Elucidating the Genetic Basis of Crystalline Biofilm Formation in Proteus mirabilis
ABSTRACT
Proteus mirabilis forms extensive crystalline biofilms on urethral catheters that occlude urine flow and frequently complicate the management of long-term-catheterized patients. Here, using random transposon mutagenesis in conjunction with in vitro models of the catheterized urinary tract, we elucidate the mechanisms underpinning the formation of crystalline biofilms by P. mirabilis. Mutants identified as defective in blockage of urethral catheters had disruptions in genes involved in nitrogen metabolism and efflux systems but were unaffected in general growth, survival in bladder model systems, or the ability to elevate urinary pH. Imaging of biofilms directly on catheter surfaces, along with quantification of levels of encrustation and biomass, confirmed that the mutants were attenuated specifically in the ability to form crystalline biofilms compared with that of the wild type. However, the biofilm-deficient phenotype of these mutants was not due to deficiencies in attachment to catheter biomaterials, and defects in later stages of biofilm development were indicated. For one blocking-deficient mutant, the disrupted gene (encoding a putative multidrug efflux pump) was also found to be associated with susceptibility to fosfomycin, and loss of this system or general inhibition of efflux pumps increased sensitivity to this antibiotic. Furthermore, homologues of this system were found to be widely distributed among other common pathogens of the catheterized urinary tract. Overall, our findings provide fundamental new insight into crystalline biofilm formation by P. mirabilis, including the link between biofilm formation and antibiotic resistance in this organism, and indicate a potential role for efflux pump inhibitors in the treatment or prevention of P. mirabilis crystalline biofilms.
INTRODUCTION
Indwelling urinary catheters are the most commonly used medical devices and are employed in a range of bladder management regimens in hospitals, community care settings, and nursing homes (1–4). It has been estimated that more than 100 million urethral catheters are fitted each year, with many used for long-term bladder management in community or nursing home settings (2–6). For the most part, these simple devices offer significant benefits to patients and improve the quality of life, but they also provide a route through which bacteria may enter and colonize the normally sterile, nutrient-rich bladder (4, 7, 8). Given the frequent use of urinary catheters in health care systems across the world, it is perhaps unsurprising that catheter-associated urinary tract infections (CAUTIs) are presently among the most common health care-associated infections in the United Kingdom and other developed countries (5, 9–11).
Although not a prominent pathogen of the normal urinary tract, Proteus mirabilis accounts for a significant proportion of CAUTIs, particularly in patients undergoing long-term catheterization (defined as a period of >28 days), where this organism may be involved in as many as 40% of CAUTIs (12, 13). A hallmark of P. mirabilis CAUTI is the encrustation and blockage of urethral catheters, which stem from the formation of unusual crystalline biofilm structures on catheter surfaces (14–17). The potent urease enzyme produced by P. mirabilis generates ammonia through the hydrolysis of urea present in the urine, leading to an elevation of urinary pH (8, 18, 19). Under these alkaline conditions, calcium and magnesium phosphates precipitate and form crystals that become trapped in the developing biofilm. The biofilm matrix itself has also been shown to attract magnesium and calcium ions and to accelerate and stabilize crystal growth (4, 20–22). Ultimately, this process results in the mineralization of the biofilm and the formation of extensive crystalline structures that occlude urine flow.
Encrustation and blockage of catheters constitute a serious risk to patient welfare and quality of life, frequently leading to the reflux of infected urine to the kidneys and culminating in episodes of pyelonephritis, septicemia, and shock (4, 8). Although advances in the care and management of catheterized patients have had a major impact in reducing the occurrence of CAUTIs in the short term, infection is almost inevitable in patients undergoing long-term treatment (3, 8, 13, 23). In addition, P. mirabilis is extremely difficult to eliminate once established in the catheterized urinary tract, often causing chronic infection and blockage; antibiotic treatment and catheter changes are generally unsuccessful in resolving these infections (8, 24, 25). Furthermore, all available catheter types, including those coated with antimicrobial agents, remain susceptible to blockage (8, 24, 25). As such, there are currently no truly effective means of preventing or treating P. mirabilis CAUTIs during long-term bladder catheterization.
Despite the clear need to develop new approaches for the prophylaxis of P. mirabilis CAUTI and the importance of biofilm formation to the pathogenesis of these infections, relatively little is known about the mechanisms underlying P. mirabilis biofilm formation in the catheterized urinary tract. Here we employed random transposon mutagenesis and in vitro models of infection to identify genes in this organism that are associated with crystalline biofilm formation and catheter blockage. These investigations provided new insight into functions underpinning biofilm formation and catheter blockage in P. mirabilis, as well as potential strategies for intervention.
MATERIALS AND METHODS
Bacterial strains, media, and routine culture.
The P. mirabilis wild-type strain B4 is a clinical isolate from an encrusted catheter and served as the parental strain for mini-Tn5 mutagenesis (26). Bacteria were routinely cultured in Luria-Bertani (LB) medium (Fisher Scientific, United Kingdom) at 37°C. Media were supplemented with 50 μg ml−1 kanamycin (Fisher Scientific) for the routine culture of mini-Tn5 transposon mutants. For the isolation of single colonies of P. mirabilis and the suppression of swarming motility, strains were grown on MacConkey agar (1.5% agar) without salt (Oxoid Ltd., United Kingdom). The artificial urine (AU) medium based on that described by Stickler et al. in 1999 (27), was composed of sodium disulfate (11.5 g liter−1), magnesium chloride (hexahydrate) (3.25 g liter−1), sodium chloride (23 g liter−1), trisodium citrate (3.25 g liter−1), sodium oxalate (0.1 g liter−1), potassium dihydrogen orthophosphate (14 g liter−1), potassium chloride (8 g liter−1), ammonium chloride (5 g liter−1), calcium chloride dihydrate (3.25 g liter−1), urea (125 g liter−1), gelatin (25 g liter−1; Fisher Scientific), and tryptone soya broth (5 g liter−1; Oxoid). Stock solutions of urea and calcium chloride dihydrate were sterilized separately by membrane filtration (0.45 μm; Sartorius, United Kingdom) and were added to the other components (which were sterilized by autoclaving) to provide the final AU medium with all ingredients at the concentrations noted above and with a final pH of 6.1.
Random transposon mutagenesis.
The random transposon mutagenesis approach utilized in this study has been described previously and has been confirmed to generate single random insertions into P. mirabilis (26). Briefly, the pUT suicide vector harboring a mini-Tn5Km2 transposon was introduced into the wild type P. mirabilis strain B4 by conjugal transfer from the donor organism, Escherichia coli S17.1λpir, on 1.5% LB agar supplemented with 10 mM MgSO4. Following matings, P. mirabilis transconjugants harboring mini-Tn5 elements were selected for by plating to single colonies on MacConkey agar without salt, supplemented with kanamycin (30 μg ml−1) and polymyxin (60 μg ml−1). Transconjugants were then arrayed into 384-well microtiter plates by using a QPix2XT colony-picking robot (Genetix, United Kingdom). Plates were sealed with breathable membranes to prevent evaporation, incubated overnight at 37°C, and subsequently stored at −80°C until required.
High-throughput screening for mutants with altered biofilm-forming abilities.
