Characterization of the Kingella kingae Polysaccharide Capsule and Exopolysaccharide

Recent evidence indicates that Kingella kingae produces a polysaccharide capsule. In an effort to determine the composition and structure of this polysaccharide capsule, in the current study we purified capsular material from the surface of K. kingae strain 269–492 variant KK01 using acidic conditions to release the capsule and a series of steps to remove DNA, RNA, and protein. Analysis of the resulting material by gas chromatography and mass spectrometry revealed N-acetyl galactosamine (GalNAc), 3-deoxy-D-manno-oct-2-ulosonic acid (Kdo), and galactose (Gal). Further analysis by NMR demonstrated two distinct polysaccharides, one consisting of GalNAc and Kdo with the structure →3)-β-GalpNAc-(1→5)-β-Kdop-(2→ and the other containing galactose alone with the structure →5)-β-Galf-(1→. Disruption of the ctrA gene required for surface localization of the K. kingae polysaccharide capsule resulted in elimination of GalNAc and Kdo but had no effect on the presence of Gal in bacterial surface extracts. In contrast, deletion of the pamABCDE locus involved in production of a reported galactan exopolysaccharide eliminated Gal but had no effect on the presence of GalNAc and Kdo in surface extracts. Disruption of ctrA and deletion of pamABCDE resulted in a loss of all carbohydrates in surface extracts. These results establish that K. kingae strain KK01 produces a polysaccharide capsule with the structure →3)-β-GalpNAc-(1→5)-β-Kdop-(2→ and a separate exopolysaccharide with the structure →5)-β-Galf-(1→. The polysaccharide capsule and the exopolysaccharide require distinct genetic loci for surface localization.


Introduction
Kingella kingae is a gram-negative bacterium that has emerged as a common cause of septic arthritis, osteomyelitis, and bacteremia in children 6-36 months of age [1]. Studies using culture-based and molecular-based detection strategies have established that K. kingae is a common commensal in the upper respiratory tract in young children [1][2][3][4]. In one report, K. kingae colonization of the pharynx was found in approximately 70% of children at some point during the first two years of life [5]. The pathogenesis of K. kingae disease is believed to involve bacterial translocation across the pharyngeal epithelial barrier, entry into the bloodstream, and dissemination to the joints, bones, or endocardium [4]. This pathogenic model is supported by data demonstrating genotypically identical isolates of K. kingae from the upper respiratory tract and either the bloodstream or joint fluid from patients with invasive disease [6,7]. Extracellular polysaccharides are produced by many pathogenic bacteria and function as important virulence factors. These polysaccharides can be categorized generally as capsular polysaccharides or exopolysaccharides. Capsular polysaccharides are lipid-anchored outer membrane-associated carbohydrates and are involved in protection from host immune mechanisms, including phagocytosis and complement-mediated killing [8,9]. These polysaccharides often exhibit a large degree of variation within a single bacterial species, with examples including Escherichia coli and Streptococcus pneumoniae [10][11][12]. In select cases, a specific capsule type can be associated with increased virulence potential, as highlighted by Haemophilus influenzae type b [13]. Exopolysaccharides are secreted carbohydrate polymers that are not anchored to the bacterial surface and that play a variety of roles, including modulation of adherence and modulation of biofilm formation [14,15].
In recent work, we identified a polysaccharide capsule in K. kingae strain 269-492 that influences adhesive interactions with host cells and is dependent upon the ctrABCD ABC-type capsule export operon for surface localization [16]. In the current study, we set out to characterize the composition and structure of this polysaccharide capsule. Glycosyl analysis of bacterial surface extracts by gas chromatography/mass spectrometry and NMR revealed two distinct extracellular polysaccharides, including the polysaccharide capsule and a separate exopolysaccharide. The polysaccharide capsule contains Nacetyl galactosamine (GalNAc) and 3-deoxy-D-manno-oct-2ulosonic acid (Kdo) and has the structure 3R)-b-GalpNAc-(1R5)-b-Kdop-(2R. In contrast, the exopolysaccharide contains only galactose and has the structure R5)-b-Galf-(1R. Targeted mutagenesis established that the capsular polysaccharide and the exopolysaccharide require separate genetic loci for surface localization.

Bacterial Strains
The strains used in this study are listed in Table 1. K. kingae strain 269-492 was originally recovered from the joint fluid of a child with septic arthritis at St. Louis Children's Hospital, St. Louis, MO. K. kingae KK01 is a stable sparsely piliated natural variant of strain 269-492 that forms nonspreading, non-corroding colonies and was used in all experiments in this study because of its colony morphology [17]. K. kingae strains were grown on chocolate agar plates at 37uC with 5% CO 2 supplemented with 50 mg/ml kanamycin or 2 mg/ml erythromycin as appropriate. K. kingae strains were stored at 280uC in brain heart infusion with 30% glycerol. E. coli strain DH5a was used for construction of gene disruption plasmids. E. coli strains were routinely grown at 37uC on Luria-Bertani (LB) agar or in LB broth with 100 mg/ml ampicillin, 50 mg/ml kanamycin, or 500 mg/ml erythromycin as appropriate. E. coli strains were stored at 280uC in LB broth with 15% glycerol.

