Construction and use of a self-cloning promoter probe vector for Gram-negative bacteria

Gene, 126 (1993) 17-23 ICY1993 Elsevier Science Publishers

GENE

0696

B.V. All rights reserved.

17

0378-I 119/93/$06.00

I

Construction and use of a self-cloning promoter probe vector for Gram-negative bacteria? (Transposon; Tn5; DNA cloning; 1acZ fusions; Pseudomonas aeruginosa; siderophore; pyoverdine; iron; transcription; gene regulation)

Tony R. Merriman and Iain L. Lamont Department

of Biochemistry. University of Otago, Dunedin, New Zealand

Received by G. Wilcox: 9 March

1992; Revised/Accepted:

23 November/24

November

1992; Received at publishers:

30 November

1992

SUMMARY

Transposon Tn5 has been used extensively for the genetic analysis of Gram- bacteria. We describe here the construction and use of a Tn.5 derivative which contains the ColEl origin of DNA replication, thereby allowing the cloning of DNA adjacent to the Tn without the need for construction of genomic libraries. The Tn is derived from Tn.&B21 [Simon et al., Gene 80 (1989) 161&169] and contains a promoter-probe 1acZ gene and genes encoding resistance to tetracycline and p-lactams. It is housed within a mobilisable suicide plasmid which can be transferred to a wide range of Grambacteria. The Tn was tested using pyoverdine siderophore-synthesis genes (pud) from Pseudomonas aeruginosa. The simple cloning procedure allowed 15.9 kb of pud-associated DNA to be cloned; in addition, the 1acZ reporter gene allowed the transcription of pvd genes to be studied. The bacteria were resistant to carbenicillin only if the Tn (and hence the B-lactamase-encoding gene) was downstream from an active promoter.

INTRODUCTION

There is increasing interest in a wide range of Grambacteria, which exhibit activities ranging from nitrogen fixation through pathogenicity of animals and plants, to the break-down of industrial pollutants, In parallel with Correspondence to: Dr. I. Lamont, Department of Biochemistry, University of Otago, P.O. Box 56, Dunedin, New Zealand. Tel. (64-3)4797869; Fax (64-3)4797866; e-mail: ilamont@‘sanger.otago.ac.nz ‘This paper is dedicated to the memory of Ian R. Merriman (193991991). Abbreviations: Ap, ampicillin; BGal, B-galactosidase; bla, gene encoding resistance to 8-lactams; bp, base pair(s); Cb, carbenicillin; cat, gene encoding Cm acetyltransferase; Cm, chloramphenicol; EDDA. ethylenediamine-di(o-hydroxyphenylacetic acid); gen, gene encoding GmR; Gm, gentamycin; IR, inverted repeat(s); kb, kilobase or 1000 bp; IacZ, promoterless gene encoding PGal; mob, c&acting mobilisation sequence; ori, origin of DNA replication of ColEl plasmids; pud, gene(s) encoding pyoverdine-synthesis enzymes; s, resistance/resistant; ‘, sensitivity/ sensitive; Tc, tetracycline; ret, gene(s) encoding Tcs; Tn, transposon; tnp, gene encoding Tn5 transposase; ::, novel junction (fusion or insertion).

the increased interest in the biology of these organisms there is increased study of their genetics, especially using the techniques of gene cloning to isolate and study genes of interest. The most usual approach used to clone genes from Gram- bacteria is to make a library of genomic DNA from the species of interest in Escherichia coli, using a vector which can subsequently be transferred by conjugation to other bacteria. The desired gene(s) is then identified by mobilising the library en masse into a recipient which is mutant for the desired gene, with selection for clones which complement the mutant gene and must therefore contain the functional version. A variation of this method which has been used in some cases is to isolate DNA which can complement an appropriate mutation in E. coli, and so must contain an analogous genetic function. An alternative approach is to mutagenise the bacteria with a Tn and then to identify mutants in which the Tn has inserted into the desired gene. Genomic DNA is prepared from the mutant and cloned in E. coli, with selec-

