Staphylococcus aureus Strains That Express Serotype 5 or Serotype 8 Capsular Polysaccharides Differ in Virulence
http://www.100md.com
感染与免疫杂志 2005年第6期
Channing Laboratory, Brigham and Women's Hospital
Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts 02115
ABSTRACT
Most isolates of Staphylococcus aureus produce a serotype 5 (CP5) or 8 (CP8) capsular polysaccharide. To investigate whether CP5 and CP8 differ in their biological properties, we created isogenic mutants of S. aureus Reynolds that expressed CP5, CP8, or no capsule. Biochemical analyses of CP5 and CP8 purified from the isogenic S. aureus strains were consistent with published structures. The degree of O acetylation of each polysaccharide was similar, but CP5 showed a greater degree of N acetylation. Mice challenged with the CP5+ strain showed a significantly higher bacteremia level than mice challenged with the CP8+ strain. Similarly, the CP5+ strain survived preferentially in the bloodstream and kidneys of infected mice challenged with a mixed inoculum containing both strains. The enhanced virulence of the CP5+ strain in vivo correlated with its greater resistance to in vitro killing in whole mouse blood. Likewise, in vitro opsonophagocytic killing assays with human neutrophils and sera revealed greater survival of the Reynolds (CP5) strain, even though the kinetics of opsonization by C3b and iC3b was similar for both the CP5+ and CP8+ strains. Electron micrographs demonstrated C3 molecules on the cell wall beneath the capsule layer for both serotype 5 and 8 strains. Purified CP5 and CP8 stimulated a modest oxidative burst in human neutrophils but failed to activate the alternative complement pathway. These results indicate that CP5 and CP8 differ in a number of biological properties, and these differences likely contribute to the relative virulence of serotype 5 and 8 S. aureus in vivo.
INTRODUCTION
Staphylococcus aureus is a major bacterial pathogen that causes a wide spectrum of clinical infections, ranging from localized soft-tissue infections to life-threatening bacteremia and endocarditis (25). Many virulence factors contribute to the pathogenesis of staphylococcal infections, including surface-associated adhesins and secreted exoproteins and toxins (35). Like many invasive bacterial pathogens, S. aureus produces a capsular polysaccharide (CP) that enhances its resistance to clearance by host innate immune defenses. Most clinical isolates of S. aureus are encapsulated, and serotype 5 and 8 strains predominate (2, 11, 40). The type 5 (CP5) and type 8 (CP8) capsular polysaccharides have similar trisaccharide repeating units comprised of N-acetyl mannosaminuronic acid, N-acetyl L-fucosamine, and N-acetyl D-fucosamine (9, 28, 43). CP5 and CP8 are serologically distinct, and this can be attributed to differences in the linkages between the sugars and in the sites of O acetylation.
Previous studies have correlated S. aureus capsule production with resistance to in vitro phagocytic uptake and killing (13, 41). Human neutrophils phagocytose capsule-negative mutants in the presence of nonimmune serum with complement activity, whereas serotype 5 isolates require both capsule-specific antibodies and complement for optimal opsonophagocytic killing (4, 41). Nilsson et al. (29) reported that peritoneal macrophages from mice phagocytosed significantly greater numbers of a CP5-negative mutant compared to the parental strain Reynolds. Once phagocytosed, the CP5-positive strain survived intracellularly to a greater extent than the mutant strain. Cunnion et al. (7) compared opsonization of isogenic S. aureus strains and demonstrated that the CP5-positive strain bound 42% less serum complement (C3) than the acapsular mutant.
Serotype 5 S. aureus strains have also been shown to be more virulent than acapsular mutants in animal models of staphylococcal infection. The CP5-positive strain Reynolds produced higher bacteremia levels in mice and resisted host clearance to a greater extent than two capsule-deficient mutants (41). Strain Reynolds was more virulent than an acapsular mutant in rodent models of renal infection or abscess formation (33, 42). Mice inoculated with the serotype 5 S. aureus strain developed more frequent and severe arthritis, demonstrated greater weight loss, and showed a higher mortality rate than mice infected with capsule-negative mutants (29).
Studies documenting the role of CP8 in virulence were lacking until recently, when Luong and Lee (26) showed that a CP8-overproducing mutant was more resistant to in vitro opsonophagocytic killing by human neutrophils than the parental strain Becker. Likewise, the CP8-overproducing strain persisted longer in the bloodstream, liver, and spleen of infected mice than strain Becker. These results were the first to show that CP8 promoted S. aureus virulence in an animal model of infection, although it was necessary to create a mutant that produced excess CP8 in order to see the effect.
Our data indicate that the serotype 8 strain Becker is less virulent for mice and rats than the serotype 5 strain Reynolds (1, 22, 23, 41 and unpublished observations), and efforts to enhance the virulence of strain Becker by cultivation under conditions that enhance capsule production failed (23). Our preliminary results with <10 clinical isolates suggest that serotype 5 S. aureus isolates produce more CP and are more virulent for mice than type 8 S. aureus isolates. However, these differences in virulence cannot be attributed to capsule type only, since the strains examined were not isogenic. The purpose of this study was to construct isogenic mutants of S. aureus that expressed CP5, CP8, or no capsule. Creation of these strains allowed us to investigate the relative contribution of capsule type to staphylococcal virulence in the genetic background of a virulent S. aureus strain.
MATERIALS AND METHODS
Bacterial strains and growth conditions. The bacterial strains and plasmids used in this study are listed in Table 1. For genetic manipulations Escherichia coli was cultured on Luria-Bertani medium, whereas S. aureus strains were cultured in tryptic soy broth (TSB) or agar (TSA). When appropriate, the culture medium was supplemented with chloramphenicol (Cm) at 10 μg/ml, erythromycin (Em) at 10 μg/ml, or ampicillin at 100 μg/ml. For all other experiments, S. aureus was cultivated for 24 h at 37°C on Columbia agar (Difco Laboratories, Detroit, Mich.) supplemented with 2% NaCl.
DNA manipulations. S. aureus chromosomal DNA was isolated using the Wizard Genomic Purification kit (Promega, Madison, Wis.), and plasmid DNA was purified with a QIAprep spin miniprep kit 250 (QIAGEN, Inc., Valencia, Calif.). Standard molecular cloning procedures were followed as detailed by Sambrook et al. (36). Restriction endonucleases and other DNA modification enzymes were obtained from Invitrogen Corp. (Carlsbad, Calif.) or New England Biolabs, Inc. (Beverly, Mass.). Electroporation and transduction of S. aureus strains were performed as described previously (15, 18).
Construction of strain Reynolds (CP–). An allelic replacement method was utilized to create an acapsular mutant of the serotype 5 strain Reynolds. The strategy was to replace the chromosomal copy of the serotype 5-specific genes (cap5HIJK) with an ermB cassette encoding Em resistance. A 2.2-kb EcoRI fragment carrying cap5FG was subcloned from pJCL24 (4) into pCL10 (39), a temperature-sensitive E. coli-S. aureus shuttle vector (Table 1). Orientation of the cloned fragment in the recombinant plasmid (pCap5FG) was determined through asymmetrical restriction digests with EcoRV, BamHI, and BstXI. The S. aureus ermB gene was subcloned from pErmB (17) into pCap5FG to create pCap5FGermB. The cap5LM gene fragment from pJCL24 (24) was amplified by PCR (25 cycles of 94°C for 30 s, 50°C for 30 s, and 68°C for 2 min, with a final extension at 72°C for 7 min) with primers cap5L-f (5'-GCGATCTAGATGACGCTTCACACGATTAC-3') and cap5M-r (5'-ACCATTCAGACCTTCTTTTCCATAAACTGCC-3'), which carry XbaI sites (underlined). The amplified fragment was cloned into pCap5FGermB to create pAP1.2 (Table 1), and the cloned cap5LM fragment was verified by sequencing. pAP1.2 was electroporated into the restriction-negative S. aureus strain RN4220 (Table 1) and then transduced into strain Reynolds, in both cases with selection for Cm-resistant (Cmr) colonies at 30°C. Reynolds(pAP1.2) was incubated at 42°C for 24 h in the presence of Cm to select for strains with plasmid integration into the chromosome by homologous recombination. A single integrant colony was confirmed by PCR and then passaged three times at 30°C in medium without antibiotics. This process allowed the integrated plasmid to be excised from the chromosome by a single crossover at the duplicated region created during plasmid integration. Depending on the site of crossover during excision, the mutation site on the insert of pAP1.2 would either be left on the chromosome, thus generating the desired mutant (Emr Cms), or be lost on the cured plasmid, thus regenerating a revertant wild-type strain (Ems Cms). A revertant (named strain JL810) was valuable for evaluating the effect of genetic manipulation on CP production. The genotypes of an Emr Cms excisant, named Reynolds (CP–), and the revertant JL810 were verified by PCR and by Southern hybridization using the ECL direct nucleic acid labeling and detection system (Amersham Biosciences, Piscataway, N.J.).
Construction of strain Reynolds (CP8). Plasmid pCL7657, generously provided by Chia Y. Lee, University of Arkansas for Medical Sciences, carries the entire cap8 locus from strain Becker (38). A 12.4-kb SalI fragment of pCL7657 harboring cap8C through cap8 M was ligated into pCL10 to create recombinant plasmid pDK1. The plasmid was electroporated into S. aureus RN4220 and then transduced into strain Reynolds (CP–) as described above, selecting for Cmr colonies at 30°C. pDK1 was integrated into the chromosome and excised as described above, except that the excisants were screened for susceptibility to Cm and Em. A single excisant, Reynolds (CP8), was further characterized by PCR and Southern blots.
Phenotypic characterization of S. aureus strains. Isogenic recombinant S. aureus strains were phenotypically evaluated by API (BioMerieux Inc., Durham, N.C.) and by hemolysis on sheep blood agar plates. Staphylococcal growth rates were determined by monitoring the optical density at 650 nm of TSB cultures of each strain.
Serotype 5 and 8 capsular antisera were prepared by immunization of rabbits with heat- or formalin-killed suspensions of S. aureus strain Reynolds (serotype 5) or PS80 (serotype 8). To render the antiserum CP5 or CP8 specific, it was absorbed with the acapsular S. aureus strain Wood 46 (protein A deficient) and trypsinized suspensions of acapsular mutants JL243 and JL252 (3). AltaStaph, kindly provided by A. Fattom (Nabi, Inc., Rockville, Md.), is a human immunoglobulin G (IgG) product derived from the pooled serum of individuals immunized with CP5 and CP8 conjugated to Pseudomonas aeruginosa exotoxoid A (8). All antibody preparations were heat inactivated for 30 min at 56°C prior to use. The capsular phenotypes of the parental and recombinant S. aureus strains were assessed by immunodiffusion and colony immunoblotting (20). We utilized an enzyme-linked immunosorbent assay (ELISA) inhibition method (23) for quantitation of S. aureus cell-associated CP.
Immunoelectron microscopy was used to visualize CP on the S. aureus strains. Suspensions of S. aureus cells were incubated in 10 ml of 0.1 M sodium phosphate buffer (pH 8.0) containing 1 mg/ml trypsin for 1 h at 37°C. The bacterial suspensions were washed with phosphate-buffered saline (PBS) before room temperature incubation for 2 h with capsular antisera. After being washed, the bacteria were fixed and processed for electron microscopy as described previously (19).
Analysis of purified CP5 and CP8. CP5 and CP8 were purified from Reynolds (CP5) and Reynolds (CP8) as described previously (42). Teichoic acid and peptidoglycan were purified as described by Peterson et al. (32). Chemical assays for O acetylation, protein, nucleic acid, and phosphorus were performed as described previously (21). The polysaccharides were treated with alkali (4) to remove O-acetyl groups. 1H nuclear magnetic resonance spectra were recorded on a Bruker 600 spectrophotometer in D2O at 60°C. The chemical shifts were given on the scale relative to the internal standard. Free amino groups on the polysaccharides were also detected using a fluorescamine assay described previously (42). Purified CP5, CP8, peptidoglycan, and teichoic acid were tested for activation of the alternative complement pathway by incubating each polymer (final concentration, 100 μg/ml) in 50% normal human serum (NHS) in gelatin-Veronal buffer (GVB) containing 8 mM EGTA and 5 mM Mg2+ for 15 min at room temperature. The mixtures were then diluted and assayed for C3a production with an ELISA kit from Quidel.
Mouse model of bacteremia and renal abscess formation. Female CD-1 mice, 7 to 8 weeks old, were obtained from Harlan Sprague Dawley, Inc. (Indianapolis, Ind.) and given food and water ad libitum. Groups of 6 to 13 mice were challenged by the intraperitoneal (i.p.) route with a 0.5-ml inoculum containing either 5 x 107 CFU or 9 x 107 CFU S. aureus. Separate groups of three or four mice were bled at different time points after inoculation. Bacteremia levels were determined by quantitative counts performed in duplicate on neat or diluted blood samples, and the results were expressed as log CFU S. aureus per ml blood.
Animals were coinfected with an equal number (5 x 107 CFU) of Reynolds (CP5) and Reynolds (CP8) in a series of competition experiments. The mixed bacterial suspension was injected i.p., and separate groups of mice were bled for quantitative culture at 20, 120, or 200 min after inoculation. The capsule phenotype of colonies from each blood culture was determined by immunoblotting (20) with CP5- or CP8-specific antiserum. Bacterial concentrations in the blood of infected mice were compared by the Welch modification of the unpaired Student t test (InStat 2.0; GraphPad Software, Inc., San Diego, Calif.). All mice were euthanized after 5 days, and their kidneys were examined grossly for the presence of abscesses. Both kidneys from animals with renal abscesses were excised, weighed, and homogenized. Dilutions of the homogenates were plated quantitatively, and CP5- or CP8-positive colonies were determined by colony immunoblotting as described above. The numbers of animals with abscesses containing either Reynolds (CP5) or Reynolds (CP8) were compared by Fisher's exact test.