Strain Construction
Targeted gene disruptions in K. kingae were generated as previously described [16,18]. Briefly, plasmid-based gene disruption constructs were created in E. coli, linearized, and introduced into K. kingae strain KK01 via natural transformation. Transformants were recovered by plating on chocolate agar plates containing the appropriate antibiotic. Correct localization of gene disruptions was confirmed by PCR. The ctrA disruption was generated as described previously [16] but with the aphA3 kanamycin resistance cassette as the marker. This disruption had a polar effect on the downstream genes in the ctrABCD operon. To delete the pamABCDE locus, fragments corresponding to the surrounding 59 and 39 regions of the locus were PCR amplified using the primers pam 59for (GCGAATTCGGCGTTGGTG-GAATATCCTG), pam 59rev (GCGGATCCACCTTCTGGT CGCTGAAATG), pam 39for (GCGGATCCTCAAAGGCTGG-TATAAACAC), and pam 39rev (GCAAGCTTCCATATCGC TTTGGCTTTGC), respectively. The resulting fragments were ligated into pUC19, creating puC19pam 59+39::BamHI. The ermC erythromycin cassette from pIDN4 was PCR amplified with flanking BamHI sites and ligated into pUC19pam 59+39, generating pUC19pam::ermC.

Polysaccharide Extraction and Purification
For small scale capsule extractions, bacteria were washed and resuspended in either Tris acetate pH 5 (acid treatment for one hour) or PBS (heat treatment at 55uC for 1 hour). Cells were removed by centrifugation, and extracts were treated with proteinase K for one hour, and then concentrated as previously described [16].
To prepare strains for extraction of extracellular polysaccharide for purification, lawns of bacteria were grown overnight on chocolate agar plates. Subsequently, bacterial growth was swabbed from plates and resuspended in 50 ml of 1% formaldehyde in PBS to fix cells and allowed to stand for 15 minutes at room temperature. Bacteria were centrifuged at 4,3556g for 10 minutes and then resuspended in 40 ml of 50 mM Tris acetate pH 5. Following vigorous shaking for 1 hour at room temperature, bacteria were pelleted by centrifugation at 12,0966g for 20 minutes. The supernatant was recovered and was then filtered using a 0.22 mm filter. The filtered material was adjusted to pH 7 with 1 M Tris pH 9. To remove contaminating DNA, RNA, and protein, the filtered material was treated with 10 units of DNase (Fermentas) and 0.1 mg of RNase (Fermentas) at 37uC for 5 hours and then with 0.18 mg of proteinase K (Roche) at 55uC overnight. Samples were concentrated to 500 ml using 100 kDa MWCO filters and extracted once with Tris-saturated phenol pH 7.4 and twice with 100% chloroform. Extracted material was dialyzed extensively overnight in deionized water, flash frozen, and lyophilized. To purify exopolysaccharide galactan, we resuspended bacteria in PBS without formaldehyde, shook for one hour, pelleted the bacteria, and subjected the supernatant to the purification described above.

Staining of Polysaccharide
Aliquots of the purified polysaccharide from K. kingae derivatives were separated on 10% SDS-PAGE gels. For Alcian blue staining, gels were stained with 0.125% Alcian blue as previously described [16]. For silver staining, gels were treated as previously described [19].

Chemical Analysis
Monosaccharide composition analysis was performed by methanolysis and trimethylsilyl derivatization as previously described [20].

NMR Spectroscopy
The isolated polysaccharide was deuterium-exchanged by lyophilization from D 2 O (99.9%D, Aldrich), dissolved in 270 ml D 2 O (99.96%D, Cambridge Isotope) containing 0.3 ml acetone, and placed in a 5-mm NMR tube with D 2 O-matched magnetic susceptibility plugs (Shigemi Inc.). One-dimensional proton and 2-D TOCSY and NOESY NMR spectra with solvent presaturation and gradient-enhanced COSY, HSQC, and HMBC NMR spectra

