Skip to main content
NCBI home page
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2012 Jun 21;40(17):8550–8557. doi: 10.1093/nar/gks537

RNA modulation of the human DNA methyltransferase 3A

Celeste Holz-Schietinger 1, Norbert O Reich 1,2,*
PMCID: PMC3458530  PMID: 22730298

Abstract

DNA methyltransferase 3A (DNMT3A) is one of two human de novo DNA methyltransferases essential for transcription regulation during cellular development and differentiation. There is increasing evidence that RNA plays a role in directing DNA methylation to specific genomic locations within mammalian cells. Here, we describe two modes of RNA regulation of DNMT3A in vitro. We show a single-stranded RNA molecule that is antisense to the E-cadherin promoter binds tightly to the catalytic domain in a structurally dependent fashion causing potent inhibition of DNMT3A activity. Two other RNA molecules bind DNMT3A at an allosteric site outside the catalytic domain, causing no change in catalysis. Our observation of the potent and specific in vitro modulation of DNMT3A activity by RNA supports in vivo data that RNA interacts with DNMT3A to regulate transcription.

INTRODUCTION

DNA methylation, in conjunction with other epigenetic controllers including histone modifications and non-coding-RNA (ncRNA), is critical to the hierarchical regulation of transcription (1,2). In mammals, de novo 5-methylcytosine (5mC) patterning is established by DNA methyltransferase 3A (DNMT3A) and the closely related DNMT3B (1,3,4) and the patterning is maintained by DNMT1 (5). The mechanisms that control genomic site selection by the de novo methyltransferases are not understood, but there is increasing evidence that binding partners, both RNA and protein, play a major role in controlling DNA methylation (6–9).

Non-coding RNAs play an essential role in transcriptional regulation, including the formation of ternary complexes (protein/RNA/DNA) that modulate genomic targeting by directing hybridization and providing scaffolding for histone modifying enzymes and transcription factors (10–13). In plants, fission yeast and flies, interaction with ncRNA directs the DNMTs to specific targets (14). Similar mechanisms have been proposed for mammalian de novo DNMTs, which are supported by recent data showing piwi RNA-directed DNA methylation in germ cells (15) and a short non-coding RNA is necessary for some DNMT3b-mediated methylation through the formation of a sequence-specific triplex structure (16). Antisense RNA to the EF1a promoter causes transcriptional silencing in cells by inducing DNA methylation (17). Additionally, RNA transcripts are essential for genomic imprinting and X-chromosome inactivation, both of which occur via DNA methylation (18). This includes Tsix RNA, which regulates X-inactivation and binds to DNMT3A in vivo but neither DNMT3B nor DNMT1 (19). Further evidence is shown by DNMTs binding various single-stranded (ss) forms of RNA and DNA, in some cases at an allosteric site (20–25). However, how RNA modulates the catalytic properties of the de novo DNMTs is not known.

We probed the interactions between the human de novo DNMT3A and three RNA transcripts previously shown to either associate with DNMT3A or change promoter-specific methylation. These include a siRNA duplex to the NUP153 mRNA (20), a ssRNA antisense to the EF1a promoter (EFS1) (26), and a single-stranded antisense RNA to the E-cadherin (CHD) promoter (27). We find these three RNA molecules bind DNMT3A in two different fashions. NUP153 duplex RNA and ssEF1a RNA bind allosterically outside the catalytic domain, causing no measurable change in DNMT3A activity. The CHD RNA can bind to the catalytic domain and is a potent inhibitor of DNMT3A, and its modulation of DNMT3A activity is dependent on the RNA structure.

MATERIALS AND METHODS

Materials

The DNA used as substrates, poly-dIdC (∼1900 bp; Sigma-Aldrich) and GCbox2 (GGGAATTCAAGGGGCGGGGCAATGTTAGGG) duplex were purchased from Midland Certified Reagent Company. Annealing was confirmed by 12% polyacrylamide gels. The recognition site for DNMT3A is underlined. The human promoter substrates were described previously (28). RNA oligonucleotides are listed in Table 1. CHD RNA (antisense E-cadherin promoter) was purchased from IDT and HPLC purified. EFS1 (antisense EF1a promoter) and NUP153 antisense and sense were purchased from Dharmacon, desalted and ethanol precipitated. Concentrations are given in number of molecules. RNA purity was visualized by 12% denaturing (12 M Urea) PAGE gel and duplex formation was confirmed by 12% native PAGE.

Table 1.

RNA sequences used in this study

RNA name Sequence Description
CHD 5′ GGGGUGACGGGGACAGGCGGGGCUG 3′ Antisense E-cadherin promoter
EFS1 5′ GAGUUUACCCCGCGCCACCUU 3′ Antisense EF1a promoter
Nup153 duplex 5′ ACCAAAUAAAACUGGCAAAdTdT 3′
siRNA 3′ dTdTUGGUUUAUUUUGACCGUUU 5′ Nuclear pore complex mRNA
Nup153-as 5′ ACCAAAUAAAACUGGCAAAdTdT 3′
Nup153-s 5′ UUUGCCAGUUUUAUUUGGUTdTd 3′
Open in a new tab

RNA preparation

RNA was prepared by adding 1 mM MgCl2, then heating to 85°C for 5 min and slow cooling for 5 h unless otherwise noted (heating slow cooled). Otherwise the RNA was fast cooled by placed directly on ice after the heating step or unheated RNA solutions were thawed by allowing them to equilibrate to room temperature before being used. In methylation and binding assays, the MgCl2 was diluted to 50 μM. The RNA after preparation was visualized on a 12% native PAGE gel in 0.5× TBE. The gel was run at 4°C, for 10 h, visualized by SYBR Gold (Invitrogen) and imaged on a Storm 840 phosphorimager.

