SEARCH

SEARCH BY CITATION

Keywords:

  • Staphylococcus epidermidis;
  • biofilms;
  • spx;
  • ClpP protease

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Infections caused by the leading nosocomial pathogen Staphylococcus epidermidis are characterized by biofilm formation on implanted medical devices. In a previous study, we found that ClpP protease plays an essential role in biofilm formation of S. epidermidis. However, the mechanism by which ClpP impacts S. epidermidis biofilms has remained unknown. Here, we show that the Spx protein accumulates in the clpP mutant strain of S. epidermidis and controls biofilm formation of S. epidermidis via a pronounced effect on the transcription of the icaADBC operon coding for the production of the biofilm exopolysaccharide polysaccharide intercellular adhesion (PIA). Notably, in contrast to Staphylococcus aureus, Spx controls PIA expression via an icaR-independent mechanism. Furthermore, Spx affected primary surface attachment, although not by regulating the production of the autolysin AtlE. Our results indicate that ClpP enhances the formation of S. epidermidis biofilms by degrading Spx, a negative regulator of biofilm formation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Staphylococcus epidermidis, previously regarded as an innocuous commensal bacterium of the human skin, has emerged as one of the most frequent causes of nosocomial infection in recent years. Staphylococcus epidermidis may cause persistent infections by forming biofilms on implanted medical devices, such as central venous catheters, urinary catheters, prosthetic heart valves and orthopedic devices. Biofilms are intercellular-attached agglomerations that have intrinsic resistance to antibiotic agents and host defense (Donlan & Costerton, 2002). Thus, infections caused by S. epidermidis biofilms are particularly hard to eradicate.

Biofilm formation by S. epidermidis is a multistep process and involves (1) attachment of the bacterial cells to a polymer surface or to the host-derived matrix that has previously coated the polymeric device and (2) accumulation to form multilayered cell clusters with cell-to-cell adherence mediated by the production of a slimy extracellular matrix (Vadyvaloo & Otto, 2005). Several genes have been identified to play important roles in the biofilm formation of S. epidermidis (Mack et al., 2007). The atlE gene encodes autolysin AtlE, which mediates the initial attachment of S. epidermidis to a polymer surface (Heilmann et al., 1997), and the ica gene locus (icaADBC) encodes the biosynthesis of polysaccharide intercellular adhesion (PIA), which is essential in the accumulation process (Heilmann et al., 1996). A few regulatory genes of biofilm formation were also identified (Mack et al., 2007). For example, the icaR gene affects the ability of biofilm formation by repressing the icaADBC operon (Conlon et al., 2002). The sarA gene encodes an activator of the icaADBC operon and positively regulates the biofilm formation of S. epidermidis (Tormo et al., 2005). The rsbU gene, a positive regulator of the alternative sigma factor, σB, positively regulates the biofilm formation of S. epidermidis by repressing icaR (Knobloch et al., 2004). Besides, LuxS (Xu et al., 2006) and Agr (Kong et al., 2006), a quorum-sensing system, also mediate biofilm formation in S. epidermidis. Recent work indicates that the regulation of biofilm formation in S. epidermidis is a complex networking and may involve mechanisms other than the ica system. The sarZ gene encodes a regulator that activates the transcription of the icaADBC operon in an icaR-independent manner and positively regulates the biofilm formation of S. epidermidis (Wang et al., 2008) Additionally, it is not uncommon to find clinical isolates that accumulate biofilm in an ica-independent mode (Ruzicka et al., 2004; Hennig et al., 2007; Qin et al., 2007), which indicates that there may be other mechanisms mediating biofilm formation.

Protein degradation is essential for cell viability and homeostasis, and this process is commonly mediated by ATP-dependent proteases. One notable case is ClpXP proteases, which function in degrading SsrA-tagged misfolded proteins (Gottesman et al., 1998), controlling the RpoS concentration in Escherichia coli (Gottesman et al., 1998) and regulating bacterial adaptation to stress (Porankiewicz et al., 1999). ClpXP proteases also play a crucial role in the biofilm formation of Pseudomonas fluorescens (O'Toole & Kolter, 1998), Streptococcus mutans (Lemos & Burne, 2002), Staphylococcus aureus (Frees et al., 2004) and S. epidermidis (Wang et al., 2007).

