当前位置: 首页 > 期刊 > 《感染与免疫杂志》 > 2006年第10期 > 正文
编号:11409602
High-Affinity Interaction between Fibronectin and the Group B Streptoc
http://www.100md.com 《感染与免疫杂志》
     Departments of Pediatrics Chemical Engineering Bioengineering, University of Washington, Seattle, Washington 98195

    ABSTRACT

    The streptococcal C5a peptidase (ScpB) of group B streptococci (GBS) is found in virtually all clinical GBS isolates and is required for mucosal colonization in a neonatal mouse model. ScpB inhibits neutrophil chemotaxis by enzymatically cleaving the complement component C5a. We previously identified a second function of ScpB as a fibronectin (Fn) adhesin using phage display. However, phage display can identify low-affinity interactions. We therefore measured the affinity of both full-length recombinant ScpB (FL-ScpB) and the 110-amino-acid phage display fragment (Scp-PDF) for immobilized Fn using surface plasmon resonance. The affinity for Fn was very high for both FL-ScpB (equilibrium dissociation constant [KD] = 4.0 nM) and Scp-PDF (KD = 4.4 nM) and is consistent with a biologically significant role for the adhesin activity of ScpB. We also studied the Fn adhesin activity of a common natural variant of ScpB (ScpB) that contains a 4-amino-acid deletion that eliminates peptidase activity. The integrity of scpB is otherwise maintained, suggesting that the Fn adhesin activity of ScpB may be responsible for its conservation in these strains. The affinities of both FL-ScpB (KD = 2.4 nM) and ScpB-PDF (KD = 1.4 nM) for Fn are unaffected by the deletion. Complementation in trans by both scpB and scpB corrected the Fn-binding defect of an scpB deletion mutant GBS strain to an identical degree. The high affinity of ScpB for Fn and the maintenance of this affinity in ScpB support our hypothesis that the Fn adhesin activity of scpB plays a role in virulence.

    INTRODUCTION

    Group B streptococci (GBS) are a leading cause of sepsis and meningitis in newborns (2) and an emerging cause of serious bacterial infections in immunocompromised adults and the elderly (12). Recent advances in prevention have significantly reduced the burden of early-onset neonatal disease (onset at <7 days of age) but have not had an impact on the incidence of late-onset disease (onset at 7 to 90 days of age) (25).

    Evasion of host defenses is a key component of bacterial virulence. One major host defense is complement, which is involved in a variety of functions including opsonization and chemotaxis. Complement activation results in the proteolytic cleavage of C5 to generate C5a. C5a is a potent chemotactic factor that plays an important role in recruiting neutrophils to sites of inflammation. The streptococcal C5a peptidase (ScpB) is a cell surface endopeptidase produced by GBS that inactivates C5a (9). The role of scpB in pulmonary disease by GBS was recently investigated (7). Isogenic scpB mutant GBS strains showed a 5-log decrease in their ability to colonize the lungs of adult mice, demonstrating that scpB plays a significant role in mucosal colonization. These results suggested that the ability to cleave C5a is required for mucosal colonization.

    However, epidemiologic studies of clinical isolates of GBS called into question the role of the peptidase activity of ScpB in virulence. The scpB gene is conserved among clinical isolates, suggesting that the scpB gene plays an important role. However, Bohnsack et al. previously studied the prevalence of peptidase activity in a set of virulent type III GBS clinical isolates (5) and found that 22% of these isolates lacked peptidase activity. Analysis of the scpB gene in these strains demonstrated that they carried an scpB allele (scpB) containing a 12-bp deletion that eliminated peptidase activity. In spite of the loss of peptidase activity, the remainder of the scpB gene was highly conserved, suggesting that the conservation of scpB in these isolates is based upon a second as-yet-unidentified function of ScpB.

