A Bacillus subtilis dnaG mutant harbours a mutation in a gene homologous to the dnaN gene of Escherichia coli

227

Gene, 45 ( 1986) 227-23 1 Elsevier GENE 1703

A ~aei~lus subtilis dnaG mutant barbours a mutation in a gene homologous to the dnaN gene of

Escherichia coli (DNA replication; replication origin; gene cloning; recombinant change)

DNA; nucleotide sequence; single base

Naotake Ogasawara”, Shigeki Moriyaa, Ciorgio Mazza b and Hiroshi Yoshikawa”* a Department of Genetics, The University of Osaka, Medical School, 3-57, Nakanoshima I-chome, Kitaku, Osaka 530 (Japan) Tel. 06-443-5531, and b Dipartimento di Genetica e Microbiologia ‘A. Buzzati-Traverse’, Universitri di Pavia, via S. Epifanio, 14-27100 Pavia (Italy) Tel. (0382)31036 (Received May 31st, 1986) (Accepted June 23rd, 1986)

SUMMARY

A dnaG mutation of Bacillus subtilis, dnaG5, was found to be liked closely to recF. We have reported previously that two putative dna genes, ‘dnaA’ and ~dna~, highly homoiogous to Escbe~cbiu c&s dnaA and dnaN, respectively, were located adjacent to recF [Ogasawara et al,, EMBO J., 4 (1985) 3345-33501. Transformation by various fragments cloned from the ‘dnaA’-recF region of the wild-type cell revealed that a 532-bp AluI fragment containing 5’-portion of the ‘dnaN’ gene could transform the dnaG5 mutation. The nucleotide (nt) sequence of the same fragment cloned from the mutant cell shows a single nt change in the ORF of ‘dnap which in turn causes a single amino acid alteration from Gly to Arg. The ‘dnahr gene is now proven to be a dna gene, mutations in which result in instant arrest of chromosomal replication.

INTRODUCTION

A number of dna mutations have been isolated and mapped on the B. subtilis chromosome (Winston and Sueoka, 1982; Piggot and Hoch, 1985). During the

* To whom correspondence addressed.

and reprint requests should be

Abbreviations: aa, amino acid(s); bp, base pair(s); EtdBr, ethidium bromide; kb, 1000 bp; nt, nucleotide(s); ORF, open reading frame; ori, origin of DNA replication.

study of the ®ion of the B. sub&s chromosome, we found two putative dna genes linked closely to each other and contiguous with the recF gene (Ogasawara et al., 1985a; Moriya et al., 1985). The aa sequences of the possible protein products from these genes are highly homologous to those of dnuA and dnaN genes of E. colt’ (Ogasawara et al., 1985b; Moriya et al., 1985). Therefore we designated them as ‘dnaA’ and ‘dnaN’ (with quotation marks) until genetic evidence become available (Ogasawara et al., 1985a,b). In E. coli, dnaA is a gene essential for initiation of chromosomal (Hansen et al., 1977;

0378-11~9/86/$03.50 0 1986 Elsevier Science Publishers B.V. (Biomedical Division)

228

1982)

and

plasmid

pSClO1

(Hasunuma

and

Sekiguchi, 1977) replication and dnaN codes for P-subunit of DNA polymerase III (Ohmori et al., 1984). Therefore

it is important

to isolate mutants

EXPERIMENTAL

(a) Linkage between dnaG and recF

in

‘dnaA’ and ‘dnaN’ genes in order to identify the biological function of these putative dna genes in

established

B. subtilis.

mation

Here we report that one of the dnaG mutations, dnaG5 (Karamata

and Gross,

mapped

between

wrongly

AND DISCUSSION

Genetic

recM and spcD around

2-3” of the 360” genetic map (Piggot and Hoch, 1985), is indeed linked to recF and located in a fragment containing the 5’-portion of the dnaN ORF. The nt sequence analyses of the cloned DNA fragment from the mutant cells revealed that a single nt change occurred in the ‘dnaN’ gene which in turn resulted in a single aa replacement in the putative ‘dnaW protein.

DNA

dnaG5 and recF33 was

by the reciprocal

crosses

using

repulsion

DNA

isolated

transfor-

from

either dnaG or recF mutations

carrying

1970) which had been

linkage between

and by transforming

strains

strains

as donor

carrying

either

recF or dnaG mutations as recipients, respectively. Selection of either recF + /dnaG * or dnaG + /recF t recombinants shows clearly that these two mutations are closely linked to each other with a cotransfer index close to 0.70 (Table I). This conclusion was further confirmed by isolation from a B. subtilis DNA bank, constructed in the vector pJH101 (Ferrari et al., 1983), of a clone containing a large chromosomal fragment (11 kb) which showed ability to transform both dnaG5 and recF33 mutations (M. Perego, E. Ferrari and G.M., manuscript in preparation).

