The Pseudomonas fluorescens SBW25 wrinkly spreader biofilm requires attachment factor, cellulose fibre and LPS interactions to maintain strength and integrity

Andrew J. Spiers1 and Paul B. Rainey1,2

1 Department of Plant Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, UK
2 School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand

Correspondence
Andrew J. Spiers
andrew.spiers{at}plants.ox.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The wrinkly spreader (WS) isolate of Pseudomonas fluorescens SBW25 forms a substantial biofilm at the air–liquid interface. The biofilm is composed of an extracellular partially acetylated cellulose-fibre matrix, and previous mutagenesis of WS with mini-Tn5 had identified both the regulatory and cellulose-biosynthetic operons. One uncharacterized WS mutant, WS-5, still expressed cellulose but produced very weak biofilms. In this work, the mini-Tn5 insertion site in WS-5 has been identified as being immediately upstream of the tol-pal operon. Like Tol-Pal mutants of other Gram-negative bacteria, WS-5 showed a ‘leaky-membrane’ phenotype, including the serendipitous ability to utilize sucrose, increased uptake of the hydrophilic dye propidium iodide, and the loss of lipopolysaccharide (LPS) expression. WS-5 cells were altered in relative hydrophobicity, and showed poorer recruitment and maintenance in the biofilm than WS. The WS-5 biofilm was also less sensitive to chemical interference during development. However, growth rate, cellulose expression and attachment were not significantly different between WS and WS-5. Finally, WS-5 biofilms could be partially complemented with WS-4, a biofilm- and attachment-deficient mutant that expressed LPS, resulting in a mixed biofilm with significantly increased strength. These findings show that a major component of the WS air–liquid biofilm strength results from the interactions between LPS and the cellulose matrix of the biofilm – and that in the WS biofilm, cellulose fibres, attachment factor and LPS are required for biofilm development, strength and integrity.


Abbreviations: A–L, air–liquid; CR, Congo red; DOC, deoxycholic acid; EPS, exopolysaccharide; Hr, relative hydrophobicity; MDM, maximum deformation mass; PI, propidium iodide


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Many bacteria are capable of forming large assemblages on plant and animal surfaces and tissues, on biological detritus, sediments, soils and other geological structures, as well as suspended flocs in water columns. Although often hard to investigate in situ, significant biological properties are attributed to these assemblages, including co-operative behaviour, competitive advantage, and defence against predators, antibiotics and immune systems, physical disturbance, etc. (for recent reviews see Costerton et al., 1995; Davey & O'Toole, 2000; Wimpenny et al., 2000; Lappin-Scott & Bass, 2001; Sutherland, 2001a; Wilson, 2001; Donlan, 2002; Dunne, 2002; Morris & Monier, 2003; Hall-Stoodley et al., 2004). These assemblages range from the largely random aggregation of bacteria growing on surfaces, in semi-solid environments or in constrained volumes, to complex structures incorporating substantial amounts of extracellular matrix material (Sutherland, 2001b; Hall-Stoodley & Stoodley, 2002; Stoodley et al., 2002; Ghigo, 2003). This latter type of assemblage represents biofilms in stricto senso, and the presence of structural matrix material provides biofilms with a cohesive physical identity that may be lacking both in colonies and in slime.

It is clear that the physical resilience of biofilms is the result of multiple interactions between matrix components (often exopolysaccharides, EPS), bacterial surface appendages (fimbriae, flagella and aggregation factors) and coatings (lipopolysaccharide, LPS) and the surface colonized by the bacteria (see references above and Dalton & March, 1998; Sutherland, 2001b; Donlan, 2002; Götz, 2002). In the case of the biofilms produced by Salmonella typhimurium and Salmonella enteritidis rdar mutants, and by the Pseudomonas fluorescens SBW25 wrinkly spreader, the expression of a cellulose matrix and a fimbrial-like attachment factor are the primary components contributing to biofilm strength and integrity (Römling & Rohde, 1999; Zogaj et al., 2001; Solano et al., 2002; Spiers et al., 2002, 2003). In each case, biofilms develop at the air–liquid (A–L) interface and are substantially larger and more robust than the archetypical submerged biofilm produced by many other bacteria, for example, Pseudomonas aeruginosa (see Wilson, 2001).

The wrinkly spreader (WS) is a niche-specialist genotype that colonizes the A–L interface of liquid cultures, forming an A–L biofilm, and grows poorly in the liquid column. It arises by spontaneous mutation from the ancestral (smooth; SM), non-biofilm-forming P. fluorescens SBW25 strain, in spatially structured microcosms, and shows significant negative frequency-dependent fitness advantage over the ancestral strain (Rainey & Travisano, 1998). Its selective advantage is attributable to cooperation among individual WS cells: overproduction of attachment factors, while costly to individual cells, results in the interests of individuals aligning with those of the group and allows colonization of the oxygen-replete A–L interface (Rainey & Rainey, 2003).

In an investigation of the genes required for biofilm formation by P. fluorescens WS (using one particular WS isolate, PR1200; Spiers et al., 2002), mini-Tn5 mutagenesis identified two major loci – the wsp chemosensory operon encoding the response regulator WspR, and the wss cellulose biosynthesis operon, which includes genes involved in the partial acetylation of the cellulose matrix (Spiers et al., 2002, 2003). WspR is required for the expression of cellulose and a putative curli or thin aggregative fimbriae (Tafi)-like attachment factor, both of which are required for normal WS biofilm development and colony formation. In addition, the cellulose acetylation-defective mutant WS-18 (WS wssF : : mini-Tn5) was found to produce weak biofilms. These findings suggest that the physical integrity of the WS biofilm results from the interactions between cellulose fibres, between fibres and attachment factor, and between attachment factor and the walls of the microcosm vial. This last interaction is required during the first phase of biofilm development when bacteria attach in the meniscus region of the liquid culture to the glass vials. Subsequent growth out over the A–L interface results in the characteristic WS biofilm (Spiers et al., 2003).

One of the previously identified WS mini-Tn5 mutants, WS-5, was found not to be associated with either of the wsp or wss operons (Spiers et al., 2002). WS-5 cells form an unusual colony morphology that is intermediate between that of WS and the non-biofilm-forming SM strain that produces smooth colonies on agar plates. Colonies of WS-5 bind Congo red, indicating that it expresses cellulose, and this strain is able to form weak biofilms when incubated in liquid microcosms. In this work, we reveal the genomic location of the transposon insertion responsible for the observed defects in WS-5 and show how defects in LPS expression affect A–L biofilm strength by significantly altering the cellulose matrix–attachment factor–bacterial cell interactions required for the normal development of WS biofilms.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids, culture media and growth conditions.
The Pseudomonas fluorescens strains used in this work are derivatives of P. fluorescens SBW25 (Table 1) and were grown using King's B (KB) medium (King et al., 1954) or minimal medium containing 20 mM sucrose at 28 °C. A KB microcosm consisted of a 35 ml Universal glass vial containing 6 ml KB, and was incubated with the lid held loosely in place with porous tape. P. fluorescens motility was assessed using 0·1x KB/0·3 % (w/v) soft-agar plates. Escherichia coli DH5{alpha} (Gibco-BRL) and S17-1-{lambda}pir (Simon et al., 1983) were used for DNA manipulation and conjugation. E. coli strains were grown using LB medium at 37 °C. The plasmids pBS+ (Stratagene) and pGEM7f (Promega) were used for subcloning and sequencing, and the P. fluorescens SBW25 cosmid library in E. coli S17-1-{lambda}pir was from Rainey (1999). p34S-Km3 was from Dennis & Zylstra (1998). WspR and WspR19 were expressed in trans using pVSP61-wspR12-{Omega}TcR (wild-type WspR) and pVSP61-wspR19-{Omega}TcR (WspR R129C); pVSP61-{Omega}TcR was used as a negative control (Goymer, 2002). Antibiotics were used at the following concentrations: ampicillin, 100 µg ml–1; chloramphenicol, 20 µg ml–1; kanamycin, 25 µg ml–1; piperacillin, 150 µg ml–1; and tetracycline, 12·5 µg ml–1. Congo red (CR) [also known as Direct Red (DR) 28; Sigma] was added to KB plates to a final concentration of 0·001 % (w/v). Direct Blue (DB) 1 (Chicago Sky Blue; Aldrich), 14 (Trypan Blue/Congo Blue; Aldrich), 15 (Sigma) and 53 (Evans Blue, Aldrich), Direct Red (DR) 2 (benzopurpurin, Aldrich) and CR were added to KB microcosms to a final concentration of 0·01 % (w/v).


