DNA methyltransferase genes of Bacillus subtilis phages: comparison of their nucleotide sequences

Gene. 42 (1986)

x9

89-96

Elsevier

GENE

1547

DNA methyltransferase (Recombinant

DNA;

genes of Bacillus subtilis phages: comparison of their nucleotide sequences

temperate

bacteriophages

SPR, SPP, @3T, Z; promoters;

molecular

evolution;

Bacillus

sphaericus M.BspRI)

A. Tran-Betcke,

B. Behrens, M. Noyer-Weidner

Mar.*_-Platlck-Institut,f~r molekulare (Received

December

4th, 1985)

(Accepted

December

30th, 1985)

Genetik. Ihnestrasse

and T.A. Trautner 63/73.

D-1000

Berlin 33 (Germanja)

Tel. 8307260

SUMMARY

The @3T DNA methyltransferase (Mtase) and most of the SPfi Mtase genes have been sequenced. With the exception of their promoters, no difference was found between the $3T and SPfl Mtase genes which code for an enzyme with a M, of 50 507, consisting of 443 amino acids (aa). Comparison of the deduced aa sequence of the $3T/SPfl type Mtase (target specificity: GGCC and GCNGC) with that of the previously established sequence of the SPR Mtase (Buhk et al., 1984) which has the target specificity GGCC and CCGG, reveals strong similarities between these two types of enzymes. There is, however, one striking difference: both the $3T/SPfl and the SPR enzymes contain at different positions inserts of 33 aa, which have no homology to each other. We suggest that the methylation specificity unique to each of the two types of Mtases (GCNGC in $3T/SP/?; CCGG in SPR) depends on these inserts, while the GGCC-specific modification potential common to all Mtases is determined by structures conserved in both types of enzymes. A DNA fragment of non-modifying phage Z, which shows homology to both flanks of the SPR Mtase gene, was also sequenced. This segment can be described as a derivative of SPR DNA, in which the Mtase gene and sequences at its 5’ end have been deleted, with the deletion extending between two direct repeats of 25 bp.

INTRODUCTION

thylates Two related kinds of Mtases encoded by a group of temperate Bacillus subtilis bacteriophages show multiple sequence specificity (Gunthert and Trautner, 1984). The Mtase of phage SPR me-

Abbreviations:

aa, amino acid(s); bp. base pair(s); Mtase, DNA

methyltransferase;

nt. nucleotide(s);

ORF, open reading

frame;

PA, polyacrylamide.

037X-I

I IY:Xh,‘S(13.50

C‘ 1986

Else&x

Science

Publlshers

B.V.

(Biomedical

(Trautner recently

cytosines

in the sequences

et al., 1980; Jentsch discovered,

GGCC,

CCGG

et al., 1981) and, as

also in the sequence

CC$GG

(U. Gunther& unpublished results). The Mtases of phages $3T, SPD, and pl 1 recognize the sequences GGCC and GCNGC; phage Z has no methylation capacity (Noyer-Weidner et al., 1983). The Mtase genes of these phages, as well as a DNA fragment of phage Z, which carries DNA sequences homologous to those flanking the Mtase genes of the Diwsion)

90

other phages, have been cloned 1983; Noyer-Weidner

(Kiss and Baldauf,

(c) Nucleotide

sequences

et al., 1985). The nucleotide

sequence of the SPR Mtase gene has been established (Buhk et al., 1984; Posfai et al., 1984) and

The approximate location of the $3T and SP/I Mtase genes in plasmids pBN16 and pBJ20 and

relatedness

their direction

been

of this gene to the $3T-type

documented

(Noyer-Weidner

by

DNA/DNA

genes has

hybridization

to define the structures

which

fragment

of transcription

(Noyer-Weidner

derived plasmid

et al., 1985).

We are interested

previously

had been determined et al., 1985). A pBR328-

(pBKU l), containing

a contiguous

of phage Z DNA, which has homology

DNA

in this communication

gene had also been described (Kupsch, 1984). Phage DNA for sequencing was obtained from these plas-

sequences

of those

in one polypeptide

is to compare Mtasegenes,

an identical

the nucleotide which

(GGCC)

encode

and a differ-

ent (CCGG/GCNGC) specificity. To this end we have sequenced DNA fragments which contain the Mtase genes of phages $3T and SPfl. A fragment of Z DNA, in which the Mtase gene had been deleted (Kupsch, 1984) was also sequenced.

