ABSTRACT
Streptococcus pneumoniae (the pneumococcus) colonizes the human nasopharynx and can cause invasive disease aided by the pneumococcal capsule. Group II nontypeable S. pneumoniae (NTSp) lacks a polysaccharide capsule, and a subgroup of NTSp carriage isolates has been found to have a novel gene, pneumococcal surface protein K (pspK), which replaces the capsule locus. A recent rise in the number of NTSp isolates colonizing the human nasopharynx has been observed, but the colonization factors of NTSp have not been well studied. PspK has been shown to play a role in mouse colonization. We therefore examined PspK-mediated immune evasion along with adherence to host cells and colonization. PspK bound human secretory immunoglobulin A (sIgA) but not the complement regulator factor H and did not decrease C3b deposition on the pneumococcal surface. PspK increased binding of pneumococci to epithelial cells and enhanced pneumococcal colonization independently of the genetic background. Understanding how NTSp colonizes and survives within the nasopharynx is important due to the increase in NTSp carriage. Our data suggest that PspK may aid in the persistence of NTSp within the nasopharynx but is not involved in invasion.
INTRODUCTION
Streptococcus pneumoniae (the pneumococcus) is a common colonizer of the human nasopharynx, with 95% of children in the United States being carriers at least once during the first 24 months of life (1). The pneumococcus is an important cause of bacterial pneumonia, meningitis, and otitis media, with nasopharyngeal colonization being a prerequisite for disease development (2, 3). A major pneumococcal virulence factor mediating disease processes is the capsular polysaccharide, which serves as the basis of the current polysaccharide and conjugate vaccines (4–6). Introduction of the 7-valent protein-polysaccharide conjugate vaccine (PCV7; Prevnar; Wyeth Lederle Vaccines) has led to an increased incidence of pneumococcal disease caused by serotypes not contained within PCV7 (7, 8). Rates of pneumococcal carriage have not decreased since the introduction of PCV7 because nonvaccine serotypes now colonize at a greater rate than before the introduction of PCV7 (9, 10). Nontypeable S. pneumoniae (NTSp) lacks capsule and is not targeted by the currently available vaccine, which may have contributed to the increase in NTSp carriage since the introduction of the conjugate vaccine (11, 12). This increase in colonization and dissemination of NTSp in the population allows a greater opportunity for horizontal transmission of virulence factors and antibiotic resistance.
NTSp isolates have been classified into group I or group II on the basis of their cps locus, with group I NTSp having a conventional cps locus that is too disrupted to produce a functional capsule and group II NTSp lacking the conventional cps locus but containing aliB-like open reading frames (12, 13). We previously characterized the pneumococcal capsule locus from 52 group II NTSp carriage isolates. Six of the isolates were found to have a novel gene, pspK, within the cps locus. Further, an isolate expressing pneumococcal surface protein K (PspK) was able to colonize the mouse nasopharynx as efficiently as an encapsulated pneumococcus (12). The mechanism of colonization of NTSp in the nasopharynx has not been well characterized, but PspK seems to aid in colonization by an undetermined mechanism. Due to pneumococcal typing based on serotype, the prevalence of NTSp in the population may have been underestimated, explaining why NTSp has not been well studied in respect to colonization factors. In the absence of capsule, colonization is typically decreased. For NTSp strains to colonize as efficiently as encapsulated strains, there must be a compensatory mechanism, such as a surface protein like PspK (14, 15). Pneumococcal surface proteins have been shown to increase colonization and mediate immune evasion, leading to the development of disease (16–18). Evidence for the importance of these proteins in NTSp colonization is currently lacking, but in nontypeable Haemophilus influenzae (NTHI), surface proteins have been shown to be important in nasopharyngeal adherence and colonization (19, 20). NTSp has been reported to be the causative agent of acute otitis media (AOM), and because colonization is a prerequisite for AOM, an increase in colonization can lead to more cases of pneumococcal disease caused by NTSp (21).
