Surface proteins that promote adherence of Staphylococcus aureus to human desquamated nasal epithelial cells

Background The natural habitat of Staphylococcus aureus is the moist squamous epithelium in the anterior nares. About 20% of the human population carry S. aureus permanently in their noses and another 60% of individuals are intermittent carriers. The ability of S. aureus to colonize the nasal epithelium is in part due to expression of surface proteins clumping factor B (ClfB) and the iron-regulated surface determinant A (IsdA), which promote adhesion to desquamated epithelial cells present in the anterior part of the nasal vestibule. S. aureus strain Newman defective in IsdA and ClfB exhibited reduced but not completely defective adherence to squamous cells in indicating that other cell surface components might also contribute. Results Surface proteins IsdA, ClfB, and the serine-aspartic acid repeat proteins SdrC, SdrD and SdrE were investigated to determine their contribution to the adherence of S. aureus to desquamated nasal epithelial cells. This was achieved by expression of ClfB, IsdA, SdrC, SdrD and SdrE on the surface of the surrogate Gram-positive host Lactococcus lactis and by isolating mutants of S. aureus Newman defective in one or more factor. The level of adherence of strains to squamous cells isolated from the nares of volunteers was measured. Results consistently showed that ClfB, IsdA, SdrC and SdrD each contributed to the ability of S. aureus to adhere to squamous cells. A mutant lacking all four proteins was completely defective in adherence. Conclusion The ability of S. aureus Newman to adhere to desquamated nasal epithelial cells is multifactorial and involves SdrD and SdrC as well as ClfB and IsdA.


Background
Staphylococcus aureus causes community-acquired and nosocomial infections. Although multiple body sites such as the axilla and the perineum can be colonized, the most frequent site of carriage is the moist squamous epithelium of the anterior nares. About 20% of the human population carry S. aureus permanently in their noses and another 60% of individuals are intermittent carriers [1]. The reasons for the variable tropism of S. aureus for the human nares are unclear. Higher carriage rates occur in white people [2], in men [2], in certain age groups [3] and in dialysis [4], diabetic [5] and AIDS patients [6]. Infection rates are higher in carriers than in non-carriers and invasive disease is often caused by a patients' carried strain [7]. However when infected, carriers suffer significantly fewer fatalities, suggesting that carriage stimulates a degree of protective immunity [8].
It has been suggested that the ability of S. aureus to adhere to human desquamated nasal epithelial cells is an impor-tant factor in determining nasal colonization [9]. Both clumping factor B (ClfB) and iron regulated surface determinant protein A (IsdA) are expressed on the bacterial cell surface and promote adhesion to desquamated epithelial cells in vitro and colonization of the nares of rodents in in vivo models [10,11], and in the case of ClfB [12], humans. Protection against colonization was elicited by active immunization of rodents with recombinant ClfB or IsdA, and in the case of ClfB, with a function-blocking monoclonal antibody.
The surface protein SasG can also promote adhesion to desquamated nasal epithelial cells in vitro [13,14]. However SasG is not expressed by many strains including Newman [14]. A mutant of S. aureus strain Newman defective in IsdA and ClfB had reduced adherence to squamous cells but still bound at about 40% of the level of the wildtype [10]. Since SasG is not expressed by strain Newman [14], other cell surface components are likely to be involved. It had been noted that the serine-aspartic acid repeat proteins SdrC and SdrD can also promote adhesion to squamous cells [11], although this has never been examined in detail. In this paper the role of surface proteins IsdA, ClfB, SdrC and SdrD in adhesion to desquamated cell has been systematically analyzed in order to determine the contribution of each under the same conditions. This was achieved by expression of ClfB, IsdA, SdrC and SdrD on the surface of the Gram-positive surrogate host Lactococcus lactis and by testing single and combined mutants of S. aureus Newman.

Expression of IsdA, ClfB, SdrC and SdrD by the surrogate host L. lactis
To determine the ability of the IsdA, ClfB, SdrC, SdrD and SdrE proteins to promote adhesion to human desquamated nasal epithelial cells, L. lactis cells expressing each protein [9] were incubated with squamous cells from the anterior nares of healthy volunteers. L. lactis containing the empty vector pKS80 adhered poorly ( Figure 1). L. lactis expressing SdrE was not significantly different to L. lactis carrying pKS80 (P = 0.2055; Figure 1) indicating that this protein cannot promote adhesion to squamous cells. In contrast, a significant increase in adherence to squamous cells was observed when L. lactis cells expressed SdrC, SdrD, ClfB or IsdA (P values of 0.0339, SdrC; P = 0.0003, SdrD; P = 0.0396, ClfB and P = 0.0178, IsdA; Figure 1) showing that each of these proteins can promote adhesion when expressed on the surface of a Gram positive coccus. It was shown previously that ClfA expressed by L. lactis did not promote adhesion [15].