TABLE

I

Linkage

between

Recipient

dnaG and recF determined

straina

Donor

DNA”

by reciprocal

Selected

transformation

Transformants/ml

Recombinant

c

Cotransfer

phenotype Classes BD54

no DNA

DnaG +

1.2 x 10”

PB1798

DnaG +

2.3 x lo5

Number

DnaG+/RecF’

74

DnaG + /RecF-

PB1798

PB168

DnaG +

no DNA

RecF +

BD54

RecF +

a References

RecF +

strains

b DNA was prepared

as described

c The DnaG + or DnaGThe 238 dnaG

l

RecF + /DnaG

by Marmur

d Cotransfer

cells were prepared

were examined

DnaG’

agar plates containing

e % Recombination

- RecF-

(or RecF’

= lOO( 1-cotransfer

of spontaneous

30

1970), B. sub&s PB1798 (t&Z,

mefBl0, recF33)

by Stewart (1969) at 30°C and treated with 0.1 ng/ml

reversions.

index)

a).

- DnaG-)

164) phenotype.

to grow at 47°C on nutrient

The RecF + or RecF-

100 pg/ml methyl-methane-sulfonate

were examined

total DnaG + (or RecF + ) r A background

0.70

44 105

by ability or inability ofthe transformants

for RecF + (74) or RecF-(

NY). The 149 RecF + transformants

index =

and Gross,

as described

(1961) (see also footnote

was determined

by ability or inability to grow on nutrient Corp., Rochester,

+

for 60 min at 30°C.

phenotype

transformants

31

1.8 x 10’

et al., 1975). Competent

of DNA of the donor

0.69

0 1.2 x lo5

B. subtilis BD54 (mecB5, ile-1, dnaG5) (Karamata

for strains:

and PB 168(trpC2) (Mazza

164

2.5 x lo5

RecF + /DnaGPB168

%Recombination’

index d

(Eastman

phenotype

Kodak Organic

for DnaG + (44) or DnaG _ (105) phenotype.

agar plates.

was identified Chemical

229

1

[

“dnaA”

] n

“dnaN”

I=

[B

EPV

Spl I

1 ERV

ERI ERV Xhol

I 201

I 1431

ERI

Xho

I

SOlI

I 5209

, II79

I 2955 , I723

, 4251

L

I

1426

2613

I

2949 Alu’ra;, 1943

inserted

reported

previously

2474

XhoI end upstream in pN021, Helinski,

into each plasmid (Ogasawara

determined

by transformation

(Ogasawara

(Ogasawara

using cloned

pSM2052

814

pSM20 I 6

424 9 63

DNA

DNA was prepared

competent

method

in ‘dnuA’-recF region.

Restriction

illustrated

above the map. Below the map,

containing

each one of the fragments

are nt numbers,

et al., 1985a). The 532-bp Ah1 fragment

1969) and then purified in CsCl + EtdBr or by the alkaline-extraction

5

DNA fragments

of dnaG5 were listed. Plasmids

at both ends of fragments

et al., 1984). Plasmid

1979). Some 10 pg/ml of DNA was used to transform EcoRI;

691

is shown. ORFs are schematically

used for transformation

et al., 1985a). Numbers

of pBR322

pSM 1001

n0

from the ‘dnuA’gene

a derivative

1840

pSM2060

1

of dnaG5 mutation

pSM2003

pSM20 14

&,;,

enzyme map of the ‘dnuA’-recF region of the chromosome fragments

IO

pSM2002

1 1207

Fig. 1. Location

Number of transformants per 0.01 ml of competent cells

Name of the plasmid used

No. 1 corresponding

were

to the nearest

was first cloned in MI3 and then recloned by the cleared-lysate

(pSM2016

and pSM2060)

method

(Clewell and

(Birnboim

and Doly,

cells of B. subtih BD54 (see legend to Table I for the procedure).

ERI,

ERV, EcoRV.

(b) Localization of the dnaG mutation by restriction fragment analysis

(c) Localization

Based on the above observation we assumed that the dnaG mutation was located within either one of the two ORFs, ‘dna.A’ or ‘dnuhJ’. As illustrated in Fig. 1, the recF gene is located between ‘dnaW and gyrB. Therefore we first constructed plasmids containing various portions of ‘dnaA’-recF region in E. coli and then tested these plasmids for transforming activity for dnaG5. We found that the dnaG mutation is located within the 532-bp AluI fragment as illustrated in Fig. 1. According to the nt sequence data (Moriya et al., 1985) this fragment contains 6-bp corresponding to the two C-terminal aa of ‘DnaA’ protein, 191-bp non-coding spacer sequence between ‘dnaA’ and ‘dnaw, and 333-bp corresponding to the N-terminal 111 aa of ‘DnaN’ protein, suggesting that the mutation may be located within ‘dnaN’ gene.