The pspK gene was submitted to a GenBank BLAST search for sequence comparison which identified the most homologous gene in the database to be pneumococcal surface protein C (PspC), with ∼15% nucleotide identity (12). The pspK gene has no homology to any other known bacterial genes in the GenBank database; therefore, we focused our investigation on possible similarities between PspC and PspK. PspC is a multifunctional pneumococcal virulence factors that aids in pneumococcal colonization and evasion of the host innate immune system through the interaction with host polymeric Ig receptor (pIgR), secretory IgA (sIgA), and by recruitment of factor H (FH) to the bacterial surface (22). The binding of PspC to pIgR has been thought to increase invasiveness due to pneumococcal translocation mediated by the pIgR, but pIgR has been shown not to be the sole receptor for pneumococcal binding and invasion (23, 24). PspC interaction with sIgA and FH aids in reducing nasopharyngeal clearance and clearance from the blood, respectively, increasing the overall persistence of the pneumococcus in the host (25–28). PspK may also act upon the innate immune system in a manner similar to that for PspC. We previously noted a YPT motif in PspK that has been shown to be vital in the binding of sIgA, increasing the likelihood that PspK functions in some ways in a manner homologous to that of PspC (12, 29). We examined the potential of PspK to affect immune evasion and epithelial cell adherence.
PspK mediates adherence to human epithelial cells in vitro and enhances colonization in a mouse model. Also, PspK binds sIgA but fails to bind FH. Epithelial cell adherence is independent of the genetic background of the pneumococcal isolate. The pneumococcus employs various mechanisms to persist within the nasopharynx. PspK may serve an important function for NTSp to colonize and persist within the human nasopharynx, imparting a reason to study such proteins, in that colonization leads to dissemination and is necessary for pneumococcal disease (3).
MATERIALS AND METHODS
Bacteria.Table 1 shows the bacterial strains and selective markers used. All pneumococcal strains were grown at 37°C in 5% CO2 either on sheep blood agar (BA) or in Todd-Hewitt medium with 0.5% yeast extract (THY) with appropriate antibiotic selection. Genomic and plasmid DNA was obtained with a DNeasy blood and tissue kit (Qiagen) and a plasmid minikit (Qiagen), respectively. Bacterial lysates for SDS-PAGE were prepared by suspending 1 ml of pelleted sample in 100 μl lysis buffer (0.01% sodium dodecyl sulfate, 0.1% sodium deoxycholate, 0.015 M sodium citrate), incubating the mixture at 37°C for 30 min, and then diluting the mixture with 100 μl phosphate-buffered saline (PBS).
S. pneumoniae strains used in this study
DNA manipulation and immunization.Recombinant PspK (rPspK) was produced by PCR amplification of MNZ11 genomic DNA using primers LSM800 (5′-CACCGCAAATCAGCCAGTAACTGTGA-3′) and LSM571 (5′-CTAATTTTTATGTTTAACAAATGGAAGA-3′) and cloning into pET100 (Invitrogen) using the manufacturer's protocols. Purification was performed with a B-PER 6× His fusion protein purification kit (Thermo Scientific) and with the addition of a third wash (wash buffer 2 with 80 mM instead of 25 mM imidazole), followed by gel filtration on Sephacryl 100-HR. A juvenile New Zealand White rabbit was injected with 100 μg rPspK emulsified with TiterMax Gold adjuvant and boosted every 3 weeks for 9 weeks. Enzyme-linked immunosorbent assay was performed 1 week after the last boost to determine the anti-PspK antibody titer before collecting the serum.
For expression in the pneumococcus, PspK was PCR amplified from MNZ11 genomic DNA with primers LSM826 (5′-CCCGGGGCATGAATAATAAGAATATCA-3′) and LSM827 (5′-GAATTCGCCTAATTTTTATGTTTAACAAATG-3′). The amplified product was cloned into the TOPO TA plasmid (Invitrogen) following the manufacturer's protocol, producing pCR2.1::PspK. The plasmid was purified and digested with XmaI and EcoRI (New England Biolab). Following gel purification of the insert with QIAquick gel extraction (Qiagen), PspK was cloned into the pneumococcal expression vector pNE1, resulting in pNE1::PspK (30).
