Decitabine promotes degradation of DNMT1 and EZH2 via the ubiquitination pathway and inhibits colorectal cancer progression Authors: Xiao-Mei Peng, Xin-Peng Shi, Han Chen, Lu-Yang Cao, Hao-Jian Zuo, Jie-Qiong Guo, Nan Jiang, Xiao-Yong Luo Published: 2025 ABSTRACT PurposeThis study aimed to investigate the role of DNMT1 in CRC progression and its regulatory relationship with TRAF6 and EZH2.MethodsDNMT1 expression was analyzed in CRC tissues and cell lines using public databases and experimental techniques, including Western blot and qRT-PCR. Functional assays, such as colony formation, transwell migration/invasion, and cell cycle analysis, were performed to assess the role of DNMT1 in CRC cell proliferation and metastasis. Mechanistic studies, including cycloheximide (CHX) chase assays, ubiquitination assays, and co-immunoprecipitation (Co-IP), were conducted to explore the regulation of DNMT1 stability and its effects on EZH2 protein stability. The regulatory axis was further validated using methylation-specific PCR (MSP), dual-luciferase assays, and immunohistochemistry (IHC) in CRC patient tissues.ResultsDNMT1 was significantly overexpressed in CRC tissues and cell lines, correlating with enhanced cell proliferation, migration, and invasion Mechanistically, DNMT1 is associated with CRC cell proliferation and regulate this process via upregulating cyclins D1/E2 and accelerating G1/S phase transition. Decitabine, a DNA methyltransferase inhibitor, induced DNMT1 degradation via the ubiquitin-proteasome pathway, with TRAF6 identified as a key E3 ubiquitin ligase mediating this process. TRAF6 was downregulated in CRC tissues and inversely correlated with DNMT1 expression. DNMT1 suppressed TRAF6 expression through promoter hypermethylation, forming a negative feedback loop. Additionally, DNMT1 stabilized EZH2 by inhibiting TRAF6-mediated ubiquitination, thereby enhancing EZH2-dependent oncogenic signaling. Functional experiments demonstrated that EZH2 was essential for DNMT1-mediated CRC progression.ConclusionThis study reveals a novel Decitabine-TRAF6-DNMT1-TRAF6-EZH2 regulatory axis in CRC. Decitabine has dual effects, suggesting a new therapy.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13402-025-01136-8. Introduction Colorectal cancer (CRC) represents a major global health burden, accounting for approximately 10% of cancer-related incidence and mortality worldwide [1]. Its pathogenesis involves progressive accumulation of genetic and epigenetic alterations, with DNA hypermethylation of tumor suppressor gene promoters emerging as a key driver of malignant transformation [2–5]. Critical genes silenced through this mechanism include p16INK4A, MLH1, and APC [6–10], a process mediated by DNA methyltransferases (DNMTs). DNMT1 is essential for maintaining DNA methylation patterns in multiple cancers and its dysregulation has been linked to tumorigenesis in various cancers. Although DNMT1 inhibitors like Decitabine (DAC) have shown promise in treating cancers such as acute myeloid leukemia and breast cancer [11–13], the specific role of DNMT1 in CRC remains poorly understood, necessitating further investigation into its molecular mechanisms and therapeutic potential. In preclinical studies of colorectal cancer, decitabine has shown potential in inhibiting cell proliferation [14, 15]. Clinical trials have preliminarily explored the disease control effect of its single-agent or combination regimens [16], laying a foundation for further research. Moreover, High-throughput screening studies have shown that epigenetic drugs such as decitabine can synergize with some FDA-approved drugs (e.g., the antiarrhythmic drug scillaren A), enhancing therapeutic efficacy by regulating epigenetic modulators including colon cancer oncogenes, while there is also a risk that some drugs may antagonize their activity, providing a basis for exploring the mechanisms of decitabine-based combination therapy in colorectal cancer [17]. These additions aim to provide a more comprehensive foundation for our research on decitabine’s mechanism of action via the ubiquitination pathway in CRC. Beyond DNA methylation, Polycomb-mediated histone modification adds another layer of epigenetic dysregulation in CRC. EZH2, the catalytic subunit of PRC2, silences tumor suppressors through H3K27 trimethylation and is overexpressed in aggressive malignancies [18]. Though EZH2 upregulation correlates with CRC progression [19], the mechanisms driving its stability and functional crosstalk with DNMT1 are unknown. Notably, while DNMTs and Polycomb proteins often cooperate in epigenetic silencing [20], their reciprocal regulation in CRC remains not fully demonstrat. Here, we identify a novel Decitabine-TRAF6-DNMT1-TRAF6-EZH2 regulatory axis controlling CRC pathogenesis. We demonstrate that DNMT1 maintains TRAF6 promoter methylation to transcriptionally repress this E3 ubiquitin ligase. Surprisingly, TRAF6 reciprocally triggers DNMT1 ubiquitination and proteasomal degradation, establishing a bidirectional feedback loop. Furthermore, DNMT1 depletion promotes TRAF6-EZH2 complex formation, destabilizing EZH2 through post-translational modification. Pharmacological inhibition of EZH2 abolishes the pro-metastatic effects of DNMT1 overexpression, while combined targeting of both epigenetic regulators synergistically suppresses tumor growth. Importantly, decitabine exerts dual effects by destabilizing DNMT1 and EZH2 via ubiquitination pathways, providing mechanistic justification for its clinical repurposing in CRC. Our findings illuminate the dynamic interplay between DNA methylation and histone modification systems, proposing novel combinatorial epigenetic therapies for CRC. Materials and methods Cell culture Human