View this table:
[in this window]
[in a new window]
 
Table 1. P. fluorescens SBW25 strains used in this work

 
General molecular biology methods.
Standard molecular biology techniques were used according to current protocols or manufacturer's instructions. Optical densities were determined using 4 mm-pathway plastic cuvettes and a Spectronic 20 Genesys (Spectronic Instruments) spectrophotometer. Conjugation and electroporation were used to transfer DNA between or into E. coli and P. fluorescens. Ligation mixtures were dialysed against deionized water for 30 min using 0·20 µm HA MF-membrane filters (Millipore) before electroporation. Plasmid and cosmid DNA was isolated from E. coli using Qiagen mini-spin kits. DNA was analysed by TBE-agarose gel electrophoresis and ethidium bromide-staining. The following subclones of the cosmid pAS256 were made for sequencing purposes: pAS257-260 and 263 were HindIII fragments in pBS+; pAS261 and 262 were SalI fragments in pBS+; pAS265, 266, 268 and 269 were partial HindIII deletions of pAS260; and pAS267, 271 and 272 were SphI fragments in pGEM7f (further mapping data are available on request). Nucleotide sequence was obtained by automated sequencing using standard vector and sequence-specific primers.

Construction of the WS tolA mutant.
This mutant was produced by the insertion of a kanamycin-resistance (KmR) cassette into tolA. This was achieved by cloning KmR from p34S-Km3 into the SphI site of pAS258, a pBS+ clone containing the three HindIII fragments covering the ybgC-tolB region (Fig. 1), to give pAS298. This plasmid was electroporated into WS, and KmR integrants were isolated. These were grown in non-selective medium then plated onto KB plus kanamycin. Individual colonies were tested for piperacillin sensitivity (bla from pAS298 confers resistance to piperacillin, Pp) indicating vector excision, and an appropriate KmR PpS isolate chosen as WS tolA.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. The mini-Tn5 transposon in WS-5 is located 24 bp upstream of the start of ybgC, the first gene of the tol-pal operon. The functions of YbgC, E and F are unknown, but the TolA, R, Q and Pal proteins are involved in maintaining the functional integrity of the inner and outer cell membranes. The tol-pal operon is highly conserved amongst Gram-negative bacteria, and the P. fluorescens SBW25 gene sequences show high levels of homology with the genes from E. coli. Downstream of tol-pal is a series of tRNA genes, followed by nadA (L-aspartate oxidase subunit). The regions sequenced in this work are shown in black. Gene positions were subsequently determined by in silico analysis of this data and of the Wellcome Trust Sanger Institute SBW25 genome project database. SBW25 tRNA sequences (small triangles) were obtained but were not mapped within the contig. The kanamycin-resistance cassette inserted in tolA at the SphI site, and the position of the mini-Tn5 insertion in WS-5, are indicated.

 
LPS analysis.
LPS samples were obtained from cultures using EDTA extraction. Overnight KB cultures were first diluted to an OD600 of ~0·5 to equalize cell numbers. Cells from 2 ml of culture were resuspended in 100 µl deionized water; 400 µl 250 mM EDTA (pH 8·0) was added and the suspension vortexed vigorously for 5 s. The suspension was incubated at 37 °C for 30 min and vortexed every 10 min. The supernatant was recovered for analysis after centrifugation at 10 000 g for 5 min. Aliquots were examined using 18 % deoxycholic acid polyacrylamide gel electrophoresis (DOC-PAGE) and silver staining (Bio-Rad) according to Reuhs et al. (1998).

The monoclonal antibody (mAb) BC12-CA4 (Meyer & Dewey, 2000) was used to detect LPS by ELISA assay. This mAb recognizes an unidentified antigen from Botrytis cinerea and binding is inhibited by rhamnose (Rha), suggesting that the target is a Rha-glycosylated protein. It was tested as a possible mAb against pseudomonads as these bacteria contain substantial Rha polymers as part of the conserved LPS A-band O-polysaccharide component (Rocchetta et al., 1999). KB-grown cells were resuspended in PBS (20 mM sodium phosphate, 150 mM NaCl, pH 8·0), adjusted to an OD600 of ~0·5 to equalize cell numbers, and used to produce a 101–103 dilution series in PBS. Aliquots of 100 µl were adsorbed to a MaxiSorp ELISA Plate (Nunc) overnight at 4 °C. After washing with PBS/0·05 % (v/v) Tween-20, mAb was added and incubated for 1 h at 37 °C. Bound mAb was detected using anti-mouse polyvalent immunoglobulin peroxidase conjugate (Sigma) and TMB substrate (Adgen) and absorbance measured at 450 nm. mAb binding was tested as {Delta}A450 OD600–1, having adjusted for cell numbers. Measurements were performed in duplicate.

Microscopy and FACS.
Propidium iodide (PI) (Sigma) was used to assess the leaky membrane phenotype of mutants after Gaspar et al. (2000). PI (20 µM) was added to KB cultures and incubated at 28 °C in the dark for 30 min before examination with an Olympus BX50 epifluorescence microscope. Calcofluor (Fluorescent Whitener 28, Sigma) was used to assess the presence of cellulose in biofilms or colony material. Calcofluor (10 µM) was added to samples resuspended in KB, then incubated at 28 °C for 2 h before washing with fresh KB and subsequent examination. A fluorescence-activated cell sorter (FACS) was used to determine the percentage of cells stained with PI. Overnight KB cultures were diluted to an OD600 of 0·100–0·150 in fresh KB and incubated with PI for 30 min. FLT-2 (orange) fluorescence and scattering was measured for 100 000 events, and the percentage above a threshold determined by preliminary comparison between WS and WS-5 was recorded for each culture.