Protein expression and purification

The plasmids used for E. coli recombinant protein expression include pET28-hDNMT3ACopt, pET28-DNMT3A2Copt and pET28-hDNMT3A_catalytic_domain. Plasmid construction, expression and purification were previously described (28).

Native gel mobility RNA shift assays

Full-length DNMT3A, DNMT3A2 and catalytic domain and M.HhaI (concentration as indicated), were incubated at room temperature for 15 min with 100 nM RNA. Binding occurred in reaction buffer with 10 μM S-Adenosyl-l-homocysteine (AdoHcy). Samples were run on native 6% polyacryamide gels in 0.5× TBE pH 7.8 at 250 V for 30 min, stained with SYBR Gold and data were analyzed using ImageJ (29). KD values where determined from a single exponential using Prism V4.0 (GraphPad). Error bars represent the mean standard deviation of at least two replicates.

Methylation assays

Methylation assays were carried out in RNase free reaction buffer with 50 mM Tris–Cl, 1 mM EDTA, 1 mM DTT, 0.2 mg/ml BSA, 20 mM NaCl and 2 μM AdoMet at pH 7.8. Reactions were initiated by addition of pre-mixed DNA and RNA, and then quenched after 1 h. Details of the assay are the same as those previously described (28). IC50 assays used enzyme at 30 nM, DNA was either poly-dIdC at 5 μM or human promoters 10 μM, and RNA concentrations were varied as indicated in the figure legends. The RNAse control used 5 mg/ml of RNAse A pre-incubated with DNA and RNA for 5 min at 37°C before carrying out the methylation assay.

Fraction Inhibition was calculated by 1 − (product formed with RNA/product formed without RNA). The data were fit to a non-linear regression (one phase decay), where half life indicates IC50 using Prism. Error bars represent the mean standard deviation of at least three replicates. IC50 values for the RNA on each promoter were determined by measuring the amount product formed at varying RNA concentrations.

Mechanism of inhibition studies used full-length DNMT3A and the catalytic domain at 30 nM; poly-dIdC and CHD RNA were used at concentrations indicated in the figure legends. The data were fit using Sigma Plot 6.1 enzyme kinetics modular 1.0, by modeling a single substrate single inhibitor for competitive, non-competitive, uncompetitive and mixed type. Modeling using partial inhibition approaches resulted in the same R2 as the full-type inhibition. The models were ranked by R2 value. Error bars represent the mean standard deviation of at least three replicates. Data are graphically represented in Lineweaver–Burk form using Sigma Plot 6.1 enzyme kinetics modular 1.0.

Processivity assays

Processivity assays were carried out as previously described (28) with DNMT3A at 50 nM. Reactions were initiated by the addition of substrate DNA, poly-dIdC (5 μM). A total of 200 nM CHD RNA was added after 21 min. The control is concurrent addition of RNA and DNA at the start of the reaction. Data were fit to a non-linear regression (one phase decay). Data were fit separately before the experimental reaction had CHD RNA added (time 0–21 min) and after the CHD RNA was added (21–260 min), using Prism. Error bars represent the mean standard deviation of at least three replicates. The number of turnovers is calculated by the amount of production formation (nM) divided by the amount of catalytic active sites (nM) total protein concentration times percent active sites (18%) as previously described (22).

M.HhaI methylation IC50 assays were carried out like DNMT3A assays, but with M.HhaI (30 nM), AdoMet (5.5 μM) and reactions were run for 20 min. M.HhaI expression and purification are described in Matje et al. (30).

RESULTS

The three RNA sequences studied in this study were previously shown to directly or indirectly associate with one or more of the DNMT enzymes or have been implicated in the epigenetic regulation of specific genes (Table 1). The antisense but not the sense strand of RNA to the human EF1a promoter (EFS1) binds DNMT3A in cells, whereas DNMT1 showed no such binding (26). The EF1a RNA also caused transcriptional silencing in cells by inducing DNA methylation (17). The CHD RNA was implicated in DNA methylation changes localized to the E-cadherin promoter (27); although this study was retracted (31), the mitigating circumstances suggest that the body of work was correct. Indeed, the CHD RNA was successfully used to direct transcriptional gene silencing in subsequent work (32). The siRNA duplex to the NUP153 mRNA binds to mouse DNMT3A and DNMT3B but not to DNMT1 or DNMT2 (20). Thus, two of the RNA molecules tested in this study regulate transcriptional gene silencing in cells but the mechanism for regulation is not known. We tested the binding of these RNA to DNMT3A and test if these RNA molecules can modulate the activity of DNMT3A.