Adherence of S. aureus mutants to desquamated nasal epithelial cells
In order to investigate the role of surface proteins in promoting adherence of S. aureus to desquamated nasal epithelial cells a set of isogenic mutants was constructed and compared. Strain Newman defective in clfA was used as the starting point in the strain construction but this mutation had no bearing on adhesion since ClfA is known not to promote adhesion to squamous cells [9]. Each strain was examined by Western immunoblotting in order to show that the relevant proteins were missing in the mutants and that the remaining proteins were expressed at the same level as in the wild type. Newman clfA grown to exponential phase in TSB expressed ClfB, SdrC and SdrE but not SdrD ( Figure 2). Since bacteria were grown in TSB they did not express Isd proteins. Introduction of the multicopy shuttle plasmid pCU1 bearing the clfB, sdrC or sdrE genes resulted in expression of proteins at levels equivalent to or higher than the wild-type. In the case of SdrD expression was not seen in the wild-type strain and was only detected when the pCU1sdrD + plasmid was present ( Figure 2C). This may be due to amplification of low level expression under these growth conditions due to a gene dosage affect by a multicopy plasmid.
Bacteria were also grown to stationary phase in RPMI. The wild-type strain expressed ClfB, IsdA, SdrD and SdrE, but not SdrC at levels that were detectable by Western immu-Adherence of L. lactis expressing different surface proteins to desquamated nasal epithelial cells Figure 1 Adherence of L. lactis expressing different surface proteins to desquamated nasal epithelial cells. L. lactis (pKS80), L. lactis (pKS80clfB + ), L. lactis (pKS80sdrC + ), L. lactis (pKS80sdrD + ), L. lactis (pKS80sdrE + ) and L. lactis (pKS80isdA + ) grown to stationary phase were tested for their ability to bind to human desquamated epithelial cells. Counts represent the number of bacterial cells adhering to 100 squamous cells. Results are expressed as the mean of triplicate experiments +/-standard deviations.  Figure 3). The Sdr proteins were detected with antibodies that recognized the conserved B domains (Figure 3C) and specific anti-A domain antibodies (not shown). Complementation of the mutant strain lacking these surface proteins with pCU1clfB + , pCU isdAB + , pCU1sdrD + or pCU1sdrE + resulted in restoration of expression of the appropriate protein at levels similar to (IsdA) or higher than wild-type (ClfB, SdrD, SdrE). In the case of pCU1sdrC + low level expression was achieved.

kDa 98 kDa
Anti-ClfB N e w m a n N e w m a n c lf not promote adhesion to squamous cells (Figure 1), it is likely that the decrease observed by disrupting the sdrCDE genes is due to the loss of SdrC.
In order to determine the role of IsdA in adherence, mutants were grown to stationary phase in the iron limited medium RPMI and tested for adhesion to squamous cells. Newman wild-type and Newman clfA adhered at similar levels of ca 1300 bacteria per 100 squamous cells N e w m a n N e w m a n c lf A -

Anti-Sdr B r epeats
clfB mutations decreased adherence further ( Figure 4B, P = 0.0352, 0.0135 and 0.0183, respectively compared to Newman clfA). Finally when a mutant lacking ClfA, ClfB, IsdA and SdrCDE was tested, only 200 bacteria adhered per 100 squamous cells. Of the Sdr proteins only SdrD and SdrE were expressed by Newman growing in RPMI ( Figure 3) (IsdA and ClfB are also expressed under these conditions [12,15]) and as SdrE does not promote adhesion it can be concluded that the decrease associated with the deletion of sdrCDE was due to the loss of the SdrD protein.
In conclusion, these results are consistent with the data obtained with L. lactis and demonstrate a similar role for ClfB, IsdA, SdrC and SdrD in adhesion to squamous cells.

