MGG

Mol Gen Genet (1988) 213:118-124

© Springer-Verlag 1988

Discoordinate gene expression in the dnaA-dnaN operon of Escherichia coil Ariel Quifiones 1 and Walter Messer 2

1 Wissenschaftsbereich Genetik, Martin-Luther-Universit/it, Domplatz 1, DDR-4020 Halle (Saale), German Democratic Republic z Max-Planck-Institut ffir Molekulare Genetik, Ihnestrage 73, D-1000 Berlin 33

Summary. The dnaN gene of Escherichia coli encodes the fl-subunit of the D N A polymerase III holoenzyme. Previous work has established that dnaN lies immediately downstream of dnaA and that both genes may be cotranscribed from the dnaA promoters; no promoter for dnaN has been described. We investigated the in vivo regulation of transcription of the dnaN gene by transcriptional fusions to the galK gene, translational fusion to the lacZ gene and S1 mapping analysis. We found that there are at least three dnaN promoters residing entirely in the reading frame of the preceding dnaA gene, and that transcription from these promoters can occur independently of dnaA transcription which, however, extends at least up to dnaN. Furthermore, we found evidence for the inducibility of the dnaN promoters in a dam background under conditions of simultaneously reduced dnaA transcription. These results are consistent with the hypothesis that although dnaA and dnaN are organized in an operon considerable discoordinate transcription can occur, thus uncoupling dnaN and dnaA regulation, when needed. Key words: dnaN transcription - dnaA operon - galK fu-

sions - S1 m a p p i n g - Gene regulation

Introduction

At 83 min of the Escherichia coIi chromosome lies a cluster of genes which play essential roles in D N A replication. The gene order, in the counter-clockwise direction on the chromosome, is dnaA dnaN recF gyrB (Bachmann 1983). The dnaA gene product has been identified as essential for initiation of the chromosomal replication of E. eoli in vivo and in vitro (Hirota et al. 1968; yon Meyenburg et al. 1979; Fuller et al. 1981). It is a basic protein of 52500 daltons which binds specifically to oriC D N A (Chakraborty et al. 1982; Fuller and Kornberg 1983) by recognition of a 9 bp repeat, the so-called dnaA box (Fuller et al. 1984), thus leading to the assembly of an initiation complex to start the replication of the chromosome (Kornberg et al. 1987). This dnaA box occurs four times within oriC, in the promoter region of the dnaA gene, and within several other genes on the chromosome (Fuller et al. 1984; Matsui et al. 1985). The expression of the DnaA protein is autogeneously regulated by binding to its own dnaA box between the two Offprint requests to: A. Quifiones

promoters on the dnaA gene (Atlund et al. 1985; Braun et al. 1985; Kficherer et al. 1986). The dnaN gene encodes the fl-subunit of the D N A polymerase III holoenzyme (Burgers et al. 1981), lies immediately downstream of the dnaA gene (Sakakibara and Mitzukami 1980) and seems to be cotranscribed with dnaA as an operon (Sako and Sakakibara 1980; Sakakibara et al. 1981) from the two dnaA promoters (Hansen et al. 1982; Ohmori et al. 1984). The fl-subunit of D N A polymerase III seems to be responsible for the high processivity and for the dissociation and association of the holoenzyme from one template to another (McHenry et al. 1987; Maki et al. 1987; Kornberg et al. 1987). The recF gene lies immediately downstream of dnaN, encodes a protein of about 40 kDa (Blanar et al. 1984) and its precise function is unknown. However, the phenotype of reeF mutants suggests that the reeF gene product is involved in SOS induction, in recombination and possibly in D N A replication (Horii and Clark 1973; Armengod 1982; Clark et al. 1979, 1982; Wang and Smith 1986; Wood and Stein 1986; Schaaper et al. 1982; Thoms and Wackernagel 1987). The gyrB gene encodes the ca. 90 kDa fl-subunit of D N A gyrase (Gellert 1981), a type II topoisomerase which catalyses the supercoiling of relaxed closed circular D N A (Gellert 1981 ; Drlica 1984) and is required in both initiation and elongation of minichromosome replication (Kaguni and Kornberg 1984; Funell et al. 1986; Kornberg et al. 1987) in vitro. Gene cloning and D N A sequence analysis have revealed that the dnaN start codon is separated by only 4 bases from the end of dnaA (Hansen et al. 1982; Ohmori et al. 1984), that dnaN and recF overlap by 1 base (Blanar et al. 1984; Adaehi et al. 1984), and that gyrB and reeF are separated by only 30 bases (Adachi et al. 1984; Blanar et al. 1984; Adachi et al. 1987). The transcription of these genes has not been completely worked out and to date little is known about the regulation of their expression. Two transcripts start from the two dnaA promoters which are negatively controlled by the DnaA protein (Atlund et al. 1985; Braun et al. 1985; Kiicherer et al. 1986) and these seem to be extended through dnaN. Thus it has been postulated that dnaA and dnaN represent an operon (Sako and Sakakibara 1980; Sakakibara et al. 1981 ; Ohmori et al. 1984). The termination site of these transcripts, if any, is not known and at present it is not clear whether reeF and perhaps gyrB are part of the dnaA operon. However, one promoter residing in the reeF reading frame has been found for gyrB and recently two poor promoters residing in the reading frame of dnaN have been found for reeF (Adachi et al.