Pneumococci were transformed with pNE1::PspK in competence medium with bacteria stimulated by 200 ng of competence-stimulating peptide 1 (CSP1) (31). Bacteria were incubated with DNA for 4 h and plated on BA with 300 μg/ml of spectinomycin. Single colonies were isolated and grown in THY containing 300 μg/ml of spectinomycin and stored at −80°C after the addition of 20% glycerol.
Interaction of PspK with human proteins.Human factor H (provided by M. Pangburn) and sIgA (provided by S. Hammerschmidt) were biotinylated using EZ-Link sulfo-NHS-LC-biotin (Thermo Scientific) following the manufacturer's protocol. Western blot analysis was performed as previously described (32). Proteins and antibodies were detected with streptavidin conjugated to horseradish peroxidase and visualized by SuperSignal West Pico chemiluminescent substrate (Pierce).
For flow cytometry, bacterial strains in log phase were incubated with either protein or antibody and fluorescently labeled by direct biotinylation of either protein or biotinylated secondary antibody bound to streptavidin conjugated to Alexa Fluor 488 (AF488). Bacteria were washed in PBS and subjected to flow cytometry using a Becton Dickinson flow cytometer with forward and side scatter parameters to gate on at least 10,000 bacteria. Data are expressed as the fluorescence index (FI; defined as the proportion of bacteria positive for binding multiplied by the geometric mean fluorescent intensity), as has been utilized previously (33, 34).
Bactericidal assay.Pneumococcal survival in mouse whole blood was performed by inoculating 1 × 105 CFU/ml of bacteria into 150 μl of fresh heparinized CBA/N mouse blood, which lacks natural antiphosphocholine antibodies but which is complement competent and can clear streptococcal infection (35). The blood was mixed on a rotating wheel for 3 h at 37°C. Samples were taken at 1 min postinoculation and plated to obtain the bacterial CFU counts at time zero. At 3 h postinoculation, samples were taken and counts were performed. Two independent experiments were performed, with each experiment performed in triplicate.
Adherence and invasion assay.Bacterial adhesion and invasion were tested by culturing human A549 and Detroit 562 epithelial cells in 10% fetal calf serum (FCS)-supplemented Dulbecco modified Eagle medium (DMEM) and Eagle's minimal essential medium (EMEM), respectively, to 90% confluency in 24-well plates at 37°C in 5% CO2. Then, 1 × 107 bacteria in 1 ml of the appropriate culture medium were incubated for 30 min for adherence or 2 h for invasion and washed with diluted PBS for A549 cells or EMEM for Detroit 562 cells. Adherence was tested by trypsinizing (100 μl 0.25% trypsin-EDTA) cells, adding 900 μl PBS, and plating on BA for determination of the numbers of CFU/ml. Invasion assay mixtures were further incubated for 1 h with penicillin (10 μg/ml) and gentamicin (200 μg/ml), followed by 3 more washes. Cells were lysed with 100 μl 0.0125% Triton X-100, followed by plating on BA. Each experiment had at least three biological replicates for each strain tested.
Biofilm formation.Production of biofilm was determined by incubating strains in 1 ml of THY containing catalase and 10% horse serum in 24-well plates. Plates were incubated for 24 h, followed by complete removal of the medium and staining with 0.1% crystal violet for 30 min at room temperature. Crystal violet was removed, the remaining crystal violet was solubilized in 350 μl 100% ethanol, and the absorbance at 540 nm was read on an xMark microplate reader.
Mouse nasopharynx colonization.Eight-week-old C57/BL6 mice were used for colonization studies. Mice were anesthetized with isoflurane and nasally infected by placing 10 μl of Ringer's lactate solution containing 1 × 107 CFU of bacteria in the nasal passages. After 5 days, the mice were sacrificed and the nasopharynx was washed with 200 μl of Ringer's lactate solution. Nasal wash samples were plated on BA with 5 μg/ml gentamicin, and the number of CFU of colonizing bacteria was determined. All strains were tested in at least two separate experiments.
