Identification of Tn10 insertions in the dsbA gene affecting Escherichia coli biofilm formation

FEMS Microbiology Letters 173 (1999) 403^409

Identi¢cation of Tn10 insertions in the dsbA gene a¡ecting Escherichia coli bio¢lm formation Pierre Genevaux a; *, Pascale Bauda b , Michael S. DuBow c , Bauke Oudega a

a

Department of Molecular Microbiology, Faculty of Biology, Free University of Amsterdam, De Boelelaan 1087, 1081 HV, Amsterdam, The Netherlands b Department of Microbiology, Center for Environmental Sciences, BP4025, 1 rue des Reècollets, 57040 Metz, France c Department of Microbiology and Immunology, McGill University, 3775 University Street, Montreal, Quebec, H3A 2B4, Canada Received 28 December 1998; received in revised form 5 February 1999; accepted 12 February 1999

Abstract Escherichia coli was used as model to study initial adhesion and early biofilm development to an abiotic surface. Tn10 insertion mutants with reduced attachment to a polystyrene surface were isolated. Three adhesion mutants harbored the transposon in the dsbA gene, whose product, DsbA, catalyses folding of numerous extracytoplasmic disulfide bond-containing proteins. All three mutants were weakly adherent and grew poorly. Cell surface structure analysis showed that motility, type 1 fimbriation and lipopolysaccharide structure were affected in these mutants. The pleiotropic effect of the dsbA mutations on biofilm formation is discussed. z 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Escherichia coli; Adhesion ; Bio¢lm formation ; Abiotic surface; Mini-Tn10; DsbA

1. Introduction The inherent tendency of microorganisms to colonize both biological and non-biological surfaces is regarded as a fundamental aspect of bacterial pathogenesis and ecology [1]. Bacterial colonization of abiotic surfaces in aqueous systems is the result of a complex interaction between the cell, the surface and the liquid phase. This process usually takes place in two principal steps; The initial attachment, which requires cell-surface interactions, and the develop-

* Corresponding author. Tel.: +31 (20) 4447178; Fax: +31 (20) 4447136; E-mail: [email protected]

ment of the bio¢lm matrix, which requires the formation of cell-to-cell bridges necessary for a more stable establishment of the bacteria onto a surface. Both, proteins and polysaccharides were found in some cases to play a role in these processes [2^5]. To obtain molecular understanding of the cell factors required for the primary step(s) of attachment and for the early development of the bio¢lm matrix, a genetic approach using the well-characterized E. coli strain and a polystyrene surface as model was chosen. Various Tn10 insertion mutants de¢cient in adhesion to polystyrene were previously described [6]. This work presents Tn10 insertions localized in the dsbA gene, which considerably a¡ected adhesion to polystyrene. Cell membrane structure perturba-

0378-1097 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 9 9 ) 0 0 1 1 0 - X

FEMSLE 8707 25-3-99

404

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

tions, such as dysfunctional motility, decrease in type 1 ¢mbriae biosynthesis or LPS alteration associated with these mutations were analyzed and their e¡ects on adhesion are discussed.