Complementation
To confirm the roles of surface proteins in S. aureus deduced from the analysis of mutants, a strain of Newman that was defective in all four adherent surface proteins (in addition to ClfA and SdrE) was complemented by introducing multicopy shuttle plasmids expressing ClfB, SdrC, SdrD, SdrE, IsdAIsdB and IsdB. When bacteria were grown to exponential phase in TSB, ClfB, SdrC and SdrD each promoted increased adhesion to squamous cells above the background level ( Figure 5A; P = 0.0554, ClfB; P = 0.0282, SdrC; P = 0.0449, SdrD; P = 0.8803, SdrE; P = 0.533, IsdA). The differences were statistically significant for SdrC and SdrD, but not for ClfB. Expression of SdrE did not promote adhesion which is consistent with results described above. The Isd proteins were not expressed in TSB-grown bacteria.
When the strains were grown under iron restricted conditions in RPMI, complementation with ClfB, IsdA, SdrC or SdrD each promoted adhesion ( Figure 5B, P = 0.029, ClfB; P = 0.0536, SdrC; P = 0.0908, SdrD; P = 0.0384, IsdA). The conclusion about IsdA was drawn by comparing the level of adhesion promoted by the plasmid expressing both IsdA and IsdB with that expressing IsdB alone. Attempts to express IsdA alone in pCU1 were unsuccessful. These results were statistically significant except for those involving SdrC and SdrD. Expression of SdrE did not promote adhesion ( Figure 5B). These results confirm that ClfB, IsdA, SdrC and SdrD are all important in adherence of S.
Adherence of Newman mutants to desquamated nasal epithelial cells

Discussion
S. aureus is a commensal of the moist squamous epithelium of the anterior nares of a significant proportion of the population. Colonization is a known risk factor for the development of staphylococcal infections in the community and hospital. The causes of intermittent and persistent carriage are believed to be different. Persistent carriers are often colonised by a single strain of S. aureus over a long period of time, while intermittent carriers tend to carry different strains for briefer time periods [16,17]. Persistent carriers also carry a higher load of bacteria in the nares than intermittent carriers [18,19]. When volunteers were decolonized and were then inoculated with a mixture of S. aureus strains non-carriers eliminated the bacteria, whereas persistent carriers selected their original S. aureus colonizing strain from the mixture [20]. These results indicate that certain host characteristics determine carriage status in particular individuals. This view is sup-ported by the fact that certain age or ethnic groups seem to be predisposed to carriage [2,3]. One determinant of varying patterns of nasal carriage may be differing expression levels of ligands for S. aureus on the surface of desquamated nasal epithelial cells.
In this study we used three donors to provide the desquamated nasal epithelial cells for adhesion experiments. They were selected because their cells supported a consistent level of adhesion. It has been noted that cells from different donors can provide widely variable levels of adhesion [21]. The reason for this is not known. One possibility is different levels of expression of the ligands responsible for adherence promoted by one or more of the S. aureus surface proteins. It is imperative to perform a detailed comparative study of the ability of the surface proteins described here to support adhesion of bacteria to squamous cells from donors who are persistent carriers and those who are non-carriers. This could contribute to the knowledge of the contribution of host factors to carriage.

B) RPMI
Surface proteins ClfB and IsdA have previously been shown to promote adhesion to squamous epithelial cells [9,15] and are required for colonization of the nares of rodents [11,15]. Both ClfB and IsdA have been shown to bind to proteins present in the envelope of cornified squamous epithelial cells. IsdA and ClfB both bind to cytokeratin 10 and loricrin [22] (Clarke, S. Walsh, E. J. Andre, G. Dufrene, Y. Foster, T. J. Foster, S. J. manuscript submitted). Loricrin accounts for 70 -85% of the cornified envelope [23][24][25]. It is possible that differences in the level of expression of these proteins could contribute to the variation in carriage of S. aureus in the nares.
To investigate the contribution of each of five surface proteins (IsdA, ClfB, SdrC, SdrD and SdrE) to squamous cell adhesion, the proteins were expressed from the surrogate host L. lactis. Expression of IsdA, ClfB, SdrC and SdrD each resulted in increased adherence. Gene disruption and complementation experiments in S. aureus also showed a role for IsdA, ClfB, SdrC and SdrD in adhesion. SdrE did not promote adhesion by either L. lactis or S. aureus. Schaffer et al 2006 investigated whether SdrC or SdrD had a role in colonization of the nares in a mouse model. Mutants defective in SdrC or SdrD colonized mice to the same extent as the wild-type indicating that these proteins do not play a role colonization of the nares of mice [11]. However, this does not necessarily mean that SdrC and SdrD have no role to play in colonization of the human nares.
Adherence to desquamated epithelial cells from the anterior nares is clearly multifactorial. When expression of IsdA, ClfB, SdrC and SdrD was disrupted in strain Newman the level of adherence was reduced to background. This does not support a role for wall teichoic acid (WTA) in promoting adhesion to squamous cells [21,26]. SasG did not play a role in adherence of Newman to squamous cells because this protein was not expressed detectably by this strain despite the intact sasG gene being present [14]. SasG might play a role in clinical isolates where expression occurs at higher levels.
It has been reported that WTA plays a prominent part in nasal colonization of the cotton rat model [26]. The authors also demonstrated that teichoic acid promoted bacterial adhesion to normal epithelial cells. However the WTA apparently does not contribute to bacterial adhesion to desquamated nasal cells epithelial cells [21]. This is consistent with the data reported here which indicates that only surface proteins are required for adhesion to squames. A multiple mutant defective in ClfB, SdrC, SdrD and IsdA did not adhere. If WTA contributed to adherence the multiple mutant would still have bound above background levels. Thus colonization of the cotton rat requires both WTA [26] and surface proteins [15] albeit at different stages in the process [21] and in different parts of the nares.

