Skip to Content
MDPIMDPI

Advanced

CancersCancers
PDF
  • Article
  • Open Access

24 February 2026

23 Pages

DNMT3B Controls Enhancer-Linked Chromatin and Cell Cycle Networks in Acute Myeloid Leukemia

,
,
and
1
Division of Hematology/Oncology, Department of Pediatrics, Arkansas Children’s Research Institute, University of Arkansas for Medical Sciences, Little Rock, AR 72202, USA
2
Department of Neuroscience, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Cancers2026, 18(5), 723;https://doi.org/10.3390/cancers18050723
This article belongs to the Section Molecular Cancer Biology
Version Notes
Order Reprints

Simple Summary

Acute myeloid leukemia (AML) cells with high levels of the epigenetic regulator DNMT3B show poor patient outcomes and appear to rely on DNMT3B-controlled DNA regulatory elements to maintain growth and survival. We found that DNMT3B overexpression is driven by changes in gene control regions rather than mutations, and that inhibiting DNMT3B selectively remodels DNA methylation and chromatin at key cell-cycle and survival genes without globally erasing methylation. This disruption induces replication stress and programmed cell death and, importantly, rewires the balance of pro- and anti-apoptotic signals to make leukemia cells markedly more sensitive to the BCL-2 inhibitor venetoclax. These findings identify DNMT3B as a central epigenetic stabilizer of leukemic fitness and support combined DNMT3B and BCL-2 targeting as a rational therapeutic strategy for high-risk AML.

Abstract

Background: DNMT3B is frequently overexpressed in molecular subsets of acute myeloid leukemia (AML) and is associated with poor prognosis. Unlike DNMT3A, DNMT3B is rarely mutated, suggesting dysregulation through epigenetic mechanisms. The regulatory basis and downstream consequences of DNMT3B overexpression in AML remain incompletely defined. Methods: We integrated analyses of BeatAML, TCGA, and BLUEPRINT cohorts with multi-omic profiling (RNA-seq, DNA methylation, ATAC-seq, and proteomics) in DNMT3B-high AML models. Nanaomycin A (NanA) was used as a DNMT3B-directed functional probe to interrogate cis-regulatory remodeling, transcriptional circuitry, and apoptotic dependencies. Results: DNMT3B overexpression was linked to enhancer-associated chromatin activation rather than recurrent genetic mutation, particularly in CEBPA- and NPM1-mutant AML. NanA exposure produced focal epigenomic remodeling, including 6900 differentially methylated CpGs, with 268 CpGs located within regions of altered chromatin accessibility. These changes were accompanied by coordinated transcriptomic and proteomic reprogramming enriched for cell-cycle, checkpoint, and stress-response pathways. Functionally, DNMT3B perturbation induced redistribution of cell-cycle phases with increased S-phase fraction and progressive apoptosis. Transcriptional profiling demonstrated induction of BH3-only sensitizers (NOXA, PUMA), repression of BCL2, and compensatory upregulation of MCL1 and BCL-XL, collectively reshaping apoptotic dependency. Combined DNMT3B perturbation and BCL2 inhibition produced cooperative cytotoxicity in DNMT3B-high AML models. Conclusion: DNMT3B functions as a context-dependent epigenetic regulator linking enhancer-associated chromatin organization with proliferative control and apoptotic resistance in AML. DNMT3B-directed epigenetic perturbation remodels cis-regulatory circuitry and is associated with increased venetoclax responsiveness, supporting DNMT3B-governed networks as a candidate co-targeting axis in high-risk AML.

1. Introduction

DNA methylation (DNAm) is a fundamental epigenetic modification that regulates gene expression, chromatin architecture, and lineage fidelity. Aberrant DNAm patterns are a hallmark of malignancy, particularly in hematologic cancers such as acute myeloid leukemia (AML), where they reinforce stemness, impair differentiation, and contribute to therapeutic resistance [1,2]. These patterns are established and maintained by the DNA methyltransferases DNMT1, DNMT3A, and DNMT3B [3,4]. DNMT1 functions primarily as a maintenance methyltransferase with high affinity for hemimethylated DNA during replication, whereas DNMT3A and DNMT3B are classically defined as de novo enzymes. However, DNMT3A/3B can contribute to maintenance methylation at selected loci, and DNMT1 can exhibit limited de novo activity, indicating coordinated rather than strictly segregated functions [5,6,7].
These enzymes differ markedly in their involvement in AML pathogenesis. DNMT3A is frequently mutated in AML, most commonly at the R882 hotspot, where dominant-negative loss of function produces focal hypomethylation and contributes to clonal hematopoiesis. In contrast, DNMT3B is rarely mutated but often overexpressed, particularly in high-risk molecular subtypes, and elevated DNMT3B expression correlates with adverse clinical outcomes. Its functional role appears context-dependent. In KMT2A-rearranged leukemia models, Dnmt3b can act as a tumor suppressor by maintaining gene-body methylation and repressing stemness-associated loci [8]. Conversely, in cytogenetically normal AML, increased DNMT3B expression predicts inferior outcomes independent of global DNAm changes, suggesting regulatory functions beyond canonical methyltransferase activity [9]. DNMT3B is also elevated in FLT3-ITD–positive AML, a transcriptionally plastic and relapse-prone subtype, raising the possibility that DNMT3B contributes to stabilization of oncogenic regulatory networks [9,10].
Emerging evidence indicates functional interplay between DNMT family members. Catalytically inactive DNMT3B isoforms can act as accessory factors that recruit or stimulate DNMT3A activity at specific genomic loci, while structural studies of DNMT3A R882 mutants show disrupted DNMT3-family oligomerization and altered methylation targeting. DNMT enzymes also exert non-catalytic effects: DNMT1 can repress transcription through interactions with histone-modifying complexes, and DNMT3B participates in chromatin regulation through protein–protein interactions and inactive splice isoforms. DNMT3A and DNMT3B exhibit distinct chromatin-targeting preferences, with DNMT3A associated with H3K36me2/3-marked gene bodies, and DNMT3B has been reported to associate with linker DNA, repetitive elements, and repressive chromatin domains in several contexts. Consequently, shifts in their relative abundance redistribute methylation activity across regulatory regions. In DNMT3A-mutant AML, loss of DNMT3A function may not be fully compensated by DNMT3B, whereas DNMT3B overexpression may reinforce methylation at normally repressed elements and stabilize aberrant epigenetic programs [11]. These context-dependent observations—DNMT3B acting as a tumor suppressor in certain leukemias yet marking aggressive disease when overexpressed—indicate that DNMT3B functions within a broader DNMT regulatory network shaped by mutational background and chromatin context. Enhancer-associated DNMT3B overexpression in specific AML subtypes, together with its distinct genomic targeting and accessory functions, supports a model in which DNMT3B stabilizes leukemic regulatory circuitry.
Here, we integrate DNA methylation, chromatin accessibility, transcriptomic, and proteomic profiling to define the cis-regulatory features underlying DNMT3B overexpression in AML. We show that DNMT3B activation is associated with enhancer-like chromatin remodeling rather than recurrent mutation. We further examine how pharmacologic DNMT3B-directed perturbation influences chromatin organization, cell-cycle control, and apoptotic signaling. These findings provide a mechanistic context for DNMT3B-dependent epigenetic circuitry and support its consideration as a rational co-target in molecularly defined AML subsets and potentially other DNMT3B-dysregulated malignancies [12].

2. Materials and Methods

2.1. Genomic and Epigenetic Analysis of AML Patient Data and Cell Lines

DNMT3B expression (n = 395) and survival associations (n = 649) were quantified in BeatAML 2.0 (Tables S1 and S2) [13]. To provide additional context for expression differences between normal hematopoietic cells and AML subtypes, an independent microarray-based AML dataset from the MILE study (accessed via BloodSpot) was analyzed for comparative DNMT3B expression across normal bone marrow and molecular AML subgroups. DNMT family mutation profiles were assessed in TCGA (n = 143), and BLUEPRINT DHS/histone-mark profiles from 37 AML samples identified four promoter and intragenic DHS modules (DHS-I–IV) at the DNMT3B locus enriched for promoter/enhancer marks (Tables S3 and S4) [14]. (Full processing, Hotspot v3/MACS2 parameters, and QC details are provided in Supplementary Information (SI) Methods.)

2.2. Functional Dependency of DNMT3B in AML

DNMT3B mRNA expression from CCLE was correlated with DEMETER2 RNAi dependency scores to assess gene essentiality in AML [15]. Protein abundance was evaluated by immunoblotting in AML cell lines and primary HSCs, and functional dependency was further tested by treating cells with the DNMT3B inhibitor Nanaomycin A (MedChemExpress, Monmouth Junction, NJ, USA) and assessed for concentration-dependent effects on viability. Detailed protocols, antibody information, and assay conditions are provided in the SI.

2.3. Transcriptomic and Pathway Enrichment Analysis

KG1A cells were treated with NanA at 50 nM for 48 h (selected as a sub-IC50 concentration based on viability curves), followed by RNA-seq on three independent biological replicates (details in SI, Table S5); interpretation emphasizes pathway-level responses and acknowledges that later time points may include secondary transcriptional effects related to stress and apoptosis. Differential expression was assessed using DESeq2 (v1.40.2) with |log2FC| ≥ 1.5 and FDR < 0.05, defining differentially expressed genes (DEGs). Pathway enrichment was performed using GSEA (v4.3.3) with MSigDB Hallmark gene sets and 1000 permutations (Table S6). Gene sets were considered enriched at nominal p-value < 0.05 with FDR q-value < 0.25, consistent with GSEA recommendations.

