RESEARCH ARTICLE
Staphylococcus epidermidis polysaccharide intercellular adhesin activates complement Elizabeth G. Aarag Fredheim1, Hildegunn Norbakken Granslo1,2, Trond Flægstad1,2, Yngve Figenschau3, Holger Rohde4, Irina Sadovskaya5, Tom Eirik Mollnes6,7 & Claus Klingenberg1,2 Paediatric Research Group, Faculty of Health Sciences, University of Tromsø, Tromsø, Norway; 2Department of Paediatrics, University Hospital of North Norway, Tromsø, Norway; 3Department of Medical Biochemistry, University Hospital of North Norway, Tromsø, Norway; 4Institut fu¨r Medizinische Mikrobiologie, Virologie und Hygiene, Universita¨tsklinikum Hamburg-Eppendorf, Hamburg, Germany; 5UMT 08, Universite du Littoral-cote d’Opale, Univ. Lille-Nord de France, Boulogne-sur-Mer, France; 6Research Laboratory, Nordland Hospital, Bodø, Norway; and 7 Department of Medical Biology, University of Tromsø, Tromsø, Norway
IMMUNOLOGY & MEDICAL MICROBIOLOGY
1
Correspondence: Elizabeth G. Aarag Fredheim, Paediatric Research Laboratory, Department of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway. Tel.: +47 77 64 53 53; fax +47 77 64 53 50; e-mail:
[email protected] Received 11 April 2011; revised 28 June 2011; accepted 12 July 2011. Final version published online 8 September 2011. DOI: 10.1111/j.1574-695X.2011.00854.x Editor: Eric Oswald Keywords Staphylococcus epidermidis; biofilm; complement system; PIA.
Abstract Staphylococcus epidermidis is a frequent cause of nosocomial infections. The central virulence factor of S. epidermidis is biofilm formation. Polysaccharide intercellular adhesin (PIA) constitutes the major biofilm matrix-component. PIA and biofilm have been implicated in S. epidermidis evasion of host immune defence. We examined the effects of S. epidermidis PIA on the inflammatory response with focus on complement activation. We used a human whole-blood ex vivo model of infection and compared the effects of a PIApositive S. epidermidis strain (SE1457) and its PIA-negative isogenic mutant (M10). The independent effect of purified PIA on complement activation was investigated. In glucose-rich media, the mutant formed a proteinacious DNArich biofilm, whereas SE1457 formed a thick PIA-biofilm. In biofilm growth, SE1457 induced a stronger activation of the complement system compared with M10. We verified that purified PIA was independently responsible for a strong activation of the complement system. In contrast, M10 induced higher granulocyte activation by expression of CD11b and higher secretion of cytokines. We conclude that PIA has potent pro-inflammatory properties by activating the complement system. However, in a complex balance of the immune response, the decreased activation of granulocytes and cytokines by a PIA biofilm may limit host eradication of S. epidermidis.
Introduction Staphylococcus epidermidis is the most frequent cause of hospital-acquired infections (NNIS 2001; Vuong & Otto, 2002). Most S. epidermidis infections are subacute or chronic and occur mainly in immunocompromised individuals or in patients with indwelling medical devices (Vuong & Otto, 2002). Biofilm formation on the surface of indwelling devices is often involved in the pathogenesis (Rupp et al., 1999a, b; Begun et al., 2007). Biofilms are adherent multicellular bacterial aggregates characterized by bacteria embedded in a self-produced extracellular matrix. In S. epidermidis, several biofilm-matrix components have been implicated as important for pathogenicity, with particular focus on the accumulation-associated FEMS Immunol Med Microbiol 63 (2011) 269–280
protein (Rohde et al., 2007; Macintosh et al., 2009) and a b-(1,6)-linked N-acetylglucosamine polysaccharide termed polysaccharide intercellular adhesin (PIA) (Arciola et al., 2005; Stevens et al., 2009b), synthesized by the gene products of the intercellular adhesion (ica) operon. Mechanisms to evade the host immune response are important in the pathogenesis of S. epidermidis infections. The two main virulence factors that contribute to S. epidermidis immune evasion are biofilm formation and production of a poly-c-DL-glutamic acid (PGA) capsule. PGA may cause resistance to polymorphonuclear neutrophil (PMN) phagocytosis (Kocianova et al., 2005). Biofilm formation and PIA have central roles in the evasion of host immune defence by protecting S. epidermidis from PMN phagocytosis, antimicrobial peptides and deposition ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
270
E.G. Aarag Fredheim et al.
of antibodies and complement (Johnson et al., 1986; Giese et al., 1994; Heinzelmann et al., 1997; Vuong & Otto, 2002; Vuong et al., 2004a; Vuong et al., 2004b; Kristian et al., 2008). The innate inflammatory response to S. epidermidis infections involves the activation of leukocytes, secretion of cytokines and activation of the complement system (Foster, 2005; Otto, 2009). Bacterial activation of the complement system can be initiated via three different pathways (Fig. 1). The classical pathway is activated by C1q, either by direct recognition of bacterial surface structures or by binding to surface-bound antibodies. The lectin pathway is activated through binding of mannosebinding lectin or ficolins to various sugar residues. The alternative pathway is activated spontaneously, but will only proceed if bacterial surfaces are present in close proximity to the activated components. This pathway may also act as an efficient amplification loop for both the classical and lectin pathways. Activation of the three initial pathways converges at C3. All complement pathways may lead to opsonisation of the bacteria for phagocytosis, particularly through C3 bound fragments, and to generation of small peptides contributing to inflammation, e.g. the anaphylatoxins C3a and C5a. Activation of the terminal pathway (C5–C9)
Lectin pathway
Classical pathway
Alternative pathway
MBL/Ficolins MASP2
C1q
C3
C1rs-C1inh*
Materials and methods Bacterial strains and culture conditions
Bb*
C4 C4d*
C3 C3a* C5 C5a* C5b-9 TCC* Fig. 1. A simplified schematic overview of the complement system, adapted from Sprong et al. (2004). Complement activation products investigated in this study are marked by *. MBL, mannose binding lectins; TCC, terminal complement complex; MASP2, mannosebinding protein-associated serine protease 2; C1inh, complement factor 1 esterase inhibitor.
ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
leads to formation of the terminal complement complex (TCC) that is present in two different forms. When activation occurs on the cell membrane, the C5b-9 complex inserts into the membrane. In Gram-negative bacteria, the TCC destroys the cell membrane leading to osmotic lysis, but in Gram-positive bacteria the thick peptidoglycan cell wall is resistant to TCC-induced lysis (Bestebroer et al., 2010). When activation of the terminal pathway occurs in the fluid phase, a soluble form of TCC (sC5b-9) is formed, which can be used as an indicator of complement activation. PMN killing of planktonic S. epidermidis is known to be complement-dependent (Clark & Easmon, 1986). A recent study (Kristian et al., 2008) showed that biofilm formation induced the release of C3a, but also protected S. epidermidis from IgG and complement opsonisation and PMN-dependent killing. Staphylococcus aureus is known to produce immune evasive molecules that also may inhibit the complement system (Foster, 2005; Laarman et al., 2010). However, there is still a lack of data on how PIA and S. epidermidis biofilms interact with the complement system. In this study, we used a human whole-blood model to perform detailed analyses of complement activation and host inflammatory response by comparing a PIA-producing S. epidermidis strain to its PIA-negative isogenic mutant. We also investigated the independent effect of PIA on complement activation.
The potent biofilm producer S. epidermidis 1457 (SE1457) and its isogenic ica-negative Tn917 insertion mutant M10 were used (Mack et al., 1994). S. epidermidis 5 (CIP 109562) (Chokr et al., 2006; Sadovskaya et al., 2006) was used for the PIA preparations. For the inflammatory response experiments with bacteria in planktonic growth, we used a final concentration of ~108 colony forming units (CFU) mL 1 blood as determined by McFarland measurements. To verify the bacterial count, the inoculum was diluted and spread on blood-agar plates, and CFU was determined by counting retrospective of the experiments. The viable count (CFU mL 1 blood) of both strains was determined after incubation in fresh human whole-blood after 10 min, 30 min and 3 h. There were no differences between the two strains in viability (data not shown). For the biofilm experiments, overnight cultures in tryptic soy broth (TSB) were diluted 1 : 100 in TSB with 1% glucose and transferred to segments of polyvinyl chloride (PVC) tubing (length 30 cm, internal diameter 3 mm; Mediplast, Malmø, Sweden). Each segment was closed end-to-end FEMS Immunol Med Microbiol 63 (2011) 269–280
271
Staphylococci and complement
and formed small loops. The loops were rotated slowly in an incubator at 37 °C for 24 h, then emptied and carefully washed one time with sterile phosphate-buffered saline (PBS) before proceeding with the experiments. To ensure that the strains had the same cell density, CFU was determined from the culture emptied from the loops as described above. PIA purification
PIA was prepared from a biofilm extract of a PIAproducing clinical strain, S. epidermidis CIP 109562, as described previously (Sadovskaya et al., 2006; Stevens et al., 2009a). Briefly, the crude extract was treated with DNase I (Sigma) for 2 h at 37 °C in 1mM Tris-HCl, in presence of 1 mM MgCl2, followed by treatment with proteinase K (Sigma, 100 lg mL 1 for 2 h at 37 °C) and H2O2 (1%, 37 °C, 4 h). This treatment leads to degradation of DNA, proteins and inactivation of highly active lipopeptides and lipoproteins (Zahringer et al., 2008; Stevens et al., 2009a). The extract was then subjected to gel-filtration chromatography on a Sephacryl S-300 column. This fractionation affords complete separation of PIA from teichoic acids and lipoteichoid acids, as well as other low molecular weight impurities. The purity of such preparations has previously been investigated and verified by 1H-Nuclear Magnetic Resonance, monosaccharide analysis and gas liquid chromatography (Sadovskaya et al., 2005). A proportion of the PIA preparation was treated with sodium periodate (NaIO4, 50 mM, 20 °C 18 h), dialysed and lyophilized to give NaIO4-treated PIA. NaIO4 aids in degrading PIA, rendering the substance inert (Stevens et al., 2009a). This NaIO4-PIA was included as a control. O-polysaccharide from Proteus mirabilis (kindly provided by Dr. Vinogradov, National Research Council, Ottawa, Canada), which is a high molecular weight zwitterionic polysaccharide (Sidorczyk et al., 2002) similar to PIA, was used as a negative control to ensure that the observed response to purified PIA was not a general response to this type of polysaccharides. Semi-quantitative determination of biofilm formation
Semi-quantitative determination of biofilm formation was done in a microtitre plate assay using crystal violet to stain the biofilm, as described previously (Christensen et al., 1985; Klingenberg et al., 2005; Fredheim et al., 2009). Biofilm formation was induced either by glucose (1%) or NaCl (3%). This biofilm assay was modified for biofilm experiments in the PVC loops. Briefly, 24 h biofilms were formed in PVC tubing loops as described above. The loops were then washed three times with PBS FEMS Immunol Med Microbiol 63 (2011) 269–280
before left to dry. Two millilitres of crystal violet stain was added to each loop and incubated for 10 min before the stain was removed and the loops washed three times with tap water. A 70/30 mix of ethanol/isopropanol was added to dissolve the remaining biofilm-bound crystal violet. Subsequently, the loops were drained and the suspension transferred in 150 lL aliquots to microtitre plates. Optical density (OD) was determined as described previously (Christensen et al., 1985; Klingenberg et al., 2005; Fredheim et al., 2009). All assays were repeated in three parallels. Analysis of biofilm-matrix components
The strains were analysed in detachment assays to determine their main biofilm-matrix components, as described previously (Fredheim et al., 2009). In short, enzymes or chemicals were added to the biofilm, and thereafter the biofilm was analysed to which degree (% based on a comparison with baseline values) these reagents caused detachment of the biofilm. In this study, the biofilms were exposed to DNase, Proteinase K or NaIO4 to determine attachment of DNA, proteins or NaIO4-sensitive polysaccharides respectively (Fredheim et al., 2009). Inflammatory response to bacteria or PIA in human whole-blood
For each experiment, blood from one healthy volunteer was collected into sterile polypropylene tubes (4.5 mL Nunc cryotubes; Nagle Nunc International) containing 50 lg mL 1 lepirudin (Refludan®; Hoechst). Lepirudin, a recombinant hirudin analogue, with highly specific thrombin inhibitory activity, was used as anticoagulant, as it has no effect on complement activation (Mollnes et al., 2002). The blood samples were obtained immediately before each experiment and kept at 37 °C until the inoculum was added. All experiments were repeated six times on separate occasions with different blood donors unless otherwise noted. Experiments were conducted using (1) bacteria in planktonic growth, (2) bacteria in biofilm or (3) purified PIA. Enzyme immunoassays for complement activation products