2. Materials and methods 2.1. Bacterial strains, plasmid, and culture conditions E. coli K-12 strains and their sources are listed in Table 1. Strains constructed for this work are miniTn10 insertion derivatives of E. coli K-12 W3110 with altered adhesion properties [6]. The ampicillin resistant plasmid pRI4 containing the dsbA gene under control of the arabinose promoter, was kindly provided by Arne Rietsch (Harvard Medical School, Boston, USA). For induction of the arabinose promoter, 0.1% arabinose was added to the culture medium. Cells were grown in static LB broth at 37³C or on LB agar plates [9]. When necessary, media were supplemented with the following antibiotics: ampicillin, 100 Wg ml31 , and kanamycin, 30 Wg ml31 . The motility assay was carried out by applying 1 Wl of a culture grown to a culture turbidity at 580 nm of 0.6 (OD580 = 0.6) onto a soft agar plate (1% Tryptone, 0.5% NaCl, 0.35% agar). Diameters of growth were measured after 6 h incubation at 37³C. Adhesion experiments were carried out using polystyrene microplates as previously described [6]. 2.2. DNA manipulations, inverse PCR, DNA sequencing Isolation of plasmid DNA, transformation of CaCl2 -treated cells and all other basic DNA procedures were carried out essentially as described by Sambrook et al. [10]. Restriction endonucleases were used according to the manufacturer's speci¢cations (New England Biolabs). For localization of the mini-Tn10 insertion mutations in the E. coli chromosome, inverse PCR was carried out. Two oligonucleotides annealing closely together at the IS903 of the mini-Tn10 derivative 104 (Kmr ) [11] were prepared (primer I: 5P-TTA CAC TGA TGA ATG TTC CG-3P, primer II: 5P-GTC AGC CTG AAT ACG CGT-3P). Chromosomal DNA of the mutant strains was isolated and digested by either AvaII or

HaeII. These restriction enzymes did not digest the region between the 3P end of primer II and the end of the IS10. DNA restriction fragments were then circularized. Subsequently, PCR using the circular fragments was carried out with 50 pmol of each primer. The reactions were run for 1 min at 96³C and for 30 cycles each consisting of 10 s at 96³C, 30 s at 55³C, and 2 min and 30 s at 65³C. PCR products were extracted from agarose gels by centrifugation through blotting paper and puri¢ed by phenol/ chloroform/isoamyl alcohol (50:48:2 by volume) extraction [12]. Nucleotide sequencing was carried out using the Thermo Sequenase dye-terminator cycle sequencing kit from Amersham, with primer I as sequencing primer and double stranded PCR DNA fragments as template. Analyses were carried out using the 373A semi-automated DNA sequencer of Applied Biosystems/Perkin-Elmer. DNA sequences were checked for homology in the Genbank library. 2.3. Type 1 ¢mbriae production Type 1 ¢mbriation was analyzed by spot blot analysis and haemagglutination tests. Bacterial cells were cultured at 37³C in static LB medium to OD580 = 0.70, collected by centrifugation (30 s at 7000Ug) and resuspended in phosphate-bu¡ered saline (136.9 mM NaCl, 2.7 mM KCl, 10.03 mM Na2 HPO4 and 1.5 mM KH2 PO4 , pH 7.2). Serial dilutions of cell aliquots were spotted onto 0.2 Wm nitrocellulose membrane paper, using the Bio-Dot apparatus from Bio-Rad. Next, the membranes were incubated with polyclonal rabbit antibodies directed against type 1 ¢mbriae, kindly provided by M. Dho-Moulin (INRA, Tours, France). The blots were then incubated with goat anti-rabbit conjugated to horse radish-peroxidase as secondary antibodies. Detection was carried out using the ECL kit from Amersham. The capacity of bacterial cells to express a D-mannose-binding phenotype was assayed by their ability to agglutinate rabbit erythrocytes. Fifty Wl of bacterial cell cultures in PBS (OD580 = 10.0) plus 50 Wl of 5% rabbit erythrocytes in KRT (128.3 mM NaCl, 5.1 mM KCl, 1.3 mM MgSO4 W7H2 O, 2.7 mM CaCl2 W2H2 O, 10 mM Tris-HCl, pH 7.4) were mixed on glass slides and the time until agglutination occurred was monitored.

FEMSLE 8707 25-3-99

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

2.4. Isolation of periplasmic extract

405

ml of phosphate-bu¡ered saline. Proteinase K (1 mg ml31 ) was added and samples were incubated for 1 h at 60³C. Outer membrane preparations were analyzed by gel electrophoresis using 18% polyacrylamide gel and a (SDS)-tricine bu¡er system [15]. Silver staining of the LPS was carried out as described by Tsai and Frasch [16].