18

tion for a genetic marker present on the Tn; if an appropriate restriction enzyme is used, a portion of the genomic DNA flanking the Tn (and therefore containing at least part of the gene of interest) is also cloned. As an extension of this approach, it is possible to include within the Tn a DNA sequence which allows DNA replication in E. coli but not in the bacteria of interest, so that the desired DNA can replicate in E. coli without needing to be joined to a cloning vector. This approach was first used with Effcillu.~subtilis (Youngman et al., 1984) and ~~~~ococc~s ~u~i~~~s(Furuichi et al., 1985) and has been termed ‘self-cloning’. Vectors which should allow this method to be used in Gram- bacteria and which are based on insertion sequence IS1 have been described by Fellay et al. (1989). Here we describe the construction and use of an alternative self-cloning vector TnS-OTl82, which is derived from Tn.5. This Tn inserts almost at random in target DNA, with no identifiable preference for particular target sequences; in addition it functions in a wide range of Gram- bacteria and is very well characterised both physically and genetically (Berg, 1989), making it ideal for use as a self-cloning vector. We have tested the vector using siderophore-synthesis genes of P. ueruginosa, and our results show that it provides a very straightforward method of cloning DNA. The Tn also contains a promoterless ZucZ reporter gene and this can be used to monitor transcription of the gene into which the Tn has been inserted; transcription of some of the siderophoresynthesis genes of P. aeruginosa is affected by the level of iron which is available to the bacteria.

pBR325 cCXJPtO2fGml

Tn5-821

Y Tn5-OT182

Fig. I. Plasmid pOTl82. pGT182 is pSUP102(Gm) containing TnSOTl82. To construct TnS-OT182, pSUP102(Gm):TnS-B21 DNA (Simon et al., 1989) was partially digested with XhoI and end-filled using the Klenow fragment of E. co/i DNA polymerase. The DNA was then ligated with the 261 I-bp PwII fragment of pBR325 DNA (Bolivar, 1978) which contains hla and ori, and the ligation mixture was transformed into E. cali strain MC1061 (Casabadan and Cohen, 1980) by the CaCL, method (Dagert and Ehrlich, 1979) with selection for ApRCmRGmR bacteria. Isolates were obtained in which the fragment of

RESULTS

AND DISCUSSION

(a) Construction

of Tn5-0T182

TnS-OT182 was constructed from the Tn.5 derivative B21 (Simon et al., 1989). TnS-B21 is a promoter probe Tn containing a truncated promoterless IacZ gene, a Tc* gene and the essential transposition functions of Tn5. Sufficient IS50L sequence is situated proximal to lac.Z to allow transposition to occur at 50% of the frequency of wild-type Tn5 (Simon et al., 1989). Tn5-01-182 was constructed by inserting a bla gene and the ColEl origin of DNA replication at the Xhof site between the tet and EucZ genes in Tn5-B2 1. A map of pOTl82, which contains TnS-OTI 82, is shown in Fig. 1. (b) Mutagenesis of Pseudomonas aeruginosa with TnS0T182 and isolation of pyoverdine-deficient mutants Plasmid pOT182 was transformed into E. coli strain S 17- 1 (Simon et al., 1983). This strain contains a chromo-

somally-integrated

derivative of RP4, and gene-products

pBR325 had been inserted at the desired Xhol site between the tet and /UC% genes. The plasmid from one such isolate was named pOTl82, and the transposon pSUP102(Gm) is about

Tn5-OT182. 8.3-kb.