    We subsequently discovered that ScpB has a second function as a fibronectin (Fn) adhesin (4). GBS infections in neonates are preceded by colonization, first in the urogenital tract of the mother and subsequently on the skin and mucosal surfaces of the infant. Receptors on mucosal surfaces to which GBS appear to adhere include fibronectin (31), laminin (27), and keratin (29). Using a phage display library of bacterial peptides that were affinity selected against immobilized Fn (iFn), we were able to identify ScpB as a potential Fn adhesin (4). We confirmed this result by demonstrating that both the 110-amino-acid (aa) phage display fragment (PDF) of ScpB (ScpB-PDF) that was identified in the screen and full-length ScpB (FL-ScpB) bind to Fn. We further demonstrated that an isogenic scpB deletion mutant showed an approximately 50% decrease in binding to Fn, confirming that scpB is an important Fn adhesin of GBS. These results suggested the possibility that both the virulence defect of scpB mutant GBS strains and the maintenance of the scpB gene among clinical isolates are due, in part, to this Fn adhesin activity. However, the naturally occurring 4-amino-acid deletion that eliminates peptidase activity is also contained within ScpB-PDF, raising the possibility that this deletion eliminates Fn adhesin activity in addition to peptidase activity.

    In this study, we further studied the Fn adhesin activity of ScpB by determining the affinity of ScpB for Fn and by comparing this to the affinity of ScpB. We then studied the ability of Scp to act as an Fn adhesin on the bacterial surface. Taken together, these studies tested our hypothesis that the maintenance of scpB in clinical isolates and the requirement for scpB in mucosal colonization are based, in part, on its role as an Fn adhesin.

    MATERIALS AND METHODS

    Bacterial strains and plasmids. The bacterial stains and plasmids used in this study are listed in Table 1. GBS were grown in Todd-Hewitt broth as previously described (28). Escherichia coli strains were grown in Luria-Bertani medium. Strains containing plasmids were cultured in medium with ampicillin at 100 μg/ml or chloramphenicol at 5 μg/ml.

    SPR analysis of fibronectin binding. The affinity of binding of various recombinant ScpB constructs was determined using surface plasmon resonance (SPR). Fusion proteins were expressed and purified according to the manufacturer's protocol, and the integrity of the proteins was verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Preliminary experiments demonstrated significant degradation of purified proteins. Bacterial lysates were therefore treated with protease inhibitors (1 mM phenylmethylsulfonyl fluoride and 10 mM EDTA), and purification procedures were completed within 2 h to limit degradation. All proteins isolated using these procedures gave a single band on SDS-PAGE gels. FL-Scp and the enzymatically inactive FL-Scp demonstrated identical mobilities on SDS-PAGE gels, indicating that autocleavage of ScpB, as described previously by Anderson et al. (1), was not occurring. Purified protein was stored at –70°C in aliquots to avoid repeated freeze-thaw cycles. Proteins were used within 6 weeks of isolation. All SPR experiments were carried out at least three times, with similar results. Curves shown are results of a single typical experiment.

    SPR was performed using an in-house-designed and -built system. The light source uses perpendicularly polarized white light passed through a planar prism (Kretschmann) configuration and projected onto the back of a gold-coated glass slide at a fixed angle (96.5°). Reflected light was collected in a spectrophotometer connected to a personal computer with monitoring software. A dual-channel flow cell was used, with one channel as the reference to track drift and the other channel as the sensing channel. A peristaltic pump was used to deliver the protein solution through the device at a rate of 50 μl/min. Glass SPR chips were cleaned prior to gold deposition. Chips were sonicated twice in RBS 35 detergent (Pierce, Rockford, IL) for 5 min; rinsed twice in 18 M water; sonicated twice in 18 M water for 5 min, twice in acetone for 5 min, and twice in methanol for 5 min; and then dried with nitrogen. Gold was deposited on the chips by electron beam evaporation at pressures below 10–6 torr. A 2-nm layer of chromium was deposited first as an adhesion layer, and a 48-nm layer of gold was subsequently deposited. Chips were UV ozone cleaned prior to use. Recycled chips were cleaned with a solution consisting of 7:3 (vol/vol) sulfuric acid and 30% hydrogen peroxide solution and then UV ozone treated prior to use. Human plasma fibronectin (Invitrogen, Carlsbad, CA), bovine serum albumin (BSA) (Sigma, St. Louis, MO), glutathione S-transferase (GST), and ScpB were all stored in degassed phosphate-buffered saline (PBS) at 4°C during experiments. Fn was passed through a 0.2-μm filter to remove protein aggregates before use.