As a preliminary screen to facilitate the identification of mutants deficient in catheter blockage, the simple CV-based microtiter plate assay described by O'Toole and Kolter (28) was utilized as a preliminary screen. This first-pass, high-throughput screen was used to eliminate mutants with general defects in growth and to identify a pool of candidate mutants likely to be enriched for those deficient in crystalline biofilm formation and suitable for assessment in the more representative in vitro bladder model assay described below. This first-pass CV-based assay was itself conducted in 2 stages.
In stage 1, mutants stored in 384-well plates were transferred to 96-well non-tissue culture-treated flat-bottom plates filled with 100 μl LB medium by using a manual replicator (Genetix, United Kingdom). Replicated mutants were incubated at 37°C for 20 h without shaking, and on the following day, the medium was removed, and wells were washed with sterile deionized water (SDW) to remove unattached cells. The wells were subsequently filled with 120 μl of a 0.5% CV solution (Fisher Scientific, United Kingdom), and the plates were incubated at room temperature for 10 min. The CV solution was subsequently decanted; wells were washed with SDW to remove excess stain; and bound stain was eluted by the addition of 120 μl dimethyl sulfoxide (DMSO). The absorbances of the resulting solutions were measured at 595 nm. To identify mutants with potential alterations in biofilm formation in this first-pass screen, the absorbance recorded for each individual well (mutants) was compared to the average plate reading (from all wells), and those wells generating absorbance readings that were at least 0.1 absorbance unit higher or lower than the plate average (biofilm enhanced or biofilm deficient) were selected for second-pass screening.
In stage 2, to validate the biofilm phenotypes of the mutants identified in stage 1, biofilm formation by selected mutants was compared directly with that by the wild-type strain B4, again using the 96-well plate CV assay as described above. For this second-pass screen, each plate also contained wells inoculated with the wild type (n = 3) and wells with uninoculated medium (n = 3). For each mutant, the average reading from 3 replicate experiments (3 distinct 96-well plates representing 3 biological replicates) was used to ascertain biofilm-forming ability in relation to that of the wild type. Mutants were also assessed for the ability to grow to wild-type levels over the duration of the assay, by measuring the optical density at 600 nm prior to the processing of plates for CV biofilm assays. Mutants displaying statistically significant differences in CV absorbance readings (P, <0.05), but no significant differences in growth (P, >0.05), were taken as biofilm-altered mutants and were characterized further.
In vitro models of the catheterized urinary tract.
Following the first -pass intermediate CV screen (described above), the more robust and more representative in vitro bladder model was utilized to characterize candidate mutants and to identify those defective in catheter blockage. Models were run as described by Stickler et al. in 1999 (27), with minor modifications, and are illustrated in Fig. S2 in the supplemental material. Models consist of a double-walled glass chamber (the bladder) maintained at 37°C by a water jacket supplied from a circulating water bath (27). Size 14 French all-silicone Foley catheters (Bard, United Kingdom) were inserted into the chamber via an outlet in the base of the glass chamber and retention balloons inflated with 10 ml sterile water. The catheter was attached to a drainage bag to form a sterile, closed drainage system. AU medium was supplied to the bladder at a constant flow rate of 0.75 ml min−1. Bladder models were inoculated with 10 ml of a bacterial culture containing ∼109 CFU ml−1, and bacterial cells were allowed to establish themselves within the model for 1 h before flow was activated. The numbers of viable cells present in the medium in bladders were enumerated at the start and end of experiments, and pH was also measured at the start and end of bladder model experiments by sampling the medium in the “bladder.”
Identification of genes disrupted in mutants with altered biofilm formation.
Prior to the identification of disrupted genes, the loss of the pUT delivery vector was confirmed by sensitivity to ampicillin (100 μg ml−1) as described previously (26). Mutants that were determined to harbor a cointegrated pUT vector were not analyzed further. Genes disrupted in mutants of interest were identified using a “cloning-free” direct PCR-based approach to amplify DNA flanking the transposon insertion, as described by Manoil in 2000 (29) (for primers, see Table S1 in the supplemental material). The resulting amplicons were sequenced by GATC Biotech (Cambridge, United Kingdom), and the sequence at the transposon-chromosome junction (20 to 40 nucleotides [nt]) was correlated to the P. mirabilis HI4320 genome sequence (GenBank accession no. NC_010554) (26, 30). Bioinformatic analysis and annotation were performed using Geneious, version 6.1.4 (Biomatters Ltd., New Zealand), and the Basic Local Alignment Search Tool (BLAST; National Center for Biotechnology Information, USA [http://blast.ncbi.nlm.nih.gov/Blast.cgi]) (31). BLAST searches were implemented in Geneious, version 6.1.4, and were used to query either a custom database of P. mirabilis HI4320 sequence data or the full nonredundant (nr) data set (online). Analysis of metabolic pathways disrupted in P. mirabilis mutants was also conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG [http://www.genome.jp/kegg/kegg2.html]) (32).
Polar effects of transposon insertions in selected mutants.
The expression of genes adjacent to those disrupted was assessed by reverse transcription-PCR (RT-PCR) (see Fig. S1 in the supplemental material). P. mirabilis cultures were grown to ∼107 CFU ml−1, and 1 ml of RNAprotect reagent (Qiagen, United Kingdom) was added to 500 μl of bacterial culture. RNA was then extracted by using the Qiagen RNeasy kit (Qiagen) according to the manufacturer's protocols and by utilizing lysozyme at 15 μg ml−1 to aid digestion. Residual DNA was eliminated using Turbo DNA-free DNase (Ambion, Life Technologies Ltd., United Kingdom), and recovered RNA was quantified using a NanoDrop ND2000 low-volume spectrophotometer. The QuantiTect reverse transcription kit (Qiagen) was then used to generate cDNA according to the manufacturer's instructions, by utilizing 8 ng of purified, DNase-treated RNA per reaction as a template. The resulting cDNA was used as the template in RT-PCRs, carried out by using Taq polymerase and standard Qiagen PCR reagents according to the manufacturer's standard protocol. Primers for genes downstream of the transposition site were designed by using Primer3 (http://frodo.wi.mit.edu), on the basis of the P. mirabilis HI4320 genome sequence, with the primers and target genes listed in Table S1 in the supplemental material.
Urease production.
The production of urease was analyzed by using a protocol modified from that of Creno et al. (33) as described previously (34). Strains were grown in LB broth supplemented with urea (0.1% [wt/vol]) at 37°C for 4 h. Cells were harvested by centrifugation (at 3,000 × g for 10 min) and were resuspended in 2.5 ml ice-cold sodium phosphate buffer (0.1 M sodium phosphate, 10 mM EDTA at pH 7.3). The total protein in cell suspensions was determined by using the Micro Total Protein kit according to the manufacturer's instructions (Sigma-Aldrich, United Kingdom). To measure urease activity, 200 μl of the cell suspension was added to 800 μl of the reaction buffer (50 mM urea, 0.1 M sodium phosphate buffer), mixed well, and incubated at 37°C for 10 min. The reaction was terminated by the addition of 2 ml of phenol sodium nitroprusside solution (0.5% phenol [wt/vol], 0.025% sodium nitroprusside [wt/vol]). Color development was initiated by the addition of 2 ml of sodium hypochlorite solution (0.2% sodium hydroxide [wt/vol], 0.21% sodium hypochlorite [wt/vol]) and incubation at 56°C for 5 min. Color change was measured in a spectrophotometer (model 6300; Jenway) at 626 nm, against a blank containing 200 μl of sodium phosphate buffer instead of the cell suspension. Absorbance readings from mutants and the wild type were compared to a standard curve of absorbance readings generated using solutions of ammonia chloride at concentrations ranging from 0.01 mM to 10 mM. Urease activity was expressed in millimoles of urea hydrolyzed per minute per milligram of protein.