TnS-OTl82

is

10.9-kb

and

of this plasmid are capable of causing conjugative transfer of mob-containing plasmids such as pOTI into a wide range of Gram- bacteria (Simon et al., 1983). Plasmid pOT182 was introduced into P. ueruginosa OT684 (Potter and Loutit, 1982) from strain S 17- 1 (pOTl82) by conjugation as described elsewhere (Rombel and Lamont, 1992), and TcRCbR transconjugants were selected on Brain Heart Infusion Agar (Oxoid) containing Tc (50 pig/ml) and Cb (300 ng/ml). Both origins of DNA replication in pOTI are inactive in P. aeruginosa so that TcR and Cb* was conferred on the bacteria only if TnS-OT182 became integrated into chromosome of the transconjugants. Tc*Cb* transconjugants were obtained at a frequency of 1.7 x IO-” per initial donor titer. A total of 1230

19

TcRCbR Tn mutants were obtained. Of these 150 were tested for growth on agar containing Cm (100 ug/ml). All of them failed to grow, indicating that the cut gene, at least, of the plasmid had not been incorporated into the chromosome with the Tn. ~hromosomal DNA was prepared using the method of Sinclair and Holloway (1982) from ten mutants which had been picked at random. The DNA was treated with restriction enzyme Xhol, electrophoresed through agarose and Southern analysis was carried out by standard methods (Maniatis et al., 1982) using pOT1 X2 DNA as a probe. The results (not shown) showed that the Tn had inserted at different sites in all ten mutants, which is in agreement with reports that insertion of Tn5 occurs essentially at random in target DNA (Berg, 1989). We tested the utility of TnS-OTl82 by using it to investigate DNA involved in the production of pyoverdine. Pyoverdine is an iron-scavenging siderophore produced by P. aer~ginosa which causes colonies grown on Kings B medium (King et al., 1954) to ffuoresce when irradiated with ultraviolet light. Mutants which fail to make pyoverdine do not fluoresce and most such mutants fail to grow on Kings B agar containing EDDA, an iron-chelating agent (Ankenbauer et al., 1986; Hohnadel et al., 1986). The 1230 TcRCbR mutants of OT684 containing TnSOTI 82 were screened for the ability to make pyoverdine and two pyoverdine-deficient mutants were identified. Strain OT2100 was non-fluorescent but able to grow on agar containing EDDA (1 mg/ ml) whilst strain OT2101 was non-fluorescent and unable to grow in the presence of EDDA (400 pg/ml). Presumably the former mutant can secrete a derivative of pyoverdine which can act as an iron-chelating agent and is defective only in the fluorescent chromophoric part of pyoverdine. (c) Cloning DNA adjacent to Tn-insertion sites

Tn5-OT182 was designed to allow cloning of DNA flanking the Tn-insertion site. As a prelude to DNA cloning, it was necessary to identify restriction sites in the chromosome near to the Tn insertion site which would give appropriately-sized DNA fragments. Genomic DNA from the pyoverdine-deficient insertion mutants OT2100 and OT2101 was digested with a range of enzymes and analysed by Southern hybridization. The results of one such experiment with OT2100 are shown in Fig. 2a, along with the restriction sites that were identified near to each insertion (Fig. 2b). The information shown in Fig. 2b was then used to identify and clone restriction fragments containing up to 7.4 kb of P. aeruginosa chromosomal DNA which were adjacent to the Tn insertion sites in mutants 0T2100 and OT2101. The strategy used for cloning DNA adjacent to the Tn insertions involved three steps (Fig. 3). Firstly

chromosomal DNA was prepared from the mutants and digested with an appropriate restriction enzyme (Hind111 in Fig. 3) to generate restriction fragments containing the ColEI ori, a selectable marker (bla and/or tet) and adjacent P. ue~~gin~s~ ~hromosomal DNA; secondly, the DNA was ligated under conditions favouring circularisation of DNA molecules; and lastly, the ligated DNA was transformed into E. coli with selection for ApR and/or TcR colonies. Only molecules containing the Tn and DNA adjacent to the insertion site could give rise to transformants. As an example of such an experiment, chromosomal DNA from mutant OT2101 was prepared by the method of Marmur (1961) and a sample was digested to completion with HindIII. Amounts of DNA varying between 6.25 and 100 ng were ligated in a volume of 20 ul using T4 DNA ligase (0.1 units) (Boehringer), with the buffer provided by the manufacturer. Portions (10 ~1) were then transformed into E. c&i strain MC1061 (Casabadan and Cohen, 1980) which had been made competent using the CaCl, method (Dagert and Ehrlich, 1979), and ApRTcR transformants were selected. The maximum number of transformants (700) was obtained following ligation of 50 ng of DNA, with both smaller and larger amounts of DNA giving fewer or no transformants. A similar procedure was carried out to clone DNA extending to the EcoRI site to the right of the insertion in strain OT2100, and to the XhoI site to the left of the insertion in strain OT2100. The fragments which were cloned are represented in Fig. 2b. We have also used this procedure to clone DNA adjacent to other Tn5-OT182 insertions. However, despite several attempts using this procedure we were unable to obtain clones extending to the Sac1 or EcoRI sites to the right of the Tn in strain OT2101. We also failed to obtain clones following transformation of the ligated DNA into strain E. co/i TAP90 which has a mutation in the recD gene (Patterson and Dean, 1987), indicating that our inability to clone this DNA was not due to recombination events involving the desired clone. We were also unable to obtain clones following transformation of the ligated DNA into strain RP7947 (Liu and Parkinson, 1989) which contains a mutation (pcnB1) that reduces the copy-number of ColEI-like plasmids. The most likely reason for our failure to clone this DNA is that it contains a gene, the product of which is toxic to E. coli even at the lower copy-number associated with the pcnB strain. The CoIEl ori is in the centre of TnS-OTl82 (Fig. 1) so that a range of enzymes are available to clone DNA on either side of the Tn. DNA fragments adjacent to the inserted ZacZ gene can be cloned following digestion of the DNA with HindIII, SalI, Smal or XhoI; fragments adjacent to the transposase can be cloned following diges-

20

I

A X

SC H

Insenion OT2.100

c(

B X SC H E B

E B

?d

-

5.1/4.9-

-

2.011.9-

4.3 3.5

-

.. . . .] SacI

,..... Him3111

1

kb

a Fig. 2. Southern

analysis

and restriction

EC&II

XhOI

.xhOI

sac1

ECORI

BamI

Hi”dIII

1

b maps of the Tn insertions.

the left end (A) or the right end (B) of TnS-0T182. of Marmur (1961) and treated with the restriction

Panel (a) Southern

Chromosomal DNA was prepared enzymes shown. Southern analysis

analysis

of DNA from insertion

mutant

OT2100

probed

with

from P. aeruginosa insertion mutant OT2100 using the method was carried out using standard methods (Maniatis et al., 1982).

The probe DNA was radio-labelled by the random-priming method (Feinberg and Weinstein, 1983) and was the 2.5kb CIaI fragment containing the left end of TnS-0T182 (A) or the 1.6-kb XhoI -BumHI fragment containing the right end of TnS-OT182 (B). The BamHl OT182 is at the extreme left end of the Tn so that both probes detect the same fragment following BamHI digestion of the chromosomal

of pOT182 site in TnSDNA. The

second fainter band present in some lanes in panel B is due to contamination of the probe DNA with other TnS-OT182 sequences. Panel (b)Restriction sites adjacent to Tn-insertions 0T2100 and OT2101. The results shown in panel (a) for mutant OT2100, and equivalent results for mutant OT2101, were used to determine the locations of restriction sites near to the Tn insertions in the chromosomal DNA. The cloned DNA fragments are indicated (1-I

) above and below the main drawing.

tion with BumHI, ClaI, EcoRI or SacI. In addition, it should be possible to clone DNA flanking both sides of the Tn by using a restriction enzyme such as BglII which does not cut within the Tn, although we have not investigated this approach. (d) Use of lad to monitor gene expression Tn5-B21 contains a promoterless lacZ gene which can generate transcriptional fusions upon insertion into target DNA (Simon et al., 1989) so that the amount of l3Gal produced corresponds to the activity of the fused promoter upstream from the inserted 1ucZ gene. This feature is retained in Tn.5-OT182 and we wished to test its usefulness in monitoring transcription. Initially we assayed PGal production from ten Tninsertion mutants selected as being TcRCbR; these were the mutants, picked at random, which had previously been shown by Southern analysis to have insertions at different sites in the chromosome of P. aeruginosa.We expected that in about half of the mutants, the Tn would have inserted in the wrong orientation for 1acZ to be expressed from a fused promoter; in addition, in some cases where insertion had occurred in the appropriate

orientation the promoter would be inactive. Unexpectedly, all of these mutants made significant amounts of PGal ranging from 136 to 168 1 units. The reason for this is discussed in section e. We next analyzed expression of 1acZ in the two pyoverdine-deficient mutants, 0T2 100 and OT2 10 1. Pyoverdine is made by P. ueruginosa when the bacteria are grown in media containing low levels of free iron; the presence of iron suppresses synthesis of pyoverdine. Added iron reduced expression of lucZ in OT2101 by about 40% when compared to iron-deficient conditions (Table I). Expression of lucZ was also slightly reduced in mutant OT2100 when the bacteria were grown in iron-supplemented medium. Thus, the effect of iron in suppressing synthesis of pyoverdine is reflected in the fact that it suppresses l3-galactosidase synthesis in the promoter-fusion mutants OT2100 and OT2101. Both mutants made significant amounts of l3Gal even when iron was present in the growth medium at levels which repressed synthesis of pyoverdine. This may indicate that the genes mutated in OT2100 and OT2101 are involved in other pathways in addition to pyoverdine synthesis.

21 TABLE

I

Production

step

3.

1 Step

of PGal by TnS-OT182

insertion

mutants

Straina

CbR’s”

PGal (units)

OT2100

R

1661

0T2101

R

OT2090 OT209 I 012092

S R

0T2093 OT2094 OT2095 0T2096

S S R S R

0T2097 OT2098 OT2099

S R S

1364* 1386 809* 54 271 34 51 236 35 182 29 322 37

4.

a Strains mutants

OT2100 and OT2101 were identified as pyoverdine-deficient following mutagenesis of P. aeruginosa OT684 with Tn5-

OT182, and strains OT2090-0T2099 were ten randomly-picked insertion mutants obtained following selection for TcR bacteria; they were subsequently tested for resistance to Cb. b Sensitive (S) or resistant (R) to 300 pg Cb/ml. c BGal was assayed using the method of Miller (1972) with bacteria which had been grown in Brain Heart Infusion broth (Oxoid); asterisks Fig. 3. Strategy for cloning DNA using TnS-OT182. Chromosomal DNA was prepared from bacteria containing the Tn integrated into the chromosome

(step 1) and was then treated

tion enzyme.

The enzyme

with an appropriate

restric-

(Hind111 in the figure) was selected

indicate

that the broth

had been supplemented

with FeCI, (370 pg/ml).

The values obtained for strains OT2100 and OT2101 of three and two experiments, respectively.

are the averages

on the

basis of restriction mapping experiments of the sort shown in Fig. 2 so that an appropriately sized fragment was released from the DNA (step 2). The DNA was diluted to a concentration of about 2.5 pg/ml and DNA ligase was added to allow intra-molecular ligation of molecules to occur (step 3). The ligated DNA was then transformed into E. coli with selection

for TcR and/or

ApR bacteria.

Such bacteria

contained

plasmids comprised partly of Tn5 0T182 and partly of chromosomal DNA which was adjacent to the Tn-insertion site (step 4).

(e) Gene bla selects for fusion to active promoters In the experiments described above bacteria containing insertions of Tn.5-0T182 into the chromosome of P. aeruginosa were selected as being TcR CbR. All of the isolates tested made significant amounts of fiGal. In subsequent experiments we selected TcR transconjugants without selection for CbR. These were obtained at a frequency of about 40-fold higher than when TcRCbR transconjugants were selected. Forty of these transconjugants were then tested for CbR, and only twelve of them were able to grow. Southern analysis of DNA from three CbS and four CbR transconjugants revealed that the bla gene was present in all cases (Fig. 4). The most likely explanation for the CbS phenotype of these mutants was that the bla gene in Tn.5OT182 when inserted into the chromosome of P. aeruginom strain 0T684 was usually expressed at too low a level to confer resistance to 300 pg Cb/ml. The Tninsertion mutants of P. aeruginosa which were resistant

12

kb 21.2

3

45

6

7

-

5.1/4.94.3 3.5 -

2.0

-

Fig. 4. Southern analysis of DNA from TcR insertion mutants. DNA was prepared from seven mutants of P. aeruginosa strain 0T684 containing TnS-OT182 and which were selected as being TcR; three of these mutants (lanes 1, 3 and 6) were CbS and the remainder were CbR. The DNA was cleaved with Hind111 + EcoRI, electrophoresed on 0.8% agarose gels, and Southern transfer was carried out using standard methods (Maniatis et al., 1982). The DNA was probed with pOT182 DNA which had been radiolabelled using the random-priming method (Feinberg and Weinstein, 1983). The 6.9-kb band containing the bla gene from Tn5 OT 182 is arrowed.

22 to Cb most likely arose due to transcription originating from outside the Tn and extending through the 1acZ gene into the bla gene (see Fig. 1) giving increased expression of bla. To further test this, we measured PGal production by six TcRCbS strains and four TcRCbR strains and the results are shown in Table I. The CbS strains made only low amounts of PGal (less than 60 units), whereas the CbR strains made high amounts of PGal (greater than 180 units) so that there was a direct correlation between production of PGal and CbR. These results indicate that expression of CbR from TnS-OT182 in P. aeruginosa strain OT684 is dependent on transcription originating from outside the Tn. This feature allows direct selection of mutants in which the Tn has inserted downstream from an active promoter. We do not know if this will also be true in other species of bacteria, although it has recently been reported that expression of the bla gene from its own promoter is not sufficient to confer Ap resistance on Rhizobium leguminosarum (Leyva et al., 1990; Kokotek and Lotz, 1991). In experiments with P. putida insertion of TnS-0T182 into the chromosome conferred resistance to Cb (300 pg/ml) without requiring transcription from outside the Tn. We have not investigated the use of Tn_5OT182 in other species, nor have we used other concentrations of Cb.

(f) Conclusions (I) We have constructed

a TnS-based promoter probe transposon, TnS-0T182, which contains an E. colispecific origin of DNA replication. Tn5-0T182 provides a very simple system for cloning several kilobases of DNA on both sides of Tn insertion sites. We have used it to clone 15.9 kb of DNA from P. aeruginosa, at least some of which is involved in synthesis of a siderophore, pyoverdine. It should be possible to use Tn5-OT182 to clone DNA from any bacteria to which pOT182 can be transferred from E. coli, but in which it cannot replicate. The only limitation is that insertion of the Tn into the DNA of interest must generate a detectable non-lethal phenotype. (2) TnS-OT182 can also be used as a promoter-probe, with insertion of the Tn downstream from an active promoter causing expression of the 1acZ gene within the Tn. We have used this feature to show that expression of at least one of the genes involved in synthesis of pyoverdine is regulated at the transcriptional level by iron. Fortuitously, expression of CbR in strain OT684 of P. aeruginosa depends on transcription originating from outside the Tn and this provides a means for direct selection of transconjugants in which the Tn is inserted into actively transcribed regions of DNA; this feature may also be

useful in studies with other species. The combination of the promoter-probe and self-cloning features of the Tn should make it a powerful tool in the analysis of genes in Gram bacteria.

ACKNOWLEDGEMENTS

We are very grateful to Reinhard Simon for providing strain S17-1 and pSUP102(Gm) carrying Tn5-B21 and to John Parkinson for providing strain RP7947. T. M. is a recipient of a post-graduate scholarship from the Health Research Council of New Zealand.

REFERENCES Ankenbauer, R., Hanne, L. and Cox, C. D.: Mapping of mutations in Pseudomonas aeruginosu defective in pyoverdine production. J. Bacteriol. 167 (1986) 7-t 1. Berg, D.E.: Transposon Tn5. In: Berg, D. E. and Howe, Mobile DNA. pp. 183-208.

Am.

Sot.

Microbial.,

Washington,

M.M. (Eds.), DC,

1989,

Bolivar, F.: Construction and characterization of new cloning vehicles, III. Derivatives of plasmid pBR322 carrying unique EcoRI sites for selection of EcoRI generated (1978) 121-136.

recombinant

DNA molecules.

Gene 4

Casabadan, M.J. and Cohen, S.N.: Analysis of gene control signals by DNA fusion and cloning in Escherichia coli. J. Mol. Biol. 138 (I 980) 179-207. Dagert, M. and Ehrlich, S.D.: Prolonged incubation in calcium chloride improves the competence of Escherichia coli cells. Gene 6 (1979) 23-28. Feinberg, A. P. and Vogelstein, B.: A technique for radiolabelling DNA restriction fragments to high specific activity. Anal. Biochem. 132 (1983) 6-13. Fellay, R., Krisch, H.M., Prentki, P. and Frey, J.: Omegon-Km: a transposable element designed for in vivo insertional mutagenesis and cloning of genes in Gram-negative bacteria. Gene 76 (1989) 2155226. Furuichi, T., Inouye, M. and Inouye, S.: Novel one-step cloning vector with a transposable element: application to the M~xococcus xanthus genome. J. Bacterial. 164 (1985) 270-275. Hohnadel, D., Haas, D. and Meyer, J.-M.: Mapping of mutations affecting pyoverdine production in Pseudomonas aeruginosa. FEMS Microbial. Lett. 36 (1986) 195-199. King, E.O., Ward, M.K. and Raney, D.E.: Two simple media for the demonstration of pyocyanin and fluorescin. J. Lab. & Clin. Med. 44 (1954)301-307. Kokotek, W. and Lotz, W.: Construction of a mobilizable cloning vector for site-directed mutagenesis of Gram-negative bacteria: application to Rhizobium leguminosarum. Gene 98 (1991) 7-13. Leyva, A., Palacios, J., Murillo, J. and Ruiz-Argueso, T.: Genetic organization of the hydrogen-uptake (hup) cluster from Rhizobium leguminosarum. J. Bacterial. 172 (1990) 1647-1655. Liu, J. and Parkinson, J.S.:.Genetics and sequence analysis of the pcnB locus, an Escherichia coli gene involved in plasmid copy number control. J. Bacterial. 171 (1989) 1254-1261. Maniatis, T., Fritsch, E. F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982.

23 Marmur, J.: A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3 (1961) 208-218. Miller, J. H.: Experiments in Molecular Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1972. Patterson, T. A. and Dean, M.: Preparation of high titer lambda phage lysates. Nucleic Acids Res. 15(1987) 6298. Potter, A.A. and Loutit, J.S.: Exonuclease activity from Pseudomonas aeruginosa which is missing in phenotypically restrictionless mutants. J. Bacterial. 151 (1982) 1204-1209. Rombel, I. T. and Lamont, 1. L.: DNA homology between siderophore genes from fluorescent Pseudomonads. J. Gen. Microbial. 138 (1992) 181-187. Simon, R., Priefer, U. and Piihler, A.: A broad host range mobihzation

system for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria. Bio/Technology 1 (1983) 784-79 I. Simon, R., Quandt, J. and Khpp, W.: New derivatives of transposon Tn5 suitable for mobilization of replicons, generation of operon fusions and induction of genes in Gram-negative bacteria. Gene 80 (1989) 161-169. Sinclair, M.I. and Holloway, B.W.: A chromosomally located transposon in Pseudomonas amuginosu. J. Bacterial. 151 (1982) 5699579. Youngman, P., Perkins, J.B. and Losick, R.: A novel method for the rapid cloning in Escherichiu co/i of Bacillus subtibs chromosomal DNA adjacent to Tn917 insertions. Mol. Gen. Genet. 195 (1984) 4244433.