A plant DNMT3 induces non-CG methylation and affects gene expression in human cells

Rotem Domb; Karina G Heskiau; Aviva Katz; Rafael Yaari; Rachel Zamostiano; Keith D Harris; Eran Bacharach; Nir Ohad; Assaf Zemach

doi:10.1016/j.isci.2026.117542

. 2026 Sep 15;29(10):117542. doi: 10.1016/j.isci.2026.117542

A plant DNMT3 induces non-CG methylation and affects gene expression in human cells

Rotem Domb ^1,⁴, Karina G Heskiau ^1,⁴, Aviva Katz ¹, Rafael Yaari ¹, Rachel Zamostiano ², Keith D Harris ¹, Eran Bacharach ², Nir Ohad ^1,^∗, Assaf Zemach ^1,^3,^5,^∗∗

PMCID: PMC13595168 PMID: 42775135

Summary

CG methylation is the predominant form of DNA methylation across eukaryotes, whereas non-CG methylation in vertebrates is largely restricted to neurons and embryonic stem cells and is mediated by DNMT3 enzymes. In contrast, plant DNMT3 homologs catalyze extensive non-CG methylation across diverse tissues. To test whether a plant DNMT3 can establish non-CG methylation and influence transcription in human cells, we expressed Physcomitrium patens DNMT3b (PpDNMT3) in HEK293 cells, which lack this modification. PpDNMT3 induced robust, genome-wide non-CG methylation in transiently and stably transfected cells, in a manner dependent on an intact catalytic domain. PpDNMT3 expression was associated with altered gene expression that coincided with methylation changes. Genes showing combined CG and non-CG hypermethylation within promoters and gene bodies exhibited the strongest transcriptional repression. These results demonstrate that plant DNMT3 enzymes can reprogram human epigenomes and provide a versatile tool for modulating DNA methylation in cell types lacking endogenous non-CG methylation.

Keywords: DNA methylation, non-CG methylation, non-CpG methylation, MeCP2, neurons

Graphical abstract

Highlights

•
PpDNMT3 induces robust non-CG methylation in human cells
•
PpDNMT3 preferentially methylates non-CG CAC sites
•
PpDNMT3-induced non-CG methylation is enriched at H3K9me3-marked regions
•
Combined CG and non-CG hypermethylation is associated with stronger gene repression

Biological sciences; Molecular biology; Molecular mechanism of gene regulation; Neuroscience; Plant biology

Introduction

Cytosine DNA methylation is a robust epigenetic mark in eukaryotes that regulates transcription and development.¹^,² Most DNA methylation in animal cells occurs at CpG dinucleotides, referred to as CG methylation or mCG.³^,⁴^,⁵ CG sites are symmetrical in the double helix context, a feature that plays an important role in the inheritance of cytosine methylation following DNA replication. At a relatively lower efficiency, cytosine methylation can also be targeted to asymmetric non-CG sites, referred to as non-CG or CH methylation (H = A, T, or C).⁶ In vertebrates, non-CG methylation is deficient in most cell types, but elevated in particular cells, such as embryonic stem cells (ESCs) and neurons.³^,⁶^,⁷^,⁸^,⁹^,¹⁰^,¹¹^,¹² Unlike CGs, which are robustly methylated across the genome, non-CG methylation levels are substantially lower, and its distribution varies across the genome.¹³ For example, non-CG methylation was found to be enriched in particular genomic regions, such as topologically associating domains (TADs).¹¹^,¹⁴ In mammalian genomes, while non-CG sites typically exhibit a low methylation level of less than 5%, their abundance is roughly 20 times higher than that of CG sites, meaning that in certain human cells, approximately 25% of all methylated cytosines are located at non-CG sites.⁸^,¹²

Both CG and non-CG methylation were shown to be involved in gene expression. A recent study suggested that histone H3 lysine 36 dimethylated (H3K36me2) controls non-CG methylation at TADs and focal expression points.¹⁵ In neurons, non-CG methylation was found to be bound by the methyl-CpG binding protein 2 (MeCP2), which regulates the expression of genes necessary for proper neuronal development and brain function.¹⁴^,¹⁶^,¹⁷^,¹⁸^,¹⁹^,²⁰ The level of mCH at gene regulatory elements was found to positively correlate with the degree of MeCP2-mediated gene repression, and depletion of non-CG methylation or MeCP2 has been associated with neurodevelopmental disorders, e.g., Rett syndrome.²¹^,²² In a plasmid reporter assay, non-CG methylation was sufficient to repress gene transcription in vitro as effectively as CG methylation.¹² Aberrant non-CG methylation is a common feature of additional disorders such as cancer.²³^,²⁴^,²⁵^,²⁶

Unlike animals, non-CG methylation in plants is more robust and plays a major role in gene and transposon regulation.⁵^,²⁷^,²⁸^,²⁹ Non-CG methylation in plants is conducted by three types of methylases: chromomethylases (CMTs) that evolved from DNMT1, and two types of DNMT3 proteins. Domain-rearranged methylases (DRMs) evolved from DNMT3, and, as the name suggests, their catalytic domains have been rearranged.³⁰ Plant DRMs function as part of the plant-specific RNA-directed DNA methylation (RdDM) pathway.³¹ Some plants encode orthologous DNMT3 methylases with a linear catalytic domain that function independently of the RdDM pathway.³²^,³³ Plant DNMT3 orthologs became extinct in flowering plants. In contrast, in the moss plant Physcomitrium patens, DNMT3b (PpDNMT3) acts as a robust non-CG methylase responsible for 95% of CHH methylation in the genome.²⁹^,³²

In this study, we aimed to test the ability of PpDNMT3 to function as a non-CG methylase in human cells. Such a tool would allow us to induce non-CG methylation in human and possibly other metazoan cells that are generally depleted of this type of DNA methylation, as well as provide a method to study the influence of newly induced non-CG methylation on gene expression and to utilize it to repair or reshape the epigenome.

Results

PpDNMT3 induces non-CG methylation in human cells

To test the ability of the plant PpDNMT3 to induce non-CG methylation in human cells, we introduced the PpDNMT3b gene fused to GFP, referred to as PpDNMT3, into HEK293T cells. HEK293T cells expressing PpDNMT3 are referred to as PpDNMT3-cells. DNA methylation of control cells (ectopically expressing GFP alone) and of PpDNMT3-cells was profiled by targeted bisulfite sequencing (SureSelectXT Methyl-Seq, Agilent), and the genomic methylation averages were quantified separately for CG and non-CG sequence contexts. Non-CG sites were further separated into CHG and CHH, the two main methylation contexts in plants. The genomic CG methylation average in PpDNMT3 transiently expressing cells (transient PpDNMT3-cells) was found to be similar to that of the control cells (Figure 1A). In contrast, non-CG methylation in the transient PpDNMT3-cells was found to be substantially increased in the CHH context but not in the CHG context (Figure 1A). Methylation in the CHH context increased by 3.4-fold, i.e., to 1.7% versus 0.67% in control cells (Figure 1A). For comparison, we requantified DNA methylation derived from neurons³⁴ and ESC³⁵ at the same targeted genome that was sequenced in this study, and found non-CG methylation at the CHH context to reach about 1.5% and 1%, respectively (Figure 1A). DNA methylation relies on de novo and maintenance activities. Hence, to test for longitudinal methylation effects of PpDNMT3, we next stably expressed PpDNMT3 in human HEK293 cells and profiled their methylome. A stable PpDNMT3-expressing HEK293 cell line (stable PpDNMT3-cells) showed a genomic methylation profile that resembled that of transient PpDNMT3-cells, i.e., a normal genomic average of CG and CHG methylation levels and a substantial increase in CHH methylation, averaging at 2.2% (Figure 1A).

PpDNMT3 induces CHH methylation in HEK293 cells

(A) Genomic average DNA methylation separated into CG (top), CHG (middle), and CHH (bottom) methylation in HEK293 cells expressing PpDNMT3, *Ppdnmt3*, and GFP control constructs. Transient and stable refer to the transient and stable expression of the indicated constructs. Neurons and ESC refer to DNA methylation in human neurons³⁴ and embryonic stem cells.³⁵ Methylation was averaged in the targeted sequenced genome (see STAR Methods).

(B) Genomic methylation averages in the different subcontexts of CG, CHG, and CHH sites in the indicated samples. PpDNMT3, Ppdnmt3, and GFP are from stable expression.

(C) DMR quantification of CG and CHH in the indicated samples. Hyper- and hypo-DMRs refer to hypermethylated and hypomethylated DMRs, respectively.

(D–F) Average methylation change. (D) DMR size and (E) the number of cytosines in DMRs of (F) CHH hyper-DMRs and CG hyper- and hypo-DMRs in transient PpDNMT3-expressing HEK293T cells, and stable HEK293 PpDNMT3, *Ppdnmt3*, and control GFP cells.

(G and H) Venn diagrams demonstrating overlaps between CG hypo-DMRs in PpDNMT3 and *Ppdnmt3* stably expressing HEK293 cells (G) and between CHH hyper-DMRs and CG hypo-DMRs in PpDNMT3 stably expressing HEK293 cells (H).