Recently, a suppressor of clpP and clpX, named as Spx, has been reported as a global regulatory protein, and is highly conserved among low-G+C-content gram-positive bacteria (Zuber, 2004). Many observed phenotypes of clpXP mutants in both Bacillus subtilis and S. aureus are caused by the accumulation of Spx (Nakano et al., 2002; Frees et al., 2004; Pamp et al., 2006). In B. subtilis, Spx activates the transcription of the trxA and trxB genes that function in thiol homeostasis (Nakano et al., 2005) and the yrrT operon that functions in organosulfur metabolism (Choi et al., 2006), whereas it represses the transcription of the srf operon involved in competence development and the hmp gene involved in anaerobic respiration (Nakano et al., 2003b; Zuber, 2004). In both B. subtilis and S. aureus, Spx is demonstrated as a substrate of ClpP proteases, and the cellular level of Spx is tightly controlled (Nakano et al., 2002, 2003b). Interestingly, Spx negatively regulates biofilm formation in S. aureus, which is likely mediated by its positive effect on the transcription of icaR (Pamp et al., 2006). Whether Spx affects the biofilm formation of S. epidermidis is unknown.

In a previous study, we found that ClpP plays an essential role in the biofilm formation of S. epidermidis (Wang et al., 2007). Here, we demonstrate that the expression level of Spx increased drastically without the degradation by ClpP protease in S. epidermidis. To explore the function of Spx in S. epidermidis, we constructed an spx-overexpressing strain. It was further found that Spx plays a role in biofilm formation, whereas it has no impact on the stress responses of S. epidermidis. In addition, we show that Spx modulates the transcription of several genes that are involved in the biofilm formation via an icaR-independent manner.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Bacterial strains and growth media

The bacteria and plasmids used are listed in Table 1. Escherichia coli DH5α was grown in Luria–Bertani medium. Plasmid-containing E. coli strains were grown in the same medium, but with ampicillin (100 μg mL−1) included. Staphylococcus epidermidis and its derivative strains were cultured in B-medium (composed of 1% peptone, 0.5% yeast extract, 0.1% glucose, 0.5% NaCl and 0.1% K2HPO4× 3H2O), and when necessary, erythromycin (10 μg mL−1) was supplemented. Media were solidified with 1.5% (w/v) agar as needed.

Table 1.   Bacterial strains and plasmids used in this study
Strain or plasmidDescriptionSource or reference
Strains
S. epidermidis 1457WT strain, biofilm positiveMack et al. (1992)
S. epidermidis 1457 clpP mutantclpP mutant (ClpP Ermr)Wang et al. (2007)
S. aureus RN4220Restriction−, modification+Kreiswirth et al. (1983)
Plasmids
pYJ90Shuttle vector, ampicillin and erythromycin resistantJi et al. (1999)
pQG53pYJ90 harboring the promoter sequence of icaADBCThis study
pQG54pYJ90 harboring the icaADBC promoter, and downstream, the spx coding sequenceThis study
pQG55pYJ90 harboring the icaADBC promoter, and downstream, a mutant allele of spx (coding for a ClpP protease-resistant form of Spx)This study
pQG56pYJ90 harboring the icaADBC promoter, and downstream, the reverse spx coding sequenceThis study

DNA manipulation

Genomic DNA of S. epidermidis 1457 was prepared using a standard protocol for gram-positive bacteria (Flamm et al., 1984). Plasmid DNA from E. coli was extracted using a plasmid purification kit (HuaShun Co.). Plasmid DNA from S. aureus and S. epidermidis was extracted using the same kit, except that the cells were incubated for at least 30 min at 37 °C in solution P1 with lysostaphin (25 μg mL−1; Sigma) before solution P2 was added. Taq DNA polymerase (Ex Taq) and restriction enzymes were obtained from TaKaRa Biotechnology Company. Staphylococcus epidermidis was transformed by electroporation as described previously (Augustin & Gotz, 1990).