EXPERIMENTAL

(a) Strains and plasmids DNA fragments containing parts of the Mtase genes of phages $3T and SPfl were obtained from pBR328 derivatives pBN16, pBN 17, pBN30, pBN32 and pBJ20, pBJ21 (Noyer-Weidner et al., 1985). Plasmid pBKU 1 (Kupsch, 1984) provided the relevant DNA of phage Z. All plasmids (see Fig. 1) were propagated on E. coli strain HB 101 (Boyer and Roulland-Dussoix, 1969). (b) DNA sequencing The Sanger dideoxy chain termination method (Sanger et al., 1977) was followed throughout. Overlapping restriction fragments were cloned into phages M 13mp8 and M13mp9 (Messing and Vieira, 1982). [35S]dATP was used in the DNA polymerase reaction starting from the Amersham universal 17-mer primer. Electrophoresis was carried out in 0.3 mm x 200 mm x 600 mm 6% PA gels. Multiple sequence determinations and/or sequencing of complementary strands were performed to maximize accuracy. Sequence evaluation was based on computer programs of Staden (1978) and Isono ( 1982).

flanking

the

3’ side

of

93T/SP[i

to

determine the specificity recognition of these enzymes. One experimental approach which we follow

and

SPR Mtase genes and the 5’ side of the SPR Mtase

mids (Fig. 1) or from plasmids

containing

subfrag-

ments of the Mtase genes. Regions sequenced (Fig. 1) include the Mtase gene of $3T and its flanking regions, part of the structural gene of SP/I and DNA on its 5’ terminus and Z sequences, which have homology to both the 5’ and 3’ flanking regions of the SPR Mtase gene. We also present in Fig. I the Mtase gene of SPR, which had previously been sequenced (Buhk et al., 1984). Within the 2360 bp of $3T DNA sequenced, the Mtase gene extends from codon TTG at position 7 14 to the termination codon TAA at position 2043 (Fig. 2). This assignment agrees with the previous localization of the Mtase gene. The sequence identitied represents the only ORF in the anticipated direction of transcription, which could code for the Mtase. This ORF codes for 443 aa. Based on the predicted aa composition, the 43T Mtase would have an M, of 50507. This value is in reasonable agreement with the M, of 47000 of the $3T Mtase derived from protein analyses of either in vitro synthesized Mtase or Mtase produced in minicells (Noyer-Weidner et al.. 1985). The @3T and SPP Mtases have the same M, and methylating specificity (Noyer-Weidner et al., 1985). Sequencing of the major part of the SPflgene revealed identity to the $3T sequences (Fig. 2). From this result and also from the finding of the same rcstriction pattern of the $3T and SPfl Mtase genes, u’c assume that also those nucleotide sequences of SP/I not sequenced are identical in SPfi and $3T. Based on restriction data, identity of $3T and SPP DNA is also likely beyond the 3’ ends of both genes. On their 5’ ends the genes are preceded by an identical region of 52 bp. Further upstream the q’,3T and SP/I sequences are different. The lack of methylation capacity of phage Z is a

Hmdlll

I I I I I

1 1 I I

I

---___-________ I

SW

SPR

HindlII I

t Fig. 1. Schematic

representation

Mtase genes and their direction analysis.

DNA/DNA

DNA regions

I

of Mtase genes and their environs. of transcription

hybridization,

sequenced.

Ikb

are indicated

or sequence

Only a limited number

analysis

Boxed regions

by open arrows.

are represented

of restriction

consequence of the absence of an Mtase gene in Z DNA (Noyer-Weidner et al., 1985). However, we have observed in Z DNA a single restriction fragment which has homology to the flanking regions of the SPR Mtase gene (Noyer-Weidner et al., 1985). When sequencing this DNA fragment (Figs. 1 and 2) and comparing it with SPR DNA, we realized that Z DNA can be described as a derivative of SPR DNA in which bp 257 to 1924 including the Mtase gene plus 348 bp to the 5’ end have been lost. This ‘deletion’ in Z extends between two quasi complete direct repeats in SPR DNA, which are found at coordinates 245 and 1913:

represent

Homologous

phage DNA, dashed

by the same kind of shadowing.

sites are shown.

Symbol

lines vector

DNA.

flanking regions, defined either by restriction d represents

Heavy lines below mark the

the deletion.

(GAAACTGATTAAAATATCTCTTTTA) (GAATCTGAATAAAATTTGTCTTTTA) (Fig. 2). Recombination leading to the deletion in Z must have occurred within the overlined region.

DISCUSSION

To facilitate comparison of the $3T, SPfl and SPR Mtases, we have aligned in Fig. 2 the sequences at the common codon TTG at locations 714 ($3T),

92

Fig. 2. Nucleotide

and predicted

aa sequences.