To check for the direct effect of PpDNMT3 on CHH methylation, we generated a PpDNMT3 mutant (Ppdnmt3) that is mutated in its ENV motif within the catalytic methylase domain³⁶ and stably introduced it into HEK293 cells. Fluorescence imaging showed that Ppdnmt3 exhibited localization similar to wild-type PpDNMT3, with both proteins localized within the nucleus and enriched at dense DNA foci (Figure S1), suggesting that it maintains its original nuclear localization. Methylome profiling of stable Ppdnmt3-cells found its genomic-averaged methylation to be similar to controls in all methylation contexts, CG, CHG, as well as CHH (Figure 1A), suggesting the disruption of its intrinsic CHH methylation activity.

Next, we assessed the methylation specificity of PpDNMT3 within specific CG and non-CG sequence contexts. We did not find a change in the genomic methylation of particular CG or CHG subcontexts in PpDNMT3 versus control cells (Figure 1B). In comparison, CHH methylation induced by PpDNMT3 had a preference for specific subcontexts (Figure 1B). For example, CAC and CCT subcontexts had the highest and lowest CHH methylation levels, with average methylation levels of 4.7% and 1.1%, respectively (Figure 1B). In comparison to PpDNMT3’s methylation preferences, natural non-CG methylation in neurons and ESCs is targeted to a smaller subset of CHH subsequences with negligible representation of CCH sites (Figure 1B). CCC and CAG were the main non-CG methylation contexts that were differentially methylated by PpDNMT3 and human DNMT3s; CCC was methylated in PpDNMT3-cells and not in neurons or ESCs, and vice versa for CAG. The most abundant non-CG methylation subcontext in both neurons and PpDNMT3-cells is the CAC sequence (Figure 1B). The other three CHH subcontexts that were hypermethylated in neurons and ESCs, i.e., CTC, CAA, and CAT, were also hypermethylated in PpDNMT3-cells. PpDNMT3 methylation of the different CHH subcontexts in human cells was generally correlated with PpDNMT3 subcontext methylation in the P. patens genome (Pearson’s r = 0.81, p = 0.008), with some deviations, such as that CAC was not the highest methylated CHH subcontext in the P. patens genome (Figure 1B).

To verify the extent of methylation effects in PpDNMT3-cells, we next identified differentially methylated regions (DMRs) in each sample relative to its relevant control using the Histogram Of MEthylation (HOME) software.³⁷ Hypermethylated and hypomethylated DMRs were designated as hyper-DMRs and hypo-DMRs, respectively. We detected 17,357 and 19,730 CHH hyper-DMRs in the transient and stable PpDNMT3-cells, respectively (Figure 1C). In contrast, only 576 and 687 CHH hypo-DMRs were identified in the transient and stable PpDNMT3-cells, respectively (Figure 1C). In contrast to PpDNMT3-cells, in Ppdnmt3-cells, the number of CHH hyper- and hypo-DMRs was similar to each other (1,329 and 979, respectively), and comparable to the numbers of hyper- and hypo-DMRs found between the control samples (1,237 and 852) (Figure 1C). While the methylation change level of CHH hyper-DMRs is similar among all samples (Figure 1D, left), PpDNMT3’s CHH hyper-DMRs, especially in stable cells, were found to be substantially bigger and to contain more CHH sites than control’s and Ppdnmt3’s (Figures 1E and 1F, left). The median size of the CHH hyper-DMRs in transient and stable PpDNMT3-cells was 196 and 286 bases, respectively, versus 82 and 79 bases in control and Ppdnmt3-cells (Figure 1E, left). CHH hyper-DMRs in transient and stable PpDNMT3-cells contained about 31 and 45 CHH sites, respectively, versus 10 and 7 sites in control and Ppdnmt3-cells, respectively (Figure 1F, left). In total, PpDNMT3’s CHH hyper-DMRs cover over 8.5 Mb, which correspond to about 10% of the targeted-sequenced genome.

To independently validate the CHH DMR pattern, we repeated the analysis using DMRcaller software.³⁸ DMRcaller confirmed pronounced CHH hypermethylation specifically in transient and stable PpDNMT3-cells, detecting 66,140 and 71,994 CHH hyper-DMRs, respectively, compared with 796 and 1,023 CHH hypo-DMRs (Figure S2A). In contrast, Ppdnmt3 and control cells showed similar and substantially lower numbers of CHH hyper- and hypo-DMRs (Figure S2A). Meta- and locus-level analyses showed that HOME-defined CHH DMRs had smaller methylation differences than DMRcaller-defined CHH DMRs (Figure S2B), consistent with the different calling stringencies of the two methods. This difference likely accounts for the higher number of CHH DMRs detected by DMRcaller, while the overall trends were similar to those detected with HOME (Figures 1C and S2A). Although DMRcaller detected approximately 3-fold more CHH hyper-DMRs than HOME, the DMRcaller-defined CHH hyper-DMRs collectively covered only approximately 20% more 50-bp windows (120,323 versus 96,070), indicating that the difference in DMR number was substantially larger than the difference in genomic coverage (Figure S2C).

Overall, these results support the robust non-CG methylation activity of PpDNMT3 in HEK293 cells.

PpDNMT3 affects CG methylation in human cells

HOME DMR analysis of the CG context showed a mixed effect on hypermethylation and hypomethylation. First, transient and particularly stable PpDNMT3-cells showed an enrichment of CG hyper-DMRs (2,863 and 4,762 DMRs, respectively) over those in Ppdnmt3-cells (2,184 DMRs) or controls (1,803 DMRs) (Figure 1C). CG hyper-DMRs in PpDNMT3 were also longer, contained more CG sites, and showed stronger hypermethylation than those in control and Ppdnmt3 (Figures 1D–1F, middle). The median size of CG DMRs in stable PpDNMT3-cells was 155 bp, about 2-fold smaller than that of the corresponding CHH-DMRs (286 bp), and included about eight fewer CG sites than CHHs (5 sites in CG DMRs versus 40 in CHH DMRs). In total, CG hyper-DMRs in stable PpDNMT3-cells covered about 1.1 Mb of the targeted genome, about 8 times less than CHH hyper-DMRs.

Besides hyper-DMRs, stable PpDNMT3- and Ppdnmt3-cells showed an enrichment of CG hypo-DMRs over controls, with 6,878, 9,384, and 1,519 DMRs, respectively (Figure 1C). CG hypo-DMRs in stable PpDNMT3 and Ppdnmt3-cells were also found to be larger, containing more CG sites, and had stronger hypomethylation than CG hypo-DMRs in controls (Figures 1D–1F, right). A significant number of CG hypo-DMRs were common between PpDNMT3 and Ppdnmt3 (hypergeometric p value < 3.289e−73; Figure 1G), suggesting that the functional and mutant PpDNMT3 proteins may be associated with a common indirect effect on CG methylation, potentially through interference with endogenous CG methylation maintenance mechanisms, such as DNMT1 activity. Most CG hypo-DMRs in PpDNMT3-cells do not overlap with CHH hyper-DMRs (Figure 1H), suggesting a distinct mechanism between the two methylation effects. Finally, the minimal enrichment of CG hypo-DMRs in transient over stable PpDNMT3-cells (Figures 1C–1F) suggests a longitudinal suppression effect of CG methylation by PpDNMT3.

To independently validate the CG DMR pattern, we repeated the analysis using DMRcaller. Although DMRcaller identified substantially fewer CG DMRs than HOME and its CG DMRs covered approximately 10-fold less of the genome, it confirmed the relative enrichment of CG hyper-DMRs in stable PpDNMT3-cells (1,408 CG hyper-DMRs versus 126 in stable controls), as well as increased numbers of CG hypo-DMRs in stable PpDNMT3-cells (1,213) and Ppdnmt3-cells (2,186) relative to stable controls (164) (Figures S2A and S2C). Meta- and locus-level analyses showed that HOME-defined CG DMRs had smaller methylation differences than DMRcaller-defined CG DMRs (Figure S2B), consistent with the different calling stringencies of the two methods. This difference likely explains why HOME identified more CG DMRs and a broader genomic coverage, whereas the overall CG methylation trends were similar between HOME and DMRcaller (Figures 1C and S2A). Overall, CG DMR data demonstrate the ability of PpDNMT3 to affect CG methylation in both directions, independently of its non-CG methylation activity.

PpDNMT3 induces non-CG methylation at heterochromatic and genic regions

Next, we tested PpDNMT3’s methylation preferences across various genetic and chromatin features of the genome. In its native P. patens genome, PpDNMT3 specifically methylates transposons; however, in HEK293 cells, PpDNMT3 induced non-CG methylation both in transposons and in genes (Figure 2A). While PpDNMT3’s CHH methylation in HEK293 cells is higher in transposons (3.2%) than in genes, genes still show substantial CHH methylation compared with control (2% vs. 0.95%; Figure 2A). The preference for methylation in transposons is not the result of enrichment in the frequencies of highly methylated CHH subcontexts. For example, the frequencies of the highly methylated CAC and CTC sites (Figure 1B) were found to be similar in genes and transposons (Figure 2B). Alternatively, we found that the individual subcontexts are preferentially hypermethylated in transposons over genes. For example, CAC sites in PpDNMT3-cells reached 4% and 8% methylation levels at genes and transposons, respectively (Figure 2C). In correlation with its methylation preference for transposons, we found PpDNMT3’s CHH methylation, rather than CAC and CTC site frequencies, to be enriched at regions modified with silent chromatin marks such as histone H3 lysine 27 trimethylated (H3K27me3) and, in particular, with that of the canonical heterochromatin, histone H3 lysine 9 trimethylated (H3K9me3) (Figures 2B and 2D).