Chromosomal DNA was isolated from a dnaG mutant strain, B. subtilis BD54, by conventional method (Marmur, 1961) and the SalI-digested fragments were separated by electrophoresis in a lowmelting 1y0 agarose gel. Fragments corresponding to about 4 kb were extracted from the gel, digested completely byAlu1 and fractionated in a low-melting 1.8% agarose gel to collect fragments of about 500 bp. They were cloned into the HincII site of M13mplO phage DNA. About 5 y0 of the phage plaques hybridized with the 32P-labeled plasmid DNA (pSM2060) containing wild-type AZuI fragment (532-bp) (see Fig. 1). About 300 nt were sequenced from both ends. The results in Fig. 2 clearly show that G at position 2358 of the wild type (Moriya et al., 1985) is substituted by A in the mutant. This gives rise to an aa substitution from Gly to Arg at aa 73 from the N-terminal of the putative ‘DnaN’ protein. No other changes were detected within the whole AluI fragment. Since the A/u1 fragment alone can transform the dnaG5 mutation, this

of the mutation site by nt sequence

determination

230

ThrIleGluGlnProArgSerIleVelLeu

indeed the B. subtilis counterpart

5'-U(CACUAUUGAACAGCCCpAGCAUCGUUUUAC$-3' 2358

\

of the /?-subunit

of

DNA polymerase III in E. coli. Further characterization of the mutant is necessary to prove this

I

possibility.

It also will be interesting

the single

aa change

to examine how

from Gly to Arg results

in

temperature sensitivity of the enzyme protein. Recently the dnaE gene of B. subtilis was cloned, sequenced

and found

to be highly homologous

dnaG, the gene for primase

2358

sequence

‘dnaW gene near

the mutation

of the wild-type site. Isolation

B. subrilis BD54 and cloning of a fragment 532-bp AluI fragment

in a Ml3 phage

section c. The nt sequence fragment

by the dideoxy chain termination

the procedure International

specified

strate the mutated is the nt number sequence

by the supplier

Ltd, Amersham,

gel of the wild-type

sequenced,

was determined

and mutant

together

of DNA

vector

from

some 1.2 billion years ago. We propose to rename the dnaG and dnaE genes of B. subtilis as dnaN and dnaG, respectively, because it is convenient to use the same names for genes homologous between B. subtilis and E. coli.

to the

are described

in

from both ends of the method

according

to

of the kit (Amersham

ofthe mutated

ACKNOWLEDGEMENTS

of the sequence

DNA are compared

as in Fig. 1. Since

the data are expressed

and the mutant

corresponding

U.K.). A portion

nt. The number

to

et al.,

1985). Thus, three E. coli dna genes, dnaA, dnaN and dnaG, have their counterparts in B. subtiiis, suggesting that proteins essential for chromosomal replication are strongly conserved between the two bacteria which are known to have evolutionally diverged

5'-tiCACUAUUGAACAGCCCGGAAGCAUCGUUUUACh-3' ThrIleGluGlnPro~luSerIleValLeu Fig. 2. Nucleotide

of E. coli (Wang

to demon-

nt (underlined)

the coding

strand

as the corresponding

was

mRNA

with aa sequence.

result indicates conclusively that the aa change from Gly to Arg is responsible for the dnaG mutation.

This work was supported by a Grant-in-aid for Special Project Research, for Cooperative Research, and for Scientific Research from Ministry of Education, Science and Culture, Japan, by ‘Progetto Finalizzato: Ingegneria Genetica e Basi Molecolari delle Malattie Ereditarie’, CNR, Rome (Italy) and by Consiglio Nazionale Delle Ricerche grant No. 83.01990.04.

(d) Conclusions REFERENCES

We present here evidence that a dnaG mutation is a mutation in the ‘dnaN’ gene of B. subtilis. We have identified the ‘dnaN’ gene by the homology of the aa sequence deduced from the DNA sequence with the sequence of the dnaN gene of E. coli. It is now proven that the ‘dnaN’ is indeed one of the dna genes in B. subtilis. The dnaG mutants were characterized as elongation type mutants, because DNA replication was arrested shortly after the shift to non-permissive temperatures (Karamata and Gross, 1970; Shivakumar and Dubnau, 1978). This fact suggests but does not prove that the gene product of dnaG is

Birnboim,

H.C. and Doly, J.: A rapid

cedure for screening

recombinant

alkaline

plasmid

extraction

Res. 7 (1979) 1513-1523. Clewell, D.B. and Helinski, D.R.: Super coiled circular protein

complex

conversion

in Escherichiu coli: purification

to an open circular

pro-

DNA. Nucl. Acids DNA-

and induced

DNA form. Proc. Natl. Acad.