Construction of Spx-expressing plasmids and an spx antisense knockdown plasmid

Because the sequence and location of the endogenous promoter that facilitates spx transcription in S. epidermidis are unknown, we utilized the promoter sequence of the icaADBC operon. This fragment was PCR amplified from S. epidermidis 1457 genomic DNA using the primers Pica1 and Pica2 (Table 2), introducing EcoRI and BamHI cleavage sites, respectively. The amplified PCR products (0.5 kb) were cloned into the shuttle plasmid pYJ90 (Ji et al., 1999), yielding pQG53. The S. epidermidis spx gene with its ribosome-binding sequence was PCR amplified using the primers spx-u and spx-d, introducing BamHI and HindIII cleavage sites, respectively. The amplified PCR products (0.4 kb) were cloned into pQG53 (placed downstream of the icaADBC promoter), yielding pQG54. A 3′ terminal mutant allele of the S. epidermidis spx gene was constructed by mutagenic PCR using the primers spx-u and spx-d2m, introducing BamHI and HindIII cleavage sites, respectively. The amplified PCR products (0.4 kb) were cloned into pQG53 (placed downstream of the icaADBC promoter), yielding pQG55 for overexpression. To inhibit the expression of Spx, the coding sequence of spx was amplified with HindIII and BamHI using the primers spxa1–spxa2, and then ligated to PQG53, yielding the antisense plasmid PQG56.

Table 2.   Primers used in this study
PrimerDescriptionDNA sequence (5′–3′)*
  • *

    Incorporated restriction sites are in italic and underlined, while an incorporated point mutation is boldfaced and underlined.

Pica1For amplification of icaADBC promoter sequenceCGGGAATTCAGTGCTTCTGGAGCACTAAAC
Pica2For amplification of icaADBC promoter sequenceCGGGATCCACCTACCTTTCGTTAGTTAGG
spx-uFor amplification of spx coding sequenceCGTGGATCCTCTATAGGAGAGTGAGATGTATGGT
spx-dFor amplification of spx coding sequenceGGAAAGCTTAATCAACTAAACGTTGCGCTTCTTG
spx-d2mMutagenic primerGGAAAGCTTAATCATCTAAACGTTGCGCTTCTTG
spx-a1For spx antisense plasmidGGAAAGCTTATGGTAACATTATTTACTTCA
spx-a2For spx antisense plasmidCGTGGATCCAATCAACTAAACGTTGCGCTTCTTG
RT-gyrB1For quantitative RT-PCRTGACGAGGCATTAGCAGGTT
RT-gyrB2For quantitative RT-PCRGTGAAGACCGCCAGATACTTT
RT-atlE1For quantitative RT-PCRGATGGATTGCTGCTAAGGATTT
RT-atlE2For quantitative RT-PCRTATCGGTTTGCTTTTGTTGG
RT-icaB1For quantitative RT-PCRGAAACAGGCTTATGGGACTTTG
RT-icaB2For quantitative RT-PCRCAAGTGCGCGTTCATTTTT
RT-icaR1For quantitative RT-PCRCATTGACGGACTTTACCAGTTTT
RT-icaR2For quantitative RT-PCRATCCAAAGCGATGTGCGTAG

Semi-quantitative biofilm assay and primary attachment assay

Semi-quantitative biofilm assays and primary attachment assays were performed as described in our previous work (Wang et al., 2007), except that B-medium, in place of TSB medium, was used.

Diamide sensitivity test

The diamide sensitivity test was performed as described previously (Larsson et al., 2007) and modified as follows: S. epidermidis strains were grown in B-medium to the stationary phase and diluted in a fresh B-medium to an OD600 nm value of 0.1. Fifty microliters of the diluted culture was plated on a B-medium plate. Three disks, each with 5 μL of 500 mM diamide, were placed on the plate. The plate was incubated at 37 °C for 18 h, and the diameters of inhibition halos were measured.

