Gene. 113 (1992) 101-106 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0378-1119/92/$05.00

101

GENE 06388

Replication and copy number control of the broad-host-range plasmid RSFI010 (Recombinant DNA; gene expression; promoters; conjugative DNA transfer; D N A primase)

Joachim Frey a, Mira M. Bagdasarianb and Michael Bagdasarianb "hlstitutefor Veterinary Bacteriology, Universityof Bern. Laenggasstr. 122, CH-3012 Bern (Switzerland) Tel. (41-31)27-44-84; Fax (41-31)2469-22 and h Department of Microbiology. Michigan State University, East Lansing, MI 48824 (U.S.A.) Received by A.M. Chakrabarty: 4 December 1991 Accepted: 12 December 1991 Received at publishers: 20 January 1992

SUMMARY Initiation of replication of the broad-host-range plasmid RSF1010, is accurately controlled by the plasmid-encoded proteins, RepB (MobA), RepB', RepA and RepC [Haring et al., Proc. Natl. Acad. Sci. USA 82 (1985) 6090-6094; Scherzinger et ai., Nucleic Acids Res. 19 (1991) 1203-1211]. The genes encoding these proteins which are essential for replication and conjugative mobilization are transcribed from a cluster of promoters, PI/P3 and P2, which partly overlap with the origin of conjugal transfer, oriT. Three regions were found where deletion mutations affect the mobilization of RSF1010 and increase its copy number in Escherichia coll. A deletion in the mobC gene increased the copy number of RSF1010 four-fold. Another deletion, that removed oriT and part of the promoter believed to be responsible for the expression of mobC, results in a three-fold increase in copy number. The third type of deletions affect the N-terminal part of RepB (MobA). A deletion that created a frame-shift results in a three-fold increase in copy number. A smaller, in-frame deletion of this region only affected the mobilization of RSF1010, but not its copy number. The extent by which RSF1010 or its deletion derivatives could repress the PI/P3 and P2 promoters has indicated that these promoters are negatively regulated by MobC and RepB (MobA), presumably by their attachment to the oriT region of RSFI010. Both MobC and RepB are required for the maximal repression of the rep operon. Optimal function of RSF1010 thus involves not only overlapping genes, but also proteins that exert multiple functions, mobilization, replication and regulation.

INTRODUCTION RSFI010 is a mobilizable, but not self transmissible plasmid of the IncQ group which has the remarkable ca-

Correspondence to: Dr. M. Bagdasarian, Department of Microbiology, S110 Plant Research Laboratory, Michigan State University, East Lansing, MI 48824 (U.S.A.) Tel. (517)353-8619; Fax (517)353-1926.

Abbreviations: bp, base pair(s); cac (Cac), see Fig. 2; ,4, deletion; EtdBr, ethidium bromide; IncQ, incompatibility group Q; kb, kilobase(s) or 1000bp; LB, Luria-Bertani (medium); nt, nucleotide(s); oriT, origin of conjugational transfer; oriV,originof vegetativeDNA replication;P, promoter; PCR, polymerasechain reaction; Sm, streptomycin;Tn, transposon; UV, ultraviolet; wt, wild type; [ ], denotes plasmid-carrier state; ::, novel joint (fusion or insertion).

pability to replicate in a broad range of bacterial hosts, including most of the Gram- bacteria (for review, see Frey and Bagdasarian, 1989) and at least some of the Gram + species (Gormley and Davies, 1991). In E. coil, RSF1010 is present at a copy number of 12 per cell (Bagdasarian et al., 1986). Previous work has identified and mapped oriV, the unique origin of vegetative DNA replication (De Graaf et al., 1978; Scherzinger et al,, 1984; Haring and Scherzinger, 1989) repA, repB, repB' and repC, the genes for essential replication proteins (Scherzinger et al., 1984; 1991; Scholz et al., 1984; Haring et al., 1985), oriT, the site of the relaxation complex and origin of conjugative DNA transfer, mobA (identical to repB), mobB and mobC, genes encoding trans-active proteins involved in plasmid mobilization (Nordheim et al., 1980; Bagdasarian et al., 1982;