Sequences

TTG, where aa numbering

begins (coordinate

indicate

to those in the $3T sequence.

nt or aa identical

ofthe Mtase genes have been aligned at the common

+ 1). Asterisks

define sequences

Regions

not determined.

where matching

nt/aa

translational

start codon

Blanks in SP/!, SPR and Z sequences

are absent

arc represented

by dashes.

The

Shine-Dalgarno bracketing

(SD) sequence

the deletion

was sequenced

and the -10 and -35 regions

in Z DNA are indicated

in regions

corresponding

by arrows.

of putative

promoters

The SPR sequence

are underlined.

presented

Two direct repeats

in SPR DNA

is taken from Buhk et al. (1984). SPfi DNA

to nt 295 to 1142 and nt 1581 to 1923 of the $3T sequence

(see Fig. 1).

94 1 AmInoacIds

A

IOC ._A

____-

___

200

300

L-

./._____i_L

400

500

@ 3T

SPR

BspRI Fig. 3. Schematic phage enzymes

representation are stippled,

of aa homologies

while those between

between

phage

and B. sp~ue~i~l~~ (BspRI)

phage and bacterial

enzymes

are hatched.

Mtases. Sequence

Regions

of homology

data for the B.yRI

hctwern

Mtascs

arc

from Posfai et al. (1983)

420 (SPfl), 605 (SPR), respectively.

This was shown to be the translational start of the SPR Mtase gene (Buhk et al., 1984; R. Lauster, unpubIished) and is most likely also the start signal in the $3T and SP/?Mtase genes. A comparison of the deduced aa sequences of the SPR and 4b3T/SPb Mtases (Fig. 2) reveals strong sjmilarities between the two types of enzymes. They have almost the same number of aa (439 in SPR; 443 in $3T/SPfl) and extensive regions of homology, ranging between 55 7; and 90% identity. There are some 50 aa both at the

N and C ends of the enzymes which are virtually identical. Amino acid sequences which show strong holnology are also found in the central portions of the enzymes (Fig. 3). The close relatedness of the enzymes, based on sequence comparison, is reflected in their immunological cross reactivity (R. Lauster and U. Gtinthert, unpublished). The most striking differences between these enzymes are two regions of 33 nonmatched aa within each Mtase. These ‘inserts’ have no homology to each other. They extend from aa 92 to 124 in the $3T/SP/3 Mtase and from aa 297 to 329 in the SPR enzyme. We speculate that the inserts represent determinants of that nlethylation specificity which is unique to each of the two types of Mtases, i.e., CCGG methylation in SPR and GCNGC methy-

lation in $3T/SP/I. Support for this speculation comes from the localization of mutations. We have found three independent mutants of SPR, which are still capable of methylating the sequence GGCC, but are unable to methylate the sequence CCGG, which is uniquely recognized by the SPR enzyme. Two of these ~nutations are caused by aa changes within the insert of SPR, the third mutation is located three aa away from the insert (Buhk et al., 1984 and unpublished results). In contrast, mutations causing a defect in both Mtase activities were found to be located all over the Mtase gene. We are presently attempting to obtain and localize additional SPR mut~ts which are defective in only CCGG methylation. Mutagenesis experiments with $3T are also in progress. In analogy to SPR, mutations within the non-homologous 99-bp insert of $3T DNA would affect the capacity of this enzyme to methylate the Fnu4HI sequence, which is uniquely recognized by the (P3T Mtase. Since no B. .wdttilis strain with ~~~4H~-speci~c restriction~modi~cation has been identified, such mutants could so far not be selected. As has been shown previously (Buhk et al., 1984; Posfai et al., 1984; Kiss et al., 1985), conserved regions characterized by 40 to 60”” identical aa, are also observed between the SPR enzyme and modifi-

95

cation

Mtases

of B. sphaericus R and B. subtilis R.

This would point to an ancestral with this evolutionary $3T/SPp

Mtase gene. Along

consideration

sequence information

the additional

provided

here shows

that conserved regions between bacterial and phage Mtases fall exclusively into regions which are shared by the phage enzymes The

nucleotide

finding

of the phage Mtases

might require

quence

as an

of

the

$3T

and

studied

a redefinition

element

here, this

of the -35 se-

in promoter

recognition.

We are presently conducting Sl mapping experiments to obtain direct information on the location of the promoter(s) Nucleotide

(Fig. 3).

sequences

expression

used by these genes. sequences

3’ to the genes and to the

deletion of phage Z DNA are almost fully conserved.