Bacterial cell cultures (OD580 = 0.7) were washed twice in phosphate-bu¡ered saline and concentrated to an OD580 of 40.0 in STE bu¡er (20% sucrose, 20 mM Tris-HCl (pH 8.0), 1 mM EDTA). Next, EDTA (5 mM ¢nal concentration) and lysozyme (50 Wg ml31 ¢nal concentration) were added. Cells were incubated for 2 h at 4³C. The supernatant fractions, containing periplasmic extracts, were then collected after centrifugation for 5 min at 10 000Ug.

3. Results 3.1. Isolation of dsbA mutants

2.5. Protein techniques

Starting with a total of 7000 randomly obtained E. coli K-12 W3110: :Tn10 insertion mutants, 72 mutants which showed a reduced ability to adhere to a polystyrene surface were isolated [6]. The precise position of the mutation in the chromosome of the adhesion mutants was determined by inverse PCR and DNA sequence analysis. Three of those adhesion-de¢cient mutants had an insertion in the dsbA gene, whose product catalyses the formation of disul¢de bonds in the periplasm (Table 1) [17]. DsbA was not detected in the mutants BGA1 and BGA4. For the mutant BGA5 a periplasmic product with an estimated molecular mass of about 20 kDa, which is about 2 kDa smaller than the wild-type protein was observed. This band also had a lower intensity than the DsbA band of the parent strain. The position of the mutation in BGA5 (68 bp before the end of the dsbA structural gene), might result in a less stable C-terminally truncated product, lacking the 22 C-terminal amino acids produced in this mutant.

Periplasmic extracts were separated by gel electrophoresis using 15% polyacrylamide gel [13]. Following gel electrophoresis, proteins were transferred onto nitrocellulose membrane, as described by Krone et al. [14]. Monoclonal rabbit antiserum raised against DsbA, kindly provided by A. Rietsch (Harvard medical school, Boston, MA, USA) was used to detect the DsbA protein. 2.6. Isolation and analysis of LPS Bacterial cells (OD580 = 0.7) were washed with phosphate-bu¡ered saline, and concentrated to OD580 = 2.0 in a solution containing 50 mM Tris and 2 mM EDTA (pH 8.5). Samples were frozen and thawed three times and sonicated three times for 10 s. Cell debris are collected by centrifugation (30 min at 13 000Ug, 4³C) and the pellets containing outer membrane fractions were resuspended in 100 Table 1 E. coli K-12 strains used in this study Strains

Revelant genotypea

Phenotype Motility

3

HB101 W3110

F

lacY1 recA13v¢m

dsbA mutants BGA1 BGA4 BGA5

F3 V3 IN(rrnD-rrnE)1 rph-1 dsbA (518) dsbA (593) dsbA (673)

b

Source or reference Haemagglutination

c

NT

3

[7]

+ 3 3 3

+ 3 3 3

[8] this study this study this study

a

For all the W3110 : :Tn10 derivative strains, the numbers in brackets referred to the nucleotide position of the mini-Tn10 in the DNA nucleotide sequence corresponding to the Z50423 Genbank accession number. b Determined by growth zones on soft agar plates and by observation under phase contrast microscopy. NT, not tested. c Ability to agglutinate rabbit erythrocytes after 5 min.

FEMSLE 8707 25-3-99

406

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

3.2. The dsbA mutants have reduced adhesion and growth Adhesion and growth were measured at various times after inoculation into polystyrene microplates. The dsbA mutants were strongly a¡ected for adhesion when compared to the parent strain W3110 (Fig. 1). As no signi¢cant di¡erences were observed between the strains BGA1 and BGA4, only BGA4 is shown in Fig. 1. In addition, the adhesion of the wild-type strain W3110 was strongly inhibited (of about 80%) when 2.5% D-mannose (w/v), the speci¢c receptor of the type 1 ¢mbriae adhesin was added to the incubation medium. Besides adhesion, growth of the dsbA mutants was considerably reduced. The mutant BGA5, which produces a truncated DsbA (see above), displayed very weak adhesion and was dramatically affected for growth when compared to the null mutants BGA1 and BGA4. When plasmid pRI4 containing the cloned E. coli K-12 dsbA gene, was introduced into each of these mutants, the ability to

