INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Staphylococcus aureus is one of the most important pathogens in humans and animals. The pathogenesis of a particular S. aureus strain is attributed to the combined effect of extracellular factors and toxins, together with the invasive properties of the strain such as adherence, biofilm formation, and resistance to phagocytosis. Despite general agreement that biofilms are the basis for persistent or chronic bacterial infections (8), the understanding of the molecular mechanisms implicated in the biofilm formation process is still growing (39). Two steps appear to be involved in this process: (i) attachment of the bacterial cells to a surface (early adherence) and (ii) growth-dependent accumulation of bacteria in multilayered cell clusters (intercellular adhesion) (18). Different proteins (24, 37), including those of the family of microbial surface components that recognize adhesive matrix molecules (MSCRAMMs) (13), are involved in S. aureus adhesion. Specifically, MSCRAMMs may mediate S. aureus attachment to different cell types (11, 44) and abiotic surfaces once the adhesive host plasma constituents have covered the target surface (10, 29). However, the possible role of known MSCRAMMs or other molecules on the binding of S. aureus to inert surfaces in the absence of host constituents has not been thoroughly studied so far.

The icaABCD cluster, an operon present in Staphylococcus epidermidis and S. aureus (9, 20), participates in the intercellular adhesion step of biofilm formation by encoding proteins involved in the synthesis of the biofilm matrix polysaccharide poly-N-succinyl -1-6 glucosamine (PIA-PNSG). The implication of ica in staphylococcal pathogenesis has been demonstrated in various animal models (30, 40). A relationship has been shown in vivo between the expression of ica in clinical S. epidermidis strains and infection (15). In addition, immunizations with PIA-PNSG efficiently protect against S. aureus infection (30).

Considerable effort has been made by different groups to associate S. aureus biofilm formation with different mechanisms of virulence and pathogenesis. In this context and using highly adherent strains, we have observed that S. aureus in vitro adherence and biofilm formation on inert or mammalian cell surfaces is associated with (i) exopolysaccharide production (6, 23); (ii) rough colony morphology phenotype in Congo red agar (CRA) (6); (iii) higher resistance to phagocytosis (32); (iv) lower susceptibility to antibiotics when forming biofilms (2); (v) higher capacity to attach to different surfaces and biomaterials used in orthopedic surgery, causing osteomyelitis (16); and (vi) higher capacity to colonize the ovine mammary gland, causing mastitis (6). In addition, active immunizations with exopolysaccharides extracted from a highly adherent S. aureus isolate have been shown to trigger protective immunity against mastitis (3). However, the genetic mechanisms underlying these observations have not been determined.

Transposon mutagenesis has been used to determine the genetics of biofilm formation in different bacterial species, including S. epidermidis (18, 33), Streptococcus gordonii (26), Escherichia coli (38), Pseudomonas fluorescens (36), and Pseudomonas aeruginosa (35), using polystyrene microtiter plates as substrate. With this methodological approach, we were able to identify in this study a new protein which is involved in S. aureus attachment to abiotic surfaces and biofilm formation in vitro and enhances infection in vivo.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, culture conditions, and plasmids. In order to find a strong biofilm forming strain, the ability of 350 S. aureus mastitis isolates to attach and form a biofilm on polystyrene microtiter plates was tested as previously described (9). Accordingly, a bovine subclinical mastitis isolate, S. aureus V329, was selected because of its strong biofilm production phenotype and antibiotic susceptibility profile, which facilitates genetic manipulations. This isolate was used to generate a genomic library and to obtain biofilm-negative transposon insertion mutants. S. aureus SA113, a restriction-negative and biofilm producer strain, and the biofilm-defective S. aureus SA113ica, which contains a tetracycline resistance cassette that replaces the ica genes, were kindly provided by F. Götz (9). S. aureus RN4220, a restriction-negative strain, unable to form a biofilm, was kindly provided by J. C. Lee. Seventy-five human S. aureus strains and 50 coagulase-negative Staphylococcus strains from animals with bovine mastitis were analyzed for the presence of bap.

DNA manipulations. Routine DNA manipulations were performed using standard procedures (41). Plasmid DNA was isolated from S. aureus strains using a Qiagen plasmid miniprep kit according to the manufacturer's protocol, except that the bacterial cells were lysed by lysostaphin (12.5 µg/ml; Sigma) at 37°C for 2 h before plasmid purification. Plasmids were transformed into staphylococci by electroporation, using a previously described procedure (25) with the following modifications: 10% glycerol was replaced by 0.5 M sucrose, and staphylococcal transformations were enhanced by inducing chloramphenicol acetyl transferase translation with subinhibitory concentrations of chloramphenicol (0.2 µg/ml) after electroporation.