Similar to native CG methylation, PpDNMT3’s CHH methylation was found to be depleted around the transcription start site (TSS) of genes (Figure 2A). Additionally, we found that PpDNMT3 CHH methylation is positively correlated with the endogenous CG methylation level (Figure 2E), supporting the association between CG methylation and PpDNMT3’s CHH methylation. Despite the general association between endogenous HEK293 genomic CG methylation and PpDNMT3’s induced non-CG methylation, the latter is also detected at lowly methylated or unmethylated CG regions (Figure 2F) and is enriched at regions depleted entirely of CG sites (Figures 2E and 2F). Overall, these results suggest that PpDNMT3 CHH methylation is generally associated with CG methylation but can be targeted to lowly methylated or depleted CG regions.

PpDNMT3 alters gene expression

To assess transcriptional effects related to PpDNMT3 expression and methylation, we profiled the transcriptomes of stable PpDNMT3, Ppdnmt3, and control cells. Principal component analysis (PCA) of gene expression separated the samples by genetic origin (Figure 3A). The two independent transgenic lines in each of the groups were clustered together, suggesting consistency in the differential gene expression in each of the transgenic lines, PpDNMT3, Ppdnmt3, and the control samples (Figure 3A). Differential gene expression analysis detected more differentially expressed genes (DEGs) in PpDNMT3 than in Ppdnmt3, having 1,935 and 829 DEGs, respectively (Figure 3B). Upregulated, downregulated, and non-regulated DEGs have been labeled as up-DEGs, down-DEGs, and non-DEGs, respectively. In both PpDNMT3 and Ppdnmt3-cells, more DEGs were upregulated than downregulated (Figure 3B). Most of the DEGs were protein-coding; however, a substantial portion belonged to long non-coding RNAs (lncRNAs), especially among PpDNMT3 down-DEGs (Figures 3B and S3). Most DEGs in PpDNMT3 and Ppdnmt3 did not overlap with each other, neither in up- nor down-DEGs (Figure 3C), suggesting largely distinct transcriptional effects caused by these functional and nonfunctional PpDNMT3s. Nonetheless, among the 15,431 expressed genes, the observed overlap was substantially greater than expected by chance for both upregulated and downregulated genes, with 6.4- and 6.9-fold enrichment, respectively (hypergeometric test; Figure 3C), suggesting a shared transcriptional effect. Downregulated genes in PpDNMT3-expressing cells, including those associated with hyper-DMRs, were enriched for nervous system development gene ontologies (Figure S4). A similar, although less pronounced, enrichment was observed among Ppdnmt3 down-DEGs (Figure S5).⁶

Combinatorial effect of PpDNMT3’s CG and non-CG methylation on gene expression

To assess the relationship between methylation changes and transcriptional effects driven by PpDNMT3, we correlated DMRs with DEGs. In both PpDNMT3- and Ppdnmt3-cells, CG hypo-DMRs in promoters and gene bodies were significantly enriched for up-DEGs over down- or non-DEGs (Figures 4A and S6). Additionally, the overlap between up-DEGs associated with CG hypo-DMRs in PpDNMT3- and Ppdnmt3-cells was found to be significantly enriched, especially in promoters (4-fold enrichment in promoters; hypergeometric test, p = 1.492e−09; Figure S7), which could explain the significant overall overlap between PpDNMT3 and Ppdnmt3 DEGs (Figure 3B).

CG and CHH hypermethylation in PpDNMT3-cells were found to be correlated with gene suppression (Pearson’s r = 0.31 and 0.39 for CG and CHH; p = 4.3e−23 and 5e−15, respectively); genes suppressed by at least 4-fold showed the most substantial increase in CG and CHH methylation (Figure 4B). Consistent with these associations, CHH hyper-DMRs were enriched in promoters relative to introns and exons, whereas CG hyper- and hypo-DMRs were enriched in promoters relative to other genomic regions (Figure S8). Additionally, we found PpDNMT3’s CG and CHH hyper-DMRs in promoters and gene bodies to be relatively enriched in down-DEGs over up- or non-DEGs (Figures 4A and S6). While gene bodies generally contained more DMRs than promoters (i.e., 30% versus 5%, respectively), the enrichment of DMRs within down-DEGs over other DEGs was found to be 10 times higher in promoters versus gene bodies (Figure 4A). Specifically, the percentage of down-DEGs containing CHH hyper-DMRs increased from 5% to 15% in promoters and from 30% to 40% in gene bodies (Figure 4A). While this result could suggest a stronger role for methylation in promoters than in gene bodies, we found that most promoter DMRs were associated with gene body DMRs (Figure 4C). Moreover, we found that the transcriptional downregulation of DEGs containing hyper-DMRs in promoters and gene bodies was substantially stronger than that of those with either gene body or promoter DMRs (Figure 4D), suggesting a combinatorial transcriptional suppression role for CG and non-CG methylation in promoters and gene bodies.

Next, we examined the relative roles of CG and non-CG methylation in transcriptional regulation. First, we quantified the number of down-DEGs associated with CHH and/or CG hyper-DMRs. We found that about 20% of the downregulated genes were associated with CG and CHH hyper-DMRs either in promoters or gene bodies; the rest had either CG- or CHH-specific hyper-DMRs (Figure 4E). Then, we assessed the level of transcriptional suppression of down-DEGs associated with different methylated DMRs. Down-DEGs associated with CG-specific DMRs (in promoters or gene bodies) were slightly more suppressed than those associated with CHH-specific DMRs (Figure 4E). However, down-DEGs associated with CG and CHH hyper-DMRs showed the strongest transcriptional suppression (Figure 4F). Together, these results suggest a combinatorial effect of CG and non-CG methylation in promoters and gene bodies on gene expression.

Discussion

Our study demonstrates the ability of a plant DNMT3 to induce non-CG methylation in human cells. PpDNMT3 non-CG methylation activity in human cells is composed of conserved and divergent methylation properties from its native P. patens genome. First, in both cellular environments, PpDNMT3 methylated only non-CG sites within the CHH subcontexts, not those within CHGs. However, among the CHH subcontexts, PpDNMT3 preference was only partially correlated between its activity in the plant’s versus the human’s genomes. Additionally, in the human genome, PpDNMT3 CHH methylation was lower than in the plant genome; in the moss plant, PpDNMT3 methylates 15% and 30% of CHH sites in the whole genome and in transposon sequences, respectively, whereas in human cells, PpDNMT3 methylated about 2% and 5% of CHH sites in the sequenced genome and transposons, respectively. Additionally, in the plant genome, PpDNMT3 methylates almost exclusively heterochromatic sequences and hardly any genes, whereas in human cells, its methylation activity is enhanced at heterochromatic sequences but remains significantly targeted at euchromatic genic sequences. Overall, these results imply that PpDNMT3 can methylate non-CG sites in human cells. The similar preferential non-CG methylation of PpDNMT3 in H3K9 trimethylation regions in HEK293 and native moss cells suggests a similar targeting mechanism to that of human and plant chromatin. The relatively smaller non-CG methylation effect of PpDNMT3 in human versus plant cells could be either due to biological factors, such as a missing cofactor or an altered post-translational modification, or technical factors, such as steric hindrance caused by the linked GFP protein.

Some of the methylation features of PpDNMT3 in human cells could be utilized for the editing of non-CG methylation in human cells. For example, PpDNMT3 methylated non-CG sites in HEK293 cells to a level comparable to that found in human ESCs and neurons (i.e., in the lower single-digit methylation percentage). Additionally, its highest methylation efficiency toward the CAC context is similar to that of neurons. PpDNMT3 non-CG methylation of genic sequences is similar to that of neurons and ESCs and can be utilized to influence gene expression.

PpDNMT3 was previously found to induce de novo CG methylation in P. patens.³² In HEK293 cells, CG hypermethylation was enriched in promoters and was largely absent in Ppdnmt3-cells, supporting catalytic dependence. However, the mechanism remains unclear and may involve altered endogenous DNMT1-mediated maintenance methylation; whether PpDNMT3 has intrinsic CG methylation activity in human cells remains to be demonstrated experimentally. Finally, PpDNMT3’s non-CG methylation in HEK293 cells at regions deficient in CG methylation or CG sites can expand the use of DNA methylation beyond, or independently of, CG methylation.

The transcriptional effect of PpDNMT3’s non-CG methylation in HEK293 cells suggests a general transcriptional suppression role. This result correlates with the transcriptional suppression effect of the mCH reader MeCP2 in neurons,¹⁴^,¹⁷^,¹⁸^,³⁹ which is expressed in HEK293 cells.⁴⁰^,⁴¹ Interestingly, PpDNMT3-downregulated genes were enriched for nervous system development-related functions. However, HEK293 cells are not neuronal cells and may not recapitulate neuron-specific regulatory mechanisms. Moreover, a similar, although weaker, enrichment was observed in cells expressing the catalytically inactive Ppdnmt3, limiting the extent to which this enrichment can be attributed specifically to PpDNMT3-induced non-CG methylation. The biological significance of this observation, therefore, remains unclear. Why only a subset of the PpDNMT3’s-induced non-CG methylated genes has been dysregulated in HEK293 cells would require further investigation. However, the stronger transcriptional effect of genes with stronger hypermethylation of CG and non-CG sites targeted to promoters and gene bodies suggests a quantitative and combinatorial mechanism. DNA methylation exerts complex and context-dependent effects on gene regulation that extend beyond transcriptional repression. Although DNA methylation is classically linked to heterochromatin formation and reduced chromatin accessibility, it may also create or stabilize binding sites for specific transcription factors, such as FoxA2.⁴² From this perspective, PpDNMT3-induced non-CG methylation could act as a programmable epigenetic signal that rewires transcriptional networks in a cell-type-specific manner, with outcomes shaped by the transcription factor landscape of each cell.