2.4. Integration of DNA Methylation and Gene Expression Profiles

Genome-wide DNAm profiling was performed on NanA-treated KG1A cells using the Illumina Infinium MethylationEPIC v2.0 BeadChip (Illumina, San Diego, CA, USA), with three independent biological replicates per condition (processing and annotation in SI). CpGs were tested with limma (BH-FDR); for downstream integration, we classified CpGs with |Δβ| ≥ 0.10 relative to controls as hyper- (increase) or hypo- (decrease) methylated, and genes with average Δβ ≥ ±10% across promoter or gene body regions were defined as differentially methylated genes (DMGs). DMGs were integrated with differentially expressed genes (|log2FC| ≥ 1, FDR < 0.05), generating four categories: hyper-up, hyper-down, hypo-up, and hypo-down. Circos plots were produced using Circa (v1.0, OMGenomics, San Francisco, CA, USA).

2.5. Proteomic Profiling and Integration with Epigenomic Data

Mass spectrometry–based proteomics was performed on five biological replicates per condition using Orbitrap DIA (ThermoFisher Scientific, Waltham, MA, USA) acquisition, with proteins identified by EncyclopeDIA and quantified via Scaffold DIA at 1% FDR. Normalization and differential abundance were analyzed using ProteiNorm and limma, with significant proteins defined as FDR < 0.05 and fold change > 2; pathway enrichment was subsequently performed. To assess epigenetic regulation, differentially methylated genes (DMGs) were integrated with RNA-seq and proteomic data, and genes showing concordant transcript and protein changes were classified as coordinately regulated. Detailed acquisition parameters, preprocessing workflows, and statistical criteria are provided in the SI.

2.6. ATAC-Seq and Integration with DNA Methylation Profiling

ATAC-seq was performed on KG1A cells treated with NanA or vehicle (n = 3 per group) to map chromatin accessibility, as previously described [16]. Because NanA induces time-dependent apoptosis, accessibility changes are interpreted as reflecting both primary regulatory remodeling and secondary stress-related chromatin responses at the 48 h time point. Differentially accessible regions were identified using DiffBind (v3.8.0; FDR < 0.001) and intersected with differentially methylated CpGs from Illumina MethylationEPIC arrays (SI). TF motif enrichment was performed using Pscan-ChIP with the JASPAR 2018 NR database [17] both globally (regions gaining accessibility after DNMT3B inhibition) and locally at DHS-I, the promoter-proximal DHS at the DNMT3B locus. For DHS-I, hg38 coordinates were analyzed with a promoter background against the TRANSFAC database to identify candidate regulators of DNMT3B expression.

2.7. Apoptosis and Cell Cycle Analysis

KG1A cells (1 × 106 per condition) were treated with NanA (25 or 50 nM) or DMSO in three biological replicates. Apoptosis was quantified using the FITC Annexin V/PI Dead Cell Apoptosis Kit (including DAPI counterstain where indicated; V13242, Thermo Fisher Scientific, Waltham, MA, USA) on a BD LSRFortessa cytometer (BD Biosciences, Milpitas, CA, USA), distinguishing early and late apoptotic populations. For cell-cycle analysis, DNA content was quantified by flow cytometry using Vybrant DyeCycle Violet (Thermo Fisher Scientific, V35003, Waltham, MA, USA) to determine G0/G1, S, and G2/M phase distribution.

2.8. Combination Treatment, Cell Viability, and qPCR Analysis of Apoptotic Regulators

M07e and KG1A cells were seeded in 96-well plates (2 × 104 and 1 × 104 cells/well, respectively) in RPMI-1640 with 10% FBS and treated with Nanaomycin-A (NanA) and Venetoclax (VEN; HY-15531, MedChemExpress, Monmouth Junction, NJ, USA), either alone or in combination, across an 11-point range (10 nM–25 µM). For combination therapy, VEN and NanA were co-administered at a 1:1 molar ratio across matched concentrations. After 48 h, cell viability was measured using the CellTiter-Glo Luminescent Assay (Promega, G9242, Madison, WI, USA) on an Agilent BioTek Gen5 Reader (Santa Clara, CA, USA), with values normalized to DMSO controls. Each condition was tested in triplicate across three biological replicates. In parallel, KG1A cells were treated for 48 h with DMSO, NanA, VEN, or NanA + VEN (each at one-fold below IC50 to ensure sub-lethal dosing). Total RNA was isolated using the RNeasy Mini Kit (Qiagen; 74104, Germantown, MD, USA), reverse-transcribed (500 ng) with iScript (Bio-Rad; 1725037, Hercules, CA, USA), and subjected to qPCR with SYBR Green Master Mix (Bio-Rad; 1725270) on a CFX96 Touch Real-Time PCR Detection (Bio-Rad; 1855195). Expression of apoptotic regulators (BCL2, MCL1, BCL2L1, PMAIP1/NOXA, BCL2L11/BIM, BBC3/PUMA, BAX, BAK1) was quantified by the ΔΔCt method, normalized to ATP5B, using technical triplicates from three biological replicates. Primer sequences are provided in Table S9.

2.9. Statistical Analysis

For targeted functional assays, statistical comparisons between groups were performed using Mann–Whitney U tests or one-way ANOVA as appropriate, with significance defined as p < 0.05 unless otherwise specified. Omics-scale analyses incorporated multiple-testing correction within each dataset using established analytical frameworks: DESeq2 for RNA-seq, limma for DNA methylation and proteomics, and DiffBind for ATAC-seq, with Benjamini–Hochberg false discovery rate (FDR) adjustment applied per analysis. Specific FDR thresholds and effect-size criteria used for each dataset are described in the relevant Methods subsections. All results reflect at least three biological replicates per condition. Data visualization and statistical analyses were performed using OriginPro (v2025; Northampton, MA, USA), GraphPad Prism (v10.5.0, Boston, MA, USA), and schematic elements in selected figures were created using BioRender (BioRender, Toronto, ON, Canada).

3. Results

3.1. DNMT3B Overexpression in AML Is Linked to Poor Prognosis and Epigenetically Reprogrammed Regulatory Accessibility

To assess the clinical relevance of DNMT3B, we analyzed BEAT-AML transcriptomic and outcome data. Patients with elevated DNMT3B expression had significantly reduced overall survival (log-rank p < 0.05; Figure 1A). Expression was highest in AML M1 (without maturation), followed by CEBPA- and NPM1-mutant AML, but it remained low in APL (PML-RARA), t(8;21), and inv(16) (Figure 1B). Independent analysis of the microarray-based MILE AML cohort showed a similar overall pattern, with higher DNMT3B expression in several AML subtypes relative to normal bone marrow, supporting subtype-specific variability consistent with the BEAT-AML observations (Figure S1).
Figure 1. DNMT3B is overexpressed in AML and epigenetically activated via cis-regulatory chromatin remodeling. (A) Kaplan–Meier survival analysis of AML patients in the BEAT-AML cohort (n = 649), stratified by DNMT3B expression above (red) versus below (blue) the median. Elevated DNMT3B expression is significantly associated with inferior overall survival (p < 0.05, log-rank test). (B) Violin plots showing DNMT3B expression (log2 RPKM) across AML molecular subtypes. DNMT3B is highly expressed in AML M1, CEBPA-mutant, and NPM1-mutant AML, but repressed in APL (PML-RARA), t(8;21), and inv(16) AML. Normal BM mononuclear cells (BM-MNC) are shown in gray. ** p < 0.01, *** p < 0.001, unpaired t-test. (C) Oncoprint summarizing DNMT3A, DNMT1, and DNMT3B mutations in AML patients (TCGA, n = 143). DNMT3A harbors frequent coding mutations, whereas DNMT1 and DNMT3B are rarely altered. Mutation types are color-coded: green = missense; orange = splice site; red = nonsense; pink = frameshift deletion; black = multi-hits. (D) DNase I hypersensitivity (DHS) across the DNMT3B locus in 37 primary AML samples (BLUEPRINT). Four discrete DHS regions (DHS-I–IV) are highlighted (red boxes) spanning the promoter and gene body. CpG island and gene structure are shown below. (E) Boxplots of DHS-enrichment scores (Hotspot v3) across AML samples. DHS-I exhibits significantly higher enrichment than DHS-IV (adjusted p = 0.0163, Dunn’s test), while DHS-I–III are statistically indistinguishable. (F) Histone-mark profiles (BLUEPRINT ChIP-seq) at the DNMT3B locus. Active marks H3K4me3 (red), H3K27ac (pink), and H3K4me1 (orange) are enriched at DHS-I, spanning the promoter/TSS. Repressive marks H3K27me3 (gray) and H3K36me3 (green) show minimal signal. (G) Boxplots of histone-mark enrichment at DHS-I. H3K4me3 is significantly higher than all other marks (**** p < 0.0001, Dunn’s test); H3K27ac is also elevated versus repressive marks, consistent with active promoter–enhancer architecture. * Asterisks denote adjusted p values from Dunn’s post hoc test following Kruskal–Wallis analysis: ns = not significant; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.
Intra-subtype Kaplan–Meier analyses in CEBPA- and NPM1-mutant AML showed that above-median DNMT3B expression trended toward poorer survival; statistical power was limited for the CEBPA subgroup (low-expression n = 2), whereas the NPM1 subgroup did not show a significant association in this dataset. These findings suggest that elevated DNMT3B expression may mark more aggressive disease behavior within genetically defined AML subsets (Figure S2).
Somatic mutation profiling of DNMT genes in TCGA confirmed frequent DNMT3A mutations but rare alterations in DNMT1 or DNMT3B (Figure 1C), indicating that DNMT3B dysregulation arises through non-genetic mechanisms. To explore chromatin involvement, we analyzed DNase I hypersensitivity data from 37 AML samples in BLUEPRINT, identifying four discrete DHS (DHS-I–IV) across DNMT3B regulatory regions (Figure 1D). Median Hotspot v3-enrichment scores were highest at DHS-I (8.17, max 14.8), DHS-II (7.38), and DHS-III (7.40) compared with DHS-IV (6.18; Table S3). Kruskal–Wallis testing confirmed significant inter-DHS differences (p < 0.05), with Dunn’s post-hoc test identifying DHS-I as significantly more enriched than DHS-IV (adjusted p = 0.0163) (Figure 1E). Histone profiling at DHS-I revealed an active chromatin state characterized by strong H3K4me3 (median 19.8) and H3K27ac (4.79), moderate H3K4me1 (3.83), and low repressive marks H3K27me3 (3.07) and H3K36me3 (2.82) (Figure 1F,G). Because histone-mark datasets for normal myeloid progenitors are limited, these results define AML-specific enrichment rather than disease-exclusive activation. The concurrent presence of H3K4me3 and H3K27ac within a CpG-island promoter indicates a promoter-proximal enhancer-like configuration rather than a canonical distal enhancer. Kruskal–Wallis testing (p < 0.0001), followed by Dunn’s comparisons, showed H3K4me3 enrichment exceeded all other marks (adjusted p < 0.01), and H3K27ac was higher than H3K4me1 and H3K27me3 (adjusted p < 0.01). Motif-enrichment analysis of DHS-I using Pscan-ChIP identified SRF as the top-enriched factor (p = 0.0103), suggesting a role in DNMT3B promoter activation (Figure S3). Together, these data indicate that DNMT3B overexpression in AML is associated with promoter-proximal chromatin accessibility and enhancer-like regulatory features rather than recurrent somatic mutation, although direct comparisons with matched normal hematopoietic chromatin datasets remain limited.

