Editor: Kornelia Smalla
The GacS sensor kinase controls phenotypic reversion of small colony variants isolated from biofilms of Pseudomonas aeruginosa PA14
Article first published online: 21 AUG 2006
FEMS Microbiology Ecology
Volume 59, Issue 1, pages 32–46, January 2007
How to Cite
Davies, J. A., Harrison, J. J., Marques, L. L. R., Foglia, G. R., Stremick, C. A., Storey, D. G., Turner, R. J., Olson, M. E. and Ceri, H. (2007), The GacS sensor kinase controls phenotypic reversion of small colony variants isolated from biofilms of Pseudomonas aeruginosa PA14. FEMS Microbiology Ecology, 59: 32–46. doi: 10.1111/j.1574-6941.2006.00196.x
- Issue published online: 21 AUG 2006
- Article first published online: 21 AUG 2006
- Received 29 March 2006; revised 21 June 2006; accepted 22 June 2006.First published online 21 August 2006.
- Pseudomonas aeruginosa;
- two-component signal transduction;
- quorum sensing;
- small colony variant;
- antimicrobial resistance
- Top of page
- Materials and methods
- Authors contribution
The GacS/GacA two-component regulatory system in pseudomonads regulates genes involved in virulence, secondary metabolism and biofilm formation. Despite these regulatory functions, some Pseudomonas species are prone to spontaneous inactivating mutations in gacA and gacS. A gacS− strain of Pseudomonas aeruginosa PA14 was constructed to study the physiological role of this sensor histidine kinase. This loss-of-function mutation was associated with hypermotility, reduced production of acylhomoserine lactones, impaired biofilm maturation, and decreased antimicrobial resistance. Biofilms of the gacS− mutant gave rise to phenotypically stable small colony variants (SCVs) with increasing frequency when exposed to silver cations, hydrogen peroxide, human serum, or certain antibiotics (tobramicin, amikacin, azetronam, ceftrioxone, oxacilin, piperacillin or rifampicin). When cultured, the SCV produced thicker biofilms with greater cell density and greater antimicrobial resistance than did the wild-type or parental gacS− strains. Similar to other colony morphology variants described in the literature, this SCV was less motile than the wild-type strain and autoaggregated in broth culture. Complementation with gacS in trans restored the ability of the SCV to revert to a normal colony morphotype. These findings indicate that mutation of gacS is associated with the occurrence of stress-resistant SCV cells in P. aeruginosa biofilms and suggests that in some instances GacS may be necessary for reversion of these variants to a wild-type state.
- Top of page
- Materials and methods
- Authors contribution
Pseudomonas aeruginosa is successful at adapting to a wide variety of environmental niches such as soil, water, plants and animals. GacS/GacA (global activator of antibiotic and cyanide synthesis) is one of 66 two-component regulatory systems that P. aeruginosa uses to alter its physiology to suit these diverse environmental conditions. This system regulates the expression of virulence factors, stress tolerance genes, enzymes for secondary metabolism, as well as periplasmic proteins important for motility (Goodman et al., 2004; Venturi, 2006). The response regulator GacA is vitally important for biofilm formation and maturation in multihost virulent P. aeruginosa PA14 (Parkins et al., 2001). This two-component system is highly conserved amongst a wide range of Pseudomonas species (de Souza et al., 2003) and has been well characterized in rhizosphere strains of P. syringae and P. chlororaphis (Spencer et al., 2003).
In natural environments or within a host, bacteria associate with surfaces to form polymer-enclosed biofilm structures that thwart biological or chemical removal. In later stages of development, community growth and behaviour is coordinated by quorum sensing, a process that relies on intercellular signalling by N-acyl-homoserine lactones (AHSLs) (recently reviewed by Juhas et al., 2005; Soberón-Chávez & Aguirre-Ramírez, 2005). In P. aeruginosa, GacA is a positive transcriptional regulator of the lasRI and rhlRI operons, which are responsible for the enzymes that synthesize N-3-oxo-dodecanoyl-homoserine lactone (3-oxo-C12-HSL) and N-butanoyl-homoserine lactone (C4-HSL), respectively. Loss-of-function mutations in gacS and/or gacA in Pseudomonas species reduce production of these autoinducers (Parkins et al., 2001; Chin-A-Woeng et al., 2005). In vitro, these quorum-sensing systems are pivotal for P. aeruginosa biofilm tolerance to hydrogen peroxide, aminoglycoside antibiotics, and polymorphonuclear leukocytes (Bjarnsholt et al., 2005). It is a paradox that, despite a role in stress tolerance and survival, spontaneous mutations in gacS and/or gacA have been observed in many pseudomonads under laboratory conditions as well as in the plant rhizosphere (Duffy & Defago, 2000; Bull et al., 2001; Sánchez-Contreras et al., 2002; van den Broek et al., 2005).
Mutation of gacS is not associated with a loss of fitness of pseudomonads in the rhizosphere (Schmidt-Eisenlohr et al., 2003). Using P. chlororaphis as an example, studies have suggested that mixtures of gacS− mutants with the wild-type population may enhance the survival of this bacterium in soil (Chancey et al., 2002). Preliminary evidence suggests that this may be linked to phenotypic variation (Sánchez-Contreras et al., 2002). Our research group has recently discovered that inactivation of gacS in P. chlororaphis gives rise to highly adherent small colony variants (SCVs) from aged biofilms exposed to silver cations (L.L.R. Marques, H. Ceri, A.J. Anderson, Y.C. Kim and M.E. Olson, unpublished data). These isolates are less motile and superior at forming biofilms, which may be an important process for root colonization (for a review, see Harrison et al., 2005c). This process has not been examined in the soil bacterium and opportunistic human pathogen Pseudomonas aeruginosa, and here we specifically hypothesize that GacS may be similarly involved in phenotypic variation. This may be important because GacS/GacA signalling in this microorganism has been implicated in attenuating virulence and establishing chronic infections in the cystic fibrosis (CF) lung (Goodman et al., 2004; Ventre et al., 2006). Further, the isolation of colony morphology variants with an increased ability for forming biofilms has been described for many laboratory and clinical strains of P. aeruginosa (Drenkard & Ausubel, 2002; Häußler et al., 2003; Häußler, 2004; Kirisitis et al., 2005).
In this study, we report the isolation of phenotypically stable SCVs from aged biofilms of multihost virulent Pseudomonas aeruginosa PA14 bearing an inactivating mutation in the sensor kinase gacS. These colony morphology variants were hyperadherent, less motile, and had a hyper-biofilm-forming phenotype compared with the wild-type strain. These variants also had elevated resistance to antimicrobials. Biofilms of PA14 gacS− gave rise to the SCV phenotype at a higher frequency (than growth controls) when exposed to some clinically used antibiotics, silver ions, or hydrogen peroxide. Furthermore, the phenotypic stability of the SCV strain suggests that GacS may control reversion of these colony morphology variants to a wild-type state.