Transposon mutagenesis. S. aureus strain V329 harboring pID408 (V329:pID408) was grown overnight in TSA-chloramphenicol at 30°C. A single colony of V329:pID408 was inoculated in 1 ml of TSB-chloramphenicol (5 µg/ml) and incubated for 1 h at 30°C. Subsequently, 100 µl of this culture were spread on TSA-erythromycin plates and incubated for 18 h at 44°C. Transposon insertion mutants were subcultured on TSA-erythromycin plates. To exclude the possibility of contamination, mutants were compared with the parental strain by coagulase DNA typing (21).

Adherence studies. (i) Early adherence to an inert surface. Early adherence to a polystyrene surface was determined as previously described (14), with the following modifications. S. aureus strains were grown overnight in TSB supplemented with 0.25% glucose at 37°C. Overnight cultures were adjusted with TSB-0.25% glucose to an optical density at 578 nm (OD₅₇₈) of 0.1. Ten milliliters of each suspension was added to two polystyrene petri dishes and incubated for 1 h at 37°C. Petri dishes were washed at least five times with phosphate-buffered saline (PBS). Cells were fixed with Bouin solution and Gram stained. Adherent bacterial cells were observed by oil immersion microscopy and counted (results represent the means of four different microscopic fields). Each experiment was repeated three times.

(ii) Biofilm assay. A late adherence assay was carried out essentially as described elsewhere (18). Briefly, S. aureus strains were grown overnight in TSB at 37°C. The culture was diluted 1:40 in TSB-0.25% glucose, and 200 µl of this cell suspension was used to inoculate sterile 96-well polystyrene microtiter plates (lwaki). After 12 h, medium was replaced, and 12 h later, the wells were gently washed three times with 200 µl of sterile PBS, dried in an inverted position, and stained with 0.1% safranin for 30 s. Wells were rinsed again, and the absorbance was determined at 490 nm (Micro-ELISA Autoreader; Elx800 Bio-Tek instruments). Each assay was performed in triplicate and repeated five times.

(a) Formation of cell aggregates in a cell suspension. Cells were grown overnight in 5 ml of TSB-0.25% glucose at 37°C and examined macroscopically and microscopically for the presence or absence of aggregates (intercellular adhesion).

(b) Macroscopic observation of biofilm on glass. Cells were grown in 50 ml of B2 at 37°C, using a glass container, without shaking, for 2 days, and the walls of the container were visually (macroscopically) examined for the presence or absence of a white biofilm layer.

(c) Colonization of other materials. The capacity to form a biofilm layer was verified using polyvinylchloride (PVC) plastic as a target surface and a phase-contrast microscope (magnification, ×1,000; Nikon Optiphot-2 microscope), as previously described (38).

(d) Colony morphology in CRA. Colony morphology was determined in CRA as previously described (6, 48), with rough colonies being indicative of biofilm formation (positive result) and smooth colonies being associated with a deficiency in biofilm formation.

PIA-PNSG detection. PIA-PNSG production in S. aureus was detected as described by Cramton et al. (9). Briefly, cells were grown overnight in TSB supplemented with 0.25% glucose, the optical density was determined, and the same number of cells (2 to 4 ml) from each culture was resuspended in 50 µl of 0.5 M EDTA (pH 8.0). Cells were then incubated for 5 min at 100°C and centrifuged to pellet the cells, and 40 µl of the supernatant was incubated with 10 µl of proteinase K (20 mg/ml; Sigma) for 30 min at 37°C. After addition of 10 µl of Tris-buffered saline (20 mM Tris-HCl, 150 mM NaCl [pH 7.4]) containing 0.01% bromophenol blue, 10 µl was spotted on a nitrocellulose filter using a Bio-Dot Microfiltration apparatus (Bio-Rad), blocked overnight with 5% skim milk in PBS with 0.1% Tween 20, and incubated for 2 h with an anti-S. aureus PIA-PNSG antibody diluted 1:10,000 (30). Bound antibodies were detected with a peroxidase-conjugated goat anti-rabbit immunoglobulin G antibody (Jackson ImmunoResearch Laboratories, Inc.) diluted 1:10,000, and the Amersham ECL (enhanced chemiluminescence) Western blotting system.