Swimming and swarming assays.
Swarming motility was characterized on dry LB agar plates. A 10-μl drop of an overnight culture was inoculated onto the center of each plate. The drops were allowed to soak into the agar at room temperature, and then the plates were incubated for 8 h at 37°C. To assess swimming motility, 2 μl of an overnight culture was stabbed into the center of LB agar (LB medium solidified with 0.15% agar) motility plates, and then the plates were incubated for 6 h at 37°C. The distances that the bacteria migrated on both types of agar from the point of inoculation were measured. The swimming and swarming abilities of each mutant were expressed as percentages of those of the wild type (taken as 100%) in order to produce swarming and swimming indices for each mutant.
Parallel plate flow chambers.
Parallel plate flow chambers (BioSurface Technologies, USA) were used to assess the abilities of test strains to adhere to silicone and were operated as described previously (34). The glass bottom slide was coated with silicone prior to use, whereas the upper slide remained uncoated. Phosphate-buffered saline (PBS) (Fisher Scientific) was flushed though the system for 10 min prior to use. Overnight cultures (20 h) of test strains were harvested by centrifugation (at 3,000 × g for 10 min), and the cell density was adjusted to ∼3 × 108 CFU ml−1 in a final volume of 1 liter of PBS. The bacterial suspension was supplied to the flow chamber at a constant flow rate of 1 ml min−1. Attachment of cells was monitored by direct visualization of the silicone-coated bottom slides at 0, 15, 30, 45, 60, 120, 180, 240, and 300 min using a BX41 phase-contrast microscope with a long working distance objective (Olympus, United Kingdom). The microscope was also fitted with an SC30 camera, and images were captured and processed using cellSens image capture and analysis software (Olympus). Images were captured as a tagged image file format (TIFF) time stack by taking 9 images over ∼0.2 s and were collated to remove unattached planktonic cells in bulk flow from images prior to enumeration using the ImageJ particle analysis tool.
SEM of encrusted urinary catheter sections.
Catheters removed from bladder models at the ends of experiments were sectioned as shown in Fig. 2a and were viewed directly as fully hydrated unprocessed samples by using either environmental scanning electron microscopy (ESEM) or standard high-vacuum scanning electron microscopy (SEM), but with minimal sample preparation. All images were obtained using a Zeiss Evo LS15 microscope (Carl Zeiss Ltd., United Kingdom). Sections of catheters were cut longitudinally across the drainage channel to expose the biofilm, mounted onto aluminum stubs using Tissue-Tek Cryo-OCT compound (Agar Scientific), placed on a liquid cooled Deben CoolStage unit at a temperature of ∼1.5°C, and viewed in extended-pressure (EP) mode using the following parameters and conditions: a 100-μm upper EP aperture, a 500-μm lower energy-dispersive X-ray spectrometer (EDS) EP aperture, a chamber pressure of ∼570 Pa with ∼85% humidity, an accelerating (extra-high-tension [EHT]) voltage of 20 kV, and a 5-quadrant back scatter detector (5Q-BSD). For SEM of catheter cross sections, samples were mounted on aluminum stubs using Leit adhesive carbon tabs (Agar Scientific, Stansted, United Kingdom) before being sputter coated with platinum using a Quorum Q150T ES system (Quorum Technologies, United Kingdom). Samples were viewed under a high vacuum (chamber pressure, 3.5 × 10−3 Pa; humidity, 0%) using an accelerating (EHT) voltage of 5 kV and using a 5Q-BSD.
Quantification of calcium on catheter sections.
Catheter sections viewed by ESEM were subsequently used directly for calcium quantification. After viewing, sections were submerged in 2 ml of a solution of ammonium oxalate (95%, wt/vol) and oxalic acid (5%, wt/vol) and were mixed vigorously using a vortex mixer for 3 min. The resulting suspensions were incubated at room temperature for 30 min before catheter sections were removed. The remaining mixture was then centrifuged (at 3,000 × g for 10 min) and the supernatant discarded. Pellets were resuspended in 5 ml perchloric acid (0.05 mol liter−1), and samples were thoroughly mixed before being centrifuged again (at 3,000 × g for 2 min). Levels of calcium dissolved in the supernatants were then determined by using a flame photometer (model 410; Corning), calibrated using calcium standards at 100, 75, 50, and 25 ppm prior to use.
Quantification of biomass on catheter surfaces.
Catheters removed from bladder models were dissected as for ESEM analysis, transferred to sterile centrifuge tubes, and gently rinsed with SDW to remove loosely attached or nonadherent debris. Catheter sections were subsequently submerged in 1 ml of 0.5% crystal violet solution (Fisher Scientific, United Kingdom) and were incubated for 10 min at room temperature. They were then transferred to fresh tubes and were gently washed 3 times with SDW to remove excess stain. SDW was completely removed before the addition of 1 ml DMSO to tubes (Fisher Scientific), and the contents were then mixed well for 1 min, using a vortex mixer, to elute crystal violet dye retained by biofilms. The absorbances of the resulting solutions were then measured at 595 nm using a spectrophotometer (model 6300; Jenway).
Antibiotic susceptibility testing.
Antibiotic susceptibility tests were conducted according to BSAC guidelines (version 12; May 2013) (35) using Iso-Sensitest agar (Oxoid). Antibiotic discs were obtained from Oxoid, Ltd., and encompassed a range of antibiotics indicated for the treatment of UTIs according to BASC guidelines (35): cephalexin (30 μg), fosfomycin (50 μg), amdinocillin (10 μg), nalidixic acid (30 μg), norfloxacin (10 μg), nitrofurantoin (300 μg), chloramphenicol (30 μg), ciprofloxacin (5 μg), gentamicin (10 μg), amoxicillin (10 μg), and penicillin G (10 μg). Zone diameters were measured according to directions for Proteus spp., in which original zones of inhibition are measured and reduced zones created by any subsequent swarming are ignored. To assess the impact of efflux pump inhibitors (EPIs) on wild-type susceptibility profiles, either thioridazine (50 μg ml−1) or 1-(1-naphthylmethyl)piperazine (NMP; 100 μg ml−1) was incorporated into the agar prior to inoculation. The concentrations selected were based on those used by Kvist et al. (36) and were confirmed not to significantly reduce the growth of P. mirabilis B4 (data not shown).
Construction of phylogenetic trees.
Full-length translated open reading frames (ORFs) from the P. mirabilis HI4320 genome sequence, representing those disrupted in the blocking-deficient mutants, were used to search the nr data set with BLASTP. The top 100 hits by bit score (E value, 1E−5 or lower) were recovered and were aligned with the HI4320 query sequence using Clustal W, implemented in Geneious, version 6.1.4 (37). N- and C-terminal regions of alignments were trimmed to the first common amino acid and were subsequently used to construct phylogenetic trees using the Geneious (version 6.1.4) tree builder, run with the following parameters: the neighbor-joining algorithm, the Jukes-Cantor substitution model, and 1,000 bootstrap replicates. The resulting trees were used to generate a bootstrap consensus tree with a node support threshold of 40, which was annotated using Dendroscope, version 3.0.1 (38).
Statistical analysis.
All statistical analysis was performed using Prism, version 6.0c for Mac OS X (GraphPad Software Inc., USA). Data were analyzed using either Student's t test or analysis of variance (ANOVA) with Dunnett's multiple-comparison test.
