Picrophilus torridus (strain DSM 9790) was cultured as described earlier (Arora et al., 2014). Briefly, cells were grown at 55°C in liquid culture with aeration, in medium of pH 1.0 comprising yeast extract (0.2%, Difco, United States), glucose (1%, Sigma, United States), potassium dihydrogen phosphate (0.3%), magnesium sulphate (0.05%), calcium chloride (0.025%) and ammonium sulphate (0.02%). For growth analysis, cultures were initiated with 1% (v/v) inoculum in media of different pH, from a starter culture at pH 1.0 that was at OD₆₀₀ ~ 1.5. Growth was monitored by measuring OD₆₀₀ every 24 h over a period of 14 days. Three experiments were set up in parallel and the average values of the three are plotted graphically, with error bars representing standard deviation. Genomic DNA was isolated from exponentially growing Picrophilus torridus cultures as described earlier (Arora et al., 2014). Briefly, harvested cells were resuspended in TEN buffer [20 mM Tris-Cl (pH 8.0), 1 mM EDTA, 100 mM NaCl], lysed by the addition of sodium lauroylsarcosine (1.6%) and Triton X-100 (0.12%), the lysate extracted with phenol, and genomic DNA precipitated using ethanol.

The ~1.3 kb gene encoding the S subunit was amplified off Picrophilus torridus genomic DNA using primers PtS-F (5′-TAAGATCTCATGAATAAAAAGGATTATAAT-3′) and PtS-R (5′-TACTCGAGATCACCACTTATTTTCAC-3′), with the proof-reading enzyme Phusion DNA polymerase (NEB, United States). The amplicon was cloned into pUC19, followed by subcloning into the BglII-XhoI sites of the pETDuet-1 vector (Novagen, United States), thus obtaining the pET-Duet/PtS clone. The ~1.7 kb gene encoding the M subunit was similarly amplified using the primers PtM-F (5′-TAGAATTCGATGGCAGATTTAAACTGG-3′) and PtM-R (5′-TAGCGGCCGCTTC ATAGTAACCAAG-3′), the amplicon cloned into pUC19, and then subcloned into the EcoRI-NotI sites of pET-Duet and pET-Duet/PtS, thus obtaining plasmids pET-Duet/PtM and pET-Duet/PtSM, respectively.

All mutageneses were carried out through PCR using Phusion DNA polymerase. For creating PtM mutants the plasmid pUC/PtM was used as template. The C-terminal deletion mutant was created by PCR using the PtM-F primer in combination with PtM_Δ537-576-R (5′-TAGCGGCCGCTTTATTGTCTATATATAAAG-3′). The ~1.6 kb amplicon thus obtained was cloned into pJET vector followed by subcloning into the EcoRI-NotI sites of pETDuet/PtS, yielding the plasmid pETDuet/PtSM_Δ537-576. The NPPW and xxGxxG mutants were created by overlap PCR. To mutate the NPPW motif, the N-terminal part of the gene was amplified using primers PtM-F and PtM-N360A-W363A-R (5′-CTGGTTCGCTGGAGGCGCCG CAAC-3′) while the C-terminal part of the gene was amplified using primers PtM-N360A-W363A-F (5′-GTTGCGGCGCCTCCAGCGAACCAG-3′) and PtM-R. The full length amplicon obtained by amplification with PtM-F and PtM-R primers was cloned into pJET vector followed by subcloning into the EcoRI-NotI sites of pETDuet/PtS, yielding the plasmid pETDuet/PtSM_N360A-W363A. To mutate the xxGxxG motif, the N-terminal part of the gene was amplified using primers PtM-F and PtM-G284S-R (5′-CCTGCAGTTGAACATGCCGG-3′) while the C-terminal part of the gene was amplified using primers PtM-G284S-F (5′-CCGGCATGTTCAACTGCAGG-3′) and PtM-R. The full length amplicon obtained using primers PtM-F and PtM-R was cloned into pJET vector followed by subcloning into the EcoRI-NotI sites of pETDuet/PtS, yielding the plasmid pETDuet/PtSM_G284S.