Statistical analysis.Flow cytometry significance was determined by Student's t test of FI with the InStat program (GraphPad Software). Results from adhesion and invasion assays and bactericidal assay were determined by Student's t test with the InStat program (GraphPad Software). For colonization, the numbers of CFU in the different experimental groups were compared using the Mann-Whitney test with the InStat program. Significant results are indicated by P values of <0.05.
RESULTS
Binding of host immune factors.Flow cytometry with anti-PspK antibody demonstrated that PspK is a surface-expressed pneumococcal protein (data not shown). The R1 and R2 regions of PspC share ∼70% amino acid sequence homology to a similar region in PspK (12). The R1 and R2 regions have not been shown to bind FH, but other regions of the protein could be responsible for FH binding, so we examined the potential role of PspK in complement evasion mediated by FH, as has been described for PspC (36–38). Western blot analysis showed that recombinant PspC (rPspC) and PspC-containing strain D39 bind FH, as previously seen (Fig. 1A and B) (22). In contrast, rPspK failed to bind FH (Fig. 1A and B). These findings were confirmed by flow cytometry of intact pneumococci incubated with biotinylated FH and visualized using streptavidin-conjugated AF488, with data being presented as the average fluorescence index of three independent experiments with standard error. MNZ11 (PspK+) and MNZ1131 (PspK−) had negligible changes in FH recruitment, with FI values of 47.06 ± 16.14 and 83.63 ± 20.72, respectively (P > 0.05). As previously reported, D39 had a greater percentage of cells with FH binding (37.3%), and its PspC− mutant ΔPAC had reduced FH recruitment (11.3%) (Fig. 1B and data not shown). The negative-control ΔPAC also lacks pneumococcal surface protein A (PspA) and pneumolysin (Ply), neither of which has been implicated in binding any of the immune factors tested (39, 40). PspC has also been demonstrated to bind the secretory component of human sIgA through a specific YPT motif, which is also found in PspK (12, 41). Western blot analysis and flow cytometry with biotinylated sIgA demonstrated that PspK mediated binding to sIgA. S. pneumoniae D39 displayed an FI of 329.12 ± 60.88 when determining sIgA binding, while MNZ11 (PspK+) had an FI of 492.65 ± 53.75 and MNZ1131 (PspK−) had an FI of 64.55 ± 9.35 (Fig. 1C and D and data not shown). The FI values for binding to sIgA did not significantly differ between strains D39 and MNZ11, but D39 displayed significantly greater sIgA binding than MNZ1131 (P < 0.01).
PspK binds human sIgA but not FH. (A) Western blot of biotinylated FH shows binding to rPspC but not rPspK; (B) flow cytometry demonstrating that FH is recruited to the surface of strain D39 (PspC+) but not strain MNZ11 (PspK+); (C) Western blot of biotinylated sIgA binding to rPspK and rPspC; (D) flow cytometry demonstrating binding of sIgA to strains D39, MNZ11, and MNZ1131.
Complement deposition.Based on the homology between PspC and PspK, we investigated whether PspK conferred immune evasion similar to what is seen in PspC+ pneumococci. It is known that there is a decrease in the deposition of complement component C3b on the bacterial surface when pneumococcal strains express PspC (22, 26). Although we have shown that PspK does not mediate FH binding, other mechanisms could be responsible for decreased complement deposition. Incubation of pneumococcal strains with normal human serum led to an FI of 1,087.45 ± 233.19 for C3b deposition on the bacterial surface of PspC− strain ΔPAC and an FI of 194.33 ± 15.41 for C3b deposition on parent strain D39 (PspC+). In contrast, MNZ11 (PspK+) had an FI of 1,019.23 ± 264.22 for C3b deposition, whereas MNZ1131 (PspK−) had an FI of 974.02 ± 250.75 for C3b deposition. These results indicate no relevant increase in C3b deposition with the absence of PspK (P > 0.05) (Fig. 2).