The secondary structures of the ssRNAs were characterized by melting curve analysis and native PAGE (Figure 1A). Three RNA preparation protocols were used for all RNA samples and were found to generate identical products for ssNUP153 and ssEF1a RNA molecules as seen by native PAGE. In contrast, the method of sample preparation dictates CHD RNA structure and aggregation (Figure 1A). Figure 1A shows the difference between unheated (lane 3), heated to 85°C then slow cooled (lane 1) or heated to 85°C followed by snap cooling on ice (lane 2). The unheated sample (lane 3) shows several forms: a highly aggregated (a) and a multimeric or secondary form of RNA (b and c). When the RNA was heated to 85°C followed by fast cooling on ice, the majority of the RNA is in the single-stranded unstructured form (lane 2). Heating the RNA to 85°C followed by slow cooling results in multiple bands (lane 1). Much of the RNA is located higher in the gel than the single-stranded form, with a band indicated as b, and smearing at c and some single-stranded unstructured form. There is also a loss in aggregation at the top of the gel (a), which occurred in the unheated sample. We found that the forms CHD RNA adopts under different preparation conditions have significant impact on the mechanisms and degree of inhibition.

Figure 1.

Figure 1.

Open in a new tab

Conformational and functional characterization of the CHD RNA. The CHD RNA takes on secondary structure that controls DNMT3A inhibition. (A) Inhibition curve for the CHD RNA on DNMT3A. Arrow bars represent standard deviation from three reactions. Black circle = CHD RNA unheated; grey circle = CHD RNA heated slowly cooled; grey square = CHD RNA heated fast cooled. (B) Native-PAGE gel of CHD RNA after different treatments, same amount of RNA loaded for each lane. Lane1: RNA heated slow cooled, 2: heated fast cooled and 3: unheated, a points to aggregated RNA still in the well, b and c are higher order structures and d is single-stranded unstructured.

We sought to examine the effect of CHD RNA on DNMT3A methylation activity using filter binding assays to measure changes in the amount of tritiated methyl groups transferred from cofactor to poly-dIdC DNA by the enzyme (28). All three preparation conditions of the CHD RNA caused inhibition on full-length DNMT3A (Figure 1B). Unheated CHD RNA shows the greatest inhibition with an IC50 of 24 ± 4.5 nM and 100% inhibition. The slow cooled CHD RNA has an IC50 of 33 ± 6.6 nM with 89% inhibition and the snap cooled sample showed the least amount of inhibition with 31% inhibition and an IC50 of 67 ± 31 nM (Supplementary Table S1). The maximal inhibition by CHD RNA on DNMT3A activity was observed with the unheated and slow cooled RNA correlating with these samples having mostly secondary bands in the native gel (Figure 1A). All subsequent CHD RNA experiments used slowly cooled RNA (Figure 1A).

CHD RNA is a potent inhibitor of DNMT3A. The inhibition seen by the CHD RNA occurs on the full-length DNMT3A (Figure 2A) with an IC50 of 40 ± 6.6 nM. We tested if the N-terminal domains mediate this inhibition, such as the PWWP domain, which was previously shown to bind DNA (22). DNMT3A2 is a truncated isoform expressed in early development (33), while the catalytic domain of DNMT3A (cDNMT3A) has been demonstrated to have kinetic properties that are essentially identical to the full length (22). The IC50 of CHD RNA on DNMT3A2 was found to be 24 ± 6.5 nM (Figure 2B) and 12 ± 1.4 nM against cDNMT3A (Figure 2C and D and Supplementary Table S2).

Figure 2.

Figure 2.

Open in a new tab

Inhibition by the ssCHD RNA. The modulation of DNMT3A activity was tested with four single-stranded and one double-stranded RNA, and only the CHD RNA altered DNMT3A activity. No inhibition was seen on the bacterial homolog. Inhibition curves for CHD RNA with the (A) full-length DNMT3A, (B) DNMT3A2 or (C) catalytic domain. (D) CHD RNA IC50 values for DNMT3A truncations. Inhibition curves for EFS1 RNA (E) and duplex NUP153 (F) on full-length DNMT3A. (G) The extent of DNMT3A and M.HhaI inhibition with different RNA molecules including RNase treatment of CHD RNA prior to DNMT3A activity assay. RNA was varied from 0 nM to 2000 nM with largest affect on activity given in the bar graph. Control has no RNA added. RNA and DNA (5 μM) were pre-mixed and combined with enzyme (50 nM) and AdoMet (2 μM) to start the reactions. (H) CHD RNA had no affect on M.HhaI activity.

We sought to determine if the CHD RNA-mediated inhibition of DNMT3A was sequence dependent via comparison to single-stranded and duplex NUP153 RNA and EF1a ssRNA. These RNAs were also prepared by slow cooling but additional experiments were done that showed preparation did not change the results for other RNA sequences tested. EF1a RNA was tested up to 2000 nM with no inhibition observed (Figure 2E). The sense, antisense and duplex NUP153 RNA also showed inhibition (Figure 2F and G). No changes in enzyme activity with EF1a or NUP153 (antisense, sense or duplex) RNA were observed with DNMT3A2 or the DNMT3A catalytic domain (data not shown).