3.2. DNMT3B Inhibition Induces Mitotic Disruption and Stress Response Pathways in High-Expressing AML Cells

We first examined the relationship between DNMT3B transcript abundance and functional dependency using the DEMETER2 RNAi dataset. DNMT3B transcript abundance was not significantly correlated with RNAi gene-effect scores (Pearson r = 0.294, p = 0.18; Figure 2A), indicating that DNMT3B expression alone does not predict functional dependency. Several DNMT3B-overexpressing AML lines used in this study (TF-1, KG1A, M07e) were absent from DEMETER2. To address this, CRISPR-based Chronos scores from the DepMap 25Q2 dataset (Table S10) revealed minimal dependency for TF-1 (+0.018) and M07e (−0.014), reinforcing that high DNMT3B expression does not necessarily confer essentiality in vitro. Immunoblotting confirmed markedly higher DNMT3B protein levels in TF-1, KG1A, and M07e compared with MV4;11, AML-193, and MOLM-13, whereas CD34+ HSPCs lacked detectable DNMT3B (Figure 2B). These findings establish TF-1, KG1A, and M07e as representative DNMT3B-overexpressing AML models for pharmacologic interrogation.
Figure 2. DNMT3B inhibition compromises AML cell viability and reprograms transcriptional networks through disruption of cell-cycle and stress-response pathways. (A) Scatterplot showing correlation between DNMT3B expression (x-axis, log2[TPM + 1]) and RNAi-based gene-effect scores (y-axis, DEMETER2) across AML cell lines (DepMap). A weak, non-significant correlation (Pearson r = 0.294) indicates expression does not predict dependency; analysis includes 23 AML cell lines available in DEMETER2, and DNMT3B-high lines used elsewhere in this study (TF-1, KG1A, M07e) are not represented. (B) Immunoblot showing DNMT3B protein in primary CD34+ HSPCs and AML cell lines. TF-1, KG1A, and M07e exhibit strong DNMT3B signal, whereas MV4;11, MOLM-13, and AML-193 show low levels. β-actin = loading control (Uncropped immunoblots: Figure S4). (C) Dose–response curves for AML cell lines treated for 48 h with the DNMT3B inhibitor Nanaomycin-A (NanA). DNMT3B-high lines (M07e, TF-1, KG1A) are relatively resistant (IC50 = 195–969 nM), while DNMT3B-low lines (MV4;11, MOLM-13, AML-193) are more sensitive (IC50 = 22–75 nM). (D) Western blot of DNMT3B in KG1A cells treated with 25, 50, or 100 nM NanA for 48 h. DNMT3B levels decline dose-dependently but remain detectable at 100 nM (Uncropped immunoblots: Figure S5). (E) Quantification of DNMT3B relative to β-actin (mean ± SEM, n = 3). Significant reduction versus DMSO control at all doses (ns = not significant; *** p < 0.001; **** p < 0.0001; one-way ANOVA + Tukey). (F) Volcano plot of differential gene expression in NanA-treated (50 nM, 48 h) versus DMSO KG1A cells. Up-regulated (red) and down-regulated (green) genes are defined by adjusted p < 0.05 and |log2FC| ≥ 1.5. (G) GSEA (Hallmark MSigDB) of NanA-responsive transcriptome. Top pathways include mitotic spindle, TNFα signaling via NF-κB, p53 pathway, and G2/M checkpoint, linking DNMT3B inhibition to cell-cycle and stress-response reprogramming. Dot size = number of genes; color = −log10(FDR q-value, GSEA).
To probe DNMT3B-associated vulnerabilities, we treated AML cell lines with Nanaomycin A (NanA), used here as a DNMT3B-directed functional probe given reported catalytic selectivity alongside potential DNMT3B-independent cytotoxic effects. Viability assays revealed a spectrum of sensitivities (IC50 = 62–969 nM; Figure 2C): TF-1 (969 nM), KG1A (334 nM), and M07e (195 nM) were relatively resistant, whereas MV4;11 (62 nM), AML-193 (74 nM), and MOLM-13 (22 nM) were more sensitive; notably, at fixed low-nanomolar dosing, NanA also elicited time-dependent stress and apoptosis, emphasizing that IC50 values should be interpreted alongside exposure duration. This pattern indicates that NanA cytotoxicity only partially tracks with DNMT3B abundance, consistent with mixed determinants of response and the known limitations of NanA as a quinone-based chemical probe [18]. Accordingly, we used NanA to interrogate DNMT3B-linked chromatin and transcriptional responses while explicitly recognizing that viability effects may not be exclusively DNMT3B-dependent. Notably, DNMT3B-high models were not uniformly more sensitive to NanA, consistent with the lack of DNMT3B dependency in RNAi/CRISPR datasets and suggesting that DNMT3B expression alone does not determine acute cytotoxic response.
In KG1A cells, NanA exposure was associated with a concentration-dependent reduction in DNMT3B protein abundance, with partial reduction at 25–50 nM and greater reduction at 100 nM (Figure 2D,E; p < 0.001). Residual DNMT3B protein at 100 nM indicates incomplete catalytic blockade, consistent with NanA’s reported pharmacodynamic profile. RNA-seq analysis of KG1A cells treated with 50 nM of NanA for 48 h revealed 1835 down-regulated and 1129 up-regulated genes (Figure 2F; Table S5). Gene-set enrichment analysis (MSigDB Hallmark) identified enrichment of mitotic spindle, TNFα signaling via NF-κB, p53 pathway, IL2–STAT5 signaling, hypoxia, and G2/M checkpoint programs (Figure 2G), consistent with integrated stress and checkpoint responses at this exposure duration. Among these, mitotic-spindle and checkpoint pathways were most enriched, linking pharmacologic DNMT3B inhibition to stress-response and apoptotic networks rather than direct cell-cycle arrest. The enrichment of p53 and inflammatory signaling signatures further suggests that DNMT3B-associated epigenetic perturbation initiates integrated stress programs that couple proliferative control with apoptotic priming in AML cells.

3.3. Disruption of DNMT3B Activity Rewires Cis-Regulatory Methylation and Protein Abundance in Cell Cycle Networks

To define the epigenomic consequences of DNMT3B inhibition, we profiled genome-wide DNA methylation in KG1A cells treated with Nanaomycin A using Illumina EPIC arrays. Circos visualization revealed widespread differential methylation with a net bias toward CpG hypomethylation (Figure 3A). In total, 6900 CpGs were significantly altered (FDR < 0.05; |Δβ| > 0.1), most within promoters (n = 3950; 3439 hypo-, 511 hypermethylated), followed by gene bodies (n = 1592; 1085 hypo-, 507 hypermethylated) and intergenic regions (n = 1358; 803 hypo-, 555 hypermethylated). Most methylation changes were modest (Δβ within ±0.25; Figure 3B), consistent with partial reconfiguration rather than global erasure. To link methylation dynamics with transcriptional output, differentially methylated CpGs were intersected with significantly altered genes from RNA-seq analysis. Correlation analysis revealed weak inverse associations between methylation and expression at promoters (r = –0.12) and gene bodies (r = –0.20; Figure 3C,D), indicating that DNMT3B-associated methylation changes act in concert with additional regulatory layers such as chromatin accessibility and transcription-factor recruitment rather than functioning as a sole determinant of transcription.
Figure 3. DNMT3B inhibition promotes DNA hypomethylation and induces coordinated transcriptomic–proteomic remodeling in AML cells. (A) Circos plot of differential CpG methylation in KG1A cells treated with NanA (50 nM, 48 h) versus DMSO. Hypomethylated CpGs (blue) predominate over hypermethylated (red). Outer ring denotes chromosomes (1–22, X). (B) Density histograms of Δβ-values across promoter, gene-body, and intergenic regions. Promoters display a left shift toward hypomethylation. (C,D) Scatterplots of differential DNA methylation (Δβ) vs. differential gene expression (log2FC) for promoter (C) and gene body (D) CpGs. Modest negative correlation (r = −0.12 and −0.20). (E) Correlation between transcript and protein fold changes (r = 0.52) for co-regulated genes showing promoter hypomethylation. (F,G) Heatmaps of Z-scored RNA (F; n = 3) and protein (G; n = 5) abundance for 17 coordinately upregulated genes after NanA treatment. Notable targets: CDK6, GSK3B, KDM2A, CENPE. Asterisks denote statistically significant upregulation at both levels.
Genes were classified by concordant or discordant methylation–expression patterns. Promoter-associated genes included 61 hypo-up, 117 hypo-down, 7 hyper-up, and 27 hyper-down. Gene-body-linked genes showed 12 hypo-up, 58 hypo-down, 3 hyper-up, and 33 hyper-down. The prevalence of hypo-down and hyper-up categories deviates from canonical repression models, supporting network-level rather than unidirectional control of transcriptional responses. To determine whether these transcriptional shifts extended to the proteome, RNA-seq data were integrated with quantitative proteomics from DNMT3B-inhibited KG1A cells. A subset of promoter-hypomethylated and transcriptionally up-regulated genes also showed increased protein abundance (Figure 3E). Seventeen genes—including TIAM1, CDK6, DENND1B, and KDM2A—were coordinately up-regulated at both transcript and protein levels (Figure 3F,G; Table S6). These genes participate in cell-cycle progression, cytoskeletal organization, checkpoint regulation, and chromatin remodeling, consistent with functional themes emerging from pathway analyses (Figure S6). Together, these findings show that DNMT3B inhibition preferentially induces hypomethylation at gene-proximal regions accompanied by selective transcriptomic and proteomic remodeling. Rather than directly “governing” transcription, DNMT3B appears to stabilize proliferative gene networks through multi-layered epigenetic interactions that integrate DNA methylation, chromatin accessibility, and transcriptional output.