M.PtoI/S_211-222(A-G) was created using Overlap PCR. The N-terminal fragment of the mutant amplicon was produced using primers PtS-F and PtS_211-222(A-G)-R (5′-GGCCCCGGCCC CAGCTCCACCTGCCCCTGCTCCTGCCAACTGTATTTCCTT-3′) and the C-terminal fragment of the mutant amplicon was produced using primers PtS_211-222(A-G)-F (5′-GCAGGAGCAGGGGCAGGTGGAGCTGGGGCCGGGGCCTTCTGAAGGCATC-3′) and PtS-R. The full-length mutated gene was amplified using PtS-F and PtS-R and ligated into the BglII-XhoI sites of pET-Duet vector to create plasmid pET-Duet/PtS_211-222(A-G). This was followed by cloning the PtM gene into the EcoRI-NotI sites of pETDuet/PtS_211-222(A-G) to create plasmid pETDuet/PtS_211-222(A-G)M.

The M.PtoI enzymes were expressed in E.coli BL21 Codon Plus cells (Stratagene, United States) that had been transformed with their pET-Duet clones. Expression of the wild type or mutant proteins was induced in cells growing exponentially (OD₆₀₀ = 0.6–0.8) at 37°C using 1 mM IPTG. Cells were allowed to grow for a further 3 h at 37°C before harvesting by centrifugation and resuspension in lysis buffer (100 mM Tris.Cl pH8, 150 mM NaCl, 10% glycerol) containing protease inhibitors and lysozyme. This was followed by incubation of the cell suspension on ice for 20 min and lysis by sonication. After clarification of the lysates by high speed centrifugation they were subjected to cobalt-affinity chromatography using TALON metal affinity resin (Clonetech). The lysates were loaded on the column resin, the column extensively washed (with 100 mM Tris.Cl (pH 8), 1 M NaCl, 10% glycerol), and the bound proteins eluted using 100 mM Tris.Cl (pH 8), 250 mM imidazole, 10% glycerol. The M subunit of M.PtoI was similarly purified from bacteria transformed with pET-Duet/PtM.

Dot blot assay was used to analyze genomic DNA that was isolated from P. torridus cultures as described above. For this, 1 μg genomic DNA was digested with EcoRI and then denatured with 0.3 N sodium hydroxide for 30 min. This was followed by neutralization with 1 M ammonium acetate before spotting on Hybond N⁺ nylon membrane (GE Healthcare, United States) using the Bio-Dot apparatus (Bio-Rad Laboratories, United States). The membrane was baked at 80°C for 2 h before probing with anti-m6A antibody (Sigma Aldrich, Cat. No. ABE572) or anti-m5C antibody (EpiGentek, Cat.No. A-1014-010). This was done by blocking the baked dot blots with 10% skim milk (Difco Laboratories) for an hour at room temperature, washing the blots twice with 1X PBS-T, and then incubating them overnight at 4°C with the anti-m6A or anti-m5C antibodies (1:1000 dil in 1X PBS). Following three washes with 1X PBS-T the blots were incubated with HRP-labelled secondary antibody (1:10000 dil, Jackson Laboratory, United States) for an hour and developed using a chemiluminescence method.

DNA methylation assays were performed using a dot blot-based assay with lambda DNA (as the substrate) that had been isolated from phage particles hosted by an E.coli dam⁻dcm⁻ strain (DNA purchased from Thermo, United States; Cat no: SD0021). Reactions (10 μl) were typically carried out in a Tris-acetate based buffer [30 mM Tris-acetate (pH 5.0), 10 mM potassium acetate, 10 mM magnesium acetate, 100 μg/ml BSA] containing 100 μM S-adenosyl methionine (AdoMet; purchased from NEB, United States), 12 mM 2-mercaptoethanol, 20 mM sodium chloride, 1 μg DNA and 1.4 μM M.PtoI, at 55°C for 1 h. Reactions were stopped with 100 mM EDTA. Reactions were analyzed by dot blot assay as described above. The blots were quantified using ImageJ analysis, and the ratios of m6A with reference to E.coli genomic DNA spotted on the same blot was determined. Each experiment was done thrice. The graphs depict average values of three experiments with error bars representing standard deviation.

CD spectroscopy was carried out as earlier (Arora et al., 2014). Briefly, proteins (100 μg in 10 mM potassium phosphate buffer (pH 5.8)) were analyzed with a Jasco J-815 spectropolarimeter using cuvette of 0.1 cm path length. CD spectra were recorded over 250–190 nm at room temperature, in 1 nm steps at a scan speed of 200 nm/min. The presented spectra are the average of 20 scans. The data is depicted as the mean residue ellipticity (MRE).