To confirm the inhibition observed with CHD RNA is specific to DNMT3A, this RNA was tested against M.HhaI, a bacterial homolog with high structural conservation in the catalytic domain. CHD RNA does not inhibit M. HhaI tested up to 2000 nM, 150-fold greater than the DNMT3A IC50 (Figure 2H). To confirm that the CHD RNA causes the potent inhibition of DNMT3A, we treated the RNA and DNA (substrate) with RNase for 5 min prior to assaying for methylation activity. This pre-treatment resulted in complete loss of the CHD RNA-mediated inhibition (Figure 2G).

We confirmed the duplex NUP153 RNA binds full-length DNMT3A, as previously shown by Jeffery and Nakielny (20). The binding affinity of the RNA to DNMT3A was determined to be 520 ± 69 nM (Figure 3A). The ssEF1a RNA showed very similar results as the NUP153 RNA, binding to the full-length DNMT3A with a KD of 690 ± 89 nM (Figure 3B), but not binding to the catalytic domain. Minimal to no RNA binding was observed for the DNMT3A catalytic domain. However, the truncated isoform DNMT3A2 showed similar binding as the full-length protein (Figure 3C). We also tested RNA binding to the bacterial homolog M.HhaI up to four times the KD of NUP153 and EF1a with minimal to no binding for either RNA (Figure 3C). EF1a and NUP153 RNA bind the enzymes but do not cause inhibition like the CHD RNA.

Figure 3.

Figure 3.

Open in a new tab

RNA binds to DNMT3A allosterically. RNA-binding curve for DNMT3A full-length with duplex NUP153 (A). or ssEF1a, (B) error bars are standard error for at least two independent samples insert is Native PAGE gel showing the loss of a RNA (100 nM) band as an indicator of DNMT3A binding (varied from 0.2 μM–2 μM). (C). The amount of RNA binding to DNMT3A full-length, truncations and M.HhaI at 1 μM, RNA (100 nM), native PAGE of the unbound RNA. RNA type as indicated, data from three independent gel shift samples.

To determine if the CHD RNA-mediated inhibition is dependent on the nature of the DNA substrate, we tested the inhibitory effect on methylation of seven human promoters (28). RNA inhibition was observed to varying degrees with all human promoters. The extent of inhibition ranged from 60% to 91% and the IC50 values ranged from 22 ± 4 nM to 68 ± 22 nM (Figure 4A–D, Supplementary Table S3 and Supplementary Figure S1). We observed that the IC50 values is inversely related to the human promoters Km for DNMT3A (Figure 4D), this suggests that the substrate and RNA are binding to the same form of the enzyme. The 3-fold variation in RNA inhibition potency observed with different DNA substrates suggests that at least some of the inhibition involves a complex between the enzyme, DNA and RNA (see below).

Figure 4.

Figure 4.

Open in a new tab

The potency of RNA inhibition on DNMT3A is dependent on DNA substrate. Inhibition curves on the full-length DNMT3A (50 nM) with CHD RNA, with different DNA substrates (10 μM), (A) Survinin human promoter, (B) p15 human promoter and (C) IC50 values for CHD RNA with different substrates, inhibition curves are in Supplementary Figure S1. (D) Inverse relationship of the substrate Km to the IC50 values from the CHD RNA on DNMT3A, log–log linear regression R2 = 0.82.

Methylation assays were carried out to probe the mechanism of inhibition by varying the substrate concentration at multiple RNA concentrations. The results are shown in Figure 5. The inhibition data on the full-length DNMT3A best fit a non-competitive or mixed type model (R2 = 0.933; Figure 5 and Supplementary Table S4), however, the fit to a competitive model was not that different (R2 = 0.933). The catalytic domain showed competitive and mixed type (R2 = 0.951) as the best-fit models (Supplementary Figure 2 and Supplementary Table S4). Since mixed type inhibition involves an extra variable, non-competitive model best fits the data for the full-length DNMT3A and competitive for the catalytic domain. The non-competitive model best fits the data for the full-length DNMT3A and a competitive model best fits the catalytic domain data. Uncompetitive inhibition was the worst fit for both enzymes. Although the difference in fit between each model was not great enough to conclusively determine the mechanism of inhibition, that data suggest the RNA interacts with the same form of the enzyme as DNA, in addition to binding the binary form (enzyme–DNA complex).

Figure 5.

Figure 5.

Open in a new tab

Inhibition Model for CHD RNA on full-length DNMT3A. The difference between models was not substantial, but the data best fit a non-competitive or mixed type model. Data shown fit to a non-competitive model R2 = 0.933 (mixed type = R2 = 0.933, competitive R2 = 0.928, uncompetitive R2 = 0.92). CHD RNA inhibitor 0, 10, 30, 50 and 70 nM. Inhibition model for CHD RNA on the catalytic domain of DNMT3A is shown in Supplementary Figure S2.