Fig. 1. Adhesion to polystyrene and growth properties of the wild-type strain and of the dsbA mutants: WT, W3110 parent strain ; A, BGA4 dsbA; B, BGA5 dsbA. Adhesion and culture turbidity were measured 2 h (O), 5 h (b), 8.5 h (E), 13 h (F) and 20 h (a) after inoculation. Error bars indicate the standard deviation based on triplicate measurements.

adhere and grow was restored for the dsbA null mutants, but was still weak for the mutant BGA5. In addition, except for the mutant BGA5, the growth alterations of the dsbA null mutants were not ob-

Fig. 2. Silver-stained gel of LPS. Proteinase K digestions of outer membrane extracts were separated on 18% polyacrylamide tricine gel, and silver stained. Lanes : 1, LPS from the parent W3110; 2, BGA1 dsbA; 3, BGA4 dsbA; and 4, BGA5 dsbA.

FEMSLE 8707 25-3-99

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

407

served under shaken culture conditions (see Section 4). 3.3. Mutations in dsbA a¡ect motility, lipopolysaccharide structures and type 1 ¢mbriae biosynthesis Extracytoplasmic proteins tend to contain disul¢de bonds that contribute to their stability and in some cases, to their catalytic activity [17]. The dsbA mutations are highly pleiotropic and have been found in many cases to block the folding of secreted proteins. To determine the e¡ect of dsbA mutations on some outer membrane structures which might be important for adhesion, cell motility, LPS and ¢mbriation were analyzed. The motility capability was determined. In contrast with the wild-type strain W3110, all the dsbA mutants did not form growth zones after 6 h when stabbed into soft agar plates. To identify modi¢cations in the LPS core in the dsbA mutants, outer membrane extracts were prepared, treated with proteinase K and then analyzed by tricine-(SDS) polyacrylamide gel electrophoresis. In Fig. 2, a comparison of the silver-stained LPS gel pro¢le of the parent strain W3110 is made with those of the dsbA mutants. Most of the LPS of the parent W3110 strain migrated as a single broad band, whereas the three dsbA mutants produced two bands, one migrating with the parent LPS (band 2) and one showing slower migration (band 1). The common band observed for the mutants and the parent strain corresponded to the rough (R) LPS exhibited by E. coli K-12 W3110 [18]. To evaluate the amount of type 1 ¢mbriae produced by the dsbA mutants, spot blot analysis using anti-type 1 antibody was carried out (Fig. 3). Strain HB101 (¢m3 ) was used as negative control. The mutants were found to produce less than 10% of the amount of type 1 ¢mbriae than the parent strain. As type 1 ¢mbriae contain a speci¢c adhesin, FimH, that binds to mannose receptor at the surface of various red blood cells (see [19] for a review), the ability of the parent strain and its dsbA mutants to agglutinate rabbit erythrocytes was analyzed. No agglutination was observed for the mutants after 5 min, whereas the wild-type strain showed a response in 30 s after contact with the red blood cells (Table 1).

Fig. 3. Type 1 ¢mbriae production by spot blot analysis. Identical dilutions of whole cell suspensions were transferred to nitrocellulose and probed using anti-type 1 antibody. Lanes: 1, HB101 Fim3 ; 2, BGA1 dsbA; 3, BGA4 dsbA; 4, BGA5 dsbA; 5, W3110 parent strain.

