Journal of Bacteriology, July 2004, p . 4665-4684, Vol . 186, No . 14

Global Gene Expression in Staphylococcus aureus Biofilms

Karen E . Beenken,¹ Paul M . Dunman,² Fionnuala McAleese,² Daphne Macapagal,² Ellen Murphy,² Steven J . Projan,³ Jon S . Blevins,⁴ and Mark S . Smeltzer^1*

Department of Microbiology and Immunology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205,¹ Wyeth Research, Pearl River, New York 10965,² Wyeth Protein Technologies, Cambridge, Massachusetts 02140,³ Department of Microbiology and Immunology, The University of Texas Southwestern Medical Center, Dallas, Texas 75390⁴

Received 13 January 2004/ Accepted 5 April 2004

 
We previously demonstrated that mutation of the staphylococcal accessory regulator (sarA) in a clinical isolate of Staphylococcus aureus (UAMS-1) results in an impaired capacity to form a biofilm in vitro (K . E . Beenken, J . S . Blevins, and M . S . Smeltzer, Infect . Immun . 71:4206-4211, 2003) . In this report, we used a murine model of catheter-based biofilm formation to demonstrate that a UAMS-1 sarA mutant also has a reduced capacity to form a biofilm in vivo . Surprisingly, mutation of the UAMS-1 ica locus had little impact on biofilm formation in vitro or in vivo . In an effort to identify additional loci that might be relevant to biofilm formation and/or the adaptive response required for persistence of S . aureus within a biofilm, we isolated total cellular RNA from UAMS-1 harvested from a biofilm grown in a flow cell and compared the transcriptional profile of this RNA to RNA isolated from both exponential- and stationary-phase planktonic cultures . Comparisons were done using a custom-made Affymetrix GeneChip representing the genomic complement of six strains of S . aureus (COL, N315, Mu50, NCTC 8325, EMRSA-16 [strain 252], and MSSA-476) . The results confirm that the sessile lifestyle associated with persistence within a biofilm is distinct by comparison to the lifestyles of both the exponential and postexponential phases of planktonic culture . Indeed, we identified 48 genes in which expression was induced at least twofold in biofilms over expression under both planktonic conditions . Similarly, we identified 84 genes in which expression was repressed by a factor of at least 2 compared to expression under both planktonic conditions . A primary theme that emerged from the analysis of these genes is that persistence within a biofilm requires an adaptive response that limits the deleterious effects of the reduced pH associated with anaerobic growth conditions .

 
Staphylococcus aureus is a prominent human pathogen that causes a wide variety of infections . Of particular interest in our laboratory are musculoskeletal infections including those associated with orthopedic implants . The hallmark characteristic of these infections is formation of a biofilm, which consists of multiple layers of bacteria encased within an exopolysaccharide glycocalyx . The presence of this glycocalyx protects the enclosed bacteria from host defenses and impedes delivery of at least some antibiotics (64) . Moreover, bacteria within biofilms adopt a phenotype that confers intrinsic resistance to many antibiotics . For example, the reduced growth rate of biofilm-associated bacteria limits the efficacy of antibiotics that target cell wall biosynthesis, while the reduced oxidative metabolism limits the uptake of aminoglycosides (33, 64, 65) . Consequently, biofilm-associated infections are recalcitrant to antimicrobial therapy and often require surgical intervention to debride infected tissues and/or remove colonized implants .

The formation of three-dimensional biofilms is a complex process that can be subdivided into the relatively distinct phases of attachment, accumulation, maturation, and dispersal (10) . With respect to staphylococcal biofilms, the primary emphasis so far has been placed on the attachment and accumulation phases, which appear to be mediated by different types of adhesins . More specifically, a group of surface-exposed proteins collectively referred to as MSCRAMMs (microbial surface components recognizing adhesive matrix molecules) (48) appear to be the primary determinants responsible for the initial attachment to both native tissues and biomaterials, while the accumulation phase appears to be dependent on polysaccharide adhesins that promote adhesive interactions between bacterial cells (26) . Although a number of candidate polysaccharides have been described, there is an emerging consensus that the primary determinant of the accumulation phase of staphylococcal biofilm formation is the polysaccharide intercellular adhesin (PIA), production of which is dependent upon the genes within the icaADBC operon (28) . Composition studies have demonstrated that PIA consists of polymeric N-acetylglucosamine, and for this reason it has also been referred to as PNAG (40) .

The ica operon was first identified in Staphylococcus epidermidis (28) and has been studied most extensively in that species . However, it is also present and appears to serve the same function in S . aureus (14) . Most S . aureus strains appear to contain the entire ica operon (14, 22, 53), although there are reports to the contrary (3), and it is clear that there are strain-dependent differences with respect to the overall capacity to form a biofilm in vitro (5, 14, 53) . The ica operon is subject to phase variation in S . epidermidis (75), and a number of studies have indicated that expression of ica in both S . epidermidis and S . aureus is also subject to environmental regulation . Perhaps most importantly, McKenney et al . (42) demonstrated that PNAG production in S . aureus is enhanced during in vivo growth . Rachid et al . (52) subsequently demonstrated that expression of ica is at least partially controlled by the stress response transcription factor ^B . In addition, anaerobic growth was found to induce expression of the ica operon and PIA production in both S . epidermidis and S . aureus (15) .

Recently, Conlon et al . (12) reported that icaR, which is located immediately upstream of the ica operon, encodes a repressor that is important for the environmental regulation of ica expression in S . epidermidis . However, studies done with S . aureus have demonstrated that regulation of ica expression and the ability to form a biofilm also involve regulatory elements other than ^B and IcaR (66) . Included among these additional regulatory loci are the accessory gene regulator (agr) and the staphylococcal accessory regulator (sarA) . The agr locus encodes a two-component quorum-sensing system that modulates production of a regulatory RNA molecule (RNAIII) in a density-dependent manner . Induction of RNAIII synthesis results in reduced production of surface proteins (e.g., MSCRAMMs) and a concomitant increase in production of exotoxins (4, 45) . Production of -toxin, which is encoded within the RNAIII locus, has been negatively correlated with biofilm formation (69, 70) . This suggests that strains expressing agr at high levels would have a reduced capacity to form a biofilm, which is consistent both with our results (5) and results from other laboratories (70) .

The sarA locus encodes a 14.5-kDa DNA-binding protein (SarA) that is required at least under some growth conditions for maximum expression from the agr and RNAIII promoters (29) . This would imply that mutation of sarA would limit production of RNAIII and thereby enhance the ability to form a biofilm . However, recent reports have confirmed that mutation of sarA results in a reduced capacity to form a biofilm (5, 66) . SarA also regulates expression of other genes in an agr-independent manner (6, 19, 72, 74), and Valle et al . (66) recently demonstrated that mutation of sarA results in reduced transcription of the ica operon and a reduced capacity to produce PNAG . They also suggested that SarA may promote biofilm formation in an indirect manner by suppressing transcription of a repressor of PNAG synthesis or a protein involved in the turnover of PNAG .