Materials and methods
- Top of page
- Materials and methods
- Authors contribution
Bacterial strains and growth conditions
Bacterial strains used in this study are summarized in Table 1. All strains were stored at −70°C in Microbank™ vials (Pro-Lab Diagnostics, Toronto, Canada) according to the manufacturer's instructions. Unless otherwise noted, P. aeruginosa and Escherichia coli were grown in tryptic soy broth or Miller Luria–Bertani broth (TSB and LB, respectively, Difco, Franklin Lakes) at 35°C. Alternatively, nutrient agar or Miller Luria–Bertani agar (Difco) was used to culture these bacteria. Antibiotics and sucrose were added as selective agents where appropriate (described below). Viable cell counting was performed by 10-fold serial dilution of cultures in phosphate-buffered saline (pH 7.2) and subsequent plating onto agar medium.
|Strain or plasmid||Genotype or description||Source|
|JM109||endA1·recA1·gyrA96·hsdR17(rk−mk−)·supE44·recA1Ä(lac-proAB); F′(traD36·proAB+·lacIq·lacZ·M15)||Yanisch-Perron et al. (1985)|
|XL1 Blue||endA1·recA1·gyrA96·thi1·hsdR17·relA1·supE44·lac; F′ [proAB, laclqZΔM15 Tn10 (Tetr)]||Bullock et al. (1987)|
|DH5α||supE44·hsdR17·Δ(lac)U169·recA1·endA1·gyrA96·thi1·relA1·deoR (φ80 lacZΔM15)||Hanahan (1983)|
|SM10||thi1·recA1·thr·leu·tonA·lacY·supE44·RP4-2-Tc::Mu::pir||Simon et al. (1983)|
|MG4||Reporter strain||Pearson et al. (1994)|
|PAO-JP2||ΔrhlI::Tn501 derivative of wild-type PAO1, ΔlasI, HgR TcR||Pearson et al. (1995)|
|UCB-PP PA14||Clinical isolate||Rahme et al. (1995)|
|PA14 gacA−||PA14ΔgacA::gmr||Rahme et al. (1995)|
|PA14 gacS−||PA14ΔgacS::gmr||This study|
|PA14 SCV||PA14ΔgacS::gmr small colony variant||This study|
|pBluescriptII ks+||Cloning and sequencing vector, ampR||Stratagene|
|pBSIIgacS||pBluescriptII ks+ containing a 2.0 kb portion of gacS amplified from PA14 genome; ampR||This study|
|pBSIIgacS::gm||pBSIIgacS containing the gmR cassette from pUCGM in the gacS region; gmR||This study|
|pEX18||Used for allelic exchange mutagenesis. Constructed by ligation of 1791 bp PvuI fragment of pUC18 to large PvuI fragment of pEX100T; ampR||Hoang et al. (1998)|
|pEX18gacS::gm||pEX18 containing the gacS::gm region from pBSIIgacS::gm; ampR, gmR||This study|
|pUCGM||Plasmid containing Tn1696 derived gmR gene flanked by pUC19 polylinker site; ampR gmR||Schweizer (1993)|
|pUCP18||1.8-kb stabilizing fragment from P. aeruginosa incorporated into pUC18||Schweizer (1991)|
|pUCP18mpgacS||pUCP18 containing a 3.4-kb fragment amplified from P. aeruginosa PA14 containing the entire gacS gene and flanking sequences||This study|
|pECP61.5||rhlA::lacZ reporter construct||Pearson et al. (1995)|
|pKDT17||lasB::lacZ reporter construct||Pearson et al. (1994)|
|Prod7 Forward||5′-GATGGTGCTTGGCGGTTACTTCAC-3′||This study|
|Prod7 Reverse||5′-ACGTCCATGAAGACCAGGTCGAAG-3′||This study|
|MPGACS Forward||5′-CGCCAACCCCTCTTCCCCGTCTC-3′||This study|
|MPGACS Reverse||5′-CGGCGACAGCGTGCGGCGAATAG-3′||This study|
Plasmid constructs and strain generation
The plasmids and PCR primers used in this study are summarized in Table 1. A 2-kb fragment of the gacS gene was amplified by PCR (94°C for 5 min, then 35 cycles consisting of 94°C for 30 s, 65°C for 30 s, 72°C for 2 min, followed by a terminal incubation at 72°C for 7 min, then held at 15°C) from P. aeruginosa PA14 genomic DNA using Platinum Pfx polymerase (Invitrogen, Carlsbad, CA) and Prod7 forward and reverse primers. Once amplified, the fragment was isolated and ligated into the EcoRV site of pBluescript II ks+(Stratagene, La Jolla, CA) to produce the interim plasmid construct pBSIIgacS.
The gentamicin resistance (gmr) cassette from pUCGM (Schweizer, 1993) was inserted into the SphI site of the gacS fragment of pBSIIgacS. A 3-kb fragment comprising the original gacS fragment and the gmr cassette was then amplified by PCR and incorporated into the SmaI site of pEX18 (Hoang et al., 1998) to produce plasmid pEX18gacS::gm, which was transformed into E. coli SM10. The plasmid was then transferred by conjugation to P. aeruginosa PA14. Overnight cultures of P. aeruginosa PA14 and E. coli SM10 (pEX18gacS::gm) were grown in LB broth (Sambrook & Russell, 1989) containing no antibiotics and 15 μg mL−1 gentamicin, respectively. Cells were pelletted by centrifugation (800 g for 5 min), gently resuspended in a small volume of phosphate-buffered saline (PBS) and combined so as to have donor cells in excess of recipients. The cell mixture was spotted onto TY plates (8 g tryptone, 5 g select yeast extract, 2.5 g NaCl L−1 agar) and incubated overnight at 37°C. Isolation of a gacS− mutant was accomplished through selection for spontaneous allelic exchange events that transferred the gmr cassette from pEX18gacS::gm into the genomic gacS gene. The resulting lawn of cells was scraped from the TY plate, resuspended in PBS, and deposited onto Vogel Bonner minimal media (Vogel & Bonner, 1956) plates containing 15 μg mL−1 gentamicin. Potential mutants were then assessed for sucrose sensitivity (5% sucrose in LB agar) to confirm loss of the donor plasmid (pEX18) (Yanisch-Perron et al., 1985).
The plasmid pUCP18mpgacS was constructed in order to complement P. aeruginosa gacS− mutants with exogenous gacS in trans. The entire gacS gene, plus c. 300 bp of flanking DNA, was amplified by PCR according to the method of Parkins et al. (2001) using Platinum Pfx polymerase (Invitrogen) and the MPGACS forward and reverse primers. This fragment was ligated into the SmaI site of pUCP18 (Schweizer, 1991) and introduced into the P. aeruginosa gacS− mutant via conjugation with transformed E. coli SM10 as described above. Pseudomonas aeruginosa clones carrying the pUCP18mpgacS construct were identified by carbenicillin resistance (500 μg mL−1), plasmid isolation, and the amplification of appropriately sized PCR products (data not shown).