S. aureus killing assays. Heparinized blood was collected by tail vein puncture of male CD-1 mice. Whole-blood killing assays were performed by mixing 200 μl of pooled mouse blood with 100 μl of an S. aureus suspension to achieve a final concentration of 105 CFU/ml. The samples were incubated on a rotator at 37°C, and 10-μl aliquots were removed for quantitative culture at time zero and after 60 min. Control tubes contained 200 μl mouse blood and 100 μl PBS. The percent killing was defined as the reduction in CFU/ml after 60 min compared with that at time zero.
The opsonophagocytic killing assay with human polymorphonuclear leukocytes (PMNs) was performed essentially as described by Xu et al. (48) with the following modifications. Normal human serum (NHS) was collected and pooled from three healthy adult volunteers. Three different pools of NHS were absorbed for 1 h on ice with all three isogenic S. aureus strains: Reynolds (CP5), Reynolds (CP8), and Reynolds (CP–). Certain NHS pools were supplemented with 10 mM EDTA prior to absorption to further limit complement activation. Following absorption, the NHS was centrifuged, filtered (0.45 μm), and stored at –80°C for not more than 3 months. Immediately prior to use, the serum was thawed and supplemented, when necessary, with 7.5 mM MgCl2 and 2.5 mM CaCl2 to restore free ions. The opsonophagocytic killing assay was performed in polypropylene tubes containing 1 x 106 PMNs, 1 x 106 CFU S. aureus, and absorbed serum (2.5 to 10% final concentration) in a total volume of 500 μl minimal essential medium (Invitrogen). Control samples contained S. aureus and PMNs (without serum) or S. aureus alone. The samples were rotated end over end (12 rpm) for 2 h at 37°C, followed by sonication (Probe Sonic Dismembranator 60; Fisher Scientific, Pittsburgh, Pa.) for 5 s at 4 W to minimize antibody-mediated bacterial agglutination. Sample dilutions were made in sterile deionized water, and bacterial killing was estimated by plating the diluted samples in duplicate on TSA. The percent killing was defined as the reduction in CFU/ml after 2 h compared with that at time zero. Results were compared by one-way analysis of variance (ANOVA) and the Tukey-Kramer multiple comparisons test (InStat).
To determine the number of viable intracellular bacteria associated with the PMNs in 10% NHS, assay samples were taken at 20 and 60 min and incubated with lysostaphin (10 μg/ml) for 20 min at 37°C to kill the extracellular bacteria. The PMNs were centrifuged at 2,700 x g for 5 min at 4°C, lysed in sterile water, and plated in duplicate on TSA. To assess the fate of intracellular S. aureus, lysostaphin was added to certain samples 20 or 60 min after mixing the bacteria and PMNs. The PMNs were incubated for an additional 2 h at 37°C, at which time numbers of viable intracellular bacteria were estimated as described above.
Flow cytometry. We measured antibody binding to the isogenic S. aureus strains by incubation of 108 CFU S. aureus with 100 μl of 10% unabsorbed or absorbed NHS for 30 min at ambient temperature. The samples were washed in PBS containing 1% bovine serum albumin (PBS-BSA) before incubation with 20 μg/ml of protein G-Alexa Fluor 488 conjugate (Molecular Probes, Eugene, Oreg.). The bacteria were fixed overnight at 4°C in 2% paraformaldehyde, washed, and stained with 10 μg/ml of hexidium iodide (Molecular Probes). Fluorescent S. aureus cells were gated on a Beckman-Coulter Epics flow cytometer with a 488-nm laser, and 10,000 events were recorded. IgG binding to S. aureus was measured in the FL1 channel. In separate experiments, S. aureus capsule expression was measured by flow cytometry after incubating the bacteria with AltaStaph antibodies [absorbed with Reynolds (CP–)] followed by the protein G-Alexa Fluor 488 conjugate. Protein A-mediated binding of IgG to S. aureus was measured by flow cytometry after incubation of the bacteria with Alexa Fluor 488 conjugated to rabbit IgG.
The respiratory burst response of human PMNs to S. aureus or purified polysaccharides was evaluated by flow cytometry. Approximtely106 PMNs were incubated with 0.8 μg/ml dihydrorhodamine (DHR123; Sigma, St. Louis, Mo.) in minimal essential medium-BSA for 20 min at 37°C. This dye diffuses inside the leukocytes and is colorless in its reduced form. When the dye is oxidized, it fluoresces and can be detected by flow cytometry. The PMN stimulants that were tested included S. aureus (106 CFU) opsonized with 10% absorbed NHS, 1 μM N-formylmethionyl-leucine-phenylalanine (fMLP; Sigma), and 100 μg/ml of either purified CP5, CP8, teichoic acid, or polygalacturonic acid (Sigma). Each stimulant was incubated with the PMNs for 30 min. The PMNs were then washed twice in PBS-BSA, filtered through a cell-strainer cap, and analyzed by flow cytometry. To account for the percentage of PMNs responding to the stimulus as well as the magnitude of the response, the results were expressed as relative PMN activation: percent activated PMNs x median fluorescence index (MFI) of the activated PMNs. To normalize the values obtained within different experiments, this value was divided by the relative PMN activation of the unstimulated PMNs.
To detect C3b and iC3b deposition, 108 CFU S. aureus were incubated at 37°C with 10% absorbed NHS for 1, 3, 5, 10, 20, or 30 min. The reaction was stopped on ice by the addition of ice-cold GVB containing 10 mM EDTA, and the samples were washed three times with PBS-BSA. Mouse monoclonal antibodies to human C3b [C-5G; IgG1()] or iC3b [G-3E; IgG2b()] (27) (kindly supplied by Sunita Gulati, Boston Medical Center) were incubated with each sample for 30 min at ambient temperature. Alexa Fluor 488-conjugated goat anti-mouse IgG (Fab)2 fragments (Molecular Probes) were added to a final concentration of 2 μg/ml and incubated for 30 min at room temperature. Samples were washed, fixed, stained, and analyzed as described above.
Visualization of C3 on S. aureus by electron microscopy. S. aureus cells were incubated with 40% pooled human serum in GVB containing 0.5 mM Mg2+ and 0.15 mM Ca2+ for 30 min at 37°C. The bacteria were washed three times with PBS containing 0.5% fish skin gelatin and incubated with murine monoclonal antibodies to human C3c (Quidel Corp., San Diego, Calif.). The staphylococci were washed before incubation with goat anti-mouse IgG conjugated to 12-nm colloidal gold particles (Jackson ImmunoResearch, West Grove, Pa.). Samples were fixed overnight at 4°C in 0.1 M sodium cacodylate buffer, pH 7.4, containing 2% paraformaldehyde and 2.5% glutaraldehyde. To visualize the capsule and preserve its integrity during the dehydration step (47), the fixed bacteria were incubated for 2 h with rabbit CP-specific antibodies. The preparations were postfixed in 1% osmium tetroxide, dehydrated with graded alcohols, and infiltrated with resin and propylene oxide (1:1) prior to embedding in Taab resin. Thin sections (90 nm) were cut with a Reichert ultramicrotome and transferred to copper grids. The sections were stained with 1% uranyl acetate in acetone followed by 0.2% lead citrate and examined on a JEOL-1200EX transmission electron microscope.
RESULTS
Construction and characterization of isogenic mutants. The genes encoding S. aureus capsule biosynthesis, transport, and assembly are clustered within the cap5 and cap8 loci. These loci are allelic and are comprised of 16 genes named cap5(8)A through cap5(8)P. Whereas the predicted gene products of cap5(8)A through cap5(8)G and cap5(8)L through cap5(8)P are virtually identical, the four open reading frames located in the central region of each locus [cap5(8)HIJK] are type specific (37). We replaced cap5HIJK in strain Reynolds (CP5) with an ermB cassette to create an isogenic mutant designated Reynolds (CP–) that was acapsular by colony immunoblotting and immunodiffusion. A CP5-positive revertant designated JL810 was shown to be identical to Reynolds (CP5) by Southern blotting and PCR analyses. To create the Reynolds (CP8) strain, we replaced the ermB gene within the cap locus of Reynolds (CP–) with the cap8HIJK genes from strain Becker. An Ems excisant named Reynolds (CP8) was analyzed by Southern blotting and PCR to confirm the authenticity of its genotype.
Strains Reynolds (CP5), Reynolds (CP–), Reynolds (CP8), and revertant JL810 showed identical colony morphology, pigmentation, and hemolysin production on blood agar plates. Likewise, the API profile of each of the strains was identical, and their in vitro growth rates in TSB were the same. PCR amplification of the pCL10-encoded cat gene from the integrants, but not from strain Reynolds (CP5), JL810, Reynolds (CP8), or Reynolds (CP–), insured complete loss of the recombinant plasmids in the excisants. CP production by the parental and recombinant strains was measured by an ELISA inhibition assay. As shown in Table 2, Reynolds (CP5), revertant JL810, and Reynolds (CP8) all produced similar quantities of CP. The inclusion of revertant JL810 in these experiments verified that the allelic replacement mutagenesis method did not have deleterious effects on CP production. Reynolds (CP–) had no detectable CP expression (<0.7 μg/1010 CFU), and Reynolds (CP8) produced approximately twice as much CP8 as strain Becker. Transmission electron micrographs of the isogenic mutants (Fig. 1) confirmed that Reynolds (CP5) and Reynolds (CP8) produced abundant surface-associated CP, whereas Reynolds (CP–) had no detectable capsule.
Comparative analysis of purified CP5 and CP8. We purified and analyzed the CPs produced by Reynolds (CP5) and Reynolds (CP8). Both polysaccharide polymers are of high-molecular-weight with Kav values of 0.03 on a Sephacryl S-300 column. (Kav represents the fraction of the stationary gel volume which is available for diffusion of a given solute species.) Purified CP5 and CP8 were both 25% O acetylated, and each polymer had <0.5% protein, nucleic acid, and phosphorus by chemical analyses. Tzianabos et al. (42) reported that purified CP8 showed strain-dependent differences in the degree of N acetylation of the fucosamine residues. Nuclear magnetic resonance analysis revealed that the CP5 from strain Reynolds was 98% N acetylated, whereas the CP8 purified from Reynolds (CP8) was 89% N acetylated. The presence of free amino groups on CP8 was confirmed using a fluorescamine assay, which demonstrated a 1.8-fold higher signal for CP8 compared with CP5. The putative gene product responsible for de-N acetylating the CP is outside of the capsule locus, and experiments to characterize its activity are ongoing in our laboratory.
Effect of capsule type on S. aureus virulence. To compare the influence of capsule type on bacterial virulence, mice were challenged i.p. with sublethal doses of Reynolds (CP5), Reynolds (CP8), or Reynolds (CP–). As shown in Table 3, animals challenged with the encapsulated strains showed a significantly (P < 0.001) higher bacteremia level 20 and 180 min after inoculation compared with animals infected with the acapsular mutant. The rapid clearance of the acapsular mutant that we observed is consistent with the results of Thakker et al., who evaluated the virulence of a capsule-deficient transposon mutant of S. aureus (41). Mice challenged with Reynolds (CP5) and Reynolds (CP8) showed similar bacteremia levels 20 min after bacterial challenge. However, by 180 min after inoculation, the blood concentrations of Reynolds (CP5) were significantly (P = 0.01) higher than those of Reynolds (CP8). Of note, there was little change in the quantitative blood culture results for animals injected with Reynolds (CP8) or Reynolds (CP–) between 20 and 180 min, whereas Reynolds (CP5) increased in concentration during the same time interval (Table 3).
To further confirm these differences in virulence and to minimize the effects of variability among individual mice, 24 additional animals were challenged i.p. with a mixed suspension containing an equal number (5 x 107 CFU) of Reynolds (CP5) and Reynolds (CP8). Separate groups of eight animals were bled after 20, 120, and 200 min, and quantitative blood cultures were performed. The percentages of CP5+ and CP8+ colonies recovered from each blood sample were assessed by immunoblotting. As shown in Fig. 2, the percentage of Reynolds (CP5) cells in the population increased over time, accounting for 75% of the recovered organisms by 200 min after bacterial challenge. Mean ± standard error of the mean (SEM) blood levels of Reynolds (CP5) were significantly higher than Reynolds (CP8) after 200 min: 3.39 ± 0.13 versus 2.89 ± 0.16 log CFU S. aureus/ml, respectively (P = 0.03). Competition assays confirmed that there was no preferential growth of either strain in vitro (data not shown).
The mice coinfected with Reynolds (CP5) and Reynolds (CP8) were euthanized after 5 days, and their kidneys were examined grossly for the presence of renal abscesses. Both kidneys from nine animals with visible abscesses were homogenized and plated quantitatively. Colonies recovered from the kidney homogenates of each animal were assessed for CP production by colony immunoblotting. All of the abscessed kidneys yielded a pure culture of either CP5- or CP8-positive colonies. Seven of the nine mice yielded CP5-positive colonies, and the two remaining mice had CP8-positive colonies (P = 0.0278 by Fisher's exact test). Even at the lowest dilution (10–1) plated, none of the kidney homogenates yielded a mixed culture.
In vitro killing assays. To determine whether there were differences in resistance to phagocytosis attributable to CP5 or CP8 production, we measured killing of each S. aureus strain in whole mouse blood. Blood pooled from nave mice killed 59% of the Reynolds (CP–) inoculum after 60 min. In contrast, only 7% and 31% of the Reynolds (CP5) and Reynolds (CP8) organisms, respectively, were killed in mouse blood within 60 min. The difference in killing between the two encapsulated strains was significant (P = 0.03; Student t test) and mirrored the difference in blood clearance that we observed in mice challenged with the same bacterial strains.
Neutrophils represent only 20% of total leukocytes in the peripheral circulation of nave mice (5), so we chose to perform opsonophagocytic killing assays using isolated human neutrophils incubated with pooled NHS absorbed with the three isogenic S. aureus strains. As shown in Fig. 3, 75% of the Reynolds (CP–) inoculum was effectively opsonized for phagocytic killing by PMNs in NHS concentrations ranging from 2.5 to 10%. In contrast, <35% of the Reynolds (CP5) inoculum was killed in 5% or 10% NHS [P < 0.001 compared with Reynolds (CP–)], and no killing was observed in 2.5% serum. Approximately 60% of Reynolds (CP8) was killed by PMNs in 10% NHS [P < 0.05 compared with killing of Reynolds (CP5)]. Although the encapsulated strains were not killed by PMNs in 2.5% serum, the addition of 4 μg/ml of AltaStaph capsular antibodies to the phagocytic assay resulted in killing of both CP-producing strains (Fig. 3). In the absence of a complement source, anti-CP5/CP8 antibodies were not opsonic, and heat inactivation of the NHS abolished its opsonic activity (data not shown). Similar results were obtained with three different pools of NHS.