To conclude, our results verify the ability of PpDNMT3 to induce non-CG methylation, along with accompanying mild CG methylation effects, in human cells that are normally depleted of non-CG methylation, and to modulate gene expression through this activity. These findings open the possibility of using plant DNMT3 methylases to trigger non-CG methylation in various human cell types and possibly in other animal cells. Importantly, such a biotechnological tool would be valuable for further dissecting the relative roles of non-CG methylation in genome regulation across different cells and conditions, as well as for practical applications by enabling the use of non-CG methylation in cells or genomic regions that are normally or abnormally deficient in non-CG methylation. In this context, fusion of PpDNMT3 to catalytically inactive Cas9, or similar programmable DNA-targeting platforms, could provide a targeted epigenome-editing strategy to deposit non-CG methylation at defined genomic loci. Such an approach would enable direct testing of whether non-CG methylation at specific promoters, gene bodies, enhancers, or repetitive elements is sufficient to alter local chromatin state or transcriptional output, thereby extending the utility of plant DNMT3 enzymes from genome-wide methylation tools to locus-specific functional interrogation.

Limitations of the study

This study has several limitations. PpDNMT3 was expressed as a GFP-fusion protein, and GFP may have affected its activity or genomic targeting, potentially contributing to the lower non-CG methylation levels observed in human cells compared with its native P. patens genome. In addition, DNA methylation was profiled using the SureSelectXT Human Methyl-Seq targeted sequencing platform rather than whole-genome sequencing. Because this platform preferentially covers gene-associated regions, whereas PpDNMT3-induced non-CG methylation was strongest in transposon- and heterochromatin-associated regions, the measured methylation levels may underestimate its genome-wide activity. Finally, methylome profiling included single PpDNMT3 and Ppdnmt3 stable lines, limiting our ability to fully distinguish reproducible PpDNMT3-associated methylation effects from line-specific changes, particularly for CG methylation.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Assaf Zemach (assafzemach@gmail.com).

Materials availability

Plasmids generated in this study are available from the lead contact upon request and can also be reconstructed using the DNA sequences provided in Figure S6 and the pCDNA3.1+N-eGFP backbone described in the STAR Methods.

Data and code availability

•
Raw and processed sequencing data generated in this study have been submitted to the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) under accession numbers GEO: GSE302209 (BS-seq data) and GEO: GSE302207 (RNA-seq data).
•
This study analyzed publicly available BS-seq data for neurons³⁴ and embryonic stem cells,³⁵ as well as HEK293 ATAC-seq⁴³ and histone marks ChIP-seq data, including H3K9me3 and H3K27me3,⁴⁴ H3K4me1, H3K4me2, H3K4me3, H3K27ac,⁴⁵ and H3K36me3.⁴⁶ Corresponding GEO accessions are detailed in Table S3.
•
This study did not generate original code.

Acknowledgments

This work was supported by the Israeli Centers for Research Excellence Program of the Planning and Budgeting Committee, Israel Science Foundation grant 757/12, Israel Science Foundation grant 1636/15, European Research Council (ERC, 679551) (to A.Z.), and Israel Science Foundation grant 767/09 (to N.O.).

Author contributions

A.Z., N.O., and E.B. directed and supervised the study. K.G.H., A.K., and R.Z. conducted the molecular and tissue culture procedures. R.D., R.Y., K.D.H., and A.Z. analyzed the genomic data. A.Z. wrote the manuscript. All authors discussed the results and commented on the manuscript.

Declaration of interests

The authors have filed a patent related to this work (US provisional patent application number 62/813,805).

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this manuscript, the authors used ChatGPT to assist with language editing and to improve the clarity and readability of the text. The authors reviewed and edited all AI-assisted content and take full responsibility for the final manuscript.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins

Agencourt Ampure XP beads	Beckman Coulter	Cat.#: A63880
DMEM	Thermo Fisher Scientific	Cat.#: 4195039
G418	Caisson LABS	Cat. #: G030
Heat-inactivated fetal calf serum	GIBCO	Cat. #: A5256801
Hoechst 33342 stain	Sigma-Aldrich	Cat.#: V875756-97-1
Nystatin	Sigma-Aldrich	Cat.#: N1638
Penicillin	Sigma-Aldrich	Cat.#: Po781
PolyJetTM In Vitro DNA Transfection Reagent	SignaGen	Cat.#: SL100688
PvuI restriction enzyme	Fermentas	Cat.#: ER0621

Critical commercial assays

cDNA kit]	Quanta-Bio	Cat. #: 95049-100
DNeasy® Blood and Tissue Kit	Qiagen	Cat.#: 69504
Q5 Site-Directed Mutagenesis kit	New England Biolabs	Cat.#: E0554
RNeasy Mini Kit	Qiagen	Cat.#: 74104
SureSelectXT Human Methyl-Seq target enrichment panel	Agilent	Cat.#: 5190-4661

Deposited data

Bulk RNA sequencing	This paper	GEO: GSE302207
Targeted bisulphite sequencing	This paper	GEO: GSE302209

Experimental models: Cell lines

Human embryonic kidney (HEK293T) cells	N/A	N/A
Human embryonic kidney (HEK293) cells	N/A	N/A

Oligonucleotides

PpDNMT3b_205EtoA, Forward: GTTTCTGGTGgccAATGTGGCCAG	This paper	N/A
PpDNMT3b_205EtoA, Reverse: CACTTAGGCTTGAACACC	This paper	N/A

Recombinant DNAS

pcDNA3.1+N-eGFP	Xiao et al.⁴⁷	Addgene Plasmid #129020; RRID:Addgene_129020
pcDNA3.1+N-PpDNMT3-eGFP	This paper	N/A
pcDNA3.1+N-PpDNMT3-eGFP	This paper	N/A

Software and algorithms

Cutadapt v1.15	Martin et al.⁴⁸	https://cutadapt.readthedocs.io/en/stable/
DESeq2 v1.46.0	Love et al.⁴⁹	https://bioconductor.org/packages/release/bioc/html/DESeq2.html
DMRcaller v1.38.0	Catoni et al.³⁸	https://www.bioconductor.org/packages/release/bioc/html/DMRcaller.html
FastQC v0.11.8	Babraham Bioinformatics	https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
HOME v1.0.0	Srivastava et al.³⁷	https://github.com/ListerLab/HOME
Methylpy v1.3.8	Schultz et al.³⁵	https://github.com/yupenghe/methylpy
Prinseq-lite v0.20.4	Shmieder et al.⁵⁰	https://prinseq.sourceforge.net/
STAR v2.7.0f	Dobin et al.⁵¹	https://github.com/alexdobin/STAR
TrimGalore v0.6.0	Babraham Bioinformatics	https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/

Other

Confocal microscope	Zeiss	LSM 780
HiSeq 2000 Sequencing System	Illumina	N/A
HiSeq 2500 Sequencing System	Illumina	N/A
Qubit fluorometer	Thermo Fisher Scientific	3.0
TapeStation	Agilent	N/A

Open in a new tab

Experimental model and study participant details

Human embryonic kidney 293T (HEK293T) cells were maintained in Dulbesco’s Modified Eagle Medium (DMEM, Thermo Fisher Scientific, cat. No. 4195039) supplemented with 10% heat-inactivated fetal calf serum, penicillin (20 U/ml), streptomycin (20 μg/ml), and nystatin (2.5 U/ml) at 37°C, 5% CO2 in a humidified incubator.

Method details

Plasmid construction

Codon-optimized PpDNMT3b gene (Figure S9) was synthesized and cloned into pCDNA3.1+N-eGFP⁴⁷ by GeneScript Biotech to be expressed in human cells under the CMV promoter. Mutant PpDNMT3b (Ppdnmt3) plasmid was generated using Q5 Site-Directed Mutagenesis kit (NEB #E0554) according to the manufacturer's protocol with the primers GTTTCTGGTGgccAATGTGGCCAG (Forward) and CACTTAGGCTTGAACACC (Reverse) introducing E to A point mutation in the catalytic ENV motif - amino acid 205 (Figure S9). Control cells were transfected with the pCDNA3.1+N-eGFP (empty) plasmid.

Transfection and cell culture

Plasmids were linearized with PvuI restriction enzyme prior to transfection. Hek293T cells and HEK293 cells were used for transient expression experiments and for espablishing stable cell lines, respectively. PolyJet In Vitro DNA Transfection Reagent (SignaGen, Cat. No. SL100688) was used to facilitate transfection according to manufacturer instructions. For transient expression experiments, cells were harvested seven days post-transfection. Stable transfectants were selected with G418 (500 μg/ml).