    Kinetic analysis. Data were fit and affinity constants were calculated using CLAMP software (http://www.cores.utah.edu/interaction/clamp.html), which employs a Levenburg-Marquad minimization routine to fit the equations (22). Where appropriate, affinity constants were calculated using a standard single-interaction model that assumes a single binding site on the receptor binding to a single binding site on the ligand (i.e., molecule A in solution binding to site B on the surface to form the bound complex AB). For systems where a single binding reaction did not model the SPR results well, reactions of the solution-phase molecule A with multiple, independent binding sites on the surface (B, C, etc.) were modeled. A model involving conformational changes (molecule A first converting to molecule A and then reacting with B to form the complex AB) was also explored.

    Affinity constants for the two-site interaction model (i.e., two parallel reactions) were calculated as shown below. We assumed that the two reactions were noncompetitive and noninhibitive and that molecule A was present in excess, since it was being continually flowed over the sample. The equations used for the two interactions were as follows:

    (1)

    (2)

    Since SPR measures mass changes at the surface, the surface concentrations are directly related to the SPR response. By casting equations 1 and 2 into differential rate equations representing the SPR response, one obtains the following:

    (3)

    (4)

    where R1 and R2 are the SPR response for each reaction (i.e., [AB] and [AC], respectively) and Rmax1 and Rmax2 are the total numbers of available sites for each reaction (i.e., [B] and [C] just before the flow of A commences). Rmax1 and Rmax2 also correspond to the saturation values for the bound species AB and AC. Integration of equations 3 and 4 yields the time-dependent uptake for each interaction. The on-rate curve represents the time the sample is exposed to molecule A and is given by the sum of the following equations:

    (5)

    (6)

    where R01 and R02 are the initial surface concentrations of the bound AB and AC complexes (typically zero for the experiments in this study) and Rmax1 and Rmax2 are the saturation values for the bound species. The off-rate curve represents the time at which the sample with the bound AB and AC complexes is flushed with pure buffer and is given by the sum of the following two equations:

    (7)

    (8)

    where Rtotal1 and Rtotal2 are the bound species concentrations at the time where the switch to pure buffer occurs. Equations 5 to 8 are fit to the response curve, and the affinities are obtained from the ratio of koff/kon.

    Creation of site-directed mutations. The 12-bp naturally occurring deletion was introduced into previously described constructs containing full-length scpB or the phage display fragment. The deletion was introduced using inverse PCR with primers containing the deletion. PCR was carried out, as previously described (30), in a 50 μl reaction volume using primers ScpB1Fwd and ScpB1Rev and 1 ng of template DNA (pBEC101, pBEC102, or pBEC103). The presence of a single product was confirmed by agarose gel electrophoresis, and the product was then treated with 20 U of DpnI for 1 h at 37°C to digest the template DNA. The PCR product was then purified using a PCR purification kit (QIAGEN, Valencia, CA) according to the manufacturer's protocol. The presence of 24 bp of overlapping sequence on the forward and reverse primers allowed for the circularization of the PCR product in vivo by recombination. The host strain (MC1061) was transformed with the linear PCR product by electroporation and plated onto LB agar containing 100 μg/ml ampicillin. Recombinant plasmids were then screened for the presence of the 12-bp deletion by PCR using pGEX-4T3 forward and reverse primers (Amersham, United Kingdom), followed by agarose gel electrophoresis. The presence of the deletion and the absence of other unintentional mutations were confirmed by nucleotide sequence analysis. Constructs containing the 12-bp deletion were derived from all three parental plasmids, pGST301 (from pBEC102), pGST302 (from pBEC103), and pGST303 (from pBEC101). The GST fusion construct plasmids were then transformed into BL21(DE3). These strains were used for the isolation and purification of GST fusion proteins for SPR studies according to the manufacturer's protocol. The GBS expression vector plasmid (pGST303) was introduced into TOH97 by electroporation as previously described (4).