For analysis of complement activation to planktonic bacteria, 1.8 mL of blood-bacteria suspension (108 CFU mL 1) was incubated in the PVC tubing loops slowly rotating for 30 min at 37 °C. For analysis of complement activation to bacteria in biofilms, 1.8 mL of blood was incubated in loops with preformed 24 h biofilms. EDTA was added to a final concentration of 10 mM at the end of the ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
272
incubation before plasma was separated and stored at 70 °C. Central complement activation products were quantified using ELISA techniques. C1rs-C1 inhibitor complex analyses The C1rs-C1 inhibitor complex was quantified as described previously (Fure et al., 1997). C4d analyses C4d was quantified using the MicroVue C4d EIA kit (Cat. no A008; Quidel Corp., San Diego, CA) according to the manufacturer’s instructions. Bb analyses Bb was quantified using the MicroVue Bb Plus EIA kit (Cat. no A027; Quidel Corp.) according to the manufacturer’s instructions. C3a analyses C3a was quantified using the MicroVue C3a EIA kit (Cat. no A015; Quidel Corp.) according to the manufacturer’s instructions. C5a analyses C5a was quantified using the OptEIA™ Human C5a ELISA kit II (Cat. no 557965; BD Biosciences, San Jose, CA) according to the manufacturer’s instructions. Soluble TCC analyses (C5b-9) Amount of the soluble TCC was tested using the ELISA test kit for TCC (HK328; Hycult biotechnologies, Uden, Netherlands) according to manufacturer’s instructions. Solid phase TCC (C5b-9) analyses TCC deposition on the biofilm in the PVC tubing was primarily determined as described earlier (Mollnes et al., 1999). All incubations were done for 30 min at room temperature in slowly rotating loops unless otherwise noted. After incubation, blood was drained from the tubing and treated for analysis of soluble complement components. The tubing was washed 3–5 times with PBS/ 0.1% Tween (also used for antibody dilutions), before further incubation with aE11 ascites (anti-TCC) diluted 1/2000. Subsequently, the tubing was washed before adding anti-mouse IgG horse radish peroxidase (Cat. no NA9310; GE Healthcare, UK) diluted 1/1000 and incuª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
E.G. Aarag Fredheim et al.
bated before being washed again as described above. Substrate solution (0.15 M sodium-acetate-buffer, pH 4.0, with 0.18 g L 1 2,2′-azino-di-3 ethylbenzothiasoline sulfoniacide and 2.4 9 10–3% H2O2) was added before incubation for 10–30 min. The samples were collected in test tubes, and 100 lL aliquots were transferred to microtitre-plates (Nunclon, Roskilde, Denmark) for determination of OD at 405/490 nm in a VERSAmax microplate reader (Molecular devices Corp., Sunnyvale, CA). Blank control and a control incubated without monoclonal aE11 antibodies were included in all experiments. Complement activation by PIA
Fresh human whole-blood was stimulated with PIA dissolved in PBS as described for planktonic bacteria above. We performed dose–response experiments, and determined 2 lg mL 1 as a suitable concentration for stimulation of complement activation. NaIO4-treated PIA (2 lg mL 1) and O-polysaccharide from P. mirabilis (2 lg mL 1) were included in each experiment. Analyses of plasma for complement activation was done as described above. CD11b expression
Investigation of CD11b expression was primarily performed as described previously (Mollnes et al., 2002). For analysis of response to planktonic bacteria, 0.6 mL of the blood-bacteria suspension was incubated for 10 min at 37 °C in polystyrene tubes (4.5 mL Nunc cryotubes; Nagle Nunc International). For analysis of response to bacteria embedded in biofilms, blood and biofilms were incubated as described for complement activation, but with an incubation time of 10 min. Thereafter, the blood-bacteria suspensions were fixed with 0.5% paraformaldehyde (PFA) in PBS for 5 min, and further stained with IgG2a anti-CD11b APC (Becton Dickenson, San Jose, CA), IgG1 anti-CD45 PerCP (Becton Dickinson) and IgG1 anti-CD64 PE (Dako, Glostrup, Denmark) for 15 min. Red blood cells were lysed using a PFA-based lysis buffer (0.04% PFA pH 7.4, 0.15 M NH4Cl, 0.01 M NaHCO3). Cells were then analysed on a FACSCalibur flow cytometer (Becton Dickenson). Threshold was set for CD45 to exclude debris and erythrocytes. Then monocytes and granulocytes were separately gated in a sidescatter (SSC)-CD64 plot (Fig. 2). CD11b expression was measured as median fluorescence intensity (MFI). Oxidative burst
Oxidative burst was measured using a Burst test kit (Phagoburst kit; ORPEGEN, Heidelberg, Germany) and the samples were treated according to the manufacturer’s FEMS Immunol Med Microbiol 63 (2011) 269–280
273
Staphylococci and complement
(a)
(b)
(c)
(d)
rotated slowly in an incubator at 37 °C for 3 h (Mollnes et al., 2002; Sprong et al., 2004) after which plasma was separated and stored at 70 °C until analysed. Quantification of 13 pro-inflammatory (TNF-a, IL-1b, MIP-1a, MIP-1b, IL-6, IL-8, IL-17, FGF, G-CSF, G-MCSF, IP-10, MCP-1, VEGF) and two anti-inflammatory (IL-1ra and IFN-c) cytokines were analysed using the Bioplex cytokine assay (Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. Ethics statement
The regional committee for medical research ethics approved the collection of blood for the immune response studies (P-REK Nord 86/2006). Informed written consent was obtained from each blood donor.
(e)
(f)
Fig. 2. Flow cytometry analyses of CD11b expression on leukocytes. Flow cytometry analyses from one representative biofilm experiment. The top panels (a, b) display the gating for an unstimulated sample. Leukocytes are separated from debris and erythrocytes with the threshold set for CD45 (a), before monocytes and granulocytes are separately gated on in a SSC-CD64 plot (b). Monocytes are green and granulocytes are red. Histograms c, d displays the CD11b expression on monocytes and granulocytes of an unstimulated sample. Histograms e, f display the CD11b expression on granulocytes stimulated by a PIA-biofilm (e) and granulocytes stimulated by a nonPIA-biofilm (f). Note that for f the Y-axis count is higher.
instructions. Incubations were done as described for CD11b investigations above. Immediately after the 10 min incubation, a lysis-buffer containing PFA was added to the samples, creating lysis of erythrocytes and fixation of leukocytes. Cells were analysed on a FACSAria flow cytometer (Becton Dickenson). Granulocytes and monocytes were identified and gated for in a forwardscatter/SSC plot by their typical patterns (Fig. 2). MFI of the total population and percentage burst cells was calculated for each cell type.