PspK does not reduce complement C3b deposition on the surface of the pneumococcus. After incubation with normal human serum, C3b deposition was measured using a biotinylated monoclonal anti-C3b antibody. The experiment was repeated three times, and a representative histogram is shown.
Bactericidal assay.A mouse blood bactericidal assay was used to determine if PspK protects NTSp from phagocytosis. Survival of MNZ11 (PspK+) in mouse whole blood was compared to that of MNZ1131 (PspK−). A 3-h incubation led to a significant decrease (P < 0.05) in pneumococcal numbers in all nonencapsulated strains tested, MNZ11, MNZ1131, and R36A. Bacterial clearance did not significantly differ (P > 0.05) among the nonencapsulated strains tested, while D39 had no significant reduction in bacterial numbers after incubation (Fig. 3). Encapsulated pneumococci, such as D39, are not cleared from CBA/N mouse blood, on the basis of the results of previous studies, and accordingly, D39 survived significantly better (P < 0.05) than MNZ11, MNZ1131, or R36A (42).
PspK does not decrease bactericidal activity of mouse whole blood. After 3 h incubation in whole blood, all nonencapsulated strains, MNZ11, MNZ1131, and R36A, had a reduction in the numbers of CFU recovered of ∼3 log units. There was no significant difference in strain D39 counts after 3 h, and there was no significant difference among the other three strains at the 3-h time point. Error bars indicate SEs. *, significant difference of P < 0.001.
Adherence and invasion of epithelial cell lines.We previously demonstrated a significant increase in mouse nasopharyngeal carriage for a PspK+ pneumococcal isolate (12). To assess the potential interaction with human cells, we incubated pneumococci with either the human lung epithelial cell line A549 or the human nasopharyngeal epithelial cell line Detroit 562. The absence of PspK resulted in a significant decrease in the number of pneumococci adhering to epithelial cells (Fig. 4A and C). MNZ1131 (PspK−) had a greater than 50% decrease in adherence compared to that for MNZ11 (PspK+), demonstrating that PspK is an important factor for epithelial adherence.
PspK increases pneumococcal adherence to but not invasion of human epithelial cells. (A and B) Pneumococcal adhesion to and invasion of A549 lung epithelial cells; (C and D) pneumococcal adhesion to and invasion of Detroit 562 nasopharyngeal epithelial cells. Bacteria recovered after each experiment were normalized to MNZ11 set at 100% binding. *, significant difference of P < 0.0001. Values were obtained from triplicate wells infected with indicated strain.
We theorized that an increase in adherence to the epithelial cell surface may also increase the proportion of pneumococci that invades epithelial cells. Plating of epithelial cell lysates after incubation with pneumococcal strains indicated no significant difference in intracellular invasion of MNZ11 compared to that of MNZ1131 (P > 0.05). These results suggest that PspK enhances binding to human epithelial cells but does not appear to facilitate invasion (Fig. 4B and D).
Enhanced colonization could be facilitated through the formation of bacterial biofilm on the epithelial cell surface (43). Testing for the presence of biofilm indicated no increase in biofilm formation for either MNZ11 (PspK+) or MNZ1131 (PspK−), but the two strains did produce significantly more biofilm than biofilm-producing strain Tigr4 (data not shown).