    Bacterial adherence assays. Fn binding assays were carried out as previously described (31). Results are expressed as a percentage of the input inoculum bound to the plate (bound cpm/input cpm x 100). Experiments were carried out three times, and results shown are those of a typical experiment.

    Statistical analysis. For comparison of affinity constants from SPR experiments, P values were calculated from the t distribution using a two-tailed test. For bacterial binding data, P values were calculated using Student's t test.

    RESULTS

    ScpB binds to iFn with high affinity. As noted above, we had identified ScpB as an Fn adhesin but had not assessed the affinity of ScpB for Fn. Assessing the affinity of this interaction was important to assess the biological significance of this interaction. Biologically significant adhesin interactions are generally of high affinity, with dissociation constants for previously identified soluble Fn adhesins in the nM range (11, 13, 15, 26).

    We elected to employ SPR because of its ability to accurately measure interactions with immobilized receptors. SPR measures mass accumulation at a surface by the resultant refractive index changes. These measurements are made in real time without the requirement for labels and thus allow the analysis of both the kinetics and the affinity of receptor-ligand interaction. The advantages of SPR for studying receptor-ligand interactions were discussed in detail elsewhere previously (14). A typical sensorgram is shown in Fig. 1. Deionized water was introduced through the cell, followed by PBS. The SPR shift seen with the change from water to PBS is due to changes in the refractive indices of the solvents, which results in shifts in SPR that are virtually instantaneous. Fn was then injected until the signal stabilized, indicating that maximal binding had occurred. The shift in SPR demonstrates that Fn had bound to the chip. PBS was then used to wash away unbound or lightly bound protein, with a resultant small decrement in SPR. A 5% solution of BSA in PBS was then injected as a blocking agent. Most of the SPR shift seen with the injection of BSA disappeared with the PBS wash, indicating that Fn coverage of the chip surface was close to a complete monolayer. Various concentrations of the recombinant full-length ScpB (rScpB)-GST fusion protein were allowed to flow over chips previously coated with Fn and blocked with BSA. As shown in Fig. 1, the mass of bound protein is much smaller than the amount of bound Fn, and thus, the SPR shift is shown on a smaller scale in the inset and in subsequent figures.

    We next determined the affinity of the interaction. Initial analysis using a single-binding-site model showed a poor fit with the data, with an initial sharp off rate followed by an extremely slow off rate (Fig. 2A). We therefore explored two additional models: a multiple binding site model, assuming a reaction of ScpB with multiple, independent binding sites on the iFn, and a conformational change model, assuming a conformational change in the iFn (iFn first converting to molecule iFn and then reacting with ScpB to form the complex iFnScpB). The data were best modeled by a two-binding-site model, one with high affinity and the other with low affinity. As shown in Fig. 2B, this model provided an excellent fit with the SPR data and allowed the calculation of a KD for both the high- and low-affinity binding sites (Table 2). The low-affinity binding site had a KD of 16 ± 0.4 mM, and the high-affinity binding site had a KD of 4.0 ± 0.1 nM. To demonstrate that the binding was specific for iFn, we also assessed the binding of rScpB-GST to control chips coated with BSA (Fig. 2). This binding did not show a biphasic dissociation curve but showed a single, very rapid dissociation curve. Thus, a single-binding-site model was used to calculate the dissociation constant for this interaction. As shown in Table 2, the KD for the binding of ScpB to BSA was 5 orders of magnitude higher than the KD of ScpB binding to Fn, indicating that the latter interaction was specific.

    We considered the possibility that binding might be altered or mediated by the GST tag. To test this possibility, we measured the binding of GST to iFn and demonstrated that the affinity of this interaction was very low, with a KD approaching 1 mM (Table 2). We also compared the binding of rScpB that was isolated with and without the GST tag and obtained similar binding affinities (data not shown). Taken together, these results demonstrate that the binding of rScpB-GST and rScpB-PDF-GST to iFn is not affected by the GST tag and that ScpB binds specifically and with high affinity to iFn. The affinity of ScpB is similar to that of other Fn adhesins that lack specificity for iFn (i.e., ones that bind to both iFn and soluble Fn [sFn]) and suggests that this interaction is biologically relevant.