Conclusion
Eradication of carriage of S. aureus has been shown to reduce infection rates in dialysis, diabetic and AIDS patients [4][5][6]. Vaccination with IsdA and ClfB was effective in reducing S. aureus carriage in animal models [11,15]. It has been suggested that immune responses in part determine the ability of humans to carry S. aureus in the nares. This study has confirmed the role of ClfB and IsdA in adhesion to desquamated nasal epithelial cells and has revealed important roles for SdrC and SdrD. Vaccination against two or more of these surface proteins could provide significant reduction in carriage and could potentially reduce the rate of infection and dissemination.
Complementation analysis was performed by transferring into DU6023 clfA5 isdA clfB::Em r ΔsdrCDE::Tc r plasmid pCU1 containing the full length structural genes for S. aureus surface proteins as follows: pCU1sdrC + , pCU1sdrD + , pCU1sdrE + , pCU1clfB + [31], pCU1isdAisdB + and pCU1isdB + . The plasmids were maintained by selecting resistance to chloramphenicol (10 μg ml -1 ). In each case the gene was amplified from genomic DNA of strain Newman to include the promoter region and the down-stream transcription terminator. In the case of isd proteins both the closely linked isdA and isdB genes and their cognate promoters were cloned together. The primers are described below.

Molecular techniques
Standard procedures were used [34]. Restriction enzymes and ligase (New England Biolabs or Roche) were used according to manufacturer's protocol, as was Pfu DNA polymerase (Roche). Oligonucleotides were purchased from Sigma-Genosys.
Primers pCU1isdBF (5'-CAGCTGCAGCCTATGTCATAGA-TATTTCATAATC-'3) and pCU1isdBR (5'-CAGGAGCTCA-GAGATTCTAAACGTATTCGTAAG-'3) which incorporate a 5' PstI and 3' XbaI site, respectively, were used to amplify the isdB coding sequence including the upstream promoter and Fur consensus sequence from strain Newman genomic DNA. The isdB coding sequence is located 203 bp downstream of the isdA coding sequence on the S. aureus chromosome. Primers pCU1isdAF (5'-CAGCTGCAGA-CATAATCCTCCTTTTTATGATTG-'3) and pCU1isdBR (5'-CAGGAGCTCAGAGATTCTAAACGTATTCGTAAG-'3) were used to amplify the isdA and isdB coding sequence including the upstream promoter and Fur consensus sequence of both genes. The 2.3 kb isdB and 3.6 kb isdAB coding sequences were digested with PstI and XbaI and cloned between the PstI and XbaI sites of plasmid pCU1. Plasmids pCU1isdB + and pCU1isdAB + were sequenced and screened by restriction mapping.

Primers
FpKisdA (5'-CGCTGATCAAACATTATT-TAAACAGTAAGTATC-'3) and RpKisdA (5'-CGCTGAT-CATTATTTAGATTCTTTTCTTTTGA-'3) which incorporate a 5' and a 3' BclI site, respectively, were used to amplify the isdA coding sequence from genomic DNA. The PCR product was digested with BclI and cloned into BclI digested pKS80. This resulted in the open reading frame of isdA being fused to the ATG codon of the expression cassette to optimize translation and created the plasmid pKS80isdA + . The plasmid was sequenced, screened by restriction mapping and electroporated into competent L. lactis strain MG1363.
swabbing of the anterior nares. One donor was a carrier of S. aureus while the other two were not. After washing in PBS, cells were adjusted to 1 × 10 5 cell ml -1 . Bacterial cells were washed with PBS and adjusted to 1 × 10 9 cells ml -1 . 100 μl volumes of bacterial and epithelial cells were mixed and incubated at 37°C for 1 h with occasional shaking before being captured on 12 μm isopore polycarbonate filters, washed with PBS, fixed and stained with 5% (w/v) crystal violet. The filters were mounted onto glass slides. The number of bacteria per 100 squames was counted using light microscopy.

Statisical Analysis
Statistical analyses were determined by the Student t-test, using the online GraphPad software. Differences were considered significant if p values were less than 0.05.
Publish with Bio Med Central and every scientist can read your work free of charge