Vol. 172, No. 10JOURNAL OF BACTERIOLOGY, Oct. 1990, p. 5697-57050021-9193/90/105697-09$02.00/0Copyright X 1990, American Society for Microbiology

Characterization of the Minimum Replicon of the Broad-Host-RangePlasmid pTF-FC2 and Similarity between pTF-FC2 and

the IncQ PlasmidsROSEMARY A. DORRINGTON AND DOUGLAS E. RAWLINGS*

Department of Microbiology, University of Cape Town, Private Bag, Rondebosch 7700, South Africa

Received 12 March 1990/Accepted 16 July 1990

The nucleotide sequence of a 3,202-base-pair fragment which contained the minimum region required forreplication of the broad-host-range plasmid, pTF-FC2, has been determined. At least five open reading framesand a region that affected the host range were identified. Proteins corresponding in size and location to four ofthe five open reading frames were produced in an in vitro transcription-translation system. The predictedamino acid sequences of two of the proteins were aligned with those of the RepA and RepC proteins of thebroad-host-range IncQ plasmid RSF1010 and found to be 43 and 60% homologous, respectively. Despite thissimilarity, neither the RepA nor the RepC protein of the IncQ plasmid was able to complement mutations inthe pTF-FC2 repA and repC genes. Although there was a considerable amount of DNA homology betweenpTF-FC2 and RSF1010 in the oriV region and the region coding for the RepA and RepC proteins, no otherhomology between the two plasmids at either the DNA or protein level could be detected.

Plasmid pTF-FC2 is a 12.6-kilobase-pair, cryptic, broad-host-range plasmid that was originally isolated from theindustrially important bacterium Thiobacillus ferrooxidans.The plasmid was cloned into the vector pBR325 and shownto replicate in Escherichia coli from an origin located on thepTF-FC2 portion of the recombinant plasmid (23). The hostrange of pTF-FC2 includes all the gram-negative bacteriathat have been tested so far, namely, Pseudomonas aerugi-nosa (24), Thiobacillus novellus (25), Klebsiella pneumo-niae, Rhizobium meliloti, Agrobacterium tumefaciens, andSalmonella typhimurium (D. E. Rawlings and R. A. Dor-rington, unpublished results). A 329-base-pair (bp) region ofthe pTF-FC2 DNA that is required for replication in cis hasbeen identified and sequenced (7).

Relatively few plasmids capable of replication in a largenumber of gram-negative bacteria have been identified. Ofthese, the most extensively studied are those belonging tothe IncQ incompatibility group, which includes the similar oridentical plasmids RSF1010 (8), R1162 (16), and R300B (3).The entire nucleotide sequence (8,685 bp) of the RSF1010has recently been published (28).

Plasmid pTF-FC2 is similar to the IncQ plasmids inseveral respects. The oriV region of pTF-FC2 includes three22-bp tandem, perfectly repeated DNA sequences (7),whereas the corresponding region of the IncQ plasmids hasthree-and-one-half 20-bp perfectly repeated DNA sequencesseparated by 2-bp spacer sequences (17, 18). The repeatedsequences of pTF-FC2 have a 60% DNA homology withthose of the IncQ plasmids, whereas the 115-bp sequenceimmediately adjacent to the repeat sequences is 75% homol-ogous (7). Both plasmids are maintained in E. coli at a copynumber of approximately 12 per chromosome (3, 7) and aremobilized at high frequency between gram-negative bacteriaby IncP plasmids such as RP4 (1, 25, 26).

Despite these similarities, no incompatibility betweenpTF-FC2 and the IncQ plasmid R300B was detected (7). Thestructural similarity between the origins of pTF-FC2 and the

* Corresponding author.

IncQ plasmids may reflect general features of a class ofbroad-host-range plasmids to which these plasmids belong.A comparison of the molecular mechanisms of replicationbetween plasmids such as pTF-FC2 and the IncQ plasmidscould result in the identification of these features, which areassociated with the phenomenon of broad host range.We report the nucleotide sequence of the minimum repli-

con of pTF-FC2 in E. coli and P. aeruginosa and theidentification of at least two proteins essential for replicationin E. coli and a region which appears to be involved in hostrange. The origin of replication of pTF-FC2 has been com-pared with that of the IncQ plasmids, and the ability of thetwo plasmids to functionally complement each other hasbeen investigated.

METHODS AND MATERIALSBacterial strains and media. E. coli strains were grown in

Luria broth or on Luria agar plates. Antibiotics were addedas required at the following concentrations: ampicillin, 100,ug/ml; chloramphenicol, 25 pug/ml; and tetracycline, 20 ,ug/ml, as required. A DNA polymerase I (polA) mutant of therecA E. coli AB1157 (15), called E. coli GW125a, was a giftfrom Graham Walker, Massachusetts Institute of Technol-ogy, Cambridge, and was used to test for PolA-independentplasmid replication. Plasmids were maintained in E. coliLK111 (31). The broad-host-range property of plasmids wastested by transformation of DNA into P. aeruginosaPA01162 (leu-38 rmo-1l), (2) which was grown on LNG(containing, per liter, 10 g of tryptone, 5 g of yeast extract, 5g of NaCl, 0.4% KNO3, and 0.4% glucose) agar platescontaining 500 ,ug of carbenicillin per ml, 200 ,ug of tetracy-cline per ml, and/or 400 pug of chloramphenicol per ml,followed by reisolation of the plasmids. Plasmids used in thisstudy were the vector pUC19 (30) and pKE462 (7), whichhas the 7.5-kilobase EcoRI fragment from R46 carrying theTcr and arsenic resistance markers cloned into R300B.Plasmid pDR412 (23) consists of the 12.3-kilobase PstI-XhoIfragment of pTF-FC2 (Fig. 1) and a PstI-SalI fragmentcarrying the Cmr gene, but not the origin of replication ofpBR325 (4).