RESULTS
Isolation of mutants with altered biofilm formation.
A total of 3,840 mini-Tn5 mutants were subjected to a first-pass precursor screen for general alterations in biofilm-forming ability and growth by use of a simple microtiter plate assay. This screen resulted in the isolation of 575 mutants exhibiting altered biofilm formation but levels of growth comparable to that of the wild type under the assay conditions (310 biofilm formation-enhanced [BFE] and 265 biofilm formation-deficient [BFD] mutants). The overall proportion of mutants identified as altered in biofilm formation by this screening process (14.97%) is similar to previous results obtained for the recovery of mini-Tn5 mutants with attenuated swarming motility (∼15% of mutants screened) (26). Swarming behavior in P. mirabilis is also a coordinated multicellular process of a complexity comparable to that of biofilm formation and is expected to involve similar numbers of genes and pathways (26, 39, 40). A subset of 12 mutants, encompassing both those with enhancements and those with deficiencies in biofilm formation in these high-throughput screens, were subsequently selected at random for further study and were subjected to more-detailed phenotypic and genetic characterization.
Characterization of selected mutants.
The genes disrupted in selected mutants were identified by amplification of the regions flanking the Tn insert and comparison of the nucleotide sequences recovered to the P. mirabilis HI4320 genome sequence (30). This revealed that the selected mutants had disruptions in a wide array of genes, which were dominated by those encoding cell wall-associated proteins and also included those with functions such as nitrate metabolism and efflux systems, which have not previously been associated with biofilm formation in this organism (Table 1; see also Fig. S1 in the supplemental material). To confirm that the disrupted genes were responsible for the phenotypes observed, and to account for possible polar effects of Tn insertions, the expression of ORFs downstream of the mini-Tn5 insert site in each mutant was examined. In all cases, the expression of downstream ORFs was unaffected, confirming that the disrupted genes were responsible for the phenotypes observed (see Fig. S1 in the supplemental material).
TABLE 1
| Phenotype and strain | P. mirabilis HI4320 locusa | Disrupted gene | Putative function/product of disrupted geneb | Urease production (mean ± SD)c | Motility (% of wild type) | |
|---|---|---|---|---|---|---|
| Swimming | Swarming | |||||
| Wild type (B4) | 5.7 ± 0.4 | 100 | 100 | |||
| Biofilm enhanced | ||||||
| NHBFG8 | PMI3402 | Unknown function; putative MuA-like DNA-binding protein (Pfam 02316) (40/40) | 5.9 ± 0.3 | 0 | 0 | |
| NHBFH4 | PMI0262d | mrpI | Fimbrial recombinase (40/40)d | 4.3 ± 0.1 | 36.96 | 0 |
| NHBFE8 | PMI2210 | Fimbrial subunit (38/40) | 5.3 ± 0.4 | 70.65 | 111.34 | |
| NHBFH5 | PMI2359 | glnE | Glutamate-ammonia-ligase adenylyltransferase (38/40) | 4.8 ± 0.4 | 51.09 | 0 |
| NHBFA5 | PMI1729 | rsbA | Regulator of swarming behavior; sensor kinase of two-component system (40/40) | 5.5 ± 0.1 | 116.3 | 32.99 |
| NHBFF9 | PMI0829 | bcr | Bicyclomycin resistance gene (sulfonamide resistance protein, MFS transporter) (40/40) | 4.7 ± 0.3 | 22.82 | 3.1 |
| Biofilm deficient | ||||||
| NHSH1 | PMI1608 | Unknown function; putative transmembrane protein (Pfam 02659) (39/40) | 5.6 ± 0.3 | 27.17 | 0 | |
| ABBF1.1C8 | PMI2867 | gltS | Sodium/glutamate symport carrier protein (20/20) | 5.4 ± 0.5 | 18.48 | 0 |
| NHSE5 | PMI1551 | Unknown function; putative lipoprotein (COG3016, DUF399) (40/40) | 5.2 ± 0.5 | 23.91 | 0 | |
| DLD1A6 | PMI2861 | Unknown function; putative membrane protein (COG2860, Pfam 03458) (40/40) | 5.7 ± 0.7 | 35.86 | 84.54 | |
| DLD1D11 | PMI0696d | lrp | Leucine-responsive regulator (40/40)d | 5.2 ± 0.2 | 77.17 | 34.02 |
| STS8.1D7 | PMI1479e | nirB | Nitrite reductase (30/30)e | 4.4 ± 0.1 | 100 | 77.32 |
a
The putative identities of disrupted genes are based on the correlation of sequences flanking mini-Tn5 insertions in P. mirabilis B4 to the P. mirabilis HI4320 chromosome (30). In all cases, 20- to 40-nt sequences across the transposon junction in B4 mutants were mapped to the HI4320 sequence.
b
Where HI4320 genome annotation did not indicate a putative function of the disrupted ORF, the full-length translated ORF from the HI4320 genome was used to search the Conserved Domain Database, and functionality was inferred from the resulting matches where possible (E value, 0.01 or lower). Numbers in parentheses are the numbers of nucleotides from mutant Tn-chromosome junctions matching the HI4320 genome sequence. See also Fig. S1 in the supplemental material for physical maps of surrounding gene neighborhoods. MFS, major facilitator superfamily.
c
Expressed as mmol urea hydrolyzed min−1 (mg protein)−1. No significant differences between mutants and the wild type were identified (P, ≥0.05).
d
The transposon insertion mapped to the intergenic region immediately upstream of the indicated ORF (maximum of 84 nt from the ORF start codon) and is presumed to have disrupted the promoter region of the ORF.
e
The precise insertion point is unknown due to sequence ambiguity at the Tn-chromosome junction but is within ∼10 nt of the indicated insertion point.
Mutants were also assessed for a range of traits relevant to crystalline biofilm formation and catheter blockage. Analysis of urease activity revealed that none of the selected mutants were attenuated in the ability to generate the urease enzyme and to hydrolyze urea (a key activity underpinning catheter blockage and encrustation) (Table 1). However, most mutants were found to be deficient in both swimming and swarming motilities; a large proportion exhibited nonswarming phenotypes under the assay conditions (incubation on 1.5% LB agar at 37°C for 8 h) (Table 1).
Role of disrupted genes in encrustation and blockage of urethral catheters.
To ascertain if high-throughput screens had recovered mutants relevant to catheter blockage, we used an in vitro model of the catheterized urinary tract to compare the abilities of selected mutants to form crystalline biofilms and encrust urethral catheters with those of the wild type (see Fig. S2 in the supplemental material) (27). Models were fitted with size 14 French all-silicone catheters and were supplied with a standardized AU medium (27). The pH of urine in “bladders” and the numbers of viable cells were also measured at the ends of experiments, to confirm that the mutants were able to elevate pH and to survive in this system at levels commensurate with that of the wild type.
Of the 12 mutants evaluated in this model, none showed significant differences in the ability to elevate urinary pH or to persist in bladder models (Fig. 1b and c), but 4 mutants exhibited significant differences in the time taken to block catheters; both deficiencies and enhancements in blockage were evident (Fig. 1a). Mutants displaying significant decreases in the time to blockage (indicating enhanced blocking ability) had disruptions in genes of unknown functions (NHBFG8, NHSH1), whereas mutants generating significantly increased blockage times (indicating attenuation of blocking ability) had disruptions in genes associated with nitrite metabolism (STS8.1D7) and a putative multidrug efflux pump (NHBFF9) (Fig. 1; Table 1).