The structures of the M and S subunits of M.PtoI were predicted using the Phyre2 (protein homology/analogy recognition engine version 2.0) online server (Kelley et al., 2015). ¹ The reliability of the predicted models were evaluated using the QMean and LG scores obtained (Benkert et al., 2009). ² In both cases the models that were considered demonstrated high coverage against their respective templates. All structures were visualized and illustrations created using PyMOL ³ (Mooers, 2016).

The study was initiated with examining the growth pattern of the organism in media of different pH. The data obtained revealed that while P. torridus grew comparably at pH 0.7 and 1.0, growth was considerably slower at pH 2.0 and no growth was discernible at pH 3–5 (Figure 1A). To determine if DNA methylation exists in this microorganism under the extreme conditions in which it thrives, genomic DNA was isolated from P. torridus cells grown in media whose pH ranged over 0.7 to 2 and the DNA analyzed by dot blot assay as described above. The results obtained clearly demonstrated that the Picrophilus torridus genome carries methylated adenine residues (m6A) regardless of the pH at which it is growing, although displaying lower levels of methylation at pH 0.7 (Figure 1B). No m5C methylation was detectable (Figure 1C), and we were unable to assess m4C methylation due to lack of quality antibodies.

A previous study by Koike et al. (2005) had established the presence of the dam gene in several archaea including P. torridus by computational analysis, but Dam-mediated adenine methylation was not experimentally verified in P. torridus. To check the methylation status of the P. torridus genome at Dam target sites (5′-GATC-3′) we exploited the property of the DpnI restriction enzyme that cleaves methylated but not unmethylated GATC sites. Genomic DNA isolated from Picrophilus cells grown in medium of pH 1.0 was thus subjected to DpnI digestion, and it was observed that while E.coli genomic DNA was cleaved by the enzyme the P. torridus genomic DNA was not (Figure 1D), nor was S.pombe genomic DNA (which lacks m6A methylation as evident from Figure 1B). These data indicate that while the P. torridus genome carries the m6A mark the organism does not harbor Dam-mediated adenine methylation.

An examination of the P. torridus genome sequence (KEGG database) ⁴ revealed that in addition to the dam gene, the euryarchaeon has genes encoding a Type II R-M system and a Type I R-M system. All Type I modification methylases characterized to date are m6A methylases, and thus we investigated the modification methylase of the P. torridus Type I R-M system. The three genes of the P. torridus Type I R-M system encoding for R, M and S subunits lie in a single operon on the lower strand of the circular genome (Figure 1E). The investigation of the modification methylase (henceforth referred to as M.PtoI) commenced with cloning the genes encoding M and S subunits.

The M.PtoI genes (designated herewith as PtM and PtS for M and S subunit genes respectively) were cloned by amplification using primers designed against the published PTO0078 and PTO0077 sequences (encoding M and S subunits respectively), as described in Materials and Methods. The pUC clones obtained were sequenced for verification. The sequences of the M and S subunits of M.PtoI were compared with those of typical Type I modification methylases EcoKI and EcoR124I as well as the identified archaeal Hvo MTase (HVO_2270–71), using Clustal Omega and BLASTp analyses (Sievers et al., 2011). ⁵ While the M subunit shared ~25–28% sequence identity over a coverage of 55–85% with the M subunits of the other enzymes (Figure 2), the S subunit shared between 23–25% identity over a coverage of 60–93% with their corresponding S subunits (Supplementary Figure S1). A comparison of the sequence of M subunit with the sequences of modification methylases of typical Type II systems using the same tools revealed that although all of them harbor the characteristic motifs discussed below (Supplementary Figure S2), overall there was no significant sequence identity (data not shown), reflecting the differences in the structure and subunit composition of Type I and Type II modification methylases.