To ensure that the observed differences in phagocytic killing between Reynolds (CP5) and Reynolds (CP8) in 10% absorbed NHS did not reflect differences in residual CP5- or CP8-specific antibodies, we used flow cytometry to detect IgG antibody binding to the three isogenic strains. The MFI values for Reynolds (CP5), Reynolds (CP8), and Reynolds (CP–) incubated in 10% unabsorbed serum were 100, 108, and 133, respectively. The MFI values for the same S. aureus strains incubated with the absorbed serum were 7.2, 9.0, and 12.4, respectively, which indicates low but comparable antibody binding to each of the isolates. We confirmed that the measured IgG binding was not mediated by protein A by evaluating protein A expression on Reynolds using Alexa Fluor 488 conjugated to rabbit IgG. Although protein A was present on staphylococci cultivated in broth, none was detected on staphylococci cultivated on Columbia salt agar plates, regardless of the capsule phenotype of the strain (data not shown).
We considered the possibility that the observed differences in phagocytic killing of Reynolds (CP5) and Reynolds (CP8) might relate to postphagocytic events. Intracellular bacteria were enumerated after the addition of lysostaphin to the phagocytic assay samples to kill the extracellular bacteria. The PMNs were then washed and lysed for quantitation of intracellular viable bacteria. As shown in Fig. 4, there were significant differences among the three isogenic strains (P = 0.0081 by ANOVA) when the numbers of intracellular bacteria were measured at the 20-min time point. Uptake of Reynolds (CP–) cells was more rapid than that of the encapsulated strains. Approximately 1% of the Reynolds (CP–) inoculum remained viable within the PMNs after 20 min, and its intracellular concentration was significantly (P < 0.05) higher than that of the two encapsulated strains. Whereas maximal recovery of Reynolds (CP–) was achieved by 20 min, the numbers of intracellular encapsulated S. aureus isolates continued to increase during the 60-min incubation period. The intracellular concentration of the three strains was also significantly different at 60 min (P = 0.0025). The CFU/ml of Reynolds (CP–) was significantly higher than that of Reynolds (CP5) (P < 0.05) at 60 min but not compared with Reynolds (CP8). The differences between the CP5- and CP8-producing strains were not significant at either time point, although the trend for greater uptake of the CP8+ strains was evident.
Additional phagocytic assay samples were further incubated for 2 h at 37°C after lysostaphin treatment to determine the fate of the intracellular bacteria. The CFU/ml did not change markedly during the 2-h incubation, and no differences in intracellular survival were observed between the three isogenic strains (data not shown).
Respiratory burst of PMNs in response to S. aureus. We measured the PMN oxidative burst in response to Reynolds (CP–), Reynolds (CP5), or Reynolds (CP8) to determine whether this response correlated with the results of the opsonophagocytic killing assays. Each strain was incubated with human PMNs in 10% absorbed NHS for 30 min. As shown in Fig. 5A, the relative PMN activation response to Reynolds (CP–) was significantly greater (P < 0.001) than that elicited by the CP-positive strains. The respiratory burst responses of PMNs to Reynolds (CP5) and Reynolds (CP8) were similar. In the absence of NHS, no PMN oxidative burst was observed in response to any of the staphylococcal strains (not shown). Purified CP5, CP8, and teichoic acid (100 μg/ml) also stimulated a PMN respiratory burst (Fig. 5B), but the response was modest compared with that of fMLP, which showed a relative PMN activation score of 161 (not shown on the graph). The control polymer polygalacturonic acid did not elicit a PMN oxidative burst (Fig. 5B).
The influence of CP on complement deposition. Cunnion et al. showed that strain Reynolds (CP5) cells bound less C3 than an acapsular mutant (7). To determine the influence of capsule serotype on opsonization by complement, we evaluated the kinetics of C3b and iC3b deposition on Reynolds (CP5) and Reynolds (CP8) cells by flow cytometry. As shown in Fig. 6, complement deposition on the bacterial cells in 10% NHS was rapid. C3b was detectable on both Reynolds (CP5) and Reynolds (CP8) after 1 min, and 50% of the bacteria were opsonized after 5 min (Fig. 6A). All the bacteria showed bound C3b on their surface 20 min after addition of 10% NHS to the bacterial suspension. As expected, deposition of iC3b on the S. aureus cells was delayed with respect to C3b, with surface-bound iC3b detectable only after 3 min (Fig. 6B). The kinetics of iC3b and C3b deposition for strain Reynolds (CP5) and Reynolds (CP8) were similar. Strain Reynolds (CP5) showed a consistent 10% reduction in C3b and iC3b binding between the 3- and 10-min time points, but it is unlikely that the difference between the two encapsulated strains is biologically relevant.
Purified S. aureus peptidoglycan and teichoic acid activate complement, although both polymers are less active than isolated staphylococcal cell walls (44-46). In our hands, 100 μg/ml of either peptidoglycan or teichoic acid activated the alternative complement pathway as measured by C3a production in 50% NHS (data not shown). In contrast, neither purified CP5 nor CP8 at the same concentration activated the alternative complement pathway in NHS. Although we can demonstrate that Reynolds (CP5) and Reynolds (CP8) bind less C3 than the acapsular mutant, the addition of 100 μg of purified CP5 or CP8 to Reynolds (CP–) cells did not affect complement deposition on the acapsular strain (not shown).
Localization of C3 on encapsulated S. aureus. Because neither purified CP5 nor CP8 activated the complement pathway, surface-associated CP might mask complement components deposited on the cell wall, thereby preventing recognition by receptors on the PMN membrane. Reynolds (CP5) and Reynolds (CP8) were incubated with absorbed NHS, and C3 deposition on the bacterial surface was detected by immunogold labeling. The CP was visualized and its integrity maintained by incubating the fixed bacterial cells with capsular antibodies prior to dehydration and embedding (47). As shown in Fig. 7, C3 molecules deposited on the bacterial cell wall beneath the capsule layer of both Reynolds (CP5) and Reynolds (CP8) cells. Flow cytometric analysis of capsule expression using AltaStaph antibodies revealed that approximately 85% of Reynolds (CP5) and Reynolds (CP8) expressed CP when grown on Columbia salt agar. Bacterial cells within the population that expressed scant capsule showed greater immunogold labeling than cells with abundant capsule (Fig. 7).
DISCUSSION
Capsules are common features of invasive bacterial pathogens. Numerous serotyping studies of S. aureus strains from diverse human sources have revealed that serotype 5 and 8 strains account for 25% and 50%, respectively, of clinical isolates (30). S. aureus Reynolds and Becker are the prototype serotype 5 and 8 strains, respectively, and both were isolated from patients with bacteremia (14). Strain Reynolds expresses abundant CP5, resists phagocytic killing in the absence of specific capsular antibodies, and is virulent for mice (41). S. aureus Becker produces less CP, is poorly virulent for mice, and is opsonized for phagocytic killing by complement alone (22, 23 and unpublished observations). These differences cannot be wholly attributed to capsule type, since the genetic backgrounds of these two strains are different. In order to define the differences in virulence between S. aureus strains expressing type 5 and 8 capsules, we performed allelic exchange experiments to create isogenic S. aureus strains that differed only in capsule type, and we chose the more virulent strain Reynolds as the parent strain. In the process of constructing Reynolds (CP8), we created an acapsular mutant Reynolds (CP–) that served as a useful isogenic control.
The results of our animal studies revealed that Reynolds (CP5) and Reynolds (CP8) caused a higher and more persistent bacteremia in mice than did the capsule-negative mutant (Table 3). This finding underscores the importance of the capsule in blood-borne staphylococcal infections and confirms the virulence of the serotype 5 strain Reynolds. This is the first report to document the greater virulence of a CP8-producing S. aureus strain compared with an isogenic acapsular mutant. A comparison of the isogenic CP5- and CP8-producing Reynolds strains in the bacteremia model revealed that Reynolds (CP5) elicited a higher and more persistent bacteremia than Reynolds (CP8). The bacteremia level in these mice is influenced by the ability of the microbes to avoid phagocytic uptake in the peritoneal cavity and their capacity to escape to and persist in the intravascular space. The greater virulence of Reynolds (CP5) was further substantiated in a mixed infection model, where it demonstrated preferential survival in the mouse blood (Fig. 2) and kidneys compared with Reynolds (CP8). None of the mice with renal abscesses yielded a mixed bacterial population, which suggests that the abscesses originated from a single microorganism seeding the kidney. Because renal abscesses formed in only nine of the infected mice, additional studies are needed to verify this supposition and to address the hypothesis that CP5-producing S. aureus strains are more nephritogenic than CP8-producing strains. We found no correlation between the ratio of the two strains recovered from the blood of individual animals given the mixed inoculum and isolation of either a CP5- or CP8-producing S. aureus strain from the kidneys after 4 days. Our phenotypic characterization of the strains indicated that Reynolds (CP5) and Reynolds (CP8) had similar growth rates in vitro.
Consistent with the mouse bacteremia data, we observed that Reynolds (CP5) was more resistant to in vitro killing in mouse blood or by human neutrophils than Reynolds (CP8) (Fig. 3). These findings suggest that CP5 production may preferentially promote bacterial survival in the bloodstream by interfering with bacterial uptake and killing by PMNs. To determine whether differences in phagocytosis by or persistence within phagocytes might contribute to the enhanced virulence of strain Reynolds (CP5), we quantified the staphylococci taken up by neutrophils after 20- and 60-min incubation periods (Fig. 4). Recovery of intracellular Reynolds (CP–) from neutrophils was greater than that of the two encapsulated strains, consistent with its increased rate of uptake and killing by phagocytic cells. In addition, there was a trend toward enhanced uptake of Reynolds (CP8) cells by PMNs compared to Reynolds (CP5), but the differences were not statistically significant. Between 0.1 and 1% of the staphylococcal inoculum survived within the phagocytes, and this percentage did not change markedly for the three S. aureus strains following an additional 2-h incubation. Gresham et al. demonstrated that staphylococci can survive within mouse PMNs and that these intracellular bacteria can contribute to persistent infection (10).
The respiratory bursts of human PMNs in response to Reynolds (CP5) and Reynolds (CP8) in NHS were similar and significantly lower than the response elicited by Reynolds (CP–) (Fig. 5A). This result is consistent with the PMNs having a greater intracellular concentration of Reynolds (CP–) compared with the encapsulated strains. The diminished response of PMNs to the encapsulated strains is unlikely to be an inhibitory effect of the capsules, since purified CP5 and CP8 stimulated a modest PMN respiratory burst similar to that elicited by teichoic acid (Fig. 5B). The control polymer polygalacturonic acid did not stimulate a respiratory burst.
Opsonization by complement is essential for phagocytosis of S. aureus by human PMNs, as none of the strains was killed by PMNs in heat-inactivated serum. Cunnion et al. demonstrated that mice depleted of complement by cobra venom factor were more susceptible to staphylococcal infection than control mice (7). Acapsular strains of S. aureus bind more C3 than encapsulated strains, although the exact mechanism for this is unclear (5, 35). To investigate the influence of CP5 and CP8 production on complement deposition, we used flow cytometry to demonstrate the kinetics and amounts of C3b and iC3b deposition on Reynolds (CP5) and Reynolds (CP8). Although the serotype 8 strain showed slightly more C3b and iC3b deposition within the initial 5 min of incubation, these subtle differences were no longer apparent by 20 min (Fig. 6). At that point, both C3b and iC3b were detected on the bacterial surface, consistent with the report by Cunnion et al. (6).
Both C3b and iC3b promote uptake of staphylococci by interacting with the CR1 (CD35) and CR3 (CD11b/CD18) receptors, respectively, on the surface of PMNs. It has been proposed that S. aureus CP renders the organism resistant to opsonophagocytic killing by masking complement molecules bound to the microbial cell wall, thus preventing ligand recognition by the phagocyte. Wilkinson et al. (47) demonstrated that complement proteins were deposited beneath the capsule layer of the highly encapsulated strain M (serotype 1). Similarly, we demonstrated by electron microscopy that C3 was localized on the cell walls beneath the capsules of S. aureus Reynolds (CP5) and Reynolds (CP8) (Fig. 7). This finding is consistent with our observation that neither purified CP5 nor CP8 activated the alternative complement pathway in NHS. Approximately 85% of Reynolds (CP5) and Reynolds (CP8) cells stained positive for capsule when analyzed by flow cytometry, an observation that is consistent with previous studies (34). Bacteria that expressed scant or no capsule on electron micrographs were characterized by more surface-associated C3 molecules than cells with abundant capsule. Although the capsule appeared to inhibit complement deposition on the staphylococcal surface, purified CP5 and CP8 did not affect complement deposition when added to Reynolds (CP–) cells in 10% NHS.
Given the differences in phagocytic uptake and killing of Reynolds (CP5) and Reynolds (CP8), despite a similar degree of opsonization, it is possible that there are differences in the ability of surface-associated CP5 and CP8 to interfere with the recognition of cell wall-bound C3 by receptors on PMNs. CP5 and CP8 are comprised of the same trisaccharide repeating units and differ only in the linkages between the sugars and sites of O acetylation. Although both CPs exhibited similar levels of O acetylation, our data indicate differences in the degree of N acetylation of the amino sugar residues that comprise CP5 and CP8. Partial de-N acetylation of CP8 results in the presence of free amino groups on the polysaccharide. To address whether this difference accounts, at least in part, for the observed differences in bacterial virulence that we have observed, we are characterizing a mutant strain carrying a deletion in the S. aureus putative FucNAc de-N-acetylase gene (Q. Cheng, A. O. Tzianabos, and J. C. Lee, unpublished studies).
