Mammalian DNA (Cytosine-5-)-methyltransferase Expressed in Escherichia coli, Purified and Characterized *

Besides modulating specific DNA-protein interac tions, methylated cytosine, frequently referred to as the fifth base of the genome, also influences DNA structure, recombination, transposition, repair, transcription, im printing, and mutagenesis. DNA (oytosine-S-j-methyl transferase catalyzes cytosine methylation in eu karyotes. We have cloned and expressed this enzyme in Escherichia coli, purified it to apparent homogeneity, characterized its properties, and we have shown that it hemimethylates DNA. The cDNA for murine mainte nance methyltransferase was reconstructed and cloned for direct expression in native form. Immunoblotting revealed a unique protein (M r = 190,000) not present in control cells. The mostly soluble overexpressed protein was purified by DEAE, Sephadex, and DNA cellulose chromatography. Peak methylating activity correlated with methyltransferase immunoblots. The purified en zyme preferentially transferred radioactive methyl moi eties to hemimethylated DNA in assays and on autora diograms, All of the examined properties of the purified recombinant DNA methyltransferase are consistent with the enzyme purified from mammalian cells. Fur ther characterization revealed enhanced in vitro meth ylation of premethylated oligodeoxynucleotides. The cloning ofhemimethyltransferase in E. coli should allow facilitated structure-function mutational analysis of this enzyme, studies of its biological effects in

The recognition sequence for DNA MTase is highly specific with almost all cytosine methylation occurring in the duplex palindrome 5'-C-p-G-3' (CpG). Over half of CpG dinucleotide palindromes are methylated in the mammalian genome (26). After semiconservative replication of DNA, both daughter duplexes are hemimethylated, and DNA MTase, which is localized to replication foci (27), fully methylates the duplex CpG dinucleotides. This process, termed maintenance methylation, restores the parental genomic methylation pattern and is consistent with the in vitro propensity of the DNA MTase for hemimethylated sequences (28)(29).
DNA MTase can also methylate certain CpGs that are not in a hemimethylated configuration, a process referred to as de novo methylation, Although the mechanisms for de novo methylation are not completely understood, a number of studies have reported the appearance of newly methylated CpG dinucleotides in the genome (29-32). Only one gene encoding mammalian DNA MTase has been found, and maintenance methylation and de novo methylation are generally believed to be catalyzed by a single enzyme (33)(34). Several studies have noted the appearance of de novo methylated cytosines in genomic regions containing preexisting methylated cytosines ti.e. methylation spreading) such as occurs in newly integrated viral DNA in the genome (31,(35)(36)(37). Since cytosine methylation can affect the DNA binding of certain transcriptional regulatory factors, the introduction of additional methylated cyto sines within gene regulatory sequences may influence gene expression (35). This spreading of cytosine methylation in gene regulatory sequences has been implicated in the gene silencing characteristic of fragile X syndrome (38)(39), cellular senescence (22), and X chromosome inactivation (13).
The importance of cytosine methylation in general and the DNA MTase in particular has led us to express this enzyme in Escherichia coli and to further study its mechanisms. Although the cloned cDNA for murine DNA MTase (33) has been expressed in mammalian COS cells (40), we report the first successful expression and purification of catalytically active mammalian DNA MTase in E. coli, providing a potential means for preserving native methylation patterns of cloned DNA in this widely used and simplified system. The purification to apparent homogeneity of DNA maintenance methyltransferase overexpressed in E. coli will facilitate mutational analysis of this enzyme and may allow its large scale production for crystallography. Studies of the effects of the recombinant methyltransferase on the prokaryotic genome and cellular processes will be useful in further elucidating the biological significance of DNA methylation.

18543
This is an Open Access article under the CC BY license.

EXPERIMENTAL PROCEDURES
Bacterial Strains and Plasmids-For standard transformations E. coli Sure cells (mcrA -, mcrCB-, mrr', hed', reels>, lacl" "; Stratagene) were used routinely. The prokaryotic expression vector pKK223-3 was obtained from Pharmacia Biotech Inc. The eDNA for murine DNA MTase (EMBL accession 14805 (corrected version» was kindly provided by Timothy Bestor (Columbia University) as overlapping coding sequences (pMG and pR2K) cloned into pBluescript SK M13+ (33). pMG includes all of the sequence from the EcoRI linker at the 5' terminus of the cDNA clone to a BglII site near the 3' end of coding (33). pR2K contains the sequence between the unique XhoI site at nucleotide 3138 and an Eco47III site just downstream of the AATAAA polyadenylation signal.