Quantitative reverse transcriptase (RT)-PCR

Quantitative RT-PCR was performed as described previously (Vetter & Schlievert, 2007) and modified as follows: Staphylococcus epidermidis strains were grown in B-medium. At an OD600 nm of 0.5, cells in 2-mL cultures were harvested and resuspended in 1 mL Trizol (Invitrogen). The cell suspensions were transferred into Conical Screw Cap Microtubes (2.0 mL; Porex Bio Products Group), where 1/3 of the volume was glass beads (0.1 mm; Biospec Products). Cells were disrupted by shaking with a Mini-Beadbeater (Biospec Products) at maximum speed for 30 s. Tubes were then incubated on ice for 5 min. This shaking/cooling cycle was repeated four times. Then, the suspension was centrifuged. Total RNA isolation from the supernatant was performed according to the instructions on Trizol (Invitrogen). Total RNA was treated using the TUBRO DNA-free kit (Ambion) to remove contaminating DNA. Approximately 1 μg of total RNA was reverse transcribed with a ReverTra Ace-α kit (Toyobo) using random primers. Of the 20-μL reverse-transcription reaction, 0.2 μL was used as a template for real-time PCR using SYBR-green PCR reagents (Toyobo), and the reactions were performed in an iCycler machine (BioRad). Reactions were performed at 95 °C for 3 min and then 95 °C for 15 s, 55 °C for 15 s and 72 °C for 15 s for a total of 40 cycles. The primers used in the quantification of the mRNAs are listed in Table 2. Constitutive gyrB transcription was used as an internal standard for RNA concentration. The transcript level of experimental genes was calculated relative to gyrB transcripts. The relative transcriptional level of experimental genes in the S. epidermidis wild-type (WT) strain was set to 1, and the level in the other strains was calculated proportionally. Data are from three independent experiments.

Immuno-dot blot assay

Immuno-dot blot assays were performed as described in our previous work (Xu et al., 2006).

Western blot

Western blot was performed as described previously (Pamp et al., 2006) and modified as follows: S. epidermidis strains were grown in B-medium. At an OD600 nm of 0.5, cells were harvested. Cell pellets were resuspended in 50 mM Tris-HCl (pH 8.0) and lysed by the addition of 25 μg mL−1 lysostaphin (Sigma) and incubation at 37 °C for 60 min. Cell debris was removed by centrifugation. The protein concentration was determined using a BCA protein Assay kit (Keygen Biotech Co.). Twenty micrograms of each sample was separated on 15% sodium dodecyl sulfate-polyacrylamide gels, and then transferred onto a Protran-BA83 nitrocellulose membrane (Whatman). Spx was probed with a 1 : 1500 dilution of the Spx antibody (a generous gift from P. Zuber), a 1 : 1000 dilution of HRP-Goat anti-Rabbit IgG (Proteintech) and the ECL Advance Western Blotting Detection Kit (GE Healthcare Life Sciences).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Identification of spx from S. epidermidis

To determine whether S. epidermidis has the spx gene, we examined the available S. epidermidis genome information (Gill et al., 2005) and identified a candidate ORF whose predicted protein product was 80% identical and 95% similar to the B. subtilis Spx protein, as well as a conserved N-terminal CXXC motif. Staphylococcus epidermidis Spx is very similar to S. aureus Spx (identity at the amino acid level of 98%) (Gill et al., 2005). According to the fact that both the upstream and the downstream genes of S. epidermidis spx are transcribed in a direction opposite to that of spx, spx is probably an independent ORF with its own promoter.

Expression of Spx is increased in the clpP mutant strain

In B. subtilis, it was demonstrated that Spx is a substrate of ClpP protease from in vitro proteolysis experiments (Nakano et al., 2002, 2003b). In S. aureus, Spx accumulates remarkably in the absence of ClpP, strongly indicating that ClpP protease degrades Spx in S. aureus (Pamp et al., 2006). To investigate whether ClpP protease degrades Spx in S. epidermidis, we examined the expression level of Spx in the S. epidermidis clpP mutant strain by Western blot. A much higher Spx level was found in the clpP mutant strain (Fig. 1). Spx accumulates with the absence of ClpP protease, indicating that Spx may also be a substrate of ClpP protease in S. epidermidis, similar to B. subtilis and S. aureus.

image

Figure 1.  Spx level in Staphylococcus epidermidis 1457 WT strain (W), its isogenic clpP mutant strain (ΔclpP) and the plasmid-harboring S. epidermidis strains with pQG53, pQG54 or pQG55, as measured by Western blot. Samples were prepared when the cells were in the exponential phase (OD600 nm=0.5).