3.4. DNMT3B Inhibition Induces Demethylation at Cis-Regulatory Elements of Cell-Cycle Genes

To assess how NanA exposure reshapes cis-regulatory architecture in DNMT3B-high AML, we intersected differentially methylated CpGs with differentially accessible chromatin regions from ATAC-seq in KG1A cells treated with 50 nM of NanA. Genome-wide annotation of differential accessibility (n = 36,942; Table S8) showed predominant localization within intronic (34%) and intergenic (33%) regions, followed by promoters (27%) and smaller fractions in exonic, untranslated, and termination sites (Figure 4A). Integration with DNAm data identified 268 CpGs located within regions of altered accessibility after DNMT3B inhibition, a small but statistically non-random overlap (hypergeometric p < 0.05) enriched at promoter-proximal and intragenic loci (Figure 4B). Promoter-associated CpGs (n = 184) were disproportionately represented compared with gene bodies (n = 40) and intergenic regions (n = 44). Most promoter-linked CpGs (n = 166) were hypomethylated, consistent with localized chromatin remodeling at regulatory regions of transcriptionally active genes.
Figure 4. DNMT3B inhibition induces focal hypomethylation and increased chromatin accessibility at cis-regulatory elements of cell-cycle genes. (A) Genomic annotation of differentially accessible chromatin regions (n = 36,942) by ATAC-seq in KG1A cells treated with NanA (50 nM, 48 h). Most elements reside in intronic (34%) and intergenic (33%) regions, followed by promoters (27%). (B) Distribution of CpGs exhibiting both differential methylation and accessibility. Promoter-linked CpGs (n = 184) dominate these overlaps. Blue = hypomethylated; red = hypermethylated. (C–F) IGV snapshots of representative cell-cycle genes (CDK6, CENPE, GSK3B, SLC7A5). DNMT3B inhibition increases chromatin accessibility (top) and induces CpG hypomethylation (middle) at promoter or enhancer elements (bottom = gene model). Arrows denote transcription direction. Axes: ATAC-seq (0–500); methylation enrichment (0–100).
IGV visualization of representative loci revealed concordant increases in chromatin accessibility and CpG hypomethylation at CDK6, CENPE, GSK3B, and SLC7A5 (Figure 4C–F). In CDK6, DNMT3B inhibition produced focal demethylation and enhanced accessibility at an intragenic DHS downstream of the TSS, displaying enhancer-like chromatin features. CENPE exhibited analogous promoter-proximal remodeling, while GSK3B and SLC7A5 showed accessibility gains within intronic, H3K27ac-positive regions defined as enhancers (>1 kb from the TSS) according to ENCODE chromatin-state maps. Motif analysis of DNMT3B-depleted accessible regions demonstrated enrichment for NRF1 and E2F1 motifs—transcription factors associated with chromatin opening and G1/S-phase progression (Figures S3 and S4). Functionally, these genes participate in mitotic-spindle organization, checkpoint control, and proliferative signaling, paralleling pathway-level reprogramming observed in transcriptomic and proteomic datasets (Figure 2 and Figure 3). Together, these findings indicate that NanA exposure is associated with selective demethylation and accessibility remodeling at cis-regulatory elements of cell-cycle genes, supporting a model in which DNMT3B maintains the epigenetic configuration of proliferative transcriptional networks in AML.

3.5. DNMT3B Inhibition Disrupts Cell Cycle Progression and Induces Time-Dependent Apoptosis in AML Cells

To functionally validate the transcriptional and epigenetic changes triggered by DNMT3B inhibition, we examined cell-cycle distribution and apoptosis in KG1A cells treated with 25 nM or 50 nM of Nanaomycin A. At 24 h, DNMT3B inhibition caused a significant redistribution across major phases (Figure 5A,B). Sub-G1 content increased at both concentrations, indicating early DNA fragmentation, whereas G1-phase cells declined at 25 nM and remained largely unchanged at 50 nM. S-phase fractions expanded at both doses, and G2/M increased at 50 nM, suggesting replication stress and checkpoint activation. By 48 h, these effects intensified: sub-G1 cells rose sharply, G1 populations fell markedly, and modest increases in S and G2/M persisted (Figure 5C,D), consistent with progression from S-phase disturbance toward G2/M checkpoint engagement and apoptotic accumulation.
Figure 5. DNMT3B inhibition disrupts cell-cycle progression and induces time-dependent apoptosis in AML cells. (A,B) Cell-cycle profiles of KG1A cells treated with DMSO or NanA (25 nM and 50 nM) for 24 h. Representative DAPI histograms (A) and quantification (B) show G1 depletion and S/G2 accumulation (* p < 0.05). (C,D) At 48 h, NanA treatment further reduces G1 fraction and increases S/G2, consistent with replication stress and checkpoint activation (** p < 0.01; *** p < 0.001). (E,F) Annexin V–FITC/PI staining at 24 h shows significant increases in early (Q2) and late (Q3) apoptotic cells (*** p < 0.001), with no change in viable (Q1) cells or necrotic (Q4) fraction. (G,H) At 48 h, apoptosis is dominant: Q2 and Q3 cells increase further (**** p < 0.0001), viable cells decrease, and necrotic cells decline. Data (mean ± SEM, n = 3) confirm that DNMT3B inhibition elicits a time-dependent programmed apoptotic response without necrotic conversion. ns = not significant; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.
Annexin V–FITC/PI staining confirmed that DNMT3B inhibition elicits a canonical, time-dependent apoptotic response (Figure 5E–H). At 24 h, early apoptotic cells (Q2; PI− AnnV+) increased by ~1 percentage point at both 25 nM and 50 nM relative to control, and late apoptotic cells (Q3; PI+ AnnV+) rose by ~2–3 points, while viable cells (Q1; double negative) and necrotic cells (Q4; PI+ AnnV−) showed no significant change. At 48 h, apoptosis became dominant: Q2 and Q3 populations increased markedly, viable cells decreased, and the distribution shifted away from the double-negative (AnnV−/PI−) population toward Annexin V–positive fractions; quadrant assignments and summary values are reported as defined in the corresponding figure panel. All changes except the minor variations in Q1 at 24 h were statistically significant (p < 0.05, Dunnett’s test). Together, these data show that NanA treatment is accompanied by accelerated loss of G1 fraction, accumulation of S/G2–M stress signatures, and a robust apoptotic response, consistent with checkpoint and p53 pathway enrichment in transcriptomic analyses (Figure S7).