Capsules from other encapsulated bacteria have been shown to have distinct biological properties; i.e., an association exists between capsule type and virulence. A Klebsiella pneumoniae isolate that produced a K2 capsule was more resistant to blood clearance than an isogenic strain that expressed a K21a capsule (12). The authors were able to correlate virulence with a capsule structure that was recognized by the mannose receptor on macrophages. Alteration of capsule type in isogenic Streptococcus pneumoniae strains had a profound effect on the virulence of some strains but not others. Expression of the type 3 capsule changed the virulence of type 5 and 6B strains but had no effect on the virulence of a type 2 strain (16). Thus, both capsule type and the genetic background of the bacterial strain influence microbial virulence.
In summary, we have created isogenic mutants of S. aureus strain Reynolds that express CP5, CP8, or no capsule. We have demonstrated that a strain expressing CP5 was more virulent than an isogenic CP8-producing strain in a mouse model of bacteremia and abscess formation. The CP5-producing strain demonstrated greater resistance to in vitro opsonophagocytic killing by neutrophils, and this likely contributes to its enhanced persistence within the mouse. However, the differences between the serotype 5 and 8 strains could not be linked to differential complement deposition, C3 processing to iC3b, or stimulation of the neutrophil oxidative burst. Additional host-bacterium interactions that we did not measure may have influenced the results of our mouse virulence studies. CP5+ and CP8+ strains may differ in their susceptibility to phagocytosis and killing within the peritoneal cavity by resident macrophages, they may exhibit variable access to the lymphatics or differential clearance by the liver and spleen, or they may vary in their activation of Toll-like receptors. Our results do confirm that CP5 and CP8 differ in a number of biological properties, and these differences likely contribute to the relative virulence of serotype 5 and 8 S. aureus in vivo.
ACKNOWLEDGMENTS
This study was supported by National Institutes of Health grant AI 29040. Danbing Ke was partially supported by the postdoctoral research fellowship from Le Fonds Quebecois de la Recherche sur la Nature et les Technologies.
We thank Robert Solinga for purifying CP8 and Linda Xie for her technical assistance. We gratefully acknowledge Ali Fattom for the AltaStaph antibodies, Chia Lee for providing pCL7657, and Sunita Gulati for mouse monoclonal antibodies to human C3b and iC3b.
A.W. and D.K. contributed equally to this work.
Present address: Bioniche Therapeutics Research Center, 6100 Royalmount Ave., Montreal, Quebec, Canada H4P 2R2.
REFERENCES
1. Albus, A., R. D. Arbeit, and J. C. Lee. 1991. Virulence of Staphylococcus aureus mutants altered in type 5 capsule production. Infect. Immun. 59:1008-1014.
2. Arbeit, R. D., W. W. Karakawa, W. F. Vann, and J. B. Robbins. 1984. Predominance of two newly described capsular polysaccharide types among clinical isolates of Staphylococcus aureus. Diagn. Microbiol. Infect. Dis. 2:85-91.
3. Baddour, L. M., C. Lowrance, A. Albus, J. H. Lowrance, S. K. Anderson, and J. C. Lee. 1992. Staphylococcus aureus microcapsule expression attenuates bacterial virulence in a rat model of experimental endocarditis. J. Infect. Dis. 165:749-753.
4. Bhasin, N., A. Albus, F. Michon, P. J. Livolsi, J.-S. Park, and J. C. Lee. 1998. Identification of a gene essential for O-acetylation of the Staphylococcus aureus type 5 capsular polysaccharide. Mol. Microbiol. 27:9-21.
5. Cotter, M. J., K. E. Norman, P. G. Hellewell, and V. C. Ridger. 2001. A novel method for isolation of neutrophils from murine blood using negative immunomagnetic separation. Am. J. Pathol. 159:473-481.
6. Cunnion, K. M., P. S. Hair, and E. S. Buescher. 2004. Cleavage of complement C3b to iC3b on the surface of Staphylococcus aureus is mediated by serum complement factor I. Infect. Immun. 72:2858-2863.
7. Cunnion, K. M., J. C. Lee, and M. M. Frank. 2001. Capsule production and growth phase influence binding of complement to Staphylococcus aureus. Infect. Immun. 69:6796-6803.
8. Fattom, A., R. Schneerson, D. C. Watson, W. W. Karakawa, D. Fitzgerald, I. Pastan, X. Li, J. Shiloach, D. A. Bryla, and J. B. Robbins. 1993. Laboratory and clinical evaluation of conjugate vaccines composed of Staphylococcus aureus type 5 and type 8 capsular polysaccharides bound to Pseudomonas aeruginosa recombinant exoprotein A. Infect. Immun. 61:1023-1032.
9. Fournier, J. M., W. F. Vann, and W. W. Karakawa. 1984. Purification and characterization of Staphylococcus aureus type 8 capsular polysaccharide. Infect. Immun. 45:87-93.
10. Gresham, H. D., J. H. Lowrance, T. E. Caver, B. S. Wilson, A. L. Cheung, and F. P. Lindberg. 2000. Survival of Staphylococcus aureus inside neutrophils contributes to infection. J. Immunol. 164:3713-3722.
11. Hochkeppel, H. K., D. G. Braun, W. Vischer, A. Imm, S. Sutter, U. Staeubli, R. Guggenheim, E. L. Kaplan, A. Boutonnier, and J. M. Fournier. 1987. Serotyping and electron microscopy studies of Staphylococcus aureus clinical isolates with monoclonal antibodies to capsular polysaccharide types 5 and 8. J. Clin. Microbiol. 25:526-530.
12. Kabha, K., L. Nissimov, A. Athamna, Y. Keisari, H. Parolis, L. A. Parolis, R. M. Grue, J. Schlepper-Schafer, A. R. Ezekowitz, D. E. Ohman, et al. 1995. Relationships among capsular structure, phagocytosis, and mouse virulence in Klebsiella pneumoniae. Infect. Immun. 63:847-852.
13. Karakawa, W. W., A. Sutton, R. Schneerson, A. Karpas, and W. F. Vann. 1988. Capsular antibodies induce type-specific phagocytosis of capsulated Staphylococcus aureus by human polymorphonuclear leukocytes. Infect. Immun. 56:1090-1095.
14. Karakawa, W. W., and W. F. Vann. 1982. Capsular polysaccharides of Staphylococcus aureus. Semin. Infect. Dis. 4:285-293.
15. Kasatiya, S. S., and J. N. Baldwin. 1967. Nature of the determinant of tetracycline resistance in Staphylococcus aureus. Can. J. Microbiol. 13:1079-1086.
16. Kelly, T., J. P. Dillard, and J. Yother. 1994. Effect of genetic switching of capsular type on virulence of Streptococcus pneumoniae. Infect. Immun. 62:1813-1819.
17. Kiser, K. B., N. Bhasin, L. Deng, and J. C. Lee. 1999. Staphylococcus aureus cap5P encodes a UDP-N-acetylglucosamine 2-epimerase with functional redundancy. J. Bacteriol. 181:4818-4824.
18. Lee, J. C. 1995. Electrotransformation of staphylococci, vol. 47. Humana Press, Totawa, N.J.
19. Lee, J. C., M. J. Betley, C. A. Hopkins, N. E. Perez, and G. B. Pier. 1987. Virulence studies, in mice, of transposon-induced mutants of Staphylococcus aureus differing in capsule size. J. Infect. Dis. 156:741-750.
20. Lee, J. C., M. J. Liu, J. Parsonnet, and R. D. Arbeit. 1990. Expression of type-8 capsular polysaccharide and production of toxic shock syndrome toxin-1 are associated among vaginal isolates of Staphylococcus aureus. J. Clin. Microbiol. 28:2612-2615.
21. Lee, J. C., F. Michon, N. E. Perez, C. A. Hopkins, and G. B. Pier. 1987. Chemical characterization and immunogenicity of capsular polysaccharide isolated from mucoid Staphylococcus aureus. Infect. Immun. 55:2191-2197.
22. Lee, J. C., J.-S. Park, S. E. Shepherd, V. Carey, and A. Fattom. 1997. Protective efficacy of antibodies to the Staphylococcus aureus type 5 capsular polysaccharide in a modified model of endocarditis in rats. Infect. Immun. 65:4146-4151.
23. Lee, J. C., S. Takeda, P. J. Livolsi, and L. C. Paoletti. 1993. Effects of in vitro and in vivo growth conditions on expression of type-8 capsular polysaccharide by Staphylococcus aureus. Infect. Immun. 61:1853-1858.
24. Lee, J. C., S. L. Xu, A. Albus, and P. J. Livolsi. 1994. Genetic analysis of type 5 capsular polysaccharide expression by Staphylococcus aureus. J. Bacteriol. 176:4883-4889.
25. Lowy, F. D. 1998. Medical progress: Staphylococcus aureus infections. N. Engl. J. Med. 339:520-532.
26. Luong, T. T., and C. Y. Lee. 2002. Overproduction of type 8 capsular polysaccharide augments Staphylococcus aureus virulence. Infect. Immun. 70:3389-3395.
27. McQuillen, D. P., S. Gulati, S. Ram, A. K. Turner, D. B. Jani, T. C. Heeren, and P. A. Rice. 1999. Complement processing and immunoglobulin binding to Neisseria gonorrhoeae determined in vitro simulates in vivo effects. J. Infect. Dis. 179:124-135.
28. Moreau, M., J. C. Richards, J. M. Fournier, R. A. Byrd, W. W. Karakawa, and W. F. Vann. 1990. Structure of the type-5 capsular polysaccharide of Staphylococcus aureus. Carbohydrate Res. 201:285-297.
29. Nilsson, I.-M., J. C. Lee, T. Bremell, C. Ryden, and A. Tarkowski. 1997. The role of staphylococcal polysaccharide microcapsule expression in septicemia and septic arthritis. Infect. Immun. 65:4216-4221.
30. O'Riordan, K., and J. C. Lee. 2004. Staphylococcus aureus capsular polysaccharides. Clin. Microbiol. Rev. 17:218-234.
31. Peng, H.-L., R. P. Novick, B. Kreiswirth, J. Kornblum, and P. Schlievert. 1988. Cloning, characterization, and sequencing of an accessory gene regulator (agr) in Staphylococcus aureus. J. Bacteriol. 170:4365-4372.
32. Peterson, P. K., B. J. Wilkinson, Y. Kim, D. Schmeling, S. D. Douglas, and P. G. Quie. 1978. The key role of peptidoglycan in the opsonization of Staphylococcus aureus. J. Clin. Invest. 61:597-609.
33. Portoles, M., K. B. Kiser, N. Bhasin, K. H. N. Chan, and J. C. Lee. 2001. Staphylococcus aureus Cap5O has UDP-ManNAc dehydrogenase activity and is essential for capsule expression. Infect. Immun. 69:917-923.
34. Poutrel, B., P. Rainard, and P. Sarradin. 1997. Heterogeneity of cell-associated CP5 expression on Staphylococcus aureus strains demonstrated by flow cytometry. Clin. Diagn. Lab. Immunol. 4:275-278.
35. Projan, S. J., and R. P. Novick. 1997. The molecular basis of pathogenicity, p. 55-81. In K. B. Crossley and G. L. Archer (ed.), The staphyloccci in human disease. Churchill Livingstone, New York, N.Y.
36. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
37. Sau, S., N. Bhasin, E. R. Wann, J. C. Lee, T. J. Foster, and C. Y. Lee. 1997. The Staphylococcus aureus allelic genetic loci for serotype 5 and 8 capsule expression contain the type-specific genes flanked by common genes. Microbiology 143:2395-2405.
38. Sau, S., and C. Y. Lee. 1996. Cloning of type 8 capsule genes and analysis of gene clusters for the production of different capsular polysaccharides in Staphylococcus aureus. J. Bacteriol. 178:2118-2126.
39. Sau, S., J. Sun, and C. Y. Lee. 1997. Molecular characterization and transcriptional analysis of type 8 capsule genes in Staphylococcus aureus. J. Bacteriol. 179:1614-1621.
40. Sompolinsky, D., Z. Samra, W. W. Karakawa, W. F. Vann, R. Schneerson, and Z. Malik. 1985. Encapsulation and capsular types in isolates of Staphylococcus aureus from different sources and relationship to phage types. J. Clin. Microbiol. 22:828-834.
41. Thakker, M., J.-S. Park, V. Carey, and J. C. Lee. 1998. Staphylococcus aureus serotype 5 capsular polysaccharide is antiphagocytic and enhances bacterial virulence in a murine bacteremia model. Infect. Immun. 66:5183-5189.
42. Tzianabos, A. O., J. Y. Wang, and J. C. Lee. 2001. Structural rationale for the modulation of abscess formation by Staphylococcus aureus capsular polysaccharides. Proc. Natl. Acad. Sci. USA 98:9365-9370.
43. Vann, W. F., M. Moreau, A. Sutton, R. A. Byrd, and W. W. Karakawa. 1988. Structure and immunochemistry of Staphylococcus aureus capsular polysaccharides, p. 187-198. In M. A. Horwitz (ed.), Bacterial-host cell interaction, vol. 64. Alan R. Liss, Inc., New York, N.Y.
44. Verbrugh, H. A., W. C. van Dijk, R. Peters, M. E. van Erne, M. R. Daha, P. K. Peterson, and J. Verhoef. 1980. Opsonic recognition of staphylococci mediated by cell wall peptidoglycan: antibody-independent activation of human complement and opsonic activity of peptidoglycan antibodies. J. Immunol. 124:1167-1173.
45. Wilkinson, B. J., Y. Kim, and P. K. Peterson. 1981. Factors affecting complement activation by Staphylococcus aureus cell walls, their components, and mutants altered in teichoic acid. Infect. Immun. 32:216-224.
46. Wilkinson, B. J., Y. Kim, P. K. Peterson, P. G. Quie, and A. F. Michael. 1978. Activation of complement by cell surface components of Staphylococcus aureus. Infect. Immun. 20:388-392.
47. Wilkinson, B. J., S. P. Sisson, Y. Kim, and P. K. Peterson. 1979. Localization of the third component of complement on the cell wall of encapsulated Staphylococcus aureus M: implications for the mechanism of resistance to phagocytosis. Infect. Immun. 26:1159-1163.