    Affinities of ScpB and ScpB-PDF for iFn are identical. We considered the possibility that the high affinity of ScpB for iFn might be based solely upon the ScpB-PDF binding site or upon the presence of multiple lower-affinity Fn-binding sites on ScpB in addition to Scp-PDF. Of note, a previous study of the Hia adhesin of Haemophilus influenzae demonstrated that two lower-affinity binding pockets on the same molecule can interact to create a higher-avidity interaction (16). To distinguish between these possibilities, we determined the affinity of ScpB-PDF for iFn using SPR. Results are shown in Fig. 3. As was the case for full-length ScpB, the binding of ScpB-PDF was saturable over time and concentration dependent and also showed low- and high-affinity binding sites. The high-affinity binding constant for Scp-PDF (4.4 ± 0.6 nM) was not statistically significantly different from that for FL-Scp. There were clear differences in the kinetics of binding that were due to differences in both the relative contribution of the low-affinity binding to the total binding and the somewhat lower affinity of Scp-PDF (24 mM versus 16 mM for Scp-FL; P < 0.05). However, the extremely low affinity of these binding interactions indicates that they are unlikely to be of biological significance. In negative control experiments, no binding of ScpB-PDF to BSA-coated chips was seen (data not shown).

    These results demonstrate that ScpB and ScpB-PDF have similar affinities for iFn. These results are inconsistent with the hypothesis that the specificity of ScpB for iFn is based upon multiple low-affinity binding sites. Instead, these results support our hypothesis that ScpB-PDF is responsible for the adhesin activity of ScpB. However, these results do not exclude the possibility that there may be other low-affinity Fn-binding sites on ScpB.

    Fn binding by recombinant ScpB and recombinant ScpB. We then compared the affinity of ScpB with that of the naturally occurring allele of ScpB (ScpB) containing a 4-aa deletion that eliminates peptidase activity.

    We first created fusion proteins containing the 4-aa deletion. Of note, the ScpB from the strains lacking the C5a peptidase activity described previously by Bohnsack et al. (5) and the ScpB of COH1 are 100% identical at the amino acid level except for the 4-aa deletion in ScpB-PDF. Introducing the 12-bp deletion into scpB from COH1 thus results in fusion proteins identical to those that would have been isolated from those peptidase-negative strains. The location of the deletion and its relationship to other regions of interest are shown in Fig. 4. The scpB deletion was introduced into the fusion protein constructs that encode full-length ScpB (pBEC102) and ScpB-PDF (pGST103) to create plasmids pGST101 and pGST102, respectively. The fidelity of these constructs was assessed by nucleotide sequence analysis, and fusion protein was purified as described in Materials and Methods.

    The affinity of full-length ScpB was tested first. We hypothesized that the 12 allele would not affect the adhesin activity of ScpB and that the affinity of ScpB and ScpB for iFn would be identical. Results of SPR experiments are shown in Fig. 5. Binding of ScpB to iFn was saturable with time and was concentration dependent, with a high-affinity KD of 2.4 ± 0.6 nM. This result was not statistically significantly different from the KD of ScpB of 4.4 nM ± 0.6 nM, demonstrating that the ScpB and ScpB alleles have similar affinities for iFn. Differences in the kinetics of binding are apparent from the curves, with a more rapid on rate and off rate for Scp-FL. This difference is due to the relatively larger contribution of the low-affinity binding to the total binding for Scp-FL (analysis not shown). However, the affinity of this interaction is extremely low and is thus unlikely to be of any biological significance.

