The mechanism of RNA methylation writing protein-related prognostic genes in lung adenocarcinoma based on bioinformatics

Abstract

Objective: RNA methylation modifications play biological roles in tumorigenicity and immune response, mainly mediated by the "writer" enzyme. Lung adenocarcinoma (LUAD) development is closely related to RNA methylation. Here, the prognostic values of the "writer" enzymes and the tumor immunosurveillance in LUAD aim to provide new theoretical references for the research of LUAD.

Methods: Genes associated with RNA methylation writer protein in LUAD were identified using The Cancer Genome Atlas Program (TCGA) data and weighted gene co-expression network analysis (WGCNA). Independent prognostic factors were screened by Cox regression and least absolute shrinkage and selection operator (LASSO) regression analyses. A prognostic risk model and a nomogram were established using these genes. Moreover, Gene Set Enrichment Analysis (GSEA) and CIBERSORTx were used to analyze the immune cell infiltration and enrichment pathways in the low- and high-risk groups, respectively. In addition, genes' potential functions and regulatory mechanisms were explored through gene-gene interaction (GGI) networks and competing endogenous RNA (ceRNA) networks.

Results: We selected 202 genes associated with RNA methylation writer proteins, from which we identified the three genes (CLEC3B, GRIA1, and ANOS1). A prognostic risk model was constructed based on genes associated with RNA methylation writer proteins and stage, demonstrating reliable predictive performance. GGI analysis revealed GRIA1 as a crucial gene. Enrichment analysis revealed that the high-risk group had upregulated pathways connected to cell division. Additionally, immune infiltration analysis revealed that the significantly higher levels of NK cells, activated mast cells, activated CD4 memory cells, and M0 and M1 macrophages displayed in the high-risk group, while the significantly lower levels of monocytes, dendritic cells, M2 macrophages, and inactive CD4 memory cells were in the low-risk group. Moreover, Spearman correlation analysis demonstrated that the three prognostic genes and risk scores correlated highly with various immune cells.

Conclusion: This study identified three prognostic genes related to RNA methylation writer proteins in LUAD. A reliable prognostic model was constructed. The identified prognostic genes also play significant roles in immune cell infiltration in LUAD. This study provides new theoretical references for subsequent in-depth research on LUAD.

Keywords: RNA methylation writer proteins; biomarker; immune cell infiltration; lung adenocarcinoma; prognostic risk mode.

FIGURE 1

Identified RNA methylation writer proteins-related genes. (A) Volcano plot of DEGs. Each dot represents a gene, red and blue dots represent upregulated and downregulated genes, respectively. There were a total of 4,951 DEGs, among which 3,156 were upregulated and 1795 were downregulated. (B,C) Heatmap of the four gene modules most correlated with RNA methylation writer proteins. (D) Venn diagram showing the 202 candidate genes related to RNA methylation writer proteins. (E) GO analysis of candidate genes. The size of the box indicates the number of genes included, and the color indicates significance. (F) KEGG analysis of candidate genes. The circle represents the enriched KEGG pathway, and the outer circle is the gene enriched in that pathway. (G) PPI network revealing 61 gene nodes. DEGs, differentially expressed genes; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; PPI, Protein-Protein Interaction.

FIGURE 2

Establishment and validation of the prognostic risk model. (A) Forest plots displaying prognostic genes identified through univariate and multivariate Cox regression. Three prognostic genes, CLEC3B, GRIA1, and ANOS1, were identified. They were all protective factors for LUAD. The significance level is set at P < 0.05. (B) Risk scores and survival time distribution for patients with LUAD in the training and validation sets. The mortality rate of patients in the high-risk group was higher than that in the low-risk group. A circle represents a sample. (C, D) Kaplan-Meier survival curves for high- (training set: n = 255; validation set: n = 82) and low-risk (training set: n = 255; validation set: n = 99) groups in training and validation sets. The survival differences between the high-risk and low-risk groups were compared using the Log-rank test. The survival probabilities of the high-risk group were significantly lower than those of the low-risk group in both the training set (P = 0.00021) and the validation set (P < 0.0001). (E, F) ROC curve AUCs for 1-, 3-, and 5-year survival. LUAD, lung adenocarcinoma; AUC, Area Under Curve.

FIGURE 3

Prognostic risk assessment for patients with LUAD. (A,B) Forest plots illustrating HR from univariate and multivariate Cox analyses. Risk score and stage were demonstrated to be an independent prognostic factor. An HR more significant than one is considered a risk factor, while an HR less than one is considered a protective factor. The significance level is set at P < 0.05. (C) Nomogram for predicting 1-, 3-, and 5-year survival probabilities of patients with LUAD based on risk score and tumor stage. Each factor corresponds to a point, and the sum of the points of each factor corresponds to the total point. The higher the total point, the lower the patient’s survival rate. (D) Calibration curves for 1-, 3-, and 5-year survival predictions based on the nomogram. The slopes of the calibration curves were all close to 1. (E) Decision curve analysis for 1-, 3-, and 5-year survival. The curve corresponding to the nomogram exceeded the “All” and “None” baselines, and the net benefit of the nomogram was higher than that of a single independent prognostic factor. HR, hazard ratio.

FIGURE 4

GSEA revealed key pathways in the high-risk and low-risk groups. (A,B) Functional enrichment pathways for different risk groups based on GO and KEGG gene sets. A negative enrichment score indicates the downregulation of the pathway, while a positive enrichment score represents the upregulation of the pathway. GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

FIGURE 5

Immune cell infiltration in different risk groups. (A) Immune cell infiltration of 21 immune cell types in patients with LUAD is categorized into low-risk and high-risk groups. The Wilcoxon test was used to compare the differences in immune cell infiltration between the high-risk and low-risk groups. 11 types of immune cells showed differences between the two groups (P < 0.05). (B) Heatmap showing correlations between prognostic genes, immune cell enrichment scores, and risk scores based on Spearman correlation analysis. The redder the color, the stronger the positive correlation; the bluer the color, the stronger the negative correlation. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns. LUAD, lung adenocarcinoma; ns, no significance.

FIGURE 6

Potential functions and regulatory networks of prognostic genes. The GGI network highlights the top 20 genes (such as CACNG2, GRID2, and CNIH2) and the related pathways of prognostic genes (such as neurotransmitter receptor complex, ionotropic glutamate receptor complex, and plasma membrane signaling receptor complex). The center circle represents the prognostic gene, and the surrounding circle represents the gene associated with the prognostic gene. The color in each circle represents the signaling pathway associated with the gene. The line between the circles represents the interaction between the two genes, and the color of the line represents the interaction pattern between the two genes. GGI, gene-gene interaction.

FIGURE 7

Potential expression regulation mode of prognostic genes. The ceRNA regulatory network analysis identified numerous potential regulatory mechanisms involving 22 miRNAs for GRIA1, 6 for CLEC3B, and 23 for ANOS1, interacting with 463 associated lncRNAs. Red is a gene, green is miRNA, and blue is lncRNA.