Decoding epigenetic and transcriptional landscapes: DNA methylome-transcriptome integration reveals novel drivers in 4NQO-Induced esophageal squamous cell carcinoma mouse model

Abstract

Background: Epigenetic alterations, particularly DNA methylation, and dysregulation of the tumor immune microenvironment (TIME) are increasingly recognized as critical factors in esophageal squamous cell carcinoma (ESCC) pathogenesis. Understanding the dynamic interplay between DNA methylation changes and TIME evolution during ESCC progression remains essential. We established a 4-nitroquinoline 1-oxide (4NQO)-induced ESCC mouse model, capturing distinct pathological stages: normal esophageal epithelium (Normal), esophageal simple hyperplasia (ESSH), intraepithelial neoplasia (IEN), and ESCC. Genome-wide DNA methylation profiling was performed using the Infinium Mouse Methylation BeadChip (285 K), coupled with transcriptome analysis via bulk RNA sequencing (RNA-seq). Immunohistochemistry (IHC) for Cd45 (leukocyte common antigen) validated immune cell infiltration.

Results: DNA methylation profiling revealed progressive genome-wide hypomethylation during ESCC development, with hypomethylated probes significantly enriched in immune response pathways. Notably, ESSH exhibited a methylation profile similar to IEN and ESCC. RNA-seq identified escalating numbers of differentially expressed genes (DEGs). Immune deconvolution analysis and IHC of Cd45 demonstrated dynamic changes in TIME composition from ESSH onwards. Furthermore, dynamic expression clustering identified an innate immune response-related gene cluster highly expressed in ESSH. Integrative analysis yielded 495 methylated regulatory genes, significantly enriched in leukocyte cell-cell adhesion and T cell activation pathways (e.g., Ptprc/Cd45, Il12rb1, Tox), with peak activity in ESSH.

Conclusions: These findings highlight ESSH as a critical window where epigenetically driven immune changes facilitate ESCC progression. Targeting these early epigenetic-immune interactions may offer a novel strategy for ESCC early detection and combination therapy.

Keywords: DNA methylation; Esophageal squamous cell carcinoma; RNA-seq; Tumor immune microenvironment.

Fig. 1

The design of this study and the differential methylation analysis. A The design of this study (n = 3 replicates each group for DNA methylation BeadChip; n = 3 replicates from 0w for Bulk-RNA sequencing; n = 4 replicates from 4w, 16w, and 28w for Bulk-RNA sequencing). B-D. The volcano plots show the number of differentially methylated probes in ESSH, IEN, and ESCC stages compared to normal group, respectively. E-G. The bar plots show the proportion of differentially hyper- and hypomethylated probes in different gene regions (Promoter, Body, IGV, and 3’UTR) (red represents hypermethylated probes, blue represents hypomethylated probes and grey represents no significant)

Fig. 2

Display of hypermethylated and hypomethylated probes. A Venn plot of differential hypermethylated probes among ESSH, IEN, and ESCC stages. B Heatmap of 3,487 hypermethylated probes among ESSH, IEN, and ESCC stages. C Venn plot of differential hypomethylated probes among ESSH, IEN, and ESCC stages. D Heatmap of 735 hypomethylated probes among ESSH, IEN, and ESCC stages. E Gene ontology enrichment of differentially hypomethylated genes. F Heatmap of 4,222 hypo- and hypermethylated probes among ESSH, IEN, and ESCC stages

Fig. 3

Disorder of immune microenvironment. A The immunohistochemistry staining of Cd45 in Normal, ESSH, IEN, and ESCC esophageal tissues. B Statistical analysis of Cd45-positive cells in Normal, ESSH, IEN, and ESCC stages. C Volcano plots show differentially expressed genes in ESSH, IEN, and ESCC stages, respectively. D Comparison boxplot of significant immune cells 13 out of 25 types of immune cell infiltration among Normal, ESSH, IEN, and ESCC stages (*, P < 0.05; **, P < 0.01; ***, P < 0.001)

Fig. 4

Bulk-RNA sequencing identified core gene clusters. A-B. Mfuzz-based dynamic expression clustering and heatmap of identified five different gene clusters (“N” stands for Normal, “H” stands for ESSH, “D” stands for IEN, “C” stands for ESCC). C-D. Gene ontology annotation analysis of C4 and C5 gene clusters

Fig. 5

Integrated analysis of DNA methylation BeadChip and Bulk-RNA sequencing. A Diagram of genomic regions and CpG sites (red represents hypermethylated probes and blue represents hypomethylated probes). B-C The Methylation and Expression Quadrant Plot shows log₂ fold changes (FC) in mRNA expression (y-axis) against Δβ in Methylation BeadChip (x-axis). The dotted lines indicate a cutoff with log₂FC > 1 or < − 1, and Δβ > 0.1 or < − 0.1. The dots in the left panel are located in promoter region; the dots in the right panel are located in body region. Q1 (n = 269) and Q3 (n = 18) represent DMPs in gene body positive regulating gene expression, Q2 (n = 117) and Q4 (n = 91) represent DMPs in promoter negative regulating gene expression. D GO enrichment of 495 genes from Quadrant Plot. E The FPKM value of 33 related genes from the top 3 biological processes in Normal, ESSH, IEN, and ESCC stages

Fig. 6

The diagram of DNA methylation changes and disorder of tumor immune microenvironment during the development of ESCC mouse model