3.6. DNMT3B Inhibition Primes AML Cells for Venetoclax Sensitization Through Apoptotic Rewiring

Because DNMT3B inhibition altered mitotic and checkpoint genes (Figure 2), induced focal hypomethylation at regulatory elements of cell-cycle regulators (Figure 4), and triggered progressive apoptosis (Figure 5), we hypothesized that DNMT3B blockade may lower the apoptotic threshold and sensitize AML cells to BCL2 inhibition. To test this, we assessed the combined effects of Nanaomycin A (NanA), used here as a DNMT3B-directed functional probe, and Venetoclax (VEN), a BCL2 antagonist. In KG1A cells, single-agent IC50 values were 334 nM for NanA and 136 nM for VEN, while the combination reduced the IC50 to 40.9 nM (95% CI 37.8–44.1) and steepened the dose-response slope (Figure 6A). In M07e cells, VEN and NanA monotherapy IC50 values were 1045 nM and 195 nM, respectively, whereas the combination decreased the IC50 to 132 nM (95% CI 121–144) (Figure 6B). Across concentrations, NanA + VEN produced greater cytotoxicity than either agent alone, consistent with cooperative activity in the DNMT3B-high models tested. Because the combination was tested using a fixed molar ratio rather than a full-dose–matrix design, formal Bliss or ZIP synergy modeling was not feasible; therefore, these data are interpreted conservatively as cooperative cytotoxicity rather than definitive pharmacologic synergy. Taken together with DNMT3B inhibition–induced replication stress and apoptosis (Figure 2, Figure 3, Figure 4 and Figure 5), these cooperative effects support a model in which DNMT3B-dependent epigenetic programming buffers apoptotic signaling; its inhibition functionally “primes” cells for BCL2 blockade. We interpret these data as apoptotic “priming” rather than direct DNMT3B–BCL2 coupling. Because apoptotic commitment is governed by the balance of pro- and anti-apoptotic BCL2-family programs, we next quantified transcriptional responses of key apoptotic regulators and interpret simultaneous induction of BH3-only genes and compensatory anti-apoptotic genes as apoptotic priming accompanied by adaptive buffering. Importantly, this interpretation aligns the combination phenotype with upstream mitotic/checkpoint disruption and apoptotic engagement following DNMT3B inhibition, supporting DNMT3B-governed circuitry as a mechanistically grounded co-targeting axis to enhance VEN efficacy in selected AML contexts.
Figure 6. DNMT3B inhibition primes AML cells for Venetoclax sensitization through transcriptional rewiring of apoptotic programs. (A,B) Dose–response curves of KG1A (A) and M07e (B) cells treated with NanA, Venetoclax (VEN), or NanA + VEN (10 nM–25 µM). Co-treatment markedly reduces viability relative to either agent alone, consistent with enhanced cytotoxicity relative to single-agent treatment. (C,D) qPCR of apoptotic effectors (BAK1, BAX). Both genes are significantly downregulated across treatments (**** p < 0.0001), consistent with post-transcriptional regulation of mitochondrial apoptosis. (E–G) BH3-only activators/sensitizers. PMAIP1/NOXA (E) is induced by NanA and further elevated with NanA + VEN (**** p < 0.0001). BBC3/PUMA (F) is strongly upregulated only in the combination arm (*** p < 0.001). BCL2L11/BIM (G) is modestly increased by single agents but plateaus in combination. (H–J) Anti-apoptotic BCL2 family genes. BCL2 (H) is repressed by NanA and VEN monotherapies and remains suppressed in combination (**** p < 0.0001). MCL1 (I) and BCL2L1/BCL-XL (J) are significantly upregulated in the combination arm (**** p < 0.0001), indicating compensatory survival activation. Expression values are normalized to ATP5B and plotted relative to DMSO control (mean ± SEM; n = 3 biological replicates, technical triplicates). Statistics: one-way ANOVA + Tukey’s post-hoc test. ns = not significant; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Panels are grouped by functional apoptotic category (effector, BH3-only sensitizer, anti-apoptotic regulator) with consistent color coding across treatment conditions.

3.7. Transcriptional Reprogramming of Apoptotic Regulators Following DNMT3B and BCL2 Inhibition

To investigate the molecular basis of VEN sensitization following DNMT3B inhibition, we profiled the expression of canonical apoptotic regulators by qPCR in KG1A cells treated with the DNMT3B inhibitor NanA, VEN, or their combination. The panel included mitochondrial apoptotic effectors (BAK1, BAX), BH3-only activators and sensitizers (PMAIP1/NOXA, BBC3/PUMA, BCL2L11/BIM), and anti-apoptotic BCL2 family members (BCL2, MCL1, BCL2L1/BCL-XL) (Figure 6C–J). BAK1 was significantly downregulated across all treatments relative to control (p < 0.0001), with no further reduction in combination, suggesting a basal reconfiguration of mitochondrial homeostasis rather than a treatment-specific effect (Figure 6C), while BAX exhibited progressive downregulation with both VEN and NanA + VEN (p < 0.01), consistent with its post-translational regulation and limited transcriptional responsiveness in AML (Figure 6D) [19,20]. In contrast, BH3-only genes showed stronger and differential responses: PMAIP1 (NOXA) was robustly induced by NanA alone and further elevated with NanA + VEN (p < 0.0001), consistent with its role in sequestering MCL1 and shifting apoptotic dependency toward BCL2 (Figure 6E) [21]. BBC3 (PUMA) expression remained unchanged under either monotherapy but was strongly upregulated in the combination arm (p < 0.001), suggesting that DNMT3B inhibition primes chromatin for inducible PUMA transcription (Figure 6F); and BCL2L11 (BIM) increased modestly under NanA or VEN alone (p < 0.01) but did not rise further in combination, indicating saturation or compensatory feedback regulation (Figure 6G). Among anti-apoptotic BCL2 family members, BCL2 was repressed by both DNMT3B inhibition and VEN individually (p < 0.05) and further suppressed in combination (p < 0.0001) (Figure 6H), consistent with transcriptional remodeling of BCL2 regulatory elements following DNMT3B inhibition [22]. In contrast, MCL1 and BCL2L1 (BCL-XL) were significantly upregulated in the NanA + VEN arm (both p < 0.0001) (Figure 6I,J), reflecting compensatory survival pathway activation commonly associated with acquired VEN resistance [23,24]. Collectively, these findings demonstrate that DNMT3B inhibition enhances the expression of BH3-only sensitizers (NOXA, PUMA), represses BCL2, and triggers compensatory MCL1/BCL-XL induction, thereby reshaping the transcriptional architecture of the apoptotic rheostat. Notably, induction of BH3-only sensitizers, together with MCL1/BCL-XL upregulation, is not mutually exclusive and is consistent with a stress-adaptation state in which apoptotic priming is counterbalanced by compensatory survival signaling. Thus, these data indicate that DNMT3B-directed perturbation shifts the apoptotic network toward increased priming while simultaneously engaging buffering pathways that are well-recognized mediators of emerging venetoclax resistance. This DNMT3B-dependent remodeling primes leukemia cells for mitochondrial apoptotic collapse upon BCL2 blockade, positioning DNMT3B as a mechanistic co-regulator and rational therapeutic target in the context of VEN-based AML therapy.