The persistence of bacteria within a biofilm also requires an adaptive response appropriate for the sessile lifestyle . The availability of complete bacterial genome sequences has facilitated the use of microarray technologies to identify genes that are differentially expressed by biofilm-encased bacteria . Using an array representing 99% of the Bacillus subtilis genome, Stanley et al . (63) identified 519 genes that were differentially expressed in biofilms as opposed to planktonic cultures . Similarly, Schembri et al . (59) found that 5 to 10% of the genes in the Escherichia coli genome were differentially expressed in biofilms, depending on which planktonic growth condition was used as a reference . Included among these genes were 30 of the 65 genes previously reported to be under the regulatory control of the general stress response regulator rpoS (36) . Schembri et al . (59) subsequently demonstrated that an E . coli rpoS mutant was incapable of forming a biofilm . However, Whiteley et al . (71) found that expression of rpoS was repressed in Pseudomonas aeruginosa biofilms and that a P . aeruginosa rpoS mutant formed a more extensive biofilm than the corresponding wild-type strain . While these results confirm that biofilms represent a unique growth state by comparison to planktonic cultures, they also suggest the existence of species-specific pathways that contribute to biofilm formation and maintenance of the sessile lifestyle . To date, no comprehensive transcriptional analysis of S . aureus biofilms has been reported . However, Prigent-Combaret et al . (51) demonstrated that biofilm-encased E . coli encounter high osmolarity, oxygen limitation, and higher cell density than cells grown under planktonic conditions, and all of these factors are known to influence gene expression in S . aureus (11, 45) .

To further investigate these issues, we generated sarA and ica mutations in a clinical isolate of S . aureus (UAMS-1) and examined their relative capacity to form a biofilm both in vitro and in vivo . We also used a custom-made Affymetrix GeneChip representing the combined genomes of six strains of S . aureus (N315, Mu50, COL, NCTC 8325, EMRSA-16 [strain 252], and MSSA-476) to investigate differential gene expression in a mature S . aureus biofilm .

 
Bacterial strains. The experiments described here focus on the S . aureus clinical isolate UAMS-1 . This strain was cultured from the bone of a patient suffering from osteomyelitis and was previously shown to form a biofilm both in vitro (5) and in vivo (21) . The UAMS-1 sarA mutant was generated by transduction as previously described (7). 11-mediated transduction from a derivative of S . aureus SA113 carrying an ica::tet mutation (14) was used to generate a UAMS-1 ica mutant . Transductants were confirmed by Southern blotting using probes corresponding to the sarA and ica loci (6, 7) .

Detection of PNAG production. To assess the production of PNAG in S . aureus clinical isolate UAMS-1 and its sarA and ica mutants, cultures were grown in tryptic soy broth with the appropriate antibiotic . After overnight incubation, the optical density at 560 nm (OD₅₆₀) was determined, and an equal number of cells (2 to 5 ml of each culture grown overnight) was harvested by centrifugation . Cells were resuspended in 50 µl of 0.5 M EDTA (pH 8.0) and boiled for 5 min . After cellular debris was removed by centrifugation, a 40-µl aliquot of the supernatant was incubated for 30 min with 10 µl of proteinase K (20 mg/ml) at 37°C to reduce nonspecific background levels . After the addition of 10 µl of Tris-buffered saline (20 mM Tris-HCl, 150 mM NaCl [pH 7.4]), 8 µl of each extract was spotted onto a nitrocellulose membrane using a BIO-dot microfiltration apparatus (Bio-Rad Laboratories, Inc., Hercules, Calif.) . After drying, the presence of PNAG in the extract was assessed using the WesternBreeze chemiluminescence immunodetection kit (Invitrogen Corp., Carlsbad, Calif.) and anti-PNAG antiserum (kindly provided by Kimberly Jefferson, Channing Laboratory, Harvard Medical School) .

Planktonic culture conditions. Because biofilm formation by strain UAMS-1 in vitro is dependent on supplementation of the medium with 0.5% glucose and 3.0% sodium chloride (5), these supplements were also added to the medium used for planktonic culture . Specifically, 5-ml samples of cultures grown overnight in tryptic soy broth at 37°C with constant aeration were used to inoculate 250 ml of fresh biofilm medium to an OD₅₆₀ of 0.05 . Cultures were incubated with aeration at 37°C . Aliquots were then removed at the mid-exponential (OD₅₆₀ = 1.0) and stationary (OD₅₆₀ = 3.5) growth phases . Aliquots were immediately mixed with an equal volume of ice-cold acetone-ethanol (1:1) and stored at –20°C prior to RNA extraction .

Biofilm cultures. Biofilms were generated in disposable flow cells (Stovall Life Science, Greensboro, N.C.) as previously described (5) . Briefly, flow cells were precoated overnight at 4°C with 20% human plasma diluted in carbonate buffer (pH 9.6) . The inlet side of the flow cell was then connected to a sterile reservoir filled with biofilm medium . The outlet side was connected to a waste reservoir . Tubing upstream of each individual cell was injected with 0.5 ml of the appropriate culture grown overnight . After the flow of medium was started and bacteria were allowed to enter the flow cell, the flow was stopped and the chamber was incubated inverted at 37°C for 1 h . After the flow cell was set upright, nonadherent bacteria were flushed by adjusting the flow rate to 0.5 ml/min, which is sufficient to replace the volume of the flow cell once every minute . Cells harvested after 1 week by aspiration from the downstream side of the flow cell were immediately mixed with acetone-ethanol as described above .

Assessment of biofilm formation. Biofilm formation in vitro was assessed using a microtiter plate assay and flow cells as previously described (5) . The murine model of catheter-associated biofilm formation described by Kadurugamuwa et al . (31) was used to assess biofilm formation in vivo . Briefly, 20- to 30-g female BALB/c mice (Charles River, Wilmington, Mass.) were anesthetized with ketamine (100 mg/kg of body weight) and xylazine (5 mg/kg), their flanks were shaved, and the skin was cleansed with Betadyne and alcohol . Under aseptic conditions, a 1-cm segment of 14-gauge Teflon intravenous catheter was implanted subcutaneously . The wound was closed with sutures and then cleansed with a Betadyne rinse . Infection was induced approximately 1 h after the implantation procedure by injecting 10⁵ CFU of the test strain into the lumen of the catheter . In some cases, mice were coinfected by injection of a mixture containing 10⁵ CFU of UAMS-1 and 10⁵ CFU of either the sarA or ica mutant . Mice were euthanized 10 days postinfection . The catheters were removed aseptically and washed with phosphate-buffered saline . Catheters were then placed in 10 ml of sterile phosphate-buffered saline and sonicated for 1 min to remove adherent bacteria . The number of bacteria in the sonicate was then determined by plating on tryptic soy agar . To confirm the identity of recovered bacteria and to determine the proportion of UAMS-1 versus the corresponding sarA and ica mutants in coinfection experiments, colonies obtained on tryptic soy agar were transferred to the appropriate selective medium and scored for growth .