FIG 1
However, the phenotypes of all selected mutants in the bladder model system (blocking deficient, blocking enhanced, or with unaltered blocking) did not correlate well with the biofilm formation phenotypes observed in the initial high-throughput microtiter plate screens (Fig. 1). Similar spectra of blockage times were observed for the two groups of mutants tested (BFD and BFE): 50% of mutants isolated as biofilm enhanced, and tested in bladder models, exhibited reductions in blocking ability, with the converse true for biofilm-deficient mutants tested (Fig. 1). It is not, however, unexpected that the phenotypes observed in the intermediate crystal violet (CV) screen do not accurately predict behavior in the more-complex bladder model system; this phenomenon confirms the impact of the physicochemical parameters of the catheterized urinary tract on biofilm formation (reviewed in reference 8) and reinforces the importance of subsequently utilizing representative models to study this process. Nevertheless, the use of the simple high-throughput CV screen as a precursor assay, to generate a pool of initial candidate mutants, likely to be enriched in mutants with defects in crystalline biofilm formation and suitable for assessment in the more rigorous (but low-throughput) bladder model system, is a useful approach for the recovery of mutants with attenuated abilities to block urethral catheters.
Characterization of crystalline biofilm formation in blocking-deficient mutants.
Because mutants with reduced abilities to encrust and block urethral catheters are most likely to provide insights relevant to preventing these processes, crystalline biofilm formation by mutants significantly attenuated in catheter blockage (NHBFF9 and STS8.1D7 [Fig. 1; Table 1]) was investigated in greater detail. To permit direct comparison of biofilm development between mutants and the wild type, in vitro bladder models were again utilized, but experiments were terminated after a set time (10 h), and levels of biofilm formation and biofilm structure were then assessed.
To investigate the general development and architecture of biofilms formed by mutants and the wild type, ESEM was used to visualize unprocessed, fully hydrated biofilms directly on catheter surfaces (Fig. 2). ESEM imaging showed levels of crystalline biofilm formation to be greatest in regions directly adjacent to the eyehole, with a progressive reduction evident toward distal catheter sections (Fig. 2b). However, in all catheter sections examined using ESEM, mutants were observed to generate less-expansive biofilms than the wild type, and alterations in architecture were apparent (Fig. 2b). Examination of platinum-coated catheter cross sections under high-vacuum SEM supported the ESEM observations and clearly demonstrated considerable reductions in biofilm thickness in both mutants examined (Fig. 2c).
FIG 2
To support the subjective ESEM and SEM observations, levels of biomass and encrustation that had developed on catheter sections were also quantified. Biomass was measured using a modified crystal violet staining procedure on representative catheter sections, whereas levels of encrustation were evaluated using flame photometric quantification of calcium on the same catheter sections viewed by ESEM. Both analyses supported the findings of ESEM and SEM, showing reduced crystalline biofilm formation in mutants, and significant reductions in the levels of biomass and encrustation were noted on all catheter sections (Fig. 2).
Collectively, image analysis and quantification of crystalline biofilm formation confirm that deficiencies in the ability of mutants to block catheters stem from attenuated biofilm formation in the bladder model system. However, with mutant NHBFF9, significant reductions in urinary pH from that of the wild type were also noted at the end of these timed experiments, and although the pH was still elevated to levels sufficient for crystal formation, these reductions may have contributed to the reductions in catheter encrustation observed for this mutant (see Table S2 in the supplemental material). In contrast, NHBFF9 also exhibited significantly greater numbers of viable cells in “bladders” at the 10-h experimental endpoint than the wild type, suggesting potentially increased survival in the bladder model system over this period, but it is notable that this did not result in greater biofilm formation.
Attachment to silicone.
To determine if mutants with significant reductions in the ability to block urethral catheters were attenuated in attachment to catheter biomaterials, adherence to silicone was evaluated using a parallel plate flow chamber system. This demonstrated that blocking-deficient mutants (STS8.1D7 and NHBFF9) showed no less ability to attach to silicone than the wild type over the duration of these experiments (Fig. 3). Therefore, initial events in biofilm formation and attachment to catheter surfaces appear to be unaffected by gene disruptions in these mutants, pointing to defects in processes underpinning the later stages of biofilm development and maturation, which occur after attachment to catheter surfaces.
FIG 3
Efflux systems modulate both biofilm formation and antibiotic susceptibility in P. mirabilis.
The disruption of a putative multidrug efflux pump in the blocking-deficient mutant NHBFF9 (P. mirabilis HI4320 locus PMI0829 [Table 1]) raises the possibility that this system may be involved in both biofilm formation and the antibiotic susceptibility profile of P. mirabilis. To investigate this possibility, the susceptibilities of NHBFF9 to a range of antibiotics used to treat UTIs were compared with those of the wild-type organism by using disc diffusion tests (35). In these experiments, NHBFF9 displayed significantly increased susceptibility to fosfomycin but significantly reduced susceptibilities to nalidixic acid and amoxicillin (Fig. 4). When sub-MIC levels of chemical inhibitors of efflux systems were incorporated into the medium, the susceptibilities of the wild type to fosfomycin and amoxicillin were altered to levels comparable with the NHBFF9 phenotype (Fig. 4).
FIG 4
The potential clinical significance of the susceptibility profiles observed was evaluated by interpreting inhibition zone diameters using breakpoint values, in order to classify results as indicative of clinical resistance or sensitivity (35). In this assessment, the wild type was classified as clinically resistant to fosfomycin, but mutant NHBFF9 was classified as sensitive. Exposure of the wild type to chemical EPIs resulted in a shift to a phenotype of clinical sensitivity to fosfomycin, comparable to that of NHBFF9 (Fig. 4). With nalidixic acid and amoxicillin, no change in the classification of the wild type as sensitive was generated by EPI exposure, and the putative efflux mutant NHBFF9 was also sensitive (Fig. 4).
Phylogenetic analysis of the efflux system disrupted in NHBFF9.
In order to explore the relationship of the gene disrupted in NHBFF9 with homologous genes in other species, phylogenetic trees were constructed from sequences related to the corresponding full-length ORF from P. mirabilis HI4320 (PMI0829 [Table 1]). Phylograms revealed that the gene disrupted in NHBFF9 was most closely related to homologues identified in other members of the Enterobacteriaceae, in particular Proteus penneri, Morganella morganii, and a Providencia sp., all of which are causative agents of CAUTI (Fig. 5). Also of note were the broader relationships indicated by this analysis: homologous sequences recovered in BLAST searches were dominated by those derived from bacterial genera and species known to cause urinary tract infections, or infection at other body sites (Fig. 5). In particular, the close relationship of the gene disrupted in NHBFF9 with homologues in other uropathogens may be indicative of common adaptations to the catheterized urinary tract, highlighting potential targets for the development of strategies to inhibit biofilm formation in a wide range of species responsible for CAUTI.
FIG 5
DISCUSSION
Despite the fact that catheterization is generally considered a last resort for bladder management, it has been estimated that ∼26% of hospital patients in the United Kingdom are fitted with urinary catheters during their stays, while the prevalence of long-term bladder catheterization among elderly nursing home patients may be as high as 40% in some locations (6, 41). Measures to prevent CAUTIs, such as the use of sterile, closed drainage systems and antimicrobial coatings or lubricants, in addition to proper aseptic handling and fitting of catheters, can have significant benefits for patients subjected to short-term catheterization. However, even with these measures and meticulous nursing care, the onset of CAUTI in patients undergoing long-term catheterization is almost a certainty (5, 23). This clearly unacceptable situation has been succinctly summarized by Saint and Chenworth (2003), who pointed out that although urinary catheters are relatively inexpensive, their use exacts an enormous toll in terms of the impact on patient welfare and the financial cost of associated treatment (42).