Gene identification. A previously described method (31) was used to clone the chromosomal DNA flanking the transposon insertion points in mutants with attenuated biofilm production. Briefly, 2.5 µg of S. aureus chromosomal DNA from each mutant was restricted with EcoRI, resuspended in 200 µl of ligation buffer (Promega), and self-ligated for 12 h at 16°C. The ligated products were transformed into E. coli DH5, plated onto Luria-Bertani agar containing ampicillin, and incubated at 37°C overnight. Plasmid DNA was extracted from a single ampicillin-resistant colony using a Qiagen plasmid miniprep kit. Chromosomal DNA sequences flanking the transposon were obtained using primer Tn917-3c (5'-AGAGAGATGTCACCGTCAAGT-3'), which corresponds to the inverted repeat region located 70 bp from the erm-proximal end of Tn917.

Construction of genomic library. The lambda vector EMBL-4 was used to construct a genomic library of S. aureus V329 according to the manufacturer's instructions (Promega). Chromosomal DNA of S. aureus V329 was digested with EcoRI and ligated into vector EMBL-4 restricted with EcoRI. A 200-bp PCR fragment of the bap flanking region was used as a DNA probe (oligonucleotides 556-1m, 5'-CTGTCCATATTTGGACTGTG-3', and 556-2c, 5'-CTTATAGATGTGCGTAGTC-3'). Labeling of the probe and DNA hybridization were performed according to the protocol supplied with the PCR-DIG DNA-labeling and chemiluminescence detection kit (Roche).

DNA sequencing and computer analysis. A genomic HindIII fragment including the bap gene was cloned in pBT2 plasmid (pBT2:Bap). The nucleotide sequence was determined by the dideoxy chain termination method, using an ABl 377 model automatic sequencer (PE Biosystems; Foster City, Calif.) at the DNA Sequencing Service of the IBMCP-UPV (Valencia, Spain). For C-repeat sequencing, a genomic DNA XbaI-EcoRI fragment containing this region was subcloned in pBT2. Nested deletions were generated (Erase-a-base system; Promega). Double DNA sequencing of this region was carried out using the primers pBT2-1c (5'-GGACGATATCCCGCAAGAGGCCCG-3') and pBT2-2c (5'-GGTGCCGAGGATGACGATGAGCGC-3'), corresponding to the flanking region of the plasmid pBT2 cloning site.

Complementation studies. To prove that the biofilm-deficient phenotype of the mutants was due to the disruption of bap, mutants were complemented with plasmid pBT2 carrying a 9.2 kb HindIII fragment from the wild-type strain, including the bap gene under the control of its own promoter (pBT2:Bap). Plasmid pBT2:Bap was transformed by electroporation into S. aureus strains m556, RN4220, SA113, and SA113ica. Stable expression of Bap was analyzed in total bacterial extracts by Coomassie staining of proteins run on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels.

SDS-PAGE and Western blot analysis. Bacteria were grown overnight in TSB (supplemented with 10 µg of chloramphenicol/ml in the case of complemented strains) at 30°C. Following centrifugation of 1 ml of culture, cells were harvested, washed, and finally resuspended in 75 µl of PBS buffer containing lysostaphin (12.5 µg/ml; Sigma). After 2 h of incubation at 37°C, cells were resuspended in 1 volume of Laemmli buffer and boiled for 10 min. After centrifugation, 20 µl of the supernatant was used for SDS-PAGE (10% separation gel, 4.5% stacking gel) and proteins were stained with Coomassie brilliant blue R250 (0.25%; Sigma).

Expression of the N-terminal region of Bap in E. coli. A 1,919-bp DNA sequence corresponding to nucleotides (nt) 1 through 1919 of the bap gene was amplified by PCR with primers bap-2m (5'-GGGGGGCATATGGGAAATAAACAAGGTTTTTTACC-3') and bap-3c (5'-GGGGGGATCCCCAACCTCGTCAATGGTTAAGTCAGC-3') (NdeI and BamHI restriction sites are shown in boldface type). The amplified product was restricted with NdeI and BamHI and cloned in frame downstream from the His tag sequence in the pET-15b vector. The nucleotide sequence of the cloned bap fragment was verified by sequence analysis. Purified plasmid DNA was used to transform the expression host BL21(DE3), and the fusion protein was produced as specified by the manufacturer (Novagen). After disrupting the cells by sonication, the recombinant protein was purified by immobilized metal affinity chromatography using a cobalt-based resin (Clontech).