Biofilm and cellulose assays.
Bacterial attachment to the glass of KB microcosms in the meniscus region was determined quantitatively using crystal violet as previously described (Spiers et al., 2003) and presented as the relative attachment with respect to the WS biofilm [A570 OD570WS–1]. The absolute strength of KB-grown biofilms was determined by placing glass balls in the centre of each biofilm until it broke, sank or was ripped from the sides of the microcosm vial, to determine the maximum deformation mass (MDM) (grams) (Spiers et al., 2003). In the case of the complementation and chemical interference experiments, biofilms were incubated on the bench at 20–22 °C to minimize physical disturbance; as a result, MDM values were lower than those obtained at 28 °C. In the complementation experiments, overnight cultures of strains to be tested were diluted to an OD600 of 1·00. KB microcosms were then inoculated with 100 µl aliquots of single strains or a mixture of two strains (in a total volume of 100 µl) in such a manner that each test used the same total number of cells. In the interference experiments, 120 µl 500 mM EDTA or water was added to mature KB-grown biofilms at the meniscus in order to avoid disruption of the biofilm. The EDTA was allowed to mix by diffusion over a period of 2 h before MDM were determined. A quantitative measure of cellulose production using CR was made according to Spiers et al. (2003) using material from KB microcosms, and data presented as {Delta}A490 OD600–1, adjusted for cell numbers.

Hydrophobicity assay.
Relative hydrophobicity (Hr) of strains was tested using the MATH assay (after van der Mei & Busscher, 2001). KB-grown cells were resuspended in 4 ml KB or 10 mM potassium phosphate (pH 5·0) to an OD600 of ~0·5 and 1 ml hexadecane was added. The samples were vortexed for 5 s and then allowed to stand for 20 min before the OD600 of the aqueous phase was determined (OD600i). The samples were revortexed for 60 s, allowed to stand for 20 min and the OD600 remeasured (OD600f). The ratio OD600f OD600i–1 gave the relative hydrophobicity (Hr) of cells of each strain (Hr determined using KB reflects real differences in hydrophobicity experienced by various strains in KB microcosms, whereas Hr determined using potassium phosphate reflects an arbitrary difference useful for comparative purposes).

Recruitment and maintenance assays.
Bacterial recruitment to the surface was assessed by using standard cuvettes containing 2 ml KB. This resulted in a liquid column of 30 mm, with the surface 20 mm above the region in which OD600 was determined. KB-grown cells were resuspended in 2 ml fresh KB with a OD600 of 0·2–0·3. OD600 measurements were taken every 10 min, with the cuvette remaining in place throughout. After 1 h, samples were mixed, and the OD600 was remeasured to determine bacterial growth, and drift checked with a blank KB sample. The mean relative OD600 was calculated using the data from the 40–60 min time-points. Bacterial maintenance within biofilms was assessed by sampling standard KB microcosms. Aliquots of 1 ml were taken from immediately below biofilms and the OD600 determined. Microcosms were then vortexed and a 1 ml aliquot removed to determine the total OD600 (OD600t). The ratio OD600 OD600t–1 gave the proportion of cells in the liquid column.

Statistical analyses.
All data are presented as the mean±standard error (SE) where appropriate. Assays were performed with five to eight replicates unless otherwise stated. ANOVA and Student's t-tests were performed using JMP Statistical Discovery Software (SAS) and P values provided where necessary.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Initial phenotypic characterization of WS-5
WS-5 was isolated from a mini-Tn5 mutagenesis of the WS strain and is defective in expression of the wrinkled colony morphology typical of wrinkly spreaders (Spiers et al., 2002). As the first step in this work, we characterized the phenotype of WS-5 on agar plates and in liquid microcosms. WS-5 colonies on both KB and LB agar were smooth-like and did not show the normal wrinkled colony morphology of the WS. After 1 day, the colonies typically were smaller and more waxy-looking than those produced by the SM strain, and over a period of 2–3 days became less SM-like but never fully WS. WS-5 colonies stained orange with CR on agar plates, and WS-5 produced very weak biofilms in KB microcosms. Examination of Calcofluor-stained colony and biofilm material by fluorescent microscopy confirmed that WS-5 expressed cellulose.

Identification and analysis of the mini-Tn5 insertion site in WS-5
In order to determine the location and genetic identity of the mini-Tn5 insertion site in WS-5, we screened a P. fluorescens SBW25 cosmid library for clones that complemented WS-5 in trans and restored the WS phenotype on KB agar plates. One cosmid was isolated (pAS256) and restriction analysis indicated that it contained a ~20 kb fragment, which was then randomly subcloned and end-sequences obtained. This sequence-sampling allowed identification of two well-conserved gene clusters at either end of the cosmid insert: the gsv glycine cleavage system (Okamura-Ikeda et al., 1993), and tol-pal (also referred to as tol-oprL) (Sturgis, 2001). When located on the unfinished SBW25 genome, the sequences identified a single contig covering the entire cosmid insert region. Using a mini-Tn5-specific primer and nested-PCR sequencing, the insertion site of mini-Tn5 in WS-5 was determined immediately upstream of ybgC, the first gene in the tol-pal cluster (i.e. WS-5 was WS tol : : mini-Tn5) (Fig. 1).

Tol-Pal proteins are involved in the normal interaction of the inner and outer membranes and are found throughout the eubacteria (Sturgis, 2001; Lazzaroni et al., 1999; Lloubes et al., 2001). Tol-Pal system mutants typically show impaired control of membrane channels (resulting in problems with uptake or export, leakage of proteins from the cytoplasm, sensitivity to pH and osmotic stress) and the disruption of outer-membrane or cell-surface components, including a reduction in or loss of LPS expression (Gaspar et al., 2000).

From the position of the mini-Tn5 insertion site in WS-5 (WS tol : : mini-Tn5), it was clear that expression of ybgC-tolQRAB-pal-ybgF would be severely affected (the Tol-Pal system per se includes tolQRAB-pal; per contra no function has been reported for ybgC or ybgF: Lloubes et al., 2001). In order to confirm that the disruption of a known tol gene was sufficient to explain the WS-5 phenotype, we made a WS tolA mutant in which tolA was disrupted and downstream tolB-ybgF expression compromised (Fig. 1). WS tolA showed a WS-5-like colony morphology, expressed cellulose and produced weak biofilms in KB microcosms, suggesting that it was the disruption of the functionally known tol genes, rather than the functionally uncharacterized ybgC gene, that was responsible for the phenotype of WS-5. In contrast to WS-5 (WS tol : : mini-Tn5), WS tolA in shaking cultures produced more floccular material, suggesting that the tolA mutation generated a more severe phenotype (through the inactivation of TolA and loss of Tol-Pal function) than the WS-5 mini-Tn5 insertion (in which tolQRAB-pal-ybgF expression was reduced, allowing some Tol-Pal function to be preserved). For this reason, the rest of our work focused on the comparative analysis of WS-5 alone.

Confirmation of the leaky-membrane phenotype
In order to determine whether WS-5 showed the leaky-membrane phenotype typical of Tol-Pal mutants, we assessed membrane integrity using the fluorescent DNA-binding dye propidium iodide (PI). This hydrophilic dye cannot pass through the bacterial membrane, and can only bind DNA if the membrane has been damaged. Fluorescent microscopy of exponential-phase WS-5 cells grown in KB with PI showed that a significant number of cells stained with the dye (and cells were misshapen), indicating that WS-5 shows the expected Tol-Pal leaky-membrane phenotype. In contrast, most WS cells did not stain with PI and showed no evidence of misshapen cell morphologies. We used FACS analysis to quantify the relative differences in PI uptake, and found that WS-5 (WS tol : : mini-Tn5) staining was 3·35-fold greater than the mean uptake for SM, WS, WS-4 (WS wspR : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) cells.