Statistical analyses
For each parallel experiment, the samples were related as we were using blood from the same donor and compared responses with exposure of different bacteria or PIA preparations. However, the immunological responses from each parallel experiment depend on individual properties of the blood. Therefore, we used the paired non-parametric Wilcoxon’s signed ranks to assess level of significance. All analyses were 2-tailed and P < 0.05 was considered to be significant. Statistical analysis was performed with SPSS for Windows, software version 16.0.
Results SE1457 formed a PIA-biofilm and M10 formed a proteinacious, DNA-rich biofilm
SE1457 was confirmed to be a strong biofilm producer with periodate-sensitive PIA as its main biofilm matrixcomponent (Table 1) (Mack et al., 1992). The Tn917 knock-out of the ica-operon abolished the ability of M10 to form biofilms in NaCl-enriched media, but not in glucose-enriched media. The main biofilm-matrix components of M10, formed in glucose-enriched media, were DNA and proteins, and not PIA. This in accordance with a previous study where knock-out of the ica-operon did not abolish biofilm formation, but resulted in a proteinacious biofilm (Hennig et al., 2007).
Multiplex assay for cytokines
The PIA-positive SE1457 induced stronger complement activation than M10, and PIA induced strong complement activation
Incubation with planktonic bacteria or biofilm was done as described for complement activation. The loops were
In planktonic growth, SE1457 induced higher levels of Bb, C3a, C5a and TCC formation compared with M10
FEMS Immunol Med Microbiol 63 (2011) 269–280
ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
274
E.G. Aarag Fredheim et al.
Table 1. Biofilm forming ability and detachment analyses with different growth supplements for the Staphylococcus epidermidis strains in this study Biofilm OD* Strain
NaCl
Glucose
Biofilm OD† Glucose
SE1457 M10
2.5 (0) 0.2 (0.01)
2.5 (0) 1.3 (0.1)
1.7 (0.2) 0.7 (0.1)
Detachment (%) of biofilm*‡ DNase
Proteinase K
NaIO4
16 (14) 33 (15)
0 (0) 37 (1.7)
54 (16) 0 (0)
OD measurements were done at 570 nm wavelength. Detachment values (%) are the mean (SD) from three separate experiments. *In microtitre plate assay. † In PVC tubing loops. ‡ 24 h old biofilms grown in glucose-enriched TSB.
Table 2. Complement activation in human whole-blood incubated for 30 min with planktonic and biofilm grown SE1457 and M10 Planktonic108 CFU
Biofilm 24 h
1 †
C1rs (AU mL ) C4d (lg ml 1) Bb (lg mL 1) C3a (lg mL 1) C5a (ng mL 1) TCC (AU mL 1)
SE1457 Median (range)
M10 Median (range)
Median difference (range)
530 16 49 65 1368 378
345 11 37 37 698 131
175 6 11 30 610 185
(370–1630) (10–32) (34–68) (36–114) (863–1868) (190–488)
(170–1000) (5–20) (14–60) (12–95) (339–815) (53–227)
(30–1070) (4–14) (3–47) (0–50) (214–1431) (87–435)
P*
SE1457 Median (range)
M10 Median (range)
Median difference (range)
P*
0.028 0.028 0.028 0.046 0.028 0.028
195 6 18 16 308 103
200 5 13 10 200 55
14 1 6 4 85 50
0.343 0.028 0.028 0.028 0.028 0.028
(135–370) (2–25) (13–40) (11–42) (235–698) (78–252)
(120–370) (1–20) (9–31) (8–42) (155–324) (44–107)
( 30 to 40) (0–5) (2–15) (0–20) (63–396) (25–145)
*P values using Wilcoxon’s signed ranks test. C1rs-C1 inhibitor complexes.
†
Table 3. Complement activation in a human whole-blood model after 30 min incubation with O-polysaccharide from Proteus mirabilis (negative control), PIA and NaIO4-treated PIA
C1rs§ (AU mL 1) C4d (lg mL 1) Bb (lg mL 1) C3a (lg mL 1) C5a (ng mL 1) TCC (AU mL 1)
Negative control* Median (range)
PIA Median (range)
NaIO4-treated PIA Median (range)
Median difference† (range)
P‡
17 1 3 2 23 2
70 5 8 16 162 31
16 0 2 1 2 2
53 4 6 15 141 30
0.028 0.028 0.028 0.028 0.028 0.028
(12–44) (0–4) (2–4) (1–5) (18–42) (1–3)
(20–170) (1–6) (6–20) (6–44) (47–644) (7–118)
(10–22) (0–0) (1–4) (0–3) (20–43) (1–3)
(5–148) (0–6) (2–17) (4–42) (4–444) (4–116)
A concentration of 2 lg mL 1 was used to stimulate the samples in all experiments. *Negative control; O-polysaccharide from Proteus mirabilis. † Median difference and P value for PIA vs. NaIO4-treated PIA. ‡ P values using Wilcoxon’s signed ranks test. § C1rs-C1 inhibitor complexes.
(Table 2). Similar findings were seen upon exposure to bacteria in biofilm; SE1457 induced higher levels of C4d, C3a, C5a and TCC formation compared with M10 (Table 2). Both SE1457 and M10 induced levels of complement activation significantly higher than the unstimulated control. There were no differences in the deposition of TCC on the biofilms of SE1457 and M10 (data not shown). PIA increased the levels of all complement activation products, including the classical pathway specific C1rs-C1 inhibitor complexes (Table 3). Complement activation ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
experiments including NaIO4-treated PIA and O-polysaccharide from P. mirabilis showed that neither of these controls significantly activated complement. M10 induced a higher upregulation of CD11b expression than SE1457 and both strains induced oxidative burst
Both SE1457 and M10 caused a significant upregulation of CD11b expression compared with the unstimulated FEMS Immunol Med Microbiol 63 (2011) 269–280
275
Staphylococci and complement
(a)
Unstimulated SE1457 (PIA pos.) M10 (PIA neg.)