Plasmid Construction-The identity of pMG and pR2K was verified with endonuclease digestion. Each plasmid contained an internal XhoI site in the DNA MTase coding sequence as well as a 3' XhoI site in the pBluescript sequence (33). Both plasmids were digested with XhoI and gel-purified, and the XhoI-XhoI sequence of pR2K was ligated into the digested and gel-purified pMG plasmid lacking this segment. This fused the coding sequences at the XhoI site (nucleotide 3138) without alteration of the original sequence as confirmed with extensive restriction digests. The newly formed plasmid containing the entire DNA MTase coding sequence in pBluescript was used as template for PCR amplification (20 cycles) of the coding sequence and 3'-untranslated region using Vent DNA polymerase (New England Biolabs) and Perfect Match DNA polymerase enhancer (Stratagene) under standard conditions (see "Polymerase Chain Reaction"). The proofreading Vent DNA polymerase was used to assure amplification fidelity of the sequence and the Perfect Match enhancer was used to facilitate PCR of the approximately 5-kilobase pair segment. PCR amplification was chosen as a cloning strategy since the ATG codon should be within 15 base pairs of the unique EcoRI site in pKK223-3 for effective subsequent ribosome binding. Since there were no unique sites in the ATG region to allow cloning within this restricted distance from the ribosome binding site, primers were synthesized that created a SmaI site just 5' of the DNA MTase ATG start codon and a HindIII site just downstream of the 3' terminus (sense primer, 5' -CCTTACCCGGGATGGCAGACTCAAATAGATC-3'; antisense primer, 5' -CGGTTAAAGCTTTTGTAAAACGACGGCCAGT-3'). The PCR product was digested with SmaI and HindIII, phenol extracted, gel purified, and ligated into pKK223-3 between its unique SmaI and HindIII sites just 3' of the ribosome binding sequence. Restriction digests with SmaI and HindIII as well as several endonucleases at DNA MTase internal unique sites confirmed successful cloning of the cDNA into the expression vector within appropriate distance of the ribosome binding site (pTOT1, see Fig. 1). Primers unique to the DNA MTase coding sequence and internal to the original set of primers amplified the expected fragment from the pTOT1 construct but not from the control pKK223-3 vector (sense primer, 5'-ATGGCAGACT-CAAATAGATCCCC-3'; antisense primer, 5'-CTGGTGTGACGTCGAA-GACT-3'). The constructed pTOT1 expression plasmid contains a tac promoter (hybrid of the strong trp and lac promoters (41», ribosome binding site, complete coding region of DNA MTase (4565 base pairs) and 3'-untranslated region, termination signal (rrnB), T7 promoter, M13 primer sequence, ampicillin resistance, and a pBR322 origin of replication (see Fig. 1).
Polymerase Chain Reaction-Amplifications were conducted in a Perkin Elmer Cetus GeneAmp PCR System 9600 thermocycler. Standard procedures were used, and cycling consisted of 1 min at 94 "C, 3 min at 55°C, and 5 min at 72 DC. Aerosol-resistant pipette tips were used for assembling all PCR reactions.
DNA MTase Assays-Mammalian DNA MTase was routinely assayed in a 100-fLI volume containing a standard assay mix (10% glycerol, 50 mM Tris acetate, pH 7.8, 10 mM EDTA, 2 mM dithiothreitol, 5 fLg/fLI RNase A, 0.7 fLglml pepstatin, 0.5 mM Pefabloc SC, 0.5 fLglml leupeptin, 2 fLg/ml aprotinin (protease inhibitors were from Boehringer Mannhsimj), 3 fLCi of'{methyl-i'Hlscdolvlet (60 Cilmmol, ICN) at 1.5 fLM final concentration, 5 fLg of DNA unless otherwise specified, and enzyme sample (27,(29)(30)42). The enzyme does not require magnesium, and 10 mM EDTA is used to prevent any possible nuclease digestion of DNA substrates. Glycerol at 10% is used due to the inherent lability of the DNA MTase. Incubations were at 37°C for 1 h unless otherwise indicated. After completion of the assay, reactions were terminated by the addition of SDS to 0.6% followed by a 30-min incubation at 60 "C with 400 fLg/ml proteinase K (42). Two volumes of 0.5 N NaOH were added, and the samples were incubated at 60°C for 10 min to hydrolyze any remaining traces of RNA (42). The samples were cooled on ice, and carrier salmon sperm DNA was added (20 fLg/assay). DNA was precip-itated in 10% trichloroacetic acid, 5 mM sodium pyrophosphate for 15 min at 4 "C and washed 5 times on a Whatman GF/C filter with 5% trichloroacetic acid, 5 mM sodium pyrophosphate, and twice with 100% ethanol (29-30). Washed filters were transferred to 5-ml Scintiverse BD (Fisher) in glass vials and counted on a scintillation counter. One unit of DNA MTase activity is defined as the amount of enzyme required to transfer 1 pmol of tritiated methyl groups to DNA in 1 h (29-30).
DNA Substrates-Polymer polycdl-dCj-polytdl-dtf) (Pharmacia) was dissolved in 10 mM Tris, pH 7.5, and 100 mM NaCI, heated to 45°C for 5 min, aliquoted at 1 fLg/fLI and stored at -20 "C. Oligodeoxynucleotides were synthesized on an Applied Biosystems 380A DNA synthesizer using standard procedures (25). Methylated 5-cytosine (Glen Research) was added as the phosphoramidite where indicated (see Table I). All synthesized oligonucleotides were gel-analyzed and used only if complete synthesis was evident. For annealing oligonucleotides, complementary strands were mixed (500 ng/ul each) and incubated for 10 min at 75°C in 20 mM Tris-HCl, pH 7.5, and 50 mM NaCI, slowly cooled to room temperature, and analyzed for annealing efficiency on 3% agarose gels stained with ethidium bromide.
SDS-Polyacrylamide Gel Electrophoresis (PAGE)-Polyacrylamide gels (5%) were prepared and run at 50 rnA. Electrophoresis was terminated when tracking dye reached the bottom of the gel. Where indicated, gels were stained for protein either with silver nitrate (43)  Protein Analysis-Except for column fractions assayed for protein by absorbance at 280 nm, protein concentration was determined using the Bio-Rad Coomassie assay kit. Standard curves were established using y-globulin.
Cell Cultures-Cultures of E. coli Sure cells (transformed either with pTOT1 or control vector pKK223-3) of 5 ml and 50 ml in 2YT medium supplemented with ampicillin (100 fLg/m!) were successively grown to saturation from a single colony at 37°C. The tac promoter is not fully suppressed by the lac suppressor in this system, and some expression of the DNA MTase occurs in the absence of isopropyl-1-thio-j3-D-galactopyranoside (lPTG). Large scale cultures were inoculated with the saturated cell suspension (3.3 mllliter) and grown at 37°C until the absorbance at 600 nm was approximately 0.5. At this point, the cells were induced with 1 mM IPTG. After 3 h, the cells were harvested by centrifugation (4,000 x g for 15 min, 4°C), washed once in phosphatebuffered saline, and recentrifuged. Cells were either immediately lysed or stored frozen at -70°C as a cell pellet. We found no difference in DNA MTase activity between cells immediately lysed and those stored as cell pellets at -70°C overnight.
Purification-Unless noted, all procedures were carried out at 4°C. The washed cell pellet was resuspended in lysis buffer (50 mMTris-HCI, pH 7.5, 5% (v/v) glycerol, 2 mM EDTA, 1 mM 2-mercaptoethanol, 0.23 M NaCI, 0.1 mM dithiothreitol, 130 fLglmllysozyme, 0.5 fLg/mlleupeptin, 2 fLg/ml aprotinin, 0.5 mM Pefabloc SC, and 0.7 fLg/ml pepstatin) at 3 mllg of cells (45). Cells were blended (Waring) at low speed for 3 min, and after 20 min sodium deoxycholate was added with stirring to 0.05%. The mixture was blended for 30 s at low speed and sonicated (5 pulses on ice for 15 s). For more complete DNA sheering, the mixture was blended for 30 s at high speed. The sample was diluted with lysis buffer (4 mllg of cell pellet) lacking lysozyme, blended at high speed for 30 s, and centrifuged at 20,000 x g for 30 min (45). The supernatant (Sl) containing soluble protein was removed, and the pellet was resuspended vigorously in 100 ml of high salt (0.4 M NaC!) lysis buffer lacking lysozyme. The resuspended high salt mixture was centrifuged as for the Sl solution. The supernatant (S2) containing protein insoluble in 0.23 M NaCI was stored frozen at -70°C. Since several preliminary purifications indicated that over 80% of the DNA MTase was present in the soluble SI fraction and combining Sl and S2 reduced the resolution and yield of DNA MTase from E. coli, all subsequent purifications were carried out using only the Sl lysate (see Fig. 5C for solubility of DNA MTase).
The 81 lysate was dialyzed 3 h with two changes of 6 liters of dialysis/column buffer (20 roM Tris-HCI, pH 7.8, 5 mM dithiothreitol, 10% glycerol, 5 mM EDTA, and protease inhibitors as for the lysis the pTOTI expression vecto r. We hav e not yet fully qu an tified the degr ee of a ppa re nt cellula r prolifera tive a nd growth impairmen t . It seems possible that its ca use may be related to Lysates from E. coli cells transfor med wit h eithe r pKK223-3 (control vector lacking the DNA MTase insert) or pTOTl (expression vector cont ain ing the DNA MTase in sert ) were resolved by 5% SOS-PAGE , transferred to nitrocellulose, a nd pro bed with DNA MTase polyclona l a ntibody. La ne M , presta in ed high molecul a r weight protein marker ; lan e 1, control lysa te from cells transform ed wit h pKK2 23-3; lan e 2 , expression lysa te from cells tra ns formed wit h pTOT1. Th e unlabeled arrow indicates th e novel protein (M, = 190,000) not pr esent in control cells. Lysate samples were pre pa re d from 1 ml of a 5-ml satu ra te d cell sus pens ion (t ra ns formed eit her wit h pTOT l or pKK223-3 a nd indu ced wit h IPTG (see "Experime nta l Procedu res")), microcen trifuged 1 min a t room temperature, resu spende d in 100 iii of 1 X SOS gel loa din g buffer , heated to 100°C for 3 min , a nd loaded (10 u l) ont o SDS-PAGE (44). Overl apping coding seque nces for th e DNA MTa se (pMG a nd pR2 K, kindly provided by Tim othy Bestor, Columbia Univ er sity) wer e endonuclease digested, ligated, amplified, an d cloned into the pKK223-3 pr okaryotic expressi on vecto r downstream of th e lac promoter an d ribosome bind ing site (see "Experime nta l Procedures"). Depicted seque nces are t he 5' a nd 3' ju nct ions of the DNA MTase insert wit hin th e cloning vector. The T7 prom oter , M13 pri mer seque nce, an d termination region (ribosoma l term inator) are a lso illustrated.

RESULTS
We chose to ex pre ss th e mainten ance DNA MTase in its na tive form to a llow it s use in future in vivo studies (e.g. pr eservin g methylati on patterns of cloned DNA) without pote ntia l in terferen ce with activity or DNA binding from a fusion pr odu ct . Plasmid pTOTI was constructe d to expre ss th e native DNA MTase from th e st rong inducible lac pr omoter (Fig. 1). Immunoblotting kin etic st udies for DNA MTase ind ica ted full express ion of this enzy me within 3 h of IPTG inducti on (data not show n). We cloned pTOTI in to mer" (modified cytos ine rest riction ) cells to pr even t poten ti al DNA degr ad ation by th e mer syste m (47). Th e lysed E. coli cells containing pTOTI (expre ssion vector) revealed a uni qu e pro tein (M" = 190 ,000 ) on immunoblots prob ed with th e DNA MTase polyclon al an tib ody (Fig. 2). Thi s pr otein wa s not presen t in lysates of cells containing pKK2 23-3 (cont rol vect or lacking t he DNA MTase insert). Th e calcula te d molecul ar ma ss of t he DNA MTase is 172,238 ba sed on it s coding seque nce. However , t his enzy me has been shown pr eviou sly to resolve a t a n a ppare nt relative molecul ar mass of 190,000 on SDS-PA GE gels, which is t hought to be du e to posttran slational modifica tion s of t he enzyme a nd/o r its molecul ar sha pe (33).
E. coli cells ex pressing mammali an DNA MTase do not appear to gro w as well as cells containing the control vector a nd typically requi re 3.5 h to reach a n A 6()() of 0.5 at 37 "C, wh ereas con trol cells reach this stage of gro wt h within 3 h. Th e pTOTI cell s produc e slightly sma ller colonies on cul ture plates a nd less tur bid overn ight cult ures compa re d wit h cells containi ng the control vector (da ta not show n). Th ese differenc es in compa rison to control cells becam e more pr onounc ed as th e cells were t ran sferred to success ive cult ure plates over a period of se vera l month s. To pr even t pr ogr essive cellula r proliferative retardation, we periodicall y t ra ns for me d fresh mer" cells with buffer). The dia lyzed Sl solution was diluted wit h a n equa l volume of column buffer an d loa ded onto a DEAE-Seph acel column (2.5 x 12 cm bed volu me). The elute d column was wash ed wit h 2 bed volumes of column buffer to remov e unbou nd prot ein a nd further elute d wit h a 200-ml 0 -400 mMNa CI gra dient. Fraction s wer e collected a nd assayed for methylating activity as indicated a bove. Active pooled fraction s were store d at -70°C in 50% glycerol, 5 mM EDTA.
For a mmoniu m su lfate precipitati on, pooled acti ve fractions from DEAE chro matogra phy were diluted wit h 1 volume of colu mn buffer and brought to 30% a mmo ni um sulfate wit h ge ntle stirring over 10 min followed by cont inued st irring over 20 min on ice. The mixtu re wa s centrifuged for 20 min a t 10,000 x g , a nd th e supernata nt was brou ght to 60% a mmonium sulfa te, st irre d, and recentrifuged. Th e 60% a mm oni um sulfate pellet containing th e DNA MTase (30) was resu spend ed in 2 ml of column bu ffer, load ed onto a Sep ha dex G-150 column (2 x 70 cm bed volume) and elute d. Active fractions wer e pooled a nd sto red as a bove. DNA cellulose (4 mg of double-stranded DNN g of solid, Sigma) ch roma tography wa s performed in a 1 x 3-cm bed volu me a nd elute d with a 0-400 rnxt Na CI gradient. Active fra ct ions wer e pooled a nd store d a t -70°C in 50% glycerol, 5 mM EDTA.
Gel S canning ofSDS -PAGE-To estimate percen tage of tota l cellu lar protein represen ted by the recombinant DNA MTa se, Coomassiestai ned 5% S DS-PAGE gels were sca nned on an Apple OneScanner, plotted , and integrated for density using the Image 1.49 program on a Macint osh Il fx compute r . A tota l of four differen t pr otein concentration s of the S l a nd S2 lysates (representing a 4-fold differ enc e in total protein load ed) from two independ en t DNA MTase puri ficatio ns were resolved on S DS-PAGE gels, sca nne d, a nd plotted in du plicate. The M , 190,000 protein band (ide nt ified with molecul ar mass mark ers) was integrated for den sity in each lan e a nd compared with the total integrated den sity of a ll proteins in the sa me la ne to obtain percentage DNA MTase of tota l E. coli protei n. For s imilar determinat ions in the mam mal ian sys te m, a photomicrograph (kindly provided in reprint form by St even Smit h, City of Hope Nationa l Med ical Center, Dua rte, CAl of a SDS-PAGE ge l resolvin g th e crude lysa te fraction of total hum an placental pr otein containi ng th e iden tified DNA MTase (46) wa s also ge l-sca nned, plotted , a nd integr a ted for den sity as for the E. coli crude Iysat es. effects of DNA MTa se express ion on the E . coli genome, although th e large size of the nove l protein product it self ma y a lso be a factor. Evid enc e for methylation of high molecu lar weight E. coli genomic DNA in vit ro can be seen (see Fig . 6B ), suggest ing tha t a simila r process may occur in vivo affecting th e growt h of th ese cells.
To assess enzyma tic catalysis by th e cloned DNA MTa se , the DNA methylating activity (as measured by transfer of tritiated methyl gr oups from AdoMet to DNA) of DEAE-purified fraction s was compared for cells transform ed with pTOT1 and pKK223-3 (Fig. 3). Th e DEAE columns were simultaneously chromatogr aphed and elute d with a sa lt gradient. Peak methylating a ctivity for th e fractions from th e pTOT1-transformed cell lysates eluted in th e range of 100 -150 mxt Na CI, consistent with resu lts of t he DNA MTase pu rifi ed from mamma lian cell s (29 -30, 48). No obviou s methylating peak was see n for th e DEAE-chromatograp hed Iysates of pKK22 3-3-transformed control cell s. Immunoblots performed on th e pTOT1 DEAE fractions indicated a M ; 190,000 protein correlating with peak methylating activity (fra ctions 32-52; pTOT1 ), which wa s not a ppa re nt below 100 mxt sa lt (fractions 8 -30; pTOT1 ) or abov e 150 mxt salt (fraction 71; pTOT1 ). Fig. 3 a lso shows that th e most inten se M I' 190,000 bands (fra ctions 38-42; pTOT1 ) correlated wit h fractions havin g th e highest methylating activity. Th e contro l DEAE column showed no evidence of th e M; 190,000 protein as indicated by the abse nce of this band at peak methylating activity for t he pTOT1 column (fraction 38; pKK223-3).
Th e cloned DNA MTase was purified to apparent homogen eity by assaying for methylating activity in a three-column system based on protein charge (DEAE), size and sha pe (Sephad ex), and DNA-affinity (DNA cellulose) (Fig. 4). Due to the pr esenc e of va rious inhibitory substances in crude fractions and la bility of th e enzyme (49), meaningful estima tes of tota l purification fact or could not be obtained consistent with reports by oth ers (29-30, 50 ). Gel filtration yielded a single peak of methylating activity in th e M; 180 ,000 -205,000 range, consistent with polya crylamide gel es tima tes. Whi le si ze se pa ra tion is efficie nt in this expression sys te m du e to th e relatively large si ze of th e mammalian DNA MTase compared with most E. coli protein s (Fig. 5G), some protein impurities remain in th e Sepha dex fraction, and a fina l purification based on th e affinity of t hi s enzyme for DNA is quite effective in producing a homog eneou s purification as asses sed by silver sta ining (Fig. 4D ). Although improvements of the purification proced ure are expect ed to increa se th e yield of recombinant DNA MTa se, we recovered a lmost a full milligram (887 J.1g) of a pparently pure enzyme from about 10 lit ers of E . coli cells. Th e apparently homo gen eou s protein exhibiti ng peak methylating activity followin g DNA cellulose chromatogra ph y reacted with th e DNA MTa se a ntibody on immunoblots (Fig. 4E ).
Partially purified recombinant DNA MTa se wa s us ed for compa rison of subs t ra te pr eferenc e with increasing DNA MTa se purit y, assessm ent of relative effectiveness of purification steps, es tima tes of solu bility a nd degr ee of expression of the cloned DNA MTase in E. coli , a nd DNA su bst ra te analysis st udies (Fig. 5). A hemimethylated oligodeoxyn ucleotide was sy nt hesized containing methyl moieti es at approximately 15base pair intervals for use as subs trate in DNA MTase assa ys (see "Experimenta l Pro cedures" for che mica l synt hesis and Table I for st ruct ure of hemimethylated oligodeoxynucleotide). Preferential transfer of radioactive methyl moieties to the oligodeoxynucleotide subs t ra te containing hemimethylated CpG sites over th e control la cking subst rate was a ppa re nt a fter DEAE purification (Fig. 5A ), a nd thi s ratio improved with gel filtration (Fig. 5B ). Ethidium bromide sta ining of agarose gels ind icated minor amounts of la rge molecular weight E . coli genomic DNA present after DEAE purification (da ta not shown), accounting for the slight activity of control assays la cking oligonucleotide subst ra te (Fig. 5A ). Th e chemically synt hesiz ed hemimethylated oligodeo xynucleot ide un derwent gr eater methylating activity in DEAE and gel filtration fraction s t ha n t he highly methylatable de novo subs t ra te , poly(dI ·dC)·poly(dI·dC), indicating pr eferentia l hemim ethylation by th e recombinant DNA MTase.
Th e pooled active fraction s as well as th e crude Iysates were assessed on polya cryl am ide gels for protein conte nt a nd purity (Fig. 5G). The soluble csn and insolu ble (S2) SDS -PAGE crude lysate fractions were estima te d for perc ent DNA MTase by scanning stained gels (see "Experi me ntal Proc edures"). Th e DNA MTa se comprised approximately 2% (ra nge of 1.0 -3.0%) of total E . coli protein in the Sl fraction and a bout 0.3% (range of 0-0.53%) for the in solub le S2 fraction , indicating that appro xim ately 85% of the enzyme is expressed in soluble form (see Fig. 5G for comparison of Sl and S2 fractions ). Th e overall expression of DNA MTa se in th ese cell s is about 2.5% of tota l E. coli protein . By contrast, mammalian cells contain a mean of 0.05% DNA MTase of total human placental protein (see "Experimental Procedures" under "Gel Scanning of SDS-PAGE" and Ref. 46 ).
To demonstrate that the recombinant DNA MTa se is ind eed active with a pr eferenc e for hemimethylated DNA, we reacted th e partia lly purified enzyme with oligodeoxynucleotides in the   Oligonucl eotid e * * * o All oligodeoxynucl eoti des are 60 ba se pa irs in length (full sequence shown at top of tabl e) except wher e ind ica ted. Ast eri sk s ind icate position of methylated cytosines placement during che mical synthesis of oligodeoxynucleotide subst ra tes .
b Assays wer e conducted with excess subst rate (5 ,.,.g of DNA) for 3 h . Oth er assay conditions wer e as indicated under "Experi menta l Procedures." Each val ue is th e mean :!: S.E. of three independent determinations. Cont rol va lues (i.e. sa mples otherw ise identical to a nd assayed side-by-side wit h substrate -containing sa mples but la cking added DNA substra te; mean = 144.1 :!: 6 pmollh) were subtracted from eac h sa mple va lue in eac h individua l experi me nt before det ermination of th e indi cated mean s a nd S.E. All assays utilized 40,.,.g of parti ally purified DNA MTase (SD fract ion, Fig.5C).
presen ce of ra dioactive AdoMet, resolved t he sa mples on agarose gels, and subje cte d the gels to a utoradiography (Fig. 6). The gel-isolated hem im ethyla ted oligonucleoti de produ ced t he most intens e band on a utoradiography , dem onst rati ng pr eferential transfer of methy l moieti es to hemimet hylated CpGs . Some radioactivity was a pparent in the otherwise identical non methyla t ed oligodeoxynucleotide (i.e. de novo methyl ation ), a nd t h is act ivity was greater t ha n t hat for t he ide ntica l full y meth yl a ted oligon ucleot ide containi ng no methyla t abl e CpGs . Th us it is a ppa re nt that the recombinant DNA MTase transfers methyl moieti es directl y to t hese oligodeoxynucleotides with a preference for hemimethylated CpG sites and with a much lower propen sity for nonmethyl at ed CpG sites. Very little methyla t ion a ppears to occur at sites oth er than CpG (Fig. 68 ,  la ne 4 ).
To furth er characterize t he enzymatic a ctivity of the DNA MTase purified from E . coli, we qu antita t ed in assays the methyl recept ivity of othe rwise id entical oligon ucleoti des differing on ly in placem en t of methyl moieties (Ta ble I). Th ese a nalyses u t ilized the mor e purified gel filtration fraction (Fig.  5C ) containi ng no evide nce of con tamin a ting E. coli DNA. Tabl e I show s t hat the h emimethyl ated oligodeoxynucleotide substrate received t he most radioactive methyl transfer catalyze d by the recombinant DNA MTa se consi stent wit h the DNA MTa se partiall y purified from mammalian cell s (5 1). Also simila r to t he mammali an cell enzyme , nonmethyl ated oligonucleotides can undergo de novo methyl a tion, and se que nces containin g no methyla t abl e CpGs (i .e. prem ethyla t ed a t a ll CpG si tes) are poor te mplates for the DNA MTa se (Table I) , dem onstrating it s strong pr eferenc e for cytosine methyl ation spec ifica lly in CpG dinucleotides (29 -30, 5 1-52). A duplex t r imethy lated oligodeoxynucleotide con taining on ly two de novo methyl atabl e CpGs on eac h st a nd (Ta ble I) is mor e recep tive t o de novo methyl a ti on (22.9 pm ol/h/CG) t ha n a n otherwise identical non methylated oligo nucleotide containi ng five de novo methyl atable CpGs on ea ch sta nd (7.3 pm ol/h/ CG ), indicating en ha nce d de novo methyla ti on of a prem ethyl ated oligodeoxyn ucleot ide conta ining methyl atabl e CpGs, DI SCUSSIO N Th e wide ly-used t echniques of DNA cloning a nd PCR a mplification st ri p mammali an ge nomic DNA of it s ori gin al cytos ine methylation. DNA th a t lack s its nat ive cytosi ne methyla ti on patt ern may give different result s in mobili ty shift a na lysis, endon uclease digestions , a nd oth er pro cedures a nalyzing it s properti es a n d beh av ior . We developed the idea th at t he methylation pa tt ern of clon ed DNA could be preser ved in host bact eri a ex pressi ng t he maintenance DNA MTase. However, the cDNA for this enzy me h as previously been ex pressed only in mammali an cells (COS-I ) (40). Wh er ea s this may be of use in stu dyi ng the effects of va ria t ions in DNA MTa se levels in mammali an cells, we chose to clone a nd ex press DNA MT ase in E. coli. We develop ed this system not only for its possibl e u se in maintaining methylation patterns of clon ed DNA in bacteria bu t a lso because of th e wid espread u se of E . coli as a prot ein expression syste m , the sim plification of cell culture a n d purification processes , t he poten tial of large sca le production of th e enzyme for crystallography , an d t he fac ilit ati on of mut agen esi s stu dies of t his enzy me .
Th e known pot ential for de novo methyla ti on a nd methyl ation spreading by the DNA MTase (29-30, 35 ) could be a factor in preservin g methyl ation patterns of ge nomic DNA in t his sys te m; how ever, both of th ese proc esses occur in proportion to greate r DNA MT ase levels (48, 5 1) a nd number of cell ge nera-Li on s (3 1, 35). Modul a ting t h e DNA M'I'a se exp ress ion by limiting IPTG inducti on a n d m in im izin g ce ll cu lt uring t im es ma y be us eful a p proach es for reducing t h e poss ibility of de /l O UO me thyl a ti on . An al ysi s of the clone d product wit h methylation -s ensitive isos ch izomers (44) or m ethyl ati on seq ue ncin g (53)(54) would be prudent to assess the possibility of ectopic methyl ation .
Previousl y it wa s t ho u ght that t h e m ammali an D A MT a s e mi ght be toxi c to E. coli s ince d e /lO UO methyl ation of t h e E. coli ge nom e m a y ac t ivate the m er sys te m leading to DN A degrad ation (4 7) , ev en th ou gh the m ammalian DNA M'I'a s e is prima r ily a m aintenance methyl t rans fe rase a n d a ppea rs to de /l O UO methyl ate onl y a s a secon da ry fun ct ion (29)(30). In orde r to circ u mv e n t thi s pot enti al problem , we clon ed the rec on structed murine MTase c DNA in m er " cell s , The m erce lls ex press ing DNA M'I'a s e a re s lig h tl y les s pr olife ra t ive th an con t ro l m er " ce lls ti,e. co n ta in ing the clo n ing vec to r a lone ), perhaps rel a t in g to de novo methyl ati on of t h e E. coli ge nome . Transforma t ion of th e vec to r into fres h m er ce lls a ppears to im p rov e ce llu la r proliferati on to ne ar cont ro l le vel s . In s p ite of thi s m in or g rowt h impairment, thes e ce lls a re a b le to ove rex pres s t h e DNA M'I'a s e to rel ati vel y h igh le vel s com pa red with the level s of t hi s e nzy me in mammali an ce lls . S t u d ies of the e ffects of exp ression of the clon ed mamma lian DNA M'I'a s e on the E. coli ge nome, on t h e con t ro l of ce llu lar proces ses in E. coli , a n d on repl ica ti on ra tes as well a s ce ll viability m ay con t r ib u te to unde rstand in g the con t ro l mech ani sm s of this e nzy me a nd its biologi cal s ig n ifica nce . A number of prok aryotic ce llu la r con t ro l proces ses cou ld be affecte d by expression of thi s recombinant e n zy me in t h es e ce lls s u ch as th e t ranscripti on of key regulatory ge nes, DNA repair, replicati on, a n d recombina ti on . We h a ve pre viou sl y reported seve ra l theoretical mo lecul ar m ech a n isms of ce ll u la r se nesce nce (22) , a h allmark of whi ch is reduced repli cati ve ca pa city , a nd h a ve s uggeste d th at d e novo methyl a t ion by t h e DNA M'I'a s e may con t r ibu te to th is ph en ome non in aging e u ka ryo t ic ce lls (22 1. P rok a ryot ic ce lls do not senesce (2 2), a n d s t u d ies a rc in progress a na lyz ing t he E. coli ce lls now ex pressing th is prote in for ev ide n ce s u ggest ive of senescing ce lls (/'.J:. morphologi ca l ch anges , s lowing of ce ll re pl ica t ion ) a nd th e DNA M'I'as e as one of t h e pu ta t ive "mortality ge ne" product s .
T h e m ainten ance DNA M'Ta se , purifi ed from mamm a lia n ce lls, is h ig hl y s us ce pt ib le to proteolytic degradation 1:l4) and los s of e nzy me activi ty d ue to it s lability 129-:l0'-Moreover, th e DNA M'I'a se is p resen t in very limited quantiti es in mamm alia n ce lls (46) . Th e ex p re s sion of th e D ' A MT nse in E. coli and purifi ca t ion of thi s e nz yme to a pp arent homogenei ty m ay h e lp overcom e some of these p ro ble ms . It is ge ne ra lly kn own t hat the use of E. coli a llow s rapi d , easy g ro wt h of large numbers of ce lls wi t h less e nd oge n ous protein heterogen eity a n d byp a ss ing of nu clear' isolation p ro toc ols . In th e case of this s peci fic e nzy me , its pu r ifica t ion from E. coli ma y a lso be fa cilita ted by its re la t ive ly large s ize compared with most E. coi i pr ote ins , allowin g more e ffective s ize se pa ra t ion a nd reducin g th e risk of proteoly t ic degrad a t ion a n d los s of e n zy me activity. Wh a te ve r t h e cho ice of purifica t ion protocol. th e ex p ressio n of mammalian DNA MTa se in E. coli s ho u ld a llow grea te r avai la bi lity of pu rified enzy me . All of th e prop erti es of th e purified n -comhin a n t DNA MT a s e examin ed in th is study includ in g rel a t ive molecu la r m a s s , elution in sa lt grad ie n ts, a ffini ty for D1\'I\, immunorea ct ivity , and s u bs t ra te preferen ce arc consistent wit h th e k nown prope r t ies of the e n zy me purified from ma mm al ia n ce lls (29-30, 33-34 , 49 , 5 1-52).
Alth ou gh it is ge ne ra lly thou ght tha t the e u karyo tic DNA M'l'a se is ca pable of maintenance a n d de /lOVO me thyl ation with out a s si s tance from associa te d mammal ian pr ot ei ns or factors , th is import an t question is s t ill not full y res ol ved 15:; ). P u r ifica t ion of t he ma m ma lia n e n zy me h a s h elped a d d ress this is sue , bu t minor contaminants that ass is t th e DNA M'I'a se coul d st ill be presen t in apparently pure fra ct ion s . O u r studies in di ca te tha t t he enzyme is in de ed ca pa b le of both ty pes of DNA methy la t ion . T h e ex pressed product in E. ('(IIi was originally de ri ved fr om a s ing le mammali an ge ne (33 ), a nd whe n this cDNA is ex pre s se d in E. coli and purified , it ca n p rfor m bot h ma in te n a nce and de novo methyl ation of D A. Wh ate ve r oth -I' prot e in s m ay be invo lved in th e e u ka ryo t ic methyl ation process, it is clear th a t t he essential features of m a inten ance a n d ell' nouo m ethyl a t ion a re not d ep endent upon associa te d prot ins u niq u e to the m amma lia n replication apparatus .
S im ila r to the mamma lia n DNA MT a se iso la te d from mamm ali an ce lls (56)(57)(58), the recombinant e nzy me purified from E. coli h a s a prefe ren ce for hemimethyl ated C pG d inucleot ides , h a s a tende ncy to d e no vo methyl ate DNA, a nd t ra ns fe rs me thy l m oieties at ve ry low lev el s in s u bs t ra tes not con tai ning methy la ta ble C pG di n ucleot ides . Althou gh so me cytosine methyl ation ca n occu r in ot h er dinucleotides in th e m am m al ia n ge nome co ntai n ing cytosin e in the S' positi on (:ll , !l H-GO I, an d s u ch ac t iv ity h a s occasionall y been reported to be at rel ative ly hi gh le vels (59) , ou r s t u d ies with th e recombin a n t e nzy me in d icate that this occ urs only very rarel y in oligo deoxynucleoti des contai n ing these d in u cleot ides .
T h e mech a n is m s for t h e prope ns ity of the e nzy me to me t hyl ate in region s a lre ady containing methyl moieties Ii.e. ge nom ic methyl a t ion s pre a d ing) a re not full y understood 13:;1. Th es e s t u d ies ind ica te e n ha nce d de n ovo methyl ati on of oligode oxy n uc leoti de s containing preexi sting methyl moie ties , whi ch s uggests in uitro methylation s preadi ng . A more det a iled s t u dy of methy la t ion s prea d ing in vi tro will he rep or ted e lsewhe re . C u r re n t ly, wo r k is aimed tow ard pr eserving methy lation patterns of clo ned DNA using our ex pre ss ion sys te m. O ther intended studies are the effect of expression of the mammalian MTase on control of biological processes in prokaryotic cells, further delineation of the functional domains of the maintenance methyltransferase in mutagenesis studies, and large scale production of this enzyme for crystallography. Finally, studies are in progress focusing on a more extensive analysis of the molecular mechanisms of methylation spreading using the defined in vitro oligodeoxynucleotide system reported in this initial study.