Some Tol-Pal system mutants are able to utilize small molecular mass molecules as sole carbon sources that diffuse across the damaged membrane, which otherwise could not cross into the cytoplasm, where they are metabolized (Llamas et al., 2003). In a test of this particular phenotype, we found that WS-5 was able to grow on minimal agar supplemented with sucrose, whereas neither the SM nor WS strains could use the disaccharide as the sole carbon source. These findings are all consistent with the leaky-membrane phenotype expected from the mini-Tn5 insertion site in WS-5.

WS-5 is insensitive to WspR reactivation of the WS phenotype
In order to determine how a disruption of the Tol-Pal system might result in weak biofilm formation by WS-5, we first examined whether the WS phenotype in WS-5 could be recovered by WspR expressed in trans. Previous work has identified WspR as a regulator of both cellulose and attachment-factor expression (Spiers et al., 2002, 2003). When expressed in trans in SM, both wild-type WspR (WspR12) and the constitutively active mutant WspR19 produce WS-like colony morphologies (Goymer, 2002). We determined the colony phenotypes of SM and WS-5 (WS tol : : mini-Tn5) carrying pVSP61-{Omega}TcR, pVSP61-wspR12-{Omega}TcR and pVSP61-wspR19-{Omega}TcR on both KB and LB agar plates. The control plasmid pVSP61-{Omega}TcR did not alter either SM or WS-5 colony morphologies, and both pVSP61-wspR12-{Omega}TcR and pVSP61-wspR19-{Omega}TcR produced WS-like colonies in SM. In contrast, neither pVSP61-wspR12-{Omega}TcR nor pVSP61-wspR19-{Omega}TcR altered the colony morphology of WS-5.

These findings indicate that WS-5 is not a mutant in which the WS phenotype has been turned off, or in which the signal that activates the WS phenotype has been interrupted. It therefore seemed most likely that a third component required for normal WS biofilm formation was no longer available. Of all of the phenotypes associated with Tol-Pal system mutants, we considered that the loss of LPS expression was most likely to have an impact on biofilm formation. We therefore directly tested this hypothesis by examining whether LPS expression was reduced in WS-5, whether WS-5 cells showed altered hydrophobicity, and whether WS biofilm strength could be changed by chemical interference targeted at LPS–cellulose fibre–attachment factor interactions.

Expression of LPS
LPS expression is strongly reduced in Tol-Pal mutants (Gaspar et al., 2000). In order to determine whether WS-5 showed a similar reduction in LPS expression, we prepared LPS EDTA-extracts from overnight KB cultures in which cell densities had been first equalized. These extracts were electrophoresed using DOC-PA gels that were then silver-stained to reveal the major LPS bands (Fig. 2). WS-5 (WS tol : : mini-Tn5) expressed insignificant amounts of LPS when compared with either WS, WS-4 (WS wspR : : mini-Tn5) or WS-18 (WS wssF : : mini-Tn5). LPS levels in WS and WS-5 were also investigated using the mAb BC12-CA4. Although the binding of mAb BC12-CA4 to P. fluorescens was weak, ELISA assays clearly showed a significantly greater (5·7x) mAb binding to WS than WS-5 cells (P=0·0384) ({Delta}A450 OD600–1±SE: WS, 0·554±0·022; WS-5, 0·097±0·011), further supporting our DOC-PAGE observations that WS-5 does not express detectable amounts of LPS.



View larger version (98K):
[in this window]
[in a new window]
 
Fig. 2. LPS expression in WS-5 is strongly reduced in comparison with the wrinkly spreader and other mutants. From left to right: LPS samples from WS, WS-4 (WS wspR : : mini-Tn5), WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5). The major LPS bands are indicated by triangles. LPS samples were prepared from overnight KB cultures adjusted to give the same OD600 cell density. Samples were extracted using EDTA, electrophoresed in an 18 % DOC-polyacrylamide gel and then silver-stained to detect LPS.

 
Relative hydrophobicity of WS strains
LPS expression is known to affect the surface charge and/or relative hydrophobicity (Hr) of bacterial cells (Rocchetta et al., 1999). We used the microbial adhesion to hydrocarbon (MATH) assay to determine whether WS-5 cells showed altered Hr compared to WS. When the assay was performed in KB, WS, WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) showed significantly different Hr from one another (P=0·0005) (Hr±SE: WS, 0·0574±0·0113; WS-5, 0·0959±0·0100; WS-18, 0·0163±0·0050). By determining Hr in KB, we have measured the real difference in cell hydrophobicities in the same environment in which the biofilms are produced. In contrast, when the assay was performed using potassium phosphate buffer (to determine an arbitrary difference in Hr), WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) Hr values were similar (P=0·9800), but different from WS (P=0·001) (Hr±SE: WS, 0·9097±0·0075; WS-5, 0·9938±0·0054; WS-18, 0·9934±0·0147). These findings show that in the context of the KB microcosm, the surface charge and/or hydrophobicity of WS, WS-5 and WS-18 bacterial cells differ, and changes in the chemical environment cause changes in relative hydrophobicity.

Comparison between WS and WS-5 biofilms
In order to quantify the differences in the physical characteristics between the WS-5 biofilm and those produced by WS and other mutants, we determined the relative attachment and maximum deformation mass (MDM, strength) of 3-day-old KB-grown biofilms. Although the ability of WS-5 (WS tol : : mini-Tn5) to attach to the surface of the glass microcosms was not significantly different from that of WS or WS-18 (WS wssF : : mini-Tn5) (P=0·0547) (Fig. 3a), the absolute strength of the WS-5 biofilm was substantially less than that of WS or WS-18 (P<0·0001) (Fig. 3b). In order to determine whether the reduced strength of WS-5 biofilms was due to a decreased rate of growth, we measured growth in KB over 24 h for WS, WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) ({Delta}OD600 h–1±SE: WS, 0·058±0·001; WS-5, 0·071±0·001; WS-18, 0·068±0·001). The three growth rates were not significantly different (P=0·4254), indicating that the reduced WS-5 biofilm strength could not be due to a decreased rate of growth. Finally, we also measured the relative amounts of cellulose produced by each strain using a CR-binding assay. WS and WS-5 (WS tol : : mini-Tn5) bound similar amounts of CR (P=0·0965) and slightly less (~1·26x) than WS-18 (WS wssF : : mini-Tn5) (P=0·0049) ({Delta}A490 OD600–1±SE: WS, 0·317±0·019; WS-5, 0·368±0·019; WS-18, 0·464±0·033). CR binds components other than cellulose, and in the case of WS and WS mutants, CR is bound by cellulose and attachment factor (Spiers et al., 2003). Having demonstrated that WS, WS-5 and WS-18 show the same degree of attachment and similar levels of CR-binding, we therefore conclude that the three strains express similar levels of cellulose. This finding indicates that the reduced strength of WS-5 biofilms is not due to a reduced level of cellulose expression.



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 3. Relative attachment and strength of WS-5 A–L biofilms. (a) Relative attachment for SM, WS, WS-4 (WS wspR : : mini-Tn5), WS-5 (WS tol : : mini-Tn5), WS-13 (WS wssB : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) (OD570 OD570WS–1); (b) Maximum deformation mass (MDM) for WS, WS-5, and WS-18 (SM, WS-4 and WS-13 do not produce biofilms). Microcosms were incubated at 28 °C for 3 days before assay. Means±SE are shown.