119 Table 1. Bacterial strains and plasmids a

Strain

Genotype

Source

ABl157 CSH-26 CSH-26 F6 GM2159 AQ200 SR1214 WM 1344 AQ265 AQ266 HBI01 EM614 WM1499 WM1500 WM1501 HC194

thr thi proA his-4 argE str-31 sup-37 lacY1 galK2 (proA-laclZYA) thi As CSH-26, but A (srl-recA)F6 As ABI 157, but daml3 : : Tn9 recF143 As ABl157, but daml3::Tn9 argE3 his-4 leuB6 thi A (gpt-proA) ara lacY1 galK rpsL31 supE44 tnaA:: tnlO dnaA46 argE3 his-4 leuB6 thi A (gpt-proA) ara lacY galK rpsL31 As WM1344, but tnaA::TnlO As ABl157, but tnaA::TnlO dnaA46 ara A (gpt-proA) galK2 lac Y1 hsdS20 leu recA 13 rpsL20 thi argE his-4 lacY1 galK2 mtl proA recA56 recB21 rpsL thi thr, carries pFD51 araD A (ara-leu) galK A (lac)X74 galU hsdR rpsL, carries pNM480 As WM1499, but carries pNM481 As WM1499, but carries pNM482 HfrC dnaN59 metB thi-1

Laboratory stock Laboratory stock S. Casaregola M.G. Marinus P1 (GM2159) x AB1157 N. Sargentini Laboratory stock P1 (SR1214) x WM1344 Pl (AQ265) x AB1157 Laboratory stock Laboratory stock Minton (1984) Minton (1984) Minton (1984) Burgers et al. (1981)

a All other plasmids used and constructed,, are described in the text 1984; Armengod and Lambies 1986). To date no promoter has been mapped for dnaN (Ohmori et al. 1984). In this paper we describe the localization of the promoter region of dnaN, which resides completely within the reading frame of the preceding dnaA gene, and the mapping of the initiation sites for dnaN transcription starting from three independent promoters. We alsc present evidence for dnaA-independent transcription of the dnaN gene.

Materials and methods

Bacterial strains, media and genetic procedures. The E. coli strains and plasmids used are listed in Table 1. Growth media and McConkey plates are described by Miller (1972). Transformations were done as described by Maniatis et al. (1982). PI transductions for strain construction were performed essentially as described by Miller (1972). Molecular cloning. Restriction enzyme digests, ligations and agarose gel electrophoresis were done in general as described by Maniatis et al. (1982). Preparation o f plasmids and D N A fragments. The boiling method of Holmes and Quigley (1981) was used for routine and rapid examination of plasmid DNA. For highly purified plasmid DNA, the cleared lysate method described by Silhavy et al. (1984) was used following caesium chloride gradient centrifugation according to Maniatis et al. (1982). Fragments were isolated from agarose gels by electroelution according to Maniatis (1982). 5' labelling o f D N A fragments. For the determination of the initiation sites of the dnaN transcripts a 1430 bp EcoRIHinfI fragment (see Fig. 1) used as specific probe was labelled, after dephosphorylation with calf intestinal alkaline phosphatase (Boehringer), with [7-32p]ATP and polynucleotide kinase (Boehringer) as described by Maniatis et al. (1982). The 1,272 bp PvuII-HinfI fragment was isolated after cutting with PvuII. R N A isolation. Total cellular R N A was isolated according to Brosius et al. (1982). Bacteria were grown to a density