MFI
controls, both when in planktonic growth and in biofilm growth mode. Upon exposure to bacteria in biofilm, M10 caused a higher upregulation of CD11b expression on granulocytes compared with SE1457 (Fig. 3). Both SE1457 and M10 caused a high percentage of monocytes and granulocytes to display oxidative burst. However, there were no significant differences in oxidative burst caused by SE1457 and M10, neither in planktonic growth nor in the biofilm growth mode (Table 4). M10 induced a higher secretion of cytokines than SE1457; most pronounced in the biofilm growth mode
Discussion In this study, we have shown that PIA, the main matrix component of S. epidermidis biofilms, is a strong activator of the complement system. We used a whole-blood model with lepirudin as anticoagulant, enabling us to study complement activation in detail. In this assay, the PIA-producing SE1457 induces more powerful complement activation than its PIA-negative isogenic mutant. Experiments with purified PIA verified that PIA is an important inducer of complement activation. PIA increased formation of C1rs-C1 inhibitor complexes and C4d, which strongly suggests activation through the classical pathway. A substantial increase in Bb indicates that the alternative pathway amplification loop is also engaged by PIA. We speculate that activation FEMS Immunol Med Microbiol 63 (2011) 269–280
Planktonic
Bioflim
(b)
Unstimulated SE1457 (PIA pos.) M10 (PIA neg.)
MFI
In planktonic growth, stimulation with M10 caused a significant higher secretion of all cytokines, except G-CSF, compared with the unstimulated control. We found no distinctive cytokine response pattern when comparing M10 and SE1457. Stimulation with M10 caused a higher secretion of IL-6 and a lower secretion of IL-8 and MIP-1b, but in general, large variations in cytokine concentrations were observed. For the other 12 cytokines, in planktonic growth, there were no differences in stimulated levels between the two isolates. In contrast, upon exposure to bacteria in biofilm, we found a pronounced difference in cytokine response pattern between M10 and SE1457. Thirteen of 15 cytokines (TNF-a, IL-1b, MIP-1a, MIP-1b, IL-6, IL-8, IL-1ra, IL-17, FGF, G-CSF, G-MCSF, IFN-c and IP-10) showed significantly higher serum concentration when exposed to the PIA-negative biofilm (M10) vs. the PIA-positive biofilm (SE1457). Figure 4 illustrates this characteristic cytokine response pattern represented by two proinflammatory (IL-6, TNF-a) and two anti-inflammatory (IL-1ra and IFN-c) cytokines.
Planktonic
Bioflim
Fig. 3. Induction of CD11b expression on leukocytes by SE1457 and M10 in planktonic or biofilm form. 3a displays CD11b expression on granulocytes and 3b displays CD11b expression on monocytes. The Box plots show median values (solid bar), interquartile range (margins of box) and 5 and 95 percentiles (whiskers). *M10 induces a significantly higher response than SE1457.
of the classical pathway occurs either by direct recognition of PIA by C1q or through opsonisation by antibodies bound to PIA (Sadovskaya et al., 2007). High anti-PIA antibody titres are not only found in patients with prosthesis infections but also in healthy controls (Sadovskaya et al., 2007). Additionally, heterogeneity in the opsonic requirements for phagocytosis of S. epidermidis has been reported. Some strains can be opsonised by either antibodies or complement alone, and some require both ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
276
E.G. Aarag Fredheim et al.
Table 4. Oxidative burst stimulated by incubation for 10 min with Staphylococcus epidermidis 1457 and M10 in biofilm- or planktonic form Planktonic 108 CFU mL
Biofilm 24 h
Granulocytes Granulocytes (MFI) Monocytes Monocytes (MFI)
†
SE1457 Median (range)
M10 Median (range)
Median difference (range)
91 (50–98) 534 (490–4226)
84 (31–97) 721 (160–7313)
51 (19–79) 300 (232–2105)
49 (20–80) 352 (126–2809)
1
P*
SE1457 Median (range)
M10 Median (range)
Median difference (range)
P*
10 ( 20 to 50) 150 ( 3087 to 740)
0.262 0.525
59 (19–74) 377 (188–508)
55 (34–75) 402 (249–805)
9 ( 42 to 14) 39 ( 334 to 81)
0.825 0.438
2 ( 29 to 41) 68 ( 704 to 106)
0.862 0.299
31 (14–61) 257 (174–350)
32 (14–55) 249 (195–361)
8 ( 24 to 21) 21 ( 129 to 89)
0.743 0.985
Unstimulated base line samples displayed median (range) values: Granulocytes 1.25 (0.7–5.5)% and granulocytes MFI 171 (76–614); Monocytes 0.45 (0–4)% and monocytes MFI 140 (82–519). *P values using Wilcoxon‘s signed ranks test. † Percent bursted granulocytes/monocytes of total amount of granulocytes/monocytes analysed.