RNA isolation and cDNA labeling. Aliquots of cells harvested from flow cells and stored as described above were pelleted by centrifugation at 7,500 x g for 10 min at 4°C . Each pellet was washed in an equal volume of TES buffer (150 mM NaCl, 78 mM disodium salt EDTA, 100 mM Tris [pH 7.5]) and resuspended to a concentration of 10⁹ CFU per ml in TES buffer containing 100 µg of lysostaphin (Ambicin L; AMBI, Inc., Lawrence, N.Y.) per ml . Samples were incubated at 37°C for 1 h prior to applying the equivalent of 10¹⁰ CFU to a Qiagen RNeasy Maxi column . Total bacterial RNA was isolated according to the manufacturer's directions (Qiagen, Inc., Valencia, Calif.) . After purification, contaminating DNA was removed with RNase-free DNase I (10 U/40 µg of total bacterial RNA at 37°C for 20 min) . RNA was then repurified using RNeasy Mini columns (Qiagen, Inc.) . The amount of recovered RNA was determined spectrophotometrically, and the absence of DNA was verified by PCR using primers (Table 1) corresponding to the collagen adhesin gene (cna) . Samples were then stored at –80°C .

 
RNA was converted to cDNA, and microarray analysis was performed according to the manufacturer's instructions (Affymetrix expression analysis technical manual, Affymetrix, Inc., Santa Clara, Calif.) for antisense prokaryotic arrays . Briefly, 10 µg of total RNA that had been mixed with random hexamer primers (Invitrogen) was denatured at 70°C for 10 min and allowed to anneal at 25°C for 10 min . cDNA was synthesized using Superscript II reverse transcriptase (Invitrogen) in 1x first-strand synthesis buffer, dithiothreitol, deoxynucleoside triphosphates, and SUPERase-In (Ambion, Inc., Austin, Tex.) . The mixture was incubated at 25°C for 10 min, 37°C for 60 min, and 42°C for 60 min . The reaction was stopped by incubating for 10 min at 70°C prior to degrading the RNA with 1 N NaOH for 30 min at 65°C and neutralizing with 1 N HCl . The cDNA was purified using a QIAquick PCR purification kit (Qiagen) and fragmented with DNase I in One-Phor-All buffer (Amersham Biosciences, Piscataway, N.J.) . DNase I was inactivated by heating the reaction mixture for 10 min at 98°C . The fragmented cDNA products were labeled with biotin on the 3' terminus using the Enzo BioArray terminal labeling kit with biotin ddUTP (Affymetrix, Inc.) .

DNA microarray hybridization and analysis. Labeled cDNA (1.5 µg) was hybridized to custom-made S . aureus GeneChips and detected according to the manufacturer's instructions for antisense prokaryotic arrays (Affymetrix, Inc.) . The GeneChip used in these experiments included 7,723 qualifiers representing the consensus open reading frame (ORF) sequences identified in the genomes of the S . aureus strains N315, Mu50, COL, NCTC 8325, EMRSA-16 (strain 252), and MSSA- 476, as well as novel GenBank entries and N315 intergenic regions greater than 50 bp, After hybridization and staining, the arrays were scanned using the Agilent GeneArray laser scanner (Agilent Technologies, Palo Alto, Calif.) . The data from duplicate experiments was normalized and analyzed using GeneSpring version 5.1 gene expression software (Silicon Genetics, Redwood City, Calif.) . Genes were considered to be induced in a biofilm if they were determined to be present by Affymetrix algorithms in the biofilm condition and they were transcribed at a level at least twofold higher than the corresponding planktonic growth condition . Genes were considered downregulated in a biofilm if they were determined to be present in either planktonic condition and had an expression level no more than half of that observed in the corresponding planktonic growth condition . Differential expression in biofilms was judged to be significant on the basis of statistical analysis, namely, the t test with a P value of 0.05 .

Real-time PCR. To confirm the results of our microarray data, the relative expression levels of the arcA, pyrR, ureA, and spa genes in each growth condition were also determined by real-time PCR . Briefly, DNase-treated RNA was reverse transcribed using the iScript cDNA synthesis kit as described by the manufacturer (Bio-Rad Laboratories) . A portion (1/20th) of each reaction mixture was then used for real-time PCR using an iCycler iQ real-time PCR detection system, gene-specific primers, and TaqMan probes corresponding to each ORF, and the iQ supermix (Bio-Rad) . The sequences of the primers and TaqMan probes are shown in Table 1 . Relative expression levels were determined by comparison to the level of gyrB expression in the same cDNA preparations .

 
Mutation of sarA, but not ica, results in a reduced capacity to form a biofilm in vitro. In a previous report from our laboratory, we demonstrated that mutation of sarA in clinical isolates of S . aureus results in a reduced capacity to form a biofilm (5) . Although our experiments did not address the mechanistic basis for this, Valle et al . (66) also observed that mutation of sarA results in a reduced capacity to form a biofilm and concluded that this was due to the impact of SarA on production of PIA (also known as PNAG) . To more definitively address the role of PNAG in biofilm formation by our clinical isolates, we generated an S . aureus UAMS-1 ica mutant and assessed its ability to form a biofilm in vitro . Mutation of the ica locus was confirmed by Southern blotting (Fig. 1A), and the inability of the ica mutant to produce PNAG was confirmed by immunoblotting using PNAG-specific antisera (Fig . 1B) . We found that mutation of ica, and the resulting inability to produce PNAG, had little impact on biofilm formation (Fig . 2) . In contrast, mutation of sarA resulted in a reduced capacity to form a biofilm . The relative capacities of the UAMS-1 sarA and ica mutants to form a biofilm were evident both in our microtiter plate assay (Fig . 2A) and in flow cells (Fig . 2B) . Under in vitro growth conditions, mutation of sarA resulted in reduced production of PNAG (Fig . 1B), which is consistent with the results of Valle et al . (66) . However, the results observed with our ica mutant make it difficult to conclude that this would explain the biofilm-deficient phenotype of the UAMS-1 sarA mutant .

 
Mutation of sarA, but not ica, results in a reduced capacity to form a biofilm in vivo. The results discussed above suggest that ica is not required for biofilm formation in at least some strains of S . aureus and that the reduced capacity of a UAMS-1 sarA mutant to form a biofilm in vitro is not a function of the impact of the sarA mutation on expression of the ica operon or production of PNAG . However, in S . aureus, it is well established that expression of the ica locus is tightly regulated and that it is preferentially expressed under in vivo conditions (42) . This leaves open the possibility that the results we observed in vitro do not reflect the situation observed in vivo .

To examine this issue directly, we assessed the relative abilities of S . aureus UAMS-1 and its sarA and ica mutants to form a biofilm in vivo using a murine model of catheter-associated biofilm formation (31) . The average numbers of bacteria obtained from explanted catheters at 10 days postinfection were 7.1 x 10⁷ CFU per catheter in mice infected with UAMS-1 and 5.9 x 10⁷ CFU per catheter in mice infected with the ica mutant (Fig . 3) . In contrast, we recovered only 2.3 x 10⁷ CFU per catheter from mice infected with the UAMS-1 sarA mutant . Although the sarA mutant was capable of colonizing the implanted catheter, the reduced recovery observed with the sarA mutant was statistically significant compared to the recovery for both UAMS-1 (P = 0.001) and its ica mutant (P = 0.022) .