 
Recruitment to the A–L interface
The differences in relative hydrophobicity (Hr) between WS, WS-5 and WS-18 cells might affect A–L biofilm strength by enhancing the recruitment of cells to the meniscus region where initial attachment to the glass surface of microcosms occurs, and by maintaining cells within the developing biofilm during growth. We examined differences in recruitment by monitoring OD600 at the bottom of the liquid column using standard spectrophotometer cuvettes. We reasoned that bacterial cells may adhere to the surface of the cuvette, or remain a homogeneous suspension of cells, and thus show no change in OD600. Alternatively, the cells may be recruited to the surface (by chemotaxis or random motion) and maintained through attachment to the walls or aggregation with other bacteria at the surface. First, however, we determined that WS, WS-4 (WS wspR : : mini-Tn5), WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) cells were motile by direct microscopic examination, and significantly different from the non-chemotactic AS24 (an evolved SM cheA) (P<0·0001) (migration through soft agar, mm h–1: WS, 0·681±0·019; WS-5, 0·394±0·019; WS-18, 0·594±0·019; cf. WS-4, which is not impeded by attachment factor, 0·836±0·022; and AS24, 0·125±0·019). The relative OD600 at the bottom of KB liquid columns was found to decrease for WS, WS-5 and WS-18 as cells migrated towards the surface over a period of 1 h (Fig. 4). WS cells showed a significantly greater (~1·5x) level of recruitment to the surface than either WS-5 (WS tol : : mini-Tn5) or WS-18 (WS wssF : : mini-Tn5) (P=0·006) (relative OD600±SE: WS, 0·8877±0·0013; WS-5, 0·9224±0·0066; WS-18, 0·9299±0·0011). In contrast, WS-4 (WS wspR : : mini-Tn5) cells, unable to express cellulose or attachment factor, showed a substantially different behaviour in which cells initially adhered to the sides of the cuvette before slowly migrating towards the surface after 20 min. We also measured the ability of WS, WS-5 and WS-18 biofilms to maintain cells within the developing biofilm during growth. After 3 days, the proportion of cells found in the liquid column under the biofilm was significantly greater (3·7–4·4x) for WS-5 (WS tol : : mini-Tn5) and WS-18 (WS wssF : : mini-Tn5) than for WS (P=0·0001) (though the total OD600 achieved by WS, WS-5 and WS-18 was the same, P=0·5022). From these findings it is clear that the weaker WS-5 and WS-18 biofilms are unable to recruit and maintain cells within the biofilm as efficiently as the WS.



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 4. WS-5 and WS-18 are unable to recruit cells into the developing biofilm as efficiently as WS. Recruitment to the surface through the liquid KB column, shown by a decrease in OD600, is faster for WS ({bullet}) than for WS-5 (WS tol : : mini-Tn5) ({triangleup}) or WS-18 (WS wssF : : mini-Tn5) ({circ}). WS-4 cells (WS wspR : : mini-Tn5) ({square}) show a different behaviour, in which cell densities initially increase at the bottom of the cuvette (these cells may be attached to the surface of the cuvette or they might remain in suspension), and then slowly decrease over time. Recruitment assays were carried out at 20–22 °C, during which 3–5 % growth-dependent increases in final OD600 were seen. Mean±SE values for 40–60 min are shown on the right.

 
Chemical interference of interactions amongst biofilm components
Previously, we have shown that the presence of the dye CR resulted in a significant decrease in the strength of WS biofilms by interfering between the normal cellulose fibre and/or attachment factor interactions (Spiers et al., 2003). In order to further elucidate interactions between biofilm components, we tested WS, WS-5 and WS-18 biofilm strengths in the presence of Ca2+, Fe3+, EDTA and various diazo dyes structurally related to CR. The metal anions are expected to bind EPS/LPS and alter the normal cell-surface charge distribution; EDTA chelates Mg2+, which is known to have a major role in LPS charge-neutralization (Groisman et al., 1997; Rocchetta et al., 1999). In contrast, the diazo dyes bind cellulose differentially due to slight structural variations (Kai & Mondal, 1997), and are expected to interact similarly with the WS attachment factor.

In initial tests, we found that both Ca2+ and Fe3+ severely affected growth in KB microcosms, whereas the addition of more Mg2+ (KB contains ~6 mM Mg2+) had no effect on MDM, and therefore these were not tested further. However, at low levels of EDTA (2–10 mM), a significant decrease in WS and WS-18 (WS wssF : : mini-Tn5) MDM was observed, whereas WS-5 (WS tol : : mini-Tn5) MDM was not significantly affected even at 10 mM EDTA (P=0·1333) (Fig. 5a). (5 mM EDTA had no effect on maximum growth rate, but 10 mM EDTA resulted in a 0·2x reduction in growth rate.) EDTA might act to prevent irreversible interactions that occur during biofilm development, or it might act to destabilize reversible interactions that maintain biofilm strength. To determine which of these possibilities was more likely, we tested the MDM of mature WS biofilms 2 h after EDTA had been added to a final concentration of 10 mM. There was no significant difference in MDM between the EDTA-treated biofilms and water-treated negative controls (P=0·6708), indicating that EDTA affects the establishment of irreversible interactions that form during biofilm development, rather than destabilizing reversible interactions (i.e. the constant association and disassociation of cellulose fibres, attachment factor and LPS) that might occur in the mature biofilm.



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 5. EDTA and diazo dyes interfere with normal interactions between biofilm components and result in altered biofilm strengths. (a) The addition of EDTA to standard KB microcosms had a significant effect on the relative maximum deformation mass (MDM) of both WS ({bullet}) and WS-18 (WS wssF : : mini-Tn5) ({circ}) biofilms, but no significant effect on the very weak WS-5 (WS tol : : mini-Tn5) ({triangleup}) biofilm except at high concentrations of EDTA. (b) Diazo dyes significantly altered the relative MDM of WS (dark bars) and WS-18 (white bars) biofilms, but had little impact on WS-5 (light grey bars) biofilms. Ctrl, no dye added. CR is also known as DR 28. Microcosms were incubated at 28 °C for 3 days before assay. Mean±SE of relative MDM shown for both assays. Note the log scale in (b).

 
We also tested CR and related diazo dyes (DR 2, DB 1, 14, 15 and 53) (Fig. 5b), having first determined that no dye showed a toxic effect on growth at the concentration tested. None of the dyes increased the relative MDM of WS-5 (WS tol : : mini-Tn5) (P=0·9804). However, DB 1 and DB 53 differentiated WS-5, DB14 differentiated WS-18 (WS wssF : : mini-Tn5), and DR 2 differentiated WS from the other two strains. These findings confirmed our expectations that EDTA and the diazo dyes would differentiate between WS, WS-5 and WS-18 biofilms through differential interference of interactions between biofilm components. This strongly suggests that biofilm strength is the result of multiple interactions between biofilm components, and that WS-5 biofilms lack some component found in both WS and WS-18 biofilms.