4. Discussion

Our study identifies DNMT3B-associated regulatory circuitry as a context-dependent epigenetic feature of AML, with pharmacologic perturbation linked to focal methylation changes at cis-regulatory elements, altered proliferative transcriptional programs, and apoptotic remodeling. Integrative multi-omic analyses demonstrate that DNMT3B inhibition initiates coordinated reprogramming of the methylome, chromatin landscape, transcriptome, and proteome, culminating in replication stress and programmed apoptosis. Rather than functioning primarily as a global methyltransferase, DNMT3B appears to act as a regulator linking local DNA methylation with chromatin organization and transcriptional stability.
We show that DNMT3B overexpression in AML is not mutation-driven but enhancer-mediated. Epigenomic profiling revealed DNase I hypersensitive sites enriched for H3K27ac and H3K4me3 across the DNMT3B promoter and gene body, consistent with active cis-regulatory regions that are most pronounced in CEBPA- and NPM1-mutated AML datasets, suggesting subtype-associated enhancer remodeling at the DNMT3B locus; however, downstream chromatin and transcriptional consequences were examined in representative DNMT3B-high AML models rather than systematically across genetically defined AML subtypes. SRF emerged as a candidate transcriptional regulator at the DNMT3B promoter, and SRF has been implicated in transcriptional and cytoskeletal programs that overlap with DNMT3B-associated regulatory networks in leukemia models, although direct SRF–DNMT3B regulatory links remain to be demonstrated [8]. In CEBPA- and NPM1-mutant AML, SRF motif enrichment may reflect broader enhancer rewiring involving lineage-defining transcription factors, although direct cooperative regulation of DNMT3B remains to be demonstrated and will require targeted perturbation and chromatin-occupancy studies [10]. Together, these findings support a model in which DNMT3B overexpression is epigenetically driven through transcription factor–mediated enhancer–promoter engagement, rather than through genetic alteration.
NanA treatment was associated with a modest, dose-dependent reduction in DNMT3B protein levels in AML cells (Figure 2D), consistent with reports in hepatocellular carcinoma and other models, but differing from A549 lung carcinoma cells where DNMT3B levels remained unchanged [25,26]. Biochemical studies have reported preferential inhibition of DNMT3B catalytic activity by Nanaomycin A, although its quinone scaffold can also induce oxidative stress- and context-dependent cytotoxicity, suggesting that NanA should be interpreted as a DNMT3B-directed functional probe rather than a fully selective inhibitor [25]. Variability in cellular redox buffering or mitochondrial stress tolerance may therefore contribute to differential NanA sensitivity independent of DNMT3B abundance, which could partly explain why DNMT3B-high AML models do not uniformly display increased NanA susceptibility. Because Nanaomycin A is a quinone-based molecule reported to induce oxidative and mitochondrial stress responses, some transcriptional or phenotypic effects may reflect DNMT3B-independent cytotoxic mechanisms; accordingly, our interpretation emphasizes loci showing concordant DNA methylation, chromatin accessibility, and gene-expression changes consistent with epigenetic regulation. DNMT3B inhibition was associated with focal rather than global hypomethylation, preferential remodeling of promoter- and enhancer-like regions, and heterogeneous transcriptional responses consistent with chromatin-context–dependent DNMT3B function; however, these associations are correlative, and we cannot exclude that accessibility changes reflect indirect effects such as replication stress, cell-cycle perturbation, or broader epigenomic remodeling rather than direct DNMT3B occupancy at these loci. Motif enrichment for NRF1 and E2F1 within DNMT3Bi-accessible regions suggests that DNMT3B influences transcription factor occupancy and cell-cycle transcriptional wiring rather than operating as a uniform gene silencer [27,28]. Consistent with this, mass spectrometry showed a trend toward increased NRF1 protein abundance following NanA (log2FC = +1.55; padj = 0.135; Table S7). While this supports NRF1 activation, direct occupancy studies (e.g., CUT&RUN or ChIP-seq) will be required to establish causality. Because NanA has been reported to inhibit DNMT3B catalytic activity, the observed epigenomic effects are consistent with catalytic inhibition; however, DNMT3B-independent cytotoxicity and secondary stress responses, particularly at later time points, cannot be excluded. Complementary genetic DNMT3B perturbation studies, including our prior work demonstrating increased venetoclax sensitivity following DNMT3B knockout in DNMT3B-high AML models [25,29], support DNMT3B-associated epigenetic regulation while underscoring that Nanaomycin A should be interpreted primarily as a functional probe rather than a fully selective inhibitor.
DNMT3Bi produced subtle yet coordinated methylation shifts sufficient to drive transcriptional and proteomic remodeling. Genes such as CDK6, CENPE, and GSK3B exhibited enhancer demethylation, chromatin opening, and protein up-regulation, implicating DNMT3B in maintaining mitotic-fidelity checkpoints [30]. Contrary to classical G1 arrest, DNMT3Bi caused G1 depletion with S-phase accumulation, consistent with premature G1 exit, replication stress, and G2/M checkpoint activation. Increased CDK6 expression following DNMT3B inhibition likely reflects an adaptive stress response rather than sustained proliferative signaling, as it occurs in the setting of checkpoint engagement and progressive apoptosis and parallels prior observations that epigenetic stress can induce compensatory upregulation of cell-cycle regulators without restoring durable proliferation. In this context, DNMT3B-directed perturbation appears to destabilize replication-associated transcriptional programs and checkpoint control mechanisms, rather than simply slowing cell-cycle progression [31,32]. This interpretation is consistent with broader epigenetic reprogramming following DNMT3B perturbation and does not establish these genes as direct DNMT3B targets; rather, candidates such as CDK6, CENPE, and GSK3B likely represent downstream effectors of DNMT3B-associated chromatin remodeling with potential therapeutic relevance that warrants functional validation in additional models (summarized schematically in Figure 7).
Figure 7. Model of DNMT3B-associated epigenetic remodeling and apoptotic sensitization in AML. Schematic summarizing relationships observed in DNMT3B-high AML models. DNMT3B overexpression is associated with enhancer-enriched chromatin features. Pharmacologic DNMT3B perturbation (Nanaomycin A) induces focal methylation remodeling and altered transcriptional programs, accompanied by premature G1 exit, S-phase accumulation, and replication stress. Transcriptional profiling shows induction of BH3-only sensitizers (PMAIP1/NOXA, BBC3/PUMA), repression of BCL2 mRNA, and compensatory upregulation of MCL1 and BCL2L1 (BCL-XL) in the combination setting. BAX and BAK1 transcripts decrease despite increased protein abundance following DNMT3B inhibition, consistent with post-transcriptional regulation. These coordinated changes are consistent with apoptotic priming that enhances sensitivity to BCL2 inhibition (venetoclax), culminating in mitochondrial outer membrane permeabilization (MOMP). The model reflects associations observed in representative AML cell lines and does not imply direct DNMT3B occupancy at depicted loci. Created in BioRender. Roy Choudhury, S. (2026) https://BioRender.com/7233l90.
To dissect the apoptotic circuitry, we profiled BCL2 family members under NanA, VEN, and combination treatment. NanA + VEN prominently up-regulated BH3-only sensitizers PMAIP1 (NOXA) and BBC3 (PUMA), while BCL2 was transcriptionally repressed across all arms, consistent with enhanced apoptotic priming. MCL1 and BCL2L1 (BCL-XL) were significantly induced at the transcript level in the combination group, mirroring compensatory survival programs commonly associated with evolving venetoclax resistance in AML and other myeloid neoplasms. BAX and BAK1 were transcriptionally suppressed by NanA (Figure 6C,D) but up-regulated at the protein level (log2FC = +1.29 for BAX; +2.77 for BAK1; Table S7), suggesting post-transcriptional regulation of effector proteins downstream of DNMT3B perturbation. Collectively, these data are consistent with an apoptotic priming model in which DNMT3B inhibition lowers the apoptotic threshold through BH3-only gene induction and BCL2 repression, although direct functional confirmation of mitochondrial priming (e.g., BH3 profiling or protein-level validation across additional models) was beyond the scope of this study; early induction of MCL1/BCL-XL may represent an adaptive response that could limit durability of venetoclax-based regimens [33,34,35]. These findings suggest that combinatorial approaches targeting MCL1 or BCL-XL could improve durability of DNMT3B–BCL2 co-targeting strategies; however, the observed enhancement of venetoclax sensitivity was demonstrated in DNMT3B-high AML cell line models, and evaluation in primary AML samples and in vivo systems will be required to establish translational generalizability.
Clinically used hypomethylating agents (HMAs) such as azacitidine and decitabine also modulate DNA methylation and improve outcomes when combined with venetoclax in older or unfit AML patients, and preclinical data suggest that HMAs can alter chromatin accessibility and enhancer activity, primarily through DNMT trapping following incorporation into DNA (and RNA for azacitidine) rather than targeted DNMT3B inhibition [33,36]. Our data suggest that DNMT3B-directed perturbation may involve more focal remodeling of enhancer-like elements compared with the broader effects of conventional hypomethylating agents, although direct comparative studies are needed [37], providing a mechanistic rationale for exploring DNMT3B-selective strategies—alone or in combination with BCL2 inhibitors—as a complementary approach to pan-hypomethylating therapy [38].
Beyond AML, DNMT3B overexpression and focal hypomethylation signatures recur in colorectal, breast, and hepatocellular carcinomas, and elevated DNMT3B or DNMT3B target-gene expression associates with adverse outcomes in several solid tumors, underscoring broader oncogenic relevance [26,39]. In TCGA breast cancer, for example, patients with below-median DNMT3B expression (RSEM < 114.28) showed significantly improved overall survival (log-rank p = 0.033; Wilcoxon p = 0.005; HR = 1.42, 95% CI 1.03–1.95; Figure S8, Table S11), consistent with a context in which DNMT3B acts predominantly as an oncogenic factor. In contrast, ectopic Dnmt3b expression can delay leukemogenesis and impair leukemia-initiating capacity in KMT2A- and Myc/Bcl2-driven mouse models, highlighting a tumor-suppressive role in specific stem-cell contexts. Collectively, these observations indicate that DNMT3B function is highly context dependent, reflecting differentiation state, cooperating lesions, and chromatin landscape rather than intrinsic oncogenic or tumor-suppressive activity. Although DNMT3B is unlikely to act as a single determinant, our data support its consideration as a context-dependent prognostic biomarker and potential therapeutic target. Clinically, DNMT3B-high AML subsets could be identified using transcriptomic profiling, quantitative PCR, or immunoblot-based assays, and available cohort analyses suggest elevated DNMT3B expression in defined high-risk AML subtypes rather than uniformly across AML. The stability of DNMT3B expression during disease evolution or relapse remains incompletely characterized and warrants longitudinal study. Importantly, while CD34+ hematopoietic progenitors show low baseline DNMT3B expression in our analyses, the consequences of pharmacologic DNMT3B inhibition under hematopoietic stress conditions remain uncertain, underscoring the need for careful toxicity evaluation in translational development.

5. Conclusions

In summary, our data support DNMT3B as a context-dependent epigenetic regulator of chromatin-governed survival programs in AML. Its overexpression appears primarily enhancer driven rather than mutation mediated, and its inhibition reconfigures focal methylation at key cis-regulatory elements, remodels transcriptional and proteomic modules governing proliferation and apoptosis, and primes AML cells for BCL2 blockade. This DNMT3B-dependent apoptotic priming—characterized by BH3-only activation, BCL2 repression, and BAX/BAK1 stabilization with compensatory MCL1/BCL-XL induction—provides a mechanistic rationale for DNMT3B-targeted epigenetic co-therapy. These findings offer a conceptual framework for extending this therapeutic paradigm to additional malignancies exhibiting aberrant DNMT3B activation.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cancers18050723/s1, Figure S1: DNMT3B expression across normal hematopoietic cells and AML molecular subtypes (MILE cohort); Figure S2: Intra-subtype survival analysis of DNMT3B expression in NPM1- and CEBPA-mutant AML; Figure S3: SRF motif enrichment within the DNMT3B promoter DHS-I region in AML; Figure S4: Uncropped immunoblots corresponding to Figure 2B; Figure S5: Uncropped immunoblots corresponding to Figure 2D; Figure S6: Predicted transcription factor binding motifs enriched in DNMT3Bi-induced open chromatin regions; Figure S7: Pathway enrichment analysis of differentially expressed genes upon DNMT3B inhibition in KG1A cells; Figure S8: Lower DNMT3B expression is associated with improved overall survival in breast cancer [13,14,15,40,41,42,43,44,45,46,47]. Table S1: Overall survival (OS) data of acute myeloid leukemia (AML) patients from the BeatAML 2.0 study, stratified by DNMT3B expression levels (log2 RNA-seq RPKM values). Table S2: DNMT3B expression levels (log2 RNA-seq RPKM values) across the major subtypes of acute myeloid leukemia in the BeatAML 2.0 study cohort, as analyzed in the current study. Table S3: Enrichment Scores of DNase I Hypersensitive Sites (DHS-I–IV) at the DNMT3B Locus Across Primary AML Samples from Bone Marrow and Peripheral Blood (BLUEPRINT Cohort). Table S4: The chromatin state at the DNMT3B locus was characterized using ChIP-Seq data from AML patient-derived samples obtained through the BLUEPRINT Epigenome Consortium. Table S5: RNA sequencing data showing the differentially expressed genes (normalized expression, log2 fold change) that are upregulated or downregulated in the KG1A cell line treated with 50 nM of Nanaomycin A, a selective DNMT3B inhibitor, compared to DMSO-treated KG1A cells used as the control. Table S6: Pathway Enrichment Analysis of Differentially Expressed Genes from RNA-Seq in Nanaomycin A–Treated KG1A Cells. Table S7: Proteomics data showing the differentially expressed proteins (normalized exclusive intensities, log2 fold change) that are upregulated or downregulated in the KG1A cell line treated with 50 nM of Nanaomycin A, a selective DNMT3B inhibitor, compared to DMSO-treated KG1A cells used as the control. Table S8: Integration of differentially methylated DNA with coinciding accessible chromatin sites in KG1A cells treated with 50 nM of Nanaomycin A. Table S9: Primers used in qPCR to determine the expression of anti-apoptotic BCL2 family members (pro-survival), pro-apoptotic “BH3-only” activators and sensitizers, and pro-apoptotic effectors. Table S10: DNMT3B Expression and CRISPR Gene Effect Scores Across AML Cell Lines from DepMap 25Q2 (Chronos). Table S11: Survival Analysis of DNMT3B Expression in Breast Cancer Patients (n = 1091) from the TCGA Cohort.