 
To further investigate the relative capacities of S . aureus UAMS-1 and its ica and sarA mutants to form a biofilm in vivo, we also performed experiments in which catheters were coinfected with equivalent numbers of both UAMS-1 and its ica mutant or UAMS-1 and its sarA mutant (in both cases, total inoculum of 2 x 10⁵ CFU) . In mice coinfected with UAMS-1 and its ica mutant, we recovered 3.4 x 10⁷ CFU per catheter of the wild-type strain and 2.3 x 10⁷ CFU per catheter of the ica mutant (Fig . 4A) . These results confirm our in vitro experiments and demonstrate that UAMS-1 and its corresponding ica mutant have an equivalent capacity to form a biofilm not only in vitro but also in vivo . In contrast, when we examined mice coinfected with UAMS-1 and its sarA mutant, the number of wild-type cells recovered was similar to the number found in previous experiments (4.9 x 10⁷ CFU), but the number of the UAMS-1 sarA mutant we recovered had decreased to an average of only 8.0 x 10⁵ CFU per catheter (Fig . 4B) . These results also confirm the results of our in vitro experiments . Moreover, the reduced recovery of the sarA mutant in coinfection experiments relative to in vivo experiments in which the sarA mutant was introduced without competition from the wild-type strain also indicates that the wild-type strain has a competitive advantage that further limits the capacity of a sarA mutant to form a biofilm in vivo .

 
Transcriptional profiling in S . aureus UAMS-1 planktonic cultures. Taken together, the results discussed above confirm that a UAMS-1 sarA mutant has a reduced capacity to form a biofilm and that this is not a function of the impact of sarA on expression of the ica operon or production of PNAG . Moreover, the results observed with our in vitro models were consistent with the results observed in our in vivo model . Because sarA is a global regulator of gene expression in S . aureus, this suggests that other elements of the sarA regulon are also important in biofilm formation both in vitro and in vivo . Presumably, these elements could be identified by defining transcriptional changes observed within S . aureus biofilms and correlating these changes with experiments aimed at defining the sarA regulon . Because comprehensive transcriptional profiling of an S . aureus sarA mutant has been reported (19), we focused our efforts on defining the transcriptional changes that occur when UAMS-1 is grown within a biofilm .

Because biofilm formation by S . aureus UAMS-1 requires supplementation of the medium with both glucose and salt (5), we also added these supplements to the medium used for planktonic culture . We first wanted to investigate whether the current paradigm (e.g., preferential expression of surface proteins during the exponential phase, followed by a shift to exoprotein production in the postexponential phase) was altered by supplementation of the medium . As expected, we found that expression of genes encoding protein A, clumping factor B, collagen adhesion, coagulase, and fibronectin-binding protein (spa, clfB, cna, coa, and fnb, respectively) were upregulated in the exponential versus stationary phase (6, 44, 45) (Table 2) . In contrast, expression of the gene (clfA) encoding a second fibrinogen-binding protein (ClfA) was upregulated in the postexponential phase . This is consistent with previous reports examining the temporal expression of clfA (72, 73) . Expression of the secreted proteins was also elevated in stationary phase as expected . Specific examples include the genes encoding several cysteine proteases, the Clp proteinase, alpha-toxin, and the genes within the accessory gene regulator (agr) operon (sspA, sspB, sspC, clpC, hla, RNAII, and RNAIII, respectively) . Taken together, these results confirm that our growth conditions accurately reflect the transition between the exponential and stationary growth phases and that this transition is not dramatically altered by supplementation of the medium in a manner that promotes biofilm formation .

 
Transcriptional profiling of S . aureus UAMS-1 biofilms. We identified a total of 580 genes that were expressed in an altered fashion when UAMS-1 was harvested from a mature biofilm . The greatest distinction, at least in terms of overall numbers of differentially expressed genes, was between the biofilm mode of growth and the exponential growth phase of planktonic culture (Table 3); however, a significant number of genes were also differentially expressed in comparison to stationary-phase cultures (Table 4) . These findings clearly imply that S . aureus biofilms represent a unique growth condition by comparison to both exponential- and stationary-phase planktonic cultures . Indeed, we identified 48 genes whose expression was enhanced at least twofold in a biofilm in comparison to both exponential- and stationary-phase planktonic cultures (Table 5) . These 48 genes included 30 genes located in six independent clusters as determined by their N315 ORF numbers . Included among these linked genes were the arginine deiminase cluster (arc; N315-SA2424-SA2428), a potassium-specific transport system (kdp; N315-SA1879-SA1881), the pyrimidine biosynthesis operon (pyr; N315-SA1041-SA1049), and the urease operon (ure; N315-SA2081-2088) . Interestingly, only one gene in the ica locus (icaD) was found to be significantly upregulated in a biofilm, and this was limited to the comparison between biofilms and the stationary phase of planktonic growth (Table 4) . This is consistent with a recent report concluding that ica expression is associated with the initial colonization of S . epidermidis in a foreign body infection model but not with its persistence (67) .

 
We also identified 84 genes whose expression was reduced by a least a factor of at least 2 by comparison with both planktonic growth conditions (Table 5) . Included were 25 genes in eight possible operons including an oligopeptide transport system (opp; N315-SA0845-SA0848) and the genes responsible for purine biosynthesis (pur; N315-SA0920-SA0925) . Most genes in the other six putative operons encode hypothetical or conserved hypothetical proteins with no known function . However, one well-defined gene that was drastically downregulated in biofilms (60 to 139 times higher in the exponential-phase cultures and 12 to 27 times higher in the stationary-phase cultures) was spa, the gene that encodes protein A .

Confirmation of transcriptional profiling by real-time PCR. To verify the results of our microarray experiments, we used real-time PCR to examine the relative expression levels of selective target genes . These comparisons were done using RNA isolated from two independent cultures representing each of three growth conditions (biofilm and exponential and stationary growth phases) . As observed in our profiling experiments, the arcA, pyrR, and ureA transcripts were present in greater quantities in the biofilm samples than in both exponential- and stationary-phase planktonic cultures (Fig . 5) . Indeed, while the patterns of gene expression observed with real-time PCR were consistent with our profiling experiments, the results from the real-time PCR experiments suggest that our profiling experiments may underestimate the actual differences . As with our profiling experiments, we also found that spa was significantly downregulated in biofilms on the basis of real-time PCR comparisons (Fig . 5) . Collectively, the real-time PCR results provide independent verification of our DNA microarray results .

 
Roles of genes regulated by sarA in biofilm formation. For the reasons discussed above, we are particularly interested in genes that are differentially expressed in biofilms and are part of the sarA regulon . Therefore, we compiled a list of genes that were reported by Dunman et al . (19) to be regulated by sarA and were either induced or repressed in a biofilm compared to either planktonic condition . This analysis revealed 27 genes that were part of the sarA regulon and were differentially expressed in biofilms (Table 6) . Because these genes may be genes that are required for biofilm formation, genes that are induced in biofilms and positively regulated by sarA would be of particular interest; however, we identified only four genes (sdhB, carA, an unidentified ORF with similarity to a major histocompatibility complex [MHC] class II analog, and a hypothetical protein) that fell into this category . At the same time, it may be equally important that specific genes be turned off to facilitate biofilm formation, and we identified eight genes (arc, phoP, pbp3, nuc, ndhG, spa, and two hypothetical proteins) that were repressed in biofilms and negatively regulated by sarA . The remaining genes were divergently regulated by sarA and in biofilms; however, the possibility that the impact of sarA is indirect in these cases cannot be ruled out .