Results
Polysaccharide can be Extracted and Purified from the Surface of K. kingae Strain 269-492 Variant KK01 In order to extract and purify the polysaccharide capsule, we incubated a bacterial suspension of K. kingae strain 269-492 variant KK01 in Tris acetate pH 5, aiming to dissociate the polysaccharide capsule from the bacterial surface by the acidic pH. To confirm that our extraction procedure yielded pure polysaccharide material, we examined the material by Alcian blue straining and silver staining ( Figure 1). Alcian blue is a cationic dye and has been used previously to demonstrate the presence of the K. kingae capsule [16]. As shown in Figure 1, the extracted and purified material from the surface of K. kingae KK01 stained prominently with Alcian blue and silver reagents (lane 3), similar to control small scale heat and acid extracts from K. kingae KK01 (lanes 1-2). The Tris acetate extract from KK01ctrA (lacking a functional capsule export locus) failed to stain with either Alcian blue or silver reagents (lane 5), while the PBS surface extract from KK01ctrA stained well with Alcian blue and silver reagents (lane 6).

Kingella kingae KK01 Extracellular Polysaccharide Contains GalNAc, Kdo, and Galactose
As shown in Table 2, composition analysis of the polysaccharide extracted from K. kingae KK01 revealed that the dominant components were GalNAc, Kdo, and galactose in roughly equimolar quantities. In addition, there were small quantities of xylose, glucose, N-acetyl glucosamine, and heptose, all common contaminants in this kind of preparation.
The 1-D proton NMR spectrum of the purified polysaccharide ( Figure 2) showed two anomeric signals at 5.18 and 4.70 ppm, with an intensity ratio of 1:3, and two upfield methylene signals at 2.40 and 1.77 ppm, each of equal intensity with the larger anomeric peak. The HSQC spectrum (Figure 3) showed that the signal at 5.18 ppm was associated with a carbon chemical shift of 109.4 ppm, implying that this residue was a furanose. Chemical shift assignment of the remaining nuclei (Table 3) demonstrated that the furanose had a galacto-configuration, and a 5.9-ppm downfield displacement of its C-5 chemical shift indicated that it was glycosylated in the 5-position. The further chemical shift assignments, based on 2-D COSY, TOCSY, and HSQC spectra, showed that the anomeric signal at 4.70 ppm belonged to a b-GalNAc residue and the methylene signals at 2.40 and 1.77 ppm belonged to a b-Kdo residue. Downfield shifts of the carbon nuclei involved in glycosidic linkages indicated that GalNAc was 3-linked and Kdo was 5-linked (see Table 3). Beside these major residues, small amounts (,5%) of non-reducing end GalNAc and reducing end 5-a-Kdo were also detected and are likely the result of cleavage of some of the very acid labile Kdo glycosidic bonds during isolation. The NOESY spectrum ( Figure 4) revealed NOE contacts between H-1 of GalNAc and H-5 of Kdo and between both H-3 s of Kdo and H-3 of GalNAc. The Galf anomeric signal gave NOE contacts only to its own protons. HMBC showed a correlation between H-1 of GalNAc and C-5 of Kdo.
Taken together, these results demonstrated that the sample contained two polysaccharides, one composed of a disaccharide repeating unit with the structure R3)-b-GalpNAc-(1R5)-b-Kdop-(2R and the other a galactofuranosyl homopolymer (galactan) with the structure R5)-b-Galf-(1R ( Figure 5). The galactan homopolymer accounted for the galactose in the composition and methylation analyses and was present at about one-third the molar concentration of the GalNAc-Kdo polymer.

The GalNAc-Kdo Polymer is the Polysaccharide Capsule and the Galactan Homopolymer is an Exopolysaccharide
In a recent report, K. kingae strain PYKK181 was found to produce an exopolysaccharide that is a galactan homopolymer and is encoded by the pamABCDE locus [22]. To elucidate the relationship between the K. kingae polysaccharide capsule and the two distinct polysaccharide structures present in our bacterial surface extracts, we began by creating a derivative of KK01 with a deletion of pamABCDE. As shown in Figure 1 (lane 4) and Table 2, the polysaccharide extract from KK01pamABCDE stained with Alcian blue, contained large amounts of GalNAc and Kdo, and had only trace quantities of galactose, suggesting that the R5)-b-Galf-(1R homopolymer corresponds to the galactan exopolysaccharide present in K. kingae strain PYKK181. As shown in Figure 6, examination of the colony morphology of KK01pamABCDE revealed mucoid colonies, indicating encapsulation and suggesting that the GalNAc-Kdo polysaccharide corresponds to the polysaccharide capsule. To confirm these conclusions, we examined polysaccharide extracts from strain KK01ctrA, recognizing that the ctrA gene is required for export and surface localization of the polysaccharide capsule. Interestingly, glycosyl composition analysis revealed no GalNAc or Kdo and minimal amounts of galactose compared to the quantities of contaminating xylose and glucose (Table 2), raising the possibility that either the absence of the GalNAc-Kdo polysaccharide results in free release of the galactan exopolysaccharide or that ctrA is essential for surface localization of both the GalNAc-Kdo polysaccharide and the galactan homopolymer. To address these possibilities, we modified the extraction procedure to eliminate both the initial fixation step and the Tris acetate treatment, aiming to recover loosely associated polysaccharide from strain KK01ctrA. Using this technique, a large quantity of galactose was detected in the purified material and only minimal amounts of GalNAc and Kdo were present (probably reflecting some bacterial lysis) ( Table 2 Table 2, and Figure 6). Together these results provide strong evidence that the GalNAc-Kdo polysaccharide is the polysaccharide capsule and that the galactan polymer is an exopolysaccharide.