Download figure to PowerPoint

Overexpression of spx in S. epidermidis

To investigate the role of Spx in the biofilm formation of S. epidermidis, we intended to construct an S. epidermidis spx mutant strain. We followed the same allelic exchange strategy (Bruckner, 1997) as that used in the construction of an S. epidermidis clpP mutant strain (Wang et al., 2007). More than 2000 clones were screened, but the desired double-crossover strain in which spx is replaced by an erythromycin-resistance cassette was not found, although we indentified single-crossover strains as determined by PCR amplifying the spx bordering regions (data not shown). The attempt to construct an spx mutant stain with a high-efficiency system through pKOR1 (Bae & Schneewind, 2006) also failed (data not shown). We further used a molecular epidemiological approach to examine the existence of spx in a collection of 80 S. epidermidis (Li et al., 2009) clinical isolates. All tested strains harbor the spx gene, indicating the possibility that spx could be an essential gene (data not shown). Instead, we constructed an spx antisense knockdown plasmid PQG56 coding reversed spx mRNA to downregulate the expression of Spx.

In a previous study, Nakano et al. (2003a) overexpressed Spx in B. subtilis to study its regulatory functions. Because the construction of an S. epidermidis spx mutant strain failed, we attempted to overexpress Spx in S. epidermidis to study its regulatory effect on biofilm formation. Attempts to overexpress Spx in B. subtilis were at first unsuccessful due to the rapid degradation of the protein by ClpP protease. Successful overexpression of Spx was achieved when an spx mutant allele that codes for a protease-resistant form of Spx (C-terminal mutant) was constructed and expressed in B. subtilis (Zuber, 2004). Thus, in addition to the expression plasmid pQG54, which carries a WT spx, we constructed another expression vector (PQG55) with an altered spx allele, with a substitution from Ala and Asn codons in the C-terminal to two Asp condons to encode a mutated SsrA peptide, in order to avoid the SsrA peptide-tagged proteolysis by ClpXP. These three plasmids were transformed into S. epidermidis. To prevent the resistance from being degraded, we compared the expression level of Spx in strains carrying PQG53, PQG54 and PQG 55 separately. As a result, little Spx protein was detected in the vector control stain harboring PQG53 and the WT expression allele harboring PQG54, whereas Spx accumulated in the strain harboring PQG55 (Fig. 1).

Effect of Spx on the biofilm formation of S. epidermidis

Biofilm formations of S. epidermidis strains harboring different plasmids were compared using semi-quantitative assays. Biofilm formation of the strain harboring pQG54 was comparable with that of the vector control strain harboring pQG53, whereas biofilm formation of the strain harboring pQG55 decreased drastically (Fig. 2). The Spx levels in these strains were examined by Western blot. The result that Spx accumulated in the strain harboring pQG55, but not in the strain harboring pQG54, indicates that Spx had a negative effect on the biofilm formation of S. epidermidis. In the strain harboring pQG54, the WT Spx expressed from the plasmid is probably degraded by ClpP protease, and the biofilm formation of the S. epidermidis stain harboring PQG56 (spx antisense knock-down plasmid) is increased substantially, in accordance with the phenotype in the homologous spx mutant strain of S. aureus (Pamp et al., 2006). This observation further supports that spx is an important regulator mediating the biofilm formation of S. epidermidis.

image

Figure 2.  Biofilm formation of Staphylococcus epidermidis 1457 WT strain (W) and the isogenic strains transformed with pQG53, pQG54, pQG55 and pQG56, as measured using semi-quantitative biofilm assays. Data are derived from two independent experiments, with four replicates in each experiment. Rations vs. 1457 (W) were showed. **P<0.01 vs. W(pQG53).

Download figure to PowerPoint

Biofilm formation by S. epidermidis is generally considered as a two-step process, including primary attachment and biofilm accumulation. To investigate which step is affected by Spx, we first compared the attachment ability of the Spx-overexpressing strain (harboring pQG55) and the vector control strain (harboring pQG53). In primary attachment assays, the Spx-overexpressing strain showed decreased attachment ability (about 34-fold) to polystyrene compared with the WT strain, whereas the strains carrying either pQG53 or pQG54 showed no difference in primary attachment (Fig. 3a and b). To investigate whether the transcription of atlE was affected by Spx, quantitative RT-PCR was performed. The result indicates that the transcriptional level of atlE in the Spx-overexpressing strain carrying pQG55 shows no difference compared with the other three strains (Fig. 3c). This indicates that Spx does not affect the attachment ability by regulating atlE. We then compared the primary attachment on 96-well polyethylene plates between WT and ica-negative strains isolated from our previous work (Li et al., 2005), and no significant difference was found (data not shown).