Given the increasing population of elderly individuals in developed countries across the world and the integral role of urinary catheters in modern medical care, it seems likely that the use of these devices in hospitals and community care settings will continue to rise, along with the incidence of CAUTIs and associated morbidity, mortality, and treatment costs. In light of this, there is an urgent need for effective countermeasures to combat or prevent CAUTI. A greater understanding of the mechanisms underlying CAUTI pathogenesis will be key to the development of such strategies, and biofilm formation is a major feature of these infections (5, 8, 14, 39, 42).
Here we employ a random transposon mutagenesis approach, coupled with representative models of the catheterized urinary tract, to elucidate genes and pathways relevant to crystalline biofilm formation by P. mirabilis, a common and particularly problematic cause of CAUTI with serious clinical consequences (4, 12–14, 27). In doing so, we reveal functions associated with both biofilm formation and antibiotic susceptibility in this organism, and we highlight potential strategies for combating CAUTI that can now be tested further.
In the case of the blocking-deficient mutant STS8.1D7, the mini-Tn5 insert was mapped to a homologue of nirB, a large-subunit nitrite reductase (P. mirabilis HI4320 locus PMI1479 [Table 1]). Disruption of nitrate and nitrite metabolism in Pseudomonas aeruginosa also results in a phenotype similar to that of our P. mirabilis mutant STS8.1D7, with swarming, biofilm formation, and virulence inhibited but no alteration in swimming motility (43). In E. coli, which is closely related to P. mirabilis and is the most common cause of UTI, NirB, along with the small-subunit protein NirD, is one of two mechanisms available for nitrite reduction (for a summary, see Fig. S3 in the supplemental material) (44). These are expressed under anaerobic conditions; the nirBD mechanism is dominant when nitrite levels are high, and ammonia is produced as an end product (44). A comparable pathway for nitrate metabolism, which also results in the production of ammonia from nirBD-driven nitrite reduction, exists in P. mirabilis (see Fig. S3 in the supplemental material), and the presence of nitrites in urine is considered indicative of urinary tract infection with E. coli, P. mirabilis, or other Enterobacteriaceae (45).
Given the pivotal role of nitrogen metabolism and ammonia production in the alkalinization of urine and crystal formation, it seems logical to hypothesize that perturbation of this pathway will impact negatively on crystalline biofilm formation and the ability to block urethral catheters. This is in keeping with the reductions in crystalline biofilm formation levels observed with STS8.1D7 and the increased time this organism takes to block urethral catheters. However, no significant differences in the ability to elevate urinary pH in bladder models was observed, indicating that ammonia production via this pathway does not contribute significantly to the initial elevation of urinary pH, which is undoubtedly driven by the potent P. mirabilis urease enzyme (reviewed in references 8 and 39). In addition, catheter encrustation is not an outcome of infection with non-urease-producing pathogens known to express the nirBD system (such as E. coli), and urease-defective mutants of P. mirabilis fail to form crystalline biofilms in laboratory models, providing a further argument against an overt and integral role of ammonia produced via nitrate metabolism in crystal formation (8, 15, 16, 46).
Alternatively, nirB disruption may lead to a more general impact on the growth and survival of cells within the bladder model system, and in other closely related species, nitrite reduction by nirBD is considered to be important for the detoxification of cellular nitrite generated during anaerobic growth (44, 47). However, evaluation of viable cell numbers in bladder model systems run for 10 h, or until catheters became blocked, indicated that the survival and persistence of P. mirabilis STS8.1D7 were not compromised. Still, there remains considerable scope for nirB to fulfill an important role specifically in biofilm development and to contribute to later stages of this process. As the community matures and cell numbers increase, it is expected that the availability of nutrients and oxygen will be reduced, while waste products accumulate in the biofilm and must be removed or neutralized (36, 48).
Therefore, it is conceivable that nirB may facilitate the detoxification of the nitrite that accumulates in maturing P. mirabilis biofilms and that loss of this function limits the rate at which biofilms can develop and expand. This hypothesis is consistent with the reductions in overall biomass observed in 10-h STS8.1D7 catheter biofilms without significant reductions in the numbers of viable planktonic cells; it is also consistent with the results of flow chamber experiments, which point to defects in biofilm maturation and expansion rather than in initial attachment to catheter surfaces (Fig. 2 and 3). While validation of these theories will require further study, characterization of the nirB-deficient mutant in this study has provided additional insight into the mechanisms underlying crystalline biofilm formation and has identified new avenues of research that can now be explored further.
This study also highlighted the involvement of efflux systems in the development of P. mirabilis crystalline biofilms; the gene disrupted in the blocking-deficient mutant NHBFF9 is predicted to form a component of a putative multidrug efflux system (PMI0829) (30). The observation of reduced crystalline biofilm formation arising from the disruption of this efflux system corresponds with recent studies highlighting a role for these extrusion systems in bacterial biofilm formation (36, 49–51). Efflux pumps have been shown to be expressed at higher levels in biofilm-associated cells than in their planktonic counterparts, while studies involving chemical inhibition of pumps or deletion of the relevant genes have demonstrated reduction or abolition of biofilm formation in uropathogenic species such as Staphylococcus aureus, E. coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa (36, 49, 52). Although the precise role of efflux pumps in bacterial biofilm formation remains to be elucidated, it has been suggested that efflux constitutes an important waste management process in the spatially challenged biofilm community, and it is likely that this is also true for P. mirabilis biofilm development (36).
Conversely, NHBFF9 also exhibited severely reduced swimming and swarming motilities, and swimming motility has been shown to be important for biofilm formation in other organisms (28, 53, 54) Although the possibility that the reduced motility of this mutant contributes to its reduced blocking ability cannot be fully excluded, previous studies with P. mirabilis mutants found neither swimming or swarming to be vital for blockage of catheters or attachment to catheter biomaterials (26, 34). In conjunction with the flow chamber experiments conducted in this study, which showed that the attachment of NHBFF9 to silicone was unaffected, these observations indicate that the motility of NHBFF9 is not a significant factor in the reduced blocking ability of this mutant and that disruption of the putative efflux pump is responsible for the phenotype observed.
Characterization of NHBFF9 also provided insights into links between biofilm formation and antibiotic susceptibility in P. mirabilis, which may be directly relevant to the prophylaxis or treatment of P. mirabilis CAUTI. This is consistent with findings in other organisms, where inhibition of efflux systems has already been shown to reduce the elevated antibiotic resistance profile characteristic of biofilms (36, 52). In the present study, disruption of efflux systems increased susceptibility to fosfomycin and resulted in a clinically sensitive phenotype, contrasting with the predicted clinical resistance of the wild type and illustrating the potential for EPIs to maximize the utility of existing antibiotics. In contrast, paradoxical effects of efflux inhibition were noted for nalidixic acid and amoxicillin: efflux disruption reduced overall susceptibility to these antibiotics. Similar paradoxical effects have been noted in other organisms when efflux systems have been disrupted or inhibited, but in this study, these effects did not result in an interpretation of clinical resistance (55, 56) Taken together, the characterization of the NHBFF9 efflux mutant and the effects of EPIs on wild-type susceptibility profiles point to a potentially viable and straightforward strategy for controlling crystalline biofilm formation in P. mirabilis, where EPIs and antibiotics could be used synergistically to reduce biofilm formation. Of particular note in this regard is the fact that some EPIs are compounds, or derivatives of compounds, already in use as approved human drugs for other purposes, which should significantly shorten the path to their clinical application in the care of catheterized patients, if they are proven to be effective in this arena.