(Clark & Easmon, 1986; Bronswijk et al., 1989). In a previous study, blocking the classical pathway had little effect on opsonisation (Clark & Easmon, 1986). However, in those experiments heparin was used as an anticoagulant, and experiments were performed with planktonic bacteria. This may have affected their results. In contrast to our findings that PIA-biofilm induced complement activation, we found lower leukocyte activation in response to the PIA-biofilm (SE1457) compared with the non-PIA-biofilm (M10). In our relatively small study, this difference was only significant for CD11b expression on granulocytes, but a clear trend towards decreased CD11b expression on monocytes was also observed. CD11b expression is regarded as a more sensitive marker of leukocyte activation than oxidative burst, which exhibit predominantly an on/off response (Lappegard et al., 2009). In addition, monocytes respond slower than granulocytes (Mollnes et al., 2002). These factors may explain the different findings between CD11b expression and oxidative burst and the differences in activation of granulocytes and monocytes. A reduction in PMN-activation in response to the PIA-biofilm is not surprising. Earlier studies have shown that S. epidermidis embedded in a biofilm is protected from PMN killing and that PIA is an important factor for immune evasion (Vuong et al., 2004a; Vuong et al., 2004b; Cerca et al., 2006; Kristian et al., 2008). Leukocyte activation can occur through different pathways. First, bacteria express pathogen associated molecular patterns (PAMPs) on their surfaces, which are recognised by cellular pattern recognition receptors (PRR, i.e. Toll-like receptors). Second, leukocyte activation may also occur via opsonins deposited on the bacterial surface or directly by anaphylatoxins (e.g. C3a, C4a and C5a). Differences in CD11b expression induced by the PIA vs. the non-PIA-biofilm may be due to less deposition of opsonins on the surface of the PIA-biofilm. PIA may also ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
physically cover and disguise PAMPs on the bacterial surface and thereby inhibit recognition by PRR’s. A recent study showed a reduced deposition of complement and IgG on the biofilm of SE1457 compared with planktonic M10 (Kristian et al., 2008). However, in these experiments, biofilm formation was not induced in M10. Thus, the results were a comparison between two isogenic isolates, but in different state of growth (biofilm vs. planktonic). In our assay we compared complement deposition on SE1457 and M10, both in a biofilm growth mode. We found no difference in complement deposition. Limited surface accessibility may explain these results (Thurlow et al., 2009), but obviously further investigations are needed to explore this in detail. Anaphylatoxins (C3a, C4a and C5a) may also activate leukocytes (Ward, 2004). We found that PIA induced a powerful complement response with high levels of C3a and C5a. Inhibition studies show that complement have a direct role in tissue damage and inflammation and that phagocytosis can be completely inhibited by blocking of the C5a receptor or C3 receptor (CD11b) (Mollnes et al., 2002). Surprisingly, the PIA-positive SE1457 induced a lower CD11b expression than its PIA-negative mutant. This somewhat contradictory result may be explained by Morris et al. (2009) who reported a negative correlation between phagocytosis and serum C3a. Furthermore, in critically ill patients, neutrophil dysfunction was directly associated with activated complement (Morris et al., 2009). We acknowledge that there is still paucity of data on the interplay between complement activation, biofilm and the cellular inflammatory response. Activation of PRR’s stimulates cytokine synthesis and secretion. In contrast, cytokine secretion is not affected by anaphylatoxins in the same manner as the phagocytic function (Morris et al., 2009). The interplay between complement activation and cytokine secretion is unclear FEMS Immunol Med Microbiol 63 (2011) 269–280
277
Staphylococci and complement
(a)
(b)
SE1457 (PIA pos.)
M10 (PIA neg.)
M10 (PIA neg.)
(pg mL–1)
Unstimulated
SE1457 (PIA pos.)
(pg mL–1)
Unstimulated
Planktonic
Bioflim
(c)
Planktonic
Bioflim
(d) Unstimulated
Unstimulated
SE1457 (PIA pos.)
SE1457 (PIA pos.) M10 (PIA neg.)
(pg mL–1)
(pg mL–1)
M10 (PIA neg.)
Planktonic
Bioflim
Planktonic
Bioflim
Fig. 4. The secretion of pro- and anti-inflammatory cytokines induced by SE1457 and M10 in planktonic or biofilm form. The PIA-negative M10 stimulates a significantly higher cytokine secretion. Box plots show median values (solid bar), interquartile range (margins of box) and 5 and 95 percentiles (whiskers). *M10 induces a significantly higher response than SE1457.
with some cytokines being complement dependent, whereas others are complement independent (Mollnes et al., 2002). Cytokines may, however, indirectly contribute to complement activation, e.g. by increasing the expression of anaphylatoxin receptors (Riedemann et al., 2003). We found a strikingly uniform pattern of lower secretion of both pro- and anti-inflammatory cytokines as a response to the biofilm from the PIA-producing SE1457 compared with M10. For some cytokines, however, the quantitative differences were relatively small, and the data FEMS Immunol Med Microbiol 63 (2011) 269–280
should therefore be interpreted with caution regarding biological significance. Among the cytokines investigated, particularly TNF-a and IL-8, are potent PMN activators. The decrease in cytokine secretion, which was seen upon exposure to a PIA-biofilm, may suggest that PIA inhibit cytokine secretion. In vivo IL-6, alone or together with IL-1b and TNF-a, induce the production of acute-phase proteins, i.e. C-reactive protein (CRP). The reduced TNF-a, IL-6 and IL-1b secretion in response to the PIA-positive strain ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
278
may therefore explain the reduced CRP-response that our group previously observed in neonatal infections caused by biofilm positive coagulase-negative staphylococci (Klingenberg et al., 2005). This study has strengths and limitations. When using PIA preparations, there are always concerns of their purity, especially in the context of recent findings that some contaminating compounds, such as glycopeptides, are active in nanogram quantities (Zahringer et al., 2008). For our study, we therefore improved the previously published PIA purification protocol (Stevens et al., 2009a) to achieve an extensive PIA purification. We also used a PIA concentration similar to what have been used in previous studies (Stout et al., 1994; Stevens et al., 2009a). Furthermore, we used proper control substances to exclude that observed effects were due to non-specific binding to polysaccharides instead of specific binding to PIA. The choice of different incubation-times for cellular activators, complement and cytokines was based on limitations of the whole-blood model in which physiological conditions can be maintained for a maximum of 4 h. Circulating leukocytes are ready to rapidly express CD11b and display oxidative burst. In this whole-blood model, oxidative burst declines if samples are incubated longer than 15 min (Mollnes et al., 2002) and early analysis is essential. It is also essential to analyse activation of free circulating complement factors early in the course. The cytokines increases in serum somewhat later and only those responding within 4 h were possible to include in our study. The concentration of IL-10, a central antiinflammatory cytokine, increases later than 12 h after an inflammatory stimulus, and is thus not possible to assess in this model. We used an ex vivo model and attempted to simulate the in vivo environment. We took advantage of a human whole-blood model developed specifically for studying host–pathogen interactions, the cytokine network, the complement system and their multidirectional pathways of inflammation (Mollnes et al., 2002; Lappegard et al., 2009). However, host-immune response is a complex process, not only dependent on the circulating immune cells but also on cell populations throughout the body, which could be investigated only in animal models or clinical studies. In addition, we did not measure anti-PIA antibody titres in the blood from the donors, and we cannot exclude that different levels might have affected our results. Several in vivo studies have shown that PIA-producing strains are persistent colonizers of indwelling medical devices and that biofilm is the major virulence factor in chronic S. epidermidis infections (Rupp et al., 1999a; Rupp et al., 1999b; Rupp et al., 2001; Begun et al., 2007). In this study, we have shown that PIA has potent proinflammatory properties by activating the complement ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
E.G. Aarag Fredheim et al.
system. However, in a complex balance of the immune response, the decreased PMN-activation and a lower cytokine secretion induced by a PIA-biofilm may limit host eradication of S. epidermidis. Further studies, including complement activation blockers (Lappegard et al., 2009) or the use of transgenic animals are needed to elucidate on the mechanism behind PIA induced complement activation, the immunological effects of this activation and to assess potential clinical significance.