of 3-5 x 10S/ml and harvested. Cells from a 500 ml culture were pelleted and resuspended in 50 mM Tris-HC1, pH 7.9. After addition of lysozyme, 250 mM EDTA, pH 8.0 and Triton-Mix (0.3% v/v Triton X-100, 0.18M EDTA, pH 8.0, 0.15 M Tris-HC1, pH 8.0), the mixture was incubated for 10 min at 0°C with gentle shaking and centrifuged at 12000 rpm for 30 rain. After repeated extraction with phenol the R N A was precipitated with ethanol and resuspended in H 2 0 (treated with diethylpyrocarbonate) aliquoted and stored at - 7 0 ° C. The R N A concentration was determined by measuring the optical density at 260 nm (1 0D260 = 40 gg/ml RNA). S1 nuclease mapping. For $1 nuclease mapping of the dnaN transcripts a 1,272 bp PvuII-HinfI fragment carrying the dnaN promoters, labelled at the 5' end of the HinfI site, was used as a specific hybridization probe for the dnaN transcripts. $1 mapping was performed essentially according to Berk and Sharp (1978) as described by Quifiones et al. (1987). Briefly, 25000 cpm of the labelled probe and 200 gg R N A were coprecipitated and resuspended in 30 gl hybridization buffer (40 mM Pipes, pH 6.3, 400 mM NaC1, I mM EDTA, pH 6.5, 80% formamide). After denaturation the samples were hybridized for 4 h at 55 ° C. Following addition of 300 gl $1 buffer (280 mM NaC1, 30 mM sodium acetate, pH 4.6, 4.5 m M zincacetate, 20 gg/ml denatured calf thymus DNA) the samples were digested with 100 U $1 nuclease (Boehringer), stopped with 0.5 M ammoniumacetate, 10 m M E D T A and precipitated with 2-propanol. The Sl-resistant nucleic acids were resuspended in loading buffer (80% formamide, 1 m M EDTA, 0.025% bromphenolblue, 0.025 xylene cyanol) and analysed by electrophoresis on 8% polyacrylamide gels containing 8 M urea. Galactokinase assay. The intracellular galactokinase activity of galK bacteria carrying promoter-galK fusion plasmids was determined according to McKenney et al. (1981) with the modifications described by Quifiones et al. (1987). Cells for galactokinase assay were grown in M9 medium (Miller 1972) containing 0.2% fructose, 0.5% casaminoacids (Difco), 10 rtg thiamine 50 gg/ml ampicillin. Units are expressed as nanomoles galactose phosphorylated per minute per millilitre cells per OD,s o- The copy number of promoter

120 H c)

w

kb5

4

I

X

~

[

C

LAJI

3

2

I

W

11

I

44 gxrB

recF

dna N

dnoA

11 a

| i

w ".t

b

',

i

•

Fig. I a, b. Structure of the dnaA-gyrB chromosomal region of Escherichia coli. The genes are indicated by dotted boxes. Promoters are symbolized by black arrowheads. The t.46 kb PvuII fragment carrying the dnaN regulatory region and the 946 bp EcoRI fragment carrying the dnaA regulatory region which were fused to the galK and lacZ genes are indicated by open boxes. The DNA probe used for S1 mapping (a) and the resulting Sl-protected fragments Co) are shown. The 5' end-labelled sites are indicated by stars. Relevant restriction sites are given. In addition, H and A on the PvuII fragment signify HinfI and AvaII, respectively

plasmids was not significantly different as judged by rapid isolation ofplasmid D N A from cells for galactokinase assay and analysis by agarose electrophoresis. fl-galactosidase assay. The intracellular fl-galactosidase activity of A (lac-pro) bacteria carrying translational fusions to the lacZ gene (Minton 1984) was determined essentially as described by Miller (1972).

Results

Fusion o f the d n a N promoter region to the galK gene As source for the dnaN promoter region we used plasmid pHPC1 (Chakraborty et al. 1982) carrying a 3.6 kb EcoRI fragment from the E. coli chromosome which contains 95% of the dnaA gene, the dnaN gene, the recF gene and the initial part of the gyrB gene (see Fig. 1). Since dnaN appears to be in the same transcriptional unit as dnaA (Sako and Sakakibara 1980; Sakakibara et al. 1981) and to date no promoter for dnaN has been mapped (Sakakibara et al. 1981; Ohmori et al. 1984), we isolated a 1,561 bp PvuII fragment from pHPC1 containing the start of the dnaN gene and a major part of the dnaA gene. It was ligated to the SmaI site of the promoter probe vectors pFD51 (Rak and von Reutern 1984) and p U T E I 3 , a derivative of pFD51 in which the polylinker region (EcoRI to SmaI) had been replaced by the polylinker of M13mp19. Intentionally we included sequences upstream of the last AluI site in the dnaA gene, because Ohmori et al. (1984) had found no promoter in an in vitro transcription system between this AluI site and the XhoI site in the dnaN gene (positions 1479-1785 in Ohmori et al. 1984; see also Fig. 1, positions 1106-2437 in Hansen et al. 1982). Galactokinase-positive clones were obtained after transformation into strains WM614 or HBI01 and the orientation of the PvuII insert was determined by mapping the AvaII site. When the potential promoter region was inserted in the transcriptional direction towards the galK gene in pFD51 and pUTE13 the plasmids were designated pLSK91-3 (R. K611ing and W. Messer, unpublished) and pAQ31, respectively. In addition to these dnaN-galK transcriptional fusions we also constructed a

