature RNA, which drives tumourigenesis and its maintenance. The characterisation of RNA processing genes in colon cancer urgently needs to be elucidated.

Methods: In this study, we obtained 1033 relevant samples from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases to explore the heterogeneity of RNA processing phenotypes in colon cancer. Firstly, Unsupervised hierarchical cluster analysis detected 4 subtypes with specific clinical outcomes and biological features via analysis of 485 RNA processing genes. Next, we adopted the least absolute shrinkage and selection operator (LASSO) as well as Cox regression model with penalty to characterise RNA processing-related prognostic features.

Results: An RNA processing-related prognostic risk model based on 10 genes including FXR1, MFAP1, RBM17, SAGE1, SNRPA1, SRRM4, ADAD1, DDX52, ERI1, and EXOSC7 was identified finally. A composite prognostic nomogram was constructed by combining this feature with the remaining clinical variables including TNM, age, sex, and stage. Genetic variation, pathway activation, and immune heterogeneity with risk signatures were also analysed via bioinformatics methods. The outcomes indicated that the high-risk subgroup was associated with higher genomic instability, increased proliferative and cycle characteristics, decreased tumour killer CD8+ T cells and poorer clinical prognosis than the low-risk group.

Conclusion: This prognostic classifier based on RNA-edited genes facilitates stratification of colon cancer into specific subgroups according to TNM and clinical outcomes, genetic variation, pathway activation, and immune heterogeneity. It can be used for diagnosis, classification and targeted treatment strategies comparable to current standards in precision medicine. It provides a rationale for elucidation of the role of RNA editing genes and their clinical significance in colon cancer as prognostic markers.

Keywords: Colon cancer; GEO; RNA processing gene; TCGA; nomogram; prognosis; risk score.

© The Author(s) 2024.

PubMed Disclaimer

Conflict of interest statement

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Figures

Figure 1.
Figure 1.

Identification of 4 different RNA processing modes in colon cancer. (A) Heatmap of 4 distinctive RNA processing modes defined by an unsupervised cluster analysis. (B) The principal component analysis (PCA) of the 4 subtypes were shown by the heatmap. (C) Kaplan-Meier survival curve analysis of overall survival for the 4 modes.

Figure 2.
Figure 2.

Identification of prognostic features related to RNA processing. (A) GO and KEGG functional enrichment analysis of differential genes among the 4 subgroups. (B) Lasso variable trajectory plot of prognosis-related RNA-processing genes. (C) Lasso coefficient filter plot of prognosis-related RNA-processing genes. (D) The corresponding hazard rations of the contained 10 genes in the signature represented by the dendrogram. (E-G) Kaplan-Meier survival curve analysis by log-rank test of high-risk and low-risk groups in combined database (E), TCGA database (F) and GEO database (G).

Figure 3.
Figure 3.

The construction of nomogram for overall survival in patients with colon cancer. (A) Prognostic nomogram for 1-year, 3-year, and 5-year overall survival of patients with colon cancer. (B, F, J) The restricted mean survival time (RMST) between the high-risk group and the low-risk group in the combined database, TCGA and GEO. (C, G, K) Concordance index (c-index) plots of TNM stage model, Clinic model, Risk score model and Composite model at different time points in the in the combined, TCGA and GEO databases respectively. (D, H, L) The calibration curve of the observation and prediction probabilities of the nomograms in the combined, TCGA and GEO databases respectively. (E, I, M) Decision curve analysis plots of TNM stage model, Clinic model and Composite model in combined, TCGA and GEO databases respectively. TNM stage model: involved in TNM stage only; Clinic model: involved in age, sex and TNM stage; Risk score model: involved in risk score only; Composite model: involved in risk score, TNM stage, age, and sex.

Figure 4.
Figure 4.

Functional enrichment analysis of RNA processing-based risk score-related genes. (A) Mountain map showed GSEA analysis results of genes associated with risk score. (B) The construction of a clustering dendrogram of the top 5000 most variable genes by an adjacency matrix. (C) Module-clinical traits relationship. Each column showed a module characteristic gene; each column corresponds to a clinical trait. Each cell contained the corresponding correlation (upper number) and P-value (lower number). (D) GO and KEGG functional enrichment analysis of genes in the brown module. (E) GO and KEGG functional enrichment analysis of genes in the turquoise module.

Figure 5.
Figure 5.

Expression and clinical features of RNA processing-related prognostic features. (A) The expression patterns of 10 prognostic-related RNA processing genes in the entire 1033 colon cancer samples shown by the heatmap. (B) The differential expression of 10 prognostic-related RNA processing genes between the low-risk group and the high-risk group. The P value was obtained by Mann-Whitney test. (C) The distribution of TNM stages and 4 different RNA processing modes between the low-risk group and the high-risk group shown by the histogram. Abbreviation: ns, no statistical significance *P < .05, **P < .01, ***P < 0.001.

Figure 6.
Figure 6.

Genetic variation in prognostic characteristics related to RNA processing gene. (A) Violin chart of the tumor mutation burden (TMB) between low-risk and high-risk groups. (B) Scatter plot of correlation analysis between tumor mutational burden risk scores. (C) Kaplan-Meier survival curve analysis combining high and low tumor burden mutations and high and low risk scores. (D) Kaplan-Meier curves analysis for low and high TMB groups of the TCGA database. (E) TMB analysis in the low-risk group. (F) TMB analysis in the high-risk group.

Figure 7.
Figure 7.

Immune heterogeneity in prognostic characteristics related to RNA processing gene. (A) Heat map showing immune cell infiltration between the high-risk and low-risk groups. (B) The differential expression of stromal score, immune score and ESTIMATE score between low-risk and high-risk groups. (C) Heat map showing immune cell infiltration between the high-risk and low-risk groups. The P value was obtained by the Mann Whitney test. Abbreviation: ns, non-statistical. *P < .05; **P < .01; ***P < .001.

Similar articles

References

    1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209-249. - PubMed
    1. Zhang YY, Chen SW, Wang PY, Liu YC. Research progress of conversion therapy in colorectal cancer liver metastases. Zhonghua Wei Chang Wai Ke Za Zhi. 2021;24:85-93. - PubMed
    1. Obeng EA, Stewart C, Abdel-Wahab O. Altered RNA processing in cancer pathogenesis and therapy. Cancer Discov. 2019;9:1493-1510. - PMC - PubMed
    1. Manning KS, Cooper TA. The roles of RNA processing in translating genotype to phenotype. Nat Rev Mol Cell Biol. 2017;18:102-114. - PMC - PubMed
    1. Tollervey D, Caceres JF. RNA processing marches on. Cell. 2000;103:703-709. - PubMed

LinkOut - more resources