Acknowledgements We thank Dorthe Christiansen and Grethe Bergseth at the Research Laboratory at Nordland Hospital, Bodø, Ingvill Harneshaug at the Flow Cytometry Core Facility at the University of Tromsø, Vivian Berg and Goran Kauric at the University Hospital of Northern Norway for excellent technical assistance and Professor Johanna Ericsson Sollid, Department of Medical biology, University of Tromsø for critical reading of the manuscript. We also thank our blood donors for their participation in the project.
References Arciola CR, Gamberini S, Campoccia D, Visai L, Speziale P, Baldassarri L & Montanaro L (2005) A multiplex PCR method for the detection of all five individual genes of ica locus in Staphylococcus epidermidis. A survey on 400 clinical isolates from prosthesis-associated infections. J Biomed Mater Res A 75: 408–413. Begun J, Gaiani JM, Rohde H, Mack D, Calderwood SB, Ausubel FM & Sifri CD (2007) Staphylococcal biofilm exopolysaccharide protects against Caenorhabditis elegans immune defenses. PLoS Pathog 3: e57. Bestebroer J, Aerts PC, Rooijakkers SH, Pandey MK, Kohl J, van Strijp JA & de Haas CJ (2010) Functional basis for complement evasion by staphylococcal superantigen-like 7. Cell Microbiol 12: 1506–1516. Bronswijk V, Verbrugh HA, Heezius HC, Renders NH, Fleer A, van der Meulen J, Oe PL & Verhoef J (1989) Heterogeneity in opsonic requirements of Staphylococcus epidermidis: relative importance of surface hydrophobicity, capsules and slime. Immunology 67: 81–86. Cerca N, Jefferson KK, Oliveira R, Pier GB & Azeredo J (2006) Comparative antibody-mediated phagocytosis of Staphylococcus epidermidis cells grown in a biofilm or in the planktonic state. Infect Immun 74: 4849–4855. Chokr A, Watier D, Eleaume H, Pangon B, Ghnassia JC, Mack D & Jabbouri S (2006) Correlation between biofilm formation and production of polysaccharide intercellular adhesin in clinical isolates of coagulase-negative staphylococci. Int J Med Microbiol 296: 381–388. Christensen GD, Simpson WA & Younger JJ (1985) Adherence of coagulase-negative staphylococci to plastic tissue culture
FEMS Immunol Med Microbiol 63 (2011) 269–280
Staphylococci and complement
plates: a quantitative model for the adherence of staphylococci to medical devices. J Clin Microbiol 22: 996– 1006. Clark LA & Easmon CS (1986) Opsonic requirements of Staphylococcus epidermidis. J Med Microbiol 22: 1–7. Foster TJ (2005) Immune evasion by staphylococci. Nat Rev Microbiol 3: 948–958. Fredheim EGA, Klingenberg C, Rohde H, Frankenberger S, Gaustad P, Flaegstad T & Sollid JE (2009) Biofilm formation by Staphylococcus haemolyticus. J Clin Microbiol 47: 1172–1180. Fure H, Nielsen EW, Hack CE & Mollnes TE (1997) A neoepitope-based enzyme immunoassay for quantification of C1-inhibitor in complex with C1r and C1s. Scand J Immunol 46: 553–557. Giese MJ, Mondino BJ, Glasgow BJ, Sumner HL, Adamu SA, Halabi HP & Chou HJ (1994) Complement system and host defense against staphylococcal endophthalmitis. Invest Ophthalmol Vis Sci 35: 1026–1032. Heinzelmann M, Herzig DO, Swain B, Mercer-Jones MA, Bergamini TM & Polk HC Jr (1997) Phagocytosis and oxidative-burst response of planktonic Staphylococcus epidermidis RP62A and its non-slime-producing variant in human neutrophils. Clin Diagn Lab Immunol 4: 705–710. Hennig S, Nyunt WS & Ziebuhr W (2007) Spontaneous switch to PIA-independent biofilm formation in an ica-positive Staphylococcus epidermidis isolate. Int J Med Microbiol 297: 117–122. Johnson GM, Lee DA, Regelmann WE, Gray ED, Peters G & Quie PG (1986) Interference with granulocyte function by Staphylococcus epidermidis slime. Infect Immun 54: 13– 20. Klingenberg C, Aarag E, Ronnestad A, Sollid JE, Kjeldsen G & Flaegstad T (2005) Coagulase-negative staphylococcal sepsis in neonates: association between antibiotic resistance, biofilm formation and the host inflammatory response. Pediatr Infect Dis J 24: 817–822. Kocianova S, Vuong C, Yao Y, Voyich JM, Fischer ER, DeLeo FR & Otto M (2005) Key role of poly-gamma-DL-glutamic acid in immune evasion and virulence of Staphylococcus epidermidis. J Clin Invest 115: 688–694. Kristian S, Birkenstock T, Sauder U, Mack D, Go¨tz F & Landmann R (2008) Biofilm formation induces C3a release and protects Staphylococcus epidermidis from IgG and complement deposition and from neutrophil dependent killing. J Infect Dis 197: 1028–1035. Laarman A, Milder F, van Strijp J & Rooijakkers S (2010) Complement inhibition by gram-positive pathogens: molecular mechanisms and therapeutic implications. J Mol Med 88: 115–120. Lappegard KT, Christiansen D, Pharo A et al. (2009) Human genetic deficiencies reveal the roles of complement in the inflammatory network: lessons from nature. P Natl Acad Sci USA 106: 15861–15866. Macintosh RL, Brittan JL, Bhattacharya R, Jenkinson HF, Derrick J, Upton M & Handley PS (2009) The terminal A
FEMS Immunol Med Microbiol 63 (2011) 269–280
279