However, despite the novelty and potential utility of the findings presented here, it should be noted that no mutants tested in bladder models were completely deficient in the abilities to encrust and block catheters and that in all cases mutants were eventually able to completely occlude urine flow through these devices. Furthermore, it is almost certain that while the urease activity of P. mirabilis remains unchecked, and this organism is able to generate conditions amenable to crystal formation, the prevention of biofilm formation will most likely serve only to reduce the time to blockage rather than affording total prevention.
Nevertheless, the experimental parameters we employed in bladder models represented a worst-case scenario, where an already heavily infected urinary tract was catheterized, and may be likened to catheter changes in patients with established high-level infections. As such, this represents a stringent and conservative test of blocking ability in these mutants, with encrustation evaluated under the conditions most favorable for biofilm formation and crystal development. Therefore, it is likely that the blocking-deficient mutants identified here would show greater attenuation under conditions simulating the early stages of infection, as the urinary tract first becomes colonized, and that intervention at this point would also be predicted to have a more pronounced impact in reducing biofilm formation and associated complications. Regardless of ability to completely abolish catheter blockage, the development of strategies that prolong the life of urinary catheters, inhibit biofilm formation, and reduce the potential for blockage and more serious complications will be of much clinical benefit.
ACKNOWLEDGMENTS
We thank Heather Catty, Rowena Berterelli, and Darren McCabe for excellent technical support, as well as Lucas Bowler and Caroline Jones for constructive criticism and comments during study development and the preparation of the manuscript.
N.H. was supported by a Ph.D. studentship from the Healthcare Infection Society (awarded to B.V.J.), with access to scanning electron microscopy supported by funding from Quorum Technologies Ltd. Work presented in this article was also supported by the Society for Applied Microbiology and the University of Brighton. L.A.O. and B.V.J. are also supported by funding from the Medical Research Council (grant G0901553 awarded to B.V.J.).
Supplemental Material
File (zii999090601so1.pdf)
- Download
- 3.45 MB
REFERENCES
1.
Darouiche RO. 2001. Device-associated infections: a macroproblem that starts with microadherence. Clin. Infect. Dis. 33:1567–1572.
2.
Saint S, Wiese J, Amory J, Bernstein ML, Patel UD, Zemencuk JK, Bernstein SJ, Lipsky BA, and Hofer TP. 2000. Are physicians aware of which of their patients have indwelling urinary catheters? Am. J. Med. 109:476–480.
3.
Getliffe K and Newton T. 2006. Catheter-associated urinary tract infection in primary and community health care. Age Ageing 35:477–481.
4.
Stickler DJ and Zimakoff J. 1994. Complications of urinary-tract infections associated with devices used for long-term bladder management. J. Hosp. Infect. 28:177–194.
5.
Jacobsen SM, Stickler DJ, Mobley HL, and Shirtliff ME. 2008. Complicated catheter-associated urinary tract infections due to Escherichia coli and Proteus mirabilis. Clin. Microbiol. Rev. 21:26–59.
6.
McNulty C, Freeman E, Smith G, Gunn K, Foy C, Tompkins D, Brady A, and Cartwright K. 2003. Prevalence of urinary tract catheterization in UK nursing homes. J. Hosp. Infect. 55:119–123.
7.
Nickel JC. 1991. Catheter-associated urinary tract infection: new perspectives on old problems. Can. J. Infect. Control 6:38–42.
8.
Stickler DJ. 2008. Bacterial biofilms in patients with indwelling urinary catheters. Nat. Clin. Pract. Urol. 5:598–608.
9.
Tambyah PA. 2004. Catheter-associated urinary tract infections: diagnosis and prophylaxis. Int. J. Antimicrob. Agents 24(Suppl 1):S44–S48.
10.
Hooton TM, Bradley SF, Cardenas DD, Colgan R, Geerlings SE, Rice JC, Saint S, Schaeffer AJ, Tambayh PA, Tenke P, and Nicolle LE. 2010. Diagnosis, prevention, and treatment of catheter-associated urinary tract infection in adults: 2009 International Clinical Practice Guidelines from the Infectious Diseases Society of America. Clin. Infect. Dis. 50:625–663.
11.
Trautner BW. 2010. Management of catheter-associated urinary tract infection. Curr. Opin. Infect. Dis. 23:76–82.
12.
Mobley HLT. 1996. Virulence of Proteus mirabilis, p 245–265. In Mobley HLT and Warren JW (ed), Urinary tract infections: molecular pathogenesis and clinical management. American Society for Microbiology, Washington, DC.
13.
Warren JW. 1991. The catheter and urinary tract infection. Med. Clin. North Am. 75:481–493.
14.
Macleod SM and Sticker DJ. 2007. Species interactions in mixed-community crystalline biofilms on urinary catheters. J. Med. Microbiol. 56:1549–1557.
15.
Stickler DJ, Ganderton L, King J, Nettleton J, and Winters C. 1993. Proteus mirabilis biofilms and the encrustation of urethral catheters. Urol. Res. 21:407–411.
16.
Stickler DJ, Morris N, Moreno MC, and Sabbuba N. 1998. Studies on the formation of crystalline bacterial biofilms on urethral catheters. Eur. J. Clin. Microbiol. Infect. Dis. 17:649–652.
17.
Stickler DJ and Morgan SD. 2008. Observations on the development of the crystalline bacterial biofilms that encrust and block Foley catheters. J. Hosp. Infect. 69:350–360.
18.
Griffith DP, Musher DM, and Itin C. 1976. Urease: the preliminary cause of infection induced urinary stones. Invest. Urol. 13:346–350.
19.
Jones BD and Mobley HL. 1987. Genetic and biochemical diversity of ureases of Proteus, Providencia, and Morganella species isolated from urinary tract infections. Infect. Immun. 55:2190–2203.
20.
Clapham L, McLean RJC, Nickel JC, Downey J, and Costerton JW. 1990. The influence of bacteria on struvite crystal habit and its importance in urinary stone formation. J. Crystal Growth 104:475–484.
21.
Dumanski AJ, Hedelin H, Edinliljegren A, Beauchemin D, and McLean RJC. 1994. Unique ability of the Proteus mirabilis capsule to enhance mineral growth in infectious urinary calculi. Infect. Immun. 62:2998–3003.
22.
Cox AJ and Hukins DWL. 1989. Morphology of mineral-deposits on encrusted urinary catheters investigated by scanning electron-microscopy. J. Urol. 142:1347–1350.
23.
Kunin CM. 1997. Urinary tract infections: detection, prevention and management, 5th ed, p 226–278. Williams & Wilkins, Baltimore, MD.
24.
Sabbuba NA, Mahenthiralingam E, and Stickler DJ. 2003. Molecular epidemiology of Proteus mirabilis infections of the catheterized urinary tract. J. Clin. Microbiol. 41:4961–4965.
25.
Morris NS, Stickler DJ, and Winters C. 1997. Which indwelling urethral catheters resist encrustation by Proteus mirabilis biofilms? Br. J. Urol. 80:58–63.