colorectal cell lines (HCT8, SW480, SW620) were purchased from the Culture Collection of the Chinese Academy of Sciences (Shanghai, China). HCT8 were cultured in RAPI-1640(Gibco, MD, USA) and SW480/SW620 were cultured in DMEM (Gibco, MD, USA) supplemented with 10% fetal bovine serum (FBS; BI) and maintained at 37 °C in a 5% CO2 atmosphere. Cell transfection, lentivirus infection, decitabine treatment Small interfering RNAs (siRNAs) were purchased from Sangon Biotechnology, Shanghai, China. siRNA sequences are as follows: For Negative control: sense, 5’- UUCUCCGAACGUGUCACGUTT − 3’; antisense, 5’- ACGUGACACGUUCGGAGAATT − 3’; For anti-DNMT1: sense, 5’- GCGGCUCAAAGAUUUGGAATT-3’; antisense, 5’-UUCCAAAUCUUUGAGCCGCTT-3’; For anti-EZH2: sense, 5’- UAGGUUAAUUGGGACCAAATT-3’; antisense, 5’-UUUGGUCCCAAUUAACCUAGC-3’; For anti-TRAF6: sense, 5’-UUCUCCGAACGUGUCACGUUU-3’; antisense, 5’-ACGUGACACGUUCGGAGAAUU-3’. Cells were grown to 70–80% confluence prior to siRNA transfection using Lipofectamine™ RNAiMAX (Thermo Scientific, 13778150) according to the manufacturer’s protocol. For TRAF6 or EZH2 overexpression, we transfected TRAF6 or EZH2 plasmids using Lipofectamine 3000(Thermo Scientific, L3000008). ShCyclinD1 and shCyclinE2 were also transfected using Lipo3000. We purchased the DNMT1-expressing lentivirus (Genechem). The viruses were used to infect HCT8 cells, the stable transfection cell lines were selected with 10ug/ml puromycin for two weeks. Colony formation assay HCT8 and SW480 cells were seeded in six-well and then cultured in medium supplemented with 10% FBS and 1% antibiotics (100 U/mL penicillin and 100 mg/mL streptomycin) at 37 °C with 5% CO2,1000 cells per well. After two weeks, cells were fixed with 4% paraformaldehyde and stained with 0.1% (m/v) crystal violet solution at room temperature. Cell cycle analysis by flow cytometry Cells were harvested at the indicated time points, washed twice with cold phosphate-buffered saline (PBS), and fixed in 70% ethanol at 4 °C overnight. After fixation, cells were centrifuged at 1000×g for 5 min, and the ethanol was carefully removed. The cell pellet was resuspended in PBS containing 50 µg/mL propidium iodide (PI; Sigma-Aldrich) and 100 µg/mL RNase A (Thermo Fisher Scientific), then incubated at 37 °C for 30 min in the dark. Cell cycle distribution was analyzed using a NovoCyte Advanteon DX. A minimum of 10,000 events per sample were collected. Data were processed with NovoExpress software to determine the percentage of cells in G1, S, and G2 phases. Transwell Experiments for determination of cell invasion, migration were performed as described previously [21]. Western blot The total protein of CRC cells and tissues were extracted using RIPA lysis buffer with phenylmethylsulfonyl fluoride (PMSF) (Beyotime, Biotechnology,). Cell lysates were centrifuged at a speed of 12,000 rpm for 20 min, then mixed with 5×loading buffer (Beyotime, Biotechnology, P0015L) and denatured at 100 °C for 10 min. Supernatant protein (20 µg) were separated by Pre - cast gels for Western Blot (WB) and electrophoretically transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore,0.45 μm). Subsequently, the PVDF membranes were subjected to blocking for 1 h by using 5% (w/v) non-fat milk at room temperature. After incubation with anti-DNMT1 (Abcam, ab19905), anti-EZH2 (CST, 5246s), anti-TRAF6 (CST, 67591s), or anti-GAPDH (ProteinTech, 60004) overnight at 4 °C, the PVDF membranes were washed with TBST. he next day, the membranes were incubated at room temperature for 1 h with corresponding horseradish peroxidase (HRP)- conjugated secondary antibody (ProteinTech, SA00001-1) or (ProteinTech, SA00001-2). The immunoreactive bands were made visible in enhanced chemiluminescence (ECL) reaction solution (Thermo Scientific, 32209). IP and Co-IP assay The level of protein ubiquitination modification can be detected by immunoprecipitation (IP) technology. The interaction between endogenous TRAF6 and DNMT1 protein or TRAF6 and EZH2 protein was detected by co-IP. For the IP assay, incubate with anti-EZH2 antibody or anti-DNMT1 antibody according to the manufacturer’s instructions of the kit (Thermo Scientific, 88805). The eluted proteins were then subjected to analysis using western blotting with Ub antibodies ProteinTech, 10201-AP)to identify the level of ubiquitination. For the Co-IP assay, the eluted proteins were then subjected to analysis using western blotting with anti-TRAF6, anti-EZH2 antibody or anti-DNMT1. QRT-PCR Total RNA was extracted using FastPure Cell/Tissue Total RNA Isolation Kit V2(Vazyme, RC112). A total of 1 µg RNA was reversed-transcribed to cDNA with FastKing RT Kit (with gDNase) (KR116). Real-time PCR was performed using the TB Green® Premix Ex Taq™ II (Takara, RR820A). Primer sequences for DNMT1, TRAF6, EZH2 and GAPDH were designed and provided by Shanghai Sangon Biotechnology (Shanghai, Table 1). Table 1Primer sequences for qRT-PCRTarget genePrimer sequence (5′→3′)DNMT1Forward primer: CCGAGCGAGCCAGAGATAGAGReverse primer: GAGATGCCTGCTTGGTGGAATCEZH2Forward primer: TTGTTGGCGGAAGCGTGTAAAATCReverse primer: TCCCTAGTCCCGCGCAATGAGCTRAF6Forward primer: ATGCGGCCATAGGTTCTGCReverse primer: TCCTCAAGATGTCTCAGTTCCATGAPDHForward primer: GGAGCGAGATCCCTCCAAAATReverse primer: GGCTGTTGTCATACTTCTCATGG Table 1Primer sequences for qRT-PCRTarget genePrimer sequence (5′→3′)DNMT1Forward primer: CCGAGCGAGCCAGAGATAGAGReverse primer: GAGATGCCTGCTTGGTGGAATCEZH2Forward primer: TTGTTGGCGGAAGCGTGTAAAATCReverse primer: TCCCTAGTCCCGCGCAATGAGCTRAF6Forward primer: ATGCGGCCATAGGTTCTGCReverse primer: TCCTCAAGATGTCTCAGTTCCATGAPDHForward primer: GGAGCGAGATCCCTCCAAAATReverse primer: GGCTGTTGTCATACTTCTCATGG Primer sequences for qRT-PCR Forward primer: CCGAGCGAGCCAGAGATAGAG Reverse