Confocal microscopy

Images were acquired using Zeiss LSM 780 confocal microscope (Zeiss, Oberkochen, Germany).

Genomic DNA extraction, methylation library preparation and sequencing

DNA methylation profiling was performed on single transient transgenic PpDNMT3 and control (GFP) lines, single stable PpDNMT3 and Ppdnmt3 lines, and two stable control lines.

Genomic DNA was extracted using DNeasy® Blood and Tissue Kit (Cat No. 69504, Qiagen) according to the manufacturer's protocol. Methylation sequencing libraries were produced using SureSelectXT Human Methyl-Seq target enrichment panel Library Preparation Kit (Agilent Part No. 5190-4661). Methyl-seq libraries were sequenced as a paired end on the Illumina HiSeq 2000 System in Beijing Genomics Institute (BGI), China. The resulting median sequencing depth of the targeted regions was above 22 for all samples.

Methylation data analysis

BS-seq reads quality was evaluated using FastQC v0.11.8 (Babraham Bioinformatics). Identical duplicate reads and low-quality reads were filtered out using Prinseq-lite v0-20-4 software.⁵² Downstream read alignment to the genome and methylation counting were performed using Methylpy v1.3.8 software³⁵ (Table S1). DMR calling was performed with HOME v1.0.0 software³⁷ using --home-pairwise function with default parameters and a minimum coverage of 5 reads and a minimum methylation difference of 5% and 25% for CHH and CG, respectively. DMR calling with DMRcaller v2.58.0 R package³⁸ was performed using computeDMRs function with the following parameters: method = “bins”, binSize = 50, test = “fisher”, pValueThreshold = 0.01, minCytosinesCount = 3, minGap = 150, minSize = 50, minReadsPerCytosine = 5, and minProportionDifference = 0.05 or minProportionDifference = 0.25 for CHH and CG, respectively [38].

For further analysis, methylation differences between samples were averaged across 50-bp genomic windows and correlated with publicly available chromatin-state mark peaks. Each 50-bp sliding window was annotated as either exhibiting or lacking the corresponding chromatin mark, as shown in Figures 2B and 2D. To test the association between methylation and transcriptional changes, we examined the presence of DMRs within 500 bp upstream of the transcription start site (TSS) or overlapping the gene body in upregulated and downregulated genes, as well as across all expressed genes (≥1 TPM).

RNA extraction, library preparation and sequencing

For each condition (Control, PpDNMT3, and Ppdnmt3), RNA was extracted to produce six replicates: two independent stably transformed lines per condition and three separate plates per line.

Total RNA was extracted from stably transformed cells using the RNeasy Mini Kit (Cat No. 74104, Qiagen) according to the manufacturer's protocol. The poly-A fraction (mRNA) was purified from 500 ng of total input RNA, followed by fragmentation and the generation of double-stranded cDNA. Then, Agencourt Ampure XP beads cleanup (Beckman Coulter), end repair, A base addition, adaptor ligation, and PCR amplification steps were performed. Libraries were quantified by Qubit (Thermo Fisher Scientific) and TapeStation (Agilent). The libraries were sequenced as single end reads on the Illumina HiSeq 2500 System (The Weizmann Institute of Science, Israel).

RNA-seq data analysis

RNA-Seq reads quality was evaluated using FastQC v0.11.8 (Babraham Bioinformatics). Adapter and quality trimming were performed with TrimGalore v0.6.0 (Babraham Bioinformatics) with Cutadapt v1.15.⁴⁸

The processed reads were aligned to the reference genome, and read counts per gene were obtained using STAR v2.7.0f46⁵¹ (Table S2). Differential expression analysis was performed with DESeq2 R package⁵¹ using the six RNA-seq datasets per sample. Genes with a raw read count below 10 were omitted from the analysis. Genes with a 2-fold change in expression level and false discovery rate (FDR) below 0.05 were considered differentially expressed.

Data visualization

Genome-wide and local methylation data was visualized using Python (Seaborn v0.13.2 library with Matplotlib v3.5.3). Transcriptomic data was visualized on Rstudio (2024.12.0+467) with ggplot2 v4.0.3 R package.

Venn diagrams were produced using BioVenn v1.1.3 Python package.

Figures were arranged in Adobe Illustrator.

Quantification and statistical analysis

Software used for NGS data processing and analysis, including versions, is listed in the key resources table. Key parameters used for the analysis are reported in the method details section of STAR Methods. Statistical tests and significance thresholds are indicated in figure legends.

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.isci.2026.117542.

Contributor Information

Nir Ohad, Email: niro@tauex.tau.ac.il.

Assaf Zemach, Email: assafzemach@gmail.com.