Since conventional modeling of the mechanism was inconclusive, we used another approach to determine if RNA and DNA (substrate) bind to the same form of the enzyme by testing if the RNA can inhibit the enzyme once bound to DNA. Our previous demonstration that DNMT3A acts processively on substrates with multiple sites of methylation provides the basis for this approach (28). An initial control was carried out by adding the CHD RNA and DNA substrate at the same time, which gave the expected 80% inhibition. When RNA is added 25 min into the reaction no inhibition was seen for 45 min, at which point a 20% decrease in inhibition is observed. A similar level of inhibition is seen when DNA lacking CpG sites is added as the chase (28), suggesting that the CHD RNA does not inhibit the enzyme when bound to DNA (Figure 6). These results are consistent with the RNA-mediated inhibition occurring largely as a result of the enzyme dissociating from the DNA and then binding the RNA. However, consideration of all of the inhibition data suggest that DNMT3A inhibition occurs through multiple complexes, as discussed below.

Figure 6.

Figure 6.

Open in a new tab

The CHD RNA does not inhibit the enzyme when bound to DNA. DNMT3A was previously shown to catalyze multiple turnovers on a polymeric substrate, so RNA was added after DNMT3A bound to the polymeric substrate, to test if the enzyme is inhibited when bound to DNA. RNA-mediated inhibition occurs largely as a result of the enzyme dissociating from the DNA and then binding the RNA. CHD RNA (200 nM) was added 25 min after the full-length DNMT3A (50 nM) was combined with the DNA (poly-dIdC at 5 μM). Filled square = DNA, open circle = DNA then RNA added at 25 min, filled triangle = DNA and RNA added at the start of reaction. Arrow indicates when CHD RNA was added.

DISCUSSION

Although compelling evidence supports direct and indirect roles of RNA in the regulation of DNA methylation (15,16), we know very little about how RNA–DNMT interactions alter the function of these enzymes. Of the four ssRNA sequences we studied, only the CHD RNA altered the function of DNMT3A (Figure 3). The potency and extent of inhibition is very dependent on the conformation of the RNA (Figure 2), further suggesting that the RNA–DNMT interface is quite specific. The CHD RNA did not inhibit the highly structurally homologous bacterial DNMT M.HhaI (34), providing further evidence of a unique RNA–DNMT3A interaction (Figure 3H). Two other RNA molecules did bind DNMT3A, the NUP153 duplex and the ssEF1a, with no measurable catalytic changes to DNMT3A, suggesting there may be several mechanisms of interaction with DNMT3A alone. Cellular data also show multiple types of RNA can direct transcriptional gene silencing through alterations in DNA methylation (16,17,32,35). Our data suggest the events observed in the cells may occur through a direct interaction of RNA to DNMT3A.

DNMT3A binds CHD RNA with similar affinity as duplex DNA, whereas small molecule inhibitors show much weaker binding (36). The full-length DNMT3A data best fit a model invoking CHD RNA binding to at least one form of the enzyme already bound to the DNA substrate (e.g. non-competitive or mixed type inhibition). This form of allosteric inhibition is similar to previously characterized inhibition of DNMT1 by ssDNA (24). The catalytic domain results are best fit to a competitive inhibition model, which is inconsistent with the full-length data, so we utilized the fact that DNMT3A catalyzes multiple turnovers on a polymeric substrate (28) to test if the CHD RNA can inhibit the enzyme once bound to DNA. The lack of immediate inhibition upon addition of the RNA (Figure 6) shows that the dissociation of DNMT3A from its DNA substrate is required for CHD RNA inhibition.

Thus, inhibition by CHD RNA during catalysis on poly-dIdC is largely or exclusively mediated by the RNA binding to the same form of the enzyme, which binds the DNA. The most straightforward interpretation is that this involves the same active site, although this is not formally required. The catalytic domain of DNMT3A forms oligomeric complexes with itself (37), where all monomers are most likely not active. The CHD RNA may bind to an inactive subunit and or bind where it alters the stability of such a complex. For example, RNA binding may disrupt oligomerization, which may not occur once the enzyme is bound to DNA. The somewhat poor fitting of the inhibition data with the full-length enzyme may be due to the CHD RNA binding to both the catalytic site and an allosteric site where the other RNA molecules are binding.

The NUP153 and EF1a RNA also bound DNMT3A but no changes in activity were observed. These two RNAs, unlike the CHD, only bound to the full-length and DNMT3A2 and not to the catalytic domain alone, indicating allosteric binding outside the catalytic domain. The binding affinity was also weaker than the CHD or DNA by about 8-fold, also indicating a different binding site. The PWWP domain of DNMT3A and DNMT3B has previously been shown to bind nucleic acids and this domain is needed for heterochromatin association (22,38,39). Recent studies of several chromatin modification complexes revealed that thousands of RNAs including antisense ncRNAs bind to chromatin modifying complexes (12,40–43). Histone modifying enzymes bind RNA to associate with chromatin or other binding partners. The RNA acts as a scaffold bridging two independent complexes together, but no changes in catalysis has been observed (44). The RNA molecules that bind DNMT3A suggest a similar scaffolding mechanism may be occurring.