image

Figure 3.  (a) Primary attachment of Staphylococcus epidermidis 1457 WT strain (W) and the isogenic strains transformed with pQG53, pQG54 or pQG55, as measured using primary attachment assays. Data are derived from two independent experiments, with four replicates in each experiment. **P<0.01 vs. W(pQG53). (b) Phase-contrast micrographs of cells attached to polystyrene microtiter plates (original magnification, × 40). (c) The relative transcriptional level of atlE in S. epidermidis 1457 WT strain (W) and the isogenic strains transformed with pQG53, pQG54 or pQG55, as measured by quantitative RT-PCR. Samples were prepared when the cells were in the exponential phase (OD600 nm=0.5). Data are derived from three independent experiments.

Download figure to PowerPoint

PIA is a key factor in the biofilm accumulation of S. epidermidis (Rupp et al., 1999). To investigate whether the production of PIA was affected by Spx, immuno-dot blot assays were performed. The Spx-overexpressing strain was found to produce significantly less PIA compared with the vector control strain (Fig. 4a). The transcription of the icaADBC operon and its repressor icaR among different strains was further examined by quantitative RT-PCR. Decreased icaADBC, but comparable icaR transcriptional levels were found in the Spx-overexpressing strain compared with the vector control strain (Fig. 4b and c). This result indicates that Spx affects PIA production by regulating the transcription of icaADBC in an icaR-independent manner.

image

Figure 4.  (a) PIA production of Staphylococcus epidermidis 1457 WT strain (W) and the isogenic strains transformed with pQG53, pQG54 or pQG55. PIA samples were isolated from the surface of cells in the exponential growth phase (4 h) and in the stationary growth phase (16 h) by boiling with 0.5 mol L−1 EDTA. PIA production was determined by immuno-dot blot analysis using anti-PIA antisera and quantified by photodigital analysis. Values are the means±SEMs from three independent experiments. *P<0.05; **P<0.01 vs. W(pQG53). (b and c) The relative transcriptional level of icaB (b) and icaR (c) in S. epidermidis 1457 WT strain (W) and the isogenic strains transformed with pQG53, pQG54 or pQG55, as measured by quantitative RT-PCR. Samples were prepared when the cells were in the exponential phase (OD600 nm=0.5). Data are derived from three independent experiments.

Download figure to PowerPoint

Overexpression of Spx contributes little to the stress adaptation of S. epidermidis

In B. subtilis and S. aureus, Spx plays an important role in the oxidative-stress adaptation. The B. subtilis and S. aureus spx mutant strains were hypersensitive to diamide, a thiol-specific oxidant (Nakano et al., 2003a; Pamp et al., 2006). To study whether the overexpression of Spx affects S. epidermidis in the adaptation to diamide, the diamide sensitivity of the Spx-overexpressing strain (harboring pQG55) and the control strain (harboring pQG53) was compared using disk diffusion tests. No significant difference in diamide sensitivity was found between these two strains; the diameter of the inhibition halo for the Spx-overexpressing strain was 19.0±1.2 mm, while that for the control strain was 18.5±0.4 mm. This indicates that the overexpression of Spx has little effect on the adaptation of S. epidermidis to diamide. To further confirm this, we extracted the RNA from WT and the spx-overexpressing strain, and determined the transcription level of trxB (encoding thioredoxin reductase and associated with thiol homeostasis). The transcription level of trxB was induced remarkably through the overexpression of Spx in B. subtilis, and decreased in both B. subtilis and S. aureus spx mutants. Consistent with our phenotypic observation, no significant difference of the trxB transcriptional level was found between the spx-overexpressing strain and WT. In addition, we found that the overexpression of Spx has little effect on the stress response of S. epidermidis to ethanol or hydrogen peroxide (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

Spx is a conserved protein in low-G+C-content gram-positive bacteria. Its cellular level is controlled by ClpP protease. In both B. subtilis and S. aureus, Spx functions as a novel type of transcriptional regulator, and was proved as a substrate of ClpP protease (Nakano et al., 2002, 2003b; Pamp et al., 2006). Many bacterial functions are regulated by Spx, such as competence, thiol homeostasis and biofilm formation (Zuber, 2004; Pamp et al., 2006). We found a much higher level of Spx in the clpP mutant strain, suggesting that Spx is likely a substrate of ClpP protease in S. epidermidis. The spx-overexpressing plasmid was constructed by modification of the C-terminal of spx gene, which also supports this view.