48. Xu, S., R. D. Arbeit, and J. C. Lee. 1992. Phagocytic killing of encapsulated and microencapsulated Staphylococcus aureus by human polymorphonuclear leukocytes. Infect. Immun. 60:1358-1362.(Andrew Watts, Danbing Ke,)
Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts 02115
ABSTRACT
Most isolates of Staphylococcus aureus produce a serotype 5 (CP5) or 8 (CP8) capsular polysaccharide. To investigate whether CP5 and CP8 differ in their biological properties, we created isogenic mutants of S. aureus Reynolds that expressed CP5, CP8, or no capsule. Biochemical analyses of CP5 and CP8 purified from the isogenic S. aureus strains were consistent with published structures. The degree of O acetylation of each polysaccharide was similar, but CP5 showed a greater degree of N acetylation. Mice challenged with the CP5+ strain showed a significantly higher bacteremia level than mice challenged with the CP8+ strain. Similarly, the CP5+ strain survived preferentially in the bloodstream and kidneys of infected mice challenged with a mixed inoculum containing both strains. The enhanced virulence of the CP5+ strain in vivo correlated with its greater resistance to in vitro killing in whole mouse blood. Likewise, in vitro opsonophagocytic killing assays with human neutrophils and sera revealed greater survival of the Reynolds (CP5) strain, even though the kinetics of opsonization by C3b and iC3b was similar for both the CP5+ and CP8+ strains. Electron micrographs demonstrated C3 molecules on the cell wall beneath the capsule layer for both serotype 5 and 8 strains. Purified CP5 and CP8 stimulated a modest oxidative burst in human neutrophils but failed to activate the alternative complement pathway. These results indicate that CP5 and CP8 differ in a number of biological properties, and these differences likely contribute to the relative virulence of serotype 5 and 8 S. aureus in vivo.
INTRODUCTION
Staphylococcus aureus is a major bacterial pathogen that causes a wide spectrum of clinical infections, ranging from localized soft-tissue infections to life-threatening bacteremia and endocarditis (25). Many virulence factors contribute to the pathogenesis of staphylococcal infections, including surface-associated adhesins and secreted exoproteins and toxins (35). Like many invasive bacterial pathogens, S. aureus produces a capsular polysaccharide (CP) that enhances its resistance to clearance by host innate immune defenses. Most clinical isolates of S. aureus are encapsulated, and serotype 5 and 8 strains predominate (2, 11, 40). The type 5 (CP5) and type 8 (CP8) capsular polysaccharides have similar trisaccharide repeating units comprised of N-acetyl mannosaminuronic acid, N-acetyl L-fucosamine, and N-acetyl D-fucosamine (9, 28, 43). CP5 and CP8 are serologically distinct, and this can be attributed to differences in the linkages between the sugars and in the sites of O acetylation.
Previous studies have correlated S. aureus capsule production with resistance to in vitro phagocytic uptake and killing (13, 41). Human neutrophils phagocytose capsule-negative mutants in the presence of nonimmune serum with complement activity, whereas serotype 5 isolates require both capsule-specific antibodies and complement for optimal opsonophagocytic killing (4, 41). Nilsson et al. (29) reported that peritoneal macrophages from mice phagocytosed significantly greater numbers of a CP5-negative mutant compared to the parental strain Reynolds. Once phagocytosed, the CP5-positive strain survived intracellularly to a greater extent than the mutant strain. Cunnion et al. (7) compared opsonization of isogenic S. aureus strains and demonstrated that the CP5-positive strain bound 42% less serum complement (C3) than the acapsular mutant.
Serotype 5 S. aureus strains have also been shown to be more virulent than acapsular mutants in animal models of staphylococcal infection. The CP5-positive strain Reynolds produced higher bacteremia levels in mice and resisted host clearance to a greater extent than two capsule-deficient mutants (41). Strain Reynolds was more virulent than an acapsular mutant in rodent models of renal infection or abscess formation (33, 42). Mice inoculated with the serotype 5 S. aureus strain developed more frequent and severe arthritis, demonstrated greater weight loss, and showed a higher mortality rate than mice infected with capsule-negative mutants (29).
Studies documenting the role of CP8 in virulence were lacking until recently, when Luong and Lee (26) showed that a CP8-overproducing mutant was more resistant to in vitro opsonophagocytic killing by human neutrophils than the parental strain Becker. Likewise, the CP8-overproducing strain persisted longer in the bloodstream, liver, and spleen of infected mice than strain Becker. These results were the first to show that CP8 promoted S. aureus virulence in an animal model of infection, although it was necessary to create a mutant that produced excess CP8 in order to see the effect.
Our data indicate that the serotype 8 strain Becker is less virulent for mice and rats than the serotype 5 strain Reynolds (1, 22, 23, 41 and unpublished observations), and efforts to enhance the virulence of strain Becker by cultivation under conditions that enhance capsule production failed (23). Our preliminary results with <10 clinical isolates suggest that serotype 5 S. aureus isolates produce more CP and are more virulent for mice than type 8 S. aureus isolates. However, these differences in virulence cannot be attributed to capsule type only, since the strains examined were not isogenic. The purpose of this study was to construct isogenic mutants of S. aureus that expressed CP5, CP8, or no capsule. Creation of these strains allowed us to investigate the relative contribution of capsule type to staphylococcal virulence in the genetic background of a virulent S. aureus strain.
MATERIALS AND METHODS
Bacterial strains and growth conditions. The bacterial strains and plasmids used in this study are listed in Table 1. For genetic manipulations Escherichia coli was cultured on Luria-Bertani medium, whereas S. aureus strains were cultured in tryptic soy broth (TSB) or agar (TSA). When appropriate, the culture medium was supplemented with chloramphenicol (Cm) at 10 μg/ml, erythromycin (Em) at 10 μg/ml, or ampicillin at 100 μg/ml. For all other experiments, S. aureus was cultivated for 24 h at 37°C on Columbia agar (Difco Laboratories, Detroit, Mich.) supplemented with 2% NaCl.
DNA manipulations. S. aureus chromosomal DNA was isolated using the Wizard Genomic Purification kit (Promega, Madison, Wis.), and plasmid DNA was purified with a QIAprep spin miniprep kit 250 (QIAGEN, Inc., Valencia, Calif.). Standard molecular cloning procedures were followed as detailed by Sambrook et al. (36). Restriction endonucleases and other DNA modification enzymes were obtained from Invitrogen Corp. (Carlsbad, Calif.) or New England Biolabs, Inc. (Beverly, Mass.). Electroporation and transduction of S. aureus strains were performed as described previously (15, 18).
Construction of strain Reynolds (CP–). An allelic replacement method was utilized to create an acapsular mutant of the serotype 5 strain Reynolds. The strategy was to replace the chromosomal copy of the serotype 5-specific genes (cap5HIJK) with an ermB cassette encoding Em resistance. A 2.2-kb EcoRI fragment carrying cap5FG was subcloned from pJCL24 (4) into pCL10 (39), a temperature-sensitive E. coli-S. aureus shuttle vector (Table 1). Orientation of the cloned fragment in the recombinant plasmid (pCap5FG) was determined through asymmetrical restriction digests with EcoRV, BamHI, and BstXI. The S. aureus ermB gene was subcloned from pErmB (17) into pCap5FG to create pCap5FGermB. The cap5LM gene fragment from pJCL24 (24) was amplified by PCR (25 cycles of 94°C for 30 s, 50°C for 30 s, and 68°C for 2 min, with a final extension at 72°C for 7 min) with primers cap5L-f (5'-GCGATCTAGATGACGCTTCACACGATTAC-3') and cap5M-r (5'-ACCATTCAGACCTTCTTTTCCATAAACTGCC-3'), which carry XbaI sites (underlined). The amplified fragment was cloned into pCap5FGermB to create pAP1.2 (Table 1), and the cloned cap5LM fragment was verified by sequencing. pAP1.2 was electroporated into the restriction-negative S. aureus strain RN4220 (Table 1) and then transduced into strain Reynolds, in both cases with selection for Cm-resistant (Cmr) colonies at 30°C. Reynolds(pAP1.2) was incubated at 42°C for 24 h in the presence of Cm to select for strains with plasmid integration into the chromosome by homologous recombination. A single integrant colony was confirmed by PCR and then passaged three times at 30°C in medium without antibiotics. This process allowed the integrated plasmid to be excised from the chromosome by a single crossover at the duplicated region created during plasmid integration. Depending on the site of crossover during excision, the mutation site on the insert of pAP1.2 would either be left on the chromosome, thus generating the desired mutant (Emr Cms), or be lost on the cured plasmid, thus regenerating a revertant wild-type strain (Ems Cms). A revertant (named strain JL810) was valuable for evaluating the effect of genetic manipulation on CP production. The genotypes of an Emr Cms excisant, named Reynolds (CP–), and the revertant JL810 were verified by PCR and by Southern hybridization using the ECL direct nucleic acid labeling and detection system (Amersham Biosciences, Piscataway, N.J.).
Construction of strain Reynolds (CP8). Plasmid pCL7657, generously provided by Chia Y. Lee, University of Arkansas for Medical Sciences, carries the entire cap8 locus from strain Becker (38). A 12.4-kb SalI fragment of pCL7657 harboring cap8C through cap8 M was ligated into pCL10 to create recombinant plasmid pDK1. The plasmid was electroporated into S. aureus RN4220 and then transduced into strain Reynolds (CP–) as described above, selecting for Cmr colonies at 30°C. pDK1 was integrated into the chromosome and excised as described above, except that the excisants were screened for susceptibility to Cm and Em. A single excisant, Reynolds (CP8), was further characterized by PCR and Southern blots.
Phenotypic characterization of S. aureus strains. Isogenic recombinant S. aureus strains were phenotypically evaluated by API (BioMerieux Inc., Durham, N.C.) and by hemolysis on sheep blood agar plates. Staphylococcal growth rates were determined by monitoring the optical density at 650 nm of TSB cultures of each strain.
Serotype 5 and 8 capsular antisera were prepared by immunization of rabbits with heat- or formalin-killed suspensions of S. aureus strain Reynolds (serotype 5) or PS80 (serotype 8). To render the antiserum CP5 or CP8 specific, it was absorbed with the acapsular S. aureus strain Wood 46 (protein A deficient) and trypsinized suspensions of acapsular mutants JL243 and JL252 (3). AltaStaph, kindly provided by A. Fattom (Nabi, Inc., Rockville, Md.), is a human immunoglobulin G (IgG) product derived from the pooled serum of individuals immunized with CP5 and CP8 conjugated to Pseudomonas aeruginosa exotoxoid A (8). All antibody preparations were heat inactivated for 30 min at 56°C prior to use. The capsular phenotypes of the parental and recombinant S. aureus strains were assessed by immunodiffusion and colony immunoblotting (20). We utilized an enzyme-linked immunosorbent assay (ELISA) inhibition method (23) for quantitation of S. aureus cell-associated CP.
Immunoelectron microscopy was used to visualize CP on the S. aureus strains. Suspensions of S. aureus cells were incubated in 10 ml of 0.1 M sodium phosphate buffer (pH 8.0) containing 1 mg/ml trypsin for 1 h at 37°C. The bacterial suspensions were washed with phosphate-buffered saline (PBS) before room temperature incubation for 2 h with capsular antisera. After being washed, the bacteria were fixed and processed for electron microscopy as described previously (19).
Analysis of purified CP5 and CP8. CP5 and CP8 were purified from Reynolds (CP5) and Reynolds (CP8) as described previously (42). Teichoic acid and peptidoglycan were purified as described by Peterson et al. (32). Chemical assays for O acetylation, protein, nucleic acid, and phosphorus were performed as described previously (21). The polysaccharides were treated with alkali (4) to remove O-acetyl groups. 1H nuclear magnetic resonance spectra were recorded on a Bruker 600 spectrophotometer in D2O at 60°C. The chemical shifts were given on the scale relative to the internal standard. Free amino groups on the polysaccharides were also detected using a fluorescamine assay described previously (42). Purified CP5, CP8, peptidoglycan, and teichoic acid were tested for activation of the alternative complement pathway by incubating each polymer (final concentration, 100 μg/ml) in 50% normal human serum (NHS) in gelatin-Veronal buffer (GVB) containing 8 mM EGTA and 5 mM Mg2+ for 15 min at room temperature. The mixtures were then diluted and assayed for C3a production with an ELISA kit from Quidel.
Mouse model of bacteremia and renal abscess formation. Female CD-1 mice, 7 to 8 weeks old, were obtained from Harlan Sprague Dawley, Inc. (Indianapolis, Ind.) and given food and water ad libitum. Groups of 6 to 13 mice were challenged by the intraperitoneal (i.p.) route with a 0.5-ml inoculum containing either 5 x 107 CFU or 9 x 107 CFU S. aureus. Separate groups of three or four mice were bled at different time points after inoculation. Bacteremia levels were determined by quantitative counts performed in duplicate on neat or diluted blood samples, and the results were expressed as log CFU S. aureus per ml blood.
Animals were coinfected with an equal number (5 x 107 CFU) of Reynolds (CP5) and Reynolds (CP8) in a series of competition experiments. The mixed bacterial suspension was injected i.p., and separate groups of mice were bled for quantitative culture at 20, 120, or 200 min after inoculation. The capsule phenotype of colonies from each blood culture was determined by immunoblotting (20) with CP5- or CP8-specific antiserum. Bacterial concentrations in the blood of infected mice were compared by the Welch modification of the unpaired Student t test (InStat 2.0; GraphPad Software, Inc., San Diego, Calif.). All mice were euthanized after 5 days, and their kidneys were examined grossly for the presence of abscesses. Both kidneys from animals with renal abscesses were excised, weighed, and homogenized. Dilutions of the homogenates were plated quantitatively, and CP5- or CP8-positive colonies were determined by colony immunoblotting as described above. The numbers of animals with abscesses containing either Reynolds (CP5) or Reynolds (CP8) were compared by Fisher's exact test.
S. aureus killing assays. Heparinized blood was collected by tail vein puncture of male CD-1 mice. Whole-blood killing assays were performed by mixing 200 μl of pooled mouse blood with 100 μl of an S. aureus suspension to achieve a final concentration of 105 CFU/ml. The samples were incubated on a rotator at 37°C, and 10-μl aliquots were removed for quantitative culture at time zero and after 60 min. Control tubes contained 200 μl mouse blood and 100 μl PBS. The percent killing was defined as the reduction in CFU/ml after 60 min compared with that at time zero.