Regulatory implications of RNA–DNMT3A interactions

We showed that the RNA-mediated inhibition of DNMT3A is highly sequence-specific, involves the catalytic domain, and involves the RNA binding to the same form of DNMT3A which binds the DNA substrate as well as to the DNMT3A:DNA complex. What is the potential relevance of these observations to the regulation of DNMT3A function in vivo? RNA may regulate DNMT3A in vivo but with distinct functional consequences as a result of occurring in a more complex context involving other proteins or factors. DNMT3A is well known to interact with numerous partner proteins (6–9), and the impact of RNA on DNMT3A in such settings could alter the function of the RNA.

Inhibitory roles for RNA in vivo are uncommon but not without precedent (45). DNMT3A inhibition may serve to limit enzymatic activity at particular loci until the enzyme is either activated or directed to a particular genomic position. RNA-mediated inhibition of DNMT3A is prevented once the enzyme initiates methylation (Figure 6), which may reconcile the inhibition with localized activity. This situation is supported by our observation that DNMT3A inhibition varies with different DNA substrates. Our in vitro observations may provide a mechanism for the recent demonstration that knockdown of two antisense ssRNAs derived from CpG islands induce methylation at the corresponding promoters (46).

Alternatively, our studies were designed to reveal the functional impact of RNA molecules on DNMT3A, whereas the primary role of the RNA may be as a scaffold for DNMT3A and its partner proteins (16). This model is consistent with some RNA sequences binding DNMT3A without functional impact. RNA scaffolding is also the mechanism used by chromatin modifying complexes and transcription factors to direct site specific binding (47). The extensive transcription of the human genome (48–50) may provide RNAs, which either directly module DNMT3A function, and/or provide a scaffold to localize the enzyme, leading to proximal methylation (16). We suggest that the potent and sequence-specific in vitro modulation of DNMT3A function by RNA supports the in vivo mechanisms in which RNA is postulated to regulate transcription (32).

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online: Supplementary Tables 1–4 and Supplementary Figures 1 and 2.

FUNDING

Funding for open access charge: California Cancer Coordinating Committee.

Conflict of interest statement. None declared.

Supplementary Material

Supplementary Data
supp_40_17_8550__index.html (810B, html)

ACKNOWLEDGEMENTS

We thank Matt Purdy for help in the original design of the project and Douglas Matje for M.HhaI and critical reading of the manuscript.