In S. aureus, Spx negatively regulates biofilm formation (Pamp et al., 2006). In our study, decreased biofilm formation was found in the S. epidermidis Spx-overexpressing strain. The primary attachment and the PIA production were severely reduced in the Spx-overexpressing strain compared with WT, and the biofilm by the strain carrying the antisense spx plasmid was decreased compared with WT. The transcription of atlE was similar to WT in the Spx-overexpressing strain, indicating that Spx does not affect the primary attachment through inhibiting the transcription of atlE. Olson et al. (2006) have previously found that PIA enhanced the adherence of S. epidermidis to several orthopedic prosthetic biomaterials, including zirconia, ultra-high-molecular-weight polyethylene and cobalt chromium, but had no impact on the primary attachment to polymethylmethacrylate and titanium. In our study, there was no notable difference in the level of the initial adherence between WT and the isogenic PIA-negative strain under the selected experimental conditions. Thus, the mechanism behind the decreased initial attachment of the Spx-overexpressing strain needs further investigation. Decreased icaADBC transcription was found in the Spx-overexpressing strain of S. epidermidis, indicating that Spx affects PIA production by regulating the transcription of icaADBC. In S. aureus, icaR was found to be positively regulated by Spx (Pamp et al., 2006). However, the transcription of icaR in the S. epidermidis Spx-overexpressing strain was at a level similar to WT, indicating that Spx does not affect the transcription of icaADBC by modulating icaR. Spx might directly repress the transcription of icaADBC or indirectly by downregulating a positive regulator of the icaADBC operon, such as SarA (Tormo et al., 2005), SarZ (Wang et al., 2008) or other unidentified factors.

In our previous work, an S. epidermidis clpP mutant strain displayed decreased primary attachment, PIA production and biofilm formation (Wang et al., 2007). This may have been due to the accumulation of Spx in the clpP mutant strain, as Spx has negative effects on primary attachment, PIA production and biofilm formation of S. epidermidis. Interestingly, the transcription of icaADBC was negatively affected by the overexpression of Spx in the clpP mutant strain (Wang et al., 2007). This implies the existence of another substrate of ClpP protease that either interferes with the regulation of icaADBC by Spx or has a positive effect on the transcription of icaADBC that counteract the effect of Spx.

An attempt to construct an S. epidermidis spx mutant strain was unsuccessful, suggesting that the spx gene might be essential in S. epidermidis. It is noteworthy that a previous attempt to delete the spx gene (denoted as yjbD) in Listeria monocytogenes also failed (Borezee et al., 2000), and in S. aureus, the spx mutant strain was only successfully constructed in strain 8325-4 (a σB-deficient strain with a small deletion in rsbU) with a low frequency and reduced size under normal growth conditions (Pamp et al., 2006). Although the author showed that the transcription of spx was at a similar level between a σB-positive WT (SH1000) and the strain 8325-4, this does not guarantee that the phenotypes modulated by Spx would be the same in these two strains. It has been demonstrated that σB affects a wide range of phenotypes in strain 8325-4 (Horsburgh et al., 2002). Whether the defect of σB has interfered with the spx knockout is unknown. Besides, the observation that all 80 tested clinical isolates of S. epidermidis in our study harbor the spx gene also supports this view. The observation that overexpression of Spx has no effect on the stress response indicates that either Spx may not be involved in the general stress response or the concentration of Spx in WT has already exceeded the threshold for bacterial cells to adapt to the selected stress conditions.

In conclusion, we found that Spx has negative effects on primary attachment, PIA production and biofilm formation and is a substrate of ClpP protease in S. epidermidis. Our results suggest that ClpP may positively contribute to the biofilm formation of S. epidermidis by degrading Spx, a negative regulator of biofilm formation. The mechanism of Spx modulating the biofilm formation of S. epidermidis will be further investigated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References

We thank Prof. Peter Zuber (Oregon Health & Science University) for kindly providing the Spx antibody. This study was supported by the Key Project of Chinese National Programs (2008ZX10003-010), National Natural Science Foundation of China 30670108 and J0730860, RFDP20060246037, and the Intramural Research Program of the National Institute of Allergy and Infectious Diseases, The National Institutes of Health, USA.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Authors' contribution
  9. References