4. Discussion The well-characterized bacterium E. coli K-12 was used as model to identify genes and membrane structures associated with adhesion to inanimate surfaces. In this study, three Tn10 insertion mutants which had inserts in the dsbA gene and which showed a reduced adhesion to polystyrene are presented. Considering the periplasmic localization of the DsbA protein, and its role in the folding of many secreted proteins, cell surface structures regarded as adhesives were analyzed. The dsbA mutations have been found to a¡ect type 1 ¢mbriae biosynthesis. Adhesive P ¢mbriae of uropathogenic E. coli were not assembled by a strain that lacked the periplasmic disul¢de isomerase DsbA [20]. In that case, DsbA was required for disul¢de bond formation in the Pap subunits and in the speci¢c periplasmic chaperone PapD itself. However, in contrast to the P ¢mbriae, the dsbA mutants

FEMSLE 8707 25-3-99

408

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

were still able to assemble type 1 ¢mbriae, but in a lower amount when compared with the parent strain. These di¡erences might be explained by the absence of cysteines in the type 1 ¢mbria periplasmic chaperone FimC. Surface components such as ¢mbriae have been in some cases associated with adhesion to non-biological surfaces. In S. typhimurium, a signi¢cant correlation between the presence of ¢mbriae and the adhesion to various substrata was observed [5]. Recently, Pratt and Kolter [2] isolated mutants of E. coli in the ¢m operon responsible for type 1 ¢mbriae biosynthesis with a strongly reduced ability to adhere to polystyrene. As observed in our case with the parent strain W3110, the wild-type adhesion was inhibited by mannose [2]. These data indicated that the loss of adhesion exhibited by the dsbA mutants might have been mainly due to the a¡ected type 1 ¢mbriation. In addition, the presence of type 1 ¢mbriae was found to strongly favor growth under static culture condition [21]. The advantage of the ¢mbriated over the non-¢mbriated bacteria in these growth conditions, is thought to lie in their ability to rapidly establish themselves in a pellicle on the surface of the broth, where their growth was promoted by the free supply of atmospheric oxygen [21]. In our case, the poor growth exhibited by the dsbA mutants BGA1 and BGA4 only in static culture conditions might have been due in large part to the considerably reduced amount of type 1 ¢mbriae found in these mutants. The dsbA mutants exhibited a non-motile phenotype. This is in accordance with Dailey et al. [22] who showed that dsbA mutants of E. coli failed to assemble the P ring of the £agella hook-basal-body because of the defective disul¢de bond formation. Motility is often cited as active participant in the adhesion process. In P. £uorescens, functional £agella permitted the rapid transport of the bacteria to the boundary layer [4]. A recent study showed that, in E. coli, motility, independently of the chemotaxis response, conferred advantage during the initial contact with the surface polystyrene substratum [2]. In a similar way, the loss of motility exhibited by the dsbA mutants could have in£uenced the adhesion patterns. Modi¢cations in LPS structures were observed in

the dsbA mutants. The appearance of the so far unidenti¢ed upper band in the dsbA mutants suggested additional component(s) on the LPS core. The absence of DsbA might a¡ect activity of several modi¢cation enzymes in the periplasm, resulting in a partial modi¢cation of the LPS. Further investigations are necessary to determine the chemical structure of band 1 in the dsbA mutants. In some cases, LPS has been associated with adhesion to inanimate surfaces, but its precise contribution in this process is not yet clear. In P. £uorescens, an increased attachment to polystyrene associated with the attenuation or lack of the antigen O was observed [3]. In the case of the dsbA mutants, the e¡ect of the modi¢ed LPS on bio¢lm formation has to be further investigated.

Acknowledgments We would like to thank Gregory M. Koningstein for technical support, Maryvonne Dho-Moulin (INRA, Tours, France) and Arne Rietsch (Harvard Medical School, Boston, MA, USA) for antiserum. This work was supported by the French National Center for Scienti¢c Research (C.N.R.S).