Author Contributions

S.R.C. conceived the study, designed the experimental strategy, and wrote the manuscript. A.C. analyzed patient-derived epigenomic datasets from the BLUEPRINT and ENCODE consortia. P.B. performed cell viability assays, gene expression analysis, and apoptosis profiling. K.S. generated the ATAC-seq libraries, analyzed chromatin accessibility data, and integrated it with DNA methylation profiles. A.C. and P.B. contributed equally to this study. All authors have read and agreed to the published version of the manuscript.

Funding

This study is primarily supported by the U.S. Department of Defense RCRP program (HT9425-24-1-0932), awarded to SRC. Additional support is provided by the National Institute of General Medical Sciences (NIGMS, 2P20GM121293), UAMS Team Science Award (AWD-104823) award, and the Marion B. Lyon New Scientist Development Award (LYON5069) awarded to SRC. SRC also receives support from UAMS start-up funds, WPRCI vouchers, and the Arkansas Bioscience Institute (ABIRC4426). Funding for the project was provided by the European Union’s Seventh Framework Program (FP7/2007-2013) under grant agreement no 282510 BLUEPRINT.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

The ATAC-seq data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE295001. The DNA methylation (DNAm) data from NanA-treated KG1A cells are available at GEO under accession number GSE296997. All additional omics datasets and supporting materials are provided in the Supplementary Tables.

Acknowledgments

We thank the UAMS Proteomics Core Facility for mass spectrometry support, the UAMS Flow Cytometry Core for assistance with cell sorting and apoptosis analysis, and the UAMS Genomics Core for RNA-seq and DNA methylation profiling. We also acknowledge the CTPR-COBRE Genomics Core at ACRI for supporting EPIC v2.0 data analysis. This study makes use of data generated by the Blueprint Consortium. A full list of the investigators who contributed to the generation of the data is available from www.blueprint-epigenome.eu.

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:
AMLAcute myeloid leukemia
DNAmDNA methylation
DNMTDNA methyltransferase
DNMT1DNA methyltransferase 1
DNMT3ADNA methyltransferase 3A
DNMT3BDNA methyltransferase 3B
DNMT3BiDNMT3B inhibitor
MLL-rMixed lineage leukemia-rearranged
KMT2A-rLysine methyltransferase 2A-rearranged
CN-AMLCytogenetically normal acute myeloid leukemia
FLT3-ITDFMS-like tyrosine kinase 3 internal tandem duplication
DHSDNase I hypersensitive site
CRECis-regulatory element
HSCHematopoietic stem cell
ATAC-seqAssay for Transposase-Accessible Chromatin using sequencing
RNA-seqRNA sequencing
DEGDifferentially expressed gene
DMGDifferentially methylated gene
Log2FCLog base 2-fold change
TSSTranscription start site
GSEAGene Set Enrichment Analysis
MSigDBMolecular Signatures Database
TCGAThe Cancer Genome Atlas
CCLECancer Cell Line Encyclopedia
DMSODimethyl sulfoxide
FDRFalse discovery rate
DIAData-independent acquisition
AGCAutomatic gain control
VSNVariance stabilizing normalization
RLRRobust linear regression
EGSEAEnsemble of Gene Set Enrichment Analyses
TFTranscription factor
PIPropidium iodide
FITCFluorescein isothiocyanate
VENVenetoclax
NanANanaomycin A
qPCRQuantitative polymerase chain reaction
ATP5BATP synthase, H+ transporting, mitochondrial F1 complex, beta subunit
ODOptical density
IC50Half-maximal inhibitory concentration
SRFSerum Response Factor