REFERENCES

  • 1.Bird A. DNA methylation patterns and epigenetic memory. Genes Dev. 2002;16:6–21. doi: 10.1101/gad.947102. [DOI] [PubMed] [Google Scholar]
  • 2.Reik W, Dean W, Walter J. Epigenetic reprogramming in mammalian development. Science. 2001;293:1089–1093. doi: 10.1126/science.1063443. [DOI] [PubMed] [Google Scholar]
  • 3.Okano M, Bell DW, Haber DA, Li E. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell. 1999;99:247–257. doi: 10.1016/s0092-8674(00)81656-6. [DOI] [PubMed] [Google Scholar]
  • 4.Chen T, Ueda Y, Dodge JE, 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. doi: 10.1128/MCB.23.16.5594-5605.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Pradhan S, Bacolla A, Wells RD, Roberts RJ. Recombinant human DNA (cytosine-5) methyltransferase. I. Expression, purification, and comparison of de novo and maintenance methylation. J. Biol. Chem. 1999;274:33002–33010. doi: 10.1074/jbc.274.46.33002. [DOI] [PubMed] [Google Scholar]
  • 6.Denis H, Ndlovu MN, Fuks F. Regulation of mammalian DNA methyltransferases: a route to new mechanisms. EMBO Rep. 2011;12:647–656. doi: 10.1038/embor.2011.110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hata K, Okano M, Lei H, Li E. Dnmt3L cooperates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal imprints in mice. Development. 2002;129:1983–1993. doi: 10.1242/dev.129.8.1983. [DOI] [PubMed] [Google Scholar]
  • 8.Klose RJ, Bird AP. Genomic DNA methylation: the mark and its mediators. Trends Biochem. Sci. 2006;31:89–97. doi: 10.1016/j.tibs.2005.12.008. [DOI] [PubMed] [Google Scholar]
  • 9.Li YQ, Zhou PZ, Zheng XD, Walsh CP, Xu GL. Association of Dnmt3a and thymine DNA glycosylase links DNA methylation with base-excision repair. Nucleic Acids Res. 2007;35:390–400. doi: 10.1093/nar/gkl1052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT. Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science. 2008;322:750–756. doi: 10.1126/science.1163045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Wang X, Arai S, Song X, Reichart D, Du K, Pascual G, Tempst P, Rosenfeld MG, Glass CK, Kurokawa R. Induced ncRNAs allosterically modify RNA-binding proteins in cis to inhibit transcription. Nature. 2008;454:126–130. doi: 10.1038/nature06992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA, Goodnough LH, Helms JA, Farnham PJ, Segal E, et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell. 2007;129:1311–1323. doi: 10.1016/j.cell.2007.05.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Martianov I, Ramadass A, Serra Barros A, Chow N, Akoulitchev A. Repression of the human dihydrofolate reductase gene by a non-coding interfering transcript. Nature. 2007;445:666–670. doi: 10.1038/nature05519. [DOI] [PubMed] [Google Scholar]
  • 14.Matzke MA, Birchler JA. RNAi-mediated pathways in the nucleus. Nat. Rev. Genet. 2005;6:24–35. doi: 10.1038/nrg1500. [DOI] [PubMed] [Google Scholar]
  • 15.Aravin AA, Bourc'his D. Small RNA guides for de novo DNA methylation in mammalian germ cells. Genes Dev. 2008;22:970–975. doi: 10.1101/gad.1669408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Schmitz KM, Mayer C, Postepska A, Grummt I. Interaction of noncoding RNA with the rDNA promoter mediates recruitment of DNMT3b and silencing of rRNA genes. Genes Dev. 2010;24:2264–2269. doi: 10.1101/gad.590910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Morris KV, Chan SW, Jacobsen SE, Looney DJ. Small interfering RNA-induced transcriptional gene silencing in human cells. Science. 2004;305:1289–1292. doi: 10.1126/science.1101372. [DOI] [PubMed] [Google Scholar]
  • 18.Reik W, Walter J. Genomic imprinting: parental influence on the genome. Nat. Rev. Genet. 2001;2:21–32. doi: 10.1038/35047554. [DOI] [PubMed] [Google Scholar]
  • 19.Sun BK, Deaton AM, Lee JT. A transient heterochromatic state in Xist preempts X inactivation choice without RNA stabilization. Mol. Cell. 2006;21:617–628. doi: 10.1016/j.molcel.2006.01.028. [DOI] [PubMed] [Google Scholar]
  • 20.Jeffery L, Nakielny S. Components of the DNA methylation system of chromatin control are RNA-binding proteins. J. Biol. Chem. 2004;279:49479–49487. doi: 10.1074/jbc.M409070200. [DOI] [PubMed] [Google Scholar]
  • 21.Ross JP, Suetake I, Tajima S, Molloy PL. Recombinant mammalian DNA methyltransferase activity on model transcriptional gene silencing short RNA-DNA heteroduplex substrates. Biochem. J. 2010;432:323–332. doi: 10.1042/BJ20100579. [DOI] [PubMed] [Google Scholar]
  • 22.Purdy MM, Holz-Schietinger C, Reich NO. Identification of a second DNA binding site in human DNA methyltransferase 3A by substrate inhibition and domain deletion. Arch. Biochem. Biophys. 2010;498:13–22. doi: 10.1016/j.abb.2010.03.007. [DOI] [PubMed] [Google Scholar]
  • 23.Hermann A, Goyal R, Jeltsch A. The Dnmt1 DNA-(cytosine-C5)-methyltransferase methylates DNA processively with high preference for hemimethylated target sites. J. Biol. Chem. 2004;279:48350–48359. doi: 10.1074/jbc.M403427200. [DOI] [PubMed] [Google Scholar]
  • 24.Flynn J, Fang JY, Mikovits JA, Reich NO. A potent cell-active allosteric inhibitor of murine DNA cytosine C5 methyltransferase. J. Biol. Chem. 2003;278:8238–8243. doi: 10.1074/jbc.M209839200. [DOI] [PubMed] [Google Scholar]
  • 25.Bacolla A, Pradhan S, Larson JE, Roberts RJ, Wells RD. Recombinant human DNA (cytosine-5) methyltransferase. III. Allosteric control, reaction order, and influence of plasmid topology and triplet repeat length on methylation of the fragile X CGG.CCG sequence. J. Biol. Chem. 2001;276:18605–18613. doi: 10.1074/jbc.M100404200. [DOI] [PubMed] [Google Scholar]
  • 26.Weinberg MS, Villeneuve LM, Ehsani A, Amarzguioui M, Aagaard L, Chen ZX, Riggs AD, Rossi JJ, Morris KV. The antisense strand of small interfering RNAs directs histone methylation and transcriptional gene silencing in human cells. RNA. 2006;12:256–262. doi: 10.1261/rna.2235106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Kawasaki H, Taira K. Induction of DNA methylation and gene silencing by short interfering RNAs in human cells. Nature. 2004;431:211–217. doi: 10.1038/nature02889. [DOI] [PubMed] [Google Scholar]