 
Valle et al . (66) was the first to demonstrate that mutation of sarA results in a reduced capacity to form a biofilm . They concluded that this was due, at least in part, to reduced expression of the icaADBC operon . At the same time, Valle et al . (66) also suggested that SarA enhances biofilm formation by suppressing production of a second, unidentified protein that was either a repressor of PNAG synthesis or was involved in the turnover of PNAG . On this basis, they proposed a model in which the impact of sarA on biofilm formation was dependent on two pathways, both of which functioned by moderating the production of PNAG .

In this report, we also demonstrated that mutation of sarA results in reduced production of PNAG . To further investigate the impact of this on biofilm formation in our clinical isolate, we generated an ica mutant and examined its capacity to form a biofilm both in vitro and in vivo . Surprisingly, our ica mutant formed a biofilm comparable to that of the parental strain under both static and flow conditions . As in our previous experiments (5), the capacity of our UAMS-1 sarA mutant to form a biofilm was reduced in comparison to both the wild-type strain and its corresponding ica mutant . This clearly indicates that the impact of sarA on biofilm formation, at least as defined under in vitro growth conditions, involves a pathway that is independent of the icaADBC operon .

The role of ica in vivo has been addressed in S . epidermidis with contradictory results (10, 23, 41, 47, 49, 55, 56, 57, 62) . However, few studies have addressed this issue in S . aureus . Cramton et al . (14) demonstrated that ica is present in S . aureus and that it is required for PIA production and biofilm formation in that species . However, the mutagenesis experiments were limited to a single strain of S . aureus (SA113) that was derived from NCTC 8325 by chemical mutagenesis, so these results may not be representative of the situation observed in clinical isolates . More recently, Vandecasteele et al . (67) analyzed expression of biofilm-associated genes, including icaA and icaC, both in vitro and in vivo . While expression of both genes was induced upon initial exposure to foreign bodies, this induction peaked shortly after the introduction of bacteria and was followed by a slow decrease over time . These results suggest that the ica operon is mainly associated with the initial colonization phase of biofilm formation, rather than maturation and persistence . This is consistent with our results and the fact that our analysis was limited to mature biofilms . Additionally, Francois et al . (23) compared an S . aureus strain and its corresponding ica mutant using a tissue cage model and found that the ica mutant retained the capacity to colonize at a level comparable to that of the wild-type strain . They concluded on this basis that biofilms were not an important factor in their model . However, we believe there are two possible alternative explanations . The first is that their model does not accurately reflect the need to form a biofilm . The second is that their ica mutant retained the capacity to form a biofilm under in vivo conditions . Our results would support the latter hypothesis . Specifically, we found that while the UAMS-1 sarA mutant demonstrated a reduced capacity to colonize catheters in vivo, the UAMS-1 ica mutant colonized the same substrates as well as the wild-type strain did . These results also suggest that PNAG is not required for in vivo colonization by S . aureus and that sarA regulates genes required for biofilm formation independent of its ability to modulate ica expression and/or the production of PNAG . We would note, however, that our results are also limited to a single strain, and it is certainly possible that UAMS-1 has an alternative means of promoting intercellular accumulation that attenuates the need for PIA . Whether this is true of other clinical isolates of S . aureus remains to be determined, but previous work in our laboratory has confirmed that all other sarA-mediated phenotypes are conserved among clinical isolates like UAMS-1 (5, 7) .

Dunman et al . (19) recently reported the results of microarray-based transcriptional profiling experiments with S . aureus sarA and agr mutants . These studies confirmed that sarA has global regulatory effects that are mediated through both agr-dependent and agr-independent pathways . In an effort to determine which of these might be involved in biofilm formation and/or maintenance of the sessile lifestyle, we performed comprehensive transcriptional profiling with RNA isolated from mature S . aureus biofilms . Although it would be preferable to do these experiments using RNA derived from biofilms grown in vivo, our attempts to isolate a sufficient quantity of high-quality RNA from in vivo samples have thus far been unsuccessful . However, our studies indicating that mutation of sarA results in a reduced capacity to form a biofilm both in vitro and in vivo suggest that biofilms grown in flow cells may also provide a relevant source of RNA for transcriptional profiling . On this basis, we isolated RNA from UAMS-1 grown in flow cells and compared the pattern of gene expression to RNA from the same strain grown in planktonic culture . The results of our comparisons to both exponential- and stationary-phase planktonic cultures clearly indicate that our flow cell biofilm model represents a unique growth environment . Indeed, we identified a total of 580 genes that are differentially expressed in biofilms by comparison to either or both planktonic conditions .

Several of the operons that were induced in biofilms have been found to be important in acid tolerance in other bacterial species . Indeed, maintenance of pH homeostasis within the bacterial cell and buffering of the surrounding microenvironment have been associated with biofilm formation in the oral bacteria Streptococcus salivarius (34) . One way bacteria combat acidic environments is to produce alkaline compounds, such as ammonia, that can neutralize the acids . Two ways in which bacteria generate ammonia are through the urease and arginine deiminase (ADI) pathways . Interestingly, we found that multiple genes from both of these pathways were induced in S . aureus biofilms by comparison to both planktonic conditions .

Under anaerobic conditions, some bacteria are also able to generate ATP as an energy source through catabolism of arginine via the ADI pathway, which is widely distributed in bacteria (16), archaea (54), and eucarya (60) . The ADI pathway is comprised of three enzymatic reactions, catalyzed by arginine deiminase (arcA), ornithine transcarbamoylase (arcB), and carbamate kinase (arcC) . Collectively, these enzymes convert arginine to ornithine, ammonia, and CO₂, yielding 1 mol of ATP per mol of arginine consumed . ArcD is an arginine-ornithine transporter that catalyzes the uptake of arginine and concomitant export of ornithine, while arcR encodes an activator of the ADI operon that is a member of the Crp-Fnr family of regulators . We found that all five of these genes were significantly induced in S . aureus biofilms (Table 5) . One additional gene that was upregulated 3.9-fold in a biofilm compared to exponential-phase cells is the arginine repressor encoded by argR (Table 3) . Under anaerobic conditions in the presence of arginine, ArgR represses anabolic ornithine carbamoyltransferase and induces the ADI pathway . In addition to its role in generating ATP anaerobically, the ADI pathway is one of two major ammonia-generating pathways utilized by oral bacteria to maintain pH homeostasis when growing in a biofilm . Ammonia generated by the deimination of arginine can neutralize acids generated by bacterial glycolysis .

The ADI system and its role in acid resistance have also been correlated with virulence in Streptococcus pyogenes . ADI in this species was originally called streptococcal acid glycoprotein (SAGP) and was characterized as being an inhibitor of stimulated human peripheral blood mononuclear cell proliferation (17, 18) . In addition, it is thought that the acid sensitivity of a SAGP-negative mutant is responsible for its reduced ability to enter and survive in epithelial cells (18) .