domain of the fibrillar accumulation-associated protein (Aap) of Staphylococcus epidermidis mediates adhesion to human corneocytes. J Bacteriol 191: 7007–7016. Mack D, Siemssen N & Laufs R (1992) Parallel induction by glucose of adherence and a polysaccharide antigen specific for plastic-adherent Staphylococcus epidermidis: evidence for functional relation to intercellular adhesion. Infect Immun 60: 2048–2057. Mack D, Nedelmann M, Krokotsch A, Schwarzkopf A, Heesemann J & Laufs R (1994) Characterization of transposon mutants of biofilm-producing Staphylococcus epidermidis impaired in the accumulative phase of biofilm production: genetic identification of a hexosaminecontaining polysaccharide intercellular adhesin. Infect Immun 62: 3244–3253. Mollnes TE, Videm V, Christiansen D, Bergseth G, Riesenfeld J & Hovig T (1999) Platelet compatibility of an artificial surface modified with functionally active heparin. Thromb Haemost 82: 1132–1136. Mollnes TE, Brekke OL, Fung M, Fure H, Christiansen D, Bergseth G, Videm V, Lappegard KT, Kohl J & Lambris JD (2002) Essential role of the C5a receptor in E coli-induced oxidative burst and phagocytosis revealed by a novel lepirudin-based human whole blood model of inflammation. Blood 100: 1869–1877. Morris AC, Kefala K, Wilkinson TS et al. (2009) C5a mediates peripheral blood neutrophil dysfunction in critically ill patients. Am J Respir Crit Care Med 180: 19–28. NNIS (2001) National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992– June 2001, issued August 2001. Am J Infect Control 29: 404–421. Otto M (2009) Staphylococcus epidermidis – the ‘accidental’ pathogen. Nat Rev Microbiol 7: 555–567. Riedemann NC, Neff TA, Guo RF, Bernacki KD, Laudes IJ, Sarma JV , Lambris JD & Ward PA (2003) Protective effects of IL-6 blockade in sepsis are linked to reduced C5a receptor expression. J Immunol 170: 503–507. Rohde H, Burandt EC, Siemssen N et al. (2007) Polysaccharide intercellular adhesin or protein factors in biofilm accumulation of Staphylococcus epidermidis and Staphylococcus aureus isolated from prosthetic hip and knee joint infections. Biomaterials 28: 1711–1720. Rupp ME, Ulphani JS, Fey PD, Bartscht K & Mack D (1999a) Characterization of the importance of polysaccharide intercellular adhesin/hemagglutinin of Staphylococcus epidermidis in the pathogenesis of biomaterial-based infection in a mouse foreign body infection model. Infect Immun 67: 2627–2632. Rupp ME, Ulphani JS, Fey PD & Mack D (1999b) Characterization of Staphylococcus epidermidis polysaccharide intercellular adhesin/hemagglutinin in the pathogenesis of intravascular catheter-associated infection in a rat model. Infect Immun 67: 2656–2659. Rupp ME, Fey PD, Heilmann C & Gotz F (2001) Characterization of the importance of Staphylococcus
ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
280
epidermidis autolysin and polysaccharide intercellular adhesin in the pathogenesis of intravascular catheter-associated infection in a rat model. J Infect Dis 183: 1038–1042. Sadovskaya I, Vinogradov E, Flahaut S, Kogan G & Jabbouri S (2005) Extracellular carbohydrate-containing polymers of a model biofilm-producing strain, Staphylococcus epidermidis RP62A. Infect Immun 73: 3007–3017. Sadovskaya I, Chaignon P, Kogan G, Chokr A, Vinogradov E & Jabbouri S (2006) Carbohydrate-containing components of biofilms produced in vitro by some staphylococcal strains related to orthopaedic prosthesis infections. FEMS Immunol Med Microbiol 47: 75–82. Sadovskaya I, Faure S, Watier D, Leterme D, Chokr A, Girard J , Migaud H & Jabbouri S (2007) Potential use of polyN-acetyl-beta-(1,6)-glucosamine as an antigen for diagnosis of staphylococcal orthopedic-prosthesis-related infections. Clin Vaccine Immunol 14: 1609–1615. Sidorczyk Z, Zych K, Toukach FV, Arbatsky NP, Zablotni A, Shashkov AS & Knirel YA (2002) Structure of the O-polysaccharide and classification of Proteus mirabilis strain G1 in Proteus serogroup O3. Eur J Biochem 269: 1406–1412. Sprong T, Moller AS, Bjerre A, Wedege E, Kierulf P, van der Meer JWM et al. (2004) Complement activation and complement-dependent inflammation by Neisseria meningitidis are independent of lipopolysaccharide. Infect Immun 72: 3344–3349. Stevens N, Sadovskaya I, Jabbouri S, Sattar T, O’Gara JP, Humphreys H & Grewal J (2009a) Staphyococcus epidermidis polysaccharide intercellular adhesin induces IL-8 expression in human astrocytes via a mechanism involving TLR2. Cell Microbiol 11: 421–432.
ª 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
E.G. Aarag Fredheim et al.
Stevens NT, Greene CM, O’Gara JP & Humphreys H (2009b) Biofilm characteristics of Staphylococcus epidermidis isolates associated with device-related meningitis. J Med Microbiol 58: 855–862. Stout RD, Li Y, Miller AR & Lambe DW Jr (1994) Staphylococcal glycocalyx activates macrophage prostaglandin E2 and interleukin 1 production and modulates tumor necrosis factor alpha and nitric oxide production. Infect Immun 62: 4160–4166. Thurlow LR, Thomas VC, Fleming SD & Hancock LE (2009) Enterococcus faecalis capsular polysaccharide serotypes C and D and their contributions to host innate immune evasion. Infect Immun 77: 5551–5557. Vuong C & Otto M (2002) Staphylococcus epidermidis infections. Microbes Infect 4: 481–489. Vuong C, Voyich JM, Fischer ER, Braughton KR, Whitney AR, DeLeo FR & Otto M (2004a) Polysaccharide intercellular adhesin (PIA) protects Staphylococcus epidermidis against major components of the human innate immune system. Cell Microbiol 6: 269–275. Vuong C, Kocianova S, Voyich JM, Yao Y, Fischer ER, DeLeo FR & Otto M (2004b) A crucial role for exopolysaccharide modification in bacterial biofilm formation, immune evasion, and virulence. J Biol Chem 279: 54881–54886. Ward PA (2004) The dark side of C5a in sepsis. Nat Rev Immunol 4: 133–142. Zahringer U, Lindner B, Inamura S, Heine H & Alexander C (2008) TLR2 – promiscous or spesific? A critical reevaluation of a receptor expressing apparent broad specificity. Immunobiology 213: 205–224.
FEMS Immunol Med Microbiol 63 (2011) 269–280