102 AsnVa•LysA•aLeuI•eMetA•aG•yVa•LeuVa•LysArgArgThrG•uAspLysArgG•uGlnG•nG•nG•nArgA•aArgG•uArgG•nI•eG•uA 2900 2920 2940 2960 2980

GTTCACCTTGGCCAAAATCATGGCCCCCACCAGCACCTTGCGCCTTGTTTCGTTCTTGCGCTCTTGCTGCTGTTCCCTTGCCCGCTCCCGCTGAATTTCG A 201 laAsnIleArgAlaArgGlnGluGluLeuLysAlaLeuArgAspAlaAlaLysAsnSerGlyLysValMET HinFI 3000 3020 r P1 -351

(-- m o b C

......

•

F

F P1 -i0 -35 1

3--->

P3

GCATTGATTCGCGCTCGTTGTTCTTCGAGCTTGGCCAGCCGATCCGCCGCCTTGTTGCTCCCCTTAACCATCTTGACACCCCATTGTTAATGTGCTGTCT r P3 -101 --> 3120 3140 3160 3180 CGTAGG~TAT~ATGGAGG~A~AG~GG~GGCAAT~C~GA~TA~TTTGTAGGGGAG~CG~A~TTA~GGTTT~TCTT~GAGAAA~TGGCCTAA~GGC~A (-.-t P2 -i0J UP2 -35 J 13 ..............................................

3200

oziT

..........................................

3220

3240

3260

3280

CCCTTCGGGCGGTGCGCTCT~CGAGGG~CATTGCATGGAGC~GAAAA~A~AA~AG~GA~G~AGCAT~GCGATTTAT~A~TTA~GG~GAAAAC~G A 13 repB -~ METAlaIleTyrHisLeuThrAlaLysThrG 3300

3320

3340

3360

HinFI

GCAG~AGGT6GGG~GG~CAAT~GG~AGGG~AAGG~GA~TA~ATC~AG~G~GAAGG~AAGTATG~CG~GA~ATGGATGAAGTCTTGCA~G~GAAT~ •y•erArgserG•yG•yG•nserA•aArgA•aLysA•aAspTyrI•eG•nArgG•uG•yLysTyrA•aArgAspMetAspG•uVa•LeuHisA•aG•uSe ACCCC

,

A 20

3400

3420

3440

3460

3480

CGGG~A~AT&~GGAGTT~GT~GAG~GG~6~G~GACTA6TGGGATG~TG~GA~TGTATGAA~G~GCCAATGGG~GGdTGTT~AAGGAGGT~GAATTT •G•yHisMet•r•G•u•heVa•G1uArg•r•A•aAspTyrTrpAspA•aA1aAspLeuTyrG•uArgA•aAsnG•yArgLeuPheLysG•uVa•G•uPhe A 2O 3500

3520

3540

3560

3580

GCCCTGCCGGTCGAGCTGACCCTCGACCAGCAGAAGGCGCTGGCGTCCGAGTTCGCCCAGCACCTGACCGGTGCCGAGCGCCTGCCGTATACGCTGGCCA

A~aLeuPr~va~G~uLeuThrLeuAspG~nG~nLysA~aLeuA~a~erG~uPheA~aG~nHisLeuThrG~yA~aG~uArgLeuPr~'~rThrLeuA~aI & 20 3600

3620

3640

3660

3680

TCCATGCCGGTGG~GGCGAGAAC~CGCACTGCCACCTGATGATCTC~GAGCGGATCAATGA~GGCATCGAGCGGCCCGCCGCTCAGTGGTTCAAGCGGTA •eHisA•aG•yG•yG•yG•uAsnpr•HisCysHisLeuMetI•eSerG•uArgI•eAsnAspG•yI•eG•uArgPr•A•aA•aG•nTrpPheLysArgTy & 20 3700

3720

3740

3760

3780

CAACGGCAAGACCCCGGAGAAGGGCGGGGCACAGAAGACCGAAGCGCTCAAGCCCAAGGCATGGCTTGAGCAGACCCGCGAGGCATGGGCCGACCATGCC •AsnG•yLysThr•r•G•uLysG•yG•yA•aG•nLysThrG•uA•aLeuLys•r•LysA•aTrpLeuG•uG1nThrArgG•uA1aTrpA•aAspHisA1a A 2O 3800