Biofilms were aerobically cultivated using the MBEC high-throughput (HTP) or Physiology and Genetics (P&G) device (Innovotech, Edmonton, Canada) as described in the manufacturer's instructions and by Ceri et al. (1999). To summarize, the parts of this batch culture apparatus were used in two ways. The top half of the plastic MBEC™-HTP device is a lid with 96 polystyrene pegs that also fits over a standard 96-well microtitre plate. For biofilm growth, the bottom half was either (1) a corrugated trough that guided 22 mL of inoculum across the pegs when the device was placed on a rocker at 3.5 rocks per min (HTP format), or (2) a microtitre plate with 150 μL of inoculum in each well that was placed on a gyrorotary shaker at 150 r.p.m. (P&G format). For either assay format, the inoculum was prepared to c. 107 CFU mL−1 of the desired strain and the inoculated device was incubated at 35°C and 95% relative humidity for the required time. Unless otherwise noted, all experiments utilized the MBEC P&G assay. Pegs from the devices were collected at specific time points for scanning electron microscopy and scanning confocal laser microscopy (see below). Biofilm cell densities were evaluated by breaking pegs from the lid of the MBEC device with sterile pliers, rinsing the peg in PBS, and subsequent viable-cell counting as described above. PBS containing the anionic surfactant Tween-20 (1% v/v) was used to assist in recovery of bacteria from the peg surfaces. Pegs were sonicated for 30 min in an Aquasonic model 250HT ultrasonic cleaner (VWR International, Mississauga, Canada). Samples of broth culture were collected at the same time points and viable cell counts were determined in a similar manner (data not shown).
Swim and swarm assays
Swim assays were performed on a semisolid medium composed of Miller LB broth amended with 0.3% agar per litre. Swarm assays were carried out on a modified M9 medium, containing per litre of double-distilled water 3.0 g KH2PO4, 6.0 g Na2HPO4, 0.5 g NaCl, 0.5 g l-glutamate, 2.0 g dextrose, and 5.0 g agar. This medium was autoclaved and enriched with 1 mL of 1 M MgSO4 and 1 mL of 0.01 M CaCl2. One microlitre aliquots of overnight bacterial cultures were spotted into the middle of the swim or swarm plates, which were incubated for 72 h at room temperature and 35°C, respectively. Swim diameter was measured and plates were photographed using a Kodak EasyShare C340 digital camera (Kodak, Toronto, Canada).
Scanning confocal laser microscopy
Three-dimensional (3-D) biofilm structure was evaluated by scanning confocal laser microscopy (SCLM). Pegs were broken from the MBEC device and immersed in 0.1% w/v acridine orange (Sigma Chemical Co., St Louis, MO) in PBS for 5 min at room temperature. Acridine orange is a membrane-permeant nucleic acid stain that interchelates dsDNA and binds ssDNA through dye-base stacking (Bernas et al., 2005). This fluorophore has an excitation wavelength of 488 nm and broad spectrum emission. Biofilms were examined using a Leica DM IRE2 spectral confocal and multiphoton microscope with a Leica TCS SP2 acoustic optical beam splitter (AOBS) (Leica Microsystems, Richmond Hill, Canada) as previously described (Harrison et al., 2006). A 63 × water immersion objective was used in all imaging experiments. Image capture and 3-D reconstruction of z-stacks were performed using leicaconfocalsoftware (LCS).
Scanning electron microscopy
Pegs broken from the MBEC device were air-dried for up to 2 h at room temperature, and then fixed for 2 h at 4°C in a solution of 5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2). Samples were air-dried overnight, attached to aluminium stubs using epoxy resin, and then sputter-coated with gold/palladium using a Technics Hummer I sputter coater. Scanning electron microscopy (SEM) was performed using a Cambridge Model 360 SEM at 20 kV emission or an environmental SEM (ESEM) Phillips XL 30 ESEM as previously described (Morck et al., 1994). Digital images were captured using omnivision 5.1 software (OmniVision Technologies Inc., Sunnyvale). Data shown are representative of over 100 fields of view for each treatment. Treatments and SEM analysis were repeated independently in triplicate, with at least three sampled pegs of each strain viewed at each time period.
N-Acyl-homoserine lactone determination
For quantification of 3-oxo-C12-AHSL and C4-AHSL, AHSL biosensors E. coli MG4 (pKDT17) (Pearson et al., 1994) and P. aeruginosa PAO-JP2 (pECP61.5) (Pearson et al., 1995) were used. These systems quantify 3-oxo-C12-AHSL and C4-AHSL production based on the measurement of β-galactosidase activity from lasB::lacZ and rhlA::lacZ reporter constructs, respectively. The method has been previously described by Pearson et al. (1994, 1995).
Stock solutions of antibiotic, metals and neutralizers
Ciprofloxacin was purchased from Bayer (Leverkusen, Germany) and 30% hydrogen peroxide from BDH Inc. With these exceptions, all other metals, antibiotics and neutralizing agents were purchased from Sigma Chemical Co. Stock solutions of metals were prepared in double-distilled water (ddH2O), syringe-filtered, and stored at room temperature. With the exception of erythromycin, antibiotics were also prepared in ddH2O but were frozen and stored at −70°C. Erythromycin was prepared in 95% ethanol. H2O2 was diluted directly from the bottle supplied by the manufacturer. Challenge media (containing the desired antibacterial) were made up in LB 30 min prior to use. Reduced glutathione (GSH) and l-cysteine were prepared at 0.25 M in ddH2O, syringe-filtered and stored at −20°C. These two compounds were used at a final concentration of 5 mM each in recovery media for all assays requiring viable cell counting.
Antimicrobial susceptibility testing
Antibiotic, metal and biocide susceptibility tests were performed as previously described (Ceri et al., 1999; Harrison et al., 2004, 2005a). Antimicrobials were arranged into arrays in microtitre plates that typically consisted of serial twofold dilutions along the rows of wells (the challenge plates). The first and last wells of every row were used as sterility and growth controls, respectively. The cultivation times for biofilms used in these assays were calibrated using growth-curve data so that the different strains of P. aeruginosa PA14 produced biofilms with similar viable cell counts (see Results). Biofilms were rinsed with PBS (to remove loosely adherent planktonic cells) and inserted into the challenge plates. After exposure, biofilms were rinsed once in PBS and inserted into microtitre plates containing 200 μL of recovery medium in each well (LB broth, 5 mM GSH, 5 mM l-cys, 1% v/v tween-20). These biofilms were disrupted into the recovery medium using a sonicator (as described above), and the recovered cells were serially diluted and plated for viable cell counting. Spot plates from these experiments were incubated for a minimum of 36 h at 35°C before enumeration. Minimum inhibitory concentration (MIC) values were determined by reading the optical density at 650 nm (OD650) of challenge plates after 72 h at 35°C using a THERMOmax microplate reader with softmax pro data analysis software (Molecular Devices, Sunnyvale, CA).
In an alternative set of experiments, the frequency of SCV cells arising from 24-h biofilms of P. aeruginosa PA14 gacS− was evaluated. Antibiotics were diluted from stock solutions into LB broth to obtain final concentrations of 5 or 1.25 μg mL−1. These were arranged in triplicate in a microtitre plate. Biofilms were incubated in these low concentrations of antibacterials for 18 h at 35°C and 95% relative humidity. Goat and human serum were also assayed in this array, and were kind gifts from The Life and Environmental Sciences Animal Care Facility at the University of Calgary, Department of Biological Sciences. After exposure, biofilms were treated in a manner identical to that described above.
One-way analysis of variance (anova) and two-sample t-tests were performed using minitab® Release 14 (Minitab Inc., State College, PA). Alternative hypotheses were tested at the 95% level of confidence. Mean and standard deviation calculations were performed using Microsoft® Excel 2003 (Microsoft Corporation, Redmond).