5697

5698 DORRINGTON AND RAWLINGS

General recombinant DNA techniques. Plasmid DNA wasprepared from E. coli and P. aeruginosa strains by using themethod of Ish-Horowicz and Burke (12), and analysis andcloning ofDNA were as described by Maniatis et al. (15). E.coli strains were transformed by using the dimethyl sulfoxidemethod of Chung and Miller (5), and transformation of P.aeruginosa PAO1162 was performed as follows. An over-night culture was harvested, suspended in 0.4 volume of cold0.15 M MgCl2, and kept on ice for 5 min, after which the cellswere pelleted and resuspended in 0.08 volume of 0.15 MMgCl2. DNA was added to 200 ,ul of cells, and the mixturewas incubated on ice for 60 min, at 37°C for 2 min, and againon ice for a further 5 min. After addition of 1 ml of LNGbroth, the cells were incubated at 37°C for 1 to 2 h and platedout.

Sequential deletions of pTV400 were made by usingexonuclease III (11), and the nucleotide sequence of thesedeletions was determined by the chain termination methodwith the Sequenase DNA-sequencing kit (United StatesBiochemical Corp., Cleveland, Ohio). The sequence wasanalyzed by using the Genepro (version 4.1) and Universityof Wisconsin Genetics Computer Group (UWGCG) (version6.1) software (6) packages.

In vitro protein synthesis. Translation products were iden-tified by using a procaryotic DNA-directed transcription-translation kit (code N.380; Amersham International plc,Amersham, England), and the products were analyzed bypolyacrylamide gel electrophoresis.

Nucleotide sequence accession number. The nucleotidesequence of pTV400 has been assigned the GenBank acces-sion number M35249.

RESULTSFunctional analysis of the pTF-FC2 origin of replication.

Recombinant plasmid pTV400 was constructed by cloning a3,202-bp partial Sau3A fragment of pTF-FC2 into pUC19(Fig. 1) and selecting for the ability to replicate in E. coliGW125a (7). Previously, we had shown that plasmidspTV4101, which had a 620-bp deletion from one side of the3,202-bp fragment, and pTV4200, with a 334-bp deletionfrom the opposite end (Fig. 1), were the smallest plasmidscapable of replication in E. coli GW125a with a normal copynumber (7). Plasmid pTV4210 (Fig. 1), with a further 150 bpdeleted from the distal end of pTV4200, was able to replicateat a reduced copy number (7). pTV400 and the deletionplasmids pTV4101, pTV4200, and pTV4210 were trans-formed into P. aeruginosa PA01162 and examined for theirability to be maintained. Plasmids pTV400 and pTV4200replicated normally, whereas pTV4210-containing colonieswere smaller on carbenicillin plates than pTV4200-contain-ing colonies were. No colonies containing either pTV4101 orpTV4100, which is 260 bp larger than pTV4101 (Fig. 1), wereisolated.

Nucleotide sequence of pTV400. The complete nucleotidesequence of pTV400 (3,202 bp) is shown in Fig. 2. Thesequence contains very few 6-bp recognition restrictionenzyme digestion sites relative to its size. Previouslymapped ApaI and AvaI restriction enzyme sites are locatedat nucleotides (nt) 1251 and 2067, respectively. Sequenceanalysis revealed unique sites for AvaII (nt 162), BalI (nt2419), DraI (nt 2107), HincIl (nt 1389), and XhoII (nt 1080).The 3,202-bp fragment has a G+C content of 59%, whichcorresponds closely to that calculated for the T. ferrooxidanschromosome (10) and to the average of 58.7% which we havecalculated for the T. ferrooxidans genes previously clonedand sequenced in our laboratory (19-22).

Computer analysis of the 3,202-bp fragment by using theCODONPREFERENCE and TESTCODE routines of theUWGCG package indicated the presence of at least fivepotential coding regions, each preceded by a consensusribosome-binding site (Fig. 3; Table 1). The amino terminusof open reading frame 2 (ORF2) overlaps with the carboxyterminus of ORFi and has two possible ATG start codons atnt 1684 and 1714, each preceded by a consensus ribosome-binding site. A protein of 33,778 daltons (Da) would beproduced if translation occurred from the first ATG, and oneof 32,659 Da would be produced if translation commencedfrom the second. The carboxy terminus of ORF2 overlapswith the first two direct repeats of the oriV.The oriV region of pTV400, which has previously been

shown to consist of three sets of 22-bp direct repeatedsequences and the 245-nt fragment immediately adjacent tothe repeats (7), is situated between nt 2540 and 2869. Otherfeatures of the nucleotide sequence are a region between nt655 and 677 containing a 14-bp match with the 22-bp re-peated sequences in the oriV and two sets of invertedcomplementary repeats (nt 2673 to 2691 and 2703 to 2738)(Fig. 2). The biological function of these structural featuresis unclear, but deletion of half of the latter set of invertedrepeats resulted in a reduced copy number (7).