References

  1. Kulis, M.; Esteller, M. DNA methylation and cancer. Adv. Genet. 2010, 70, 27–56. [Google Scholar] [PubMed]
  2. Esteller, M.; Herman, J.G. Cancer as an epigenetic disease: DNA methylation and chromatin alterations in human tumours. J. Pathol. A J. Pathol. Soc. Great Br. Irel. 2002, 196, 1–7. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  3. Jurkowska, R.Z.; Jurkowski, T.P.; Jeltsch, A. Structure and function of mammalian DNA methyltransferases. Chembiochem 2011, 12, 206–222. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  4. Bestor, T.H. The DNA methyltransferases of mammals. Hum. Mol. Genet. 2000, 9, 2395–2402. [Google Scholar] [CrossRef] [Scilit]
  5. Ley, T.J.; Ding, L.; Walter, M.J.; McLellan, M.D.; Lamprecht, T.; Larson, D.E.; Kandoth, C.; Payton, J.E.; Baty, J.; Welch, J.; et al. DNMT3A mutations in acute myeloid leukemia. N. Engl. J. Med. 2010, 363, 2424–2433. [Google Scholar] [CrossRef] [Scilit]
  6. Hayette, S.; Thomas, X.; Jallades, L.; Chabane, K.; Charlot, C.; Tigaud, I.; Gazzo, S.; Morisset, S.; Cornillet-Lefebvre, P.; Plesa, A.; et al. High DNA methyltransferase DNMT3B levels: A poor prognostic marker in acute myeloid leukemia. PLoS ONE 2012, 7, e51527. [Google Scholar] [CrossRef] [Scilit]
  7. Chen, T.; Ueda, Y.; Dodge, J.E.; Wang, Z.; Li, E. Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b. Mol. Cell Biol. 2003, 23, 5594–5605. [Google Scholar] [CrossRef] [Scilit]
  8. Schulze, I.; Rohde, C.; Scheller-Wendorff, M.; Bäumer, N.; Krause, A.; Herbst, F.; Riemke, P.; Hebestreit, K.; Tschanter, P.; Lin, Q.; et al. Increased DNA methylation of Dnmt3b targets impairs leukemogenesis. Blood 2016, 127, 1575–1586. [Google Scholar] [CrossRef] [Scilit]
  9. Niederwieser, C.; Kohlschmidt, J.; Volinia, S.; Whitman, S.P.; Metzeler, K.H.; Eisfeld, A.K.; Maharry, K.; Yan, P.; Frankhouser, D.; Becker, H.; et al. Prognostic and biologic significance of DNMT3B expression in older patients with cytogenetically normal primary acute myeloid leukemia. Leukemia 2015, 29, 567–575. [Google Scholar] [CrossRef] [Scilit]
  10. Monteferrario, D.; Noordermeer, S.M.; Bergevoet, S.; Huls, G.; Jansen, J.H.; van der Reijden, B.A. High DNA-methyltransferase 3B expression predicts poor outcome in acute myeloid leukemia, especially among patients with co-occurring NPM1 and FLT3 mutations. Blood Cancer J. 2014, 4, e233. [Google Scholar] [CrossRef] [Scilit]
  11. Duymich, C.E.; Charlet, J.; Yang, X.; Jones, P.A.; Liang, G. DNMT3B isoforms without catalytic activity stimulate gene body methylation as accessory proteins in somatic cells. Nat. Commun. 2016, 7, 11453. [Google Scholar] [CrossRef] [Scilit]
  12. Wong, K.K.; Lawrie, C.H.; Green, T.M. Oncogenic Roles and Inhibitors of DNMT1, DNMT3A, and DNMT3B in Acute Myeloid Leukaemia. Biomark. Insights 2019, 14, 1177271919846454. [Google Scholar] [CrossRef] [Scilit]
  13. Bottomly, D.; Long, N.; Schultz, A.R.; Kurtz, S.E.; Tognon, C.E.; Johnson, K.; Abel, M.; Agarwal, A.; Avaylon, S.; Benton, E.; et al. Integrative analysis of drug response and clinical outcome in acute myeloid leukemia. Cancer Cell 2022, 40, 850–864.e9. [Google Scholar] [CrossRef] [Scilit]
  14. Yi, G.; Wierenga, A.T.J.; Petraglia, F.; Narang, P.; Janssen-Megens, E.M.; Mandoli, A.; Merkel, A.; Berentsen, K.; Kim, B.; Matarese, F.; et al. Chromatin-Based Classification of Genetically Heterogeneous AMLs into Two Distinct Subtypes with Diverse Stemness Phenotypes. Cell Rep. 2019, 26, 1059–1069.e6. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  15. Marcotte, R.; Sayad, A.; Brown, K.R.; Sanchez-Garcia, F.; Reimand, J.; Haider, M.; Virtanen, C.; Bradner, J.E.; Bader, G.D.; Mills, G.B.; et al. Functional genomic landscape of human breast cancer drivers, vulnerabilities, and resistance. Cell 2016, 164, 293–309. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  16. Frett, B.; Stephens, K.E.; Koss, B.; Melnyk, S.; Farrar, J.; Saha, D.; Roy Choudhury, S. Enhancer-activated RET confers protection against oxidative stress to KMT2A-rearranged acute myeloid leukemia. Cancer Sci. 2024, 115, 963–973. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  17. Zambelli, F.; Pesole, G.; Pavesi, G. Pscan: Finding over-represented transcription factor binding site motifs in sequences from co-regulated or co-expressed genes. Nucleic Acids Res. 2009, 37, W247–W252. [Google Scholar] [CrossRef] [Scilit]
  18. de Koning, C.B.; Rousseau, A.L.; van Otterlo, W.A.L. Modern methods for the synthesis of substituted naphthalenes. Tetrahedron 2003, 59, 7–36. [Google Scholar] [CrossRef] [Scilit]
  19. Youle, R.J.; Strasser, A. The BCL-2 protein family: Opposing activities that mediate cell death. Nat. Rev. Mol. Cell Biol. 2008, 9, 47–59. [Google Scholar] [CrossRef] [Scilit]
  20. Tait, S.W.G.; Green, D.R. Mitochondria and cell death: Outer membrane permeabilization and beyond. Nat. Rev. Mol. Cell Biol. 2010, 11, 621–632. [Google Scholar] [CrossRef] [Scilit]
  21. Lomonosova, E.; Chinnadurai, G. BH3-only proteins in apoptosis and beyond: An overview. Oncogene 2008, 27, S2–S19. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  22. Vogler, M.; Braun, Y.; Smith, V.M.; Westhoff, M.-A.; Pereira, R.S.; Pieper, N.M.; Anders, M.; Callens, M.; Vervliet, T.; Abbas, M.; et al. The BCL2 family: From apoptosis mechanisms to new advances in targeted therapy. Signal Transduct. Target. Ther. 2025, 10, 91. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  23. Stevens, B.M.; Jones, C.L.; Pollyea, D.A.; Culp-Hill, R.; D’Alessandro, A.; Winters, A.; Krug, A.; Abbott, D.; Goosman, M.; Pei, S.; et al. Fatty acid metabolism underlies venetoclax resistance in acute myeloid leukemia stem cells. Nat. Cancer 2020, 1, 1176–1187. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  24. Ramsey, H.E.; Fischer, M.A.; Lee, T.; Gorska, A.E.; Arrate, M.P.; Fuller, L.; Boyd, K.L.; Strickland, S.A.; Sensintaffar, J.; Hogdal, L.J.; et al. A Novel MCL1 Inhibitor Combined with Venetoclax Rescues Venetoclax-Resistant Acute Myelogenous Leukemia. Cancer Discov. 2018, 8, 1566–1581. [Google Scholar] [CrossRef] [Scilit]
  25. Kuck, D.; Caulfield, T.; Lyko, F.; Medina-Franco, J.L. Nanaomycin A Selectively Inhibits DNMT3B and Reactivates Silenced Tumor Suppressor Genes in Human Cancer Cells. Mol. Cancer Ther. 2010, 9, 3015–3023. [Google Scholar] [CrossRef] [Scilit]
  26. Lai, S.C.; Su, Y.T.; Chi, C.C.; Kuo, Y.C.; Lee, K.F.; Wu, Y.C.; Lan, P.-C.; Yang, M.-H.; Chang, T.-S.; Huang, Y.-H. DNMT3b/OCT4 expression confers sorafenib resistance and poor prognosis of hepatocellular carcinoma through IL-6/STAT3 regulation. J. Exp. Clin. Cancer Res. 2019, 38, 474. [Google Scholar] [CrossRef] [Scilit]
  27. Kaluscha, S.; Domcke, S.; Wirbelauer, C.; Stadler, M.B.; Durdu, S.; Burger, L.; Schübeler, D. Evidence that direct inhibition of transcription factor binding is the prevailing mode of gene and repeat repression by DNA methylation. Nat. Genet. 2022, 54, 1895–1906. [Google Scholar] [CrossRef] [Scilit]
  28. Kent, L.N.; Leone, G. The broken cycle: E2F dysfunction in cancer. Nat. Rev. Cancer 2019, 19, 326–338. [Google Scholar] [CrossRef] [Scilit]
  29. Choudhury, S.R.; Kaushal, A.; Biswas, P.; Padilla, C.; Sarthy, J.F.; Chavan, A.; Gonzalez, G.A.; Meshinchi, S.; Farrar, J.E. Transcriptional rewiring by enhancer methylation in CBFA2T3-GLIS2-driven pediatric acute megakaryoblastic leukemia. Genes. Dis. 2026, 13, 101843. [Google Scholar] [CrossRef] [Scilit]
  30. Sherr, C.J.; Roberts, J.M. CDK inhibitors: Positive and negative regulators of G1-phase progression. Genes. Dev. 1999, 13, 1501–1512. [Google Scholar] [CrossRef] [Scilit]
  31. Qie, S.; Diehl, J.A. Cyclin D1, cancer progression, and opportunities in cancer treatment. J. Mol. Med. 2016, 94, 1313–1326. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  32. Tadesse, S.; Yu, M.; Kumarasiri, M.; Le, B.T.; Wang, S. Targeting CDK6 in cancer: State of the art and new insights. Cell Cycle 2015, 14, 3220–3230. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  33. DiNardo, C.D.; Pratz, K.; Pullarkat, V.; Jonas, B.A.; Arellano, M.; Becker, P.S.; Frankfurt, O.; Konopleva, M.; Wei, A.H.; Kantarjian, H.M.; et al. Venetoclax combined with decitabine or azacitidine in treatment-naive, elderly patients with acute myeloid leukemia. Blood 2019, 133, 7–17. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  34. Zhang, Q.; Han, L.; Shi, C.; Pan, R.; Ma, M.C.J.; Ryan, J.; E Kurtz, S.; Mu, H.; Ruvolo, V.; Ma, H.; et al. Upregulation of MAPK/MCL-1 Maintaining Mitochondrial Oxidative Phosphorylation Confers Acquired Resistance to BCL-2 Inhibitor Venetoclax in AML. Blood 2016, 128, 101. [Google Scholar] [CrossRef] [Scilit]
  35. Liu, J.; Chen, Y.; Yu, L.; Yang, L. Mechanisms of venetoclax resistance and solutions. Front. Oncol. 2022, 12, 1005659. [Google Scholar] [CrossRef] [Scilit]
  36. Bouligny, I.M.; Murray, G.; Doyel, M.; Patel, T.; Boron, J.; Tran, V.; Gor, J.; Hang, Y.; Alnimer, Y.; Zacholski, K.; et al. Venetoclax with decitabine or azacitidine in the first-line treatment of acute myeloid leukemia. EJHaem 2023, 4, 381–392. [Google Scholar] [CrossRef] [Scilit]
  37. Benetatos, L.; Vartholomatos, G. Enhancer DNA methylation in acute myeloid leukemia and myelodysplastic syndromes. Cell Mol. Life Sci. 2018, 75, 1999–2009. [Google Scholar] [CrossRef] [Scilit]
  38. Kim, D.J. The Role of the DNA Methyltransferase Family and the Therapeutic Potential of DNMT Inhibitors in Tumor Treatment. Curr. Oncol. 2025, 32, 88. [Google Scholar] [CrossRef] [Scilit]
  39. MacKenzie, D.J.; Robertson, N.A.; Rather, I.; Reid, C.; Sendzikaite, G.; Cruickshanks, H.; McBryan, T.; Hodges, A.; Pritchard, C.; Blyth, K.; et al. DNMT3B Oncogenic Activity in Human Intestinal Cancer Is Not Linked to CIMP or BRAFV600E Mutation. iScience 2020, 23, 100838. [Google Scholar] [CrossRef] [Scilit]
  40. Kim, D.; Paggi, J.M.; Park, C.; Bennett, C.; Salzberg, S.L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 2019, 37, 907–915. [Google Scholar] [CrossRef] [Scilit]
  41. Anders, S.; Huber, W. Differential expression analysis for sequence count data. Genome Biol. 2010, 11, R106. [Google Scholar] [CrossRef] [Scilit]
  42. Searle, B.C.; Pino, L.K.; Egertson, J.D.; Ting, Y.S.; Lawrence, R.T.; MacLean, B.X.; Villén, J.; MacCoss, M.J. Chromatogram libraries improve peptide detection and quantification by data independent acquisition mass spectrometry. Nat. Commun. 2018, 9, 5128. [Google Scholar] [CrossRef] [Scilit]
  43. Graw, S.; Tang, J.; Zafar, M.K.; Byrd, A.K.; Bolden, C.; Peterson, E.C.; Byrum, S.D. proteiNorm—A User-Friendly Tool for Normalization and Analysis of TMT and Label-Free Protein Quantification. ACS Omega 2020, 5, 25625–25633. [Google Scholar] [CrossRef] [Scilit]
  44. Huber, W.; von Heydebreck, A.; Sültmann, H.; Poustka, A.; Vingron, M. Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002, 18, S96–S104. [Google Scholar] [CrossRef] [Scilit]
  45. Bolstad, B. R Package, version 1.46.0. PreprocessCore: A collection of pre-processing functions. R Foundation for Statistical Computing: Vienna, Austria, 2019.
  46. Ritchie, M.E.; Phipson, B.; Wu, D.; Hu, Y.; Law, C.W.; Shi, W.; Smyth, G.K. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015, 43, e47. [Google Scholar] [CrossRef] [Scilit]
  47. Chawade, A.; Alexandersson, E.; Levander, F. Normalyzer: A Tool for Rapid Evaluation of Normalization Methods for Omics Data Sets. J. Proteome Res. 2014, 13, 3114–3120. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
Two-Stage Surgery in Patients with Diffuse Glioma—Indications, Implications and Outcome
Previous
Improving Lung Cancer Screening Selection: A Comparative Analysis of Risk Models and Traditional Criteria in a Western European General Population
Next