m6A and m4C methylases are typified by the presence of nine conserved motifs, identified more than 25 years ago by analyzing the sequences of over 40 DNA MTases (Malone et al., 1995). By examining these motifs in the context of published structures of MTases it is now known that Motifs I-III and X are involved in the binding of the methyl group donor S-adenosyl methionine (AdoMet). It has been suggested that Motif I and motif X form a pocket that lodges the methionine group of AdoMet while motifs II and III are involved in interactions with the ribose and adenine moieties of the same. On the basis of structural as well as biochemical analyses it is also understood that Motifs IV-VIII are part of the active site pocket and are involved in catalysis of the methyl-transfer reaction, while the Target Recognition Domain (TRD) is responsible for recognition of the specific DNA (Malone et al., 1995; Bheemanaik et al., 2006). In Type I enzymes where DNA recognition is not a part of the functions of the M subunit the TRD lies in the S subunit. The recognition sequence of Type I enzymes is bipartite, and accordingly, the S subunits harbor two TRD domains. The MTases have been classified into six categories: α, β, γ, δ, ε, and ζ based on the order of these signature motifs in their primary sequence (Bujnicki, 2002). The majority of the m6A and m4C methylases fall in the α, β, and γ categories. All the Type I modification methylases as well as several Type II methylases are γ MTases, whose order of the motifs is X-I-II-III-IV-V-VI-VII-VIII, and M.PtoI conforms to this as seen in Figure 2. No clearcut motif II and motif III were identifiable in M.PtoI.

While the structures of the individual subunits of Type I R-M systems have been solved in case of several enzymes, the holoenzyme (M2S1) structures of only two Type I modification methylases are available to date. The EcoKI modification methylase structures (pdb IDs: 2Y7C and 2Y7H) were proposed based on single particle electron microscopy constructions at 18 Å resolution (Kennaway et al., 2009). The more recent EcoR124I modification methylase and restriction enzyme structures were proposed based on single particle cryo-EM at 4.54 Å resolution (Gao et al., 2020). The basic architecture of the M and S subunits of the two MTases and the interfaces of interaction among the different subunits are largely conserved between the two enzymes. The structures of the M and S subunits of M.PtoI were predicted by modelling using Phyre2. The modelled structures were viewed using PyMOL, which allowed us to superimpose the obtained structures on their respective templates to analyze structural alignments.

The most reliable model obtained for the M subunit used the EcoR124I enzyme as template (pdb ID: 7BTQ; Gao et al., 2020). The predicted structure of the M subunit displayed 27% identity with the structure of the M subunits in the 7BTQ cryo-EM structure (4.54 Å), over a coverage of ~83% (residues 89–153 and 164–569). As seen in Figure 3, seven motifs that were aligned in the primary sequence (Figure 2) were also aligned structurally. The M subunits of Type I RM systems are typified by a C-terminal helix that is involved in interactions with the S subunit, and the M subunit of M.PtoI too harbored this conserved helical C-terminal domain (Figure 3).

The most reliable structure obtained for the S subunit used the S subunit of the Methanocaldococcus jannaschii Type I R-M system (pdb ID: 1YF2; Kim et al., 2005) as template. The S subunit of M.PtoI showed 30% identity with the 1YF2 X-ray crystal structure of the M. jannaschii S subunit (2.4 Å), over ~97% coverage (amino acids 6–106 and 121–437). The S subunit of Type I systems harbors two TRDs (target recognition domains) that are oriented inverted with respect to each other. Each TRD carries a globular domain and an alpha-helical dimerization domain (Loenen et al., 2014a). The globular domains (TRDI and TRDII in Figure 4) are responsible for DNA binding, with the N-terminal one binding to the 5′ segment (specific sequence) of the bipartite recognition site and the C-terminal one binding to the 3′ segment (specific sequence) of the same. The alpha-helical domains of the two TRDs (named central conserved region or CCR, and distal conserved region or DCR; see Figure 4) associate with each other such that the two globular domains are held apart at a fixed distance, thus fixing the length of the non-specific spacer sequence in the bipartite recognition site. The architecture of the S subunit of M.PtoI was predicted to be similar, with two TRDs each harboring a globular domain and an alpha helical domain (Figure 4).

The S and M subunits of Type I enzymes characteristically interact with each other (either in the M2S1 or in the R2M2S1 complex) through a four-helix bundle that comprises the C-terminal helical domains of the two M subunits (depicted as CTD in Figure 3) and the two alpha-helical domains of the S subunit (depicted as CCR and DCR in Figure 4). As these regions are conserved in the M and S subunits of M.PtoI, the structure of the M.PtoI holoenzyme (M2S1) was predicted using template-based docking, via superimposition of the predicted 3D structures of M and S of M.PtoI (Figures 3, 4) on the M.EcoR124I holoenzyme structure (pdb:7BTQ). The M.PtoI docked structure obtained predicted the two M subunits of the M2S1 holoenzyme to interact with each other via their N-terminal domains, and the M and S subunits to interact with each other via the four-helix bundle (Figure 5).