Analysis of polypeptides encoded by the minimum replicon.Polypeptides from the pTF-FC2 minimum replicon wereanalyzed both in vivo and in vitro to ascertain whetherpolypeptides equivalent to those predicted from the nucleo-tide sequence could be detected. Polypeptides were ex-pressed from cloned fragments and deletion plasmids byusing an in vitro DNA-directed transcription-translation kit,and the products were analyzed by polyacrylamide gelelectrophoresis. Since the predicted polypeptides varied insize, different concentrations of acrylamide were used tomaximize resolution. Translation of the pUC19 vector re-sulted in the 30,500-Da P-lactamase protein (Fig. 4A, lane 5;Fig. 4B, lane 1), which was present in all the lanes. PlasmidpTV400, with the intact 3,202-bp fragment, produced addi-tional polypeptides of molecular masses 33,500 Da, 32,000Da (Fig. 4A, lane 1; Fig. 4B, lane 2), 10,000 Da, and 8,000 Da(Fig. 4B lane 2). These values correspond closely to thepolypeptides predicted for ORF2, ORFi, ORF4, and eitherORF3 or ORF5, respectively. Plasmid pTV4220 (nt 1 to2353) has the oriV and 77 amino acid residues deleted fromthe carboxy terminus of ORF2. The 33,500 Da correspond-ing to ORF2 has disappeared, and a new 28,500-Da polypep-tide, which corresponds closely to the predicted molecularmass of 25,679 Da (Fig. 4A, lane 2), has appeared, while theother polypeptides remained unchanged. Plasmid pTV4260(nt 1 to 1739), which has the oriV and all except the first 17amino acid residues of ORF2 deleted, showed a completeloss of the 33,500-Da protein while the other peptides wereunchanged (Fig. 4A, lane 3; Fig. 4B, lane 4). PlasmidspTV4283 (nt 1 to 801) and pTV4282 (nt 1 to 870), which haveboth ORFi and ORF2 deleted, no longer produced the33,500- or 32,000-Da proteins (Fig. 4A, lane 4; Fig. 4B, lane5) but still produced the 10,000- and 8,000-Da proteins (Fig.4B, lane 5). From the codon preference data (Fig. 3) it waspredicted that the ORF3 (8,000-Da) and ORF4 (10,000-Da)proteins are located within the first 571 bp of pTV400. Thispredicton correlates well with what was found in vitro.Plasmid pTV4290 (nt 1 to 660) produced proteins of 10,000and 8,000 Da (Fig. 4B, lane 6), while pTV4101 (nt 615 to3202) no longer produced these proteins (Fig. 4B, lane 3).Additional proteins of 30,000 and 28,000 Da (Fig. 4A, lanes1 to 3) were observed. These proteins disappeared when

J. BACTERIOL.

REPLICATION OF BROAD-HOST-RANGE PLASMID pTF-FC2

lul

pTF-FC212600

8000'

Sau 3A

0 1

Apa I Ava Sau 3ApTV400

2kb

ORF4

ORF1 (repA)

ORF2 (repC)

ORI V

FIG. 1. Restriction endonuclease cleavage map showing the position and direction of potential ORFs and subclones used in this study.Only pTF-FC2-derived fragments are shown. The locations of the regions required for mobilization (MOB) and the vegetative origin ofreplication (oriV) are shown as open boxes.

I

pTV4200

pTV4210

pTV4220

pTV4260

pTV4282

pTV4283

pTV4290

pTV4100

pTV4101

pTV4161

a s

I

m

5

VOL. 172, 1990 5699

3 Xod

I_

1 GATCGAGCGCGACAACTCGCCGGGAATGAGTCTCTAGCGTTGCGTGGTGGTTGTGATATACTTGTATAGCGTTTTCAGAAQGGAGCCGAAACATGCTTG101 CAATCCGACTGCCCGCCGAAGTGGAAACCCGCCTTGAAGCACTGGCGCAGGCCACAGGGCGGACCAAGACTTTCTATGCCCGCGAAGCCATCCTTGAGCA201 CTTGGATGACCTCGAAGATTTGTACCTTGCAGAGCAACGCCTGATCGACATTCGCGCAGGCAAAACCCAAACCGTGCCACTCGAAGAAGTGATGAAACGC301 TATGGCATGGAAGGTTGAACTCGACCCAGCCGCCGAGCGCGAGCTAGGCAAGATCGACCAGCAGACCGCCCGCCGCATCCTCGCTTTTTTGCATGGCCGT401 GTCGCCCAGCTCGACGACCCGCGCAGCATTGGCGAAGCCCTCAAAGGCTCCAAACTGGGAGCCTTCTGGAAATACCGCGTTGGGGATTGGCGAATCATCG501 CCAGCATCGAGGACGGTGCTTTGCGCATCCTCGTTATGCGCATCGGCAATCGTAAGGAGGTTTACCGCCAATGATCGAATACAGCTACCAGATCGACCCG601 CGCCCCTCCGACCTTGGCGGCGGCTGGCGGTTGCGCCTGTTGGAAAGCGGCGAGGAALGTCGGCGGCGGAGTGTTCCGTTGTCCGAGTACGCCACAGCAG701 AGAACGCAGAAGAAGCGGCCACGTACGCCTATGAGGACGCCTTGGCCGAGGCTTCGGCGTGGCTGGCATCGAGGGGCGAAAATTGAGCGGCGCGGCAGGG801 GATTGCGGCCCCGGCAGCGCCTAACCACAACTGTCTGAA ACAAGCATGGCTTTAGACATTATGGCGGCCTTCACCAATGAGCCGCCAGAACTTG

RepA-, M A L D I M A A F T N E P P E L

901 ATTTCATCTGGCCCGGATTCTTGGCCGGAACCGTGGGCGCACTTGTCGCACCTGGCGCAACTGGCAAGAGCTTTTTTGCTCTTGAAGCGGCCATGTCAATD F I W P G F L A G T V G A L V A P G A T G K S F F A L E A A M S I

1001 CGCTTGCAGTGTGGCAGGCGGCGACCTTGTGGGACTAACCCCGGCGCACACCGGGCGCGTGGTTTATCTCGCTGGCGAAGATCCACAGCCCGCCCTTGTGA C S V A G G D L V G L T P A H T G R V V Y L A G E D P Q P A L V

1101 CGACGTGTCCACGCCATCGGCCAGCACCTCAACCAGTCGGCCCGCGAAGCCATCGCTGAGAACCTGATGCTTGAGCCGATCATGGGCAAGCGGCTAAACGR R V H A I G Q H L N Q S A R E A I A E N L M L E P I M G K R L N

1201 TGATGGACGACGCGCACTTGCGCCGCGTCATCGACTACAGCGCAGGGGCCCGGCTGATTGTGCTGGACACCCTGAGCCGGATTCACATCCTCGACGAGAAV M D D A H L R R V I D Y S A G A R L I V L D T L S R I H I L D E N