Complementation of WS-5 biofilms with WS-4
If biofilm development and final strength are the result of multiple cellulose fibre–attachment factor–LPS interactions, we reasoned that the weak WS-5 biofilm may be complemented by a second strain capable of expressing LPS, but which cannot express cellulose or attachment factor (e.g. WS-4). In order to test this expectation, we determined the MDM of KB-grown WS-4 (WS wspR : : mini-Tn5)/WS-5 (WS tol : : mini-Tn5) mixed biofilms (Fig. 6). Biofilms produced from WS-5 alone or 1 : 9 WS-4/WS-5 were significantly weaker than 1 : 4 WS-4/WS-5-mixed biofilms (P=0·0465, 0·0486), but none of the mixed biofilms reached the strength of WS biofilms. Nevertheless, this result demonstrates that WS-4 can partially complement WS-5 biofilm strength. Furthermore, similar partial complementation of JB01 (SM NPTII : : wss) biofilms with WS-13 (WS wssB : : mini-Tn5) was also seen (JB01 is an SM derivative overexpressing cellulose but not expressing attachment factor, which produces a particularly weak biofilm: Spiers et al., 2002). When JB01 was complemented with WS-13 expressing attachment factor, the MDM of the 1 : 1 WS-13/JB01 biofilm was significantly greater than that of a biofilm of JB01 alone (P<0·0001). In both the WS-4/WS-5 and WS-13/JB01 tests, the increased MDM of the mixed biofilms is not the result of differences in growth rates, as in both cases, significant complementation was only seen with higher initial ratios of the strain that could not produce a biofilm alone (i.e. the significant comparisons are between 1 : 9 and 1 : 4 WS-4/WS-5, and between 1 : 9 and 1 : 1 WS-13/JB01). These findings strongly suggest that the strength of the WS biofilm is due to the combination of cellulose, attachment factor and LPS, and that partial complementation can be achieved by expressing all three components by two strains in different combinations.



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 6. WS mutant A–L biofilms can be partially complemented and show significant increases in strength. The WS-5 (WS tol : : mini-Tn5) biofilm can be complemented by WS-4 (WS wspR : : mini-Tn5), leading to a 1·7x increase in maximum deformation mass (MDM) (white bars). WS-5 expresses attachment factor and cellulose, but not LPS; WS-4 expresses LPS, but not cellulose or attachment factor. Similarly, the JB01 (SM NPTII : : wss) biofilm is complemented by WS-13 (WS wssB : : mini-Tn5), leading to a 3·9x increase in MDM (grey bars). JB01 expresses cellulose and LPS, but not attachment factor; WS-13 expresses LPS and attachment factor, but not cellulose. Neither WS-4 nor WS-13 produces biofilms. The MDM of WS in this experiment was 0·151±0·93 g (dark bar off-scale). Microcosms were inoculated with single strains, or with mixtures of strains, such that each test used the same total number of cells. Microcosms were incubated at 20–22 °C for 3 days before assay. Means±SE are shown.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In previous analyses of factors required for development of the WS A–L biofilm, we demonstrated that partially acetylated cellulose and a fimbrial-like attachment factor were significant determinants of the strength and structural integrity of the WS biofilm (Spiers et al., 2003). In this work, we have extended our analysis to the wrinkly spreader-defective mini-Tn5 mutant WS-5, which produces weak biofilms despite the production of partially acetylated cellulose and fimbrial attachment factor.

The site of the mini-Tn5 insertion in WS-5 was shown to be immediately upstream of ybgC, the first gene in the highly conserved tol-pal gene cluster, and it exerts polar effects on downstream gene expression. The Tol-Pal system proteins are involved in maintaining the correct functional relationship between the inner and outer membranes, and mutants often show a variety of pleiotropic effects, including the loss of LPS expression (Gaspar et al., 2000). We confirmed that this was the case in WS-5 by DOC-PAGE and ELISA assays.

LPS plays a major role in determining the surface-charge or relative hydrophobicity (Hr) of the cell (Rocchetta et al., 1999), and is involved in bacterial attachment to surfaces and biofilm formation; it differentiates between biofilm and planktonic cells, as well as affecting colony morphology in a number of bacteria (Giwercman et al., 1992; Genevaux et al., 1999; Mireles et al., 2001; Nesper et al., 2001; Landini & Zehnder, 2002; de Lima Pimenta et al., 2003; Rashid et al., 2003). WS-5 cells showed a significantly different Hr from WS cells, and maintenance of WS-5 cells in the biofilm was less efficient than that of WS. In addition, the strength of the WS biofilm was more sensitive to chemical interference than that of the WS-5 biofilm. Each of these findings, along with the demonstration that the WS-5 biofilm can be partially complemented by an LPS-expressing strain, strongly suggests that LPS-dependent and charge-sensitive interactions are important in WS biofilm development, and that complex interactions between cellulose fibres, attachment factor and LPS determine the final strength of WS biofilms.

We have previously noted that the P. fluorescens WS A–L biofilm and colony morphology are similar to those produced by Escherichia coli and Salmonella sp. (Spiers et al., 2002), and in each case, cellulose fibres and curli/Tafi fimbriae are required for both biofilm strength and the rdar colony morphology (Römling & Rohde, 1999; Zogaj et al., 2001; Solano et al., 2002). In S. enterica biofilms, Tafi fibres appear as a tangled amorphous matrix when cellulose is present, but when it is not, the fibres adopt a more normal, slightly curled linear structure (White et al., 2003). Further analysis of E. coli and S. enterica biofilms has also revealed the presence of a third matrix component, an anionic extracellular polysaccharide, which requires cellulose in order to maintain a close association with cells (White et al., 2003). These findings, along with our observations regarding the WS biofilm, indicate that the structure and physical properties of bacterial biofilms are the result of multiple interactions between various matrix components – the main EPS matrix fibres, proteinaceous attachment fibres (fimbriae and flagella), LPS and additional polysaccharides. In parallel research, we are undertaking a comprehensive screen for new WS mutants, in which ISphoA/hah disruptions of the cellulose acetylation genes, putative LPS biosynthesis and membrane-associated genes have been identified (S. Gehrig, A. Spiers & P. Rainey, work in progress), further underlining the importance of these interactions in the WS phenotype.

LPS is generally anchored to the bacterial outer membrane via lipid A (Rocchetta et al., 1999). However, LPS is also known to be released from cells during normal growth, and cell-free LPS accumulates in cultures after cell lysis (Cadieux et al., 1983; Ishiguro et al., 1986; Al-Tahhan et al., 2000). This suggests that some of the cellulose fibre–attachment factor–LPS interactions important to WS biofilm strength may be cell-independent, insofar that once LPS has been produced and released at one location, bacterial cells may not be required to remain in place to maintain biofilm strength. Indeed, the WS biofilm might result from the aggregation of locally expressed and largely cell-free cellulose, attachment factor and released LPS, with the bacterial cells free to move within the biofilm as it develops and as environmental conditions change. This possibility is supported by WS biofilm microscopy, in which few bacteria were found to be closely associated with the cellulose matrix, and most found to be mobile within the spaces of the biofilm (Spiers et al., 2003).

Biofilm matrices are known to be chemically complex, with 85–98 % of the total organic carbon present as excreted polymers and products from cell lysis, and the balance in intact cells (Sutherland, 2001a). It is becoming increasingly apparent that the physical structure of many biofilms is not primarily the result of the expression of one matrix component, but of several interacting elements. Furthermore, although the main component might be specifically expressed during biofilm development, the expression of other components may not be restricted to the biofilm. The added complexity of matrix components and expression patterns has an obvious impact on the study of biofilm development. While the formation of biofilms may confer a growth advantage at surfaces, biofilm matrix compounds need not be biofilm-specific; for example, flagella and pili play a role in motility as well as initial attachment, and EPS glycocalyx may play a protective role as well as contribute to formation of the biofilm matrix.