Production of rabbit polyclonal antiserum. Polyclonal antiserum to purified recombinant protein was raised as previously described (43). For the initial dose (day 1), 100 µg of antigen in complete Freund's adjuvant was injected subcutaneously. Booster doses (50 µg) were administered intramuscularly on days 14 and 42 in incomplete adjuvant. Blood was collected from the marginal ear vein at 2-week intervals after booster administration, and antibody titer in serum was determined.

Experimental infection. A mouse foreign body infection model was used to determine the role of bap in the pathogenesis of S. aureus. A total of 53 adult male mice (Swiss-Albino, B&K Universal, Barcelona, Spain) were used. A 1-cm segment of intravenous catheter (22G1"; B. Braun) was aseptically implanted into the subcutaneous interscapular space. Each group of nine mice was inoculated with 1.5 × 10⁵ CFU of either S. aureus V329 or the Bap-defective S. aureus strain m3591. Three animals were euthanatized by cervical dislocation on days 4, 7, or 10 postinfection. The catheter was aseptically removed, placed in a sterile microcentrifuge tube containing 1 ml of PBS, and vortexed at high speed for 3 min. The number of bacteria was determined by plate count. The experiment was repeated three times.

Statistical analysis. A two-tailed Student's t test was used to determine the differences between groups in biofilm formation. A nonparametric test (Mann-Whitney U test) was used to assess significant differences in bacterial recovery within groups in primary adherence and experimental infection. For analysis of the cure ratio, a two-by-two contingency table was produced and Fischer's exact test was applied. Differences were considered statistically significant when P was <0.05 in all cases.

Nucleotide sequence accession number. The DNA sequence reported in this article has been deposited in the GenBank nucleotide sequence database under accession number AF288402.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Production and characterization of Tn917 mutants. Upon transfer of plasmid pID408 into the adherent S. aureus strain V329, a collection of approximately 4,000 random Em^r Cm^s transposon insertions was screened for their ability to form a biofilm. Two mutants, designated m556 and m3591, had lost the ability to form a biofilm (Fig. 1) and exhibited a growth rate indistinguishable from that of the wild type. Southern hybridization analysis of EcoRI-digested chromosomal DNA using a Tn917-specific probe revealed that each mutant had a single transposon insertion (data not shown).

View larger version (75K): [in this window] [in a new window]   FIG. 1.   Biofilm formation phenotype. Differences between the wild-type strain V329 and mutants m556 and m3591 in the capacity to form a 24-h biofilm on polystyrene microtiter plates after staining with safranin. The ELISA plate colors correspond to OD490s of 1.3, 0.17, and 0.18, respectively.

Bap expression allows and enhances biofilm formation. A double band that migrated at a position corresponding to 230 and 240 kDa was reproducibly detected by SDS-PAGE of the total protein extract from the wild-type bacteria but not from the m556 mutant (Fig. 2A). Bands of similar mobility were also detected in S. aureus strains which were complemented with the bap gene in pBT2:Bap. In addition, a band of the expected size was recognized by polyclonal antibodies raised against the first 640 aa of Bap (Fig. 2B), strongly suggesting that the band present in the wild type and complemented strains is the product of the bap gene.

View larger version (71K): [in this window] [in a new window]   FIG. 2.   (A) SDS-PAGE of protein extracts of wild-type strain V329, m556, RN4220, SA113, and the corresponding complemented strains: m556 pBT2:Bap (m556, lane +), RN4220 pBT2:Bap (RN4220, lane +), and SA113 pBT2:Bap (SA113, lane +). Note that a double protein band of 230 and 240 kDa is present in the wild-type strain as well as in all the strains harboring plasmid pBT2:Bap. (B) Study of the presence of Bap by Western blotting. An approximately 240-kDa band is recognized by polyclonal antibodies against the first 640 aa of Bap only in the wild-type strain V329 and complemented strains: m556 pBT2:Bap (m556, lane +), SA113ica pBT2:Bap (SA113ica, lane +), and SA113 pBT2:Bap (SA113, lane +).