- Top of page
- Materials and methods
- Authors contribution
Creation of a gacS− cell line from P. aeruginosa PA14
The gacS gene was inactivated by allelic exchange for a gentamicin resistance marker from a donor plasmid containing sacB. PA14 cells that were sucrose-resistant (which selected for loss of the donor plasmid) and gentamicin-resistant were assessed for interruption of gacS by determining the size and sequence of the PCR product based on primers corresponding to the gacS gene. The resulting cell line was thus verified by the production of appropriately sized PCR products, and this engineered strain was denoted P. aeruginosa PA14 gacS− (data not shown).
Origin of the SCV
When disrupted and plated onto agar, solid-surface-attached biofilms of P. aeruginosa PA14 gacS− that were older than 24 h produced two distinct colony morphologies. After overnight incubation at 35°C, the majority of these colonies were shiny, smooth, light yellow or pale green, and 3–5 mm in diameter. These colonies were similar to those produced by the wild-type organism, with the exception of the slightly greater colony diameter produced by the mutant. A minority of colonies exhibited abrupt edges and were much smaller than colonies produced by either wild-type or gacS− strains of P. aeruginosa PA14. We hypothesized that these colonies represented a SCV of the original gacS− strain.
These SCV isolates were evaluated for growth on Pseudomonas isolation agar and for gentamicin resistance (the marker for the gacS− mutation), as well as by PCR analysis and Gram-staining. These tests were consistent with the premise that the variants were derived from the parental gacS− strain (data not shown). The SCVs were stable and no reversion to normal colony morphology was observed, even after three days' incubation in broth medium or 45 days' serial culture on nutrient agar at room temperature. Phenotypically stable SCVs were not observed originating from cultures of wild-type P. aeruginosa PA14 or the isogenic PA14 gacA− mutant. Rather, these strains produced colony variants that reverted to the normal colony morphotype after subculture on LB agar (this was replicated five to 20 times for each strain). Further, when PA14 SCV was transformed with the plasmid pUCP18mpgacS (bearing the wild-type gacS gene and flanking DNA sequences), the SCV reverted to the wild-type colony morphology with a frequency of c. 10−1.
Strain characterization and biofilm formation
Inactivation of the response regulator GacA is known to affect the ability of P. aeruginosa PA14 to form biofilms (Parkins et al., 2001). Thus, a first logical step was to evaluate biofilm development by P. aeruginosa PA14 wild-type (PA14 wt), the gacS− sensor kinase mutant (PA14 gacS−), and the isolated SCV (PA14 SCV). Relative to either PA14 wt or PA14 gacS−, PA14 SCV produced biofilms of greater cell density between 4 and 10 h of growth in LB medium (Fig. 1a). By 24 h, these three strains produced biofilms with an equivalent mean viable cell count. However, the biomasses produced by these three strains were not equal. For instance, the extracellular polymeric substance produced by the SCV strain was visible to the naked eye (data not shown). This difference in biomass was also observed by microscopy (see below). There were no strain differences in the rates of planktonic cell growth (data not shown). In broth culture, PA14 SCV had a qualitatively greater tendency to form aggregates as well as a surface pellicle (data not shown).
Each strain of P. aeruginosa was tested for a potential to swarm (Fig. 1b–d) or swim (Fig. 1e–g). PA14 gacS− was highly motile relative to the other two strains, and showed an increased ability to swarm. This strain also had significantly greater motility on semisolid swim agar than either PA14 wt or SCV strains (P<0.01, by a two-sample t-test, based on four replicates each). Conversely, PA14 SCV showed significantly decreased swim motility relative to the wild-type strain (P<0.02, by a two-sample t-test, based on four replicates each). In summary, these results show that PA14 gacS− is hypermotile and a poor biofilm former, whereas the isogenic SCV strain is less motile but an excellent biofilm former. Streak plates of the PA14 wt, gacS− and SCV strains are also pictured in Fig. 1h–j, respectively.
Biofilms were examined in situ on pegs from the MBEC device using scanning confocal laser microscopy (SCLM). All bacteria were stained with acridine orange, a membrane-permeant nucleic acid interchelator that has broad spectrum fluorescence (Bernas et al., 2005). This compound stains all cells in a biofilm, live or dead, and may also bind to nucleic acids that are present in the extracellular matrix. Thus, acridine orange may function as a general indicator of biomass present on pegs. Here, surface-adherent growth from P. aeruginosa PA14 wt, gacS− and SCV strains was evaluated after 10 and 24 h. Every image presented here is a representative of at least three independent replicates.
By 10 h, wild-type P. aeruginosa PA14 had formed thin layers of bacteria that were 5–7 μm in height at the air–liquid-surface interface of the polystyrene peg (Fig. 2a and d). In contrast, the gacS− strain had adhered to the surface as scattered cells or small cellular aggregates (Fig. 2b and e). Under the same conditions, the PA14 SCV strain had formed biofilms with greater surface coverage than the wild-type strain and developed into flat layers of densely packed cells that were also 5–7 μm in height (Fig. 2c and f).
After 24 h of growth, the wild-type strain had formed layers up to 15 μm in height, with the greatest amount of biomass present at the air–liquid-surface interface (Fig. 3a and d). Pseudomonas aeruginosa PA14 gacS− formed little more than flat microcolonies and clumps that were heterogeneously distributed across the entire surface (Fig. 3b and e). However, the SCV strain had formed undulating layers of cells that were 20–25 μm thick and that again gave greater surface coverage of the polystyrene pegs than wild-type PA14 (Fig. 3c and f). The structure of biofilms was also examined using SEM at 27 h growth (Fig. 4). These results correlated well with SCLM data at 24 h. In particular, PA14 SCV formed very thick biofilms that lifted away from the surface of the peg when fixed and dehydrated (Fig. 4c and d). At lower magnifications this strain was again observed to produce undulating surface growth owing to the uneven thickness of the biofilm (data not shown). As a control, the revertant strain PA14 SCV (pUCP18mpgacS) was similarly imaged using SEM. Biofilms of this revertant covered less surface area and had lost the undulating surface characteristic of the PA14 SCV biofilm (data not shown). Each SEM image examined here was representative of at least three independent replicates.
