Systematic Characterization of DNA Methyltransferases Family in Tumor Progression and Antitumor Immunity

Abstract

Objective: DNA methylation is an essential epigenetic marker governed by DNA methyltransferases (DNMTs), which can influence cancer onset and progression. However, few studies have provided an integrated analysis of the relevance of DNMT family genes to cell stemness, the tumor microenvironment (TME), and immunotherapy biomarkers across diverse cancers. Methods: This study investigated the impact of five DNMTs on transcriptional profiles, prognosis, and their association with Ki67 expression, epithelial-mesenchymal transition signatures, stemness scores, the TME, and immunological markers across 31 cancer types from recognized public databases. Results: The results indicated that DNMT1/DNMT3B/DNMT3A expression increased, whereas TRDMT1/DNMT3L expression decreased in most cancer types. DNMT family genes were identified as prognostic risk factors for numerous cancers, as well as being prominently associated with immune, stromal, and ESTIMATE scores, as well as with immune-infiltrating cell levels. Expression of the well-known immune checkpoints, PDCD1 and CILA4, was noticeably related to DNMT1/DNMT3A/DNMT3B expression. Finally, we validated the role of DNMT1 in MCF-7 and HepG2-C3A cell lines through its knockdown, whereafter a decrease in cell proliferation and migration ability in vitro was observed. Conclusion: Our study comprehensively expounded that DNMT family genes not only behave as promising prognostic factors but also have the potential to serve as therapeutic targets in cancer immunotherapy for various types of cancer.

Keywords: DNMT family; immunomodulators; prognosis; tumor microenvironment; various cancers.

Figure 1.

Expression patterns of DNMTs in distinct cancer types. (A) Gene expression in diverse cancers. (B) DNMT family gene expression levels among 23 different cancer types from the TCGA database. The gradation of color indicates high or low expression based on the log2 fold change (FC) value. (C-F) Differential expression of DNMT1, DNMT3A, DNMT3B, and TRDMT1 between tumor and normal tissues in diverse cancers. The red box represents the expression level of tumor tissues and the blue box represents that of normal tissues. (G and H) Protein expression of DNMT1 in BrCa and liver cancer tissues compared to that in paracancerous tissues via IHC staining. *P < .05; **P < .01; ***P < .001.

Figure 2.

Univariate Cox regression model for the prognostic role of DNMTs in various cancers. (A-D) Correlations between DNMT1, DNMT3A, DNMT3B, and TRDMT1 expression in OS, PFS, DFS, and DSS are summarized in the heat map by Cox regression analysis of multiple cancers. (E-H) Forest maps showing the prognostic value of DNMTs for OS in TCGA cancers via the univariate Cox proportional hazard regression model. A hazard ratio greater than 1 (HR > 1) is marked in red and implies a risk factors for patient prognosis. Hazard ratios below 1 (HR < 1) are marked in blue and implies a protective factor for patient prognosis. White color is considered statistically significant. OS, overall survival; PFS, progression-free survival time; DFS, disease-free survival time; DSS, disease-specific survival time.

Figure 3.

Association of DNMTs expression with proliferation and the epithelial–mesenchymal transition phenotypes among diverse cancer types. (A-D) Spearman's correlation analysis was employed to analyze the relationship of DNMT1, DNMT3A, DNMT3B, and TRDMT1 with Ki67 expression across different TCGA cancer types. FDR_level and rho_level denotes significant differences and the Spearman's rank correlation coefficient, respectively. (E-H) The heat map shows the relationship between DNMT1, DNMT3A, DNMT3B, and TRDMT1 expression and the EMT score in 31 cancer types. Orange color represents a positive correlation and blue color represents a negative correlation. *P < .05; **P < .01.

Figure 4.

Correlation of DNMT family genes expression with cell stemness. (A, B) Relationship between DNMT expression and stemness indices, including the mRNA expression-based stemness index (mRNAsi) and DNA methylation-based stemness index (mDNAsi), in 31 TCGA cancers. Red squares indicate a positive correlation and blue squares a negative correlation. *P < .05; **P < .01.

Figure 5.

Association analysis between DNMT family genes expression and immune, stromal, and ESTIMATE scores, as well as with tumor purity, in different cancers.

Figure 6.

Relationship analysis of DNMT1 expression with the tumor-infiltrating levels of immune cells in various cancer types. Tumor-infiltrating immune cells: B cells, T cells (CD4⁺ and CD8⁺), dendritic cells (DCs), endothelial (Endo) and eosinophil (Eos) cells, cancer-associated fibroblasts (CAFs), hematopoietic stem cells (HSCs), progenitors of lymphoid/myeloid/monocyte (progenitor) cells, macrophages, mast cells, monocytes, neutrophils, natural killer (NK) and T cell follicular helper (Tfh) cells, natural killer T cells (NKTs), regulatory T cells (Tregs), gamma delta T cells (γδTs), and myeloid-derived suppressor cells (MDSCs). A positive correlation is represented in red and a negative correlation represented in blue.

Figure 7.

Relationship of DNMT1/DNMT3A/DNMT3B expression to immune-related modulators, MSI, and TMB in pan-cancer. (A-D) Relationships between DNMT1/DNMT3A/DNMT3B expression and MHC (A), immunoinhibitors (B), immunostimulators (C), and immunocheckpoints (D) based the TCGA database. Red color represents a positive correlation and blue a negative correlation. (E-F) Correlation analysis of DNMT1/DNMT3A/DNMT3B expression with MSI and TMB in distinct cancers. MSI, microsatellite instability; TMB, tumor mutation burden. *P < .05; **P < .01.

Figure 8.

Proliferative and behavioral roles of DNMT1 in cancer cells. (A) Knockdown efficiency of virus-infected shDNMT1 measured in MCF-7 and HepG2-C3A cells via western blotting. (B) Colony formation ability of MCF-7 and HepG2-C3A was determined after knocking down DNMT1 expression, and the colony numbers thereafter counted. (C) MCF-7 cells infected with shDNMT1-virus were permitted to migrate for 30 h, and the wound healing area thereafter measured and counted. Bar = 10 μm. Objective lens magnification = ×20. (D) A Boyden chamber assay was used to analyze the effect of knocking down DNMT1 expression on cell migration for 48 h, and the number of migratory cells per field microscopically counted. Bar = 10 μm. Objective lens magnification = ×20. (E, F) A cell viability assay was applied to analyze the effect of knocking down DNMT1 expression on cell proliferation. *P < .05; **P < .01, ****P < .0001.