dnaN-lacZ translational fusion using the p N M vectors developed by Minton (1984) by ligation of the PvulI fragment into the Sinai site of pNM481 resulting in plasmid pAQ47. Since fl-galactosidase-positive clones were isolated with the expected orientation of the insert in the Sinai site of pNM481, this revealed that the translation initiation site must be the A T G codon at position 2295 (Hansen et al. 1982) as suggested by Ohmori et al. (1984) and McHenry et al. (1987), because only this start site and not the start Table 2. Expression level of the dnaN promoters in comparison with several reference promoters Strain a

Plasmid

Promoter fusion

Galactokinase units b

AQlI0 AQ198 AQ199 AQI96 AQI97 AQlll AQII2 AQ191

pFD51 pUTE13 pAQ31 pLSK91-3 pLSK88-1 pAQ2 pAQ4 pLSK-1

dnaN-galK dnaN-galK dnaA-galK dnaQ-galK rnh-galK lacUV5-galK

0.8 0.5 12.1 10.8 25.7 31.5 12.5 95.3

Strain °

Plasmid

Gene fusion

fl-Galactosidase units d

AQ250 AQ251 AQ252

pNM481 pAQ47 pAQ 184

dnaN-lacZ rnh-lacZ

185 t 240

a All strains in AB1157 background b Cells for galactokinase assay were grown to an OD45o=0.4. Samples 1 ml were frozen in liquid nitrogen and assayed for galactokinase activity. Galactokinase units are mean values of four independent experiments with triplicate samples for each strain All strains in CSH-26 background d Cells for fl-galactosidase assay were grown to an OD4so = 1.0 and assayed for fl-galactosidase activity. Units are the average of two independent experiments with duplicate samples

121 Table 3. Discoordinate transcription from the dnaA and dnaN promoters Strain a

AQ198 AQ199 AQ196 AQ197 AQ201 ] AQ202 Ae203 [ AQ204 l

Relevant genotype a

Wild type

dam : : Tn9

AQ267 ] AQ268 AQ269 | AQ270 )

dnaA46

Plasmid

Promoter fusion

pUTE13 pAQ31 pLSK91-3 pLSK88-I pUTE13 pAQ31 pLSK91-3 pLSK88-1 pUTE13 pAQ31 pLSK91-3 pLSK88-1

dnaN-galK dnaN-galK dnaA-galK dnaN-gaIK dnaNgalK dnaA-galK dnaN-galK dnaN-galK dnaA-galK

Galaktokinase 30° C

activity b 42° C

0.5 11.5 10.5 26.0 0.4 15.8 14.0 18.2 0.5 t 1.2 10.5 27.3

0.6 18.2 16.9 17.5 0.5 35.8 33.1 11.5 0.7 17.8 16.4 38.9

All strains were constructed in an ABl157 background b Bacteria for galactokinase assay were grown to stationary phase in galactokinase assay medium containing 50 gg/ml ampicillin. Cultures were diluted I : 100 in the same medium and growth to an OD~5o of 0.2. The cultures were divided into 30° C and 42° C samples and grown to a n O D 4 5 o of 0.5. Samples were frozen and assayed for galactokinase activity. Units are mean values of three independent experiments with duplicate samples

site proposed by Hansen et al. (1982) lies in frame with the truncated lacZ gene of pNM481. The galactokinase levels obtained with E. coli cells carrying the dnaN-galK fusions are shown in Table 2. These revealed unambiguously that dnaN possesses its own promoter(s) which must be localizated between the PvuII site and the AluI site within the reading frame of the dnaA gene (positions 1102 2133 in the sequence of Hansen et al. (1982), and positions 450-1480 in the sequence of Ohmori et al. (1984)). Table 2 shows also, by comparison of the strength of several promoters from our collection with the dnaN promoter, that the dnaN promol:er is weak and possesses about half the activity of the dnaA promoters. Furthermore, comparison of the galactokinase activities of transcriptional galK fusions and the fl-galactosidase activities of translational lacZ fusions for dnaN and rnh, two genes with very similar promoter strengths, shows an unexpectedly low fl-galactosidase activity for the former probably due to the absence of a good ribosome binding site immediately 5' to the translation start. In order to examine whether the transcription activity starting from this promoter region possesses any translational relevance for the dnaN gene we used plasmid pHPC1 (see Fig. 1) to transform strain HC194 carrying the dnaN59 mutation. As expected, this plasmid which lacks the dnaA promoter region was able to complement the dnaN59 mutation as judged by the ability to confer temperature-resistant growth.