N-Acyl-homoserine lactone production
To determine whether there was a correlation between gacS inactivation and AHSL levels, we compared the production of these metabolites between wild-type PA14, gacS− and SCV strains. Pseudomonas aeruginosa PA14 gacA− was also assayed, as we have previously reported that this strain produces lower levels of 3-oxo-C12-AHSL than does the wild-type PA14 strain (Parkins et al., 2001). Escherichia coli MG4 and P. aeruginosa PAO-JP2, bearing plasmids with either a lasB::lacZ (pKDT17) or rhlA::lacZ (pECP61.5) reporter construct, respectively (Pearson et al., 1994, 1995), were used to quantify 3-oxo-C12-AHSL and C4-AHSL levels to β-galactosidase activity. These data are summarized in Table 2, and each value presented is the mean and standard deviation of three trials. C4-AHSL production was similar between P. aeruginosa PA14 wt and its isogenic gacA−, gacS− and SCV strains. However, there were noticeable strain differences in 3-oxo-C12-AHSL production. Induction of lasB::lacZ expression by PA14 wt was approximately twofold greater than that of PA14 SCV or PA14 gacA−, and at least eight times greater than that of PA14 gacS−. In other words, inactivation of gacA produced a different phenotype than did inactivation of gacS. Further, as part of the SCV phenotype, 3-oxo-C12-AHSL production was partially restored (Table 2). These results were corroborated by thin-layer chromatography (data not shown).
|AHSL||Reporter||Pseudomonas aeruginosa PA14 (all values are in Miller units)|
|3-oxo-C12-HSL||lasB-lacZ||590 ± 17||74 ± 14||322 ± 30||300 ± 50|
|C4-HSL||rhlA-lacZ||297 ± 3||250 ± 9||245 ± 5||313 ± 20|
Biofilms are less susceptible to many antimicrobial agents than the corresponding planktonic cells. Mutations in gacA were shown to reduce the resistance of P. aeruginosa PA14 biofilms to some antibiotics (Parkins et al., 2001). Therefore, we determined whether the biofilms of PA14 gacS− or PA14 SCV strains had altered resistance to antibacterials relative to the wild-type strain. Here, the inhibitory and bactericidal actions of metal cations (Cu2+ and Ag+), hydrogen peroxide (H2O2) and ciprofloxacin were evaluated. Cu2+ and Ag+ are industrial pollutants that are also used as disinfectants, H2O2 is produced by plant and animal hosts, and ciprofloxacin is an antibiotic clinically used to treat P. aeruginosa infections.
For susceptibility testing, growth-curve data were used to calibrate incubation times to produce biofilms of similar cell density. For these assays, PA14 wt, gacS− and SCV were incubated at 35°C for 6.0, 7.0, and 5.5 h, respectively, to produce biofilms with cell densities of 5.0±0.7, 5.3±0.5, and 5.5±0.4 log10 CFU peg−1 (based on the mean and standard deviation of 50–55 pooled replicates each). Biofilms formed by individual strains in the MBEC P&G device were statistically equivalent between the different rows of pegs (0.09<P<0.91 by one-way analysis of variance, data not shown). In this model system, planktonic cells shed from the surface of biofilms served as the inoculum for MIC determinations. The advantage of this system is that it may reflect infections or environmental settings where biofilms and planktonic cells form integrated parts of the microbial lifestyle. These data are summarized in Table 3, and each value represents the mean and standard deviation of four independent trials. There were no significant differences in planktonic cell susceptibility to either Cu2+, Ag+ or ciprofloxacin between the different strains (i.e. there was a log2 difference or less between these values). However, planktonic PA14 gacS− was hypersensitive to H2O2, whereas (by comparison) PA14 SCV was highly resistant.
|MIC72 h||MBEC99.9||MIC72 h||MBEC99.9||MIC72 h||MBEC99.9|
|Cu2+ (mM)||8 ± 0||8 ± 0||8 ± 0||7 ± 2||4 ± 0||64 ± 0|
|Ag+ (mM)||0.04 ± 0.02||0.08 ± 0.05||0.04 ± 0.01||0.06 ± 0.02||0.04 ± 0||4.8 ± 0|
|H2O2 (ppm)||938 ± 0||352 ± 135||45 ± 16||22 ± 14||1875 ± 0||293 ± 203|
|Ciprofloxacin (μg mL−1)||0.4 ± 0.2||0.16 ± 0||0.7 ± 0.3||0.12 ± 0.05||0.8 ± 0.4||0.64 ± 0.45|
The antibiofilm activity of Cu2+, Ag+, H2O2, and ciprofloxacin was evaluated by determining mean viable cell counts and log-killing of biofilm populations of P. aeruginosa PA14 wt, gacS− and SCV strains. Consistent with the American Clinical and Laboratory Standards Institute's definitions (CLSI, http://www.nccls.org/), the bactericidal threshold was defined as a 3 log10 reduction in viable cells in the bacterial population. This value will be denoted here as the minimum biofilm eradication concentration required to kill 99.9% of the bacterial cells (MBEC99.9). These values are summarized in Table 3. Although the MBEC99.9 values for H2O2 are similar for PA14 wt and SCV strains (Table 3), the biofilms of PA14 SCV showed increased survival at sub-MBEC99.9 concentrations relative to the wild-type strain (see below). For ciprofloxacin, Cu2+ and Ag+, biofilms of the SCV strain were approximately four, eight and 60 times more tolerant to these toxic factors than the wild-type strain.
We noted that in some instances MIC values obtained using this method were greater than MBEC99.9 values. This represents an expected normality, not peculiarity, to the method. For example, over the course of incubation, peroxide would be gradually degraded in the challenge plates, especially by biofilms during exposure. After removing the biofilms from the challenge media, bacteria were allowed to recover for 72 h prior to MIC determination. In contrast, biofilm cell density was enumerated immediately after exposure to the peroxide (when its in vitro concentration would have been highest). Because there was no corresponding period of recovery for biofilms, this would result in the comparatively lower MBEC99.9 value.
Mean viable cell counts and log-killing data for Cu2+ and Ag+ are presented in Fig. 5, where each point represents the mean and standard deviation of four independent replicates. Similarly, data for H2O2 and ciprofloxacin are presented in Fig. 6. In these two figures, we see the general trend that P. aeruginosa PA14 gacS− is much more susceptible to antimicrobials than the wild-type strain. However, in every case, the PA14 SCV strain produced biofilms that were more tolerant to antimicrobial exposure than those of the wild-type strain. For example, biofilms of the SCV were resistant to 2.4 mM Ag+, whereas the vast majority of cell viability was lost from PA14 wt and gacS− biofilms at 0.04 and 0.02 mM Ag+, respectively. Collectively, these data indicate that deletion of gacS reduces the antimicrobial tolerance of P. aeruginosa PA14. However, phenotypic variation in biofilms of this mutant population gives rise to SCV cells that are much more tolerant to antimicrobials than either the wild-type or parental gacS− strain.
Frequency of phenotypic variation
During the course of susceptibility assays, we noted that the proportion of SCV cells recovered from biofilms after exposure to Ag+ or H2O2 was increased relative to the corresponding growth controls. We thus queried whether this may be true for other antimicrobial agents or growth conditions. An array of clinically used antibiotics, saline, and goat and human sera were examined for an ability to select for these SCVs from 24-h biofilms of P. aeruginosa PA14 gacS−. Biofilms were exposed to these agents for 18 h and each assay was performed in triplicate. Viable cell counts were determined for each exposure condition, and log-survival was determined. The proportion of SCV cells in bacterial populations recovered from these exposure conditions was calculated as the mean of the proportions from each individual trial. The data from these assays are summarized in Table 4.