    We then considered the possibility that even though the affinity of full-length ScpB was not affected by the 12-bp deletion, the affinity of the PDF binding site might be affected. It is possible that there are other iFn-binding sites of equal or lower affinity that are masking an effect of the 12 deletion on the affinity of the ScpB-PDF binding site. To eliminate this possibility, we measured the affinity of ScpB-PDF for iFn, and results are shown in Fig. 5. Binding was concentration dependent and saturable with time, with a KD of 1.4 ± 0.5 nM. These findings eliminate the possibility that an effect of the 4-amino-acid deletion on Fn binding by Scp-PDF is being masked by other Fn-binding sites. Taken together, these results demonstrate that the 12-bp deletion has no effect on the affinity of either ScpB or the ScpB-PDF binding site and are consistent with our hypothesis that the scpB gene is retained on the basis of its Fn adhesin activity.

    Complementation of the Fn-binding defect of the ScpB deletion mutant by scpB and scpB. All of the above-described results were obtained using purified recombinant ScpB in solution. However, adhesins act while attached to the surface of bacterial cells, raising the possibility that the 12 allele may be altering the adhesin activity of ScpB by altering the availability of ScpB or other Fn-binding sites on the surface of GBS. We hypothesized that the 12 allele does not affect the adhesin activity of ScpB when it is expressed on the surface of GBS. To test this hypothesis, we compared the ability of scpB and scpB to complement the Fn-binding defect of TOH97, a derivative of wild-type strain COH1 that does not express scpB. Site-directed mutagenesis was used to create pGST303 containing the full-length ScpB gene with the 12-bp deletion in the shuttle expression vector pDC123. This construct was used to transform TOH97 to create GBS strain COH1-GT3. The Fn binding of COH1-GT3 was then compared to those of COH1 (wild type), the deletion mutant TOH97, and the complemented strain BEC971 (TOH97 with pBEC102 containing scpB expressed in pDC123). Results are shown in Fig. 6. As previously demonstrated (3), Fn binding by TOH97 was less than that of COH1 at all Fn coating concentrations tested (P < 0.05), and this defect could be complemented by scpB expressed in trans. PDC123 alone has no effect (4). There was no statistically significant difference between the binding abilities of COH1, BEC971, and COH1-GT3, indicating that wild-type scpB and scpB are equivalent in their ability to complement the Fn-binding defect of TOH97. These results demonstrate that the 12 allele has no effect on the Fn adhesin activity of ScpB expressed on the surface of GBS.

    Taken together, the results we have obtained demonstrate that the 12 allele has no effect on the Fn binding of recombinant ScpB, recombinant ScpB-PDF, or ScpB expressed by GBS. This finding supports our hypothesis that the high-affinity interaction between ScpB and iFn is responsible for the maintenance of the scpB gene in strains lacking C5a peptidase activity.

    DISCUSSION

    A number of bacteria with the ability to bind specifically to iFn have been described previously, initially the P fimbriae for uropathogenic E. coli strains and subsequently a number of streptococcal strains, including GBS, Streptococcus sanguis (17), and Streptococcus pneumoniae (33).

    The ability to bind specifically to iFn may play a significant role in the pathogenesis of diseases caused by these organisms. First, specific adherence to iFn may enhance binding to epithelial cells. Soluble Fn is present in plasma and exudates at very high concentrations (>0.25 g/liter) (19). Fn is also present at lower concentrations in bronchial fluid, and this concentration is increased markedly with inflammation (21). sFn blocks the interaction of bacteria bearing sFn receptors from binding to iFn (10), suggesting that these bacteria would not bind to iFn in the presence of bodily fluids. A recent study with Staphylococcus aureus demonstrated that a blockade of sFn receptors occurs when they are grown in peritoneal dialysis fluid isolated from humans (18), suggesting that this phenomenon occurs in vivo. This blockade would not occur with bacteria bearing adhesins that bind specifically to iFn. Furthermore, sFn can act as an opsonin (34). A study with Staphylococcus aureus demonstrated that sFn adhesins actually led to a decrease in virulence (20), suggesting that opsonization by Fn may play an important role in vivo. Interestingly, Fn does not opsonize GBS, which lack sFn receptors (34). Taken together, these findings suggest two separate advantages of adhesins specific for iFn. First, such adhesins allow bacteria to bind to Fn present on epithelial cells in the presence of bodily fluids. Second, they allow bacteria to avoid the opsonic effects of sFn.