The finding that the structural aspect of biofilm matrices is more complex than originally thought has implications for the ‘architectural’ nature of biofilm development (Wimpenny et al., 2000; Ghigo, 2003). If biofilm development is a genetically programmed growth mode, then far more biosynthetic pathways would need to be controlled than currently recognized. On the other hand, if biofilm growth is a consequence of bacterial attachment, the resulting community structure may be better defined by a small number of specific biosynthetic pathways, plus a number of less-specific systems that contribute to the overall physical-chemical structure of the biofilm matrix. If this is so, then we predict that biofilm growth will show a variable requirement for secondary matrix components and a degree of redundancy, as these will not be uniquely required and may be complemented by other components.

In the case of natural biofilms, ranging from those growing on dental surfaces or other human tissues to true water-, soil- or plant-associated environmental biofilms (Davey & O'Toole, 2000; Wilson, 2001; Morris & Monier, 2003), different members of the biofilm may contribute to different parts of the biofilm matrix. Given the continuously changing composition of biofilms during establishment, growth and maturity, a dynamic and complex physical-chemical matrix can only be expected. Many of the phenotypes associated with biofilms, such as increased resistance to stress and antimicrobial agents, may be the consequence of the microenvironment heterogeneity within biofilms (Ghigo, 2003). The complexity of both the biofilm matrix and community underlies the ecological success of this type of assemblage, and may also explain the value of such structures to both pathogenic and opportunistic bacteria.


   ACKNOWLEDGEMENTS
 
We thank J. Stansfield for her technical assistance and A. Whiteley for his help in the FACS analyses. mAb BC12-CA4 was kindly made available by M. Dewey. DNA sequences obtained in this work were used to interrogate the unfinished P. fluorescens SBW25 genome sequence (Wellcome Trust Sanger Institute, UK) to aid gene identification. Funding for this work came in part from the BBSRC (UK).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Al-Tahhan, R. A., Sandrin, T. R., Bodour, A. A. & Maier, R. M. (2000). Rhamnolipid-induced removal of lipopolysaccharide from Pseudomonas aeruginosa: effect on cell surface properties and interaction with hydrophobic substrates. Appl Environ Microbiol 66, 3262–3268.[Abstract/Free Full Text]

Cadieux, J. E., Kuzio, J., Milazzo, F. H. & Kropinski, A. M. (1983). Spontaneous release of lipopolysaccharide by Pseudomonas aeruginosa. J Bacteriol 155, 817–825.[Medline]

Costerton, J. W., Lewandowski, Z., Cladwell, D. E., Korber, D. R. & Lappin-Scott, H. M. (1995). Microbial biofilms. Annu Rev Microbiol 49, 711–745.[CrossRef][Medline]

Dalton, H. M. & March, P. E. (1998). Molecular genetics of bacterial attachment and biofouling. Curr Opin Biotechnol 9, 252–255.[CrossRef][Medline]

Davey, M. E. & O'Toole, G. A. (2000). Microbial biofilms: from ecology to molecular genetics. Microbiol Mol Biol Rev 64, 847–867.[Abstract/Free Full Text]

De Lima Pimenta, A., Di Martino, P., Le Bouder, E., Hulen, C. & Blight, M. A. (2003). In vitro identification of two adherence factors required for in vivo virulence of Pseudomonas fluorescens. Microbes Infect 13, 1177–1187.[CrossRef]

Dennis, J. J. & Zylstra, G. J. (1998). Improved antibiotic-resistance cassettes through restriction site elimination using Pfu DNA polymerase PCR. Biotechniques 25, 772–776.[Medline]

Donlan, R. M. (2002). Biofilms: microbial life on surfaces. Emerg Infect Dis 8, 881–890.[Medline]

Dunne, W. M. (2002). Bacterial adhesion: seen any good biofilms lately? Clin Microbiol Rev 15, 155–166.[Abstract/Free Full Text]

Gaspar, J. A., Thomas, J. A., Marolda, C. L. & Valvano, M. A. (2000). Surface expression of O-specific lipopolysaccharide in Escherichia coli requires the function of the TolA protein. Mol Microbiol 38, 262–275.[CrossRef][Medline]

Genevaux, P., Bauda, P., DuBow, M. S. & Oudega, B. (1999). Identification of Tn10 insertions in the rfaG, rfaP, and galU genes involved in lipopolysaccharide core biosynthesis that affect Escherichia coli adhesion. Arch Microbiol 172, 1–8.[CrossRef][Medline]

Ghigo, J.-M. (2003). Are there biofilm-specific physiological pathways beyond a reasonable doubt? Res Microbiol 154, 1–8.[CrossRef][Medline]

Giwercman, B., Fomsgaard, A., Mansa, B. & Hoiby, N. (1992). Polyacrylamide gel electrophoresis analysis of lipopolysaccharide from Pseudomonas aeruginosa growing planktonically and as biofilm. FEMS Microbiol Immunol 4, 225–229.[Medline]

Götz, F. (2002). Staphylococcus and biofilms. Mol Microbiol 43, 1367–1378.[CrossRef][Medline]

Goymer, P. J. (2002). The role of the WspR response regulator in the adaptive evolution of experimental populations of Pseudomonas fluorescens SBW25. DPhil thesis, University of Oxford.

Groisman, E. A., Kayser, J. & Soncini, F. C. (1997). Regulation of polymyxin resistance and adaptation to low-Mg2+ environments. J Bacteriol 179, 7040–7045.[Abstract/Free Full Text]

Hall-Stoodley, L. & Stoodley, P. (2002). Developmental regulation of microbial biofilms. Curr Opin Biotechnol 13, 228–233.[CrossRef][Medline]

Hall-Stoodley, L., Costerton, J. W. & Stoodley, P. (2004). Bacterial biofilms: survival and propagation on surfaces from the environment to infectious diseases. Nat Rev Microbiol 2, 95–108.[CrossRef][Medline]

Ishiguro, E. E., Vanderwel, D. & Kusser, W. (1986). Control of lipopolysaccharide biosynthesis and release by Escherichia coli and Salmonella typhimurium. J Bacteriol 168, 328–333.[Medline]

Kai, A. & Mondal, I. H. (1997). Influence of substituent of direct dye having bisphenylenebis(azo) skeletal structure on structure of nascent cellulose produced by Acetobacter xylinum [I]: different influence of Direct Red 28, Blue 1 and 15 on nascent structure. Int J Biol Macromol 20, 221–231.[CrossRef][Medline]

King, E. O., Ward, M. K. & Raney, D. C. (1954). Two simple media for the demonstration of pyocyanin and fluorescin. J Lab Clin Med 44, 301–307.[Medline]

Landini, P. & Zehnder, A. J. (2002). The global regulatory hns gene negatively affects adhesion to solid surfaces by anaerobically grown Escherichia coli by modulating expression of flagella genes and lipopolysaccharide production. J Bacteriol 184, 1522–1529.[Abstract/Free Full Text]

Lappin-Scott, H. M. & Bass, C. (2001). Biofilm formation: attachment, growth and detachment of microbes from surfaces. Am J Infect Control 29, 250–261.[CrossRef][Medline]