Discussion
In previous work we showed that K. kingae produces a polysaccharide capsule and contains a capsule export locus designated ctrABCD with high homology to the capsule export locus in N. meningitidis. In addition, we demonstrated that the K. kingae capsule is associated with a mucoid colony phenotype on chocolate agar and is visible by thin section transmission electron    Table 3 (C3e and C3a designate the equatorial and axial proton, respectively, of the C-3 methylene group). doi:10.1371/journal.pone.0075409.g002 microscopy after staining with cationic ferritin [16]. In the present study we set out to characterize the composition and structure of the K. kingae polysaccharide capsule. As a first step, we developed a capsule extraction and purification procedure that generated material amenable to analysis by GC/MS and NMR. Examination of the extracted material from K. kingae strain KK01 revealed GalNAc, Kdo, and galactose in roughly equimolar amounts. Additional analysis of the extracted material from derivatives of strain KK01 containing targeted mutations in the capsule export locus, the galactan biosynthesis locus, or both loci established that GalNAc and Kdo make up the capsule and that galactose makes up a galactan exopolysaccharide. In considering our results, it is noteworthy that Bendaoud et al. reported two polysaccharides in the supernatant from K. kingae strain PYKK181, including a polysaccharide with the structure R6)-a-D-GlcpNAc-(1R5)-b-Kdop-(2R and a galactan homopolymer with the structure R3)-b-Galf-(1R6)-b-Galf-(1R that was designated poly-DNA-containing anti-adhesive material extract (PAM galactan) and was found to inhibit biofilm formation by a variety of organisms [22]. In the report by Bendaoud and coworkers, the specific relationship between these two polysaccharides and the organism was unclear. Based on the results of our mutational studies, we believe that the GlcNAc-Kdo polymer from strain PYKK181 is a polysaccharide capsule and that the PAM galactan is an exopolysaccharide.
It is interesting that the strain KK01 polysaccharide capsule has the structure R3)-b-GalpNAc-(1R5)-b-Kdop-(2R and the strain PYKK181 polysaccharide capsule has the structure R6)-a-D-GlcpNAc-(1R5)-b-Kdop-(2R [22]. The fact that the polysaccharide capsules expressed by these strains have different structures raises the possibility that there may be multiple K. kingae capsule types, similar to the case with E. coli, S. pneumoniae, N. meningitidis, H. influenzae, and a number of other bacterial pathogens [10,11,23]. Along these lines, it is notable that Amit et al. observed an association between K. kingae PFGE clonal types and specific clinical presentations [24]. In ongoing work, we are examining  Table 3. doi:10.1371/journal.pone.0075409.g003 whether there is variability in capsule type among different clones and conservation of capsule type within a given clone. It is possible that certain capsule types confer greater virulence than others and are associated with specific disease processes.
In considering the genetic determinants of K. kingae polysaccharide capsule biosynthesis, it is notable that the ctrABCD capsule export locus is not flanked by a capsule synthesis operon, thus differing from the configuration of capsule genes in most other bacteria that produce a polysaccharide capsule [8,9,31]. Based on available K. kingae genome data, there is no apparent capsular polysaccharide synthesis operon. We speculate that the genes required for synthesis of the capsular polysaccharide are not arranged in an operon and instead are located at separate loci throughout the genome.
Sequence analysis of the PAM locus in K. kingae strain KK01 and the PAM locus in K. kingae strain PYKK 181 reveals almost complete identity. Despite this sequence conservation, the structures of the galactan homopolymers in these strains are different, with distinct linkages. One possibility is that there are other genes beyond pamABCDE involved in biosynthesis of the galactan exopolysaccharides. Alternatively, the limited sequence variation may result in significant changes in enzymatic specificity.
To summarize, in this study we have established that K. kingae expresses two extracellular polysaccharides that require distinct genetic loci for surface localization, including a polysaccharide capsule and an exopolysaccharide. In future studies we will examine the contribution of these polysaccharides to the pathogenesis of K. kingae disease.