Supplemental information

Document S1. Figures S1–S9 and Tables S1–S4

mmc1.pdf^{(7MB, pdf)}

References

1.Zeng Y., Chen T. DNA Methylation Reprogramming during Mammalian Development. Genes (Basel) 2019;10:257. doi: 10.3390/genes10040257. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Cedar H., Stein R., Gruenbaum Y., Naveh-Many T., Sciaky-Gallili N., Razin A. Effect of DNA Methylation on Gene Expression. Cold Spring Harb. Symp. Quant. Biol. 1983;47 Pt 2:605–609. doi: 10.1101/SQB.1983.047.01.071. [DOI] [PubMed] [Google Scholar]
3.Lister R., Pelizzola M., Dowen R.H., Hawkins R.D., Hon G., Tonti-filippini J., Nery J.R., Lee L., Ye Z., Ngo Q.M., et al. Human DNA methylomes at base resolution show widespread epigenomic differences. Nature. 2009;462:315–322. doi: 10.1038/nature08514. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Bird A.P. DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Res. 1980;8:1499–1504. doi: 10.1093/nar/8.7.1499. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Feng S., Cokus S.J., Zhang X., Chen P.Y., Bostick M., Goll M.G., Hetzel J., Jain J., Strauss S.H., Halpern M.E., et al. Conservation and divergence of methylation patterning in plants and animals. Proc. Natl. Acad. Sci. USA. 2010;107:8689–8694. doi: 10.1073/pnas.1002720107. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Brethouwer T., de Mendoza A., Bogdanovic O. Non-CG DNA methylation in animal genomes. Nat. Genet. 2025;57:2395–2407. doi: 10.1038/s41588-025-02303-1. [DOI] [PubMed] [Google Scholar]
7.He X.-J., Chen T., Zhu J.-K. Regulation and function of DNA methylation in plants and animals. Cell Res. 2011;21:442–465. doi: 10.1038/cr.2011.23. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.He Y., Ecker J.R. Non-CG Methylation in the Human Genome. Annu. Rev. Genomics Hum. Genet. 2015;16:55–77. doi: 10.1146/annurev-genom-090413-025437. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Butcher L.M., Ito M., Brimpari M., Morris T.J., Soares F.A.C., Ährlund-Richter L., Carey N., Vallier L., Ferguson-Smith A.C., Beck S. Non-CG DNA methylation is a biomarker for assessing endodermal differentiation capacity in pluripotent stem cells. Nat. Commun. 2016;7 doi: 10.1038/ncomms10458. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Guo W., Zhang M.Q., Wu H. Mammalian non-CG methylations are conserved and cell-type specific and may have been involved in the evolution of transposon elements. Sci. Rep. 2016;6 doi: 10.1038/srep32207. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Arand J., Spieler D., Karius T., Branco M.R., Meilinger D., Meissner A., Jenuwein T., Xu G., Leonhardt H., Wolf V., et al. In Vivo Control of CpG and Non-CpG DNA Methylation by DNA Methyltransferases. PLoS Genet. 2012;8 doi: 10.1371/journal.pgen.1002750. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Guo J.U., Su Y., Shin J.H., Shin J., Li H., Xie B., Zhong C., Hu S., Le T., Fan G., et al. Distribution, recognition and regulation of non-CpG methylation in the adult mammalian brain. Nat. Neurosci. 2014;17:215–222. doi: 10.1038/nn.3607. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Lee J.H., Park S.J., Nakai K. Differential landscape of non-CpG methylation in embryonic stem cells and neurons caused by DNMT3s. Sci. Rep. 2017;7 doi: 10.1038/s41598-017-11800-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Clemens A.W., Wu D.Y., Moore J.R., Christian D.L., Zhao G., Gabel H.W. MeCP2 Represses Enhancers through Chromosome Topology-Associated DNA Methylation. Mol. Cell. 2020;77:279–293.e8. doi: 10.1016/j.molcel.2019.10.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Hamagami N., Wu D.Y., Clemens A.W., Nettles S.A., Li A., Gabel H.W. NSD1 deposits histone H3 lysine 36 dimethylation to pattern non-CG DNA methylation in neurons. Mol. Cell. 2023;83:1412–1428.e7. doi: 10.1016/j.molcel.2023.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Cheng C., Tarutani Y., Miyao A., Ito T., Yamazaki M., Sakai H., Fukai E., Hirochika H. Loss of function mutations in the rice chromomethylase OsCMT3a cause a burst of transposition. Plant J. 2015;83:1069–1081. doi: 10.1111/tpj.12952. [DOI] [PubMed] [Google Scholar]
17.Lagger S., Connelly J.C., Schweikert G., Webb S., Selfridge J., Ramsahoye B.H., Yu M., He C., Sanguinetti G., Sowers L.C., et al. MeCP2 recognizes cytosine methylated tri-nucleotide and di-nucleotide sequences to tune transcription in the mammalian brain. PLoS Genet. 2017;13 doi: 10.1371/journal.pgen.1006793. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Christian D.L., Wu D.Y., Martin J.R., Moore J.R., Liu Y.R., Clemens A.W., Nettles S.A., Kirkland N.M., Papouin T., Hill C.A., et al. DNMT3A Haploinsufficiency Results in Behavioral Deficits and Global Epigenomic Dysregulation Shared across Neurodevelopmental Disorders. Cell Rep. 2020;33 doi: 10.1016/j.celrep.2020.108416. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Tillotson R., Cholewa-Waclaw J., Chhatbar K., Connelly J.C., Kirschner S.A., Webb S., Koerner M.V., Selfridge J., Kelly D.A., De Sousa D., et al. Neuronal non-CG methylation is an essential target for MeCP2 function. Mol. Cell. 2021;81:1260–1275.e12. doi: 10.1016/j.molcel.2021.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Moore J.R., Nemera M.T., D’Souza R.D., Hamagami N., Clemens A.W., Beard D.C., Urman A., Razia Y., Rodriguez Mendoza V., Law T.E., et al. MeCP2 and non-CG DNA methylation stabilize the expression of long genes that distinguish closely related neuron types. Nat. Neurosci. 2025;28:1185–1198. doi: 10.1038/s41593-025-01947-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Beard D.C., Zhang X., Wu D.Y., Martin J.R., Erickson A., Boua J.V., Hamagami N., Swift R.G., McCullough K.B., Ge X., et al. Distinct disease mutations in DNMT3A result in a spectrum of behavioral, epigenetic, and transcriptional deficits. Cell Rep. 2023;42 doi: 10.1016/j.celrep.2023.113411. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Amir R.E., Van den Veyver I.B., Wan M., Tran C.Q., Francke U., Zoghbi H.Y. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat. Genet. 1999;23:185–188. doi: 10.1038/13810. [DOI] [PubMed] [Google Scholar]
23.Ramasamy D., Deva Magendhra Rao A.K., Rajkumar T., Mani S. Non-CpG methylation—a key epigenetic modification in cancer. Brief. Funct. Genomics. 2021;20:304–311. doi: 10.1093/bfgp/elab035. [DOI] [PubMed] [Google Scholar]
24.Inoue S., Oishi M. Effects of methylation of non-CpG sequence in the promoter region on the expression of human synaptotagmin XI (syt11) Gene. 2005;348:123–134. doi: 10.1016/j.gene.2004.12.044. [DOI] [PubMed] [Google Scholar]
25.Yang A., Yan S., Yin Y., Chen C., Tang X., Ran M., Chen B. FZD7, Regulated by Non-CpG Methylation, Plays an Important Role in Immature Porcine Sertoli Cell Proliferation. Int. J. Mol. Sci. 2023;24:6179. doi: 10.3390/ijms24076179. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Malone C.S., Miner M.D., Doerr J.R., Jackson J.P., Jacobsen S.E., Wall R., Teitell M. C m C(A/T)GG DNA methylation in mature B cell lymphoma gene silencing. Proc. Natl. Acad. Sci. USA. 2001;98:10404–10409. doi: 10.1073/pnas.181206898. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Zemach A., McDaniel I.E., Silva P., Zilberman D. Genome-Wide Evolutionary Analysis of Eukaryotic DNA Methylation. Science. 2010;328:916–919. doi: 10.1126/science.1186366. [DOI] [PubMed] [Google Scholar]
28.Zhang H., Lang Z., Zhu J.-K. Dynamics and function of DNA methylation in plants. Nat. Rev. Mol. Cell Biol. 2018;19:489–506. doi: 10.1038/s41580-018-0016-z. [DOI] [PubMed] [Google Scholar]
29.Domb K., Katz A., Harris K.D., Yaari R., Kaisler E., Nguyen V.H., Hong U.V.T., Griess O., Heskiau K.G., Ohad N., et al. DNA methylation mutants in Physcomitrella patens elucidate individual roles of CG and non-CG methylation in genome regulation. Proc. Natl. Acad. Sci. USA. 2020;117:33700–33710. doi: 10.1073/pnas.2011361117. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Cao X., Springer N.M., Muszynski M.G., Phillips R.L., Kaeppler S., Jacobsen S.E. Conserved plant genes with similarity to mammalian de novo DNA methyltransferases. Proc. Natl. Acad. Sci. USA. 2000;97:4979–4984. doi: 10.1073/pnas.97.9.4979. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Cao X., Aufsatz W., Zilberman D., Mette M.F., Huang M.S., Matzke M., Jacobsen S.E. Role of the DRM and CMT3 Methyltransferases in RNA-Directed DNA Methylation. Curr. Biol. 2003;13:2212–2217. doi: 10.1016/j.cub.2003.11.052. [DOI] [PubMed] [Google Scholar]
32.Yaari R., Katz A., Domb K., Harris K.D., Zemach A., Ohad N. RdDM-independent de novo and heterochromatin DNA methylation by plant CMT and DNMT3 orthologs. Nat. Commun. 2019;10:1613. doi: 10.1038/s41467-019-09496-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Malik G., Dangwal M., Kapoor S., Kapoor M. Role of DNA methylation in growth and differentiation in Physcomitrella patens and characterization of cytosine DNA methyltransferases. FEBS J. 2012;279:4081–4094. doi: 10.1111/febs.12002. [DOI] [PubMed] [Google Scholar]
34.Heyn H., Vidal E., Ferreira H.J., Vizoso M., Sayols S., Gomez A., Moran S., Boque-Sastre R., Guil S., Martinez-Cardus A., et al. Epigenomic analysis detects aberrant super-enhancer DNA methylation in human cancer. Genome Biol. 2016;17:11. doi: 10.1186/s13059-016-0879-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Schultz M.D., He Y., Whitaker J.W., Hariharan M., Mukamel E.A., Leung D., Rajagopal N., Nery J.R., Urich M.A., Chen H., et al. Human body epigenome maps reveal noncanonical DNA methylation variation. Nature. 2015;523:212–216. doi: 10.1038/nature14465. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Reither S., Li F., Gowher H., Jeltsch A. Catalytic Mechanism of DNA-(cytosine-C5)-methyltransferases Revisited: Covalent Intermediate Formation is not Essential for Methyl Group Transfer by the Murine Dnmt3a Enzyme. J. Mol. Biol. 2003;329:675–684. doi: 10.1016/S0022-2836(03)00509-6. [DOI] [PubMed] [Google Scholar]
37.Srivastava A., Karpievitch Y.V., Eichten S.R., Borevitz J.O., Lister R. HOME: A histogram based machine learning approach for effective identification of differentially methylated regions. BMC Bioinf. 2019;20 doi: 10.1186/s12859-019-2845-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Catoni M., Tsang J.M., Greco A.P., Zabet N.R. DMRcaller: A versatile R/Bioconductor package for detection and visualization of differentially methylated regions in CpG and non-CpG contexts. Nucleic Acids Res. 2018;46:e114. doi: 10.1093/nar/gky602. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Chen L., Chen K., Lavery L.A., Baker S.A., Shaw C.A., Li W., Zoghbi H.Y. MeCP2 binds to non-CG methylated DNA as neurons mature, influencing transcription and the timing of onset for Rett syndrome. Proc. Natl. Acad. Sci. USA. 2015;112:5509–5514. doi: 10.1073/pnas.1505909112. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Jin Y., Su K., Kong H.E., Ma W., Wang Z., Li Y., Li R., Allen E.G., Wu H., Jin P. Cell type-specific DNA methylome signatures reveal epigenetic mechanisms for neuronal diversity and neurodevelopmental disorder. Hum. Mol. Genet. 2023;32:218–230. doi: 10.1093/hmg/ddac189. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Uhlén M., Fagerberg L., Hallström B.M., Lindskog C., Oksvold P., Mardinoglu A., Sivertsson Å., Kampf C., Sjöstedt E., Asplund A., et al. Tissue-based map of the human proteome. Science. 2015;347 doi: 10.1126/science.1260419. [DOI] [PubMed] [Google Scholar]
42.Miodownik I., Solozabal R., O’Hagan M.P., Albeck S., Peleg Y., Takac M., Afek A. DNA methylation shapes transcription factor binding beyond canonical CpG contexts. Proc. Natl. Acad. Sci. USA. 2025;122 doi: 10.1073/pnas.2520814122. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Karabacak Calviello A., Hirsekorn A., Wurmus R., Yusuf D., Ohler U. Reproducible inference of transcription factor footprints in ATAC-seq and DNase-seq datasets using protocol-specific bias modeling. Genome Biol. 2019;20 doi: 10.1186/s13059-019-1654-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Patel D., Patel M., Datta S., Singh U. CGGBP1 regulates CTCF occupancy at repeats. Epigenetics Chromatin. 2019;12 doi: 10.1186/s13072-019-0305-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Morgan M.A.J., Rickels R.A., Collings C.K., He X., Cao K., Herz H.M., Cozzolino K.A., Abshiru N.A., Marshall S.A., Rendleman E.J., et al. A cryptic tudor domain links BRWD2/PHIP to COMPASS-mediated histone H3K4 methylation. Genes Dev. 2017;31:2003–2014. doi: 10.1101/gad.305201.117. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Zhu K., Lei P.J., Ju L.G., Wang X., Huang K., Yang B., Shao C., Zhu Y., Wei G., Fu X.D., et al. SPOP-containing complex regulates SETD2 stability and H3K36me3-coupled alternative splicing. Nucleic Acids Res. 2017;45:92–105. doi: 10.1093/nar/gkw814. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Xiao M.S., Wilusz J.E. An improved method for circular RNA purification using RNase R that efficiently removes linear RNAs containing G-quadruplexes or structured 3’ ends. Nucleic Acids Res. 2019;47:8755–8769. doi: 10.1093/NAR/GKZ576. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 2011;17:10. doi: 10.14806/ej.17.1.200. [DOI] [Google Scholar]
49.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15 doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Schmieder R., Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27:863–864. doi: 10.1093/bioinformatics/btr026. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Dobin A., Davis C.A., Schlesinger F., Drenkow J., Zaleski C., Jha S., Batut P., Chaisson M., Gingeras T.R. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21. doi: 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Schmieder R., Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27:863–864. doi: 10.1093/bioinformatics/btr026. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S9 and Tables S1–S4