26.
Jones BV, Young R, Mahenthiralingam E, and Stickler DJ. 2004. Ultrastructure of Proteus mirabilis swarmer cell rafts and role of swarming in catheter-associated urinary tract infection. Infect. Immun. 72:3941–3950.
27.
Stickler DJ, Morris NS, and Winters C. 1999. Simple physical model to study formation and physiology of biofilms on urethral catheters. Methods Enzymol. 310:494–501.
28.
O'Toole GA and Kolter R. 1998. Flagellar and twitching motility are necessary for biofilm development. Mol. Microbiol. 30:295–304.
29.
Manoil C. 2000. Tagging exported proteins using Escherichia coli alkaline phosphatase gene fusions. Methods Enzymol. 326:35–47.
30.
Pearson MM, Sebaihia M, Churcher C, Quail MA, Seshasayee AS, Luscombe NM, Abdellah Z, Arrosmith C, Atkin B, Chillingworth T, Hauser H, Jagels K, Moule S, Mungall K, Norbertczak H, Rabbinowitsch E, Walker D, Whithead S, Thomson NR, Rather PN, Parkhill J, and Mobley HL. 2008. Complete genome sequence of uropathogenic Proteus mirabilis, a master of both adherence and motility. J. Bacteriol. 190:4027–4037.
31.
Altschul SF, Gish W, Miller W, Myers EW, and Lipman DJ. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403–410.
32.
Kanehisa M, Goto S, Sato Y, Furumichi M, and Tanabe M. 2012. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res. 40(Database issue):D109–D114.
33.
Creno RJ, Wenk RE, and Bohlig P. 1970. Automated micromeasurement of urea using urease and the Berthelot reaction. Am. J. Clin. Pathol. 54:828–832.
34.
Jones BV, Mahenthiralingam E, Sabbuba NA, and Stickler DJ. 2005. Role of swarming in the formation of crystalline Proteus mirabilis biofilms on urinary catheters. J. Med. Microbiol. 54(Part 9):807–813.
35.
British Society for Antimicrobial Chemotherapy. May 2013. BSAC methods for antimicrobial susceptibility testing, version 12. http://bsac.org.uk/wp-content/uploads/2012/02/Version-12-Apr-2013_final.pdf.
36.
Kvist M, Hancock V, and Klemm P. 2008. Inactivation of efflux pumps abolishes bacterial biofilm formation. Appl. Environ. Microbiol. 74:7376–7382.
37.
Thompson JD, Higgins DG, and Gibson TJ. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680.
38.
Huson DH and Scornavacca C. 2012. Dendroscope 3: an interactive tool for rooted phylogenetic trees and networks. Syst. Biol. 61:1061–1067.
39.
Armbruster CE and Mobley HL. 2012. Merging mythology and morphology: the multifaceted lifestyle of Proteus mirabilis. Nat. Rev. Microbiol. 10:743–754.
40.
Pearson MM, Rasko DA, Smith SN, and Mobley HLT. 2010. Transcriptome of swarming Proteus mirabilis. Infect. Immun. 78:2834–2845.
41.
Glynn A, Ward V, Wilson J, Charlett A, Cookson B, Taylor L, and Cole. 1997. Hospital-acquired infection: surveillance, policies and practice. Public Health Laboratory Service, London, United Kingdom.
42.
Saint S and Chenworth CE. 2003. Biofilms and catheter-associated urinary tract infection. Infect. Dis. Clin. North Am. 17:411–432.
43.
Van Alst NE, Picardo KF, Iglewski BH, and Haidaris CG. 2007. Nitrate sensing and metabolism modulate motility, biofilm formation, and virulence in Pseudomonas aeruginosa. Infect. Immun. 75:3780–3790.
44.
Wang H and Gunsalus RP. 2000. The nrfA and nirB nitrite reductase operons in Escherichia coli are expressed differently in response to nitrate than to nitrite. J. Bacteriol. 182:5813–5822.
45.
Franz M and Hörl WH. 1999. Common errors in diagnosis and management of urinary tract infection. I. Pathophysiology and diagnostic techniques. Nephrol. Dial. Transplant. 14:2746–2753.
46.
Stickler DJ, Lear JC, Morris NS, Macleod SM, Downer A, Cadd DH, and Feast WJ. 2006. Observations on the adherence of Proteus mirabilis onto polymer surfaces. J. Appl. Microbiol. 100:1028–1033.
47.
Page L, Griffiths L, and Cole JA. 1990. Different physiological roles of two independent pathways for nitrite reduction to ammonia by enteric bacteria. Arch. Microbiol. 154:349–354.
48.
Donlan R and Costerton WJ. 2002. Biofilms: survival mechanisms of clinically relevant microorganisms. Clin. Microbiol. Rev. 15:167–193.
49.
Matsumura K, Furukawa S, Ogihara H, and Morinaga Y. 2011. Roles of multidrug efflux pumps in the biofilm formation of Escherichia coli K12. Biocontrol Sci. 16:69–72.
50.
Soto SM. 2013. Role of efflux pumps in the antibiotic resistance of bacteria embedded in a biofilm. Virulence 4:223–229.
51.
Baugh S, Ekanayaka AS, Piddock LJV, and Webber MA. 2012. Loss of or inhibition of all multidrug resistance efflux pumps in Salmonella enterica serovar Typhimurium results in impaired ability to form a biofilm. J. Antimicrob. Chemother. 67:2409–2417.
52.
Zhang L and Mah T-F. 2008. Involvement of a novel efflux system in biofilm-specific resistance to antibiotics. J. Bacteriol. 190:4447–4452.
53.
Pratt LA and Kolter R. 1998. Genetic analysis of Escherichia coli biofilm formation: roles of flagella, motility, chemotaxis and type I pili. Mol. Microbiol. 30:285–293.
54.
Watnick PI, Lauriano CM, Klose KE, Croal L, and Kolter R. 2001. The absence of a flagellum leads to altered colony morphology, biofilm development and virulence in Vibrio cholerae O139. Mol. Microbiol. 39:223–235.
55.
Sulavik MC, Houseweart C, Cramer C, Jiwani N, Murgolo N, Greene J, DiDomenico B, Shaw KJ, Miller GH, Hare R, and Shimer G. 2001. Antibiotic susceptibility profiles of Escherichia coli strains lacking multidrug efflux pump genes. Antimicrob. Agents Chemother. 45:1126–1136.
56.
Turner RJ, Taylor DE, and Weiner JH. 1997. Expression of Escherichia coli TehA gives resistance to antiseptics and disinfectants similar to that conferred by multidrug efflux pumps. Antimicrob. Agents Chemother. 41:440–444.
Information & Contributors
Information
Published In
Infection and Immunity
Volume 82 • Number 4 • April 2014
Pages: 1616 - 1626
Editor: S. M. Payne
PubMed: 24470471
Copyright
© 2014 American Society for Microbiology.
History
Received: 23 December 2013
Accepted: 22 January 2014
Published online: 19 March 2014
Contributors
Editor
S. M. Payne
Editor
Metrics & Citations
Metrics
Note:
- For recently published articles, the TOTAL download count will appear as zero until a new month starts.
- There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
- Citation counts come from the Crossref Cited by service.
Citations
Citation
2014.
Elucidating the Genetic Basis of Crystalline Biofilm Formation in Proteus mirabilis. Infect Immun
82:.
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.