The role of the C-terminal helical domain of M in mediating interactions with the S subunit was experimentally validated by creating a C-terminal deletion mutant of M: M.PtoI/M_Δ537-576, and analyzing it for loss of interactions with S. For this, the recombinant M.PtoI proteins were purified from E.coli. Recombinant M.PtoI (full length) was expressed in E.coli using the pET-Duet expression vector which allows the simultaneous expression of two proteins from two separate T7lac promoters, with one of the two proteins being expressed in fusion with the His tag at the N-terminus. The M and S subunits were thus simultaneously expressed in BL21 Codon Plus E. coli cells from the pETDuet/PtSM clone (as described in Methods) such that M was His-tagged. The cell lysates were subjected to cobalt affinity-based chromatography. SDS-PAGE analysis revealed the eluate to carry both subunits in stoichiometric amounts, leading us to conclude that the M.PtoI holoenzyme (M2S1) assembles in the host bacterium, allowing the S subunit to co-purify with the His-tagged M (Figure 6A). In creating the C-terminal deletion of M, amino acids 537–576 were excluded (mutagenesis detailed in Methods) and the M_Δ537-576 co-expressed with S in BL21 Codon Plus cells from the pETDuet/PtSM_Δ537-576 clone. When these cell lysates were put through TALON beads the eluate carried only the truncated M subunit, clearly demonstrating loss of interaction with S in M.PtoI/M_Δ537-576 (Figure 6B, left panel, fourth lane). The deletion did not cause any gross structural changes in the M subunit based on a comparison of the CD spectroscopy profiles of purified M (Figure 6B, right panel, fourth lane) and M.PtoI/M_Δ537-576 (Figure 6C). Taken together, these results indicate that the C-terminal helical domain of M is functionally conserved in M.PtoI.

The role of the central helical region of the S subunit in mediating M-S interactions was examined by creating a mutant S subunit where 12 amino acids of the central conserved region (CCR) were replaced by a 12 amino acid stretch of alternate glycine and alanine residues to create M.PtoI/S_211-222(A-G) (described in “Methods”). When the recombinant M.PtoI/S_211-222(A-G) protein was purified from E.coli using cobalt affinity chromatography it was found that there was a substantial increase in the ratio of M:S in the eluate fractions (Figure 6D, fourth lane), implying that the replaced region played a significant role in mediating M-S interactions. The difference in CD spectroscopy profiles of the wild type and mutant M.PtoI proteins most likely reflects the differences in M:S stoichiometry in the two proteins (Figure 6E).

M-S interactions in wild type and mutant M.PtoI proteins were also checked using purified proteins. For this, M and S subunits were separately expressed in E.coli cells from respective pET/Duet clones carrying either M or S genes. As all M subunits were tagged with His at their N-terminus, they were immobilized on Co-affinity resin. Whole cell lysates harboring recombinant S subunit (wild type or mutant) were added to the reaction, and after washing off the unbound fractions, the bound fractions (M subunits plus any interacting S subunits) were eluted using imidazole and analyzed. As seen in Supplementary Figure S3, while S subunit interacts with wild type M subunit (lane 5), it fails to interact with mutant M_Δ537-576 subunit (lane 6). The mutant S_211-222(A-G) subunit interacts with M subunit to a significantly lesser extent than wild type S (compare lane 7 with lane 5). These data confirm the role of the C-terminus of M subunit and residues 211–222 of S subunit in maintaining M-S interactions.