primer: GAGATGCCTGCTTGGTGGAATC Forward primer: TTGTTGGCGGAAGCGTGTAAAATC Reverse primer: TCCCTAGTCCCGCGCAATGAGC Forward primer: ATGCGGCCATAGGTTCTGC Reverse primer: TCCTCAAGATGTCTCAGTTCCAT Forward primer: GGAGCGAGATCCCTCCAAAAT Reverse primer: GGCTGTTGTCATACTTCTCATGG Immunofluorescence Cells were seeded on glass cover-slips in 24-well plates, left overnight before the treatment. Cells were fixed with phosphate-buffered saline (PBS) containing 4% paraformaldehyde for 30 min and permeabilized with 0.5%Triton X-100 in PBS for 10 min at room temperature. Cells were washed three in PBS and blocked for 1 h with 5% BSA in PBS and then incubated with primary antibodies DNMT1 (abcam, ab188453) and TRAF6 (santa cruz, sc-8409) at 4 °C overnight. After washing with PBS, cells were incubated with secondary antibodies Alexa Fluor-488 (ab150077) and Alexa Fluor-594 (abcam, ab150116) for 1 h, washed three times in PBS and cell nuclei were counterstained with DAPI (Beyotime, C1006). Finally, take pictures under a fluorescence microscope (leica). Dual luciferase reporter gene assay Luc-TRAF6 promoter plasmid, DNMT1 overexpression plasmid and Renlilla were purchased from Genechem. HEK293T cells were maintained in DMEM supplemented with 10% FBS at 37 °C under 5% CO₂. At 80% confluency, cells were transfected with 500 ng of TRAF6-reporter plasmid and 50 ng of Renlilla plasmid using Lipofectamine 3000 following the manufacturer’s protocol. For experimental groups, DNMT1 overexpression plasmid were co-transfected. 48 h post-transfection, cells were lysed with 1× Passive Lysis Buffer (Promega, USA) for 15 min at room temperature. 11 µL of lysate was mixed with 70 µL of Luciferase Assay Reagent II (LAR II). Luminescence was immediately measured using a Microplate Luminometer. 70 µL of Stop & Glo Reagent (Promega, USA) was added to quench firefly luciferase activity and simultaneously activate Renilla luciferase. Luminescence was recorded within 2 s. Methylation specific PCR Genomic DNA was isolated from SW480 cell lines using the QIAamp DNA Mini Kit (Qiagen, Germany) according to the manufacturer’s protocol. Bisulfite conversion of genomic DNA (500 ng) was performed using the DNA Bisulfite Conversion Kit (TIANGEN, DP215). PCR amplification was performed using the Methylation-Specific PCR (MSP) Kit. TRAF6 promoter M primer Forward: TATTTATTGATATATTCGGGTTCGG, TRAF6 promoter M primer Reverse: GAAAAAAATACCCCCTAACGAT, TRAF6 promoter U primer Forward: TATTTATTGATATATTTGGGTTTGG, TRAF6 promoter U primer Reverse: TCAAAAAAAATACCCCCTAACAAT. Immunohistochemistry Paraffin sections (4 μm) of colorectal cancer tissues were used for immunohistochemical detection of the expression of DNMT1, TRAF6, and EZH2. After dewaxing with xylene and rehydration with graded ethanol, antigen retrieval was performed using citrate buffer (pH 6.0) to expose the target proteins. Endogenous peroxidase activity was blocked by incubating the sections with 3% hydrogen peroxide at room temperature for 10 min. Subsequently, the sections were blocked with 5% BSA for 30 min, followed by the addition of the corresponding primary antibodies (DNMT1, TRAF6, EZH2, diluted at 1:100) and incubation at 4 °C overnight. The next day, HRP-conjugated secondary antibodies were added, and the sections were incubated at room temperature for 1 h. DAB chromogenic reagent was used for color development, followed by hematoxylin counterstaining. After dehydration and clearing with graded ethanol and xylene, the sections were mounted with neutral balsam. Statistical analysis Data are presented as mean ± standard deviation (SD) at least three biological experiments. Two Groups comparisons were analyzed by Student’s t-test. Statistical significance was set at p < 0.05. Statistical analyses were performed using the GraphPad Prism software. Results DNMT1 overexpression drives CRC cell proliferation and metastasis Analysis of the EGO database revealed that the expression of the DNMT1 gene was significantly higher in colorectal cancer (CRC) tissues compared to adjacent normal tissues (Fig. 1A). This finding was further validated by western blot analysis in CRC cell lines (HCT8 and SW480) and normal intestinal epithelial cells, which also showed markedly elevated DNMT1 levels in CRC cells (Fig. 1B, C). Consistent with these observations, DNMT1 expression was significantly upregulated in clinical CRC specimens compared to adjacent non-tumor tissues (Fig. S1). To investigate the functional role of DNMT1 in CRC cells, we established stable DNMT1-overexpressing HCT8 and SW480 cell lines and performed siRNA-mediated knockdown of DNMT1. The efficiency of overexpression and knockdown was confirmed at both mRNA and protein levels (Fig.S2A, B). Colony formation assays demonstrated that ectopic expression of DNMT1 significantly enhanced CRC cell proliferation (Fig. 1D). Conversely, knockdown of DNMT1 or treatment with the DNA methyltransferase inhibitor DAC markedly reduced the proliferative capacity of HCT8 and SW480 cells (Fig. 1D, E). Flow cytometry analysis revealed that DNMT1 overexpression significantly decreased the proportion of CRC cells in G0/G1 phase and increased the percentage in S and G2/M phases relative to controls, indicating accelerated cell cycle progression. Conversely, DNMT1 knockdown resulted in G0/G1 phase arrest and a reduction of cells in S and G2/M phases (Fig. 1F, G). Consistent with these observations, protein levels of cyclin D1 and cyclin E2 were elevated in DNMT1-overexpressing cells (Fig. 1H, I), whereas their expression was reduced following DNMT1 knockdown or DAC treatment (Fig. 1H–J). Rescue experiments showed that silencing cyclin D1/cyclin E2 partially rescued DNMT1-overexpression-induced proliferation in colorectal cancer