1301 CAGCAATGGCGACTGGCCCACCTTGTTTCCGTGTTGGAACACATCGCGGCGACCACCGGGGCGGCTGTCCTGTACCTGCACCACGTCAACAAGGGCAGCS N G D M A H L V S V L E H I A A T T G A A V L Y L H H V N K G S

1401 GCCCGCGACGGCCAGACCGACCAGCAGCAGGCAGCGCGGGGCGCGTCTGCCCTGATCGACAACGCCAGATGGTGCGGCTATGTCGCCAAAATGACGGAGCA R D G Q T D Q Q Q A A R G A S A L I D N A R W C G Y V A K M T E

1501 AGGAAGCCGAGCGQTGAGTGACCGGGGCTTTGATCGTTCGCCATCCGGCAACGAGCGGCGCGGCCTTTTTGTCCGCTTTGGCGTGAGCAAGCAGAACTAQ E A E R M S D R G F D R S P S G N E R R G L F V R F G V S K Q N Y

1601 CGACGCGACCCCGCTAGACCGCTGGTATCGCGGCAGCGGCGGGGTGTTGTTGCCCGTTGAACTACAG CAGTQGCAATGGAGCGGAAAAA&D A T P L D R W Y Q R H S G G V L L P V E L Q E A I S N G A G K K

RepC* M E Q E K K

1701 G GCAATGAGCTATGACCTQCCATGCGCGGQCGACCCCGCGQTTGCCTQCGCCGGGGCTTTTCCGQGTCTCAAGCGCGGAGAACGAG G K R N E LG E S A M S Y D L T H A R H D P A H C L T P G L F R S L K R G E R

1801 AAGAGGCTCAAGCTCGATGTGACCTACAACTACGGAGATGACTCAATCCGTTTTTGGGGGCCTGAACCACTTGGCGGCGATGACTTGCGCGTATTGCAAGK R L K L D V T Y N Y G D D S I R F W G P E P L G G D D L R V L Q

1901 GGCTGGTGGCAATGGCTGCAATTTCCGGAGATAACGGGCGCGGQTCGTGCTACGGQCGAAACGGAATQGAAGQGGCQGCAACTCCGCCTATGGCTG L V A M A A I S G D N G R G I V L R H E T E S E A G Q Q L R L W L

2001 TGATATGCGGTGGGACGCQTAGAGAAAGATACGATGGTAGCCAAGGGCAGCTTCCGCCAGTTGGCCCGAGAACTTGGCTACGCCGAAGATGGAGGAAGTD M R W D A I E K D T M V A K G S F R Q L A R E L G Y A E D G G S

2101 CAGTTTAAAACQTCCGGGAAAGCATCGAACGGCTTTGGGCGGTATCGGTGATTGTCGAAAGAGGTGGTAAGCGGCAAGGGTTCCGCATTCTGTCCGAGTQ F K T I R E S I E R L W A V S V I V E R G G K R Q G F R I L S E

2201 ACGCGAGCGACGAGCAAGAAGGCAAGTTATTTGTTGCGCTTAATCCCCGGCTGGCGGATGCGGTTATGGGAGAGCGCCCGQQCCCGCATCAACATGGCY A S D E Q E G K L F V A L N P R L A D A V M G E R P H T R I N M A

2301 AGAAGTTCGCAAGCTGGAAACAGACCCGGCACGGCTGCTACCACCGCGGCTATGTGGCTGGATTGACCCCGGAAAGTCTGGCAAAGCTGAAATCGAQCGE V R K L E T D P A R L L H Q R L C G W I D P G K S G K A E I D T

2401 CTGTGCGGTTATGTATGGCCAGACGCAGCCAACGATGAAGCAATGAAAAAGCGCCGCCAGACCGCGCGCAAGGCGCTTGTCGAGCTTGCCGCCGTTGGTTL C G Y V W P D A A N D E A M K K R R Q T A R K A L V E L A A V G

2501 GGACGGTGAACGAGTACGCCAAGGGCAAGTGGGAAATCAGCAGGCCCAACCCCCGGCGTAACGTTCCCCAACCCCCGGCGTAACGTTCCCAACCCCCGG

2601

2701

2801

2901

W T V N E Y A K G K W E I S R P N P R R N V P Q P P A

CGTAACGTTCCCCGCCGAAATCTGAAAAACCTAGCAACGGCGCGGGTTTGCGGGCGATTCGCAAAATCCCTCCA-TGAT-CATCA42TCATCCACTAGG__________________--_---------------------oriV -----------------------------------------------------

----.~~~~CGCGGTACTTTTCGCCGCCCTTGA%g GCGTAAAA~GTC~CTTACC.AAAAACCCCAAAAGAAACGGCCGGGACAAGCCCGGCCAGACACACCCCGCCCCGCC

CCGCTCGCCTTCATTCTTCCACCGGGACAATGGACACCATCACCCGGTAGCGTTTGGCCCTTCCGGCAGGTACGCAGACGGCCAGCTTGGCGAGCGTTTC

GGCTGGCTTGTCGGTCGTGCTTGTGGAGCACATCGCCTCATACCCGAACAGAAGCCATCAGAATCGCCTACAGCGGATTTTTGGATGTTCTGGCTGCCTT3001 GAGCTAGGGTTGGTAAAGAAAACGCCTATGGCTGTTTGCGGGGCTTCTGCGAGCATTGCCGGGACGGTCTTGGGCTTGCTTGTGCCGTTGAGGCGAAAAA3101 CGCCACCGCCAGGACAAGCAGGGGTCGTCTCAGAAAACGGAAAATAAAGCACGCTAAGCCGGTTGCAGCGGCGGTAGCGGCCTGAACTCGCCCGCGCCGA3201 TC

FIG. 2. Nucleotide sequence of the 3,202-bp Sau3A DNA fragment of pTF-FC2 contained in pTV400 and the predicted amino acidsequences of ORF1 (repA) and (repC). Solid lines below the nucleotides denote the possible ribosome-binding sites and start codons of eachORF. The extent of oriV is indicated by a dotted line below the sequence. Thick arrows above the sequence denote the three tandemlyrepeated 22-bp DNA sequences of oriV and the partially homologous sequence (14 of 22 bp) upstream of repA. Thin arrows above and belowthe nucleotide sequence show the two sets of inverted complementary repeats in oriV.

5700

REPLICATION OF BROAD-HOST-RANGE PLASMID pTF-FC2

2.0

1.5

1.0

0.5

0.0

G)

C:0)

w

4)

4-4

0L)

0

c00la)

2.0

1.5

1.0

0.5-

0.0oI.

?.o

1.5

1.0

o.s

0.0oi

1,000 2,000 3,000

him* l.-NMI I hIIII hlmi l ullII II I I tII I . IR I o n 1lul I IIIJlli III I II 1 II 1 huh ml I*

0 1, 000 2,000 3,000

FIG. 3. Codon preference plot and rare codon usage for the predicted ORFs of the 3,202-bp fragment of pTV400. Rare codons are

indicated by bars below open boxes, which represent the ORFs. ORFs were identified by using the UWGCG package codon preferencesubroutine. A codon preference table based on ORF1 and ORF2 was used to help identify ORF3, ORF4, and ORF5.

ORFi was deleted (Fig. 4A, lane 4) and appeared to bedegradation products of ORF1.Comparison of ORF1 and ORF2 of pTF-FC2 with RepA

and RepC of the IncQ plasmids. The predicted amino acidsequences of the ORF1 and ORF2 polypeptides were alignedand compared with the amino acid sequences of RepA andRepC of RSF1010 (28). The predicted ORF1 protein is 11amino acids larger than RepA, and on alignment the twoproteins have a similarity of 43% based on the perfectlyconserved amino acid residues (Fig. 5). An alignment be-tween the larger of the two possible ORF2 polypeptides andRepC of RSF1010 is also shown in Fig. 5, and the similaritywas calculated at 60%. There is no homology between theremaining pTF-FC2 ORFs and any of the other RSF1010proteins. Owing to the similarity between the ORF1 andORF2 proteins of pTF-FC2 and RepA and RepC of

RSF1010, the corresponding pTF-FC2 genes were namedrepA and repC, respectively.

Since there was a high degree of homology between theRepA and RepC proteins of pTF-FC2 and RSF1010, we

wished to determine whether the IncQ proteins were able tocomplement mutations and deletions in the pTF-FC2 origin.Deletion plasmid pTV4161 contained only the region nt 2543to 3202 (Fig. 1), which included the intact oriV but none ofthe ORFs, and was able to replicate in E. coli GW125a onlywhen pDR412 was present (7). We substituted pDR412 withpKE462, a derivative of R300B, but were unable to detectreplication of pTV4161. The R300B replicon was thereforeunable to complement oriV of pTF-FC2.The ability of the IncQ plasmid proteins to complement

the pTF-FC2 RepA and RepC proteins individually was

investigated by inactivating the proteins through the intro-

TABLE 1. Location of potential protein-coding regions on the origin of replication of pTF-FC2

ORF Position of coding Translational initiation sitesa No. of amino acid Mol mass (Da)sequence (nt) residues

ORF1 852-1721 CUGUCUGAAAALGAGACAAGCAUG 290 31,252ORF2 1684-2580 AACUACAGGAGGCAAUCAGCAAUG 299 33,778

1714-2580 GAAAAAAAGGGGGAAA CGCAA.LQ 289 32,659ORF3 94-315 UUUCAGAACAGGAGCCGAAACAUG 74 8,462ORF4 302-571 UCGAAGAAGUGAUGAAACGCUAUG 90 10,318ORF5 571-783 UCGUAAGGAGGUUUACCGCCAALLG 71 7,702

a The start codons are underlined, as are the nucleotides in the pre-start region that could pair with the 3' end of the E. coli 16S rRNA (AUUCCUCCACU...5').

eORF3 ORF5 I I h ORF2 (repC) , h Ii

ORF4l 1 7 ORI* *(rePA U lllU IlEl I II III 11111 l 1h1111 h1 l ll

ORF1 (rePA)-I .- I . h h h __

VOL. 172, 1990 5701

0

5702 DORRINGTON ANDRAWLINGSJ.BCEOL

A ~~~~~~~~~~~~~~RSF1O1Oreplicon. A nucleotide sequence alignment betweenthe 3,202-bp fragment -of pTF-FC2 and the completely se-

kD 1 2 34 5quenced InCQ plasmid, RSF1O1O, is shown in Fig. 6. Thereare two regions wher'e DNA homology of more than 60% is33.5-~ ~ ~ ~ *3glf Shl fr h eAadRpCpoen,adte te orsod305- * 13apparent. One of these corresponds to the region that codes

ao.6- ~~~~~~~~~~tothe oriV regio'n. 'There was no significant homologyaS0 betwe'en the first 800 bp and the last 500 bp of the 3,202-bp

fragme'nt and the RSF1O1O replicon.

B

F]resi:

dele

pTVwer

pTV20%f

duc,

restand

digcend

Pro

Neilabs4trainmutpTI3341tion

(dat)A

offPequipT's

pDJ

rest

plaswer

that

eith,

A

DISCUSSIONThere is a marked lack of 6-bp restriction endonuclease

recognition site's in the nucleotide sequence of the 3,202-bpkD 1 2 3 4 5 6 origin of replication of pTF-FC2. This feature has also been

noted in other 'broad-host-range 'plasmids (29), and it has335 ~~~~~~~~~~~~beensuggested that this phe'nomonon may be essential for30.5~~ ~ ~ ~ ~ ~ ~ 4~ the survi"val, of promiscuous- plasmids.

Plasmid pTF-FC2 and the IncQ plasmids are similar in thestructure and nucleotide sequence of their oriV regions (7),

sm and fromp the nucleotide sequence reported here, it is clearthat the similarity extends to the region encoding their

Df ~~~~respective RepA and RepC proteins. This high degree ofhomology is maintained at both the DNA and amino acidlevels; this, along with the observation that the similarity isdistributed evenly acr'oss both proteins, indicates that thetwo sets of proteins have similar functions in the'replication

ion-- of their re'spective origins. Despite these'similarities, theIncQ RepA and RepC proteins were unable to complement

* ~~~~~~~~~mutations in their equivalent proteins on the p'TF-FC2The RepA protein of RSF1O1O has been reported to be a

DNA-dependent ATPase (27), and there is strong evidence,[IG. 4. Sodium dodecyl sulfate-polyacrylamide gel electropho- ta h eCpoenfntosa niiitro elctosanalysis of polypeptides translated in vitro from pTV400 andtion.()Lnes1to5 sow olyepties rompTV00, by binding to the 20-bp repeated DNA sequences in oriV of4220pT426,pV423, nd UC19 repecivey; rotins the IncQ pl'asmids (9, 14). Lin et al. (14) introduced point

e separated on a 12% acrylamide gel. -(B). Lan'es I to 6 show muaininte2-pretsoR16whcrslednarpeptides produced Iby pUC19, pTV400, pTV4101, pTV4260, reduction in the degree of incompatibility between plasmids(4282, and pTV4290, respectively; proteins were separated in a carrying these mutations and the wild-type R1162 origin a'sacrylamide gel. compared with plasmids carrying wild-type repeated se-

quences with'the wild-type R1162 origin. They postulatedthat this was due to an -altered binding efficiency of the RepC

-tion of frameshift mutations by using the ApaI and AvaI protein to the mutant repeated sequences. The ab'sence oftriction enzyme sites -which are located within the repA detectable incompatibility between the pTF-FC2 oriV andIrepC genes, respectively. Mutations were introduced by R300B (7), together with the finding that the two types ofr,stion of pTV4100 with ApaI or AvaI and flushing the dir'ectly repeat'ed sequences are only 60% homologous,Is with the Klenow fragment of DNA polymerase I to suggests that the inability of the IncQ RepC protein toduce plasmids 'pTV4102 (repA) and pTV41O3 (repC). complement a pTF-FC2 repC mutant is due to its inability toither mutant plasmid couild be replicated in E. coli in the bind to pTPF-C2 repeated sequences.ence of pDR412. An' in vitro DNA-directed transcription- Besid'es the repA and repC gene products, the IncQislation system was used to test whether the frameshift plas'mids have an absolute requirement -for the repB genetations had inactivated the respectiv'e genes. Plasmid product for replication to occur in vitro in E. coli (27). The14102, with a mutation- in repA, produced only the 35,900-Da protein produced by this gene is involved in500-Da RepC protein, whereas pTV41O3, with the muta- primer synthesi's at oriV during initiation of replication (28).Iin repC, produced only the 32,000-Da RepA protein The carboxy terminus of the repE gene is s'ituated within 500

:a n'ot shown). bp of the start of repA. No equivalent of this protein hasVe tested whether the individual RepA and RepC proteins been found 'in the 850 bp of spquence upstream of the.tSF1O1O could complement mutations in their respective pTF-FC2 repA start.iivalents on the pTF-FC2 origin. Plasmids pTV41O2 and We have identified a 615-bp region which affects the ability14103 were transformed into E. coli GW125a containing' of the pTF-FC2 replicon to be maintained in P. aeruginosa.R412 or pKE462. Transformation with either plasmid This'rgocodes for proteins of apoiately 8,000 andLilted in mor'e than 1,500'transformants per 100 ng of 10,000 Da, which correspond closely to the peptides pre-3mid when pDR412 was present, but no transformants dicted fo'r ORF3 or ORES and ORF4, respectively. In thee detected in the presence of pKE462. We concluded IncQ'plasmid, RSF1O1O, a small (10,000-Da.) protein hastthe IncQ replicon on pKE462 was unable to complement been implicated in the determination of host range (28).ier the repA or repC Imutant. There was' no significan't homology between this protein andlignnment of the o'rin of replication of pTF-FC2 with the the products of ORF3, ORF4, or ORES

J. BACTERIOL.

REPLICATION OF BROAD-HOST-RANGE PLASMID pTF-FC2

TF-FC2 RepA 1 MALDIMAAFTNEPPELDFIWPGFLAGTVGALVAPGATGKSFRSF1010 RepA 1 MATHKPINILEAFAAAPPPLDYVLPNMVAGTVGALVSPGGAGKSM

42 FALEAAMSIACSVAGGDLVGLTPAHTGRVVYLAGEDPQPALVRRV:s s s: .s 3:. . :s s.ss .s:: 3. 3.

46 LALQLAAQIA.---GGPDLLEVGELPTGPVIYLPAEDPPTAIHHRL

87 HAIGQHLNQSAREAIAENLMLEPIMGKRLNVMDDAHLRRVIDYSA23.2 ss 3 3.3. 3 . 3. 3 :.s 0

89 HALGAHLSAEERQAVADGLLIQPLIGSLPNIMAPEWFDGLKRAAE

132 GARLIVLDTLSRIHILDENSNGDMAHLVSVLEHIAATTGAAVLYLs ss sssss s ss .ss s ss .. s s:s ss ...s

133 GRRLMVLDTLRRFHIEEENASGPMAQVIGRMEAIAADTGCSIVFL

177 HHVNKGSARDGQTDQQQAARGASALIDNARWCGYVAKMTEQEAER178 HHASKGAAMMGAGDQQQASRGSSVLVDNIRWQSYLSSMTSAEAE-

222 MSDRGFDRSPSGNERRGLFVRFGVSKQNYDATPLDRWYQRHSGGV

222 ------EWGVDDDQRR-FFVRFGVSKANYGAPFADRWFRRHDGGV

267 LLPVELQEAISNGAGKKGGKRNEL33 3s .s

260 LKPAVLERQRKSKGVPRGEA

TF-FC2 RepC 1 MEQEKKGESAMSYDLTHARHDPAHCLTPGLFRSLKRGERKRLKLDVTYNY3.: :3:::::: 33::: :ssssss::::::3:s

RSF1010 RepC 1 VKPKNKHSLSHVRHDPAHCLAPGLFRALKRGERKRSKLDVTYDY

51 GDD-SIRFWGPEPLGGDDLRVLQGLVAMAAISGDNGRGIVLRHETESEAG:3 3 3 SSSSSS.SSSS.SSSSSSSS. 3 .3: 3 .S.:

43 GDGKRIEFSGPEPLGADDLRILQGLVAMAGPNG-----LVLGPEPKTEGG

100 QQLRLWLDMRWDAIEKDTMVAKGSFRQLARELGYAEDGGSQFKTIRESIEsss. .s.3. . sss.s ss.ss 3s .s

88 RQLRLFLEPKWEAVTAECHVVKGSYRALAKEIGAEVDSGGALKHIQDCtE

150 RLWAVSVIVERGGKRQGFRILSEYASDEQEGKLFVALNPRLADAVMGERP3SS SS.S SSS5SS.SSSSSS .3.3.SSS .3 33SS

138 RLWKVSIIAQNGRKRQGFRLLSEYASDEADGRLYVALNPLIAQAVMGGGQ

200 HTRINMAEVRKLETDPARLLHQRLCGWIDPGKSGKAEIDTLCGYVWPDAA3 3: 3 3s 3... sssssss:sss.s:s sssssss

188 HVRISMDEVRALDSETARLLHQRLCGWIDPGKTGKASIDTLCGYVWPSEA

250 NDEAMKKRRQTARKALVELAAVGWTVNEYAKGKWEISRPN-----PRRNV: 3:: 3 : 33ss s.ss: ss s .:.:s s

238 SGSTMRKRRQRVREALPELVALGWTVTEFAAGKYDITRPKAAGLPPPLYC

295 PQPPA

288 KQDTF

FIG. 5. Alignment of the amino sequences of the RepA and RepC proteins of pTF-FC2 and RSF1010 by using the Genepro computersequence analysis package. Identical and conservative amino acid changes are indicated by double and single dots, respectively, between thesequences.

VOL. 172, 1990 5703

5704 DORRINGTON AND RAWLINGS

1I

w1r

oa -1a

om X

ORF3 C!MORF4 repAl.

2 3 Kb

oriV

repC

pTF-FC2FIG. 6. Alignment of the nucleotide sequences of the 3,202-bp

fragment ofpTF-FC2 and nt 2000 to 7800 of RSF1010, which includeall sequences involved in replication and mobilization but excludethe sulfonamide and streptomycin coding sequences (28).

The IncQ RepA and RepC proteins are expressed from a

polycistronic mRNA (9), and their expression is controlledat the level of translation by an antisense RNA complemen-tary to the ribosome-binding site and start of the repA gene(13). In addition, expression of IncQ repC appears to betranslationally coupled to expression of repA (9). We wereable to detect the pTF-FC2 RepC product when the front endof repA was deleted and replaced by a T7 promoter (Dor-rington, unpublished results). In addition, the frameshiftmutation in repA did not affect RepC production in vitro.Thus, unlike for its IncQ counterpart, the translation ofRepC appears to take place even when uncoupled from thatof RepA.A comparison of the structure of the pTF-FC2 replicon

with that of the IncQ plasmids is interesting. In both plas-mids the repA genes, repC genes, and oriV regions haveconsiderable nucleotide homology. In pTF-FC2, the tworegions that show homology are contiguous, whereas inRSF1010, oriV is separated from repA and repC by a2,484-bp fragment that includes the sulfonamide and strep-tomycin resistance genes. These genes have a differentcodon usage pattern from the rest of the genes found onRSF1010, and Scholz et al. (28) suggest that these genes mayhave been incorporated at a later stage in the development of

the IncQ plasmids. The finding that the two homologousregions are contiguous in pTF-FC2 indicates that the tworeplicons may be derived from a common ancestor fromwhich they have susequently diverged. In the IncQ plas-mids, the repA and repC genes may have been separatedfrom oriV by insertion of a transposable element such asTn4, which carries sulfonamide and streptomycin resistancemarkers.

LITERATURE CITED

1. Bagdasarian, M., M. M. Bagdasarian, R. Lutz, A. Nordheim, A.Frey, and K. N. Timmis. 1982. Molecular and functional analysisof the broad-host-range plasmid RSF1010 and construction ofvectors for gene cloning in Gram-negative bacteria, p. 183-197.In S. Mitsubishi (ed.) Bacterial Drug Resistance. Japan Scien-tific Societies Press, Tokyo.

2. Bagdasarian, M., and K. N. Timmis. 1982. Host-vector systemsfor gene cloning in Pseudomonas. Curr. Top. Microbiol. Immu-nol. 96:47-67.

3. Barth, P., and N. J. Grinter. 1974. Comparison of the deoxyri-bonucleic acid molecular weight and homologies of plasmidsconferring linked resistance to streptomycin and sulfonamides.J. Bacteriol. 120:618-630.

4. Bolivar, F. 1978. Construction and characterization of newcloning vectors. III. Derivatives of plasmid pBR322 carryingunique EcoRI sites for selection of EcoRI generated recombi-nant DNA molecules. Gene 4:121-136.

5. Chung, C. T., and R. H. Miller. 1988. A rapid and convenientmethod for the preparation and storage of competent bacterialcells. Nucleic Acids Res. 16:3580.

6. Devereux, I., P. Haeberli, and 0. Smithies. 1984. A comprehen-sive set of sequence programs for the VAX. Nucleic Acids Res.12:387-395.

7. Dorrington, R. A., and D. E. Rawlings. 1989. Identification andsequence of the basic replication region of a broad-host-rangeplasmid isolated from Thiobacillus ferrooxidans. J. Bacteriol.171:2735-2739.

8. Guerry, P., J. van Emden, and S. Falkow. 1974. Molecularnature of two nonconjugative plasmids carrying drug resistancegenes. J. Bacteriol. 117:619-630.

9. Haring, V., P. Scholz, E. Scherzinger, J. Frey, K. Derbyshire, G.Hatfull, N. S. Willetts, and M. Bagdasarian. 1985. Protein RepCinvolved in copy number control of the broad-host-range plas-mid RSF1010. Proc. Natl. Acad. Sci. USA 82:6090-6094.

10. Harrison, A. P., Jr. 1984. The acidophilic thiobacilli and otheracidophilic bacteria that share their habitat. Annu. Rev. Micro-biol. 38:265-292.

11. Henikoff, S. 1984. Unidirectional digestion with exonuclease IIIcreates targeted breakpoints forDNA sequencing. Gene 28:351-369.

12. Ish-Horowicz, D., and F. Burke. 1983. Rapid and efficientcosmid cloning. Nucleic Acids Res. 9:2989-2998.

13. Kim, K., and R. J. Meyer. 1986. Copy number of broad-host-range plasmid R1162 is regulated by a small RNA. NucleicAcids Res. 14:8027-8046.

14. Lin, L.-S., Y.-J. Kim, and R. J. Meyer. 1987. The 20-base pair,directly repeated DNA sequence of broad host range plasmidR1162 exerts incompatibility in vivo and inhibits R1162 DNAreplication in vitro. Mol. Gen. Genet. 208:390-397.

15. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratories,Cold Spring Harbor, N.Y.

16. Meyer, R., M. Hinds, and M. Brasch. 1982. Properties of R1162,a broad-host-range, high-copy-number plasmid. J. Bacteriol.150:552-562.

17. Meyer, R., L.-S. Lin, K. Kim, and M. Brasch. 1984. Broad-host-range plasmid R1162: replication, incompatibility and copy-number control, p. 173-188. In D. R. Helinsky, S. N. Cohen,D. B. Clewell, D. A. Jackson, and A. Hollaender (ed.), Plas-mids in bacteria. Plenum Publishing Corp., New York.

18. Persson, C., and K. Nordstrom. 1986. Control of replication of

a- ~/

~~7.

J. BACTERIOL.

REPLICATION OF BROAD-HOST-RANGE PLASMID pTF-FC2

the broad host range plasmid RSF1010: the incompatibilitydeterminant consists of directly repeated DNA sequences. Mol.Gen. Genet. 203:189-192.

19. Pretorius, I.-M., D. E. Rawlings, E. G. O'Neil, W. A. Jones, andD. R. Woods. 1987. Nucleotide sequence of the gene encodingthe nitrogenase iron protein of Thiobacillus ferrooxidans. J.Bacteriol. 169:367-370.

20. Ramesar, R. S., V. Abratt, D. R. Woods, and D. E. Rawlings.1989. Nucleotide sequence and expression of a cloned Thioba-cillus ferrooxidans recA gene in Escherichia coli. Gene 78:1-8.

21. Rawlings, D. E. 1988. Sequence and structural analysis of the a-and ,-dinitrogenase subunits of Thiobacillusferrooxidans. Gene69:337-343.

22. Rawlings, D. E., W. A. Jones, E. G. O'Neil, and D. R. Woods.1987. Nucleotide sequence of the glutamine synthetase gene andits controlling region from the acidophilic autotroph Thiobacil-lusferrooxidans. Gene 53:211-217.

23. Rawlings, D. E., I.-M. Pretorius, and D. R. Woods. 1984.Expression of a Thiobacillus ferrooxidans origin of replicationin Escherichia coli. J. Bacteriol. 158:737-738.

24. Rawlings, D. E., I.-M. Pretorius, and D. R. Woods. 1986.Expression of Thiobacillus ferrooxidans plasmid functions andthe development of genetic systems for the thiobacilli. Biotech-nol. Bioeng. Symp. 16:281-287.

25. Rawlings, D. E., R. Sewcheran, and D. R. Woods. 1986. Char-

acterization of a broad-host-range mobilizable Thiobacillus fer-rooxidans plasmid and the construction of Thiobacillus cloningvectors, p. 419-427. In R. W. Lawrence, R. M. R. Branion, andH. G. Ebnes (ed.), Fundamental and applied biohydrometal-lurgy. Elsevier Biomedical Press, Amsterdam.

26. Rawlings, D. E., and D. R. Woods. 1985. Mobilization ofThiobacillus ferrooxidans plasmids among Escherichia colistrains. Appl. Environ. Microbiol. 49:1323-1325.

27. Scherzinger, E., M. M. Bagdasarian, P. Scholz, R. Lurz, B.Ruckert, and M. Bagdasarian. 1984. Replication of the broadhost range plasmid RSF1010: requirement for three plasmidencoded proteins. Proc. Natl. Acad. Sci. USA 81:654-658.

28. Scholz, P., V. Haring, B. Wittmann-Liebold, K. Ashman, M.Bagdasarian, and E. Scherzinger. 1989. Complete sequence andgene organization of the broad-host-range plasmid RSF1010.Gene 75:271-288.

29. Walker, G. C. 1986. Plasmid biology of pKM101 of the mucABgenes. Banbury Rep. 24:313-321.

30. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quence of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

31. Zabeau, M., and K. K. Stanley. 1982. Enhanced expression ofcro-beta-galactosidase fusion products under the control of thePR promoter of bacteriophage lambda. EMBO J. 1:1217-1224.

VOL. 172, 1990 5705