    However, in contrast to multiple studies measuring the affinity of sFn adhesins, previous studies have not quantified the affinity of bacterial adhesins for iFn, raising the question of whether these interactions might be of low affinity and thus of questionable biological significance. Our results demonstrate that the interaction between ScpB and iFn is best explained using a two-binding-site model. The low-affinity binding constant is 6 to 7 orders of magnitude higher than that of other bacterial adhesins and is unlikely to be of biological significance. In contrast, the high-affinity site is similar to those of other bacterial adhesins, including the soluble Fn adhesins FnBPA of Staphylococcus aureus (26) and F1 and F2 of Streptococcus pyogenes (11, 15), which also have affinity constants in the nM range, and is considerably higher than the affinity of fimbriae of Porphyromonas gingivalis for Fn, which is in the μM range (23). These results demonstrate for the first time that the affinity of an iFn adhesin is similar to the affinities of sFn adhesins and support our hypothesis that the interaction between ScpB and iFn is a biologically significant interaction.

    Previous data demonstrated that ScpB is required for pulmonary colonization in a neonatal mouse model (7). However, ScpB is a bifunctional molecule, acting both as an iFn adhesin and in evading host defenses by inactivating the complement component C5a, and it is not clear which of these activities of ScpB are required for pulmonary colonization. As noted above, the low activity of ScpB against rodent C5a and the lack of C5a peptidase activity of many virulent GBS isolates have called into question the role of the enzymatic activity of ScpB in pathogenesis and suggest the possibility that the iFn adhesin activity of ScpB might be important in virulence. The identification of strains of GBS that cause invasive disease that contain a 12-bp deletion in scpB (scpB) that eliminates peptidase activity further suggested that the iFn adhesin activity of ScpB might play an important role in the pathogenesis of human infection. In this study, we demonstrate that this 12-bp deletion does not affect the Fn adhesin activity of ScpB when ScpB is expressed as a fusion protein. Of note, scpB from COH1 and that from the wild-type strains containing the 12-bp deletion described previously by Bohnsack et al. (GW and I25) (5) are otherwise identical at the amino acid level. Thus, our experiments have truly demonstrated that the affinities of ScpB from COH1 and ScpB from stains lacking peptidase activity are identical and support our hypothesis that the Fn adhesin activity of ScpB plays a role in the maintenance of scpB in strains that lack C5a peptidase activity.

    Another issue of great interest is the biophysical basis for the ability of GBS to bind to iFn while not binding to sFn. One potential explanation is that ScpB binding to Fn is a low-affinity/high-avidity interaction and that GBS bind specifically to iFn based upon a multivalent interaction between multiple ScpB molecules on the surface of GBS and multiple iFn molecules, while monovalent binding of ScpB to sFn is not detected in the assays employed because of its low affinity. The ability of recombinant monovalent ScpB in solution to bind to iFn suggested that this was not the case but did not exclude the possibility that this was a very-low-affinity interaction. Our findings in this study demonstrate that the binding of ScpB to iFn is of very high affinity and conclusively determine that the specificity of GBS for iFn is not based upon a low-affinity/high-avidity interaction.

    In future work, we plan to further investigate the role of the two activities of ScpB in pathogenesis. Because our experiments comparing the adhesin activities of ScpB and ScpB on the bacterial surface were carried out in the background of COH1, we plan to extend our findings by confirming that these alleles also have identical adhesin activities in the background of strains that naturally lack peptidase activity. In addition, we plan to create GBS mutants that lack iFn adhesin activity specifically activity while retaining peptidase activity. We will then be able to specifically test the relative importance of Fn adhesin and C5a peptidase activities of ScpB in pathogenesis by comparing the abilities of these strains to colonize the lung in a neonatal mouse model. Data presented here demonstrate that the naturally occurring 4-aa deletion specifically eliminates the peptidase activity of ScpB while retaining its adhesin activity, making it an appropriate mutation for study in such experiments.