|Test medium||Antibiotic||Concentration (μg mL−1)||No. of survivors (log10 CFU peg−1)||Log-survival (log10 CFU peg−1)||SCV frequency (%)|
|Culture conditions (overnight)|
|0.9% NaCl||None||NA||3.8 ± 0.3||−2.5 ± 0.3||0|
|Goat serum||None||NA||5.4 ± 0.2||−1.0 ± 0.2||4|
|LB broth||None (growth control)||NA||6.7 ± 0.2||+0.3 ± 0.2||4|
|Human serum||None||NA||5.1 ± 0.2||−1.3 ± 0.2||19|
|Antibiotic exposure (overnight)|
|LB broth||Erythromicin||5||6.4 ± 0.3||−0.1 ± 0.3||0|
|Imipenem||1.25||2.7 ± 1.0||−3.8 ± 1.0||0|
|Tobramicin||1.25||5.4 ± 0.5||−1.0 ± 0.5||14|
|Amikacin||5||3.6 ± 0.6||−2.8 ± 0.6||23|
|Azetronam||1.25||4.5 ± 0.5||−2.6 ± 0.5||24|
|Ceftrioxone||1.25||6.8 ± 0.5||+0.4 ± 0.5||29|
|Oxacilin||1.25||5.6 ± 0.6||−0.9 ± 0.6||32|
|Piperacillin+Tazobactam||5||2.7 ± 1.7||−3.8 ± 1.8||33|
|Rifampicin||5||5.7 ± 0.3||−0.7 ± 0.3||58|
|Antibacterial exposure (2 h exposure, representative example from susceptibility assays)|
|LB broth||None (growth control)||NA||4.2 ± 0.3||−0.9 ± 0.3||0|
|Ciprofloxacin||0.16||2.4 ± 0.8||−2.6 ± 0.8||0|
|Copper cations (Cu2+)||16||4.2 ± 0.1||−0.9 ± 0.1||0|
|Silver cations (Ag+)||4||2.4 ± 0.1||−2.6 ± 0.1||7|
|Hydrogen peroxide||30||1.2 ± 1.1||−3.8 ± 1.1||8|
Rifampicin, an RNA polymerase inhibitor, was a strong selective agent for SCVs from P. aeruginosa PA14 gacS− biofilms. At a concentration of 5 μg mL−1, this drug killed 0.7 log10 cells from the biofilm population. On average, approximately three of five surviving cells from biofilms exposed to this concentration of rifampicin were phenotypic variants. Similarly, the β-lactams piperacillin, oxacillin and ceftrioxone selected for SCVs at a frequency of approximately one in three. This occurred regardless of cell growth (ceftrioxone) or cell death (oxacillin or piperacillin). The aminoglycosides tobramycin and amikacin, both of which find high clinical use in combating P. aeruginosa infections, selected for SCVs at a frequency of approximately one in five. This in vitro selection was compound-specific, as in no instances were saline, erythromycin, imipenem, or ciprofloxacin observed to increase the frequency of SCV cells from PA14 gacS− biofilms. Human serum, but not goat, also gave rise to phenotypic variants at elevated frequencies compared with growth controls. These assays indicate that environmental conditions, such as antibacterial exposure or host factors, may select for SCVs from biofilms of P. aeruginosa PA14.
- Top of page
- Materials and methods
- Authors contribution
In this work we engineered a strain of P. aeruginosa PA14 by creating a mutation inactivating the sensor kinase gacS. This mutant was hypermotile and a poor biofilm former relative to the wild-type strain. While characterizing this strain, we discovered that biofilms of this mutant gave rise to phenotypically stable SCVs at a proportion that was increased by three factors, namely (1) age of the biofilm, (2) by in vitro culture in human serum, and (3) by exposure of biofilms to certain antibacterial agents. This SCV strain had a hyper-biofilm-forming phenotype, and was less motile and more tolerant to bactericidal agents than the parental gacS− and wild-type strains. With the exception of phenotypic stability, all of these traits have been described for P. aeruginosa colony morphology variants in the literature (Deziel et al., 2001; Drenkard & Ausubel, 2002; Häußler et al., 2003; Kirisitis et al., 2005). Although there may be multiple mechanisms that give rise to SCVs in P. aeruginosa, we have concluded that GacS regulates the reversion of variants to normal colony morphotypes for at least one of these pathways. This premise was supported by complementation analysis, in which SCVs reverted to normal colony phenotypes when transformed with a plasmid bearing wild-type gacS. Thus, the inactivation of gacS, which frequently occurs in laboratory and rhizosphere populations of pseudomonads (Sánchez-Contreras et al., 2002; van den Broek et al., 2005), may lead to the accumulation of stress-resistant SCV cells in P. aeruginosa biofilms.
These findings may be important with respect to the phenotypic variation of P. aeruginosa and other Pseudomonas species in soil. For instance, phenotypic variation in P. fluorescens is mediated by two site-specific recombinases, XerD and Sss, which appear to introduce mutations into gacA and/or gacS (Martinez-Granero et al., 2005). Overexpression of xerD and sss has been used to generate highly motile variants that have an enhanced ability to colonize the alfalfa rhizosphere (Martinez-Granero et al., 2006). Pseudomonas aeruginosa PA14 similarly possesses a homologue of sss (Pseudomonas Genome Database version 2, Locus ID PA14_69710, http://v2.pseudomonas.com/) and xerD (PA14_16040). It is worth noting that other rhizosphere Pseudomonas species show phenotypic variation that is based on spontaneous mutation of the gacA and gacS genes that may enhance plant-root colonization (Achouak et al., 2004; van den Broek et al., 2005). To the best of our knowledge, this is the first report that explicitly and directly links an inactivating mutation in gacS to the production of stable SCVs in P. aeruginosa. Our laboratory group has similarly observed the production of these stable colony variants in a ΔgacS strain of P. chlororaphis, which characteristically occurs by exposing biofilms to Ag+ (L.L.R. Marques, H. Ceri, A.J. Anderson, Y.C. Kim and M.E. Olson, unpublished data). SCVs of P. chlororaphis O6 generated in this manner show enhanced resistance to certain heavy metals (G.R. Foglia, J.J. Harrison, H. Ceri and R.J. Turner, unpublished data). In conjunction with the data presented in this paper, this affirms the notion that the SCV phenotype may play a role in stress tolerance.
Genes of the GacS regulon strongly influence the later stages of biofilm formation in P. aeruginosa PA14. Biofilms formed by the PA14 gacS− mutant did not proceed far beyond the irreversible attachment and proliferation stages of development (reviewed by Donlan & Costerton, 2002). Biofilms of this mutant remained flat and lacked the characteristic layered structures of the mature biofilms formed by the parental strain. The biofilm growth process observed here for P. aeruginosa PA14 gacS− also differed from that previously reported by our research group for PA14 gacA−, which failed to form surface-adherent aggregates under similar laboratory conditions (Parkins et al., 2001).