3820

384

3860

3880

~u~CCGGGCATTAGAGCGGGCTGGCCACGACGCCCGCATTGACCACAG~CACTTGAGGCGCAGGGCATCGAGCGCCTGCCCGGTGTTCACCTGGGGCCGA AsnArgA~aLeuG~uArgA~aG~yHisAspA~aArgI~eAspHisArgThrLeuG~uA~aGlnG~yI~eG~uArgLeuPr~G~yva~HisLeuG~yPr~A ~ 20 3900

3920

3940

3960

3980

ACGTGGTGGAGATGGAAGGCCGGGGCATCCGCACCGACCGGGCAGACGTGGCCCTGAACATCGACACCGCCAACGCCCAGATCATCGA~TACAGGAATA snVa~Va~G~uMetG~uG~yArgG~yI~eArgThrAspArgA~aAsDVa~A~aLeuAsnI~eAspThrA~aAsnA~aG~nI~eI~eAspLeuG~nG~uTy

A 20 4000

4020

4040

4060

4080

CCGGGAGGCAATAGACCATGAACGCAATCGACAGAGTGAAGAAATCCAGAGGCATCAACGAGTTAGCGGAGCAGATCGAACCGCTGGCC~AGAGCATGGC rArgG•uA•aI•eAspHisG•uArgAsnArgG•nSerG•uG•uI•eG•nArgHisG•nArgValSerGlyA•aAspArgThrA•aG•yPr•G•uHisG•y

A 20 4100

4120

4140

4160

4180

GACACTGGCCGACG~GCCCGGCAGGTCATGAGCCAGACC~GCAGGCCAGCGAGGCGCAGGCGGCGGAGTGGC'~2%GCCCAGCGCC~GACAGGGGCG AspThrG~yArgArg~erPr~A~aG~yHisG~uPr~AspG~nA~aG~yG~r~rgG~yA~aG~yG~yG~yva~A~aG~u~erPr~A~aPr~AspArgG~yG A 20

& 18

4200

4220

4240

4260

4280

GCATGGGTGGAGCTGGCC~GAGTTGCGGGAGGTAGCCGCCGAGGTGAGCAGCGCCGC~CAGAGCGCCCGGAGCGCGTCGCGGGGGTG~CACTGG~G( ~yMetG~yG~A~aG~yG~nArgVa~A~aG~yG~yserArgArgG~yG~uG~nArgArgA~aG~uArgPr~G~uArgVa~A~aG~yva~A~aLeuG~uA~ 4300

4320

4340

4360

4380

TA'~GCT/~CCG'Ix3ATGCTGGC'P~`CCATGA~CCTACGGTGGTGCTGCTGATCGCATCGTTGcTCTTGCTCGACCTGACGCcACTGAC/~CCGAGGACGG aMetA~aAsnArgAspA~aG~yPheHisAspA~aTyrG~yG~yA~aA~aAspArgI~eVa~A~aLeuA~aArgPr~AspA~aThrAspAsnArgG~yArg A 18 rRBS1 4420 4440 CTCGATCTGGCTGCGCTTGGTGGCCCGATGAAGAACGACAGGACTTTGC~GGCCATAGGCCGACAGCTCA

LeuAspLeuAlaAlaLeuGlyGlyProMETLysAsnAspArgThrLeuGlnAlaileGlyArgGlnLeu

Fig. 1.

repB"

--)

103 0

1

;~

3

4

5

6

7

8

8.684 kb

I

I

I

1

I

I

I_

I

1

l

~pl

-~ P4 3

\

P6 P5

\

°1 Fig. 2. Physical and genetic map of the plasmid RSFI010. The map is linearized at the unique Hpa! site. Genes encoding identified products are indicated by open boxes and functional sites by hatched boxes. E, is a gene of unknown function; cac (control of A and C genes) encodes a repressor of P4 promoter (Scholz et al., 1989; Maeser et al., 1990). Promoters are indicated by P with numbers. Wavy lines represent transcripts of the replication and mobilization operons. Dash-dot lines indicate negative regulatory function.