The opsonophagocytic killing assay with human polymorphonuclear leukocytes (PMNs) was performed essentially as described by Xu et al. (48) with the following modifications. Normal human serum (NHS) was collected and pooled from three healthy adult volunteers. Three different pools of NHS were absorbed for 1 h on ice with all three isogenic S. aureus strains: Reynolds (CP5), Reynolds (CP8), and Reynolds (CP–). Certain NHS pools were supplemented with 10 mM EDTA prior to absorption to further limit complement activation. Following absorption, the NHS was centrifuged, filtered (0.45 μm), and stored at –80°C for not more than 3 months. Immediately prior to use, the serum was thawed and supplemented, when necessary, with 7.5 mM MgCl2 and 2.5 mM CaCl2 to restore free ions. The opsonophagocytic killing assay was performed in polypropylene tubes containing 1 x 106 PMNs, 1 x 106 CFU S. aureus, and absorbed serum (2.5 to 10% final concentration) in a total volume of 500 μl minimal essential medium (Invitrogen). Control samples contained S. aureus and PMNs (without serum) or S. aureus alone. The samples were rotated end over end (12 rpm) for 2 h at 37°C, followed by sonication (Probe Sonic Dismembranator 60; Fisher Scientific, Pittsburgh, Pa.) for 5 s at 4 W to minimize antibody-mediated bacterial agglutination. Sample dilutions were made in sterile deionized water, and bacterial killing was estimated by plating the diluted samples in duplicate on TSA. The percent killing was defined as the reduction in CFU/ml after 2 h compared with that at time zero. Results were compared by one-way analysis of variance (ANOVA) and the Tukey-Kramer multiple comparisons test (InStat).
To determine the number of viable intracellular bacteria associated with the PMNs in 10% NHS, assay samples were taken at 20 and 60 min and incubated with lysostaphin (10 μg/ml) for 20 min at 37°C to kill the extracellular bacteria. The PMNs were centrifuged at 2,700 x g for 5 min at 4°C, lysed in sterile water, and plated in duplicate on TSA. To assess the fate of intracellular S. aureus, lysostaphin was added to certain samples 20 or 60 min after mixing the bacteria and PMNs. The PMNs were incubated for an additional 2 h at 37°C, at which time numbers of viable intracellular bacteria were estimated as described above.
Flow cytometry. We measured antibody binding to the isogenic S. aureus strains by incubation of 108 CFU S. aureus with 100 μl of 10% unabsorbed or absorbed NHS for 30 min at ambient temperature. The samples were washed in PBS containing 1% bovine serum albumin (PBS-BSA) before incubation with 20 μg/ml of protein G-Alexa Fluor 488 conjugate (Molecular Probes, Eugene, Oreg.). The bacteria were fixed overnight at 4°C in 2% paraformaldehyde, washed, and stained with 10 μg/ml of hexidium iodide (Molecular Probes). Fluorescent S. aureus cells were gated on a Beckman-Coulter Epics flow cytometer with a 488-nm laser, and 10,000 events were recorded. IgG binding to S. aureus was measured in the FL1 channel. In separate experiments, S. aureus capsule expression was measured by flow cytometry after incubating the bacteria with AltaStaph antibodies [absorbed with Reynolds (CP–)] followed by the protein G-Alexa Fluor 488 conjugate. Protein A-mediated binding of IgG to S. aureus was measured by flow cytometry after incubation of the bacteria with Alexa Fluor 488 conjugated to rabbit IgG.
The respiratory burst response of human PMNs to S. aureus or purified polysaccharides was evaluated by flow cytometry. Approximtely106 PMNs were incubated with 0.8 μg/ml dihydrorhodamine (DHR123; Sigma, St. Louis, Mo.) in minimal essential medium-BSA for 20 min at 37°C. This dye diffuses inside the leukocytes and is colorless in its reduced form. When the dye is oxidized, it fluoresces and can be detected by flow cytometry. The PMN stimulants that were tested included S. aureus (106 CFU) opsonized with 10% absorbed NHS, 1 μM N-formylmethionyl-leucine-phenylalanine (fMLP; Sigma), and 100 μg/ml of either purified CP5, CP8, teichoic acid, or polygalacturonic acid (Sigma). Each stimulant was incubated with the PMNs for 30 min. The PMNs were then washed twice in PBS-BSA, filtered through a cell-strainer cap, and analyzed by flow cytometry. To account for the percentage of PMNs responding to the stimulus as well as the magnitude of the response, the results were expressed as relative PMN activation: percent activated PMNs x median fluorescence index (MFI) of the activated PMNs. To normalize the values obtained within different experiments, this value was divided by the relative PMN activation of the unstimulated PMNs.
To detect C3b and iC3b deposition, 108 CFU S. aureus were incubated at 37°C with 10% absorbed NHS for 1, 3, 5, 10, 20, or 30 min. The reaction was stopped on ice by the addition of ice-cold GVB containing 10 mM EDTA, and the samples were washed three times with PBS-BSA. Mouse monoclonal antibodies to human C3b [C-5G; IgG1()] or iC3b [G-3E; IgG2b()] (27) (kindly supplied by Sunita Gulati, Boston Medical Center) were incubated with each sample for 30 min at ambient temperature. Alexa Fluor 488-conjugated goat anti-mouse IgG (Fab)2 fragments (Molecular Probes) were added to a final concentration of 2 μg/ml and incubated for 30 min at room temperature. Samples were washed, fixed, stained, and analyzed as described above.
Visualization of C3 on S. aureus by electron microscopy. S. aureus cells were incubated with 40% pooled human serum in GVB containing 0.5 mM Mg2+ and 0.15 mM Ca2+ for 30 min at 37°C. The bacteria were washed three times with PBS containing 0.5% fish skin gelatin and incubated with murine monoclonal antibodies to human C3c (Quidel Corp., San Diego, Calif.). The staphylococci were washed before incubation with goat anti-mouse IgG conjugated to 12-nm colloidal gold particles (Jackson ImmunoResearch, West Grove, Pa.). Samples were fixed overnight at 4°C in 0.1 M sodium cacodylate buffer, pH 7.4, containing 2% paraformaldehyde and 2.5% glutaraldehyde. To visualize the capsule and preserve its integrity during the dehydration step (47), the fixed bacteria were incubated for 2 h with rabbit CP-specific antibodies. The preparations were postfixed in 1% osmium tetroxide, dehydrated with graded alcohols, and infiltrated with resin and propylene oxide (1:1) prior to embedding in Taab resin. Thin sections (90 nm) were cut with a Reichert ultramicrotome and transferred to copper grids. The sections were stained with 1% uranyl acetate in acetone followed by 0.2% lead citrate and examined on a JEOL-1200EX transmission electron microscope.
RESULTS
Construction and characterization of isogenic mutants. The genes encoding S. aureus capsule biosynthesis, transport, and assembly are clustered within the cap5 and cap8 loci. These loci are allelic and are comprised of 16 genes named cap5(8)A through cap5(8)P. Whereas the predicted gene products of cap5(8)A through cap5(8)G and cap5(8)L through cap5(8)P are virtually identical, the four open reading frames located in the central region of each locus [cap5(8)HIJK] are type specific (37). We replaced cap5HIJK in strain Reynolds (CP5) with an ermB cassette to create an isogenic mutant designated Reynolds (CP–) that was acapsular by colony immunoblotting and immunodiffusion. A CP5-positive revertant designated JL810 was shown to be identical to Reynolds (CP5) by Southern blotting and PCR analyses. To create the Reynolds (CP8) strain, we replaced the ermB gene within the cap locus of Reynolds (CP–) with the cap8HIJK genes from strain Becker. An Ems excisant named Reynolds (CP8) was analyzed by Southern blotting and PCR to confirm the authenticity of its genotype.
Strains Reynolds (CP5), Reynolds (CP–), Reynolds (CP8), and revertant JL810 showed identical colony morphology, pigmentation, and hemolysin production on blood agar plates. Likewise, the API profile of each of the strains was identical, and their in vitro growth rates in TSB were the same. PCR amplification of the pCL10-encoded cat gene from the integrants, but not from strain Reynolds (CP5), JL810, Reynolds (CP8), or Reynolds (CP–), insured complete loss of the recombinant plasmids in the excisants. CP production by the parental and recombinant strains was measured by an ELISA inhibition assay. As shown in Table 2, Reynolds (CP5), revertant JL810, and Reynolds (CP8) all produced similar quantities of CP. The inclusion of revertant JL810 in these experiments verified that the allelic replacement mutagenesis method did not have deleterious effects on CP production. Reynolds (CP–) had no detectable CP expression (<0.7 μg/1010 CFU), and Reynolds (CP8) produced approximately twice as much CP8 as strain Becker. Transmission electron micrographs of the isogenic mutants (Fig. 1) confirmed that Reynolds (CP5) and Reynolds (CP8) produced abundant surface-associated CP, whereas Reynolds (CP–) had no detectable capsule.
Comparative analysis of purified CP5 and CP8. We purified and analyzed the CPs produced by Reynolds (CP5) and Reynolds (CP8). Both polysaccharide polymers are of high-molecular-weight with Kav values of 0.03 on a Sephacryl S-300 column. (Kav represents the fraction of the stationary gel volume which is available for diffusion of a given solute species.) Purified CP5 and CP8 were both 25% O acetylated, and each polymer had <0.5% protein, nucleic acid, and phosphorus by chemical analyses. Tzianabos et al. (42) reported that purified CP8 showed strain-dependent differences in the degree of N acetylation of the fucosamine residues. Nuclear magnetic resonance analysis revealed that the CP5 from strain Reynolds was 98% N acetylated, whereas the CP8 purified from Reynolds (CP8) was 89% N acetylated. The presence of free amino groups on CP8 was confirmed using a fluorescamine assay, which demonstrated a 1.8-fold higher signal for CP8 compared with CP5. The putative gene product responsible for de-N acetylating the CP is outside of the capsule locus, and experiments to characterize its activity are ongoing in our laboratory.
Effect of capsule type on S. aureus virulence. To compare the influence of capsule type on bacterial virulence, mice were challenged i.p. with sublethal doses of Reynolds (CP5), Reynolds (CP8), or Reynolds (CP–). As shown in Table 3, animals challenged with the encapsulated strains showed a significantly (P < 0.001) higher bacteremia level 20 and 180 min after inoculation compared with animals infected with the acapsular mutant. The rapid clearance of the acapsular mutant that we observed is consistent with the results of Thakker et al., who evaluated the virulence of a capsule-deficient transposon mutant of S. aureus (41). Mice challenged with Reynolds (CP5) and Reynolds (CP8) showed similar bacteremia levels 20 min after bacterial challenge. However, by 180 min after inoculation, the blood concentrations of Reynolds (CP5) were significantly (P = 0.01) higher than those of Reynolds (CP8). Of note, there was little change in the quantitative blood culture results for animals injected with Reynolds (CP8) or Reynolds (CP–) between 20 and 180 min, whereas Reynolds (CP5) increased in concentration during the same time interval (Table 3).
To further confirm these differences in virulence and to minimize the effects of variability among individual mice, 24 additional animals were challenged i.p. with a mixed suspension containing an equal number (5 x 107 CFU) of Reynolds (CP5) and Reynolds (CP8). Separate groups of eight animals were bled after 20, 120, and 200 min, and quantitative blood cultures were performed. The percentages of CP5+ and CP8+ colonies recovered from each blood sample were assessed by immunoblotting. As shown in Fig. 2, the percentage of Reynolds (CP5) cells in the population increased over time, accounting for 75% of the recovered organisms by 200 min after bacterial challenge. Mean ± standard error of the mean (SEM) blood levels of Reynolds (CP5) were significantly higher than Reynolds (CP8) after 200 min: 3.39 ± 0.13 versus 2.89 ± 0.16 log CFU S. aureus/ml, respectively (P = 0.03). Competition assays confirmed that there was no preferential growth of either strain in vitro (data not shown).
The mice coinfected with Reynolds (CP5) and Reynolds (CP8) were euthanized after 5 days, and their kidneys were examined grossly for the presence of renal abscesses. Both kidneys from nine animals with visible abscesses were homogenized and plated quantitatively. Colonies recovered from the kidney homogenates of each animal were assessed for CP production by colony immunoblotting. All of the abscessed kidneys yielded a pure culture of either CP5- or CP8-positive colonies. Seven of the nine mice yielded CP5-positive colonies, and the two remaining mice had CP8-positive colonies (P = 0.0278 by Fisher's exact test). Even at the lowest dilution (10–1) plated, none of the kidney homogenates yielded a mixed culture.
In vitro killing assays. To determine whether there were differences in resistance to phagocytosis attributable to CP5 or CP8 production, we measured killing of each S. aureus strain in whole mouse blood. Blood pooled from nave mice killed 59% of the Reynolds (CP–) inoculum after 60 min. In contrast, only 7% and 31% of the Reynolds (CP5) and Reynolds (CP8) organisms, respectively, were killed in mouse blood within 60 min. The difference in killing between the two encapsulated strains was significant (P = 0.03; Student t test) and mirrored the difference in blood clearance that we observed in mice challenged with the same bacterial strains.
Neutrophils represent only 20% of total leukocytes in the peripheral circulation of nave mice (5), so we chose to perform opsonophagocytic killing assays using isolated human neutrophils incubated with pooled NHS absorbed with the three isogenic S. aureus strains. As shown in Fig. 3, 75% of the Reynolds (CP–) inoculum was effectively opsonized for phagocytic killing by PMNs in NHS concentrations ranging from 2.5 to 10%. In contrast, <35% of the Reynolds (CP5) inoculum was killed in 5% or 10% NHS [P < 0.001 compared with Reynolds (CP–)], and no killing was observed in 2.5% serum. Approximately 60% of Reynolds (CP8) was killed by PMNs in 10% NHS [P < 0.05 compared with killing of Reynolds (CP5)]. Although the encapsulated strains were not killed by PMNs in 2.5% serum, the addition of 4 μg/ml of AltaStaph capsular antibodies to the phagocytic assay resulted in killing of both CP-producing strains (Fig. 3). In the absence of a complement source, anti-CP5/CP8 antibodies were not opsonic, and heat inactivation of the NHS abolished its opsonic activity (data not shown). Similar results were obtained with three different pools of NHS.