SPPMtase genes including the insert pattern are identical. The pl 1 Mtase gene has most likely the

This includes

same sequence,

present a transcriptional stop signal, and an ORF of 72 aa extending between coordinates 2097 and 23 13

since its restriction

to those of $3T/SPB.

Furthermore,

map is identical it has previously

been reported that there is strong DNA/DNA homology between the cloned ~111 and $3T Mtase genes (Noyer-Weidner et al., 1985). The restriction maps of the $3T type Mtase genes are distinct from that of the corresponding SPR gene. This is due to both real aa differences between the two types of enzymes and the degeneracy of the third position of the codon used within regions of aa homology within regions of aa homology. Looking at sequences surrounding the Mtase genes (Figs. 1 and 2), we recognize in all sequences a highly conserved region starting 52 bp upstream from the translational start signal. However, the preceding sequences are different in all three phage DNAs. The DNA beyond the 5’ end of all three genes analyzed here must contain signal elements (underlined in Fig. 2) which define the transcriptional start, since the fragments sequenced (Fig. 1) express the Mtase gene when cloned into heterologous vectors irrespective of their orientation with respect to the vector promoters (Noyer-Weidner et al., 1985; Giinthert et al., 1986). Within the conserved 52-bp region we can recognize at a consensus distance of 6 bp from the translational start codon a ribosomal binding site GGAGGT. Several sequences in this region would qualify to be the -10 promoter sequence. This would imply that the -35 regions of all genes sequenced fell into the region of non-homology and hence would be different for the three genes analyzed here. The putative -35 regions’ (underlined in Fig. 2) resemble, in the case of SPR and $3T the consensus sequence TTGACA of the vegetative obl RNA polymerase, whereas the corresponding sequence CTAAA in SPBis identical to the “ -35 region” described for the 0” RNA polymerase (Gilman et al., 1981; Moran et al., 1982). Apart from the significance for the

following

of $3T for the ed in et al.,

a palindromic

sequence

the TAA stop codon,

immediately

which

might

re-

DNA. It is conceivable that this DNA codes protein of approx. 11 kDa, which was observprevious studies with $3T (Noyer-Weidner 1985). The function of this protein is not

known. Through DNA/DNA hybridization experiments and electron microscopic inspection of appropriate heteroduplex molecules we had previously demonstrated homology between the fragment of Z DNA shown in Fig. 1 and both flanks of the SPR Mtase gene. The sequence comparison between the cloned Z and SPR DNAs demonstrated the absence in Z DNA of 1668 bp of DNA, including the Mtase gene. Since neither phage Z nor for that matter any of the SPR or $3T mutants defective in Mtase activity have any loss of viability, the Mtase genes must be dispensable. Still these genes remain conserved in the methylating phages in the absence of apparent selective pressure for gene maintenance. Such stability is particularly surprising in phage SPR, where repetitious DNA sequences bracketing the Mtase gene would obviously facilitate deletions similar to that found in phage Z.

ACKNOWLEDGEMENTS

We thank R. Lurz for cheerful and patient help in computer work and R. Lauster, U. Gtinthert, J. Kupsch, K. Wilke and P.A. Terschiiren for discussions and the communication of unpublished results. M. N.-W. is supported by the Deutsche Forschungsgemeinschaft (Tr 25/10- 1).

96

REFERENCES

Maniatis,

T., Fritsch,

A Laboratory Boyer,

H.W. and

analysis

Roulland-Dussoix,

of the restriction

Spring

D.: A complementation

and

modification

of DNA

in

Escherichia coli. J. Mol. Biol. 41 (1969) 459-472. Buhk,

H.-J.,

Behrens,

Gtinthert,

B., Tailor,

U.. Noyer-Weidner,

T.A.: Restriction tide sequence,

R.. Wilke, M., Jentsch,

and modification functional

DNA methyltransferase

J.J..

nucleo-

and product

gene of bacteriophage

of the

SPR. Gene 2Y

M.Z., Wiggs,

sequences

J.L. and Chamberlin.

RNA polymerase.

M.A.,

strand

of double-digest

Sonnenschein, quences

A.L., Pero, J. and Losick.

that signal the initiation M., Jentsch, T.A.:

phages Z,

SPR,

L.-F.

and

exists between

Doi,

R.H.:

the major

A strong

sequence

RNA polymerase

4 fac-

Noyer-Weidner, Trautner,

U. and Trautner,

T.A.: DNA methyltransfcrases

subtilis and its bacteriophages.