View larger version (61K): [in this window] [in a new window]   FIG. 3.   pBT2:Bap complementation studies involving biofilm formation in biofilm-defective mutant m556, non-biofilm-forming strain RN4220, and biofilm-forming strain SA113 and its ica-defective mutant SA113ica. Biofilm formation capacity differences correspond to 24-h biofilm formed on polystyrene microtiter plates after staining with 0.1% safranin. The ELISA plates and mean OD490s obtained are shown. (A) ELISA wells corresponding to wild type V329, m556 () and m556 pBT2:Bap (+), RN4220 () and RN4220 pBT2:Bap (+), SA113ica () and SA113ica pBT2:Bap (+), and SA113 () and SA113 pBT2:Bap (+). (B) Mean optical density. Bars represent the mean values, and error bars represent the standard errors of the means (n = 5). Significant differences in adherence (P < 0.01) were noted between complemented and noncomplemented strains, as well as between the wild type and isogenic mutants.

Bap is involved in primary adherence and intercellular adhesion. Primary adherence of S. aureus strains V329, m556, SA113, and SA113ica is illustrated in Fig. 4. S. aureus V329 adhered to polystyrene much more efficiently than the isogenic bap mutant m556 and strains SA113 and SA113ica. In addition, the presence of cell-to-cell clusters was observed in V329 and SA113 cultures (incubated overnight in TSB-0.25% glucose) by phase-contrast microscopy (data not shown), but not in the mutant m556 and SA113ica cultures. Accumulation of cell aggregates at the bottom of the tube was macroscopically observed only in the case of wild type strains. These results strongly suggest that Bap is not only involved in intercellular adhesion and accumulation in multilayered cell clusters, as the product of the ica operon does, but also in primary attachment to an abiotic surface.

View larger version (12K): [in this window] [in a new window]   FIG. 4.   Primary attachment assay. Significant (P < 0.01) differences were detected between wild-type strain V329 and isogenic mutant m556. No differences were found between S. aureus SA113 and isogenic mutant SA113ica. Bars represent the mean values, and error bars represent the standard errors of the means (n = 3).

Relationship between Bap and PIA-PNSG. To determine Bap-PIA-PNSG interaction we tested PIA-PNSG production in Bap⁺ and Bap strains. The strain V329 showed a low level of in vitro PIA-PNSG production. Similar results were obtained whith other natural Bap⁺ strains (data not shown). Inactivation of the bap gene in the mutant strain m556 reduced the levels of PIA-PNSG (Fig. 5). Complementation of the PIA-PNSG-producing strain SA113 with the bap gene strongly increased PIA-PNSG accumulation (Fig. 5). As expected, PIA-PNSG was no longer produced in the ica knockout strain SA113ica (Fig. 5).

View larger version (84K): [in this window] [in a new window]   FIG. 5.   Dot blot analysis of S. aureus PIA-PNSG acummulation induced by Bap expression. S. aureus V329 (blot A1) and m556 (blot B1) produced low levels of PIA-PNSG. Complementation of S. aureus SA113 with pBT2-Bap plasmid strongly increased the levels of PIA-PNSG (blot B2) with respect to the wild-type strain (blot A2). PIA-PNSG was not detected in S. aureus SA113ica (blot A3).

Further phenotypic characterization of biofilm-defective mutants. Consistent with the results on polystyrene microtiter plates, the wild-type strain (V329) grown on CRA exhibited a rough colony morphology typically associated with biofilm producers, whereas mutants m556 and m3591 produced a smooth colony morphology commonly found in non-biofilm-producing strains (Fig. 6A). Phase-contrast microscopic observation of late adherence to PVC plastic discs showed that the wild-type strain produced multiple layers of cells almost completely covering the PVC surface (Fig. 6B). In contrast, very few cells of the biofilm-defective mutant were attached to PVC plastic. Macroscopic examination of biofilms in a glass container revealed that upon 2 days of culture, the wild-type strain formed an obvious biofilm on the glass surface, but the mutants did not (Fig. 6C).

View larger version (98K): [in this window] [in a new window]   FIG. 6.   Phenotypic differences between the wild-type strain V329 and the defective mutant m556. (A) CRA colony morphology; (B) capacity to form a 24-h biofilm on PVC plastic, as observed by phase-contrast microscopy (magnification, ×1,000); (C) capacity to form a 48-h biofilm on the surface of a glass container (visual observation).