GacS represents another piece of the cell specialization puzzle that exists in the literature for P. aeruginosa biofilms. GacS shares a high degree of identity with an open reading frame (ORF3) downstream and adjacent to pvrR (phenotype variant regulator), a hypothetical response regulator for a two-component system (PubMed accession number AF482691; http://www.ncbi.nih.gov/). Together, ORF3 and pvrR form a hybrid, putative sensor kinase and response regulator. Overexpression of PvrR from a plasmid reduces the frequency of phenotypic variation in P. aeruginosa biofilms (Drenkard & Ausubel, 2002). GacS/GacA are also upstream regulators of the pel (pellicle) operon (Goodman et al., 2004). This cluster of seven adjacent genes is postulated to encode polysaccharide biosynthetic enzymes important for matrix formation in P. aeruginosa PA14 (Friedman & Kolter, 2004). These genes are implicated in surface adherence, and in general the pel locus shows increased expression in SCVs derived from biofilms of P. aeruginosa PAO1 (Kirisitis et al., 2005). The data from this study suggest that a functional gacS limited the generation of SCVs in biofilms (although the exact mechanism remains elusive), and that this phenomenon was specific to the gacS− mutant, as phenotypically stable SCVs were not produced from an isogenic gacA− strain of P. aeruginosa PA14. Further indicative of the low-fidelity relationship between GacA and GacS are decreases in AHSL levels of gacS− relative to the gacA− strain and the differences in biofilm structure (Parkins et al., 2001). Two other sensor kinases, RetS and LadS, are known to modify intracellular signalling through GacA (Goodman et al., 2004; Ventre et al., 2006). It is interesting to note that deletion of retS is similarly associated with the occurrence of hyper-biofilm-forming colony morphology variants in P. aeruginosa (Zolfaghar et al., 2005).
Quorum-sensing systems may be involved in the process of phenotypic variation, and consequently may be indirectly and partly responsible for alterations in antimicrobial susceptibility. Amongst many other genes, these autoinducers control the expression of superoxide dismutase and catalase (Hassett et al., 1999), which may account for the hypersensitivity of PA14 gacS− to H2O2. Compared with the gacS− strain, we noted that AHSL levels were partially restored in the SCV, which coincided with increased tolerance to H2O2. The increased production of extracellular polymers associated with the SCV strain may further enhance the protective activity of these enzymes. This may contribute to resistance through a reaction–diffusion phenomenon in which the substrate (H2O2) is degraded in the extracellular matrix before penetrating into the depths of the biofilm (Stewart, 2003).
The extra biomass in SCV biofilms may also play a role in Cu2+ and Ag+ sorption. Sequestration of divalent copper cations in P. aeruginosa biofilms has been evaluated using the organic chelator sodium diethyldithiocarbamate to cause coloured precipitation of the metal (Harrison et al., 2005b). Using this approach, we have observed that biofilms of the PA14 SCV strain qualitatively adsorb greater Cu2+ than either the PA14 wt or gacS− strain (J.J. Harrison, G.R. Foglia and H. Ceri, unpublished data). This implies that the production of hyper-biofilm-forming SCVs from a genotypically diverse Pseudomonas population represents a strategy that may give rise to elevated heavy metal resistance at the population level. A similar statement may be made for H2O2 and ciprofloxacin, although in the case of the latter more precise mechanistic detail is not known, as this antibiotic rapidly equilibrates across the P. aeruginosa biofilm matrix (Lewis, 2001).
Because the biofilm mode of growth is thought to be responsible for persistent infections, these P. aeruginosa SCVs may play an additional role in pathogenesis, in particular the destructive infections of the CF lung (Drenkard & Ausubel, 2002; Häußler et al., 2003). Pseudomonas aeruginosa is also known for causing infections associated with burn wounds and the use of catheters (reviewed by O'Toole & Stewart, 2005). Here we have identified that low concentrations of clinically used antibiotics may select for hyper-biofilm-forming SCVs from biofilms of the gacS− strain of this nosocomial pathogen. A similar trend has been previously shown for CF isolates of P. aeruginosa (Drenkard & Ausubel, 2002). The present study indicates that silver ions may be added to this list of triggers and/or selective agents. This is important, as silver compounds are finding renewed use in medicine as antimicrobial surface coatings for bandages and catheters (Ovington, 2004). Thus, an emerging and provocative theme is that antimicrobial chemotherapy may be triggering or selecting for the phenotypic variation in P. aeruginosa biofilms that contributes to drug resistance and the destruction of the chronically infected tissue(s) (Hoffman et al., 2005; O'Toole & Stewart, 2005). This type of response would also be advantageous in soil environments, where P. aeruginosa, similar to other pseudomonads, would encounter other antibiotic-producing microorganisms, toxic metals, or H2O2 produced by plants.
The stability of many types of biological systems is increased by diversity. For instance, phenotypic diversity arises from genetically identical founding populations of P. fluorescens grown in spatially heterogeneous microcosms. In this instance, the emergence of the hyper-biofilm-forming wrinkly-spreader phenotype allows highly efficient colonization of the air–liquid interface (reviewed by Kassen & Rainey, 2004). In another example, Boles et al. (2004) have recently identified that P. aeruginosa biofilm communities self-generate genetic diversity through a recA-dependent mechanism. Spontaneous mutations in gacS of P. fluorescens introduced by the site-specific recombinases Sss and XerD are analogous and also contribute to phenotypic variation as well as to fitness (Martinez-Granero et al., 2005). This work suggests that P. aeruginosa biofilm formation and antibacterial resistance are interrelated with phenotypic variation, which itself may be linked to the underlying genetic diversity of these bacterial populations.
- Top of page
- Materials and methods
- Authors contribution
This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) to H.C. and M.E.O., Canadian Institute for Health Research (CIHR) grant MOP15680 to D.G.S. and H.C., and a University Technologies Incorporated (UTI) Fellowship in Science to L.L.R.M. M.E.O. was supported by the ASRA/Westaim Chair for Biofilm Research. J.J.H. was supported by an NSERC Canada Graduate Scholarship and an Alberta Heritage Foundation for Medical Research (AHFMR) studentship. Additional funding has been provided by MBEC Bioproducts Incorporated. SCLM was made possible by a Canadian Foundation for Innovation (CFI) Bone and Joint Disease Network grant to H.C. We are thankful for technical assistance, expert advice and input from Dr C.-M. Ryu, Dr Michael Surette, Dr Anne J. Anderson, Katy Ward, Byron Sessons, Jon McGovern, Jaime MacDonald, Liz Middlemiss and Elsie Lee.
- Top of page
- Materials and methods
- Authors contribution
J.A.D. and J.J.H. contributed equally to this work.