Also included among the genes induced in biofilms (Table 5) were seven genes that comprise the urease operon (ureABCEFGD) . Urease (urea amidohydrolase) is a nickel-containing enzyme that catalyzes the hydrolysis of urea to yield two molecules of ammonia and one molecule of CO₂ . Ureases of most bacteria are composed of three distinct subunits encoded by three contiguous genes, ureA, ureB, and ureC . Urease gene clusters also encode accessory genes, in addition to these structural genes, that are required for the de novo synthesis of active urease . Urease activity is essential for the colonization of the gastric mucosa by Helicobacter pylori and colonization of the urinary tract by both Proteus mirabilis and Staphylococcus saprophyticus (20, 25, 30) . In addition, urease is thought to play a central role in the pathogenesis of Ureaplasma urealyticum urinary and respiratory tract infections (27, 35) .

Recently, Saïd-Salim et al . (58) found that genes of the urease operon in S . aureus are negatively regulated by the SarA homolog Rot (repressor of toxin) . Interestingly, expression of rot was repressed in biofilms, although this was limited to the comparison with exponential-phase planktonic cultures . The trigger for induced transcription of the urease operon in S . aureus has not yet been studied . However, urease synthesis by Klebsiella aerogenes is stimulated under conditions of nitrogen starvation, such as when the bacteria are cultured in minimal medium containing a poor nitrogen source, such as proline, arginine, or histidine (24) . In H . pylori, expression of the urease operon is upregulated by a pH-dependent, posttranscriptional regulatory mechanism . More specifically, Akada et al . (1) showed that a shift to an acidic pH resulted in a significant increase in the level of ure operon mRNA, even in the presence of inhibitors of transcription . Until recently, it was believed that urease was associated with the cell surface and that it directly neutralized the microenvironment surrounding the cell (50) . However, it is now thought to play a more important role as an intracellular enzyme required for acid resistance (61) .

In addition to the production of ammonia, cation transport ATPases, such as the high-affinity K⁺-specific transport system encoded by the kdp operon, can also contribute to pH homeostasis through the exchange of K⁺ for H⁺ (13) . In this study, we found that three genes of the kdp operon (kdpABC) were induced in biofilms by comparison to both planktonic growth conditions (Table 5) and two other genes (kdpDE) were induced by comparison to the exponential growth phase (Table 3) . In E . coli, the Kdp system is composed of the ion motive P-type ATPase encoded by the kdpFABC operon, and expression of the kdpFABC operon is regulated by an adjacent operon, kdpDE (2) . KdpD is a membrane-spanning sensor kinase, and KdpE is a cytosolic transcriptional activator . In response to an appropriate signal(s) (membrane stretch, alteration in turgor pressure, and external and internal potassium levels), KdpD transphosphorylates KdpE, which in turn upregulates transcription of the kdpFABC operon (2, 43) . Interestingly, van der Laan et al . (68) recently reported that NH₄⁺ ions strongly stimulate the ATPase activity of the KdpFABC complex in E . coli .

Several operons of the pyrimidine nucleotide biosynthetic (pyr) pathway (pyrRPBC, carAB, and pyrFE) were also induced in biofilms (Table 5) . The pathway for the de novo synthesis of pyrimidines consists of six enzymatic steps leading to the formation of UMP . The first step in the pyrimidine biosynthetic pathway is the formation of carbomyl-phosphate (CP) from bicarbonate, glutamine (or ammonia), and ATP by CP synthase, which is encoded by the carAB genes . Interestingly, CP is also required for the biosynthesis of arginine . In B . subtilis, PyrR regulates transcription of the pyr operon by binding in a uridine nucleotide-dependent fashion to pyr mRNA and altering the secondary structure of the downstream mRNA (37, 38) . Binding of PyrR to the downstream mRNA stabilizes a binding loop and prevents the formation of the antiterminator (39) . In the absence of PyrR or when levels of the nucleotides UMP and UTP are low, the antiterminator is a stable secondary structure, and transcription of the downstream genes continues (39) . While we do not know if the function of B . subtilis and S . aureus PyrR is conserved, our results showing that all of the genes in the pyr operon were induced in biofilms suggests that the level of UMP in cells growing in a biofilm is severely limited . In addition, upregulation of the pyr operon and subsequent CP production may be required for synthesis of sufficient levels of arginine to be used by the ADI pathway during anaerobic growth .

Taken together, the results of our array experiments suggest that mature biofilms are growing anaerobically and that genes of the acid tolerance response are upregulated in response to an acidic environment . While the regulatory pathway for the acid tolerance response in S . aureus has not been well characterized, there have been studies suggesting that the global regulators SigB and SarA are involved . Specifically, mutation of sigB results in strains with impaired abilities to respond to acid and to induce a stationary-phase acid tolerance response (8, 32) . In addition, in the absence of RsbU, expression from the SigB-dependent sarA promoter was significantly reduced at pH 5.5 (46) .

Our analysis revealed 27 genes that were differentially expressed in biofilms and were part of the sarA regulon as defined by Dunman et al . (19) (Table 6) . Given the different mechanistic possibilities of how SarA modulates biofilm production, it is difficult to determine which of these genes might be the most relevant candidates for further consideration . For example, because SarA mutants have a reduced capacity to form a biofilm, SarA would presumably be required for production of an activator of biofilm formation or a required effector molecule, in which case transcription of the relevant gene(s) would be upregulated by SarA and in a biofilm . However, it is also possible that SarA represses production of an effector that is deleterious to biofilm formation, in which case the SarA target would be downregulated within a biofilm . One possibility in that regard is that the increased production of proteases observed in sarA mutants (7) results in degradation of a required surface protein . However, Valle et al . (66) concluded on the basis of both mutagenesis experiments and experiments employing specific inhibitors that protease production was not responsible for the decreased capacity of a sarA mutant to form a biofilm . Nevertheless, it remains possible that a similar scenario in which SarA represses transcription of some factor that has a negative impact on biofilm formation is involved . Moreover, both of these scenarios are based on the assumption that the SarA-dependent regulation of the relevant effector(s) is direct, and it is also possible that sarA modulates biofilm formation in an indirect fashion, perhaps via one of the increasing number of SarA homologs (4, 9, 45) . It is also possible that sarA is required for production of an adhesin or some other effector protein that is required only for initial stages of biofilm formation . In this case, expression of the relevant gene may be transient in a fashion that would not have been apparent in our profiling experiments focusing on mature biofilms .

Finally, it should be noted that biofilms are not homogenous populations of cells and because the experiments we performed did not address this issue, it is certainly possible that we failed to detect genes that are within the sarA regulon but are differentially expressed only within certain regions of the biofilm . At the same time, this would imply that the genes we did identify are either differentially expressed throughout the biofilm or that our results actually underestimate the degree of differential gene expression observed within more limited regions . It should also be noted that the profiling experiments of Dunman et al . (19) were done using a first-generation GeneChip that was limited to only 85% of the genes identified in a single strain (COL) of S . aureus, and it is certainly possible that this limited the identification of relevant genes that are conserved among clinical isolates like UAMS-1 . Moreover, the profiling experiments of Dunman et al . (19) were done using RNA isolated from derivatives of the S . aureus strain RN6390, and we have previously demonstrated that the regulatory circuits observed in RN6390 and its corresponding sarA and agr mutants are different from those of other strains, including UAMS-1 (7) . To address these issues, we are currently performing transcriptional profiling experiments using RNA isolated from UAMS-1 sarA and agr mutants and the more comprehensive chips used for the biofilm profiling experiments described here .