R36A expression of PspK leads to increased adhesion.Utilizing the Escherichia coli-pneumococcus shuttle expression vector pNE1, we expressed PspK in R36A, a nonencapsulated laboratory strain (44). The presence of pNE1::PspK was verified by PCR, and expression was confirmed by Western blotting (Fig. 5A). Flow cytometry demonstrated that PspK was surface expressed on R36A upon transformation with pNE1::PspK. Figure 5B shows an FI of 202.12 ± 38.79 for surface expression of PspK in transformed R36A when stained with anti-PspK antibody. This value is a significant increase from the FI of 24.47 ± 7.51 before transformation (P < 0.01). The R36A cells observed to be positive for PspK can be accounted for by antibody cross-reactivity that was noted with at least PspC. This transformed R36A strain, designated LEK01, was used in the adhesion and invasion assay to determine the effect of PspK in a different genetic background. Figure 5C shows a significant increase in the number of adherent LEK01 cells compared to that for R36A (P < 0.0001). The number of adherent LEK01 cells did not differ significantly from that for MNZ11, demonstrating that induction of expression of PspK in strains that do not contain PspK increases adherence to levels seen in strains that acquired PspK in the environment. As with MNZ11, expression of PspK in R36A did not result in an increase in epithelial cell invasion compared to that for wild type (Fig. 5D).
PspK mediates adhesion independently of the pneumococcal genetic background. (A) Western blot with anti-PspK rabbit antiserum and strains MNZ11 (PspK+), LEK01 (R36A transformed with pNE1::PspK), and R36A (PspK−); (B) flow cytometry with antiserum demonstrating that PspK is expressed on the surface of LEK01; (C) LEK01 shows increased adhesion to Detroit 562 cells compared to that of R36A but no difference compared to that of MNZ11; (D) invasion of LEK01 did not significantly differ from that of parent strain R36A or MNZ11. *, P < 0.0001.
LEK01 has increased epithelial cell adherence due to the expression of PspK and a significant increase in sIgA binding compared to those for parent strain R36A, as determined by Western blotting and flow cytometry (Fig. 6). The FI of sIgA binding pneumococci significantly increased in strain LEK01 (PspK+) compared to that in parent strain R36A, with FIs of 355.45 ± 28.55 and 163.88 ± 47.71 for positive cells binding sIgA, respectively (P < 0.05) (Fig. 6B).
Binding of sIgA mediated by PspK expressed in R36A. Western blot analysis of strain LEK01 and parent strain R36A shows that LEK01 binds human sIgA upon expression of PspK (A), and the result was verified by flow cytometry (B).
Mouse nasopharyngeal colonization with LEK01.We have previously demonstrated a loss of mouse nasopharyngeal colonization upon deletion of PspK in MNZ11 (12). Therefore, we wanted to determine if the increase in adherence to human epithelial cells seen in LEK01 would translate into an increase in mouse nasopharyngeal colonization. PspK expression in R36A resulted in a greater than 9-fold increase in the number of colonizing pneumococci compared to that for wild type (Fig. 7). These results indicate that PspK significantly increases colonization independently of the pneumococcal genetic background.
R36A nasopharyngeal colonization is increased by expression of PspK. At 5 days postinfection, nasal lavage specimens show a significant increase in the number of pneumococci recovered when PspK is expressed. *, P < 0.0001.
Genetic complementation.A PspK deletion mutant, LEK04 (Kanr), derived from MNZ11, was used in complementation studies. This allowed for selection of spectinomycin resistance conferred by pNE1::PspK. Adhesion to Detroit 562 epithelial cells, as described above, was used to test the genetic complementation of LEK04 with pNE1::PspK. Percent adhesion of LEK04 was comparable to that for the PspK mutant MNZ1131, and complementation with PspK restored the number of adhesive pneumococci to wild-type levels (Fig. 8).
Genetic complementation of the PspK mutant. Strain LEK04 has decreased adherence compared to strain MNZ11 but no difference in adherence from that for strain MNZ1131. Complementation of LEK04 with pNE1::PspK increased adherence comparable to that for wild type. Bacteria recovered after each experiment were normalized to MNZ11 set at 100% binding. *, significant difference of P < 0.0001. Values were obtained from triplicate wells infected with the indicated strain.