Sci. USA 62 (1969) 1159-l 166. Ferrari,

F.A., Nguyen, A., Lang, D. and Hoch, J.A.: Construction

and properties J. Bacterial. Hansen,

of an integrable

plasmid

for Bacillus subtilis.

154 (1983) 1513-1515.

E.B., Hansen,

otide sequence

F.G. and von Meyenburg,

gene of Escherichia coli K-12. Nucl. Acids 7373-7385.

K.: The nucle-

of the dnuA gene and the first part of dnaN Res.

10 (1982)

231

Hansen,

F.G. and Rasmussen,

K.U.:

of the dnaA

Regulation

in Escherichia coli. Mol. Gen.

product

Genet.

155 (1977)

Hasunuma,

K. and

Sekiguchi,

in Eschenkhia

pSClO1 function.

Mol. Gen. Genet. D. and Gross,

Marmur,

and

G., Fortunato, Polsinelli,

recombination

for dnaA

mutants

of

of Bacillus subtilis defective

in

M.: Genetic

E., Canosi, and

U., Falaschi,

enzymic

studies

A.

on the

in Bacillus subtilis. Mol. Gen. Genet.

process

S., Ogasawara,

function

N. and Yoshikawa,

of the region of the replication

subtilischromosome,

III. Nucleotide

base pairs in the origin region.

H.: Structure

and

origin of the Bacillus

sequence

of some 10000

Nucl. Acids

Res. 13 (1985)

origin

N., Mizumoto,

S. and Yoshikawa,

of the Bacillus subtilis chromosome

hybridization

of the first replicating

ment from the replication

H.: Replication determined

DNA with cloned

by frag-

origin region of the chromosome.

S., von Meyenburg,

chromosomal

function

N., Moriya,

S. and Yoshikawa,

of the replication

H.: Structure

and

origin region of the Bacillus subtilis

K., Hansen,

F.G.

of genes and their organi-

replication

origin

region

of

Escherichia coli. EMBO J. 4 (198513)

H., Kimura,

M., Nagata,

T. and Sakakibara,

Y.: Struc-

of the dnaA and dnaN genes of Escherichia coli.

Gene 28 (1984) 159-170. Piggot,

P.J. and Hoch,

J.A.: Revised

Bacillus subtilis. Microbial. Shivakumar, Stewart,

Wang,

D.: Plasmid

of Bacillus subtilis. Plasmid

C.R.:

Physical

deoxyribonucleic markers.

genetic

heterogeneity

acid molecules

J. Bacterial.

linkage

map

of

Rev. 49 (1985) 158-179.

A.G. and Dubnau,

replication

in dna

1 (1978) 405-416. Bacillus subtilis

among carrying

particular

genetic

98 (1969) 1239-1297.

L-F., Price, C.W. and Doi, R.H.: Bacillus subtilis dnaE a

protein

homologous

to

DNA

primase

of

Escherichia coli. J. Biol. Chem. 260 (1985) 3368-3372. Winston,

S. and Sueoka,

In Dubnau, Academic

N.: DNA replication

D.A. (Ed.), The Molecular Press, New York,

Gene 30 (1984) 173-182. Ogasawara,

frames.

2267-2279.

3345-3350.

encodes

2251-2265. Ogasawara,

in the

ts mutants

136 (1975) 9-30. Moriya,

genes and other open reading

H.: Conservation

Bacillus subtilis and

tural analysis

of deoxyribonucleic

J. Mol. Biol. 3 (1961) 208-218.

A., Ferrari,

N., Moriya,

and Yoshikawa,

Ohmori,

108 (1970) 277-287.

for the isolation

Ogasawara, zation

and genetic analysis

Mol. Gen. Genet.

J.: A procedure

of plasmid

requirement

154 (1977) 225-230.

acid from a micro-organism. Mazza,

Replication

J.D.: Isolation

temperature-sensitive DNA synthesis.

M.:

coli K-12:

ofthe oriC region and expres-

IV. Transcription

sion of DNA gyrase

Nucl. Acids Res. 13 (1985a)

219-225.

Karamata,

chromosome,

Communicated

by J.A. Hoch.

in Bacillus subti1i.s. Biology

1982, pp. 35-69.

of Bacilli.