Structural features of Bap protein. Analysis of the Bap primary amino acid sequence revealed the presence of a typical gram-positive amino-terminal signal sequence for extracellular secretion (first 44 aa) and a putative carboxy-terminal segment containing an LPXTG motif and a hydrophobic membrane-spanning domain followed by a series of positively charged residues typical of cell-wall-anchored surface proteins of gram-positive bacteria (34) (Fig. 7A). Following the putative signal peptide, the N-terminal region of Bap can be divided into two regions. Region A (aa 45 to 360) contains two short repeats of 32 aa (repeats A₁ and A₂) separated by 26 aa (Fig. 7B). BlastP searches of this region revealed no significant similarity scores among GenBank sequences. Region B, the remaining part of the N-terminal domain (aa 361 to 818), exhibited a significant similarity with an Enterococcus faecalis surface protein (Esp) (43), which is mostly found in infection-derived isolates.

View larger version (58K): [in this window] [in a new window]   FIG. 7.   (A) Structural analysis of Bap. S represents the signal sequence. The positions of the LPXTG motif and the A, B, C, and D domains are shown. Region A (aa 45 to 360) contains two short repeats of 32 aa (A1 and A2) separated by 26 aa. Region B (aa 361 to 818) represents the remaining part of the N-terminal domain. A spacer region (aa 819 to 947) followed by region C (aa 948 to 2139) consisting of 258-nt tandem repeats constitutes the central domain of Bap. The carboxy-terminal region of Bap consists of region D (aa 2148 to 2208), which contains three short repeats of 18 aa followed by an incomplete D repeat, and the LPXTG motif. (B) Alignment of the amino acid residues within repeat blocks A, C, and D of the Bap protein. Substitutions are shown in boldface type.

Distribution of the bap gene among staphylococcal species. PCR amplification and Southern blot analysis using specific probes for the bap gene revealed that bap is present in only 5% of the 350 S. aureus bovine mastitis isolates tested. The presence of bap could not be detected in any of the 75 human S. aureus isolates studied. All the strains harboring bap were strong biofilm producers. Sequences that hybridized with the bap probe were present in 10% of the 50 coagulase-negative Staphylococcus isolates tested from animals with bovine mastitis (data not shown).

Experimental infection. Bap contributed to the pathogenesis of S. aureus in the murine catheter-induced infection model (Fig. 8). Although at day 4 postinoculation differences between wild-type and mutant strains in the number of recovered bacteria per catheter were non significant, at day 7 postinoculation, the number of recovered bacteria (CFU) was 1.2 × 10⁶ and 2.8 × 10⁵ for the wild-type and mutant strains, respectively (P > 0.05). This difference increased by day 10, when the values were 1.7 × 10⁶ and 3 × 10⁴ CFU, respectively (P < 0.05).

View larger version (24K): [in this window] [in a new window]   FIG. 8.   Recovery of S. aureus V329 and its isogenic Bap mutant m3591 from implanted subcutaneous catheter in a mouse foreign body infection model. Bars represent the mean of CFU collected from catheters, and the error bars represent the standard errors of the means (n = 9). At day 10, differences between wild type and mutant were detected (P < 0.05). Bacteria were not detectable in control animals at the end of the experimental period.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In S. epidermidis, several surface proteins involved in biofilm formation have been described, three of which (SSP1, SSP2, and the AtlE autolysin [19, 45]) contribute to the primary attachment, and the other, a 140-kDa protein mediating intercellular adhesion, has been proposed as essential for accumulation of sessile bacteria on glass or polystyrene surfaces (22). The absence of homology among these proteins strongly suggests that staphylococci may use different approaches to form a biofilm. To our knowledge, data presented in this report describe the first S. aureus surface protein (Bap) directly involved in biofilm formation on abiotic surfaces in the absence of host plasma constituents.

To identify Bap, we have used the standard biofilm assay on microtiter plates for the screening of transposon mutants unable to adhere to the polystyrene surface. Surprisingly, in contrast to similar studies with other microorganisms like P. aeruginosa (35), E. coli (38), and Vibrio cholerae (47), where several genes were found to be involved in the biofilm formation process, this assay allowed identification of only two mutants in which the transposon affected the same 6,831-bp open reading frame designated bap.