Transcription of the d n a N gene occurs independently o f d n a A transcription In order to elucidate whether expression of dnaN from its own promoter(s) is related to transcription starting from the dnaA promoters, we measured in vivo galactokinase activity using the dnaN-galK fusions and a dnaA-galK fusion under conditions in which dnaA transcription is changed in comparison with the wild-type pattern. Previous studies on the regulation of dnaA transcription have revealed that in dnaAts mutants transcription from the two

dnaA promoters is derepressed at 42 ° C and that overproduction of D n a A protein represses promoter activity, showing the autoregulation of dnaA expression. Additionally, in dam mutants dnaA transcription is reduced (Atlund et al. 1985; Braun et al. 1985; Braun and Wright 1986; Kficherer et al. 1986). Therefore we compared dnaA and dnaN expression in the wild type, in the dnaA46 mutant, and in the dam19:: Tn9 mutant. These results, which are summarized in Table 3, show clearly that transcription starting from the dnaN promoters is uncoupled from dnaA transcription, because there are conditions which show reduced dnaA transcription but simultaneously enhanced dnaN transcription and vice versa. Furthermore, these results reveal that expression of the dnaN gene is induced at 42 ° C. The most pronounced differential effect was observed in dam mutant strains, in which dnaN transcription was induced at 42 ° C relative to 3 0 ° C or to dam +, whereas dnaA transcription showed the opposite effect. Mapping of the transcription initiation sites of the d n a N gene In order to determine the transcription initiation site(s) of the dnaN gene within the preceding reading frame of the dnaA gene in wild-type bacteria, $1 nuclease mapping analysis of the dnaN transcripts was performed as described in Materials and methods. SI nuclease mapping was done using a 1,272 bp HinfI-PvuII D N A fragment 5' terminally labelled at the Hinfl site as specific hybridization probe for the dnaN transcripts (see Fig. 1). Total cellular R N A isolated from exponentially growing cells carrying pAQ31 or pLSK4 (a pBR322 derivative with the dnaA gene with its own promoters; C. Kiicherer and W. Messer, unpublished) was hybridized with an excess o f the specific dnaN probe. Following S1 nuclease digestion, the Sl-resistant fragments were visualized by gel electrophoresis and autoradiography. The results show (Fig. 2) that there are three different transcription initiation sites for the dnaN gene. The starts of these transcripts are at 325,350 and 490 nucleotides upstream of the 5' end-labelled HinfI site in the dnaN

122

Fig. 2A, B. Mapping of the transcription initiation sites of the dnaN gene. SI nuclease mapping of the dnaN transcripts was performed with a 1272bp HinfI-PvuII fragment labelled at the 5' end of the HinfI site as a dnaN-specific hybridization probe (Fig. 1) as described in Materials and methods. A Lane 1, ~bx t74/HaeIII standard; lane 2, dnaN probe; lane 3, tRNA control; lanes 4-6, RNA of strains carrying pAQ31; lane 7, RNA of bacteria carrying pLSK4. B Lane 1, ~bx 174/HaeIII standard; lane 2, dnaN probe; lane 3, tRNA control; lane 4, RNA of strain ABlI57 carrying pLSK4 isolated after growth at 30°C; lane 5, RNA of the coisogenic dam: :Tn9 strain isolated after growth at 42° C reading frame (position 2376 in the sequence of Hansen et al. 1982), the 490 nucleotide start site being the predominant one (P1). A promoter search in the relevant region of the dnaA sequence revealed that these three different transcription initiation sites correlate with three potential promoter structures with poor homology to the promoter consensus sequence at positions 1846-1876 (dnaNP1), 1990-2019 (dnaNP2) and 2021 2049 (dnaNP3) according to the sequence of Hansen et al. (1982). An additional 820 nucleotide protected fragment was not reproducibly observed. It probably does not represent a transcription initiation site, since there is no promoter consensus sequence in this region, and cloning of a 564 bp HaeIII fragment (positions 1249-1813) carrying the putative promoter into the SmaI site of pFD51 or pUTE13 did not produce a GalK + phenotype. Figure 2 also shows that the increased transcription of the dnaN gene in dam bacteria at 42 ° C, as shown by the galK fusion experiment, can also be visualized by the S1 mapping technique. The increase in dnaN transcription is based essentially on the induction of the principal promoter dnaNP1. In addition, transcription from the dnaA promoters in the pLSK4-carrying strain, seen as a full length protected probe fragment in Fig. 2, was reduced in dam strains, as described before (Braun etal. 1985; Kficherer etal. 1986). This dnaA transcript was also present, albeit in smaller amounts, in pAQ31-carrying bacteria, presumably originating from the chromosome. The full-length protected probe cannot be due to D N A - D N A hybridization, since