- Top of page
- Materials and methods
- Authors contribution
- 2004) Phenotypic variation of Pseudomonas brassicacearum as a plant root-colonization strategy. Mol Plant Microbe Interact 17: 872–879. , , & (
- 2005) Confocal fluorescence imaging of photosensitized DNA denaturation in cell nuclei. Photochem Photobiol 81: 960–969. , , , & (
- 2005) Pseudomonas aeruginosa tolerance to tobramycin, hydrogen peroxide and polymorphonuclear leukocytes is quorum-sensing dependent. Microbiology 151: 373–383. , , et al. (
- 2004) Self-generated diversity produces “insurance effects” in biofilm communities. Proc Natl Acad Sci USA 101: 16630–16635. , & (
- 2001) Characterization of spontaneous gacS and gacA regulatory mutations of Pseudomonas fluorescens biocontrol strain CHA0. Antonie Van Leeuwenhoek 79: 327–336. , , , , & (
- 1987) E. coli XL-Blue: a high efficiency plasmid transforming recA Escherichia coli strain with beta-galactosidase selection. Biotechniques 5: 376–378. , & (
- 1999) The Calgary Biofilm Device: new technology for rapid determination of antibiotic susceptibilities in bacterial biofilms. J Clin Microbiol 37: 1771–1776. , , , , & (
- 2002) Survival of GacS/GacA mutants of the biological control bacterium Pseudomonas aureofaciens 30-84 in the wheat rhizosphere. Appl Environ Microbiol 68: 3308–3314. , , & (
- 2005) The Pseudomonas chlororaphis PCL1391 sigma regulator psrA represses the production of the antifungal metabolite phenazine-1-carboxamide. Mol Plant Microbe Interact 18: 244–253. , , & (
- 2003) Conservation of the response regulator gene gacA in Pseudomonas species. Environ Microbiol 5: 1328–1340. , & (
- 2001) Initiation of biofilm formation by Pseudomonas aeruginosa 57RP correlates with emergence of hyperpiliated and highly adherent phenotypic variants deficient in swimming, swarming, and twitching motilities. J Bacteriol 183: 1195–1204. , & (
- 2002) Biofilms: survival mechanisms of clinically relevant microorganisms. Clin Microbiol Rev 15: 167–193. & (
- 2002) Pseudomonas biofilm formation and antibiotic resistance are linked to phenotypic variation. Nature 416: 740–743. & (
- 2000) Controlling instability in gacS-gacA regulatory genes during inoculant production of Pseudomonas fluorescens biocontrol strains. Appl Environ Microbiol 66: 3142–3150. & (
- 2004) Genes involved in matrix formation of Pseudomonas aeruginosa PA14 biofilms. Mol Microbiol 51: 675–690. & (
- 2004) A signalling network reciprocally regulates genes associated with acute infection and chronic persistence in Pseudomonas aeruginosa. Dev Cell 7: 745–754. , , , , & (
- 1983) Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166: 557–580. (
- 2004) Biofilm susceptibility to metal toxicity. Environ Microbiol 6: 1220–1227. , , & (
- 2005a) High-throughput metal susceptibility testing of microbial biofilms. BMC Microbiol 5: 53. , & (
- 2005b) Persister cells, the biofilm matrix and tolerance to metal cations in biofilm and planktonic Pseudomonas aeruginosa. Environ Microbiol 7: 981–994. , & (
- 2005c) Biofilms: a new understanding of these microbial communities is driving a revolution that may transform the science of microbiology. Am Sci 93: 508–515. , , & (
- 2006) Metal resistance in Candida biofilms. FEMS Microbiol Ecol 55: 479–491. , , , , & (
- 1999) Quorum sensing in Pseudomonas aeruginosa controls expression of catalase and superoxide dismutase genes and mediates biofilm susceptibility to hydrogen peroxide. Mol Microbiol 34: 1082–1093. , , et al. (
- 2004) Biofilm formation by the small colony variant phenotype of Pseudomonas aeruginosa. Environ Microbiol 6: 546–551. (
- 2003) Highly adherent small colony variants of Pseudomonas aeruginosa in cystic fibrosis lung infection. J Med Microbiol 52: 295–301. , , , , , , , & (
- 1998) A broad host range Flp-Frt recombination system for site-specific excision of chromosomally located DNA sequences: application for isolation of unmarked Pseudomonas aeruginosa mutants. Gene 212: 77–86. , , & (
- 2005) Aminoglycoside antibiotics induce bacterial biofilm formation. Nature 436: 1171–1175. , , , , & (
- 2005) Quorum sensing: the power of cooperation in the world of Pseudomonas. Environ Microbiol 7: 459–471. , & (
- 2004) The ecology and genetics of microbial diversity. Annu Rev Microbiol 58: 207–231. & (
- 2005) Characterization of colony morphology variants isolated from Pseudomonas aeruginosa biofilms. Appl Environ Microbiol 71: 4809–4821. , , & (
- 2001) Riddle of biofilm resistance. Antimicrob Agents Chemother 45: 999–1007. (
- 2005) Two site-specific recombinases are implicated in phenotypic variation and competitive rhizosphere colonization in Pseudomonas fluorescens. Microbiology 151: 975–983. , , , & (
- 2006) Rhizosphere selection of highly motile phenotypic variants of Pseudomonas fluorescens with enhanced competitive colonization ability. Appl Environ Microbiol 72: 3429–3434. , & (
- 1994) Comparative evaluation of fleroxacin, ampicillin, trimethoprim-sulfamethoxazole, and gentamicin as treatments of catheter associated urinary tract infections in a rabbit model. Am J Med 94: 23S–30S. , , , , , , & (
- 2005) Biofilms strike back: sublethal concentrations of antibiotics increase bacterial drug tolerance by promoting biofilm formation. Nat Biotechnol 23: 1378–1379. & (
- 2004) The truth about silver. Ostomy/Wound Manage 50: 1S–10S. (
- 2001) Pseudomonas aeruginosa GacA, a factor in multihost virulence, is also essential for biofilm formation. Mol Microbiol 40: 1215–1226. , & (
- 1994) Structure of the autoinducer required for the expression of Pseudomonas aeruginosa virulence genes. Proc Natl Acad Sci USA 91: 197–201. , , , , , & (
- 1995) A second N-acyl-homoserine lactone signal produced by Pseudomonas aeruginosa. Proc Natl Acad Sci USA 92: 1490–1494. , , & (
- 1995) Common virulence factors for bacterial pathogenicity in plants and animals. Science 268: 1899–1902. , , , , & (
- 1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Press, New York, NY. & (
- 2002) Phenotypic selection and phase variation occur during alfalfa root colonization by Pseudomonas fluorescens F113. J Bacteriol 184: 1587–1596. , , , , & (
- 2003) Inactivation of gacS does not affect the competitiveness of Pseudomonas chlororaphis in the Arabidopsis thaliana rhizosphere. Appl Environ Microbiol 69. , & (
- 1991) Escherichia–Pseudomonas shuttle vectors derived from pUC18/19. Gene 97: 109–112. (
- 1993) Small broad-host-range gentamicin resistance gene cassettes for site-specific insertion and deletion mutagenesis. Biotechniques 5: 831–832. (
- 1983) A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Biotechnology 1: 784–791. , & (
- 2005) Is Pseudomonas aeruginosa only “sensing quorum”? Crit Rev Microbiol 31: 171–182. & (
- 2003) Induced defence in tobacco by Pseudomonas chlororaphis strain O6 involves at least the ethylene pathway. Physiol Mol Plant Pathol 63: 27–34. , , , , & (
- 2003) Diffusion in biofilms. J Bacteriol 185: 1485–1491. (
- 2005) The role of phenotypic variation in rhizosphere Pseudomonas bacteria. Environ Microbiol 7: 1686–1697. , & (
- 2006) Multiple sensors control reciprocal expression of Pseudomonas aeruginosa regulatory RNA and virulence genes. Proc Natl Acad Sci USA 103: 171–176. , , , , , , , , & (
- 2006) Regulation of quorum sensing in Pseudomonas. FEMS Microbiol Rev 30: 274–291. (
- 1956) Acetylornithinase of Escherichia coli: partial purification and some properties. J Biol Chem 218: 97–106. & (
- 1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33: 103–119. , & (
- 2005) Mutation of retS, encoding a putative hybrid two-component regulatory protein in Pseudomonas aeruginosa, attenuates multiple virulence mechanisms. Microb Infect 7: 1305–1316. , , , , & (