cells(Fig. S3A, B). These findings collectively suggest that DNMT1 plays a critical role in promoting CRC cell proliferation in vitro. Fig. 1DNMT1 promotes CRC cell proliferation through cyclin-dependent pathways. A. DNMT1 mRNA expression in CRC tissues versus adjacent normal tissues from TCGA database (***p < 0.001). B, C. DNMT1 protein levels in CRC cell lines (SW480, HCT8) and normal colon epithelial cells (NCM460) detected by Western blot (*p < 0.05,**p < 0.01). D, E. Colony formation assays showing proliferation changes in DNMT1-overexpressing and DNMT1-knockdown/DAC-treated CRC cells (*p < 0.05,**p < 0.01, ***p < 0.001). F, G. DNMT1 overexpression promotes cell cycle progression by reducing G0/G1 phase and increasing S/G2/M phases, while knockdown induces G0/G1 arrest(**p < 0.01). H-J. Cell cycle protein levels of cyclinD1/cyclinE2 in DNMT1-modulated cells DNMT1 promotes CRC cell proliferation through cyclin-dependent pathways. A. DNMT1 mRNA expression in CRC tissues versus adjacent normal tissues from TCGA database (***p < 0.001). B, C. DNMT1 protein levels in CRC cell lines (SW480, HCT8) and normal colon epithelial cells (NCM460) detected by Western blot (*p < 0.05,**p < 0.01). D, E. Colony formation assays showing proliferation changes in DNMT1-overexpressing and DNMT1-knockdown/DAC-treated CRC cells (*p < 0.05,**p < 0.01, ***p < 0.001). F, G. DNMT1 overexpression promotes cell cycle progression by reducing G0/G1 phase and increasing S/G2/M phases, while knockdown induces G0/G1 arrest(**p < 0.01). H-J. Cell cycle protein levels of cyclinD1/cyclinE2 in DNMT1-modulated cells Next, we explored whether DNMT1 influences CRC cell motility. Transwell assays demonstrated that DNMT1 knockdown or treatment with DAC significantly impaired the migration and invasion capabilities of CRC cells (Fig. 2A-H). Conversely, ectopic expression of DNMT1 enhanced the migratory and invasive properties of these cells (Fig. 2A, B, E, F). These results collectively indicate that DNMT1 plays a critical role in promoting CRC cell migration and invasion in vitro. Fig. 2DNMT1 drives CRC cell migration and invasion in vitro. A-H. Suppressed migration and invasion in DNMT1-knockdown or DAC-treated cells (HCT8/SW480) (*p < 0.05,**p < 0.01, ***p < 0.001; scale bar: 100 μm). A-B, E-F. Transwell migration and Matrigel invasion assays in DNMT1-overexpressing cells (HCT8/SW480) (*p < 0.05, **p < 0.01 vs. vector control; scale bar: 100 μm) DNMT1 drives CRC cell migration and invasion in vitro. A-H. Suppressed migration and invasion in DNMT1-knockdown or DAC-treated cells (HCT8/SW480) (*p < 0.05,**p < 0.01, ***p < 0.001; scale bar: 100 μm). A-B, E-F. Transwell migration and Matrigel invasion assays in DNMT1-overexpressing cells (HCT8/SW480) (*p < 0.05, **p < 0.01 vs. vector control; scale bar: 100 μm) Decitabine targets DNMT1 stability via ubiquitin-proteasome pathway Following treatment with decitabine, the protein levels of DNMT1 were significantly reduced(Fig. 3A), while no significant changes were observed in its mRNA levels(Fig. 3B). To explore the mechanism underlying this degradation, HCT8 cells treated with decitabine were co-treated with the proteasome inhibitor MG132. The results showed that the protein levels of DNMT1 were restored(Fig. 3C). Furthermore, cycloheximide (CHX) chase assays revealed that the half-life of DNMT1 was significantly shortened in DAC-treated HCT8 cells compared to untreated cells(Fig. 3D). To further investigate the mechanism, HCT8 cells were treated with 500 nM DAC, and immunoprecipitation (IP) assays demonstrated an increase in the ubiquitination levels of DNMT1 following DAC treatment(Fig. 3E). These findings suggest that decitabine-induced DNMT1 degradation is mediated through the proteasome pathway. Fig. 3Decitabine induces DNMT1 degradation via the ubiquitin-proteasome pathway. (A) Western blot showing DNMT1 protein downregulation in HCT8 cells treated with DAC (500 nM, 48 h). (B) qRT-PCR analysis of DNMT1 mRNA levels in DAC-treated cells (ns: not significant). (C) Rescue of DNMT1 protein levels by co-treatment with proteasome inhibitor MG132 (10 µM, 6 h) and DAC. (D) Cycloheximide (CHX, 50 µg/mL) chase assay showing shortened DNMT1 half-life in DAC-treated cells. (E) Immunoprecipitation (IP) of DNMT1 followed by anti-ubiquitin (Ub) Western blot demonstrating enhanced DNMT1 ubiquitination after DAC treatment (input controls: 5% lysate) Decitabine induces DNMT1 degradation via the ubiquitin-proteasome pathway. (A) Western blot showing DNMT1 protein downregulation in HCT8 cells treated with DAC (500 nM, 48 h). (B) qRT-PCR analysis of DNMT1 mRNA levels in DAC-treated cells (ns: not significant). (C) Rescue of DNMT1 protein levels by co-treatment with proteasome inhibitor MG132 (10 µM, 6 h) and DAC. (D) Cycloheximide (CHX, 50 µg/mL) chase assay showing shortened DNMT1 half-life in DAC-treated cells. (E) Immunoprecipitation (IP) of DNMT1 followed by anti-ubiquitin (Ub) Western blot demonstrating enhanced DNMT1 ubiquitination after DAC treatment (input controls: 5% lysate) Reciprocal Regulation of TRAF6 and DNMT1 in Colorectal Cancer: TRAF6 Targets DNMT1 for Ubiquitination While Being Epigenetically Repressed by DNMT1 Analysis of TCGA data revealed significantly reduced TRAF6 expression in colorectal cancer tissues. Kaplan-Meier survival analysis paradoxically demonstrated that patients with higher TRAF6 expression exhibited prolonged post-progression survival (PPS). Furthermore, a significant inverse correlation was identified between DNMT1 and TRAF6 expression levels (Fig. 4A-C). We identified TRAF6, an E3 ubiquitin ligase, as a key mediator of DAC-induced DNMT1 degradation. Immunofluorescence confirmed DNMT1-TRAF6 co-localization (Fig. 4D), and DAC treatment