  • 28.Holz-Schietinger C, Reich NO. The inherent processivity of the human de novo methyltransferase 3A (DNMT3A) is enhanced by DNMT3L. J. Biol. Chem. 2010;285:29091–29100. doi: 10.1074/jbc.M110.142513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Abramoff MD, Magelhaes PJ, Ram SJ. Image processing with ImageJ. Biophoton. Int. 2004;11:36–42. [Google Scholar]
  • 30.Matje DM, Coughlin DF, Connolly BA, Dahlquist FW, Reich NO. Determinants of precatalytic conformational transitions in the DNA cytosine methyltransferase M.HhaI. Biochemistry. 2011;50:1465–1473. doi: 10.1021/bi101446g. [DOI] [PubMed] [Google Scholar]
  • 31.Taira K. Induction of DNA methylation and gene silencing by short interfering RNAs in human cells. Nature. 2006;441:1176. doi: 10.1038/nature04952. [DOI] [PubMed] [Google Scholar]
  • 32.Li LC, Okino ST, Zhao H, Pookot D, Place RF, Urakami S, Enokida H, Dahiya R. Small dsRNAs induce transcriptional activation in human cells. Proc. Natl Acad. Sci. USA. 2006;103:17337–17342. doi: 10.1073/pnas.0607015103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Chen T, Ueda Y, Xie S, Li E. A novel Dnmt3a isoform produced from an alternative promoter localizes to euchromatin and its expression correlates with active de novo methylation. J. Biol. Chem. 2002;277:38746–38754. doi: 10.1074/jbc.M205312200. [DOI] [PubMed] [Google Scholar]
  • 34.Cheng X, Blumenthal RM. Mammalian DNA methyltransferases: a structural perspective. Structure. 2008;16:341–350. doi: 10.1016/j.str.2008.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Aravin AA, Bourc'his D. Small RNA guides for de novo DNA methylation in mammalian germ cells. Genes Dev. 2008;22:970–975. doi: 10.1101/gad.1669408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Stresemann C, Brueckner B, Musch T, Stopper H, Lyko F. Functional diversity of DNA methyltransferase inhibitors in human cancer cell lines. Clin. Cancer Res. 2005;11:9156S–9157S. doi: 10.1158/0008-5472.CAN-05-2821. [DOI] [PubMed] [Google Scholar]
  • 37.Jia D, Jurkowska RZ, Zhang X, Jeltsch A, Cheng X. Structure of Dnmt3a bound to Dnmt3L suggests a model for de novo DNA methylation. Nature. 2007;449:248–251. doi: 10.1038/nature06146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Qiu C, Sawada K, Zhang X, Cheng X. The PWWP domain of mammalian DNA methyltransferase Dnmt3b defines a new family of DNA-binding folds. Nat. Struct. Biol. 2002;9:217–224. doi: 10.1038/nsb759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Chen T, Tsujimoto N, Li E. The PWWP domain of Dnmt3a and Dnmt3b is required for directing DNA methylation to the major satellite repeats at pericentric heterochromatin. Mol. Cell. Biol. 2004;24:9048–9058. doi: 10.1128/MCB.24.20.9048-9058.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Yap KL, Li S, Munoz-Cabello AM, Raguz S, Zeng L, Mujtaba S, Gil J, Walsh MJ, Zhou MM. Molecular interplay of the noncoding RNA ANRIL and methylated histone H3 lysine 27 by polycomb CBX7 in transcriptional silencing of INK4a. Mol. Cell. 2010;38:662–674. doi: 10.1016/j.molcel.2010.03.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Guttman M, Amit I, Garber M, French C, Lin MF, Feldser D, Huarte M, Zuk O, Carey BW, Cassady JP, et al. Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature. 2009;458:223–227. doi: 10.1038/nature07672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Khalil AM, Guttman M, Huarte M, Garber M, Raj A, Rivea Morales D, Thomas K, Presser A, Bernstein BE, van Oudenaarden A, et al. Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression. Proc. Natl Acad. Sci. USA. 2009;106:11667–11672. doi: 10.1073/pnas.0904715106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Spitale RC, Tsai MC, Chang HY. RNA templating the epigenome: long noncoding RNAs as molecular scaffolds. Epigenetics. 2011;6:539–543. doi: 10.4161/epi.6.5.15221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, Lan F, Shi Y, Segal E, Chang HY. Long noncoding RNA as modular scaffold of histone modification complexes. Science. 2010;329:689–693. doi: 10.1126/science.1192002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Anderson SL, Minard KI, McAlister-Henn L. Allosteric inhibition of NAD+-specific isocitrate dehydrogenase by a mitochondrial mRNA. Biochemistry. 2000;39:5623–5629. doi: 10.1021/bi000272e. [DOI] [PubMed] [Google Scholar]
  • 46.Tomikawa J, Shimokawa H, Uesaka M, Yamamoto N, Mori Y, Tsukamura H, Maeda K, Imamura T. Single-stranded noncoding RNAs mediate local epigenetic alterations at gene promoters in rat cell lines. J. Biol. Chem. 2011;286:34788–34799. doi: 10.1074/jbc.M111.275750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Ponting CP, Oliver PL, Reik W. Evolution and functions of long noncoding RNAs. Cell. 2009;136:629–641. doi: 10.1016/j.cell.2009.02.006. [DOI] [PubMed] [Google Scholar]
  • 48.Huttenhofer A, Schattner P, Polacek N. Non-coding RNAs: hope or hype? Trends Genet. 2005;21:289–297. doi: 10.1016/j.tig.2005.03.007. [DOI] [PubMed] [Google Scholar]
  • 49.Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, Oyama R, Ravasi T, Lenhard B, Wells C, et al. The transcriptional landscape of the mammalian genome. Science. 2005;309:1559–1563. doi: 10.1126/science.1112014. [DOI] [PubMed] [Google Scholar]
  • 50.Yelin R, Dahary D, Sorek R, Levanon EY, Goldstein O, Shoshan A, Diber A, Biton S, Tamir Y, Khosravi R, et al. Widespread occurrence of antisense transcription in the human genome. Nat. Biotechnol. 2003;21:379–386. doi: 10.1038/nbt808. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Data
supp_40_17_8550__index.html (810B, html)
supp_gks537_nar-00904-m-2012-File001.docx (126.8KB, docx)

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES