The DNA [adenine-N6]methyltransferase (Dam) of bacteriophage T4

517

Gene, 73 (1988) 517-530 Elsevier GEN 02751

The DNA [adenine-Wjmethyltransferase (Dam) of bacteriophage T4 (Open reading frame(s); phage T4 promoter(s);

recombinant

DNA; sequence specificity; Saccharomyces

cerevtkiae)

Samuel L. Schlagman,Zoe Miner, ZsigmondFeh&* and Stanley Hattman Department of Biology. Universityof Rochester, Rochester, NY 14627 (U.S.A.) Received 28 June 1988 Accepted 8 August 1988 Received by publisher 6 September 1988

SUMMARY

A functional bacteriophage T4 dam + gene, which specifies a DNA [ adenine-P] methyltransferase (Dam), was cloned on a 1.8-kb Hind111 fragment [Schlagman and Hattman, Gene 22 (1983) 139-1561. Sequence analysis [Macdonald and Mosig, EMBO J. 3 (1984) 2863-28711 revealed two overlapping in-phase open reading frames (ORFs). The 5’ proximal ORF initiates translation at an AUG and encodes a 30-kDa polypeptide, whereas the downstream ORF initiates translation at a GUG and encodes a 26-kDa polypeptide. Analysis of BAL 31 deletions in our original dam+ clone has verified that at least one of these overlapping ORFs, in fact, encodes T4 Dam. To investigate where T4 Dam translation is initiated, we have constructed plasmids in which a tuc or LIP, promoter is placed 5’ to either the longer ORF or just the shorter ORF. Only clones which contain a promoter in front of the longer ORF produce active T4 Dam. This indicates that the 26-kDa polypeptide alone cannot be T4 Dam. Additional experiments suggest that only the 30-kDa polypeptide is required for enzyme activity and that the shorter ORF is not translated in plasmid-carrying cells. We also present evidence that T4 Dam is capable of methylating 5’-GATC-3’, GATm’C, and GAThmC sequences; non-canonical sites (e.g., GACC) are also methylated, but much less efficiently.

INTRODUCTION

Bacteriophages T2 and T4, but not the related bacteriophage T6, encode a Dam (Gold et al., 1964; CorrespondenceIO: Dr. S. Hattman, Department of Biology, University of Rochester, NY 14627 (U.S.A.) Tel. (716)275-3846. * Current address: Department of Biology, University Medical School, Debrecen, H-4012 (Hungary) Tel. 36-52-16531. Abbreviations: AdoMet, S-adenosyl-methionine; AP, 2-aminopurine; Bap, bacterial alkaline phosphatase; bp, base pair(s); BSA, bovine serum albumin; Dam, DNA [adenine-@Imethyltransferase which methylates 5’-GATC-3’; dam, gene coding 0378-l119/88/$03.500 1988Elsevier

Fujimoto et al., 1965; Gefter et al., 1966; Hausmann and Gold, 1966; Hattman, 1970). Both T2 and T4 Dam methylate the sequence, GATC (Van Ormondt et al., 1975; Brooks, 1977; Hattman et al., 1978b; for Dam; dam”, mutant producing Damh which can methylate sequences other than 5’-GATC-3’; EGTA, ethyleneglycolhis@aminoethyl ether) NJVJV,iV’-tetraacetic acid; hmC, 5-hydroxymethylcytosine; IPTG, isopropyl-thio-/I-D-galactopyranoside; kb, 1000 bp; m6A, N6-methyladenine; mSC, 5-methylcytosine; MTase, DNA methyltransferase; ORF, open reading frame; PolIk, Rlenow (large) fragment of E. coli DNA polymerase I; SDS, sodium dodecyl sulfate; Tris, Tris[hydroxymethyl] aminomethane; [ 1, designates plasmid-carrier state.

Science Publishers B.V. (Biomedical Division)

518

Schlagman and Hattman, 1983; Doolittle and Sirotkin, 1988), the E. coli Dam recognition site (Lacks and Greenberg, 1977; Hattman et al., 1978a; Geier and Modrich, 1979). T4 Dam methylates mainly 5’-GATC (Schlagman et al., 1986; Doolittle and Sirotkin, 1988), whereas T2 Dam recognizes additional sequences (Brooks and Hattman, 1978; Hattman et al., 1978b; Doolittle and Sirotkin, 1988). The normal substrate for the phage MTase, however, is hmC-containing DNA because the phages have substituted hmC for C in their DNA. Furthermore, both the T4 dam + and the T2 dam + genes mutate to T4 damh and T2 damh, respectively, variants which code for enzymes capable of methylating additional sequences (e.g., 5’-AGACC) (Hattman, 1970; Brooks, 1977; Brooks and Hattman, 1978; Hattman et al., 1978b). The damh phage can mutate further to damh dam-x forms which lack enzyme activity and are devoid of DNA methylation (Revel and Hattman, 1971). Since the double mutants are viable, the Dam function appears to be non-essential (Georgopoulos, 1969; Hattman, 1970). To study the T4 Dam protein, Schlagman and Hattman (1983) cloned a functional T4 dam + gene on aHindIII-fragment of approx. 1.8 kb. Macdonald and Mosig (1984) sequenced this fragment and found two overlapping in-phase ORFs. The 5’ proximal ORF initiates translation at an AUG and codes for a 30-kDa polypeptide; the downstream ORF initiates translation at a GUG and codes for a 26-kDa polypeptide. In this communication, we present evidence that only the 30-kDa polypeptide is required for T4 Dam activity, and that the shorter ORF is not translated in plasmid-containing cells. In addition, we present evidence that in vivo the T4 Dam recognizes sequences in addition to GATC, and that it is also capable of methylating m5Ccontaining DNA in vitro.

MATERIALS AND METHODS

(a) Bacteria, bacteriophages, plasmids, and yeast Escherichia coli SK1036 dam-4 (Herman and Modrich, 1981; 1982) and E. coli GM119 dam-3 dcm-6 were from J.E. Brooks. E. coli 1674 [F’ pro + lacIq 1acZ AM15]A(lac-pro),,,,dcm-6 dam-3 and

E. coli GM124 dam-4

were from M.G. Marinus. E. coli W3 110 lacIq was from R. Brent. E. coli K704 sup11 rglA q$B was from B. de Groot. E. coli 1100 sup endI- thi hsdR rgZA rglB was from H.R. Revel and made recipient of an F factor. E. coli CSR 603 (Sancar et al., 1979) was from the Yale University Stock Culture Center. Bacteriophages T2 agt- 1 damhdam-1 (designated T2 gt- damhdam-1), T4 cqt - ,figt - ,,dum + (designated T4 gt - dam + ), T4 crgt- ,Bgt-,&amh(designatedT4gt-damh),andT6 ag- z (designated T6 gt- ) were from H.R. Revel. T4gt - dam hdam- 1 was isolated from T4gt - damh by M. Myers in this laboratory. Bacteriophage T4 amN55 x 5 (gene 42), amE5l(gene 56), nd28 (gene denA), D2a2 (gene denB), ale 10, dam + (designated T4c819 dam + ) was from J. Wiberg. Plasmids pSSH-la and pSSH-lb have been described by Schlagman and Hattman (1983). Plasmid pTP1 66, containing the E. coli dam+ gene under the tat promoter, was obtained from M.G. Marinus (Marinus et al., 1984). Plasmids pBR322Sma1, pKK223-3 (containing the tat promoter), and pUC18 were from P-L Biochemicals. Plasmid pBR322 was from J.E. Brooks. Plasmid pGW7 (containing the & and & promoters and the temperature-sensitive kI857-coded repressor) was from G. G. Wilson. Plasmid YEp51 (Broach et al., 1983), containing the yeast LEU2 gene, the 2 p origin of replication, the GAL10 promoter and the bla gene, was from L. Prakash. The haploid yeast strain Saccharomyces cerevisiae GRF18 (MATa his3-11 hisj-15 led-3 led-112 c&-101) was from A. Hinnen. (b) Chemicals and media

Restriction endonucleases, BAL 3 1 exonuclease, Bap, PolIk, T4 DNA ligase, DNA linkers, and BSA were from New England Biolabs (NEB), Bethesda Research Laboratories (BRL), or International Biotechnologies, Inc. (IBI). All enzymes were used under the conditions specified by the manufacturer, unless otherwise stated. Zymolase 20T (from Arthrobatter luteus) was from Seikagaku Kogyo Co., Tokyo, Japan. EGTA, AP, D-( + )-raffinose, amino acids, EDTA, diethylpyrocarbonate, Tris, AdoMet and pancreatic RNase A were from Sigma. SorbitolEDTA solution is 1 M sorbitol and 0.1 M EDTA (pH 8.0). Yeast minimal-rat&rose medium contains

519

0.6% yeast nitrogen base without amino acids (D&o), 2% (w/v) raflimose, and the appropriate amino acid supplants at 20 fig/ml. (c) Subcloning of the T4 dam+ gene into derivative plasmids (see Fig. 1 for a description of the various subclones) Plasmid pSSH-lb was cleaved with HgiAI and incubated for 5 min at 30°C with ‘slow’ BAL 3 1 exonuclease (IBI); BumHI (phosphorylated) linkers (NEB) were ligated to the BAL 31 blunt ends. The EcoRI-BamHI fragment was ligated into theEcoRIBamHI sites of pBR322. The resulting plasmids were transformed (Mandel and Higa, 1970) into E. coli SK1036. Transformants were checked for AP resistance/sensitivity. An AP-resistant clone was identified and designated pSSH-10. Sequence analysis (Sanger et al,, 1977) revealed that the BamHI site was 66 bp downstream from the T4 dam + gene TAA stop codon. Plasmids pSSH-15, pSSH-16, pSSH-17 and pSSH-18 were derived by further ‘slow’ BAL 31 exonuclease digestion of pSSH-10 at the BamHI site (see Fig. 1 for a description). Nucleotide sequence analysis revealed that pSSH-15 was deleted 60 bp into the T4 dam+ gene, pSSH-16 was deleted 42 bp into the T4 dam’ gene and pSSH-17 was deleted 25 bp into the T4 dam + gene. The BAL 3 1 deletion in plasmid pSSH-18 extended to 14 bp downstream from the T4 dam* gene TAA stop codon. Plasmid pSSH-10 DNA was digested with RsaI + BamHI, and the T4 dam+ RsaI-BamHI fragment was ligated between the SmaI and BamHI sites of pBR322SmaI to form plasmid pSSH-Il. The T4 RsaI-BamHI fragment was transferred into pUC18 and then back into pBR322 between the EcoRI and PstI sites to form plasmid pSH-11. The ~i~dI~I-BumHI T4 dam+ fragment from pSSH-10 was digested with ?%a1 and the Z%uIBamHI fragment was ligated between the SmaI and BamHI sites of pBR322SmaI to form plasmid pSSH-12. Plasmids pSSH-12c, pSH-12c and pSH-12 contain the same T4 insert as pSSH-12 but are derived from pUC18 ~te~~iates; inserts are cloned between the EcoRI and PstI sites of pBR322 in the case of pSH-12 and pSH-12c and between the EcoRI and Z&d111 sites in the case of plasmid

pSSH-12~. TheT4 Thai-BamHI fragment was also excised with BumHI from one of the pUC18 intermediates and ligated into the BumHI (Bap-seated) site of pGW7 to form plasmid pSLS-12L, in which the T4 fragment is oriented such that the (smaller) T4 Dam ORF is transcribed from the 1P,_ promoter. In a similar manner the T4 NdeI-BumHI fragment from pSSH-10 was cloned into pUC18 excised with BamHI and ligated into the BamHI (Bap-treated) site of pGW7 to form plasmid pSLS13L. The T4 NdeI-BamHI fragment was also excised with EcoRI + PstI from the same pUC18 intermediate and ligated into pBR322 to form pSH-13 (see Fig. 1). The T4 insert of plasmid pSH- 13 was excised with XbaI (sites in the vector) and the larger plasmid vector fragment was ligated closed to form plasmid pSH-586. Plasmid pSH-586 contains the EcoRIPstI portion of the m~ticlo~g site from pUC18 inserted between the EcoRI and PstI sites of pBR322. The T4 NdeI-BamHI fragment from pSSH-10 was cloned into pUC18, excised with SuZI and BamHI, and ligated between the SalI and BamHI sites of plasmid YEp5 1 to form pYSH-19 (see Fig. 1). (d) Yeast transformation and iuduction of the T4 Dam MTase Transformation of pYSH-19 into yeast strain GRFl8 was performed as described by Hinnen et al. (1978). To induce T4 Dam, GRFl8 cells were grown overnight to early log phase (A 600 nm = 0.5-0.7) in ~~~-r~mose medium. Galactose was then added to a fmal concentration of 10 nM, to induce transcription from the GALZO promoter. After 10 h additional growth, the yeast cells were harvested by centrifiigation. (e) Preparation of yeast DNA GRFl8 cells were suspended in sorbitol-EDTA solution. Spheroplasts were prepared by adding 100 pg/ml zymolyase-20T and incubating the cells at 37°C for 2 h. The spheroplasts were washed once with sorbitol-EDTA solution and resuspended in 50 mM EDTA (pH 8.0), 0.3% (wt/vol) SDS. After the addition of diethylprocarbonate to 0.4%

520

(vol/vol) final concentration,

the mixture was incubated at 65 ‘C for 10 min and cooled on ice. Onefti (l/5) vol of 5 M KS acetate was added and the precipitate centrifuged in an Eppendorf microfuge for 5 min. The supernatant was extracted twice with phenol-chloroform (l/l) and once with chloroform. The nucleic acids were precipitated in ethanol, dissolved in 10 mM Tris (pH 8.0), 1 mM EDTA (pH 8.0), and incubated at 37°C for 1 h with pancreatic RNase A at a final concentration of 25 pg/ml. Following digestion with restriction enzymes the samples were analyzed on a 1.0% agarose gel containing TAE buffer (Maniatis et al., 1982). (f) Purification, in vitro methyiation, and preparation of DNA for analysis by high-performance liquid chromatography E. coli K704 host cells were grown to 1 x lo* per ml in LB broth at 37°C and infected with T2gtrdumhdum-1 at an input ratio of one phage per cell. Eight ruin after infection the cells were infected a second time with T2gr- damhdam-1 at an input ratio of one phage per cell. After ~cubatio~ for 5 h at 37 oC, the phage were puri%d as described previously (Schlagman and Hattman, 19X3), except that the CsCl gradient step was eliminated. Virion DNA was purified as described previously (Hehhuann and Ha&man, 1972). T2g~-d~mhdum-l DNA and phage XP12 DNA (the generous gift of M. Ehrlich) were each methylated in vitro with T4 Dam enzyme partially purified from E. coli dam - cells harboring a T4 dam + plasmid. (The puritication procedure will be published elsewhere.) The me~ylation reaction was carried out at 30 *C for 1 h in a total volume of 400 ,ul containing 25 pg of DNA, 50 mM Tris * HCl (pH 8.0), 150 mM NaCl, 5 mM EDTA (pH 8.0), 5 mM fi-mercaptoethanol, 25 PM AdoMet and 10 ~1 of partially purified T4 Dam. Following in vitro me~ylation, the DNA was precipitated at -20” C in 70% ethanol, collected by centrifugation, washed once with 70% ethanol and air dried. The DNA was digested to deoxyribonucleosides as described previously (Scarbrough et al., 1984). Control DNA which had not been methylated in vitro was also digested to deoxyribonucleosides. Twenty ~1 of the digest were applied to a Supelcosil reverse-phase column (25 cm by 4.6 mm) type

LC-18-DB, fitted with a guard column (5 cm by 4.6 mm; Supelco Chromatography Supplies), and the deox~bonuel~sides were separated isocratitally by a modification of the method of Gama-Sosa et al. (1983) in 50 mM KH,PO,-7% methanol (pH 4.1) at 37°C.

RESULTS

(a) T4 Dam activity of various subclones To determine which ORF(s) is required for T4 Dam activity, we constructed several subclones of our original plasmid, pSSH-lb (Schlagman and Hattman, 1983) and determined by three different methods their ability to produce T4 Dam. In the fast method, we determined the AP resistance/sensitivity of E. coli dam - cells hying the plasmid subclone; dam - ceils harboring a plasmid producing T4 or E. coli Dam are AP resistant (Herman and Modrich, 1982; Arraj and Marinus, 1983; Brooks et al., 1983; Schlagman and Hattman, 1983), whereas dam - containing a plasmid defective in T4 or E. co/i Dam activity are AP sensitive. The second method is based on the in vitro resistance/sensitivity of isolated plasmid DNA to the restriction enzymes, Mb01 and DpnI. T4 Dam methylation of GATC sequences protects the DNA against Mbof, but renders it susceptible to Dpni when the A’s on both strands are methylated. In the third method, we determined the m6A content of DNA from E. coli dam - cells harboring the T4 plasmid in question. All three methods measure enzyme activity on C-DNA rather than on hmC-DNA (the normal substrate for the enzyme); in general, the results of these methods correlate well with each other. It should be noted that the first two methods assess GATC methylation only; the third method measures A-methylation independent of the target sequence. (None of the methods measures quantitatively the MTase activity in the cell.) We have found, however, that the second procedure is the most sensitive one for detecting low levels of MTase activity, particularly if the plasmid DNA is isolated from saturated overnight cultures (S.L.S., Z.M. and S.H., ~pub~sh~ observation). Unless specified otherwise, plasmid DNA isolated from overnight cultures was analyzed by the second method.

521

?-

PLASMIDor DERIVATIVE SERlE8

&mORF

?_

P_17.40

pSSH-1b

Dam’

BAL 31 deletion at HgiAl; add BamHl + linker; reclone into pBR322 Hindlll 1

I

8amHl I I TAA m

Rsal Ndel Thai m II I I I AUG UJG

pSSH-10 Dam’

H

Hindlll w

Hindlll ”

Rsal Ndel Thal i-1 ,I 1. AUG

I

I

subclone ii

Rsal Ndel I ,I

mf

BamHl I*

I I

AUG

Hindlll I

BamHl

I I Gt_lG

Rsal Ndel Thal l-1 11

GUG

i AU0

(Ndel) Thai

d

pSSH-17

Dam-

pSSH-16

Darn-

pSSH-15

Dam-

pSSH-11 pSH-11

Dam

.

BarnHI I J I TAA m I subclone 13

c

BarnHI

AUG w subclone 12~ (Thal) ;’

pSSH-13 pSH-13 pSLS-13L pYSH-13

Dam+

BMIHI AAm

subclonel2

Dam+

BamHl I I Gixj

(Rsal) Ndel Thal II I ,‘-I AUG GUG

71

pSSH-18

1_42

Thai II

BAL 31 deletions of pSSH-10 at the BamHl site; BamHl linker added at the site of the BAL 31 deletion and at the Pvull site of vector pBR322

pSSH-12 pSH-12 pS.SH-12c pSH-12c pSLS-121

Dam

Fig. 1. Derivation aud structure of subciones of pIasmid pSSH-lb. Positions of the T4 promoters, PJ7.75 and PJ7.40, indicated by the fihed-in squares were determined from in vitro experiments by Macdonafd and Mosig (198r9).The position of restriction enzyme sites and location of the ORPs are also taken from their sequencing data. The location of the Shine-Dalgarno (SD.) sequence in the ORFs is indicated only on clone pSSH-lb. Parentheses around a restriction sequence denote loss of that site as a result of ‘blunt-ending’ and/or cloning. The numbers in the boxes at the 3’-end of the subclones represent the distance from the TAA stop codon at which the BAL 31 deletion ends. A positive number represents a deletion ending downstream of the TAA stop codon, whereas a negative number represents a deletion extending into the 3’ end of the gene.

522 TABLE I

Fig. 1 shows the de~vation of the various subclones. As indicated, starting at the BumHI site of pSSH-10, which is 66 bp downstream of the TAA stop codon, several BAL 31 deletions were made at the 3’ end of the two ORFs, generating plasmids pSSH-15, pSSH-16, pSSH-17, and pSSH-18 (see

Mol % m6A in bacterial DNA of E. coli GM1 19 dam-3 dcm-6 harboring various plasmids Plasmida

[pBR322] [pTP166] [pSSH-la] [pSSH-101 [pSSH-1 I] [pSSH-181

MTase gene present on the plasmid

Mol % m6Ab

none

0.14

E. coli dam + T4 dam + T4 dam + T4 dam + T4 dam +

1.9 2.0 2.0 2.0 2.6

MATERIALS

* The plasmids are pBR322 derivatives. For a description see MATERIALS AND METHODS, sections a and c, and Fii. 1.

b Mel % m6A refers to the percentage of adenine bases (adenine t m6A) which are m6A. The values listed are the means obtained from analyses of, at least, two independent labeling experiments; in most cases, each preparation was subjected to duplicate chromatographic analysis. In most cases, the values for each labeling experiment deviated by less than + 10% of the mean. A saturated overnight culture of E. coli GM1 19 plasmidcontaining cells, grown at 37°C in LB broth plus ampicillin (50 &ml), was diluted 1 to 100 in 5 ml of fresh medium and grown at 37*C to a titer of 1 to 2 x lo8 cells per ml. Fifty pCi of [2-3H]adenine (~ersh~; 161 mCi/mg) was added and the cells were incubated for an additional 2 h at 37 ’ C. The cells were collected, the DNA purified and DNA bases analysed as previously described (Hattman et al., 1973), except that the RNA was degraded by treatment with 0.2 N NaOH for 16 to 18 h at 37°C.

abcdefghi

AND

METHODS,

SeCtiOn

c).

E. cob

strains harboring plasmid pSSH-15, pSSH-16, or pSSH17 were AP sensitive, and the plasmid DNA isolated from these cells were MboI sensitive and &rI resistant (data not shown). E. coli dam- strains harboring plasmid pSSH-18 were AP resistant (data not shown) and chromosomal DNA isolated from E. co/i GM119 dam-[pSSH-181 contained 2.6 mol% m6A (Table I). Nucleotide sequence analysis revealed that the BAL 3 1 deletion extends upstream of the TAA stop codon in plasmids pSSH-15, pSSH-16, and pSSH-17 but is downstream of the TAA stop codon in plasmid pSSH-18 (see Fig. 1 for the extent of the BAL 31 deletions). Thus, deletions extending into the 3’ end of the two ORFs resulted in loss of T4 Dam methylation, indicating that one or both of these ORFs is required for enzyme activity. dam-

j

k

lmnopqr

st

kb 23.13 9.42 6.66 4.36

-

2.32 2.03 1.351.060.87 0.60 -

Fig. 2. Restriction enzyme susceptibility ofplasmid pSSH-1 l,pSSH-13,pSSH-12 andpSSH-12c DNAs.E. coli 1674 plasmid containing cells were grown at 37 “C overnight in LB broth containing spit (50 &ml). The plasmid DNA was puritied by alkali lysis (Maniatis et al., 1982), with the addition of one phenol extraction prior to ethanol precipitation. Following restriction endonuclease digestion, the samples were treated with pancreatic RNase A (10 &ml) for 30 mm prior to loading in a 1.2% agarose gel containing TAE buffer (Maniatis et al., 1982). Lanes: a, and t, ,U?indIII marker DNA; b and s, $X174 RFI/&reIII marker DNA; c, pSSH-11; d, pSSH-11 cutwithEcoRItofo~alinearmolecule;e,pSSH-l1/MboI;f,pSSH-ll/DpnI;g,pSSH-13;h,pSSH-13cutwithEcoRI;i,pSSH-13/~~oI; j, pSSH-13/DpnI; k, pSSH-12; 1; pSSH-12 cut with EcoRI; m, pSSH-lZ/MboI; n, pSSH-lZ/DpnI: o, pSSH-12~; p, pSSH-12c cut with EcoRI; q, pSSH-lZc/MboI; r, pSSH-lZc/DpnI.

523

As seen in Figs. 1 and 2, clone pSSH-11, containing both ORFs and only the proximal T4 promoter (P,17.40), retained T4 Dam activity. When the P,17.40 promoter was removed to form clone pSSH-13, a low level of T4 Dam activity was observed (Pig. 2, lanes g-j), We believe that this activity is due either to transcription from a weak promoter within the vector or one formed at the vector-insert junction. Subclones pSSH-12 and pSSH-12c, containing only the shorter ORF and no upstream promoter, had no T4 Dam activity (Fig. 2). These data suggest that T4 promoter PJ7.40 is used in vivo and that one or both of the ORFs identified by Macdon~d and Mosig (1984) is required for T4 Dam enzyme activity. Some of the subclones that are described above may produce different levels of T4 Dam than our original plasmids pSSH-la and pSSH-lb (Schlagman and Hattman, 1983). For instance, plasmid pSSH-11 contains the proximal promoter P,17.40, but it is missing the distal PJ7.75 promoter (see Fig. 1). Thus, a priori, it might produce less enzyme than the original pSSH-la (or pSSH-lb) clone or subclone pSSH-10, each of which contains both promoters. On the other hand, plasmid pSSH-18 contains both promoters, and it also has a higher copy number than plasmids pSSH-la, pSSH-lb, pSSH-10 and pSSH-11 (unpublished observation); thus, it might produce more T4 Dam enzyme. To ascertain whether plasmid copy n~ber or the loss of the distal promoter affects the level of T4 Dam, we analyzed the m6A content of host E. coli DNA. As seen in Table I, E. coli dam - cells containing plasmid pSSH-la, pSSH-10 or pSSH-11 have similar levels of m6A (2.0 mol%); thus, loss of the distal PJ7.75 promoter in plasmid pSSH-11 has no aBect on m6A level of cellular DNA. Furthermore, the m6A level is essentially the same as that observed in E. cuk dam -cells containing plasmid pTP166, which has the E. cali dam + gene under the control of the tat promoter (1.9 mol% ). In contrast, E. colidam - cells containing plasmid pSSH-18 have a significantly higher DNA m6A level (2.6 mol%). This is consistent with the notion that pSSH-18 produces more T4 Dam because of its higher copy number.

TABLE II Mol % m6A iu bacterial DNA from E. coli GM 1674&rn-3 dcm-6 containing various T4 a&n subclones Plasmid*

[pSLS-13L] [pSLS-12L] [PGW-71

MoI % m6A following growth at 30”Cb

following heat-induction at 42’CQ

0.15 0.09 0.08

1.80 0.09 0.09

a See Table I, footnote a. b The procedure was as described in Table I, footnote b, except that bacteriaI growth was at 30°C. c The procedure was as described in Table I, footnote b, except for the following modifications: the cells were grown at 30°C until 1 h after addition of the [2-‘Hladenine and then shifted to 42°C for l h prior to DNA purification.

(b) T4 Dam activity requires the larger ORF To determine which ORP(s) is required for T4 Dam activity, we constructed subclones in which a tat or a APL promoter was placed 5’ to either the longer ORF or just the shorter ORF. Only clones which contained a promoter in front of the longer ORF produced active T4 Dam enzyme (Table II), For example, cells containing plasmid pSLS-13L had very little enzyme activity at 30 ’ C, a temperature at which the &I857 repressor (encoded by the plasmid) is functional and the ,%F, promoter is repressed; however, enzyme activity was observed at 42”C, a temperature at which the IzcI857 repressor is not factional and the U, promoter is active (Table II). In contrast, cells cont~g plasmid pSLS-12L, which has the PL promoter 5’ to just the shorter ORF, did not have T4 Dam activity at 30” C or 42” C (Table II). Similar observations were made with plasmids containing the tat promoter and the same T4 fragments (data not shown). These results indicate that the 26-kDa protein alone has no T4 Dam activity. Therefore, MTase activity requires either a combination of the 26-kDa and 30-kDa polypeptides or the 30-kDa polypeptide alone.

524

Polypeptides produced in ‘maxicells by various T4 dam subclones

(c)

In order to determine whether both ORFs are in fact being translated, we analyzed the polypeptides produced in labeled ‘maxicells’ (Sancar et al., 1979) containing various T4 dam plasmids. Subclones pSH-11 and pSH-13 contain both ORFs, whereas subclones pSH-12 and pSH-12c contain only the shorter ORF (see MATERIALS AND METHODS, section c and Fig. 1). The T4 dam’+ ORF(s) in all four subclones is (are) transcribed from a promoter within the tet gene. In pSH-11, the T4 promoter P,17.40 also transcribes both ORFs. These four subclones, as well as the control vector, pSH-586 (see MATERIALS AND METHODS, section c), are deleted for most of the blu gene, which would produce polypeptides of 3 1-kDa and 28-kDa in maxicells (Sancar et al., 1979). As seen in Fig. 3, only the 30-kDa protein is observed in maxicells containing a plasmid with both ORFs (plasmid pSH-11 or plasmid pSH-13). Furthermore, the 26-kDa protein cannot be detected in maxicells containing a plasmid with only the shorter ORF (plasmid pSH-12 or plasmid pSH-12~). This indicates that the shorter ORF within the T4 dam + gene is not translated appreciably, and that T4 Dam consists of only the 30-kDa polypeptide.

kDa

abcdefg

92.5 66.2 -

45.0 -

37.0 +, 31.0 -p 31.0 r" 28.0

21.5 -

14.0 -

(d) T4 Dam activity in yeast cells To provide further evidence that only the 30-kDa polypeptide is necessary for T4 Dam activity, plasmid pYSH-19 was constructed in a yeast expression VeCtOr (See

MATERIALS

AND METHODS,

SeCtiOn c

and Fig. 1). This clone has the inducible yeast GALlO promoter upstream of the dam + gene. Since yeast uses AUG as a start codon, but not GUG (Bairn et al., 1985), it shonld produce only the 30-kDa polypeptide; as seen in Fig. 4, yeast cells containing pYSH-19 have T4 Dam activity. This result is similar to that previously observed for a T4 dam+ clone constructed in a different yeast shuttle vector (Z.M. and S.H., unpublished observation). Furthermore, the level of T4 Dam MTase activity increased following induction of the yeast GALlO promoter, as indicated by the increased sensitivity of the yeast DNA to the restriction enzyme, @XX The low level of T4 Dam activity observed in uninduced

Fig. 3. Analysis of the polypeptides encoded and translated by various plasmids harbored in E. coli CSR 603. To obtain labeled plasmid polypeptides in ‘maxicells’, the procedure of Sancar et al. (1979) was used with the following modifications. AAer cells were exposed to a UV dose of 2.5 to 3.0 joules, they were incubated for 3 b at 37°C. D-cycloserine was added to a fmal concentration of 0.1 m&ml and incubation was continued for an additional 14 to 18 h at 37°C. The cells were washed twice with and resuspended in M-9 minimal medium lacking sulfate and incubated for 1 h at 37°C; 50 FCi of [3sS]met~o~ne (New England Nuclear; 1151 Ci/mmol) were added and the cells incubated for 1 h at 37OC. The remainder of the procedure was as described by Sancar et al. (1979). Samples were electrophoresed on a 0.1% SDS t 12% polyacrylamide gel (Laemmli, 1970) along with the following unlabeled molecular ratio markers: phosphorylase B (92500), bovine serum albumin (66200), ovalbumin (45000), carbonic anhydrase (31000), soybean trypsin inhibitor (21500) and lysozyme (14000). The position and molecular ratios of the unlabeled markers, as well as the labeled bla gene polypeptides (3 1000 and 28 000) and the labeled tetgene polypeptide (37000), are indicated. a, pBR322; b, pSH-586; c, pSH-11; d, pSH-13; e, pSH-12; f, pSH-12~; g, pBR322.

525

i

abcdefahi

k

I

kb 23.13 9.42 6.56 4.36 -

2.32 2.03

r

yeast cells containing plasmid p?SH-19 may be due to transcription from a promoter other than GALZO. Recent experiments suggest that a TATA site 5’ to the T4 dam+ gene is recognized by yeast (Z.F., S.L.S., Z.M. and S.H., submitted for publication). These data are consistent with our conclusion that only the 30-kDa polypeptide is required for T4 Dam activity. (e) Activity of T4 dam+ subclones on hmC-containing DNA

Fig. 4. Restriction enzyme susceptibility of yeast DNA purified from uninduced and galactose-induced yeast cells containing plasmid pYSH-19 or YEpSl. Lanes: a, uninduced [pYSH-191; b, uninduced [pYSH-19]/MboI; c, uninduced [pYSH-19]/DpnI; d, uninduced [pYSH-19]/Sou3AI; e, galactose-induced [pYSH191;f, galactose-induced [pYSH-19]/MboI; g, galactose-induced [pYSH-19]/DpnI; h, galactose-induced [pYSH-19]/Suu3AI; i, uninduced [YEpS l] ; j, uninduced vEp5 l]/MboI; k, uninduced ~EpSl]/DpnI; 1, uninduced [YEpSl]/Suu3AI. The position of the lWindII1 DNA bands in the gel are indicated (in kb) on the left margin.

We have shown that some of our subclones produce an MTase which is active on C-DNA. In a previous study (Schlagman and Hattman, 1983) we propagated phage T2gf-damhdam-1 in E. coli harboring clone pSSH-la and detected methylation of the phage hmC-DNA (0.4% of the adenines were methylated). Phage T2gr-dumhdam-1 (as well as phage T4gt-damhdum-1) produces a defective Dam polypeptide; therefore, MTase activity in plasmidcontaining cells might be affected by positive or negative intragenic complementation between plasmid- and phage-encoded Dam polypeptides. However, since phage T6gt- does not appear to have a dam gene (Miner and Hattman, 1988), any methylation of T6 hmC-DNA in plasmid-containing cells must result exclusively from methylation by the plasmid encoded T4 Dam. Therefore, to determine whether our subclones produce a T4 Dam that is

TABLE III Mol % m6A in phage DNA after propagation in various plasmid-containing E. coli K704 dam + hosts Plasmid”

[pBR322] [pTPl66] [pSSH-la] [pSSH-IO] [pSSH-111 [pSSH-181

MTase gene present on the plasmid

none E. wli dam + T4 dam + T4 dam + T4 dam + T4 dam +

Mol % m6A in virionb T2g1damhdam-1

T6gt_

T4gtdamhdam-1

T4gt dam +

T4gi damh

0.04 0.28 0.44 0.45 0.42 N.D.

0.03 0.22 0.44 0.45 0.42 0.51

0.03 N.D. 0.36 0.42 0.42 N.D.

0.52 N.D. N.D. 0.56 N.D. 0.63

1.25 N.D. N.D. N.D. N.D. 1.29

a See Table I, footnote a. b A culture of E. coli K704 plasmid containing cells grown at 37°C in LB broth containing ampicillin (50 pg/ml) was diluted into 5 ml of fresh medium and grown at 37°C to 3 x lo8 cells per ml. One hundred (100) pCi of [2-3H] adenine (Amersham; 161 mCi/mg) and bacteriophage at an input ratio of 5 to 10 per cell were added. Incubation was continued for an additional 5 h at 37°C. The bacteriophage and virion DNA were purified and analyzed as described previously (Hattman, 1970). See also Table I, footnote b. N.D. = not done.

526

active on hmC-DNA, and to determine whether intragenic complementation occurs, we measured the mol% m6A in packaged DNA of phages T2gt-&mhdum-1, T4gt-damhdum-1 and T6gtpropagated in plasmid-containing cells. As seen in Table III, the DNAs of phages T2gt-dumhdam-1, T4gt- dam hdum - 1 and T6gt - are virtually devoid of m6A when propagated in E. coli dam + cells which lack a T4 dam + plasmid or a plasmid which overexpresses E. coli Dam. When grown in a host containing plasmid pTP166 which overexpresses E. coli Dam, the phage DNAs contain only about 50 % of the m6A content found in the DNA of phages propagated in a host containing a T4 dam + plasmid (Table III). Since cells containing plasmid pTP166 produce at least as much, if not more, Dam activity than cells containing the T4 dam’ plasmid pSSH-lb, when assayed on C-containing DNA (Schlagman et al., 1986), we conclude that hmCcontaining DNA is a poor substrate for E. coli Dam. Furthermore, cells containing plasmid pSSH-la (original clone), plasmid pSSH-10 (subclone), or plasmid pS SH- 11 (subclone) produce DNA with the same m6A content in all three phages (Table III). This is similar to what we observed for bacterial DNA (Table I); i.e., the absence of the distal promoter (P,17.75) in plasmid pSSH-11, or its presence in plasmids pSSH-la, pSSH-lb and pSSH-10, has no apparent effect on the m6A levels of the virion DNA. In addition, cells containing plasmid pSSH-18, which was shown to produce DNA with a higher m6A content in bacterial DNA than the other plasmids (Table I), also produces DNA with a slightly higher m6A content in T6gt- virion DNA. We conclude that there is no evidence for intragenic complementation, that subclones pSSH-10, pSSH-11, and pSSH-18 (see Fig. 1) produce a T4 Dam enzyme capable of methylating hmC-DNA, and that these plasmids show the same relative methylation ability with hmC-containing T4 DNA as they do with C-containing bacterial DNA. (f) Production of T4 Damb by infecting pbage is not influenced by the presence of T4 Dam in the infected cells

We decided to investigate whether the production of T4 Damh by infecting phage is influenced by the presence of T4 Dam in infected cells harboring a T4

dam+ plasmid. To accomplish this, T4gtrdamh phage infections were carried out in E. coli harboring or lacking a T4 dam+ plasmid. The Damh enzyme produced by T4@-damh methylates sites in virion DNA which are not methylated by the T4 dam + encoded enzyme; this results in a higher m6A content (1.25mol%) in T4gr-dumb virion DNA as compared to that in T4gr-dam + virion DNA (0.51 mol%) (Hattman, 1970; Table III). As seen in Table III, the m6A content of the T4grdumb virion DNA is the same following infection of cells whether or not they carry a T4 dam + plasmid. This indicates that the presence of plasmid-encoded T4 Dam enzyme did not a&t production or function of Damh encoded by the infecting phage. (g) GATbmC sites are fully methylated in T4 DNA

To determine whether T4gt- dam + DNA is fully methylated in vivo at GAThmC sites, we measured the m6A content in virion DNA following infection of hosts harboring a T4 dam+ plasmid vs a host lacking such a plasmid. We have shown above that the presence of plasmid-encoded T4 Dam does not afiect production of Dam by infecting T4 phage. Thus, cells harboring a T4 dam’ plasmid should contain more T4 Dam activity following T4gt-dam’ phage infection than cells which lack such a plasmid. As shown in Table III, the m6A content of T4gt- dam + virion DNA after growth in a host lacking a T4 dam + plasmid is similar to that of T4gt- dam hdam- 1 after growth in a host harboring

TABLE IV Mol % m6A of T4c8 19 dam + virion DNA following propagation in various E. coli GM124 dam-4 strains containing listed plasmids Plasmid ’

MTase gene carried on the plasmidb

Mol % m6AE

none [pSSH-IO] [pSSH-181

none T4 dam + T4 dam +

0.40 0.49

0.51

a See Table I, footnote a. b T4c819dam+ phage produce C-containing DNA in E. coli GM124dam-4 cells. c The procedure was as described in Table III, footnote b, except that bacteriophages were added at an input ratio 1 per cell at t = 0 and at t = 8 min; see also Table I, footnote b.

521

a T4 dam + plasmid. Furthermore, there is only a slight, but reproducible, increase in the m6A content of T4gt - dam + virion DNA following propagation in a host harboring a T4 dam’ plasmid. This result suggests that infecting T4gt - dam + phages produce enough T4 Dam to fully methylate most of the sites on T4 DNA before packaging occurs. We cannot, however, exclude the possibility that certain noncanonical sites may be methylated at a low frequency. The m6A content of T4c dam+ C-containing DNA following propagation in various E. coli dam hosts is shown in Table IV. As with T4gt- dam + @C-DNA) phage, propagation in hosts containing a T4 dam + plasmid resulted in a slight (20x), but reproducible, increase in the m6A content in virion DNA. However, in all cases, the m6A content of T4c dam’ phage DNA was slightly lower than in T4gt - dam + containing DNA. This may reflect a lower affinity of T4 Dam for C- vs hmC-containing DNA; a lower level of T4 Dam produced by T4c dam + phage (or a combination of both factors) is not ruled out at this time. (h) T4 Dam methylates 5-methylcytosine DNA

containing

Since T4 Dam can methylate both hmC- and C-containing DNAs, we investigated whether this enzyme is also capable of methylating %methylTABLE

V

Mol % of m6A in T4gt-damhdam-1 and XP-12 virion DNAs prior to and following in vitro methylation by T4 Dam

Phage DNA”

XP-12 T4gt- damhdam-l

Mol % m6A prior to in vitro methylation b*c

0.28 ~0.05

Mol % m6A following in vitro methylation by T4 Dambd 3.69 0.46

a See MATERIALS AND METHODS, sections a and f, as well

as Kuo et al. (1968) and Ehrlich et al. (1975). b See MATERIALS AND METHODS, section f, for details of puriilcation, in vitro methylation conditions, and preparation of the DNA for analysis by high performance liquid chromatography (HPLC). See also Table I, footnote b. c The values listed are the averages of at least two independent DNA digests and HPLC analysis. d The values listed are from a single DNA digest and single HPLC run.

cytosine (m5C>containing DNA. For this purpose we used Xanthomonas phage XP-12 DNA, in which m5C completely replaces C (Kuo et al., 1968; Ehrlich et al., 1975), as a substrate for in vitro methylation (See MATERIALS AND METHODS, SeCtiOII f). As seen in Table V, a ten-fold increase in the m6A content of XP-12 DNA was observed following in vitro methylation by T4 Dam. A similar result was obtained for T4gt - dumhdum- 1 hmC-containing DNA following in vitro methylation by T4 Dam (Table V). Furthermore, although XP-12 DNA is MboI-sensitive and DpnI-resistant, it becomes MboI-resistant and DpnIsensitive following in vitro methylation by T4 Dam (data not shown). We conclude that m5C-containing DNA can also serve as a substrate for T4 Dam.

DISCUSSION

Schlagman and Hattman (1983) cloned a functional T4 dam + gene on a 1.8-kb HindIIIfragment. Nucleotide sequence analysis revealed the presence of two in-frame overlapping ORFs (Macdonald and Mosig, 1984). We show here that the 30-kDa polypeptide, encoded by the longer ORF, corresponds to the T4 Dam MTase. This enzyme is capable of methylating GATC sequences in C-DNA, as well as DNAs in which C is completely replaced with hmC or m5C. Cells harboring plasmid pSSH-lb contain ten times more Dam activity than E. coli dam+ cells lacking this plasmid; nevertheless, the m6A content in the DNA of cells harboring pS SH- lb is not sign& cantly higher than that of cells harboring plasmid pTP 166, which contains the E. coli dam + gene under the tuc promoter. Schlagman et al. (1986) concluded that T4 Dam methylates mainly the E. cob Dam recognition sequence, GATC (Lacks and Greenberg, 1977; Hattman et al., 1978a; Geier and Modrich, 1979). With the exception of plasmid pS SH- 18, we have obtained similar m6A levels in the DNA of cells containing T4 dam + plasmids or plasmid pTP166 (Table I). Plasmid pSSH-18 has a higher copy number than the other T4 dam + plasmids and, thus, might produce more Dam enzyme; cells containing this plasmid, in fact, have a higher m6A content in their DNA (Table I). From this we conclude that pSSH- 18 produces more enzyme than

528

the other T4 dam+ clones, and that T4 Dam is capable of methylating sites in addition to GATC. This is consistent with the results of Doolittle and Sirotkin (1988) who analyzed the sequences in pBR322 DNA methylated in vitro by T2 Dam and T4 Dam. They found that both enzymes methylated the sequence, GATC, very efficiently, but both enzymes were also capable, at a much lower efficiency, of methylating certain GACC containing sequences; in this regard, T2 Dam was more efficient than T4 Dam, and the ability to methylate GACC sequences depended on the flanking 5’- and 3’-nucleotides. The ability of the T2 Dam to methylate secondary sites has been reported previously (Hattman et al., 1978b; Brooks and Hattman, 1978). To investigate whether T4 Dam also methylates non-canonical sites in T4 hmC-containing DNA, we analyzed the available T4 sequences from GenBank and determined the frequency of occurrence of GAThmC. This sequence occurred 65 times in the 50 963 bp, representing approximately 30% of the entire phage genome, with an A + T content of 64.3%. We calculate [(65/(0.322 x 50963) x loo)] that complete methylation of GAThmC would result in 0.40 mol% m6A in T4 virion DNA. The observed mol% m6A for T4gt- dam + virion DNA is between 0.5 and 0.6, depending on the host in which the phage was propagated (Table III). If the sequences in the data base are representative of the entire T4 genome, then T4 Dam methylates sequences in addition to GAThmC. As seen in Table III, when T2gt- dumhdum- 1 or T6gt- phage is propagated in a host harboring a plasmid which overexpresses E. coli Dam, the virion DNA contains only 40-60x of the m6A content observed for phage propagated in a host harboring a plasmid producing T4 Dam. These results are similar to those previously reported by Schlagman and Hattman (1983). From the T4 sequence analysis presented above, it would appear that, to some extent, this difference is due to the abitity of T4 Dam, but not E. coli Dam, to recognize sequences in addition to GAThmC. However, E. coli Dam methylation produces only 0.22 and 0.28 mol% m6A in T6gf - and T2gr- dumhdum-1 virion DNAs, respectively; this is significantly lower than the calculated value of 0.4 mol% m6A. Thus, unless the occurrence of the sequence GAThmC differs drastically in the

genomes of phages T2, T4, and T6, it would appear that E. coli Dam does not methylate ah the available GAThmC sequences in the phage. In this regard it should be noted that in plasmid-containing cells, T4 Dam methylates all three virion DNAs to the same extent (Table III) indicating that the frequency of GAThmC sites (and non-canonical sequences) are probably similar for the three phage genomes. Furthermore, since the relative amount of MTase activity produced in cells containing pTP 166 is equal to or greater than the amount produced in cells containing pSSH-lb, when assayed on C-containing DNA (Schlagman et al., 1986), we conclude that hmC -containing DNA is a poor substrate for E. coli Dam. It is interesting to note that phage propagated in cells harboring clone pSSH-11 [with both ORFs and only the proximal T4 promoter (PJ7.40) (see Fig. l)] have the same m6A content as phage propagated in cells harboring plasmid pSSH-10 (or pSSH-la) (Table III), which contains both the proximal T4 promoter (PJ7.40) and the distal T4 promoter (P,17.75). (Plasmid pSSH-10 was derived from pSSH-lb and pSSH-11 was derived from pSSH-10; see Fig. 1.) Similar results were also observed for methylation of cellular DNA (Table I). As stated above, Schlagman et al. (1986) found that cells with plasmid pSSH-lb contain ten times more Dam activity than dam + cells lacking this plasmid; nevertheless, mainly GATC sequences are methylated. It is, therefore, conceivable that absence of the distal promoter in plasmid pSSH-11 could result in a significant reduction in T4 Dam enzyme concentration but still not affect the m6A content of cellular or virion DNAs. There is a slight, but reproducible, increase in m6A content of T4gt - dam + virion DNA following propagation in a host harboring a T4 dam’ plasmid, as compared to a host lacking such a plasmid (Table III). We interpret this to mean that T4gt- dam + phage produces enough Dam enzyme to fully methylate the GAThmC sites in DNA before packaging into the virion occurs. Because non-canonical sequences are methylated much less efficiently in vitro (Doolittle and Sirotkin, 1988), it is likely that they are not all methylated in T4 DNA in vivo. In this regard, the increment in m6A content observed for T4gt- dam + virion DNA following propagation in a host harboring a T4 dam + plasmid probably represents methylation of some of these sequences.

529

We have previously shown that T4 Dam and E. coli Dam share regions of amino acid sequence homology (Hat&man et al., 1985). It is interesting to note that the smaller 26kDa untranslated ORF in T4 Dam is missing one of these regions of amino acid sequence homology; in fact, these regions appear to be common to a variety of DNA-adenine MTases (Mamxuelli et al., 1985; Lauster et al, 1987; Loenen et al., 19871, some of which don’t even methylate GATC. The genetic and physical study of these enzymes is a promising direction for gaining insight into protein-DNA interactions.

We thank Dr. Valakunja Nagaraja for his help in the SDS-PAGE analysis of labeled ‘maxicells’, and M. Myers for isolating the T4g~-~u~h~~-l mutant. This work was supported by Public Health grant GM-29227 to S.H.

REFERENCES Arraj, J.A., and Ma&us, M.G.: Phenotypic reversal in dam mutants of L&herichiu co&.Gene 25 (1983) 165-178. Bairn, S.B., Goodhoe, C.T., Pietras, D.F., Eustice, DC., Labhard, M., Friedman, L.R., Hampsey, D.M., Stiles, J.I. and Sherman, F.: Rules of translation: studies with altered forms of yeast CYCl gene. In Calendar, R. and Gold, L. (Eds.), Sequence Specificity in Transcription and Translation. UCLA Symposia on Molecular and Cellular Biology, Alan R. Liss, Inc., New York, Vol. 30 (1985) pp. 351-362. Broach, J.R., Li, Y.-Y., Wu, L.-C. C. and Jayaram, M.: Vectors for high-level, inducible expression of cloned genes in yeast, In Inouye, M. (Ed.), Experimental Manipulation of Gene Expression. Academic Press, New York, 1983, pp. 83-l 17. Brooks, J.E.: A Comparative Study of the Wild Type and Mutant Forms of Phage T2 DNA-Adenine Methylase. Ph.D. Thesis, University of Rochester, NY., 1977. Brooks, J.E., Bhnnenthal, R.M. and Gmgeras, T.R.: The isolation and characterization of the Escherichiu cob DNA adenine methylase (dam) gene. Nucleic Acids Res. 11 (1983) 837-851. Brooks, J.E. and Hattman. S.: In vitro methylation of bacteriophap 1 DNA by wild-type (dhm * ) and mutant (damh) forms of phage TZ DNA adenine methylase. J. Mol. Biol. 126 (1978) 38 l-394. Doolittle, M.M. and Sirotkin, K.: Bacteriophage T2 and T4 dam + and da& and Ew dam + methylation: preference at different sites. Biochim. Biophys. Acta 949 (1988) 240-246.

Ehrlich, M., Ehrlich, K. and Mayo, LA.: Unusual properties of the DNA fromXunrhomono,s phage XP-12 in which %methylcytosme complexly replaces cytosine. Biochim. Biophys. Acta 395 (1975) 109-I 19. Fujimoto, D., Stinivasan, P.R. and Borek, E.: On the nature of the deoxyribonucleic acid metbylases. Biological evidence for the m~tiple nature of the enzyme. Bi~he~s~ 4 (1965) 2849-2855. Gama-Sosa, M.A., Midgett, R.M., Slagel, V.A., G&hens S., Kuo, K.C., Gehrke, C.W. and Ehrlich, M.: Tissue-specific differences in DNA methylation in various mammals. Biochim. Biophys. Acta 740 (1983) 212-219. Getter, M., Hausmamr, R., Gold, M. and Hurwitx, J.: The enzymatic methylation of ribonucleic acid and deoxyribonucleic acid, X. Bacteriophage T3-induced S-adenosyl-met~o~ne cleavage. J. Biol. Chem. 241 (1966) 1995-2006. Geier, G.E. and Modrich, P.: Recognition sequence of the dam methylase of Esche~&ia coli K-12 and mode of cleavage of @II endonuclease. I. Biol. Chem. 254 (19’79) 1408-1413. Georgopoulos, C.P.: Genetic Studies on the Glucosylation of T-even Phage DNA. Ph.D. Thesis, Massachusetts Institute of Technology, Cambridge, MA, 1969. Gold, M., Hausmann, R., Maitra, U. and Hurwitx, J.: The enzymatic methylation of RNA and DNA, VII. Effects of bacteriophage infection on the activity of the methylating enxymes. Proc. Natl. Acad. Sci. USA 52 (1964) 292-291. Hattm~, S.: DNA methylation of T-even b~te~oph~es and of Vernon-~ucosylat~ mutants: its rote in Pl-directed restriction. Virology 42 (1970) 359-370. Hattman, S., Brooks, J.E. and Masurekar, M.: Sequence specificity of the P l-m~cation methylase (M . Eco P 1) and the DNA methylase (M +Ew dum) controlled by the E. coli dam gene. J. Mol. Biol. 126 (1978a) 367-380. Hattman, S., Sehlagman, S. and Cousens, L.: Isolation of a mutant of Esters wli defective in cytosine-specific deoxyribonucleic acid methylase activity and in partial protection of bacteriophage 1 against restriction by cells containing the N-3 drug-resistance factor. J. Bacterial. 115 (1973) 1103-l 107. Hattman, S., Van Ormcmdt, H. and dewaard, A.: Sequence specificity ofthe wild-type (dum+) and mutant (dumb) forms of bacteriophage T2 DNA adenine methylase. J. Mol. Biol. 119 (1978b) 361-376. Hattman, S., Wi~inson, J., Swinton, D., Schla~~, S., Macdonald, P.M. and Mosig, G.: Common evolutionary origin of the phage T4 dam and host Escherichiacoli dam DNA-adenine methyltransferase genes. J. Bacterial. 164 (1985) 932-937. Hausmann, R. and Gold, M.: The enzymatic methylation of ribonucleic and deoxyribonucleic acid, IV. Deoxyribonucleic acid methylase in bacteriophage-infected Escherichiacoli.J. Biol. Chem. 241 (1966) 1985-1994. Hehlmann, R. and Hattman, S.: Mutants of bacteriophage T2gf with altered DNA methylase activity. J. Mol. Biol. 67 (1972) 351-360. Herman, G. E. and Modrich, P.: E.&wick co&K-12 clones that overproduce dam methylase are hypermutable. J. Bacterial. 145 (1981) 644-646.

530

Herman, G.E. and Modrich, P.: Esehrrichia coli dum methylase. Physical and catalytic properties of the homogeneous enzyme. J. Biol. Chem. 257 (1982) 2605-2612. Hinnen, A., Hicks, J.B. and Fink, G.R.: Transformation of yeast. Proc. Natl. Acad. Sci. USA 75 (1978) 1929-1933. Kuo, T.-T., Huang, T.-C. and Teng, M.-H.: 5-Methylcytosine replacing cytosine in the deoxyribonucleic acid of a bacteriophage for Xunthomonas oryzue. J. Mol. Biol. 34 (1968) 373-375. Lacks, S. and Greenberg, B.: Complementary specificity of restriction endonuclease of Dipbcoccus pneumoniae with respect to DNA methylation. J. Mol. Biol. 114 (1977) 153-168. Laemmli, U.K.: Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227 (1970) 680-685. Lauster, R., Kriebardis, A. and Guschlbauer, W.: The GATATCmodification enzyme EcoRV is closely related to the GATCrecognizing methyltransferases DpnII and dam from E. coli and phage T4. FEBS Lett. 220 (1987) 167-176. Loenen, W.A.M., Daniel, AS., Braymer, H.D. and Murray, N.E.: Organization and sequence of hsd genes of Escherichia coli K-12. J. Mol. Biol. 198 (1987) 159-170. Macdonald, P.M. and Mosig, G.: Regulation of a new bacteriophage T4 gene, 69, that spans an origin of DNA replication. EMBO J. 3 (1984) 2863-2871. Mandel, M. and Higa, A.: Calcium-dependent bacteriophage DNA infection. J. Mol. Biol. 53 (1970) 159-162. Maniatis,T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982, pp. 54-96. Marmarelli, B.M., BaIganesh, T.S., Greenberg, B., Springhom, S.S. andLacks, S.A.: Nucleotide sequence ofthe DpnII DNA methylase gene of Streptococcus pneumoniae and its relation-

ship to the dam gene ofEscheri&a coli. Proc Natl. Acad. Sci. USA 82 (1985) 4468-4472. Marinus, M.G., Poteete, A. and Arraj, J.A.: Correlation of DNA adenine methylase activity with spontaneous mutability in Escherichia coli K-12. Gene 28 (1984) 123-125. Miner, Z. and Hattman, S.: Molecular cloning, sequencing, and mapping of the bacteriophage T2 dam gene. J. Bacterial. 170 (1988) in press. Revel, H.R. and Hattman, S.: Mutants ofT2gr with altered DNA methylase activity: relation to restriction by prophage Pl. Virology 45 (1971) 484-495. Sancar, A., Hack, A.M. and Rupp, W.D.: Simple method for identification of plasmid-coded proteins. J. Bacterial. 137 (1979) 692-693. Sanger F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Scarbrough, K., Hattman, S. and Nur, U.: Relationship of DNA methylation level to the presence of heterochromatin in mealybugs. Mol. Cell. Biol. 4 (1984) 599-603. Schlagman, S.L. and Hattman, S.: Molecular cloning of a functional dam+ gene coding for phage T4 DNA adenine methylase. Gene 22 (1983) 139-156. Schlagman, S.L., Hattman, S. and Marinus, M.G.: Direct role of the Escherikhia cob Dam DNA methyltransferase in methylation-directed mismatch repair. J. Bacterial. 165 (1986) 896-900. Van Ormondt, H., Gorter, J., Havelaar, K.J. and deWaard, A.: Specificity of a deoxyribonucleic acid transmethylase induced by bacteriophage T2, I. Nucleotide sequence isolated from Micrococcus luteus DNA methylated in vitro. Nucleic Acids Res. 2 (1975) 1391-1400. Communicated by M. Belfort.