upregulated TRAF6 protein levels (Fig. 4E). TRAF6 knockdown significantly elevated DNMT1 protein levels(Fig. 4F) and reduced DNMT1 ubiquitination (Fig. 4G). Moreover, knockdown of TRAF6 rescues DNMT1 from DAC-mediated degradation(Fig. 4H). These demonstrating that TRAF6 is essential for DNMT1 destabilization. We also found that DNMT1 knockdown or DAC treatment or overexpression inversely correlated with TRAF6 levels at both protein(Fig. 4I, J) and mRNA levels (Fig. 4K-M). DNMT1 and TRAF6 mutually regulate each other through a negative feedback mechanism. Moreover, our methylation-specific PCR (MSP) (Fig. 4N), dual-luciferase assays(Fig. 4O), confirm that DNMT1 silences TRAF6 expression by increasing the methylation level of its promoter. Fig. 4TRAF6 mediates DAC-induced DNMT1 ubiquitination and degradation. (A) Reduced TRAF6 mRNA expression in colorectal cancer tissues compared to adjacent normal tissues (TCGA dataset; RNA-seq analysis; ****p < 0.0001). (B) Kaplan-Meier analysis reveals prolonged post-progression survival (PPS) in patients with high TRAF6 expression versus low expression (log-rank test, p = 0.0061). (C) Inverse correlation between DNMT1 and TRAF6 mRNA expression in TCGA colorectal cohorts (Spearman’s r = -0.1603, p = 0.0008061). (D) Immunofluorescence showing co-localization of DNMT1 (green) and TRAF6 (red) in HCT8 and SW620 (DAPI: blue; scale bar:100 μm). (E) TRAF6 protein upregulation in DAC-treated cells. (F) TRAF6 knockdown (siTRAF6) significantly elevated DNMT1 protein levels. (G) IP-Western blot showing reduced DNMT1 ubiquitination upon TRAF6 knockdown (anti-Ub antibody; IgG as IP negative control). (H) TRAF6 knockdown blocks DAC-induced DNMT1 degradation I, J. TRAF6 protein levels inversely correlate with DNMT1 modulation. K-M. qRT-PCR reveals TRAF6 transcriptional upregulation following DNMT1 knockdown (siDNMT1) or DAC treatment (500nM, 48 h), and suppression by DNMT1 OE(*p < 0.05,**p < 0.01, ***p < 0.001, ****p < 0.0001). N. DNMT1 induces hypermethylation of the TRAF6 promoter. Methylation-specific PCR (MSP) using methylated (M) and unmethylated (U) primers. DNMT1 OE enhances methylation (intensified M-band). O. Dual-luciferase assays show DNMT1 knockdown or treated with DAC increases TRAF6 promotor activity in HEK293T cells(**p < 0.01) TRAF6 mediates DAC-induced DNMT1 ubiquitination and degradation. (A) Reduced TRAF6 mRNA expression in colorectal cancer tissues compared to adjacent normal tissues (TCGA dataset; RNA-seq analysis; ****p < 0.0001). (B) Kaplan-Meier analysis reveals prolonged post-progression survival (PPS) in patients with high TRAF6 expression versus low expression (log-rank test, p = 0.0061). (C) Inverse correlation between DNMT1 and TRAF6 mRNA expression in TCGA colorectal cohorts (Spearman’s r = -0.1603, p = 0.0008061). (D) Immunofluorescence showing co-localization of DNMT1 (green) and TRAF6 (red) in HCT8 and SW620 (DAPI: blue; scale bar:100 μm). (E) TRAF6 protein upregulation in DAC-treated cells. (F) TRAF6 knockdown (siTRAF6) significantly elevated DNMT1 protein levels. (G) IP-Western blot showing reduced DNMT1 ubiquitination upon TRAF6 knockdown (anti-Ub antibody; IgG as IP negative control). (H) TRAF6 knockdown blocks DAC-induced DNMT1 degradation I, J. TRAF6 protein levels inversely correlate with DNMT1 modulation. K-M. qRT-PCR reveals TRAF6 transcriptional upregulation following DNMT1 knockdown (siDNMT1) or DAC treatment (500nM, 48 h), and suppression by DNMT1 OE(*p < 0.05,**p < 0.01, ***p < 0.001, ****p < 0.0001). N. DNMT1 induces hypermethylation of the TRAF6 promoter. Methylation-specific PCR (MSP) using methylated (M) and unmethylated (U) primers. DNMT1 OE enhances methylation (intensified M-band). O. Dual-luciferase assays show DNMT1 knockdown or treated with DAC increases TRAF6 promotor activity in HEK293T cells(**p < 0.01) DNMT1 suppresses TRAF6 to stabilize EZH2 in CRC Previous studies link EZH2 overexpression to CRC progression [22]. Emerging evidence from prostate cancer research suggests that DNMT1 regulates EZH2 protein stability through the ubiquitin-proteasome pathway [23]. Our findings demonstrate that DNMT1 overexpression significantly upregulates EZH2 protein levels, whereas DNMT1 knockdown or DAC treatment markedly reduces EZH2 protein expression in CRC cells(Fig. 5A-C). To gain insight into the mechanism by which DNMT1 alters EZH2 protein levels, we tested the presence of EZH2 mRNA in HCH8 and SW480 cells. Notably, DNMT1 did not regulate EZH2 transcription (Fig. 5D-G). Then, we performed protein stability assays in HCT8 cells using cycloheximide chase analysis (50 µg/mL). Western blot analysis revealed that DNMT1 knockdown significantly shortened the half-life of EZH2 protein (Fig. 5H, I). Furthermore, proteasome inhibition by MG132 treatment (10 µM, 6 h) effectively increased EZH2 protein levels in DNMT1-depleted cells (Fig. 5J), suggesting that DNMT1 regulates EZH2 stability through proteasome-mediated degradation. Endogenous ubiquitination assays confirmed that DNMT1 deficiency or treated with DAC increased EZH2 polyubiquitination (Fig. 5K, L), suggesting DNMT1 protects EZH2 from proteasomal degradation. Through the analysis of TCGA data, it was found that EZH2 expression was markedly elevated in colorectal cancer tissues (Fig. 5M). Moreover, TRAF6 was significantly negatively correlated with EZH2 (Fig. 5N). Research has demonstrated that TRAF6, an E3 ubiquitin-ligase, mediates the ubiquitination-dependent degradation of EZH2 [24–27]. Thus, we hypothesized whether DNMT1 is involved in TRAF6-mediated ubiquitination of EZH2. Critically, TRAF6 overexpression enhanced EZH2 ubiquitination, while TRAF6 knockdown reduced it (Fig. 5O, P). Co-immunoprecipitation (Co-IP) assays revealed that DNMT1 