The ability of the purified M.PtoI to methylate DNA at adenine residues was analyzed using unmethylated lambda DNA isolated from phage that had been grown in a dam⁻dcm⁻ E.coli strain. The reactions were analyzed in dot blot assays using anti-m6A antibody, as described in Methods. Activity assays were initiated at 55°C for 1 h in Tris-acetate buffers whose pH ranged over 3 to 7, in the presence of sodium chloride ranging over 0–100 mM. As seen in Figure 7A M.PtoI was able to methylate lambda DNA (producing m6A) at pH ranging from 4 to 7, with optimum activity being detected at pH 5, which is near the organism’s intracellular pH of 4.6. Higher salt concentrations were detrimental to activity, with optimal activity seen at 0–20 mM NaCl. Thus, all further assays were carried out in Tris-acetate buffer (pH 5) containing 20 mM NaCl. The optimal concentration of magnesium ions that supported the DNA methylation reaction was determined over 0–50 mM magnesium acetate. It was found that while magnesium was necessary for the methylation reaction, concentrations of magnesium over 5–25 mM supported the reaction more or less equivalently, with a higher concentration of 50 mM being inhibitory (Figure 7B). The methylation reaction displayed an initial lag before proceeding almost linearly for approximately the first 45 min and plateaued beyond 60 min (Figure 7C). The ability of M.PtoI to methylate DNA was concentration-dependent, increasing almost linearly between 200 and 600 nM before plateauing (Figure 7D). High concentrations of enzyme (2,000 nM and beyond) proved to be detrimental, perhaps due to protein aggregation occurring at such high concentrations. In carrying out the assays at different temperatures ranging over 37–70°C it was found that M.PtoI was active at temperatures between 37–65°C with maximal activity at 55 and 60°C, in keeping with the conditions in which it thrives (Figure 7E). AdoMet supported the methylation reaction over concentrations varying from 10–500 μM, with 50–100 μM being optimal (Figure 7F). Interestingly, the enzyme displayed substrate inhibition at higher AdoMet concentration of 1 mM. Furthermore, the enzyme was able to methylate DNA, although to a much lesser extent, even in the absence of exogenously added AdoMet, suggesting that endogenous AdoMet remains bound to the M subunit through the purification process. This is not unusual for such enzymes, with similar findings in multiple enzymes such as in BpmI, BseMII, and EcoR124II (Dreier et al., 1996; Jurenaite-Urbanaviciene et al., 2001; Bath et al., 2002). Unsurprisingly, the addition of ATP had no effect on methylation activity (data not shown). Taken together, the results presented in Figure 7 establish M.PtoI as an active m6A MTase.

Type I modification methylases, like most adenine MTases, are typified by two primary domains they harbor in their M subunits: the AdoMet binding domain (motifs X, I-III in Figures 2, 3), and the catalytic domain that is responsible for the transfer of the methyl group from the donor to the nucleotide which is to be methylated (motifs IV-VIII in Figures 2, 3). Two motifs that are of singular importance are the AdoMet-binding motif I (FxGxxG) and the catalytic motif IV (NPPY/F/W). The catalytic motif IV lies in a pocket on the surface of the M subunit and at the time of methylation the specific adenine residue flips out of the duplex DNA substrate into the pocket where it stacks with the aromatic residue as well as forms hydrogen bonds with the first amino acid of the motif (which could by D, S, or N), stabilizing it for receiving the methyl group which is directly transferred to the target amino group on the adenine (Bheemanaik et al., 2006). Based on the sequence alignment with other Type I MTases (Figure 2), the FxGxxG motif in M is located at residues 282–287 and the NPPW motif lies at residues 360–363 (Figure 8A). The importance of these two motifs for the methyltransferase activity of M.PtoI was assessed by developing mutant enzymes and analyzing their activity. Accordingly, the glycine residue at position 284 was mutated to serine, and the asparagine and tryptophan residues at positions 360 and 363 were mutated to alanine (described in Methods). The mutated M subunits were co-expressed with S subunit in E.coli cells from pETDuet/PtSM_G284S and pETDuet/PtSM_N360A-W363A, respectively. The mutant enzymes M.PtoI/M_G284S and M.PtoI/M_N360A-W363A were purified (Figure 8B) and assessed for any gross structural changes using CD spectroscopy; no gross structural changes were detectable (Figure 8C).

The two mutant enzymes were assessed for their ability to methylate DNA in Tris-acetate buffer at pH 5 over sodium chloride concentrations ranging from 0–50 mM. M.PtoI/M_N360A-W363A was found to be virtually inactive under all tested conditions, while M.PtoI/M_G284S was less active than wild type enzyme at all tested conditions (Figure 8D). The activity of M.PtoI/M_G284S was compared with that of the wild type M.PtoI at different concentrations of AdoMet. It was observed that while the wild type enzyme supported DNA methylation more or less equivalently over 50–500 μM AdoMet, the mutant enzyme while displaying an overall lower activity showed maximal support of DNA methylation at 100–200 μM AdoMet (Figure 8E). These data establish that the NPPW and xxGxxG motifs in M.PtoI are critical to M.PtoI activity.