Lazzaroni, J. C., Germon, P., Ray, M. C. & Vianney, A. (1999). The Tol proteins of Escherichia coli and their involvement in the uptake of biomolecules and outer membrane stability. FEMS Microbiol Lett 177, 191–197.[CrossRef][Medline]

Llamas, M. A., Rodríguez-Herva, J. J., Hancock, R. E. W., Bitter, W., Tommassen, J. & Ramos, J. L. (2003). Role of Pseudomonas putida tol-oprL gene products in uptake of solutes through the cytoplasmic membrane. J Bacteriol 185, 4707–4716.[Abstract/Free Full Text]

Lloubes, R., Cascales, E., Walburger, A., Bouveret, E., Lazdunski, C., Bernadac, A. & Journet, L. (2001). The Tol-Pal proteins of the Escherichia coli cell envelope: an energized system required for outer membrane integrity? Res Microbiol 152, 523–529.[CrossRef][Medline]

Meyer, U. & Dewey, F. M. (2000). Efficacy of different immunogens for raising monoclonal antibodies to Botrytis cinerea. Mycol Res 104, 979–987.[CrossRef]

Mireles, J. R., Toguchi, A. & Harshey, R. M. (2001). Salmonella enterica serovar typhimurium swarming mutants with altered biofilm-forming abilities: surfactin inhibits biofilm formation. J Bacteriol 183, 5848–5854.[Abstract/Free Full Text]

Morris, C. E. & Monier, J.-M. (2003). The ecological signifcance of biofilm formation by plant-associated bacteria. Annu Rev Phytopathol 41, 429–453.[CrossRef][Medline]

Nesper, J., Lauriano, C. M., Klose, K. E., Kapfhammer, D., Kraiss, A. & Reidl, J. (2001). Characterization of Vibrio cholerae O1 El tor galU and galE mutants: influence on lipopolysaccharide structure, colonization, and biofilm formation. Infect Immun 69, 435–445.[Abstract/Free Full Text]

Okamura-Ikeda, K., Ohmura, Y., Fujiwara, K. & Motokawa, Y. (1993). Cloning and nucleotide sequence of the gcv operon encoding the Escherichia coli glycine-cleavage system. Eur J Biochem 216, 539–548.[Abstract]

Rainey, P. B. (1999). Adaptation of Pseudomonas fluorescens to the plant rhizosphere. Environ Microbiol 1, 243–257.[CrossRef][Medline]

Rainey, P. B. & Bailey, M. J. (1996). Physical map of the Pseudomonas fluorescens SBW25 chromosome. Mol Microbiol 19, 521–533.[CrossRef][Medline]

Rainey, P. B. & Travisano, M. (1998). Adaptive radiation in a heterogeneous environment. Nature 394, 69–72.[CrossRef][Medline]

Rainey, P. B. & Rainey, K. (2003). Evolution of cooperation and conflict in experimental bacterial populations. Nature 425, 72–74.[CrossRef][Medline]

Rashid, M. H., Rajanna, C., Ali, A. & Karaolis, D. K. (2003). Identification of genes involved in the switch between the smooth and rugose phenotypes of Vibrio cholerae. FEMS Microbiol Lett 227, 113–119.[CrossRef][Medline]

Reuhs, B. L., Geller, D. P., Kim, J. S., Fox, J. E., Kolli, V. S. K. & Pueppke, S. G. (1998). Sinorhizobium fredii and Sinorhizobium meliloti produce structurally conserved lipopolysaccharides and strain-specific K antigens. Appl Environ Microbiol 64, 4930–4938.[Abstract/Free Full Text]

Rocchetta, H. L., Burrows, L. L. & Lam, J. S. (1999). Genetics of O-antigen biosynthesis in Pseudomonas aeruginosa. Microbiol Mol Biol Rev 63, 523–553.[Abstract/Free Full Text]

Römling, U. & Rohde, M. (1999). Flagella modulate the multicellular behaviour of Salmonella typhimurium on the community level. FEMS Microbiol Lett 180, 91–102.[CrossRef][Medline]

Simon, R., Priefer, U. & Puhler, A. (1983). A broad host range mobilisation system for in vivo genetic engineering: random and site-specific transposon mutagenesis in gram-negative bacteria. Biotechnology 1, 784–791.[CrossRef]

Solano, C., Garcia, B., Valle, J., Berasain, C., Ghigo, J.-M., Gamazo, C. & Lasa, I. (2002). Genetic analysis of Salmonella enteritidis biofilm formation: critical role of cellulose. Mol Microbiol 43, 793–808.[CrossRef][Medline]

Spiers, A. J., Kahn, S. G., Travisano, M., Bohannon, J. & Rainey, P. B. (2002). Adaptive divergence in experimental populations of Pseudomonas fluorescens. 1. Genetic and phenotypic bases of wrinkly spreader fitness. Genetics 161, 33–46.[Abstract/Free Full Text]

Spiers, A. J., Bohannon, J., Gehrig, S. & Rainey, P. B. (2003). Colonisation of the air-liquid interface by the Pseudomonas fluorescens SBW25 wrinkly spreader requires an acetylated form of cellulose. Mol Microbiol 50, 15–27.[CrossRef][Medline]

Stoodley, P., Sauer, K., Davies, D. G. & Costerton, J. W. (2002). Biofilms as complex differentiated communities. Annu Rev Microbiol 56, 187–209.[CrossRef][Medline]

Sturgis, J. N. (2001). Organisation and evolution of the tol-pal gene cluster. J Mol Microbiol Biotechnol 3, 113–122.[Medline]

Sutherland, I. W. (2001a). The biofilm matrix – an immobilized but dynamic microbial environment. Trends Microbiol 9, 222–227.[CrossRef][Medline]

Sutherland, I. W. (2001b). Biofilm exopolysaccharides: a strong and sticky framework. Microbiology 147, 3–9.[Medline]

Van der Mei, H. C. & Busscher, H. J. (2001). Electrophoretic mobility distributions of single-strain microbial populations. Appl Environ Microbiol 67, 491–494.[Free Full Text]

White, A. P., Gibson, D. L., Collinson, S. K., Banser, P. A. & Kay, W. W. (2003). Extracellular polysaccharides associated with thin aggregative fimbriae of Salmonella enterica serovar enteritidis. J Bacteriol 185, 5398–5407.[Abstract/Free Full Text]

Wilson, M. (2001). Bacterial biofilms and human disease. Sci Prog 84, 235–254.[Medline]

Wimpenny, J., Manz, W. & Szewzyk, U. (2000). Heterogeneity in biofilms. FEMS Microbiol Rev 24, 661–671.[CrossRef][Medline]

Zogaj, X., Nimtz, M., Rohde, M., Bokranz, W. & Römling, U. (2001). The multicellular morphotypes of Salmonella typhimurium and Escherichia coli produce cellulose as the second component of the extracellular matrix. Mol Microbiol 39, 1452–1463.[CrossRef][Medline]

Received 22 February 2005; revised 13 July 2005; accepted 13 July 2005.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Spiers, A. J.
Articles by Rainey, P. B.
Articles citing this Article
PubMed
PubMed Citation
Articles by Spiers, A. J.
Articles by Rainey, P. B.
Agricola
Articles by Spiers, A. J.
Articles by Rainey, P. B.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2005 Society for General Microbiology.