References [1] Costerton, J.W., Cheng, K.-J., Geesey, G.G., Ladd, T.I., Nickel, J.C., Dasgupta, M. and Marrie, T.J. (1987) Bacterial bio¢lms in nature and disease. Annu. Rev. Microbiol. 41, 435^464. [2] Pratt, L.A. and Kolter, R. (1998) Genetic analysis of Escherichia coli bio¢lm formation: roles of £agella, motility, chemotaxis and type 1 pili. Mol. Microbiol. 30, 285^293. [3] Williams, V. and Fletcher, M. (1996) Pseudomonas £uorescens adhesion and transport through porous media are a¡ected by lipopolysaccharide composition. Appl. Environ. Microbiol. 62, 100^104. [4] Korber, D.R., Lawrence, J.R. and Caldwell, D.E. (1994) Effect of motility on surface colonization and reproductive success of Pseudomonas £uorescens in dual-dilution continuous culture and batch culture systems. Appl. Environ. Microbiol. 60, 1421^1429. [5] Stenstroëm, T.A. and Kjelleberg, S. (1985) Fimbriae mediated non-speci¢c adhesion of Salmonella typhimurium to mineral particles. Arch. Microbiol. 143, 6^10. [6] Genevaux, P., Muller, S. and Bauda, P. (1996) A rapid screening procedure to identify mini-Tn10 insertion mutants of

FEMSLE 8707 25-3-99

P. Genevaux et al. / FEMS Microbiology Letters 173 (1999) 403^409

[7]

[8] [9]

[10]

[11]

[12] [13]

[14]

[15]

Escherichia coli K-12 with altered adhesion properties. FEMS Microbiol. Lett. 142, 27^30. Boyer, H.W. and Roulland-Dussoix, D. (1969) A complementation analysis of the restriction and modi¢cation of DNA in Escherichia coli. J. Mol. Biol. 41, 459^472. Bachmann, B.J. (1972) Pedigrees of some mutant strains of Escherichia coli K-12. Bacteriol. Rev. 36, 525^557. Miller, J.H. (1992) A Short Course in Bacterial Genetics; A Laboratory Manual and Handbook for E. coli and Related Bacteria. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Kleckner, N., Blender, J. and Gottesman, J. (1991) Uses of transposons with emphasis on Tn10. Methods Enzymol. 204, 139^180. Weichenhan, D. (1991) Fast recovery of DNA from agarose gels by centrifugation through blotting paper. TIG 7, 109. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680^685. Krone, W.J.A., De Vries, P., Koningstein, G., De Jong, A.J.R., De Graaf, F.K. and Oudega, B. (1986) Uptake of cloacin DF13 by susceptible cells: removal of immunity protein and fragmentation of cloacin molecules. J. Bacteriol. 166, 260^268. Schagger, H. and von Jagow, G. (1987) Tricine-sodium do-

[16]

[17] [18]

[19]

[20]

[21]

[22]

409

decyl sulfate-polyacrylamide gel electrophoresis for the separation of the proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368^379. Tsai, C.M. and Frasch, C.E. (1982) A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119, 115^119. Missiakas, D. and Raina, S. (1997) Protein folding in the bacterial periplasm. J. Bacteriol. 179, 2465^2471. Klena, J.D. and Schnaitman, C.A. (1994) Genes for TDPRhamnose synthesis a¡ect the pattern of lipopolysaccharide heterogeneity in Escherichia coli K-12. J. Bacteriol. 176, 4003^ 4010. Klemm, P. and Krogfelt, K.A. (1994) Type 1 ¢mbriae of Escherichia coli. In: Fimbriae: Adhesion, Genetics, Biogenesis and Vaccines (Klemm, P., Ed.), pp. 9^26. CRC Press, Ann Arbor, MI. Jacob-Dubuisson, F., Pinkner, J., Xu, Z., Striker, R., Padmanhaban, A. and Hultgren, S.J. (1994) PapD chaperone function in pilus biogenesis depends on oxidant and chaperone-like activities of DsbA. Proc. Natl. Acad. Sci. USA 91, 11552^11556. Old, D.C. and Duguid, J.P. (1970) Selective outgrowth of ¢mbriate bacteria in static liquid medium. J. Bacteriol. 103, 447^456. Dailey, F.E. and Berg, H.C. (1993) Mutants in disul¢de bond formation that disrupt £agellar assembly in Escherichia coli. Proc. Natl. Acad. Sci. USA 90, 1043^1047.

FEMSLE 8707 25-3-99