mmc1.pdf^{(7MB, pdf)}

Data Availability Statement

•
Raw and processed sequencing data generated in this study have been submitted to the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) under accession numbers GEO: GSE302209 (BS-seq data) and GEO: GSE302207 (RNA-seq data).
•
This study analyzed publicly available BS-seq data for neurons³⁴ and embryonic stem cells,³⁵ as well as HEK293 ATAC-seq⁴³ and histone marks ChIP-seq data, including H3K9me3 and H3K27me3,⁴⁴ H3K4me1, H3K4me2, H3K4me3, H3K27ac,⁴⁵ and H3K36me3.⁴⁶ Corresponding GEO accessions are detailed in Table S3.
•
This study did not generate original code.

[bib1] 1.Zeng Y., Chen T. DNA Methylation Reprogramming during Mammalian Development. Genes (Basel) 2019;10:257. doi: 10.3390/genes10040257. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] 2.Cedar H., Stein R., Gruenbaum Y., Naveh-Many T., Sciaky-Gallili N., Razin A. Effect of DNA Methylation on Gene Expression. Cold Spring Harb. Symp. Quant. Biol. 1983;47 Pt 2:605–609. doi: 10.1101/SQB.1983.047.01.071. [DOI] [PubMed] [Google Scholar]

[bib3] 3.Lister R., Pelizzola M., Dowen R.H., Hawkins R.D., Hon G., Tonti-filippini J., Nery J.R., Lee L., Ye Z., Ngo Q.M., et al. Human DNA methylomes at base resolution show widespread epigenomic differences. Nature. 2009;462:315–322. doi: 10.1038/nature08514. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Bird A.P. DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Res. 1980;8:1499–1504. doi: 10.1093/nar/8.7.1499. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Feng S., Cokus S.J., Zhang X., Chen P.Y., Bostick M., Goll M.G., Hetzel J., Jain J., Strauss S.H., Halpern M.E., et al. Conservation and divergence of methylation patterning in plants and animals. Proc. Natl. Acad. Sci. USA. 2010;107:8689–8694. doi: 10.1073/pnas.1002720107. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Brethouwer T., de Mendoza A., Bogdanovic O. Non-CG DNA methylation in animal genomes. Nat. Genet. 2025;57:2395–2407. doi: 10.1038/s41588-025-02303-1. [DOI] [PubMed] [Google Scholar]

[bib7] 7.He X.-J., Chen T., Zhu J.-K. Regulation and function of DNA methylation in plants and animals. Cell Res. 2011;21:442–465. doi: 10.1038/cr.2011.23. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8.He Y., Ecker J.R. Non-CG Methylation in the Human Genome. Annu. Rev. Genomics Hum. Genet. 2015;16:55–77. doi: 10.1146/annurev-genom-090413-025437. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] 9.Butcher L.M., Ito M., Brimpari M., Morris T.J., Soares F.A.C., Ährlund-Richter L., Carey N., Vallier L., Ferguson-Smith A.C., Beck S. Non-CG DNA methylation is a biomarker for assessing endodermal differentiation capacity in pluripotent stem cells. Nat. Commun. 2016;7 doi: 10.1038/ncomms10458. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Guo W., Zhang M.Q., Wu H. Mammalian non-CG methylations are conserved and cell-type specific and may have been involved in the evolution of transposon elements. Sci. Rep. 2016;6 doi: 10.1038/srep32207. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Arand J., Spieler D., Karius T., Branco M.R., Meilinger D., Meissner A., Jenuwein T., Xu G., Leonhardt H., Wolf V., et al. In Vivo Control of CpG and Non-CpG DNA Methylation by DNA Methyltransferases. PLoS Genet. 2012;8 doi: 10.1371/journal.pgen.1002750. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Guo J.U., Su Y., Shin J.H., Shin J., Li H., Xie B., Zhong C., Hu S., Le T., Fan G., et al. Distribution, recognition and regulation of non-CpG methylation in the adult mammalian brain. Nat. Neurosci. 2014;17:215–222. doi: 10.1038/nn.3607. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] 13.Lee J.H., Park S.J., Nakai K. Differential landscape of non-CpG methylation in embryonic stem cells and neurons caused by DNMT3s. Sci. Rep. 2017;7 doi: 10.1038/s41598-017-11800-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Clemens A.W., Wu D.Y., Moore J.R., Christian D.L., Zhao G., Gabel H.W. MeCP2 Represses Enhancers through Chromosome Topology-Associated DNA Methylation. Mol. Cell. 2020;77:279–293.e8. doi: 10.1016/j.molcel.2019.10.033. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Hamagami N., Wu D.Y., Clemens A.W., Nettles S.A., Li A., Gabel H.W. NSD1 deposits histone H3 lysine 36 dimethylation to pattern non-CG DNA methylation in neurons. Mol. Cell. 2023;83:1412–1428.e7. doi: 10.1016/j.molcel.2023.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib16] 16.Cheng C., Tarutani Y., Miyao A., Ito T., Yamazaki M., Sakai H., Fukai E., Hirochika H. Loss of function mutations in the rice chromomethylase OsCMT3a cause a burst of transposition. Plant J. 2015;83:1069–1081. doi: 10.1111/tpj.12952. [DOI] [PubMed] [Google Scholar]

[bib17] 17.Lagger S., Connelly J.C., Schweikert G., Webb S., Selfridge J., Ramsahoye B.H., Yu M., He C., Sanguinetti G., Sowers L.C., et al. MeCP2 recognizes cytosine methylated tri-nucleotide and di-nucleotide sequences to tune transcription in the mammalian brain. PLoS Genet. 2017;13 doi: 10.1371/journal.pgen.1006793. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib18] 18.Christian D.L., Wu D.Y., Martin J.R., Moore J.R., Liu Y.R., Clemens A.W., Nettles S.A., Kirkland N.M., Papouin T., Hill C.A., et al. DNMT3A Haploinsufficiency Results in Behavioral Deficits and Global Epigenomic Dysregulation Shared across Neurodevelopmental Disorders. Cell Rep. 2020;33 doi: 10.1016/j.celrep.2020.108416. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19.Tillotson R., Cholewa-Waclaw J., Chhatbar K., Connelly J.C., Kirschner S.A., Webb S., Koerner M.V., Selfridge J., Kelly D.A., De Sousa D., et al. Neuronal non-CG methylation is an essential target for MeCP2 function. Mol. Cell. 2021;81:1260–1275.e12. doi: 10.1016/j.molcel.2021.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Moore J.R., Nemera M.T., D’Souza R.D., Hamagami N., Clemens A.W., Beard D.C., Urman A., Razia Y., Rodriguez Mendoza V., Law T.E., et al. MeCP2 and non-CG DNA methylation stabilize the expression of long genes that distinguish closely related neuron types. Nat. Neurosci. 2025;28:1185–1198. doi: 10.1038/s41593-025-01947-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] 21.Beard D.C., Zhang X., Wu D.Y., Martin J.R., Erickson A., Boua J.V., Hamagami N., Swift R.G., McCullough K.B., Ge X., et al. Distinct disease mutations in DNMT3A result in a spectrum of behavioral, epigenetic, and transcriptional deficits. Cell Rep. 2023;42 doi: 10.1016/j.celrep.2023.113411. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Amir R.E., Van den Veyver I.B., Wan M., Tran C.Q., Francke U., Zoghbi H.Y. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat. Genet. 1999;23:185–188. doi: 10.1038/13810. [DOI] [PubMed] [Google Scholar]

[bib23] 23.Ramasamy D., Deva Magendhra Rao A.K., Rajkumar T., Mani S. Non-CpG methylation—a key epigenetic modification in cancer. Brief. Funct. Genomics. 2021;20:304–311. doi: 10.1093/bfgp/elab035. [DOI] [PubMed] [Google Scholar]