To ensure that the observed differences in phagocytic killing between Reynolds (CP5) and Reynolds (CP8) in 10% absorbed NHS did not reflect differences in residual CP5- or CP8-specific antibodies, we used flow cytometry to detect IgG antibody binding to the three isogenic strains. The MFI values for Reynolds (CP5), Reynolds (CP8), and Reynolds (CP–) incubated in 10% unabsorbed serum were 100, 108, and 133, respectively. The MFI values for the same S. aureus strains incubated with the absorbed serum were 7.2, 9.0, and 12.4, respectively, which indicates low but comparable antibody binding to each of the isolates. We confirmed that the measured IgG binding was not mediated by protein A by evaluating protein A expression on Reynolds using Alexa Fluor 488 conjugated to rabbit IgG. Although protein A was present on staphylococci cultivated in broth, none was detected on staphylococci cultivated on Columbia salt agar plates, regardless of the capsule phenotype of the strain (data not shown).
We considered the possibility that the observed differences in phagocytic killing of Reynolds (CP5) and Reynolds (CP8) might relate to postphagocytic events. Intracellular bacteria were enumerated after the addition of lysostaphin to the phagocytic assay samples to kill the extracellular bacteria. The PMNs were then washed and lysed for quantitation of intracellular viable bacteria. As shown in Fig. 4, there were significant differences among the three isogenic strains (P = 0.0081 by ANOVA) when the numbers of intracellular bacteria were measured at the 20-min time point. Uptake of Reynolds (CP–) cells was more rapid than that of the encapsulated strains. Approximately 1% of the Reynolds (CP–) inoculum remained viable within the PMNs after 20 min, and its intracellular concentration was significantly (P < 0.05) higher than that of the two encapsulated strains. Whereas maximal recovery of Reynolds (CP–) was achieved by 20 min, the numbers of intracellular encapsulated S. aureus isolates continued to increase during the 60-min incubation period. The intracellular concentration of the three strains was also significantly different at 60 min (P = 0.0025). The CFU/ml of Reynolds (CP–) was significantly higher than that of Reynolds (CP5) (P < 0.05) at 60 min but not compared with Reynolds (CP8). The differences between the CP5- and CP8-producing strains were not significant at either time point, although the trend for greater uptake of the CP8+ strains was evident.
Additional phagocytic assay samples were further incubated for 2 h at 37°C after lysostaphin treatment to determine the fate of the intracellular bacteria. The CFU/ml did not change markedly during the 2-h incubation, and no differences in intracellular survival were observed between the three isogenic strains (data not shown).
Respiratory burst of PMNs in response to S. aureus. We measured the PMN oxidative burst in response to Reynolds (CP–), Reynolds (CP5), or Reynolds (CP8) to determine whether this response correlated with the results of the opsonophagocytic killing assays. Each strain was incubated with human PMNs in 10% absorbed NHS for 30 min. As shown in Fig. 5A, the relative PMN activation response to Reynolds (CP–) was significantly greater (P < 0.001) than that elicited by the CP-positive strains. The respiratory burst responses of PMNs to Reynolds (CP5) and Reynolds (CP8) were similar. In the absence of NHS, no PMN oxidative burst was observed in response to any of the staphylococcal strains (not shown). Purified CP5, CP8, and teichoic acid (100 μg/ml) also stimulated a PMN respiratory burst (Fig. 5B), but the response was modest compared with that of fMLP, which showed a relative PMN activation score of 161 (not shown on the graph). The control polymer polygalacturonic acid did not elicit a PMN oxidative burst (Fig. 5B).
The influence of CP on complement deposition. Cunnion et al. showed that strain Reynolds (CP5) cells bound less C3 than an acapsular mutant (7). To determine the influence of capsule serotype on opsonization by complement, we evaluated the kinetics of C3b and iC3b deposition on Reynolds (CP5) and Reynolds (CP8) cells by flow cytometry. As shown in Fig. 6, complement deposition on the bacterial cells in 10% NHS was rapid. C3b was detectable on both Reynolds (CP5) and Reynolds (CP8) after 1 min, and 50% of the bacteria were opsonized after 5 min (Fig. 6A). All the bacteria showed bound C3b on their surface 20 min after addition of 10% NHS to the bacterial suspension. As expected, deposition of iC3b on the S. aureus cells was delayed with respect to C3b, with surface-bound iC3b detectable only after 3 min (Fig. 6B). The kinetics of iC3b and C3b deposition for strain Reynolds (CP5) and Reynolds (CP8) were similar. Strain Reynolds (CP5) showed a consistent 10% reduction in C3b and iC3b binding between the 3- and 10-min time points, but it is unlikely that the difference between the two encapsulated strains is biologically relevant.
Purified S. aureus peptidoglycan and teichoic acid activate complement, although both polymers are less active than isolated staphylococcal cell walls (44-46). In our hands, 100 μg/ml of either peptidoglycan or teichoic acid activated the alternative complement pathway as measured by C3a production in 50% NHS (data not shown). In contrast, neither purified CP5 nor CP8 at the same concentration activated the alternative complement pathway in NHS. Although we can demonstrate that Reynolds (CP5) and Reynolds (CP8) bind less C3 than the acapsular mutant, the addition of 100 μg of purified CP5 or CP8 to Reynolds (CP–) cells did not affect complement deposition on the acapsular strain (not shown).
Localization of C3 on encapsulated S. aureus. Because neither purified CP5 nor CP8 activated the complement pathway, surface-associated CP might mask complement components deposited on the cell wall, thereby preventing recognition by receptors on the PMN membrane. Reynolds (CP5) and Reynolds (CP8) were incubated with absorbed NHS, and C3 deposition on the bacterial surface was detected by immunogold labeling. The CP was visualized and its integrity maintained by incubating the fixed bacterial cells with capsular antibodies prior to dehydration and embedding (47). As shown in Fig. 7, C3 molecules deposited on the bacterial cell wall beneath the capsule layer of both Reynolds (CP5) and Reynolds (CP8) cells. Flow cytometric analysis of capsule expression using AltaStaph antibodies revealed that approximately 85% of Reynolds (CP5) and Reynolds (CP8) expressed CP when grown on Columbia salt agar. Bacterial cells within the population that expressed scant capsule showed greater immunogold labeling than cells with abundant capsule (Fig. 7).
DISCUSSION
Capsules are common features of invasive bacterial pathogens. Numerous serotyping studies of S. aureus strains from diverse human sources have revealed that serotype 5 and 8 strains account for 25% and 50%, respectively, of clinical isolates (30). S. aureus Reynolds and Becker are the prototype serotype 5 and 8 strains, respectively, and both were isolated from patients with bacteremia (14). Strain Reynolds expresses abundant CP5, resists phagocytic killing in the absence of specific capsular antibodies, and is virulent for mice (41). S. aureus Becker produces less CP, is poorly virulent for mice, and is opsonized for phagocytic killing by complement alone (22, 23 and unpublished observations). These differences cannot be wholly attributed to capsule type, since the genetic backgrounds of these two strains are different. In order to define the differences in virulence between S. aureus strains expressing type 5 and 8 capsules, we performed allelic exchange experiments to create isogenic S. aureus strains that differed only in capsule type, and we chose the more virulent strain Reynolds as the parent strain. In the process of constructing Reynolds (CP8), we created an acapsular mutant Reynolds (CP–) that served as a useful isogenic control.
The results of our animal studies revealed that Reynolds (CP5) and Reynolds (CP8) caused a higher and more persistent bacteremia in mice than did the capsule-negative mutant (Table 3). This finding underscores the importance of the capsule in blood-borne staphylococcal infections and confirms the virulence of the serotype 5 strain Reynolds. This is the first report to document the greater virulence of a CP8-producing S. aureus strain compared with an isogenic acapsular mutant. A comparison of the isogenic CP5- and CP8-producing Reynolds strains in the bacteremia model revealed that Reynolds (CP5) elicited a higher and more persistent bacteremia than Reynolds (CP8). The bacteremia level in these mice is influenced by the ability of the microbes to avoid phagocytic uptake in the peritoneal cavity and their capacity to escape to and persist in the intravascular space. The greater virulence of Reynolds (CP5) was further substantiated in a mixed infection model, where it demonstrated preferential survival in the mouse blood (Fig. 2) and kidneys compared with Reynolds (CP8). None of the mice with renal abscesses yielded a mixed bacterial population, which suggests that the abscesses originated from a single microorganism seeding the kidney. Because renal abscesses formed in only nine of the infected mice, additional studies are needed to verify this supposition and to address the hypothesis that CP5-producing S. aureus strains are more nephritogenic than CP8-producing strains. We found no correlation between the ratio of the two strains recovered from the blood of individual animals given the mixed inoculum and isolation of either a CP5- or CP8-producing S. aureus strain from the kidneys after 4 days. Our phenotypic characterization of the strains indicated that Reynolds (CP5) and Reynolds (CP8) had similar growth rates in vitro.
Consistent with the mouse bacteremia data, we observed that Reynolds (CP5) was more resistant to in vitro killing in mouse blood or by human neutrophils than Reynolds (CP8) (Fig. 3). These findings suggest that CP5 production may preferentially promote bacterial survival in the bloodstream by interfering with bacterial uptake and killing by PMNs. To determine whether differences in phagocytosis by or persistence within phagocytes might contribute to the enhanced virulence of strain Reynolds (CP5), we quantified the staphylococci taken up by neutrophils after 20- and 60-min incubation periods (Fig. 4). Recovery of intracellular Reynolds (CP–) from neutrophils was greater than that of the two encapsulated strains, consistent with its increased rate of uptake and killing by phagocytic cells. In addition, there was a trend toward enhanced uptake of Reynolds (CP8) cells by PMNs compared to Reynolds (CP5), but the differences were not statistically significant. Between 0.1 and 1% of the staphylococcal inoculum survived within the phagocytes, and this percentage did not change markedly for the three S. aureus strains following an additional 2-h incubation. Gresham et al. demonstrated that staphylococci can survive within mouse PMNs and that these intracellular bacteria can contribute to persistent infection (10).
The respiratory bursts of human PMNs in response to Reynolds (CP5) and Reynolds (CP8) in NHS were similar and significantly lower than the response elicited by Reynolds (CP–) (Fig. 5A). This result is consistent with the PMNs having a greater intracellular concentration of Reynolds (CP–) compared with the encapsulated strains. The diminished response of PMNs to the encapsulated strains is unlikely to be an inhibitory effect of the capsules, since purified CP5 and CP8 stimulated a modest PMN respiratory burst similar to that elicited by teichoic acid (Fig. 5B). The control polymer polygalacturonic acid did not stimulate a respiratory burst.
Opsonization by complement is essential for phagocytosis of S. aureus by human PMNs, as none of the strains was killed by PMNs in heat-inactivated serum. Cunnion et al. demonstrated that mice depleted of complement by cobra venom factor were more susceptible to staphylococcal infection than control mice (7). Acapsular strains of S. aureus bind more C3 than encapsulated strains, although the exact mechanism for this is unclear (5, 35). To investigate the influence of CP5 and CP8 production on complement deposition, we used flow cytometry to demonstrate the kinetics and amounts of C3b and iC3b deposition on Reynolds (CP5) and Reynolds (CP8). Although the serotype 8 strain showed slightly more C3b and iC3b deposition within the initial 5 min of incubation, these subtle differences were no longer apparent by 20 min (Fig. 6). At that point, both C3b and iC3b were detected on the bacterial surface, consistent with the report by Cunnion et al. (6).
Both C3b and iC3b promote uptake of staphylococci by interacting with the CR1 (CD35) and CR3 (CD11b/CD18) receptors, respectively, on the surface of PMNs. It has been proposed that S. aureus CP renders the organism resistant to opsonophagocytic killing by masking complement molecules bound to the microbial cell wall, thus preventing ligand recognition by the phagocyte. Wilkinson et al. (47) demonstrated that complement proteins were deposited beneath the capsule layer of the highly encapsulated strain M (serotype 1). Similarly, we demonstrated by electron microscopy that C3 was localized on the cell walls beneath the capsules of S. aureus Reynolds (CP5) and Reynolds (CP8) (Fig. 7). This finding is consistent with our observation that neither purified CP5 nor CP8 activated the alternative complement pathway in NHS. Approximately 85% of Reynolds (CP5) and Reynolds (CP8) cells stained positive for capsule when analyzed by flow cytometry, an observation that is consistent with previous studies (34). Bacteria that expressed scant or no capsule on electron micrographs were characterized by more surface-associated C3 molecules than cells with abundant capsule. Although the capsule appeared to inhibit complement deposition on the staphylococcal surface, purified CP5 and CP8 did not affect complement deposition when added to Reynolds (CP–) cells in 10% NHS.
Given the differences in phagocytic uptake and killing of Reynolds (CP5) and Reynolds (CP8), despite a similar degree of opsonization, it is possible that there are differences in the ability of surface-associated CP5 and CP8 to interfere with the recognition of cell wall-bound C3 by receptors on PMNs. CP5 and CP8 are comprised of the same trisaccharide repeating units and differ only in the linkages between the sugars and sites of O acetylation. Although both CPs exhibited similar levels of O acetylation, our data indicate differences in the degree of N acetylation of the amino sugar residues that comprise CP5 and CP8. Partial de-N acetylation of CP8 results in the presence of free amino groups on the polysaccharide. To address whether this difference accounts, at least in part, for the observed differences in bacterial virulence that we have observed, we are characterizing a mutant strain carrying a deletion in the S. aureus putative FucNAc de-N-acetylase gene (Q. Cheng, A. O. Tzianabos, and J. C. Lee, unpublished studies).