The bap gene displayed little sequence similarity with known genes. However, the 2,276-aa bap product displayed an organizational similarity with an outer membrane protein of P. putida involved in adhesion to seeds (12) and with a surface protein of E. faecalis (Esp) of unknown function (43), but whose presence is highly correlated with the ability of this bacteria to produce a biofilm on abiotic surfaces (our unpublished results). The most remarkable feature of Bap is the presence of an extensive repeat region, in which the repeats are identical even at the nucleotide level. Although the biological function of the repeats has not been established, it is tempting to speculate that this region could serve to project the amino-terminal part of the protein from the cell surface and to promote interaction with abiotic surfaces, host cell components, or other bacteria. Analogous long repeats have been described in the alpha C protein of group B streptococci (46), where addition or deletion of repeat units leads to the expression of variant proteins which allow bacteria to escape the host immune response (27). Similar pathogenic mechanisms might occur in different Bap⁺ strains, in which a variable number of C repeats can be found in the structure of the bap gene.

Primary attachment, intercellular aggregation, and biofilm formation studies showed that Bap promotes primary attachment as well as the second step of biofilm formation, where intercellular adhesion plays an important role. In this context, since complementation of S. aureus SA113 with bap resulted in a significantly increased ability to form a biofilm and PIA-PNSG accumulation, it is tempting to speculate that PIA-PNSG and Bap may cooperatively affect cell-to-cell aggregation during the biofilm formation process. A possible explanation of the observation that biofilm formation capacity was not completely restored in the complemented m556 mutant is that the insertion of the transposon in mutant m556 resulted in the production of a truncated protein lacking the last 15 aa, and therefore the protein would be unable to anchor to the cell wall. The soluble truncated Bap protein may cover the abiotic surface and compete with the cell-associated Bap for the interaction with the abiotic surface.

CRA has been used to discriminate by colony morphology between biofilm and non-biofilm-forming strains of S. aureus (2, 6) and S. epidermidis (48), since they produce rough and smooth colonies, respectively. Congo red interacts with several proteins and proteinaceous fibrillar structures, such as curli fimbriae from E. coli (17) and the type III secretion machinery of Shigella flexneri (4). In the case of bap-harboring S. aureus, loss of the Bap protein resulted in transformation of the rough colony morphology (of the wild-type V329 strain) into the smooth colony morphology (of the m556 mutant). A similar phenotypic variation has been described in S. epidermidis strains deficient in the production of PIA-PNSG (the ica product) (48) and S. aureus strain SA113ica (our unpublished results). When the expression of bap was restored in complemented strains, the rough phenotype appeared, suggesting the implication of Bap in this phenotypic variation. Why the deficiency in either the polysaccharide intercellular adhesin (PIA-PNSG) or a surface protein (Bap) results in the same phenotypic change is unknown.

Animal models have been useful to study the importance of different genes in the pathogenesis and virulence of S. aureus (30, 31). In this report, we used a mouse foreign body infection model to evaluate the role of Bap and the resulting biofilm formation process in the pathogenesis of S. aureus. Differences between the Bap-deficient mutant S. aureus m3591 and the parental strain in the capacity to colonize the catheter became more obvious at late stages of infection (by day 10), when the mutant strain showed a decreased persistence relative to the wild type. In the closely related species S. epidermidis, the expression of specific bacterial cell surface components appears to hinder the interaction of particular bacterial cell receptors with host proteins (5). In our model, the catheter may have become rapidly coated in vivo by host proteins after implantation and Bap might have hindered the interaction between bacterial receptors and the host proteins on the catheter. This may explain why a Bap-deficient mutant may be more prevalent than the wild-type strain at the initial stages of infection (up to day 4). Later on (days 7 to 10), biofilm formation would be strongly enhanced by Bap as infection proceeds, by promoting cell-to-cell interactions, bacterial accumulation, and persistence of infection.

In conclusion, this study describes a novel protein of S. aureus involved in biofilm formation on abiotic surfaces. Attachment to abiotic surfaces might not be necessarily related to attachment to biotic surfaces. In fact, there are examples that support the idea that bacterial biofilm formation can proceed through divergent pathways, depending on whether or not bacteria settle on a surface or in an environment that can provide nutrients (47). Other results demonstrate that the same factors may be involved in attachment to both types of surfaces (12). Probably, during the development of infections such as subclinical mastitis, biofilm formation could be an efficient way of persisting in the microenvironment of the udder, where shear forces arise during milking (6). The presence of Bap in S. aureus strains responsible for subclinical mastitis suggests that Bap may enhance intramammary adherence and biofilm formation, leading to the inefficacy of antibiotic treatment against biofilm bacteria and chronicity. Further studies to demonstrate this hypothesis are warranted.