depletion robustly enhanced the interaction between EZH2 and TRAF6 (Fig. 5Q). Notably, dual inhibition of TRAF6 and DNMT1 rescued the DNMT1 knockdown-driven hyperubiquitination of EZH2(Fig. 5R). These results suggesting that DNMT1 silences TRAF6, thereby blocking TRAF6-mediated EZH2 ubiquitination and stabilizing oncogenic EZH2. Fig. 5DNMT1 silences TRAF6 to block EZH2 ubiquitination. A-C. DNMT1 modulation regulates EZH2 protein levels in CRC cells: Western blot analysis of EZH2 in DNMT1-overexpressing and DNMT1-knockdown(A, B), DAC-treated (C) cells. D-G. qRT-PCR analysis of EZH2 mRNA in HCT8 (D) and SW480 (E) cells with DNMT1 modulation (ns: not significant). H-I. Cycloheximide (CHX, 50 µg/mL) chase assay showing reduced EZH2 half-life in DNMT1-knockdown HCT8 cells. J. Proteasome inhibitor MG132 (10 µM, 6 h) rescues EZH2 protein levels in DNMT1-depleted cells. K-L. Endogenous ubiquitination assays showing increased polyubiquitinated EZH2 in DNMT1-knockdown or treated with DAC cells (IP with anti-EZH2 followed by anti-Ub Western blot; input: 5% lysate). M. TCGA data analysis reveals significantly elevated EZH2 mRNA expression in colorectal cancer (CRC) tissues compared to adjacent normal tissues. N. Expression correlation analysis demonstrates a significant inverse relationship between TRAF6 and EZH2 mRNA levels in colorectal cohorts (p < 0.0001). O, P. Overexpression of TRAF6 enhances polyubiquitination of EZH2, as detected by ubiquitin (Ub) immunoblotting in CRC cells. TRAF6 knockdown reduces EZH2 ubiquitination levels. Lysates were immunoprecipitated (IP) with anti-EZH2 and immunoblotted (IB) with anti-Ub. Q. Co-IP assay demonstrating strengthened EZH2-TRAF6 interaction in DNMT1-depleted cells (lysates immunoprecipitated with anti-EZH2, blotted for TRAF6; input: 5% lysate). R. Dual inhibition of TRAF6 and DNMT1 rescues EZH2 hyperubiquitination (IP-Western blot with siTRAF6/siDNMT1 co-treatment DNMT1 silences TRAF6 to block EZH2 ubiquitination. A-C. DNMT1 modulation regulates EZH2 protein levels in CRC cells: Western blot analysis of EZH2 in DNMT1-overexpressing and DNMT1-knockdown(A, B), DAC-treated (C) cells. D-G. qRT-PCR analysis of EZH2 mRNA in HCT8 (D) and SW480 (E) cells with DNMT1 modulation (ns: not significant). H-I. Cycloheximide (CHX, 50 µg/mL) chase assay showing reduced EZH2 half-life in DNMT1-knockdown HCT8 cells. J. Proteasome inhibitor MG132 (10 µM, 6 h) rescues EZH2 protein levels in DNMT1-depleted cells. K-L. Endogenous ubiquitination assays showing increased polyubiquitinated EZH2 in DNMT1-knockdown or treated with DAC cells (IP with anti-EZH2 followed by anti-Ub Western blot; input: 5% lysate). M. TCGA data analysis reveals significantly elevated EZH2 mRNA expression in colorectal cancer (CRC) tissues compared to adjacent normal tissues. N. Expression correlation analysis demonstrates a significant inverse relationship between TRAF6 and EZH2 mRNA levels in colorectal cohorts (p < 0.0001). O, P. Overexpression of TRAF6 enhances polyubiquitination of EZH2, as detected by ubiquitin (Ub) immunoblotting in CRC cells. TRAF6 knockdown reduces EZH2 ubiquitination levels. Lysates were immunoprecipitated (IP) with anti-EZH2 and immunoblotted (IB) with anti-Ub. Q. Co-IP assay demonstrating strengthened EZH2-TRAF6 interaction in DNMT1-depleted cells (lysates immunoprecipitated with anti-EZH2, blotted for TRAF6; input: 5% lysate). R. Dual inhibition of TRAF6 and DNMT1 rescues EZH2 hyperubiquitination (IP-Western blot with siTRAF6/siDNMT1 co-treatment EZH2 is essential for DNMT1-mediated CRC progression via the TRAF6 pathway To validate the functional significance of EZH2 in this axis, we knocked down EZH2 in DNMT1-overexpressing CRC cells. EZH2 depletion reversed DNMT1-induced upregulation of cyclinD1 and cyclinE2 (Fig. 6A) and abolished DNMT1-driven proliferation (Fig. 6B). TRAF6 overexpression reversed DNMT1/EZH2-induced proliferation (Fig. 6C, D). Similarly, transwell assays showed that EZH2 knockdown suppressed DNMT1-enhanced migration and invasion, and these enhancements were also negated by TRAF6 overexpression (Fig. 6E-G). Furthermore, TRAF6 ablation rescues the impaired proliferation, migration, and invasion of CRC cells induced by DNMT1 or EZH2 silencing(Fig. S4A-D). These findings conclusively demonstrate that EZH2 is a downstream effector of DNMT1, mediating its Tumor-promoting effects. Fig. 6EZH2 is a downstream effector of DNMT1 in CRC progression. (A) Western blot analysis showing that EZH2 knockdown reverses DNMT1-induced upregulation of cyclinD1 and cyclinE2 in HCT8 and SW480 cells. (B) Colony formation assays demonstrating that EZH2 depletion abolishes DNMT1-driven proliferation in HCT8 cells (**p < 0.01 vs. DNMT1-OE + siControl). C-D. Colony formation assays demonstrating that TRAF6 overexpression abolishes DNMT1/EZH2-driven proliferation in HCT8 cells (**p < 0.01 vs. DNMT1-OE + Vector, **p < 0.01 vs. EZH2-OE + Vector). E. Transwell migration assays showing suppression of DNMT1-enhanced motility by EZH2 knockdown in HCT8 cells(***p < 0.001 vs. DNMT1-OE + siControl); Matrigel invasion assays confirming EZH2-dependent regulation of DNMT1-induced invasiveness in SW480 cells(**p < 0.01 vs. DNMT1-OE + siControl; scale bar: 100 μm). F. Transwell migration assays showing suppression of DNMT1-enhanced motility by TRFA6 overexpression in HCT8 cells(**p < 0.01 vs. DNMT1-OE + Vector); Matrigel invasion assays confirming TRAF6-dependent regulation of DNMT1-induced invasiveness in HCT8 cells(***p < 0.001 vs. DNMT1-OE + Vector; scale bar: 100 μm). G. Transwell migration assays showing suppression of EZH2-enhanced motility by TRFA6 overexpression in HCT8 cells(***p < 0.001 vs. EZH2-OE + Vector); Matrigel invasion assays confirming TRAF6-dependent