 
* Corresponding author . Mailing address: Department of Microbiology and Immunology, Mail Slot 511, University of Arkansas for Medical Sciences, 4301 W . Markham, Little Rock, AR 72205 . Phone: (501) 686-7958 . Fax: (501) 686-5359 . E-mail: smeltzermarks@uams.edu .

1. Akada, J . K., M . Shirai, H . Takeuchi, M . Tsuda, and T . Nakazawa. 2000 . Identification of the urease operon in Helicobacter pylori and its control by mRNA decay in response to pH . Mol . Microbiol . 36:1071-1084.
2. Altendorf, K., P . Voelkner, and W . Puppe. 1994 . The sensor kinase KdpD and the response regulator KdpE control expression of the kdpFABC operon in Escherichia coli . Res . Microbiol . 145:374-381.
3. Arciola, C . R., L . Baldassarri, and L . Montanaro. 2001 . Presence of icaA and icaD genes and slime production in a collection of staphylococcal strains from catheter-associated infections . J . Clin . Microbiol . 39:2151-2156 .
4. Arvidson, S., and K . Tegmark. 2001 . Regulation of virulence determinants in Staphylococcus aureus . Int . J . Med . Microbiol . 291:159-170.
5. Beenken, K . E., J . S . Blevins, and M . S . Smeltzer. 2003 . Mutation of sarA in Staphylococcus aureus limits biofilm formation . Infect . Immun . 71:4206-4211 .
6. Blevins, J . S., A . F . Gillaspy, T . M . Rechtin, B . K . Hurlburt, and M . S . Smeltzer. 1999 . The staphylococcal accessory regulator (sar) represses transcription of the Staphylococcus aureus collagen adhesin gene (cna) in an agr-independent manner . Mol . Microbiol . 33:317-326.
7. Blevins, J . S., K . E . Beenken, M . O . Elasri, B . K . Hurlburt, and M . S . Smeltzer. 2002 . Strain-dependent differences in the regulatory roles of sarA and agr in Staphylococcus aureus . Infect . Immun . 70:470-480 .
8. Chan, P . F., S . J . Foster, E . Ingram, and M . O . Clements. 1998 . Staphylococcus aureus alternative sigma factor ^B controls the environmental stress response but not starvation survival or pathogenicity in a mouse abscess model . J . Bacteriol . 180:6082-6089 .
9. Cheung, A . L., and G . Zhang. 2002 . Global regulation of virulence determinants in Staphylococcus aureus by the SarA protein family . Front . Biosci . 1:1825-1842.
10. Christensen, G . D., L . Baldassarri, and W . A . Simpson. 1994 . Colonization of medical devices by coagulase-negative staphylococci, p . 45-78 . In A . L . Bisno and F . A . Waldvogel (ed.)., Infections associated with indwelling medical devices, 2nd ed . ASM Press, Washington, D.C.
11. Clements, M . O., and S . J . Foster. 1999 . Stress resistance in Staphylococcus aureus . Trends Microbiol . 7:458-462.
12. Conlon, K . M., H . Humphreys, and J . P . O'Gara. 2002 . icaR encodes a transcriptional repressor involved in environmental regulation of ica operon expression and biofilm formation in Staphylococcus epidermidis . J . Bacteriol . 184:4400-4408 .
13. Cotter, P . D., and H . Colin. 2003 . Surviving the acid test: responses of gram-positive bacteria to low pH . Microbiol . Mol . Biol . Rev . 67:429-453 .
14. Cramton, S . E., C . Gerke, N . F . Schnell, W . W . Nichols, and F . Götz. 1999 . The intercellular adhesion (ica) locus is present in Staphylococcus aureus and is required for biofilm formation . Infect . Immun . 67:5427-5433 .
15. Cramton, S . E., M . Ulrich, F . Götz, and G . Döring. 2001 . Anaerobic conditions induce expression of polysaccharide intercellular adhesin in Staphylococcus aureus and Staphylococcus epidermidis . Infect . Immun . 69:4079-4085 .
16. Cunin, R., N . Glansdorff, A . Pierard, and V . Stalon. 1986 . Biosynthesis and metabolism of arginine in bacteria . Microbiol . Rev . 50:314-352.
17. Degnan, B . A., M . A . Kehoe, and J . A . Goodacre. 1997 . Analysis of human T cell responses to group A streptococci using fractionated Streptococcus pyogenes proteins . FEMS Immunol . Med . Microbiol . 17:161-170.
18. Degnan, B . A., M . C . Fontaine, A . H . Doebereiner, J . J . Lee, P . Mastroeni, G . Dougan, J . A . Goodacre, and M . A . Kehoe. 2000 . Characterization of an isogenic mutant of Streptococcus pyogenes Manfredo lacking the ability to make streptococcal acid glycoprotein . Infect . Immun . 68:2441-2448 .
19. Dunman, P . M., E . Murphy, S . Haney, D . Palacios, G . Tucker-Kellogg, S . Wu, E . L . Brown, R . J . Zagursky, D . Shlaes, and S . J . Projan. 2001 . Transcription profiling-based identification of Staphylococcus aureus genes regulated by the agr and/or sarA loci . J . Bacteriol . 183:7341-7353 .
20. Eaton, K . A., J . V . Gilbert, E . A . Joyce, A . E . Wanken, T . Thevenot, P . Baker, A . Plaut, and A . Wright. 2002 . In vivo complementation of ureB restores the ability of Helicobacter pylori to colonize . Infect . Immun . 70:771-778 .
21. Evans, R . P., C . L . Nelson, W . R . Bowen, M . G . Kleve, and S . G . Hickmon. 1998 . Visualization of bacterial glycocalyx with a scanning electron microscope . Clin . Orthop . Relat . Res . 347:243-249.
22. Fowler, V . G., Jr., P . D . Fey, L . B . Reller, A . L . Chamis, G . R . Corey, and M . E . Rupp. 2001 . The intercellular adhesin locus ica is present in clinical isolates of Staphylococcus aureus from bacteremic patients with infected and uninfected prosthetic joints . Med . Microbiol . Immunol . 189:127-131.
23. Francois, P., P . H . Tu Quoc, C . Bisognano, W . L . Kelly, D . P . Lew, J . Schrenzel, S . E . Cramton, F . Götz, and P . Vaudaux. 2003 . Lack of biofilm contribution to bacterial colonisation in an experimental model of foreign body infection by Staphylococcus aureus and Staphylococcus epidermidis . FEMS Immunol . Med . Microbiol . 35:135-140.
24. Friedrich, B., and B . Magasanik. 1977 . Urease of Klebsiella aerogenes: control of its synthesis by glutamine synthetase . J . Bacteriol . 131:446-452.
25. Gatermann, S., and R . Marre. 1989 . Cloning and expression of Staphylococcus saprophyticus urease gene sequences in Staphylococcus carnosus and contribution of the enzyme to virulence . Infect . Immun . 57:2998-3002.