Capsules from other encapsulated bacteria have been shown to have distinct biological properties; i.e., an association exists between capsule type and virulence. A Klebsiella pneumoniae isolate that produced a K2 capsule was more resistant to blood clearance than an isogenic strain that expressed a K21a capsule (12). The authors were able to correlate virulence with a capsule structure that was recognized by the mannose receptor on macrophages. Alteration of capsule type in isogenic Streptococcus pneumoniae strains had a profound effect on the virulence of some strains but not others. Expression of the type 3 capsule changed the virulence of type 5 and 6B strains but had no effect on the virulence of a type 2 strain (16). Thus, both capsule type and the genetic background of the bacterial strain influence microbial virulence.
In summary, we have created isogenic mutants of S. aureus strain Reynolds that express CP5, CP8, or no capsule. We have demonstrated that a strain expressing CP5 was more virulent than an isogenic CP8-producing strain in a mouse model of bacteremia and abscess formation. The CP5-producing strain demonstrated greater resistance to in vitro opsonophagocytic killing by neutrophils, and this likely contributes to its enhanced persistence within the mouse. However, the differences between the serotype 5 and 8 strains could not be linked to differential complement deposition, C3 processing to iC3b, or stimulation of the neutrophil oxidative burst. Additional host-bacterium interactions that we did not measure may have influenced the results of our mouse virulence studies. CP5+ and CP8+ strains may differ in their susceptibility to phagocytosis and killing within the peritoneal cavity by resident macrophages, they may exhibit variable access to the lymphatics or differential clearance by the liver and spleen, or they may vary in their activation of Toll-like receptors. Our results do confirm that CP5 and CP8 differ in a number of biological properties, and these differences likely contribute to the relative virulence of serotype 5 and 8 S. aureus in vivo.
ACKNOWLEDGMENTS
This study was supported by National Institutes of Health grant AI 29040. Danbing Ke was partially supported by the postdoctoral research fellowship from Le Fonds Quebecois de la Recherche sur la Nature et les Technologies.
We thank Robert Solinga for purifying CP8 and Linda Xie for her technical assistance. We gratefully acknowledge Ali Fattom for the AltaStaph antibodies, Chia Lee for providing pCL7657, and Sunita Gulati for mouse monoclonal antibodies to human C3b and iC3b.
A.W. and D.K. contributed equally to this work.
Present address: Bioniche Therapeutics Research Center, 6100 Royalmount Ave., Montreal, Quebec, Canada H4P 2R2.
REFERENCES
1. Albus, A., R. D. Arbeit, and J. C. Lee. 1991. Virulence of Staphylococcus aureus mutants altered in type 5 capsule production. Infect. Immun. 59:1008-1014.
2. Arbeit, R. D., W. W. Karakawa, W. F. Vann, and J. B. Robbins. 1984. Predominance of two newly described capsular polysaccharide types among clinical isolates of Staphylococcus aureus. Diagn. Microbiol. Infect. Dis. 2:85-91.
3. Baddour, L. M., C. Lowrance, A. Albus, J. H. Lowrance, S. K. Anderson, and J. C. Lee. 1992. Staphylococcus aureus microcapsule expression attenuates bacterial virulence in a rat model of experimental endocarditis. J. Infect. Dis. 165:749-753.
4. Bhasin, N., A. Albus, F. Michon, P. J. Livolsi, J.-S. Park, and J. C. Lee. 1998. Identification of a gene essential for O-acetylation of the Staphylococcus aureus type 5 capsular polysaccharide. Mol. Microbiol. 27:9-21.
5. Cotter, M. J., K. E. Norman, P. G. Hellewell, and V. C. Ridger. 2001. A novel method for isolation of neutrophils from murine blood using negative immunomagnetic separation. Am. J. Pathol. 159:473-481.
6. Cunnion, K. M., P. S. Hair, and E. S. Buescher. 2004. Cleavage of complement C3b to iC3b on the surface of Staphylococcus aureus is mediated by serum complement factor I. Infect. Immun. 72:2858-2863.
7. Cunnion, K. M., J. C. Lee, and M. M. Frank. 2001. Capsule production and growth phase influence binding of complement to Staphylococcus aureus. Infect. Immun. 69:6796-6803.
8. Fattom, A., R. Schneerson, D. C. Watson, W. W. Karakawa, D. Fitzgerald, I. Pastan, X. Li, J. Shiloach, D. A. Bryla, and J. B. Robbins. 1993. Laboratory and clinical evaluation of conjugate vaccines composed of Staphylococcus aureus type 5 and type 8 capsular polysaccharides bound to Pseudomonas aeruginosa recombinant exoprotein A. Infect. Immun. 61:1023-1032.
9. Fournier, J. M., W. F. Vann, and W. W. Karakawa. 1984. Purification and characterization of Staphylococcus aureus type 8 capsular polysaccharide. Infect. Immun. 45:87-93.
10. Gresham, H. D., J. H. Lowrance, T. E. Caver, B. S. Wilson, A. L. Cheung, and F. P. Lindberg. 2000. Survival of Staphylococcus aureus inside neutrophils contributes to infection. J. Immunol. 164:3713-3722.
11. Hochkeppel, H. K., D. G. Braun, W. Vischer, A. Imm, S. Sutter, U. Staeubli, R. Guggenheim, E. L. Kaplan, A. Boutonnier, and J. M. Fournier. 1987. Serotyping and electron microscopy studies of Staphylococcus aureus clinical isolates with monoclonal antibodies to capsular polysaccharide types 5 and 8. J. Clin. Microbiol. 25:526-530.
12. Kabha, K., L. Nissimov, A. Athamna, Y. Keisari, H. Parolis, L. A. Parolis, R. M. Grue, J. Schlepper-Schafer, A. R. Ezekowitz, D. E. Ohman, et al. 1995. Relationships among capsular structure, phagocytosis, and mouse virulence in Klebsiella pneumoniae. Infect. Immun. 63:847-852.
13. Karakawa, W. W., A. Sutton, R. Schneerson, A. Karpas, and W. F. Vann. 1988. Capsular antibodies induce type-specific phagocytosis of capsulated Staphylococcus aureus by human polymorphonuclear leukocytes. Infect. Immun. 56:1090-1095.
14. Karakawa, W. W., and W. F. Vann. 1982. Capsular polysaccharides of Staphylococcus aureus. Semin. Infect. Dis. 4:285-293.
15. Kasatiya, S. S., and J. N. Baldwin. 1967. Nature of the determinant of tetracycline resistance in Staphylococcus aureus. Can. J. Microbiol. 13:1079-1086.
16. Kelly, T., J. P. Dillard, and J. Yother. 1994. Effect of genetic switching of capsular type on virulence of Streptococcus pneumoniae. Infect. Immun. 62:1813-1819.
17. Kiser, K. B., N. Bhasin, L. Deng, and J. C. Lee. 1999. Staphylococcus aureus cap5P encodes a UDP-N-acetylglucosamine 2-epimerase with functional redundancy. J. Bacteriol. 181:4818-4824.
18. Lee, J. C. 1995. Electrotransformation of staphylococci, vol. 47. Humana Press, Totawa, N.J.
19. Lee, J. C., M. J. Betley, C. A. Hopkins, N. E. Perez, and G. B. Pier. 1987. Virulence studies, in mice, of transposon-induced mutants of Staphylococcus aureus differing in capsule size. J. Infect. Dis. 156:741-750.
20. Lee, J. C., M. J. Liu, J. Parsonnet, and R. D. Arbeit. 1990. Expression of type-8 capsular polysaccharide and production of toxic shock syndrome toxin-1 are associated among vaginal isolates of Staphylococcus aureus. J. Clin. Microbiol. 28:2612-2615.
21. Lee, J. C., F. Michon, N. E. Perez, C. A. Hopkins, and G. B. Pier. 1987. Chemical characterization and immunogenicity of capsular polysaccharide isolated from mucoid Staphylococcus aureus. Infect. Immun. 55:2191-2197.
22. Lee, J. C., J.-S. Park, S. E. Shepherd, V. Carey, and A. Fattom. 1997. Protective efficacy of antibodies to the Staphylococcus aureus type 5 capsular polysaccharide in a modified model of endocarditis in rats. Infect. Immun. 65:4146-4151.
23. Lee, J. C., S. Takeda, P. J. Livolsi, and L. C. Paoletti. 1993. Effects of in vitro and in vivo growth conditions on expression of type-8 capsular polysaccharide by Staphylococcus aureus. Infect. Immun. 61:1853-1858.
24. Lee, J. C., S. L. Xu, A. Albus, and P. J. Livolsi. 1994. Genetic analysis of type 5 capsular polysaccharide expression by Staphylococcus aureus. J. Bacteriol. 176:4883-4889.
25. Lowy, F. D. 1998. Medical progress: Staphylococcus aureus infections. N. Engl. J. Med. 339:520-532.
26. Luong, T. T., and C. Y. Lee. 2002. Overproduction of type 8 capsular polysaccharide augments Staphylococcus aureus virulence. Infect. Immun. 70:3389-3395.
27. McQuillen, D. P., S. Gulati, S. Ram, A. K. Turner, D. B. Jani, T. C. Heeren, and P. A. Rice. 1999. Complement processing and immunoglobulin binding to Neisseria gonorrhoeae determined in vitro simulates in vivo effects. J. Infect. Dis. 179:124-135.
28. Moreau, M., J. C. Richards, J. M. Fournier, R. A. Byrd, W. W. Karakawa, and W. F. Vann. 1990. Structure of the type-5 capsular polysaccharide of Staphylococcus aureus. Carbohydrate Res. 201:285-297.
29. Nilsson, I.-M., J. C. Lee, T. Bremell, C. Ryden, and A. Tarkowski. 1997. The role of staphylococcal polysaccharide microcapsule expression in septicemia and septic arthritis. Infect. Immun. 65:4216-4221.
30. O'Riordan, K., and J. C. Lee. 2004. Staphylococcus aureus capsular polysaccharides. Clin. Microbiol. Rev. 17:218-234.
31. Peng, H.-L., R. P. Novick, B. Kreiswirth, J. Kornblum, and P. Schlievert. 1988. Cloning, characterization, and sequencing of an accessory gene regulator (agr) in Staphylococcus aureus. J. Bacteriol. 170:4365-4372.
32. Peterson, P. K., B. J. Wilkinson, Y. Kim, D. Schmeling, S. D. Douglas, and P. G. Quie. 1978. The key role of peptidoglycan in the opsonization of Staphylococcus aureus. J. Clin. Invest. 61:597-609.
33. Portoles, M., K. B. Kiser, N. Bhasin, K. H. N. Chan, and J. C. Lee. 2001. Staphylococcus aureus Cap5O has UDP-ManNAc dehydrogenase activity and is essential for capsule expression. Infect. Immun. 69:917-923.
34. Poutrel, B., P. Rainard, and P. Sarradin. 1997. Heterogeneity of cell-associated CP5 expression on Staphylococcus aureus strains demonstrated by flow cytometry. Clin. Diagn. Lab. Immunol. 4:275-278.
35. Projan, S. J., and R. P. Novick. 1997. The molecular basis of pathogenicity, p. 55-81. In K. B. Crossley and G. L. Archer (ed.), The staphyloccci in human disease. Churchill Livingstone, New York, N.Y.
36. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
37. Sau, S., N. Bhasin, E. R. Wann, J. C. Lee, T. J. Foster, and C. Y. Lee. 1997. The Staphylococcus aureus allelic genetic loci for serotype 5 and 8 capsule expression contain the type-specific genes flanked by common genes. Microbiology 143:2395-2405.
38. Sau, S., and C. Y. Lee. 1996. Cloning of type 8 capsule genes and analysis of gene clusters for the production of different capsular polysaccharides in Staphylococcus aureus. J. Bacteriol. 178:2118-2126.
39. Sau, S., J. Sun, and C. Y. Lee. 1997. Molecular characterization and transcriptional analysis of type 8 capsule genes in Staphylococcus aureus. J. Bacteriol. 179:1614-1621.
40. Sompolinsky, D., Z. Samra, W. W. Karakawa, W. F. Vann, R. Schneerson, and Z. Malik. 1985. Encapsulation and capsular types in isolates of Staphylococcus aureus from different sources and relationship to phage types. J. Clin. Microbiol. 22:828-834.
41. Thakker, M., J.-S. Park, V. Carey, and J. C. Lee. 1998. Staphylococcus aureus serotype 5 capsular polysaccharide is antiphagocytic and enhances bacterial virulence in a murine bacteremia model. Infect. Immun. 66:5183-5189.
42. Tzianabos, A. O., J. Y. Wang, and J. C. Lee. 2001. Structural rationale for the modulation of abscess formation by Staphylococcus aureus capsular polysaccharides. Proc. Natl. Acad. Sci. USA 98:9365-9370.
43. Vann, W. F., M. Moreau, A. Sutton, R. A. Byrd, and W. W. Karakawa. 1988. Structure and immunochemistry of Staphylococcus aureus capsular polysaccharides, p. 187-198. In M. A. Horwitz (ed.), Bacterial-host cell interaction, vol. 64. Alan R. Liss, Inc., New York, N.Y.
44. Verbrugh, H. A., W. C. van Dijk, R. Peters, M. E. van Erne, M. R. Daha, P. K. Peterson, and J. Verhoef. 1980. Opsonic recognition of staphylococci mediated by cell wall peptidoglycan: antibody-independent activation of human complement and opsonic activity of peptidoglycan antibodies. J. Immunol. 124:1167-1173.
45. Wilkinson, B. J., Y. Kim, and P. K. Peterson. 1981. Factors affecting complement activation by Staphylococcus aureus cell walls, their components, and mutants altered in teichoic acid. Infect. Immun. 32:216-224.
46. Wilkinson, B. J., Y. Kim, P. K. Peterson, P. G. Quie, and A. F. Michael. 1978. Activation of complement by cell surface components of Staphylococcus aureus. Infect. Immun. 20:388-392.
47. Wilkinson, B. J., S. P. Sisson, Y. Kim, and P. K. Peterson. 1979. Localization of the third component of complement on the cell wall of encapsulated Staphylococcus aureus M: implications for the mechanism of resistance to phagocytosis. Infect. Immun. 26:1159-1163.
48. Xu, S., R. D. Arbeit, and J. C. Lee. 1992. Phagocytic killing of encapsulated and microencapsulated Staphylococcus aureus by human polymorphonuclear leukocytes. Infect. Immun. 60:1358-1362.(Andrew Watts, Danbing Ke,)