regulation of EZH2-induced invasiveness in HCT8 cells(***p < 0.001 vs. EZH2-OE + Vector; scale bar: 100 μm) EZH2 is a downstream effector of DNMT1 in CRC progression. (A) Western blot analysis showing that EZH2 knockdown reverses DNMT1-induced upregulation of cyclinD1 and cyclinE2 in HCT8 and SW480 cells. (B) Colony formation assays demonstrating that EZH2 depletion abolishes DNMT1-driven proliferation in HCT8 cells (**p < 0.01 vs. DNMT1-OE + siControl). C-D. Colony formation assays demonstrating that TRAF6 overexpression abolishes DNMT1/EZH2-driven proliferation in HCT8 cells (**p < 0.01 vs. DNMT1-OE + Vector, **p < 0.01 vs. EZH2-OE + Vector). E. Transwell migration assays showing suppression of DNMT1-enhanced motility by EZH2 knockdown in HCT8 cells(***p < 0.001 vs. DNMT1-OE + siControl); Matrigel invasion assays confirming EZH2-dependent regulation of DNMT1-induced invasiveness in SW480 cells(**p < 0.01 vs. DNMT1-OE + siControl; scale bar: 100 μm). F. Transwell migration assays showing suppression of DNMT1-enhanced motility by TRFA6 overexpression in HCT8 cells(**p < 0.01 vs. DNMT1-OE + Vector); Matrigel invasion assays confirming TRAF6-dependent regulation of DNMT1-induced invasiveness in HCT8 cells(***p < 0.001 vs. DNMT1-OE + Vector; scale bar: 100 μm). G. Transwell migration assays showing suppression of EZH2-enhanced motility by TRFA6 overexpression in HCT8 cells(***p < 0.001 vs. EZH2-OE + Vector); Matrigel invasion assays confirming TRAF6-dependent regulation of EZH2-induced invasiveness in HCT8 cells(***p < 0.001 vs. EZH2-OE + Vector; scale bar: 100 μm) Antitumor effects of decitabine and tazemetostat and regulatory network of DNMT1, EZH2 and TRAF6 in CRC To evaluate the anti-tumor effects of decitabine and tazemetostat, both as monotherapies and in combination, we first assessed their impact on colorectal cancer cell proliferation. As shown in Fig. 7A, the combination treatment resulted in significantly greater suppression of proliferation compared to either agent alone. We further examined their effects on metastatic potential using migration and invasion assays; consistent with the proliferation data, the combination markedly reduced both migration and invasion compared to monotherapy groups (Fig. 7B), indicating a synergistic inhibitory effect on malignant behaviors. Immunohistochemistry results of colorectal cancer patient tissues showed that DNMT1 and EZH2 were highly expressed in tumor tissues compared with adjacent non-tumor tissues, while the expression of TRAF6 in tumor tissues was lower than that in adjacent non-tumor tissues(Fig. 7C). Decitabine can increase the expression of TRAF6. Subsequently, the increased TRAF6 promotes the ubiquitination and degradation of DNMT1. Notably, DNMT1 negatively regulates the expression of TRAF6, thus forming a negative feedback regulatory loop between DNMT1 and TRAF6. Moreover, DNMT1 stabilizes EZH2 by reducing the expression of TRAF6(Fig. 7D). Fig. 7Effects of Decitabine (DAC) Alone and in Combination with Tazemetostat (TAZ) on the Biological Behaviors of SW480 Cells and Analysis of Related Molecular Expressions. (A) Representative images of colonies formed by SW480 cells treated with DMSO (vehicle control), decitabine (DAC, 500 nM) alone, tazemetostat (TAZ, 1 µM) alone, or the combination (DAC + TAZ) (***p < 0.001). (B) Representative images (left) and quantitative analysis (right) of the migration (upper panel) and invasion (lower panel) of SW480 cells after treatment with DMSO (vehicle control), decitabine (DAC, 500 nM) alone, tazemetostat (TAZ, 1 µM) alone, or the combination (DAC + TAZ) (**p < 0.01, ***p < 0.001). (C) The expression levels of DNMT1, EZH2 and TRAF6 in colorectal cancer tissues and adjacent non-tumor tissues. (D) Regulatory Relationships among Decitabine, TRAF6, DNMT1 and EZH2 Effects of Decitabine (DAC) Alone and in Combination with Tazemetostat (TAZ) on the Biological Behaviors of SW480 Cells and Analysis of Related Molecular Expressions. (A) Representative images of colonies formed by SW480 cells treated with DMSO (vehicle control), decitabine (DAC, 500 nM) alone, tazemetostat (TAZ, 1 µM) alone, or the combination (DAC + TAZ) (***p < 0.001). (B) Representative images (left) and quantitative analysis (right) of the migration (upper panel) and invasion (lower panel) of SW480 cells after treatment with DMSO (vehicle control), decitabine (DAC, 500 nM) alone, tazemetostat (TAZ, 1 µM) alone, or the combination (DAC + TAZ) (**p < 0.01, ***p < 0.001). (C) The expression levels of DNMT1, EZH2 and TRAF6 in colorectal cancer tissues and adjacent non-tumor tissues. (D) Regulatory Relationships among Decitabine, TRAF6, DNMT1 and EZH2 Discussion Colorectal cancer (CRC) progression is driven by epigenetic dysregulation, with DNMT1 and EZH2 emerging as central players. This study reveals for the first time the TRAF6-mediated DNMT1-EZH2 bidirectional regulatory axis. By integrating the DNA methylation enzymes and protein ubiquitination degradation network, it reshapes the malignant phenotype of CRC. Our findings provide novel insights into the crosstalk between DNA methylation enzymes and ubiquitin-proteasome pathways in CRC pathogenesis. Our initial observations data, including the EGO database, consistently demonstrated DNMT1 overexpression in CRC tissues and cell lines. Functional studies have confirmed that DNMT1 plays a role in promoting colorectal cancer (CRC) cell proliferation. Mechanistic investigations further showed that DNMT1 can enhance the expression of cyclin D1 and cyclin E2, and this upregulation of cyclin D1/E2 may contribute to the acceleration of G1/S phase transition, thereby potentially driving uncontrolled cell division in CRC cells. This aligns with prior studies showing DNMT1 