DISCUSSION
We have determined that PspK, which is encoded within the capsule locus of some NTSp isolates, increases adherence of pneumococci to human epithelial cells in vitro. PspK also binds human sIgA, which could confer several possible advantages, such as pIgR binding or immune evasion (28, 45). Although we have shown that PspK does not aid with the resistance of complement deposition or FH binding, our data demonstrate that pneumococcal colonization and adherence are increased as a function of PspK independently of the pneumococcal genetic background. The prevalence of NTSp has been increasing since the introduction of the capsular conjugate vaccine, and NTSp strains that contain PspK may have an advantage in establishing nasopharyngeal colonization (11). A prerequisite for pneumococcal disease is the effective establishment of a population in the nasopharynx that can then disseminate to other areas of the host (3). The presence of PspK may help mediate this initial step, possibly increasing the risk for pneumococcal disease caused by NTSp.
Pneumococcal colonization has previously been shown to be dependent upon the presence of capsule in naturally encapsulated strains, and capsule is necessary for full virulence (14, 15). Weiser et al. (46) showed that pneumococcal phase variants alter colonization fitness. Transparent variants had reduced capsule expression, which led to increased colonization (46). While reduced capsule expression aids with colonization, the complete absence of capsule abrogates colonization in naturally encapsulated strains (14). In contrast, MNZ11 (PspK+), which expresses no capsule, colonizes and persists as readily as encapsulated strains (12). While adherence has been shown to be decreased by the presence of capsule, colonization is enhanced by the pneumococcal capsule through decreasing agglutination in the mucosa, which reduces clearance (15). Our demonstration in a mouse model that PspK increases colonization of NTSp to a level equal to that of encapsulated pneumococci is in contrast to previous reports (14, 15). This may be due to the increased adhesive properties of PspK. Alternatively, PspK may provide protection against clearance from the nasopharynx by an undetermined mechanism.
We have demonstrated that PspK is capable of binding human sIgA, as has been shown for PspC (27). PspC binds the secretory component of sIgA and not dimeric serum IgA (27, 29). The secretory component of sIgA is obtained from the proteolytic cleavage of human pIgR on the mucosal surface after epithelial translocation of serum polymeric Ig (47). It is known that PspC binds pIgR via an interaction between a YPT motif of PspC and the secretory component of pIgR (23, 29). This interaction increases the adherence and invasion of PspC+ pneumococcal strains (23). Given that PspK possesses the same YPT motif, the increased association of PspK-expressing strains with epithelial cells could possibly be due to this mechanism. However, rates of invasion did not change upon deletion of PspK. Also, the two epithelial cell lines used for adhesion and invasion assays have differential pIgR expression, with Detroit 562 cells expressing high levels of pIgR and A549 cells expressing undetectable amounts (23, 29). This indicates that while PspK may interact with pIgR via the secretory component, an additional epithelial cell attachment point likely exists. While sIgA binding by PspK does not necessarily mediate pIgR binding or increase nasopharyngeal persistence in this manner, there are other benefits conferred by interaction with sIgA. Human sIgA aids with the clearance of bacteria from the mucosal membrane through increased agglutination, reduced pneumococcal adherence, and increased phagocytosis (48–51). Binding of sIgA by PspK can reduce these immune processes, allowing greater pneumococcal persistence on mucosal surfaces. Pneumococci also contain an IgA protease, and sIgA binding may allow greater access of the IgA protease to sIgA, thereby increasing immunoglobulin deactivation and allowing greater pneumococcal colonization (52).