26. Götz, F. 2002 . Staphylococcus and biofilms . Mol . Microbiol . 43:1367-1378.
27. Hedelin, H., J . E . Brorson, L . Grenabo, and S . Pettersson. 1984 . Ureplasma urealyticum and upper urinary tract stones . Br . J . Urol . 56:244-249.
28. Heilmann, C., O . Schweitzer, C . Gerke, N . Vanittanakom, D . Mack, and F . Götz. 1996 . Molecular basis of intercellular adhesion in the biofilm-forming Staphylococcus epidermidis . Mol . Microbiol . 20:1083-1091.
29. Heinrichs, J . H., M . G . Bayer, and A . L . Cheung. 1996 . Characterization of the sar locus and its interaction with agr in Staphylococcus aureus . J . Bacteriol . 178:418-423.
30. Jones, B . D., C . V . Lockatell, D . E . Johnson, J . W . Warren, and H . L . Mobley. 1990 . Construction of urease-negative mutant of Proteus mirabilis: analysis of virulence in a mouse model of ascending urinary tract infection . Infect . Immun . 58:1120-1123.
31. Kadurugamuwa, J . L., L . Sin, E . Albert, J . Yu, K . Francis, M . DeBoer, M . Rubin, C . Bellinger-Kawahara, T . R . Parr, Jr., and P . R . Contag. 2003 . Direct continuous method for monitoring biofilm infection in a mouse model . Infect . Immun . 71:882-890 .
32. Kullik, I., P . Giachino, and T . Fuchs. 1998 . Deletion of the alternative sigma factor ^B in Staphylococcus aureus reveals its function as a global regulator of virulence genes . J . Bacteriol . 180:4814-4820 .
33. Lewis, K. 2001 . Riddle of biofilm resistance . Antimicrob . Agents Chemother . 45:999-1007.
34. Li, Y . H., Y . Y . Chen, and R . A . Burne. 2000 . Regulation of urease gene expression by Streptococcus salivarius growing in biofilms . Environ . Microbiol . 2:169-177.
35. Ligon, J . V., and G . E . Kenny. 1991 . Virulence of ureaplasmal urease for mice . Infect . Immun . 59:1170-1171.
36. Lowen, P., B . Hu, J . Strutinsky, and R . Sparling. 1998 . Regulation in the rpoS regulon of Escherichia coli . Can . J . Microbiol. 44:707-717.
37. Lu, Y., and R . L . Switzer. 1996 . Transcriptional attenuation of the Bacillus subtilis pyr operon by the PyrR regulatory protein and uridine nucleotides in vitro . J . Bacteriol . 178:7206-7211.
38. Lu, Y., R . J . Turner, and R . L . Switzer. 1995 . Roles of the three transcriptional attenuators of the Bacillus subtilis pyrimidine biosynthetic operon in the regulation of its expression . J . Bacteriol . 177:1315-1325.
39. Lu, Y., R . J . Turner, and R . L . Switzer. 1996 . Function of RNA secondary structures in transcriptional attenuation of the Bacillus subtilis pyr operon . Proc . Natl . Acad . Sci . USA 93:14462-14467 .
40. Maira-Litran, T., A . Kropec, C . Abeygunawardana, J . Joyce, G . Mark III, D . A . Goldmann, and G . B . Pier. 2002 . Immunochemical properties of the staphylococcal poly-N-acetylglucosamine surface polysaccharide . Infect . Immun . 70:4433-4440 .
41. McDermid, K . P., D . W . Morck, M . E . Olson, N . D . Boyd, A . E . Khoury, M . K . Dasgupta, and J . W . Costerton. 1993 . A porcine model of Staphylococcus epidermidis catheter-associated infection . J . Infect . Dis. 168:897-903.
42. McKenney, D., K . L . Pouliot, Y . Wang, V . Murthy, M . Ulrich, G . Doring, J . C . Lee, D . A . Goldmann, and G . B . Pier. 1999 . Broadly protective vaccine for Staphylococcus aureus based on an in vivo-expressed antigen . Science 284:1523-1527 .
43. Nakashima, K., A . Sugiura, K . Kanamaru, and T . Mizuno. 1993 . Signal transduction between the two regulatory components involved in the regulation of the kdpABC operon in Escherichia coli: phosphorylation-dependent functioning of the positive regulator, KdpE . Mol . Microbiol . 7:109-116.
44. Ni Edhin, D., S . Perkins, P . Francois, P . Vaudaux, M . Hook, and T . J . Foster. 1998 . Clumping factor B (ClfB), a new surface-located fibrinogen-binding adhesin of Staphylococcus aureus . Mol . Microbiol . 30:245-257.
45. Novick, R . P. 2003 . Autoinduction and signal transduction in the regulation of staphylococcal virulence . Mol . Microbiol . 48:1429-1449.
46. Palma, M., and A . L . Cheung. ^B activity in Staphylococcus aureus is controlled by RsbU and an additional factor(s) during bacterial growth . Infect . Immun . 69:7858-7865.
47. Patrick, C . C., S . V . Hetherington, P . K . Roberson, S . Henwick, and M . M . Sloas. 1995 . Comparative virulence of Staphylococcus epidermidis isolates in a murine catheter model . Ped . Res . 37:70-74.
48. Patti, J . M., B . L . Allen, M . J . McGavin, and M . Hook. 1994 . MSCRAMM-mediated adherence of microorganisms to host tissues . Annu . Rev . Microbiol . 48:585-617.
49. Perdreau-Remington, F., M . A . Sande, G . Peters, and H . F . Chambers. 1998 . The abilities of Staphylococcus epidermidis wild-type strain and its slime-negative mutant to induce endocarditis in rabbits are comparable . Infect . Immun . 66:2778-2781 .
50. Phadnis, S . H., M . H . Parlow, M . Levy, D . Ilver, C . M . Caulkins, J . B . Connors, and B . E . Dunn. 1996 . Surface localization of Helicobacter pylori urease and heat shock protein homolog requires bacterial autolysis . Infect . Immun . 64:905-912.
51. Prigent-Combaret, C., O . Vidal, C . Dorel, and P . Lejeune. 1999 . Abiotic surface sensing and biofilm-dependent regulation of gene expression in Escherichia coli . J . Bacteriol . 181:5993-6002 .
52. Rachid, S., K . Ohlsen, U . Wallner, J . Hacker, M . Hecker, and W . Ziebuhr. 2000 . Alternative transcription factor ^B is involved in regulation of biofilm expression in a Staphylococcus aureus mucosal isolate . J . Bacteriol . 182:6824-6826 .
53. Rohde, H., J . K . Knobloch, M . A . Horstkotte, and D . Mack. 2001 . Correlation of biofilm expression types of Staphylococcus epidermidis with polysaccharide intercellular adhesin synthesis: evidence for involvement of icaADBC genotype-independent factors . Med . Microbiol . Immun .