there was no such band in the tRNA control under the same hybridization conditions.

Discussion The dnaN gene of E. coli which encodes the fl-subunit of the D N A polymerase III holoenzyme (Burgers et al. 1981) lies within a gene cluster at 83 min of the E. coli genetic map (Bachmann 1983) which is involved in D N A replication. Gene cloning and sequencing have revealed that dnaN lies immediately downstream of dnaA (Sako and Sakakibara 1980; Hansen et al. 1982; Ohmori et al. 1984). On the basis of complementation analysis with specialized transducing lambda phages carrying the dnaA and dnaN genes it has been postulated that dnaN is cotranscribed with dnaA as an operon, because some deletions or Tn3 insertions in dnaA have a polar effect on dnaN transcription. Furthermore, a promoter region has been found only for dnaA and attempts to localize the promoter for dnaN have been unsuccessful (Sakakibara et al. 1981; Hansen et al. 1982; Ohmori et al. 1984). Therefore, it has been suggested these genes be designated the dnaA operon. In this report we have demonstrated the existence of at least three promoters for dnaN which reside within the reading frame of the preceding dnaA gene. The upstream dnaNP1 promoter seems to be the principal promoter compared with dnaNP2 and dnaNP3. The existence of dnaA transcripts extending into dnaN show that dnaN is a compo-

123 nent of the postulated dnaA operon. Therefore, the dnaN promoters must be responsible for the maintenance and/or enhancement of dnaN transcription when dnaA expression is decreased or repressed. We found conditions in which dnaA transcription was decreased bu~; dnaN transcription was simultaneously increased (see Table 3; Fig. 2). These results are the first evidence that although dnaA and dnaN seem to be an operon the regulatiol~, of these genes can also occur independently, thus, uncoupling dnaN transcription (and possible translation) from dnaA transcription. Furthermore, the increased dnaN expression in a dam background at 42 ° C shows clearly that dnaN expression is inducible. The inducibility in dam bacteria might be related to the presence of a G A T C site (recognition site for the dam methylase) within the - 10 region of the dnaNP1 promoter. M a n y E. coli promoters containing G A T C sites in the - 1 0 or - 3 5 region show altered expression in dam mutants, e.g. the expression of trpR is increased (Marinus 1985) while the expression of mioC (Schauzu et al. /987) and dnaA is reduced (Braun and Wright 1986; Kficherer et al. 1986). However, the mechanism(s) of this differential dependence o f promoter activity on methylation of G A T C sites is still not understood (Marinus 1987). In addition, dnaN transcription has been found to be induced by 2-aminopurine in contrast to that of dnaA (Quifiones et al. 1988). Two phenomenon have to be emphasized concerning gene organization in the dnaA region: first, with the exception of the dnaA gene itself, the promoter structures of all the genes of this region (gyrB, reeF, dnaN) reside entirely within the translated region of the respective preceding gene (Adachi et al. 1984; Menzel and Gellert 1987; Armengod and Lambies 1986; this paper) and the genes themselves are organized in a very compact fashion; thus dnaA and dnaN are separated by 4 bases (Hansen et al. 1982; Ohmori et al. 1984), dnaN and reeF overlap by 1 base (Blanar et al. 1984; Adachi et al. 1984) and there are 30 bases in the recF-gyrB intercistronic region (Adachi et al. 1984; Blanar et al. 1984; Yamagishi et al. 1986). Secondly, all genes except gyrB (which seems not to be a component of the postulated operon) contain a large leader region and dnaN possesses a very p o o r ribosome binding site, if any. This unusual gene organization might allow regulation of these genes independently of each other and also of dnaA regulation. The M M S operon in E. coli, which contains the genes rpsU (involved in translation), dnaG (coding for D N A primase) and rpoD (coding for the sigma-70 subunit of R N A polymerase) has a similar structure. This operon possesses three promoters but the distal gene rpoD contains in addition three promoters which reside entirely in the reading frame of the preceding dnaG gene and are independently regulated. One o f them is a heat shock promoter (Lupski and G o d s o n / 9 8 4 ) . After this manuscript was prepared we learnt of work by M. Armengod et al. (in press) on the transcriptional organization of dnaN and recF with very similar S1 mapping results for the transcriptional initiation sites of the dnaN promoters. Furthermore, they have presented evidence that dnaN and recF are organized in an operon. Since, as shown here, dnaA and dnaN are also organized in an operon their results support the hypothesis that the dnaA operon includes dnaN and recF which, however, permits individual control of its several components, thus leading to differential expressi6n and discoordinate regulation of the genes of the operon.

References

Adachi T, Mizuuchi K, Menzel R, Gellert M (1984) DNA sequence and transcription of the region upstream of the Escherichia coli gyrB gene. Nucleic Acids Res 12: 6389-6395 Adachi T, Mizuuchi M, Robinson E, Appella E, O'Dea M, Gellert M, Mizuuchi K (1987) DNA sequence of the E. coli gyrB: aplication of a new sequence strategy. Nucleic Acids Res 15:771-785 Armengod ME (1982) ReeF-dependent recombination as a SOS function. Biochimie 64: 629-632 Armengod ME, Lambies E (/986) Overlapping arrangement of the reeF and dnaN operons of E. coli; positive and negative control sequences. Gene 43 : 183 196 Atlund T, Clausen E, Hansen F (1985) Autoregulation of the dnaA gene of E. coli. Mol Gen Genet 200:442450 Bachmann B (1983) Linkage map of Escheriehia coli K-t2. Microbiol Rev 47:J80-230 Berk A, Sharp PA (1977) Sizing and mapping of early adenovirus mRNAs by gel electrophoresis of $1 endonuclease-digested hybrids. Cell 12:72-732 Blanar M, Sandler S, Armengod ME, Ream L, Clark A (1984) Molecular analysis of the reeF gene of E. coli. Proc Natl Acad Sci USA 81:46224626 Braun RE, Wright A (1986) DNA methylation differentially enhances the expression of one of the two E. coli dnaA promoters in vivo and in vitro. Mol Gen Genet 202:246-250 Braun RE, O'Day K, Wright A (/985) Autoregulation of the dnaA gene in E. coli. Cell 40:459-469 Brosius J, Care R, Perlmutter P (1982) Precise location of two promoters for the fl-lactamase gene of pBR322. J Biol Chem 257:9205-9210 Burgers PM, Kornberg A, Sakakibara Y (198/) The dnaN gene codes for the fl-subunit of DNA polymerase III holoenzyme of E. coli. Proc Natl Acad Sci USA 78:5391-5395 Chakraborty T, Yoshinaga K, Lother H, Messer W (1982) Purification of the E. coli dnaA gene product. EMBO J 1 : t 545-1549 Clark A, Volkert M, Margossian L (1979) A role for reeF in repair of UV damage to DNA. Cold Spring Harbor Symp Quant Biol 43 : 887-892 Clark A, Volkert M, Margossiau L, Nagaishi H (1982) Effects of a reeA operator mutation on mutant phenotypes conferred by lexA and reeF mutations. Mutat Res 106:/1-26 Drlica K (1984) Biology of bacterial DNA topoisomerases. Microbiol Rev 48 : 273-289 Fuller R, Kornberg A (1983) Purified dnaA protein in initiation of replication at the E. coli chromosomal origin of replication. Proc Natl Acad Sci USA 80:58/7-5821 Fuller R, Kaguni J, Kornberg A (198/) Enzymatic replication of the region of the E. coli chromosome. Proc Natl Acad Sci USA 78:7370 7374 Fuller R, Funnell B, Kornberg A (1984) The dnaA protein complex with the E. eoli chromosomal replication origin (oriC) and other sites. Cell 38 : 889-900 Fuunell B, Backer T, Kornberg A (1986) Complete enzymatic replication of plasmids containing the origin of the E. coli chromosome. J Biol Chem 261:5616-5624 Gellert M (/981) DNA topoisomerases. Annu Rev Biochem 50: 879-910 Hansen EB, Hansen FG, von Meyenburg K (1982) The nucleotide sequence of the dnaA gene and the first part of the dnaN gene of E. eoli. Nucleic Acids Res 10:7373 7385 Hirota Y, Ryter A, Jacob F (1968) Thermosensitive mutants of E. coli affected in the process of DNA synthesis and cellular division. Cold Spring Harbor Symp Quant Biol 33 : 677-693 Holmes D, Quigley M (198/) A rapid boiling method for the preparation of bacterial plasmids. Anal Biochem 114:193-197 Horii Z, Clak A (1973) Genetic analysis of the recF pathway to genetic recombination in E. coli K12: isolation and characterization of mutants. J Mol Biol 80:327-344 Kaguni JM, Kornberg A (1984) Replication initiated at the origin

124 (oriC) of the E. coli chromosome reconstitued with purified enzymes. Cell 38:183-190 Kornberg A, Backer T, Bertsch L, Bramhill D, Funnell B, Lasken R, Maki H, Maki S, Sekimizu K, Wahle E (1987) Enzymatic studies of replication of oriC plasmids. UCLA Symp Mol Cell Biol New Set 47:429-439 Kiicherer C, Lother H, K611ingR, Schauzu MA, Messer W (1986) Regulation of transcription of the chromosomal dnaA gene of E. coli. Mol Gen Genet 205 : 115-121 Lupski J, Godson N (1984) The rpsU-dnaG-rpoD macromolecular synthesis operon of E. coli. Cell 39:251 252 Maki H, Maki S, Lasken R, Kornberg A (1987) DNA polymerase III holoenzyme: subunits and functions. UCLA Syrup Mol Cell Biol New Ser 47: 429-439 Maniatis T, Fritsch EF, Sambrock J (1982) Molecular cloning A laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Marinus M (1985) DNA methylation influences trpR promoter activity in E. eoli. Mol Gen Genet 200:185-186 Marinus M (1987) DNA methylation in E. coli. Annu Rev Genet, in press Matsui M, Oka A, Takanami M, Yasuda S, Hirota Y (1985) Sites of dnaA protein-binding in the replication origin of the E. coli chromosome. J Mol Biol 184:529-533 McHenry C, Oberfelder R, Johanson K, Tomasiewicz H, Franden M (1987) Structure and mechanism of the DNA polymerase III holoenzyme. UCLA Syrup Mol Cell Biol New Ser 47:47-61 Meyenburg K von, Hansen FG, Rfise E, Bergmans H, Meijer M, Messer W (1979) Origin of replication. Cold Spring Harbor Syrup Quant Biol 43:121-128 Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Minton NP (1984) Improved plasmid vectors for the isolation of translational lac gene fusions. Gene 31:269-273 Ohmori H, Kimura M, Nagata T, Sakakibara Y (1984) Structural analysis of the dnaA and dnaN genes of E. eoli. Gene 28:159 170 Quifiones A, Kiicherer C, Piechocki R, Messer W (1987) Reduced transcription of the rnh gene in E. coli mutants expressing the SOS regulon constitutively. Mol Gen Genet 206:95-100

Quifiones A, Piechocki R, Messer W (1988) Expression of the Escherichia coli dnaQ (mutD) gene is inducible. Mol Gen Genet 211 : 106-112 Rak B, von Reutern M (1984) Insertion element JS5 contains a third gene. EMBO J 3:802811 Sakakibara Y, Mitzukami T (1980) A temperature sensitive E. coli mutant defective in DNA replication: dnaN a new gene adjacent to the dnaA gene. Mol Gen Genet 178:541 553 Sakakibara Y, Tsukano H, Sako T (1981) Organization and transcription of dnaA and dnaN genes of E. coll. Gene 13:47-55 Sako T, Sakakibara Y (1980) Coordinate expression of E. coli dnaA and dnaN genes. Mol Gen Genet 179:521-526 Schaaper R, Glickman B, Loeb LA (1982) Mutagenesis resulting from depurination is an SOS process. Mutat Res 106:1-9 Schauzu MA, Kficherer C, K611ingR, Messer W, Lother H (1987) Transcripts within the replication origin, oriC, of E. coll. Nucleic Acids Res 15:2479-2497 Silhavy T, Berman M, Enquist L (1984) Experiments with gene fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Thorns B, Wackernagel W (1987) Regulatory role of reeF in the SOS response of E. coil: Impaired induction of SOS genes by UV irradiation and nalidixic acid in a recF mutant. J Bacteriol 169:1731-1736 Wang TC, Smith K (1986) RecA suppression of recF deficiency in the postreplication repair of UV-irradiated E. coli K-12. J Bacteriol 168:940-946 Wood RD, Stein J (1986) Role of the recF gene product in UV mutagenesis of lambda phage. Mol Gen Genet 204:82-84 Yamagishi J, Yoshida H, Yamayoshi M, Makamura S (1986) Nalidixic acid resistant mutation in the gyrB gene of E. coli. Mol Gen Genet 204: 367-373

C o m m u n i c a t e d by H. B6hme

Received February 1, 1988