[bib24] 24.Inoue S., Oishi M. Effects of methylation of non-CpG sequence in the promoter region on the expression of human synaptotagmin XI (syt11) Gene. 2005;348:123–134. doi: 10.1016/j.gene.2004.12.044. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Yang A., Yan S., Yin Y., Chen C., Tang X., Ran M., Chen B. FZD7, Regulated by Non-CpG Methylation, Plays an Important Role in Immature Porcine Sertoli Cell Proliferation. Int. J. Mol. Sci. 2023;24:6179. doi: 10.3390/ijms24076179. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] 26.Malone C.S., Miner M.D., Doerr J.R., Jackson J.P., Jacobsen S.E., Wall R., Teitell M. C m C(A/T)GG DNA methylation in mature B cell lymphoma gene silencing. Proc. Natl. Acad. Sci. USA. 2001;98:10404–10409. doi: 10.1073/pnas.181206898. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Zemach A., McDaniel I.E., Silva P., Zilberman D. Genome-Wide Evolutionary Analysis of Eukaryotic DNA Methylation. Science. 2010;328:916–919. doi: 10.1126/science.1186366. [DOI] [PubMed] [Google Scholar]

[bib28] 28.Zhang H., Lang Z., Zhu J.-K. Dynamics and function of DNA methylation in plants. Nat. Rev. Mol. Cell Biol. 2018;19:489–506. doi: 10.1038/s41580-018-0016-z. [DOI] [PubMed] [Google Scholar]

[bib29] 29.Domb K., Katz A., Harris K.D., Yaari R., Kaisler E., Nguyen V.H., Hong U.V.T., Griess O., Heskiau K.G., Ohad N., et al. DNA methylation mutants in Physcomitrella patens elucidate individual roles of CG and non-CG methylation in genome regulation. Proc. Natl. Acad. Sci. USA. 2020;117:33700–33710. doi: 10.1073/pnas.2011361117. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30.Cao X., Springer N.M., Muszynski M.G., Phillips R.L., Kaeppler S., Jacobsen S.E. Conserved plant genes with similarity to mammalian de novo DNA methyltransferases. Proc. Natl. Acad. Sci. USA. 2000;97:4979–4984. doi: 10.1073/pnas.97.9.4979. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.Cao X., Aufsatz W., Zilberman D., Mette M.F., Huang M.S., Matzke M., Jacobsen S.E. Role of the DRM and CMT3 Methyltransferases in RNA-Directed DNA Methylation. Curr. Biol. 2003;13:2212–2217. doi: 10.1016/j.cub.2003.11.052. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Yaari R., Katz A., Domb K., Harris K.D., Zemach A., Ohad N. RdDM-independent de novo and heterochromatin DNA methylation by plant CMT and DNMT3 orthologs. Nat. Commun. 2019;10:1613. doi: 10.1038/s41467-019-09496-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Malik G., Dangwal M., Kapoor S., Kapoor M. Role of DNA methylation in growth and differentiation in Physcomitrella patens and characterization of cytosine DNA methyltransferases. FEBS J. 2012;279:4081–4094. doi: 10.1111/febs.12002. [DOI] [PubMed] [Google Scholar]

[bib34] 34.Heyn H., Vidal E., Ferreira H.J., Vizoso M., Sayols S., Gomez A., Moran S., Boque-Sastre R., Guil S., Martinez-Cardus A., et al. Epigenomic analysis detects aberrant super-enhancer DNA methylation in human cancer. Genome Biol. 2016;17:11. doi: 10.1186/s13059-016-0879-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.Schultz M.D., He Y., Whitaker J.W., Hariharan M., Mukamel E.A., Leung D., Rajagopal N., Nery J.R., Urich M.A., Chen H., et al. Human body epigenome maps reveal noncanonical DNA methylation variation. Nature. 2015;523:212–216. doi: 10.1038/nature14465. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Reither S., Li F., Gowher H., Jeltsch A. Catalytic Mechanism of DNA-(cytosine-C5)-methyltransferases Revisited: Covalent Intermediate Formation is not Essential for Methyl Group Transfer by the Murine Dnmt3a Enzyme. J. Mol. Biol. 2003;329:675–684. doi: 10.1016/S0022-2836(03)00509-6. [DOI] [PubMed] [Google Scholar]

[bib37] 37.Srivastava A., Karpievitch Y.V., Eichten S.R., Borevitz J.O., Lister R. HOME: A histogram based machine learning approach for effective identification of differentially methylated regions. BMC Bioinf. 2019;20 doi: 10.1186/s12859-019-2845-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] 38.Catoni M., Tsang J.M., Greco A.P., Zabet N.R. DMRcaller: A versatile R/Bioconductor package for detection and visualization of differentially methylated regions in CpG and non-CpG contexts. Nucleic Acids Res. 2018;46:e114. doi: 10.1093/nar/gky602. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39.Chen L., Chen K., Lavery L.A., Baker S.A., Shaw C.A., Li W., Zoghbi H.Y. MeCP2 binds to non-CG methylated DNA as neurons mature, influencing transcription and the timing of onset for Rett syndrome. Proc. Natl. Acad. Sci. USA. 2015;112:5509–5514. doi: 10.1073/pnas.1505909112. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Jin Y., Su K., Kong H.E., Ma W., Wang Z., Li Y., Li R., Allen E.G., Wu H., Jin P. Cell type-specific DNA methylome signatures reveal epigenetic mechanisms for neuronal diversity and neurodevelopmental disorder. Hum. Mol. Genet. 2023;32:218–230. doi: 10.1093/hmg/ddac189. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Uhlén M., Fagerberg L., Hallström B.M., Lindskog C., Oksvold P., Mardinoglu A., Sivertsson Å., Kampf C., Sjöstedt E., Asplund A., et al. Tissue-based map of the human proteome. Science. 2015;347 doi: 10.1126/science.1260419. [DOI] [PubMed] [Google Scholar]

[bib42] 42.Miodownik I., Solozabal R., O’Hagan M.P., Albeck S., Peleg Y., Takac M., Afek A. DNA methylation shapes transcription factor binding beyond canonical CpG contexts. Proc. Natl. Acad. Sci. USA. 2025;122 doi: 10.1073/pnas.2520814122. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43.Karabacak Calviello A., Hirsekorn A., Wurmus R., Yusuf D., Ohler U. Reproducible inference of transcription factor footprints in ATAC-seq and DNase-seq datasets using protocol-specific bias modeling. Genome Biol. 2019;20 doi: 10.1186/s13059-019-1654-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44.Patel D., Patel M., Datta S., Singh U. CGGBP1 regulates CTCF occupancy at repeats. Epigenetics Chromatin. 2019;12 doi: 10.1186/s13072-019-0305-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] 45.Morgan M.A.J., Rickels R.A., Collings C.K., He X., Cao K., Herz H.M., Cozzolino K.A., Abshiru N.A., Marshall S.A., Rendleman E.J., et al. A cryptic tudor domain links BRWD2/PHIP to COMPASS-mediated histone H3K4 methylation. Genes Dev. 2017;31:2003–2014. doi: 10.1101/gad.305201.117. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 46.Zhu K., Lei P.J., Ju L.G., Wang X., Huang K., Yang B., Shao C., Zhu Y., Wei G., Fu X.D., et al. SPOP-containing complex regulates SETD2 stability and H3K36me3-coupled alternative splicing. Nucleic Acids Res. 2017;45:92–105. doi: 10.1093/nar/gkw814. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] 47.Xiao M.S., Wilusz J.E. An improved method for circular RNA purification using RNase R that efficiently removes linear RNAs containing G-quadruplexes or structured 3’ ends. Nucleic Acids Res. 2019;47:8755–8769. doi: 10.1093/NAR/GKZ576. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] 48.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 2011;17:10. doi: 10.14806/ej.17.1.200. [DOI] [Google Scholar]

[bib49] 49.Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15 doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] 50.Schmieder R., Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27:863–864. doi: 10.1093/bioinformatics/btr026. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 51.Dobin A., Davis C.A., Schlesinger F., Drenkow J., Zaleski C., Jha S., Batut P., Chaisson M., Gingeras T.R. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21. doi: 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 52.Schmieder R., Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27:863–864. doi: 10.1093/bioinformatics/btr026. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

A plant DNMT3 induces non-CG methylation and affects gene expression in human cells

Rotem Domb

Karina G Heskiau

Aviva Katz

Rafael Yaari

Rachel Zamostiano

Keith D Harris

Eran Bacharach

Nir Ohad

Assaf Zemach

Summary

Graphical abstract

Highlights

Introduction

Results

PpDNMT3 induces non-CG methylation in human cells

Figure 1.

PpDNMT3 affects CG methylation in human cells

PpDNMT3 induces non-CG methylation at heterochromatic and genic regions

Figure 2.

PpDNMT3 alters gene expression

Figure 3.

Combinatorial effect of PpDNMT3’s CG and non-CG methylation on gene expression

Figure 4.

Discussion

Limitations of the study

Resource availability

Lead contact

Materials availability

Data and code availability

Acknowledgments

Author contributions

Declaration of interests

Declaration of generative AI and AI-assisted technologies in the writing process

STAR★Methods

Key resources table

Experimental model and study participant details

Method details

Plasmid construction

Transfection and cell culture

Confocal microscopy

Genomic DNA extraction, methylation library preparation and sequencing

Methylation data analysis

RNA extraction, library preparation and sequencing

RNA-seq data analysis

Data visualization

Quantification and statistical analysis

Footnotes

Contributor Information

Supplemental information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases