Construction of GFP Vectors for Use in Gram-Negative Bacteria Other Than

MICROBIOLOGY LETTERS ELSEVIER FEMS Microbiology Letters 145 (1996) 87-94

Construction of GFP vectors for use in Gram-negative bacteria Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 other than Escherichia coli

Ann G. Matthysse 8, Serina Stretton ‘, Catherine Dandie b, Nicholas C. McClure b, Amanda E. Goodman b.s

” Department of Bio1og.y. Lkversity of North Cuvolina, Chapel Hill, NC 27599-3280, D’SA I’ School of Biological Sciences. Thr Flinders bnniversitv of South Australia GPO Box 2100, Adelaide 5001. Austruliu

Receivzd 14 August 1996: accepted 12 September 1996

Abstract

A set of vectors containing a mutated gfj gene was constructed for use with Gram-negative bacteria other than Escherichia coli. These constructs were: pTn3gfp for making random promoter probe g@ insertions into cloned DNA in E. coli for subsequent introduction into host strains; pUTmini-TnSgfp for making random promoter probe &fp insertions directly into host strains: p519gfp and p519rzgfb. broad host range mob+ plasmids containing gfp expressed from a ILK and an npf promoter, respectively.

Keywords: Green fluorescent protem: Bacterial promoter probe; Transposon; Plant pathogenic bacteria

1. Introduction generally yielded disappointing results in that fluorescence was poor, and not detectable in single cells. The gene encoding green fluorescent protein Cormack et al. [4] found that poor fluorescence in (GFP), cloned from the jellyfish Aquovia victoria [l] bacteria was caused by GFP folding incorrectly and has been used as a visual marker of gene expression precipitating in the cells. They mutated sfp so that and cell structure in studies of eukaryotic organisms resulting mutant proteins folded correctly and re- [l-3]. Its use in prokaryotic systems has not, how- mained soluble in the cells. In addition, the excita- ever, paralleled its use in eukaryotes, even though tion optimum was shifted to between 481 and 501 the gene has been available commercially for the nm [4]. These mutations resulted in GFP fluores- past few years in Escherichia coli vectors (Clontechj. cence enhanced about 100 times compared to the Indeed, many initial attempts to use the commer- wild-type protein when expressed in E. coli. In addi- cially available clones in bacteria other than E. coli tion, Cormack et al. [4] cloned each mutated gene into an E. coli vector such that GFP was expressed from a tuc promoter with the T7 (genelo) ribosome binding site optimally placed.

* Corresponding author. Work from our laboratories is focussed on gaining E-mail: [email protected] a better understanding of microbial behaviour and

0378-1097 /96/ $12.00 Copyright 8 1996 Published by Elsevier Science B.V. All rights reserved PZZSO378-1097(96)00392-X 88 A.G. Matthysse et al. I FEMS Microbiology Letters 145 (1996) X7-94 activity in situ, generally in complex communities [5- be a simpler system for visualisation by microscopy 91. Since a microorganism’s activity is controlled at of specific gene expression in single bacterial cells in the genetic level, a way to achieve this aim is to situ. Our aim in this work was to construct several investigate the genetic regulation of microbial activ- vectors containing the GFPmut2 mutant gene [4] ity, preferably in single cells in situ. The use of trans- which will be useful for studies of Gram-negative poson mutagenesis, combined with reporter gene fu- bacteria other than E. coli. These vectors can be sion, is a powerful method with which to pursue used to generate reporter gene fusions or to tag bac- such studies [9-131. teria with a simple phenotypic marker for visualisa- IucZ and lux reporter gene fusions have been used tion and detection in environmental samples. Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 to investigate gene expression in single cells in cell smears [ 141, as well as in live cells in situ [15-171. These systems, however, require that the substrates/ 2. Materials and methods cofactors for enzyme action penetrate each cell. In addition, enzymatic activity of luciferase requires 2.1. Bacterial strains and plasmids oxygen and is dependent on cellular ATP levels. GFP requires no substrates or cofactors for activity The bacterial strains and plasmids used in this [4], thus GFP reporter fusions have the potential to study are described in Table 1. Plasmid constructs

Table 1 Strains and plasmids used in this study

Strain Genotype Reference/source

E. coli DHSa supE, Aluc ($80 lacZAM15), hsdR, recA endA, gyrA, thi, relA 1201 C21lOR RpR spontaneous mutant of C2110, NxR, polA this study and [24] Sl7-I hpir SmR. pro, thi, h&-M+, RP4-2-Tc:Mu-Km:Tn7, hpir ~271 C600 supE, hsdR, thi, thr, leu, lacY, tonA 1201

A. tumefaciens C58 wild type, virulent P51

A. rhkogenes K1347 KS4 derivative, plasmid free WI

P. putida UWCl RpR spontaneous mutant, plasmid free v91

Plasmid GFPmut2 ApR, &+ [41 pBC SK+ Cma Stratagene pNJ5000 Tea, tra+ 1301 pRK2013 Nm’<, tra+ [311 pTn3HoHol Ap” (Tn3. Amp with 1acZ reporter) ~241 pSShe Cm” (contains tnp for Tn3) [241 pUTmini-TnSluxAB ApH; TcR and lux reporter on mini-Tn5; mob+ u21 pDSKS19 RSFlOlO derivative, KmR, lac promoter, mob’ ~321 pCP13.101 TcR, A. tumefaciens eel operon, mob+ [251 ~BCgfp Cma. gfp+ this study PTnW> as for pTn3HoHo1, with g@ replacing lacZ this study pUTmini-TnSgfp as for pUTmini-TnSluxAB, with sfp replacing 1uxAB this study P519gfp pDSK519 with gfp cloned behind lac promoter this study P519fitiP p519& with pnpt2 cloned directly in front of gfj this study A. G. Matfh,sse et al. I FEMS Microbiology Letters 145 (1996) 87-94 89 made during this study are shown in Fig. 1. Each methods. Oligonucleotide primers were purchased plasmid construct, in E. coli DHSc(, has been lodged from AMRAD Pharmacia Biotech. with the American Type Culture Collection. The 740 bp gfp fragment, including the RBS, was amplified from pBCgfp at an annealing temperature 2.2. Growth conditions und media of 47°C using the following primers: gfpHindIII-F (5’-CTCAAGCTTGATTTCTAGATTTAAGAAG- E. coli strains were grown in Luria broth (LB, [18]) G), and g@EcoRI-R (5’-CTCGAATTCTCATTAT- at 37°C Pseudomonas putida UWCl in LB at 28°C TTGTATAGTTCATCCATGCC). The npt2 promo-

Agrobacterium tumefaciens C58 in LB or in RK ter was amplified as a 158 bp fragment from Tn5 Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 minimal medium [19] at 26°C A. rhizogenes strains (Genbank accession no. L19385) at an annealing K1347 in RK medium or in Nutrient Broth (Difco) temperature of 60°C using the following primers: at 26°C. Agar plates contained 15 g 1-i Bitek agar pnpt2HindIII-F (5’-CTCAAGCTTGCAGGTAGC- (Sigma). The following antibiotics (Sigma) and con- TTGCAGTGGG), and pnpt2XbaI-R (5’-CTCTCT- centrations were used when appropriate (pg ml-‘): AGAGCGCCATCAGATCCTTGGCG). E. coli, ampicillinlcarbenicillin (Ap/Cb, 50) chloram- phenicol (Cm, 10) kanamycin (Km, 50), naladixic 2.5. Conjugation acid (Nx, 20), rifampicin (Rp, 100) and tetracycline (Tc, 10); P. putida, Rp (50) and Tc (50); Agrobac- Conjugations between A. tumejkciens or A. rhizo- teria, Cb (50), neomycin (Nm, 60 in plates and 20 in genes and E. coli using plate matings were carried liquid) and Tc (5). out as previously described [21-231, using either pRK2013 or pNJ5000 (Table 1) as a helper plasmid 2.3. DNA cloning where appropriate.

Plasmid extractions, restriction enzyme digests, 2.6. Microscopy gel-isolated DNA fragment purifications, ligations, transformations and agarose gel electrophoresis An epifluorescence Olympus BX50 microscope, were carried out using standard methods [20], and fitted with a halogen lamp, a 100 W mercury burner, following the manufacturers’ instructions where ap- and a PM-30 automatic photomicrographic system, propriate. Restriction and other enzymes were from was used. Photomicrographs were generated using New England BioLabs Inc. either Nomarski (differential interference contrast) or epifluorescence (excitation 488 nm, emission 520 2.4. Polymeruse chain reaction (PCR) nm) optics, using a 40X objective with a numerical aperture of 1.0. Living bacterial cells were suspended PCR amplification was achieved using an FTS-320 in 100% glycerol on gelatin coated slides to prevent thermal sequencer (Corbett Research, Australia) motility. Cut roots of 1-g-week-old Arabidopsis thali- with Tth+ DNA polymerase (Biotech International, ana plants were incubated with lo”-lo7 bacteria Australia). The PCR solution contained 1 X reaction ml-’ in a 1 in 10 dilution of Murashige and Skoog buffer (67 mM Tris-HCl pH 8.8, 16.6 mM salts medium (MS, Gibco BRL) containing 0.4% su- (NH&Sod, 0.45% Triton X-100, 0.2 mg ml-i gela- crose. tin), 2 mM MgC12, 200 mM dNTPs, 0.05 mM of each primer, 1 unit Tth+ and 10 ng template DNA. A total of 30 cycles were run using the follow- 3. Results and discussion ing programme; denaturation at 94°C for 30 s, annealing at appropriate temperature for 30 s and ex- 3.1. pBCg;fp with varying restriction enzyme sites for tension at 72°C for 2 min, followed by 1 cycle at convenient subcloning of dp 72°C for 10 min. Mineral oil was removed from the PCR product which was extracted once with &fp with its ribosome binding site and most of the chloroform/isoamyl alcohol and purified by standard surrounding restriction enzyme sites was excised 90 A.G. Matthysse et al. I FEMS Microbiology Letters 145 (1996) 87-94

I (MC!3 oBCSE11 Sac1 Konl Smal BamHl Xbal RBS3 Ndel I

B Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 1 EcoRl Sac1 Kpnl Smal BamHl Xbal RBS Ndel 1

ISfil EcoRl BamHl Hindllll I Ndel RBS Xbal BamHl Smal Koni Sac1 I

I Sdhl lHindlll BamHl EcoRl Sal1 Hindlll Sfil Xhol Sal1 Clal Hindlll Sphl Pstll

D Accl Hincll 1Hindlll ’ Sphl Pstl Sall Xbal RBS NdelI

EdoRl

Fig. 1. Restriction enzyme maps of gfj constructs. (A) pBC,g@; ‘multiple cloning site (MCS) from pBC SK+ from Sac1 to EcoRI inclusive, ‘multiple cloning site from pBC SK+ from Hind111 to KpnI inclusive, :‘ribosome binding site (RBS). (B) Tn3gJp; the right end of pTn3HoHol from the Cl01 site to the right inverted repeat (IRK) is unchanged. and includes the b/u gene and a SacIsite. (C) pUTmini- TnSgfp: lux in pUTmini-TnSluxAB was replaced by gfp. (D) p519&, &fp replaced part of the polylinker downstream of plac in pDSK519; ‘there is a second Hind111 site in the NmR gene in the vector. (E) p519ngfj~. pnpt2 was inserted between the Hind111 and XbaI sites in front of sfi, in p519&. Diagrams are not to scale and only show altered parts of the vectors. A. ti. Matthysse ct al. I FEMS Microbiology Letters 145 f 1996) 87-94 91 from GFPmut2 using EcoRI and HindIII. The re- site in each of the 4 isolates, as determined by resulting approx. 750 bp &fp fragment was isolated striction mapping. These 4 plasmids were introduced by agarose gel electrophoresis and ligated into a into A. tumefaciens C58 by conjugation (using helper CmR bluescript vector, pBC SK+, which had also plasmid pRK2013), with transconjugants being se- been digested with EcoRI and NindIII. The resulting lected by growth on minimal RK medium contain- plasmid, pBCdp (Fig. lA), was transformed into ing Cb and Tc. Wild-type A. tumefaciens E. coli DHSa, selecting for CmR transformants. C58(pCP13.101) is not fluorescent at an excitation DHSa(pBC&p) cells from liquid cultures were wavelength of 488 nm. A. tumefaciens C58 transcon- brightly fluorescent under epifluorescence micros- jugants containing each of the 4 pCP13.101: :Tn3dp Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 copy (Fig. 2A,B). Bacteria grown on agar plates plasmids were fluorescent when incubated with A. and resuspended in water showed variable fluores- thaliana root segments (data not shown). cence, but all cells were fluorescent to some degree (data not shown). 3.3. pUTmini-TnSgfp for transposition to random genomic sites in non-E. coli bacteria 3.2. pTn3dp for transposition to cloned DNA in E. coli pUTmini-TnSluxAB was digested with NotI, blunt ended with DNA polymerase I (Klenow) and phos- The transposon mutagenesis system developed by phatased with calf intestinal alkaline phosphatase. Stachel et al. [24] using pSShe and pTn3HoHol has An approx. 800 bp fragment encoding the promoter- been used to introduce a promoterless P-galactosi- less &fp was isolated from pBC&p by cutting with dase gene into random sites in cloned DNA in E. EcoRI and XhoI and blunt ending with DNA poly- coli. We have modified this system for use with a merase I (Klenow). Vector and insert fragments were promoterless ,gfp reporter gene. The plasmids pSShe isolated after separation in an agarose gel, ligated and pTn3HoHol were prepared from E. coli and and transformed into E. coli S17.l(hpir). Plasmid separated by gel electrophoresis. pSShe was excised preparations from Ap and Tc resistant transformants from the gel and used without modification to trans- were digested with XbaI to determine the orientation form E. coli DHSa. Isolated pTn3HoHol was di- of s;fp in relation to the Tc resistance marker in the gested with EcoRI and Clal and fragments separated mini-Tn5afp. One plasmid containing tip in the cor- by gel electrophoresis. The 7 kb vectorlTn3 fragment rect orientation was named pUTmini-TnSgfp (Fig. was excised, purified, and ligated to the approx. 800 1C). bp fragment resulting from the digestion of pBCgfp E. coli S17.l(hpir)(pUTmini-TnS&) was conju- with EcoRI and CM. The ligation mixture was gated with recipient A. tumefuciens C58, A. rhizo- transformed into E. coli DHSo(pSShe) and transcon- genes K1347 or P. putida UWCl. From plating 200 jugants selected for resistance to both Cb and Cm. yl of each mating mixture, > 300 transconjugants The resulting plasmid construct was named pTn3& were recovered on each selection plate. Fifty trans- (Fig. 1B). conjugants from each plate were patched onto the To test the ability of E. coli DHSa(pSShe, same selective media and incubated at 28°C over- pTn3gfp) to introduce the transposon into cloned night. Colonies were resuspended in a drop of water DNA, a library clone from A. tumefaciens containing on microscope slides and viewed under epifluores- ccl genes (pCP13.101 [25]) was transformed into the cence microscopy at an excitation of 488 nm. Cells E. coli strain, selecting for resistance to Cm (pSShe), which showed a high level of fluorescence were Cb (pTnSdp), and Tc (pCP13.101). The procedure noted, and their corresponding colonies were re- described by Stachel et al. [24] was used to obtain E. streaked on the selective media. From the 50 colonies coli C2110R carrying pCP13.101 with Tn3gfp trans- tested from each mating, approx. 45 containing poson insertions. Cell smears of transconjugant col- strongly fluorescent cells were isolated (Fig. 2C-F). onies were examined using epifluorescence micros- In order to examine the potential use of these bac- copy and 4 fluorescent isolates were retained. The teria, A. tumefaciens C58 : :mini-TnSgfb tagged cells transposon had inserted into pCP13.101 at a single were incubated with roots of A. thaliana. Fluorescent Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021

Fig. 2. Photomicrographs of bacteria containing gjj contsructs. (A,C,E,G) Taken with Nomarski optics; (B,D,F,H) show the same Images. respectively, under epifluorescence microscopy. (A,B) E. coli(pBCgfp);(C,D) P. putidu: :mini-TnSgfp: (E-H) A. tumefaciens: :mini- Tn5&p; (E,F) free bacteria; (G,H) bacteria attached to the surface of roots of Arabidopsis thakzna. Note in H the bacterial fluorescence is bright in comparison to root autofluorescence. bacteria adhering to the root surface were clearly the 158 bp npt2 promoter PCR product, also di- visible (Fig. 2G,H). gested with Hind111 and XbaI. .PstI (the unique site which had been eliminated in the new construct, 3.4. p519gfb and p519ng&, broad host range mob+ compare Fig. 1D and E) was added to the ligation plasmids mixture immediately prior to transformation into E. coli DHSa, to prevent transformation by p519dp. The g;fp and pnpt2 PCR products were amplified Transformants were selected for NmR on LB con- with both primers containing restriction enzyme sites taining glucose (0.2%) to inhibit lac promoter activ- introduced at the 5 prime end to facilitate cloning. In ity. One transformant that showed green fluoresence order to construct a broad host range mob+ plasmid under epifluorescence microscopy was checked by containing an expressed gfp, the approx. 740 bp &fp PCR amplification with the primers pnpt2HindIII-F PCR fragment was digested with EcoRI and XbaI and dpEcoRI-R at an annealing temperature of and ligated to pDSK.519, also digested with the 50°C and was shown to contain the correctly sized, same restriction enzymes. The ligation mixture was approx. 900 bp, fragment containing pnpt2-tip. This transformed into E. coli DH5a cells and transfor- plasmid was named p519ng/b (Fig. 1E). E. coli mants were selected for Nm”. Plasmid DNA was DH5a(p519gfp) grown on LB containing glucose isolated from one transformant that showed green (0.2%) to reduce expression from the luc promoter, fluoresence under epifluorescence microscopy. This fluoresced less brightly than when grown without plasmid, p519gfp (Fig. lD), was shown to contain glucose. The addition of glucose to the medium a single copy of the J&P fragment in the correct posi- had no affect on the fluorescence of E. coli tion by restriction enzyme analysis. GFP is expressed DH5a(p519ngfp) cells indicating that the npt2 pro- in pDSKg&p from the luc promoter which is known moter in p519ngfp was functional. to have poor activity in other Gram-negative bacte- E. coli DH5a carrying either pDSK519, p519gfp, ria [26]. Therefore, another construct was made in or p519ngfp, were conjugated to A. tumefaciens C58 which the npt2 constitutive promoter was used to using helper plasmid pNJ5000. Transconjugants express GFP. p519gfp DNA was digested with were selected for NmR on minimal medium without XbaI and partially digested with HindIII. The ap- glucose and were assessed under epifluorescence prox. 9 kb vector fragment was separated by agarose microscopy. A. tumefaciens cells containing gel electrophoresis, excised, purified and ligated to pDSK5 19 (no gfi7) did not fluoresce. Cells containing A. G. Mutthwse et ul. I FEMS Microbiology Letters 145 (1996) X7-94 93

p519ngfp fluoresced more brightly than cells contain- (1996) Characterisation of carbon dioxide-inducible genes of ing p519dp. This is consistent with the work of the marine bacterium, Pseudomonas sp. S91. FEMS Micro- biol. Lett. 140, 3742. Labes et al. [26] who showed that the npt2 promoter Herrero, M., De Lorenzo. V. and Timmis. K.N. (1990) Trans- was approximately twice as strong as the luc promo- poson vectors containing non-antibiotic resistant selection ter in Rhizobia. markers for cloning and stable chromosomal insertion of for- eign genes in gram-negative bacteria. J. Bacterial. 172, 65577 6567. Berg, CM. and Berg. D.E. (1987) Uses of transposable ele- Acknowledgments f111 ments and maps of known insertions. In: Escherichia coli and Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 Salmonellu typhimurium: Cellular and Molecular Biology We thank Brendan Cormack for the generous gift (Neidhardt, F.C., Ingraham, J.L., Low. L.B., Magasamik, of plasmid GFPmut2, and Corbett Research Austra- B., Schaechter, M. and Umbarger, H.E., Eds.), pp. 1071l lia for supplying the thermal cycler. A.G.M. thanks 1109, American Society for Microbiology, Washington, DC. de Lorenzo. V., Herrero, M., Jakubzik. U. and Timmis: K.N. the University of North Carolina for a Keenan leave [I21 (1990) Mini-TnS transposon derivatives for insertion muta- which allowed her to undertake this work at the genesis. promotor probing, and chromosomal insertion of Flinders University. Part of this work was supported cloned DNA in gram-negative eubacteria. J. Bacterial. 172, by NSF grant number MCB-9405844 to A.G.M., by 6568-6572. a grant from the Patawalonga and Torrens Catch- u31 de Bruijn, F. and Rossbach, S. (1994) Transposon mutagenesis. In: Methods for General and Molecular Bacteriology ment Water Management Boards to N.C.M., and by (Gerhardt, P., Murray. R.G.E., Wood, W.A. and Kreig, a grant from the Flinders University to A.E.G. S.S. N.R., Eds.), pp. 387405, American Society for Microbiology, was supported by a Flinders University postgraduate Washington, DC. scholarship. u41 Nwoguh, C.E., Harwood. C.R. and Barer. M.R. (1995) De- tection of induced B-galactosidase activrity in individual non- culturable cells of pathogenic bacteria by quantitative cytolog- ical assay. Mol. Microbial. 17. 545-554. References u51 Shaw, J.J. and Kado, C.I. (1986) Development of a Vihrio bioluminescence gene-set to monitor phytopathoghenic bac- [I] Chalfie, M., Tu, Y., Euskirchen, G.. Ward. W.W. and teria during the ongoing disease process in a nondisruptive Prasher. D.C. (1994) Green Auorescent protein as a marker manner. BiofIechnology 4. 560-564. for gene expression. Science 263, 802-805. [I61 Waterhouse, R.N.. White, H., Silcock, D.J. and Glover, L.A. [2] Kolberg, R. (1994) Jellyfish protein lights up cells’ lives. (1993) The cloning and characterisation of phage promoters, J. NIH Res. 6, 4441. directing high expression of luciferase in Psrudumonas syrin- [3] Stearns, T. (1995) The green revolution. Curr. Biol. 5, 2622 gae pv. phuseolicola, allowring single cell and microcolony de- 264. tection in planta. Mol. Ecol. 2, 285-294. [4] Cormack, B.P., Valdivia, R.H. and Falkow. S. (1996) FACS- [17] Davies. D.G. and Geesey, G.G. (1995) Regulation of the al- optimized mutants of the green fluorescent protein (GFP). ginate biosynthesis gene ufgC in Pseudomonas aeruginosu dur- Gene 173, 33338. ing biofilm development in continuous culture. Appl. Environ. [5] Angles, M.L., Marshall, K.C. and Goodman, A.E. (1993) Microbial. 61. 860-867. Plasmid transfer between marine bacteria in the aqueous [18] Miller, J.H. (1972) Experiments in Molecular Genetics, Cold phase and biofilms in reactor microcosms. Appl. Environ. Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Microbial. 59, 8433850. [19] Roberts, W.P. and Kerr, A. (1974) Crown gall induction: [6] Dalton. H.M., Poulsen, L.K.. Halasz, P., Angles, M.L., serological reactions, isozyme patterns, and sensitivity to mi- Goodman, A.E. and Marshall. K.C. (1994) Substratum-in- tomycin C and to bacteriocin of pathogenic and non-patho- duced morphological changes in a marine bacterium and their genic strains of A. udiohucter. Physiol. Plant Pathol. 4, 81-91. relevance to biofilm structure. J. Bacterial. 176, 6900-6906. PO1 Sambrook, J., Fritsch, E.J. and Maniatis, T. (1989) Molecular [71 Matthysse. A.G. (1985) Mechanisms of bacterial adhesion to Cloning, A Laboratory Manual. Cold Spring Harbor Labora- plant surfaces. In: Bacterial Adhesion: Mechanisms and Phys- tory Press. Cold Spring Harbor, NY. iological Significance (Savage, D.W. and Fletcher M., Eds.), Pll Robertson-Crews. J.L., Colby, S. and Matthysse. A.G. (1990) pp. 2555275, Plenum Press, New York. Agrohacterium rhizogenes mutants that fail to bind to plant PI Goodman. A.E. and Marshall. K.C. (1995) Genetic responses cells, J. Bacterial. 172, 6182-6188. of bacteria at surfaces. In: Microbial Biofilms (Lappin-Scott, WI Matthysse, A.G. (1983) Role of bacterial cellulose fibrils in H.M. and Costerton J.W., Eds.), pp. 80-98, Cambridge Uni- Agrohacterim tum~faciens infections. J. Bacterial. 154, 9066 versity Press, Cambridge. 915. [91 Stretton, S., Marshall, K.C.. Dawes, I.W. and Goodman, A.E. ~231 Gstling, J., Goodman, A. and Kjelleberg, S. (1991) Behaviour 94 .4.G. Mutthysse et al. IFEMS Microbiology Letters 145 (1996) 87-94

of IncP-1 plasmids and a mimMu transposon in a marine and derivatives in the biological control of Agrobacterium I’ibrio sp. : isolation of starvation inducible lac operon fusions. biovar 2 pathogens. In: Improving Plant Productivity with FEMS Microbial. Ecol. 86, X3-94. Rhirosphere Bacteria. Proceedings of the 3rd International [24] Stachel, SE., An, G.. Flares. C. and Nester, E.W. (1985) A Workshop on Plant Growth Promoting Rhizobacteria (Ryder, Tn3 IacZ transposon for the random generation of P-galacto- M., Stephens, P.M. and Bowen, G.D., Eds.), pp. 1255127, sidase fusions: application to the analysis of gene expression CSIRO Division of Soils, Australia. in Agrobacterium. EMBO J. 4. 891-898. [29] McClure, N.C., Weightman, A.J. and Fry, J.C. (1989) Sur- [25] Matthysse, A.G., Lightfoot, R. and White, S. (1995) Genes vival of Pseudomonas putida UWCI containing cloned cata- required for cellulose synthesis in Agrobacterium tumefaciens. bolic genes in a model activated-sludge unit. Appl. Environ. J. Bacterial. 177, 1069-1075. Microbial. 55, 2627-2634. Downloaded from https://academic.oup.com/femsle/article/145/1/87/481817 by guest on 30 September 2021 [26] Labes, M., Piihler, A. and Simon, R. (1990) A new family of [30] Grinter, N.J. (1983) A broad host range cloning vector trans- RSFlOlO-derived expression and lac-fusion broad-host-range posable to various replicons. Gene 21, 133-143. vectors for Gram-negatrve bacteria. Gene 89, 3746. [31] Figurski, D.H. and Helinski, D.R. (1979) Replication of an [27] Simon, R., Priefer, U. and Ptihler, A. (1983) A broad host origin-containing derivative of plasmid RK2 dependent on a range mobilization system for in viva genetic engineering: plasmid function provided in trans. Proc. Natl. Acad. Sci. transposon mutagenesis in gram negative bacteria. BiofTech- USA 761, 1648-1652. nology 1, 784-791. [32] Keen, N.T., Tamaki, S., Kobayashi, D. and Trollinger, D. [28] McClure, N.C., Ahmadi, A.R. and Clam, B.G. (1994) The (1988) Improved broad host-range plasmids for DNA cloning role of agrocin 434 produced by Agrobacterium strain K84 in Gram-negative bacteria. Gene 70, 191-197.