Several different strains of pneumococcus, some pathogenic and others benign, can simultaneously reside within the human nasopharynx (1, 53). pspK is contained within the capsule locus, which is known to undergo genetic rearrangement and is easily transferred to other pneumococcal strains due to natural competency (54, 55). Due to the genetic location of pspK, it may be possible to transfer this gene to other NTSp strains that are present in the nasopharyngeal passage (55). We have demonstrated that the acquisition of the pspK gene, which is naturally contained within the cps locus of some NTSp strains, increases the adherence and colonization of NTSp strains that lack pspK. Thus, if NTSp strains that lack pspK cocolonize with pspK-containing NTSp strains, such as MNZ11, and incorporate pspK into their genomes, a selective advantage would be conferred by an increase in colonization fitness. If an increase in colonization occurs through the acquisition of pspK, this would allow NTSp strains that would normally be cleared to persist, increasing the possibility that they will cause disease and spread to other human hosts. The original NTSp isolates containing PspK were initially found in South Korea, but other isolates containing PspK have also been identified in Thailand and the Netherlands under the name nontypeable pneumococcal surface protein (nspA) (12, 18). This may indicate either that PspK is naturally ubiquitous in NTSp isolates or that it has been rapidly disseminating in the population. MNZ11 (PspK+) also contains several antibiotic resistance genes that could be transferred to other bacteria. Chromosomal exchange between various pneumococcal strains occurs in the nasopharynx, and increasing the number of strains containing antibiotic resistance, such as the NTSp strain MNZ11, allows a greater chance of spread of these resistance genes (56). Therefore, dissemination of colonization factors such as PspK to other pneumococci could expand the repertoire of pneumococcal DNA in the nasopharynx, increasing the possible spread of antibiotic resistance to previously sensitive strains (56, 57). A decline in antibiotic resistance has been observed since the introduction of the pneumococcal vaccine because pathogenic strains are more likely to contain resistance markers, but antibiotic-resistant NTSp strains could shift this dynamic back to a higher rate of antibiotic resistance (58, 59).
NTSp has been observed to cause disease at a low frequency, but with an increase in the number of NTSp strains circulating in humans and the acquisition of novel colonization factors such as PspK, this could potentially lead to an increase in disease caused by NTSp (21, 60). With the transfer of genetic material arises the possibility that NTSp will acquire virulence factors aiding with the spread of NTSp-related disease. The pneumococcus is the most common cause of bacterial otitis media, but with the introduction of the conjugate vaccine, there has been a decrease in otitis media caused by serotypes protected against by the vaccine (61, 62). While this decrease in pneumococcal disease caused by vaccine serotypes has been observed, NTSp cannot be protected against by any variation of the current vaccine, and NTSp has been shown to cause invasive disease and otitis media (13, 21, 60). A rise in the incidence of otitis media caused by serotypes not contained within the vaccine has been reported, opening the possibility that strains of NTSp that can efficiently colonize, possibly because of the acquisition of PspK, could spread and more readily cause otitis media (61). NTSp cannot be vaccinated against with the current vaccines, leading to the possibility of a lack of protection against otitis media caused by NTSp. This underlines the need for increased study into novel colonization factors and a pneumococcal vaccine that is reliant not on polysaccharide capsule but, rather, on common pneumococcal proteins (13, 21, 60, 63, 64).
In summary, characterization of PspK can increase our understanding of its role in NTSp colonization. We have determined that PspK increases pneumococcal epithelial adherence and aids with colonization. PspK does not bind human FH or decrease C3b deposition like PspC does, but it does bind to sIgA. These findings increase the need for further research into NTSp colonization related to PspK due to its potential to spread antibiotic resistance and cause disease. The elucidation of the binding receptor of PspK on epithelial surfaces can increase our understanding of the colonization process of these NTSp strains. This novel protein has the potential to increase the occurrence of NTSp within the human nasopharynx. Thus, the study of PspK is necessary to increase our understanding of pneumococcal evolution under the current selective pressure against historically virulent serotypes.
ACKNOWLEDGMENTS
We acknowledge Mary Marquart for aiding with the production of the rabbit polyclonal antibody, along with discussions about the current project. We also acknowledge Carlos Orihuela for help with cell culture.
Flow cytometry was performed in the UMMC Cancer Institute Flow Cytometry Core Facility, which is supported by institutional funds and by the Department of Microbiology, UMMC. This work was supported in part by National Institutes of Health grant AI-31473 to M.H.N.
FOOTNOTES
- Received 2 August 2012.
- Returned for modification 9 September 2012.
- Accepted 18 October 2012.
- Accepted manuscript posted online 31 October 2012.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