Derbyshire et al., 1987), as well as the sulfonamide and Sm resistance genes sui and str (Scholz et al., 1989). Expression of the rep genes, in particular repC, the gene for the initiator protein, was shown to regulate the extent of replication at oriV (Hating et al., 1985). It was suggested that transcription of the replication genes, repB (mobA), repB', encoding plasmid specific D N A primases, as well as the transcription of repA and repC, encoding a helicas¢ and an oriV.binding protein, respectively, was initiated from a site upstream from the repB gene, located 3 kb from the unique Hpal site of the plasmid. In addition, transcription of the repA and repC genes may be initiated from a site preceding the repA gene (Bagdasarian et al., 1986; Maeser et al., 1990). Translation of repB' gene proceeds in the same reading frame as that of the repB but is initiated at nt 4408. Both RepB proteins exhibit D N A primase activities (Hating and Scherzinger, 1989; Honda et al., 1991). Transcription promoters on RS F l010 physical map were recognized initially by electron microscopy visualizing the

E. coli RNA polymerase binding sites (Bagdasarian et al., 1981). After the complete nt sequence of the plasmid became available, E. cog promoter consensus sequences were identified at the sites previously located by electron microscopy (Scholz et ai., 1989). In the region nt 3040-3120, three overlapping promoter sequences may be recognized (Figs. 1 and 2). They may be responsible for the transcription of the replication genes repB (mobA) and repB' as well as genes repA and repC and a small gene mobC transcribed in the opposite orientation (Scholz et al., 1989; Hating and Scherzinger, 1989). We have analyzed plasmid-borne factors that repress promoters PI/P3 and P2 thus providing negative regulation of replication and copy number of RSF1010. The results have indicated that RSF1010 has evolved mechanisms of regulation that ensure not only the economy of space, by taking advantage of overlapping genes, but also the economy of function by taking advantage of multifunctiona[ proteins and specific arrangement of functional sites on its genome.

Fig. 1. Physical aud genetic map of the PI/P2/P3 region of RSFI010. The nt are numbered starting at the unique Hpal site (see Fig. 2) of the plasmid (Schoiz et al., 1989). Translation of r,,pB and repB' is indicated below and that of mobCabove the nt sequence. Promoter sequences are indicated by boxes and their functional orientation by arrows. Deletions are marked by solid lines and oriT by dotted line below the nt sequence. RBS, ribosomebinding site. The nt sequence of the deletion derivativeswas determined by dideoxychain-termination method (Sanger ¢t al., 1977)using double-stranded plasmid DNA and synthetic oligodeoxynucleotideprimers corresponding to the following coordinates on the RSFI010 sequence: nt 3234-3251, nt 3490-3509 and nt 3766-3783.

104 EXPERIMENTAL AND DISCUSSION

(a) Deletions in the rep and mob regions The mobilization defective deletion mutants, A18 and A20, of RSF1010 were obtained by BAL 31 digestion of a Tn3 d :rivative, pKT260, linearized at its unique B a m H I site (Bagdasarian et al., 1982). Deletion d13 was a spontaneous deletion of RSF1010 selected as conferring increased Sm resistance to the host. The in vitro generated deletion A201 was obtained by removing the 3' recessed ends, in a partial BstXI digest of RSF1010 DNA, by T4 DNA polymerase and religating the blunt ends. The exact location of the deletions A 13, A 18 and d20 was determined by nt sequence analysis (Fig. 1). Deletion A13 is 124 bp long and extends over the oriT and the -35 region of the P2 promoter. Deletion d 18 is an in-frame deletion of 267 bp. It is entirely within the first third of the repB gene and does not extend into the repB' coding sequence. Deletion A20 is 809 bp long. It contains a 4 bp insertion at the start and extends through most of the N-terminal part of the repB gene. Since this deletion confers a frame-shift to the reading frame of the repB gene, RSF 1010/120 derivatives are assumed to produce no functional RepB primase protein which is essential for conjugative replication. The RepB' primase should still be produced from its own start codon at nt 4408 (Fig. 1) and be able to provide essential priming functions for vegetative replication initiating at oriV. As expected (Bagdasarian et al., 1982; Derbyshire et al., 1987), all deletion mutants described here were defective in their conjugational mobility function. Deletions d 13, d 18 and /120 had a frequency of mobilization 4.0x 10 -t', 1.7 x 10 -t' and 9.0 x 10 -8, respectively, whereas the wt RSFI010 was mobilized at frequencies of 10- i when RK2/ RP1/RP4 plasmid was used as helper. Deletion/1201 generates a frame-shift and is thus expected to produce a truncated protein MobC. Mobilization frequency of RSF 1010 /1201 was 2.0 x 1 0 - 3 It is not known at present whether the truncated product of the mobCd201 gene has any residual activity. Each of the deletion derivatives described above was a functional plasmid, stably maintained in E. coil bacteria. This confirms previous findings that indicated the ability of protein RepB' to supply the essential primase function for the replication starting from the RSF 1010 oriV (Scherzinger et al., 1984; Honda et al., 1991) and supports our conclusion that RepB' protein is indeed produced, and can exert its DNA priming function, in rive.

(b) Copy number of RSFI010 deletion mutants Copy number of the RSF1010 derivatives was determined by comparison of the intensities of plasmid D N A bands, resolved by agarose gel electrophoresis, stained with EtdBr and photographed in the UV light, with the inten-

sities of D N A bands of coresident plasmids pSC101 and pACYC177. As shown in Table I, derivatives of RSF1010 having deletions in the repB gene and in the mobC gene had copy numbers three- to four-fold higher than the wt plasmid. The copy number of deletion plasmid A201 could be reduced to wt R S F 1010 level if the mobC gene product was provided in trans by the coresident plasmid pJFF162. The mutant R S F 1010A 18 although mobilization-defective retained the wt copy number. It should be noted, however, that the deletion A 18 is a relatively small deletion that does not alter the reading frame of the gene repB. The properties of this mutant could, therefore, be the result of production of a truncated RepB protein, inactive as mobilization factor, but still active as represser. Previous results have indicated that the cellular concentration of protein RepC is mainly responsible for the regulation of RSF1010 copy number (Haring et al., 1985). It has also been shown that a transcript long enough to comprise all three rep genes is produced in vivo by RSFI010 TABLE I Copy numbers of the RSFI010 deletion derivatives RSF 1010 derivative~

RSFI010wt RSFI010AI8 RSFI010zlI3 RSFI010A20 RSFI010A201

Copy number with coresident plasmidt, None

pACYCI84 pJFF162~ (pACYC184::mobC÷)

12 13 38 36 52

12 13 38 36 52

12 13 38 37 l0

Extent of deletions is shown in Fig. 1. b M e t h o d s . Determination of plasmid copy number: plasmid DNA was extracted from E. coliC600[pSC 10! ], containing the RS F 1010 derivative indicated (Holmes and Quigley, 1981),digested with Pstl+Pvull, resolved by electrophoresis in 0.7% agarose gels and stained with EtdBr. The gels were photographed in UV light and the intensities of the largest RSFI010 fragment band and of the pSCI01 linear band were measured on the negatives with a Joyce Loebl Microdensitometer and Nummonics Integrator. Copynumbers were determined by a linear comparison of the band intensities (after correction for the size differences) using a copy number of 5.3 copies per cell (8 copies per chromosome equivalent) for pSCI01 (Hasunuma and Sekiguchi, 1977; Frey et al., 1979). Control experiments with E. coli C600[pSCI01, pACYCI77] and C600[pSCI01, pACYC177, RSFI010] were made to show that RSFI010 and pSCI01 did not mutually affect their copy numbers if present in the same cell (Frey et al., 1979). Plasmid pJFF162 was constructed by PCR amplification of the RSFI010 region between nt 2757 and 3063, comprising the entire mobC gene and its ribosome-binding site. Primers were designed such that Him dill and BamHl sites were included at the ends of the synthesized fragment. The fragment was then inserted between the Hindlll and BamHl sites of the vector pACYCI84 (Chang and Cohen, 1978). The resulting recombinant plasmid pJ FF 162, contained the wt mobCgene inserted such that it could be expressed from the promoter of the kanamycin-resistance gene. ~'

105 (Bagdasarian et al., 1986). Moreover, it has been shown that deletions which remove the promoter P4 (Fig. 2) yield viable plasmids (unpublished). This indicated that the transcription of the entire repBB'AC operon from the PI/P3 promoters can satisfy the replication requirements of RSF 1010. The results in Table I are interpreted, therefore, as indication that proteins RepB and MobC are involved in the repression of the rep operon of RSF1010. Deletion RSF1010 A13 does not extend into any known coding sequence. It may, however, affect the expression of the mobC gene by deleting the -35 region of the P2 promoter. Alternatively, it may damage a site to which proteins MobC or RepB attach to repress the PI/P3 promoters. The oriT, which is entirely removed by the deletion 313, seems the most likely site for the attachment of proteins MobC and RepB involved in conjugal mobilization (Fig. 1).

(c) Regulation of the PI/P2/P3 promoter functions To estimate the activity and the function of PI/P2/P3 promoter cluster we have inserted a Hinfl DNA fragment of RS F I 010 (nt 2985-3376), in both orientations, upstream from a promoterless lacZ gene in the plasmid pJL260, a derivative of pGA39 (pACYC184-based plasmid; An and Friesen, 1979) containing the entire lac operon without its promoters (S. Molin, personal communication). The expression of lacZ was then measured in the cells containing the resultant recombinant plasmids and in the cells containing, in addition, a coresident derivative of RS F 1010. As shown in Table II the Hinfl fragment containing the PI/ P2/P3 sequences could provide promoter functions in both orientations. The presence of the wt RSFI010 as coresident plasmid repressed the function of the promoter clusTABLE !I Repression of the PI/P3 and P2 promoters by deletion derivatives of plasmid RSFI010 Co-resident plasmid '~

None RSFI010wt RSFI010AI8 RSFI010A 13 RSFI010A20 RSFI010A201

pJFF162(mobC ±)

p-galactosidase units h expressed from:

PI/P3

P2

1321 + 134 19 + 3 92 + 13 159 + 17 266 + 53 664 + 87 908 + 91

610+42 17 + 4 65 +4 80 + 8 150 + 15 334 + 17 646 + 41

" E. coil araD139 d(ara-leu)7697 d(iacX74-phoA)20 gale galK thi rpsL ~lmB argE recA l[pJL260::Hinfl(nt 2985-3376) of RSF1010 in the orientation indicated] (see also Figs. 1 and 2), was tranformed with the D N A of individual RSF1010 deletion derivative plasmids. The transformants were grown in LB to a mid log phase and used for enzyme activity determination. b The activity offl-galactosidase was determined in the cells and the units were defined as described by Miller (1972).

ter in both directions, towards the rep operon and towards the mobC gene. RSFI010 derivative/1201 could only reduce the expression of either promoter by approx. 50% indicating that MobC protein was an important component of the repressor. However, the presence of plasmid pJFFI62, containing the mobC + gene inserted in the vector pACYC184, was not able to repress the promoters in a significant way, showing that MobC protein alone is not sufficient to achieve repression. Deletion derivative RSF1010/120, which should only produce one functional primase, the RepB' protein, could exert a partial repression of the promoters in either orientation. On the other hand, RSF1010 A18, which should produce a truncated RepB protein, in addition to RepB', was able to repress almost as well as the wt RSF1010. These results are interpreted to indicate that both MobC and RepB proteins are essential for maximal repression at the PI/P2/P3 promoter clusters. They also correlate well with the results presented in Table I showing that /120 and d201 have a higher copy number whereas d18 maintains a wt number of copies in the cell. It is rea.sonable to suggest, therefore, that proteins MobC and RepB repress the functions of the P1/P2/P3 promoter cluster by attaching to this region, most probably to the oriT region of the RSF1010 DNA. Whether these proteins attach independently to their respective domains of the oriT or form a protein-protein complex before or after attachment is at present unknown. There seemingly is a discrepancy between the elevated copy number of the deletion RSF1010 A13 and its ability to exert considerable repression of the rep promoters functions in both directions, if this repression is attributed to the decreased production of protein MobC by this plasmid. We interpret the properties of A 13, therefore, as follows. In the wt RSFI010 repression by the MobC and RepB is exerted by attachment of these proteins to the oriT region. In d 13, the entire oriT region is deleted, the target for the repressors is, therefore, missing, but the promoters PI and P3 are intact. This plasmid is thus derepressed for the rep operon, but is able to provide the MobC and RepB proteins for the repression in trans. Since at present we do not have a way to determine directly the amounts of Mob proteins in the cell, we cannot be certain that AI3 produces less of the protein MobC than does the wt RSF1010. The results of Table II, however, are consistent with this interpretation. It has been demonstrated by Maeser et al. (1990) that protein Cac of RSFI010 can specifically attach to the region of the P4 promoter. It is thus reasonable to believe that the gene cluster repA/C is transcribed from, and negatively regulated at, this promoter. However, as we have demonstrated by deletion mapping (to be published), promoter P4 is not essential for the independent replication of RSF 1010. We have to conclude, therefore, that two transcription and regulation units for the rep genes exist on this plasmid, one

106 comprising the repA/C cluster and one spanning the entire repBB' A C operon.

(d) Conclusions (1) The operon comprising the replication genes of the broad-host-range plasmid RS F 1010 is transcribed and regulated at two distinct sites. O n e of them is the PI/P2/P3 promoter cluster, responsible for the expression of the repBB'AC operon and the mobC gene (Fig. 2). In addition, the repAC operon m a y be transcribed from, and regulated at, the P4 p ro m o ter (Maeser et al., 1990). (2) Promoters PI/P2/P3 are negatively regulated by M o b C and RepB proteins presumably by attachment of these M o b proteins to the oriT region of RSF1010. (3) Both M o b C and RepB proteins are required for optimal regulation of the repBB'AC operon,

ACKNOWLEDG EM ENTS We are grateful to S. Molin for the gift of the pJL260 plasmid and to E. Scherzinger for numerous helpful discussions. This work was supported by grants from the Swiss National Science Foundation (No. 3.627.087) to J.F., U.S. D e p a r t m e n t of Agriculture (No. 89-01053) and Research Excellence Fund from the State of Michigan (to M.B.).

REFERENCES An, G. and Friescn, J.D.: Plasmid vehicles tbr direct cloning of Escherichia coil promoters. J. Bacteriol. 140 (1979) 400-407. Bagdasarian, M,, Lurz, R., Rtlckert, B., Franklin, F.C.H., Bagdasarian, M.M. and Timmis, K.N.: Specific-purpose plasmid cloning vectors, II. Broad host range, high copy number, RSFl010,derivcd vectors, and a host-vector system for gene cloning in Pseudomonas. Gene 16 (1981) 237-247, Bagdasarian, M,, Bagdasarian, M.M., Lurz, R., Nordhcim. A., Frey, J. and Timmis, K.N.: Molecular and functional analysis of the broadhost-range plasmid RSFI010 and construction of vectors for gene cloning in Gram-negative bacteria, In: Mitsuhashi, S. (Ed.), Bacterial Drug Resistance. Japan Sci. Soc, Press, Tokyo, 1982, pp. 183-197. Bagdasarian, M.M,, Scholz, P,, Frcy, J. and Bagdasarian, M.: Regulation of the rep operon expression of the broad-host-range plasmid RSFI010. In: Novick, R. and Levy, S, (Eds.), Evolution and Environmental Spread of Antibiotic Resistance Genes. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1986, pp. 209-223, Chang, A.C.Y.C. and Cohen, S.N.: Construction and amplification of amplifiable multicopy DNA cloning vehicles derived from the P 15A cryptic miniplasmid, J, Bacteriol, 134 (1978) 1141-1156. Cohen, S,N. and Chang, A.C.Y.C.: Revised interpretation of the origin of the pSCl01 plasmid. J. Bacteriol. 132 (1977) 734-737.

De Graaf, J., Crossa, J.H., Hefron, F. and Falkow, S.: Replication of the non-conjugative plasmid RSFI010 in Escherlchia coli K-12. J. Bacteriol. 134 (1978) 1117-1122. Derbyshire, K.M., Hatfull, G. and Willetts, N.P.: Mobilization of the non-conjugative plasmid RSFI010: a genetic and DNA sequence analysis of the mobilization region. Mol. Gee. Genet, 206 (1987) 161-168. Frey, J. and Bagdasarian, M: The molecular biology of IneQ plasmids. In: Thomas, C.M. (Ed.), Promiscuous Plasmids of Gram Negative Bacteria. Academic Press, London, 1989, pp. 79-94. Frey, J., Chandler, M. and Caro, L.: The effects of an Escherichia coil dnaAts mutation on the replication of the plasmids ColEI, pSCI01, RI00.1 and RTF-TC. Moi. Gen. Genet. 154 (1979) 117-126. Gormley, E.P. and Davies, J.: Transfer of plasmid RSFI010 by conjugation from Escherichiacoilto Strepmmj'ceslividansand Mycobacterium smegmatL¢.J, Bacteriol. 173 (1991) 6705-6708. Haring, V. and Scherzinger, E.: Replication proteins of the IncQ plasmid RSFi010. In: Thomas, C.M. (Ed.), Promiscuous Plasmids of Gram Negative Bacteria. Academic Press, London, 1989, pp. 95-124. Haring, V., Scholz, P., Scherzinger, E., Frey, J., Derbyshire, K., Hatfull, G., Willets, N.S. and Bagdasarian, M.: Protein RepC is involved in copy number control ofthe broad-host-range plasmid RSFI010. Proc. Natl. Acad. Sci, USA 82 (1985) 6090-6094. Holmes, D.S. and Quigley, M.: A rapid boiling method for the preparao tion of bacterial plasmids. Anal. Biochem. !!4 (1981) 193-197. Honda, Y., Sakai, H., Hiasa, H., Tanaka, K., Komano, T. and BagdasarJan, M.: Functional division and reconstruction of a plasmid replication origin: molecular dissection of the oriV of the broad-host-range plasmid RSFIOI0. Proc. Natl. Acad, Sci. USA 88 (1991) 179-183. Maeser, S., Scholz, P., Otto, S. and Scherzingcr, E,: Gene F of plasmid RSFi010 codes for a low-molecular-weight reprcssor protein that autorcgulates expression of the repAC operon. Nucleic Acids Res. 18 (1990) 6215-6222. Miller, L: Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1972, pp. 352-355. Nordheim, A., Hashimoto-Gotoh, T. and Timmis, K.N,: Location of two relaxation nick sites in R6K and single sites in pSCI01 and RSFI010 close to origins of vegetative replication: implication for conjugal transfer of plasmid deoxyribonucleic acid. J. Bacteriol. 144 (INS0) 923-932. Sanger, F., Nicklen, S. and Coulson A.R.: DNA sequencing with chainterminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 54636467. Scherzingcr, E., Bagdasarian, M.M., Scholz, P., Lurz, R. Rtlckcrt, B. and Bagdasarian, M.: Replication of the broad-host-range plasmid RSFi010: requirement for three plasmid-encoded proteins, Proc. Natl. Acad. Sci. USA 81 (1984) 654-658. Scherzinger, E,, Haring, V., Lurz, R. and Otto, S.: Plasmid RSFI010 DNA replication in vitro promoted by purified RSFI010 RepA, RepB and RcpC proteins. Nucleic Acids Res. 19 (1991) 1203-1211. Scholz, P., Haring, V., Scherzinger, E., Lurz, R., Bagdasarian, M.M., Schuster, H, and Bagdasarian, M.: Replication determinants of the broad-host-range plasmid RSFI010. In: Helinski, D.R., Cohen, S.N., Clewell, D., Jackson, D.A. and Hollaender, A. (Eds.), Plasmids in Bacteria. Plenum Press, New York, 1984, pp. 243-259. Scholz, P., Haring, V., Wittmann-Liebold, B., Ashman, K., Bagdasarian, M. and Scherzinger, E.: Complete nucleotide sequence and gene organization of the broad-host-range plasmid RS F 1010. Gent 75 (1989) 271-288.