induces cell cycle disorders in pancreatic/lung cancer by silencing tumor suppressors HOXB9/P53 via abnormal methylation [28, 29]. Notably, the extended value of this study lies in the fact that it is the first to expand the function of DNMT1 from “DNA methylation-mediated transcriptional repression” to “regulation of protein ubiquitination and degradation”, thereby providing dual mechanistic support for explaining its potent oncogenic role in colorectal cancer (CRC). Beyond proliferation, we uncovered DNMT1’s critical involvement in CRC metastasis. Transwell assays demonstrated that DNMT1 suppression or DAC treatment impaired migration and invasion. This result echoes recent findings that DNMT1 inhibits the expression of downstream miR-137 through methylation, thereby enhancing the ability of colorectal cancer cells to metastasize to the liver [30]. However, we identified TRAF6, an E3 ubiquitin ligase, as a key downstream target mediating DNMT1’s degradation. This finding bridges DNA methylation regulation with protein degradation pathways, a less explored axis in CRC metastasis. The reciprocal regulatory relationship between DNMT1 and TRAF6 emerged as a central theme. TCGA data analysis revealed inverse correlations between their expressions: DNMT1 overexpression suppressed TRAF6 levels, while TRAF6 upregulation destabilized DNMT1 via proteasomal degradation. DNMT1 is known to be abnormally overexpressed in various tumors, promoting tumorigenesis through aberrant DNA methylation—for example, in liver cancer, where DNMT1 overexpression correlates with poor prognosis and drives progression by methylating the suppressor gene BEX [31]. Notably, despite DNMT1’s global DNA methyltransferase activity, our methylation-specific PCR and dual-luciferase assays identified promoter hypermethylation of TRAF6, suggesting DNMT1 may regulate TRAF6 through epigenetic mechanisms. This aligns with recent studies showing DNMT1-mediated silencing of E3 ligases as a common oncogenic strategy [32], but expands it to a context-specific role in CRC. Another key finding is DNMT1’s function in maintaining EZH2 protein stability by inhibiting TRAF6-mediated ubiquitination of EZH2. This mechanism forms a bidirectional loop with DNMT1’s classical epigenetic function: DNMT1 catalyzes methylation of the TRAF6 promoter to suppress transcription, while DNMT1 protein directly binds TRAF6, sterically hindering its interaction with EZH2. This ensures continuous EZH2 accumulation. It is hypothesized that it is precisely because the DNMT1-mediated EZH2 stabilization mechanism is not blocked that EZH2 cannot be completely eliminated. This study echoes the recently reported finding that the combination of EZH2 inhibitors and demethylating agents can improve the efficacy in hepatocellular carcinoma [33]. Notably, the E2-conjugating enzyme UBE2L6 and E3 ligase UBR4 regulate EZH2 stability in melanoma [34]. but our study identifies DNMT1 as a novel EZH2 stabilizer in CRC through TRAF6-dependent ubiquitination, highlighting context-specific regulation of EZH2 proteostasis. Decitabine, a hypomethylating agent clinically used for hematological malignancies [35, 36], is shown here to exert a novel role in triggering DNMT1 degradation via the ubiquitin-proteasome pathway, beyond its classical DNA demethylation function. This dual mechanism may explain why low-dose decitabine, which has limited single-agent activity in solid tumors, synergizes with other therapies: combined treatment with decitabine and HDAC inhibitors, CDA inhibitors, or PD-1 blockade shows superior efficacy in inhibiting malignant phenotypes [14, 37]. Our findings suggest these synergies may partly stem from concurrent DNMT1/EZH2 degradation, complementing decitabine’s known role in reactivating tumor antigens [38]. This mechanism holds particular promise for microsatellite-stable colorectal cancer (MSS CRC), a subtype that is inherently refractory to immunotherapy—largely due to its low tumor mutational burden (weak immunogenicity) and an immunosuppressive tumor microenvironment [39]; Notably, previous studies have reported that EZH2 inhibition reduces regulatory T cell (Treg) infiltration (alleviating immunosuppression) [40], while DNMT1 depletion enhances antigen presentation (strengthening antitumor immune activation) [41]. Together, these observations highlight the DNMT1-TRAF6-EZH2 axis as a potential therapeutic target to reverse MSS CRC’s resistance to immunotherapy by simultaneously addressing its dual defects in immunogenicity and immune suppression. Our study has several limitations. First, the in vitro nature of most experiments precludes direct validation in vivo; thus, the therapeutic strategies of decitabine alone or in combination with an EZH2 inhibitor for colorectal cancer remain unvalidated. Second, it remains unclear whether EZH2, as a histone methyltransferase, inhibits the expression of downstream tumor suppressor genes through H3K27me3 modification. Additionally, we have not yet analyzed the correlation between the expression pattern of this regulatory axis and patients’ clinicopathological characteristics. In conclusion, our data support a model where decitabine exerts dual antitumor effects by destabilizing both DNMT1 and EZH2 through TRAF6 activation. The synergistic efficacy of combined DNMT1/EZH2 inhibition highlights the therapeutic promise of co-targeting interconnected epigenetic nodes. This study redefines DNMT1 as a multifunctional oncoprotein that coordinates DNA methylation and protein stability networks via TRAF6. By elucidating the mechanistic underpinnings of the DNMT1-TRAF6-EZH2 axis, we provide a rationale for innovative combinatorial therapies in CRC. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2