(Eds.), Methylation

of DNA. Current

and Immunology,

of

In Trautncr,

‘I .A.

Berlin, 1984, pp.

L. and Lauster,

R.: Cloning

I l-22.

and expres-

sion of Bacillus subtilis phage DNA methyltransferasc

genes

Isono, K.: Computer

programs

to analyze DNA and aa sequence

M., Jentsch,

S., Gtinthert, affecting

U. and Trautner, the sequence

(19X3)

J.. Bergbauer,

M. and

F.: Molecular

coli of two

Nucl. .4cids Res. 9

cloning

modification

G., Keller,

R.J.: Nucleotide

sequence

CC.,

and expression

methylase

genes

in of

11l-l 19.

Bacillus subrilis. Gene 21 (1983)

tion system.

S., Posfai,

gene

expression.

J. and Venetianer,

of the BaciNus .rphuericus R modification

Posfai, G.. Baldauf,

P.:

methylase

F.. Erdei. S.. Posfai. J.. Venetianer,

Kiss, A.: Structure

P. and

of the gene coding for the sequence-speci-

fic DNA-methyltransferase

of the B. subdis

Sanger, I;., Nicklen, S. and Coulson. inhibitors.

Staden,

R.: Further

procedures

Venetianer.

Trautncr. and

T.A., Pawlek. Freund.

phagc

SPR.

A.R.: DNA sequencing

with

Proc. Natl. Acad. Sci. USA 74 for sequence

subtilis: Identification P. and Roberta.

of the BsuRI restriction-modifca-

coding

B., Gtinthert,

M.: Restriction

analysis

Mol. Gen. Gcnct.

J.: Untersuchungen

zur Evolution

temperenter

Freie Universitat,

Berlin,

1984.

Diploma Commumcatcd

modification

1X0 (1980) 361-367.

der DNA-Methyl-

B. subrilis-Phagcn.

U., Canosi,

by com-

U.. Jcntsch,

and modification

of a gene in the temperate

for a BsuR specific

Nucl. Acids Res. 13 (1985) 6403-6421.

transferase-Gene

and

puter. Nucl. Acids Res. 5 (1978) 1013-1015.

Kiss, A. and Baldauf,

Kiss, A., Posfai,

genes ofBcrci/lus suh-

relatedness

G., Kiss, A., Erdei,

Structure

S., Kupsch.

(1977) 5463-5467.

T.A.: DNA methyltrans-

5’CCGG.

( I98 I) 2753-2759. Escherichiu

Posfai,

structural

chain-termmating

data. Nucl. Acids Res. 10 (1982) 85-89.

Thesis,

bactcrio-

J. Virol.46

Nucl. Acids Res. I2 (1984) YO39-YO4Y.

in Escherichia coli and B. subtilis. Gene 41 (1986) 261-270.

Kupsch.

of the related pll.

U. and

in Bacillrrs

gcnc. J. Mol. Biol. 170 (1983) 597-610.

Topics in Microbiology

Vol. 108. Springer,

U., Reiners,

ferases

B.. Gtinthert,

T.A.: DNA methyltransfcrasc

rills phages:

Jentsch,

and

se-

and trans-

modification

potential

SPB, $3T

Gene 35 (1985) 143-150.

Gtinthert,

and

M..

R.: Nuclcotide

IX6 (1982) 339-346.

S., Pawlek,

Restriction

tors et Bucillus subrilis and Escherichicr co/i. J. Biol. Chem.

Budus

fragments.

of transcription

lation in B. subtilis. Mol. Gen. Genet.

260 (1985) 7178-7185. Gtinthert,

restriction

446-453.

Wang,

homology

DNA

subrilis: DNA methylation

used by Bad/w

Nucl. Acids Rcs. Y (1981)

599 1-6000. Gitt,

Cold

NY, 1982.

Gene 19 (1982) 269-276.

Trautner,

M.J.: Nucleotidc

of two Bacillus suhtilis promoters

.subtilis sigma-28

Cloning.

Laboratory,

Messing, J. and Vieira, J.: A new pair of MI3 vectors for selecting

Noyer-Weidner,

(1984) 51-61. Gilman,

J.: Molecular

Moran, Jr., C.P., Land, N., Leguce, S.F.J., Lee, G., Stephens,

S. and Trautner.

in Brrcillus suhrik

organization

Harbor,

either

K., Pradn,

D.F. and Sambrook,

Manual. Cold Sp,,ing Harbor

by T. Bicklc

S.

in Bucillus phagc

SP/?

methyltransferase.