    We are also currently planning further investigations into the basis for the specificity of GBS for iFn. We propose two possible explanations for the specificity of GBS for iFn. First, GBS may bind to a conformational determinant present on iFn that is not present on sFn. Consistent with this hypothesis, it is well known that Fn undergoes significant conformational changes upon adsorption to a solid surface (3, 24, 32). Second, it is possible that GBS bind to a determinant created by the juxtaposition of multiple Fn molecules on a solid surface. The definition of ScpB as a high-affinity adhesin with specificity for iFn provides us with a powerful tool for investigating these two possibilities.

    ACKNOWLEDGMENTS

    Partial support for this research was provided by the National ESCA and Surface Analysis Center for Biomedical Problems (funded by NIH grant EB-002027) and the University of Washington Engineered Biomaterials program (funded by NSF grant EEC-95291661). J.R.H. was supported by an NIH predoctoral traineeship from grant GM-065098.

    FOOTNOTES

    Corresponding author. Mailing address: Department of Pediatrics, University of Washington, Box 359300, Seattle, WA 98195. Phone: (206) 987-1918. Fax: (206) 987-7311. E-mail: gtamura@u.washington.edu.

    REFERENCES

    1. Anderson, E. T., M. G. Wetherell, L. A. Winter, S. B. Olmsted, P. P. Cleary, and Y. V. Matsuka. 2002. Processing, stability, and kinetic parameters of C5a peptidase from Streptococcus pyogenes. Eur. J. Biochem. 269:4839-4851.

    2. Baker, C. J. 1997. Group B streptococcal infections. Clin. Perinatol. 24:P59-P70.

    3. Baugh, L., and V. Vogel. 2004. Structural changes of fibronectin adsorbed to model surfaces probed by fluorescence resonance energy transfer. J. Biomed. Mater. Res. A 69:525-534.

    4. Beckmann, C., J. D. Waggoner, T. O. Harris, G. S. Tamura, and C. E. Rubens. 2002. Identification of novel adhesins from group B streptococci by use of phage display reveals that C5a peptidase mediates fibronectin binding. Infect. Immun. 70:2869-2876.

    5. Bohnsack, J. F., S. Takahashi, L. Hammitt, D. V. Miller, A. A. Aly, and E. E. Adderson. 2000. Genetic polymorphisms of group B streptococcus scpB alter functional activity of a cell-associated peptidase that inactivates C5a. Infect. Immun. 68:5018-5025.

    6. Chaffin, D. O., and C. E. Rubens. 1998. Blue/white screening of recombinant plasmids in gram-positive bacteria by interruption of alkaline phosphatase gene (phoZ) expression. Gene 219:91-99.

    7. Cheng, Q., S. Debol, H. Lam, R. Eby, L. Edwards, Y. Matsuka, S. B. Olmsted, and P. P. Cleary. 2002. Immunization with C5a peptidase or peptidase-type III polysaccharide conjugate vaccines enhances clearance of group B streptococci from lungs of infected mice. Infect. Immun. 70:6409-6415.

    8. Chmouryguina, I., A. Suvorov, P. Ferrieri, and P. P. Cleary. 1996. Conservation of the C5a peptidase genes in group A and B streptococci. Infect. Immun. 64:2387-2390.

    9. Cleary, P. P., J. Handley, A. N. Suvorov, A. Podbielski, and P. Ferrieri. 1992. Similarity between the group B and A streptococcal C5a peptidase genes. Infect. Immun. 60:4239-4244.

    10. Courtney, H. S., I. Ofek, W. A. Simpson, D. L. Hasty, and E. H. Beachey. 1986. Binding of Streptococcus pyogenes to soluble and insoluble fibronectin. Infect. Immun. 53:454-459.

    11. Ensenberger, M. G., B. R. Tomasini-Johansson, J. Sottile, V. Ozeri, E. Hanski, and D. F. Mosher. 2001. Specific interactions between F1 adhesin of Streptococcus pyogenes and N-terminal modules of fibronectin. J. Biol. Chem. 276:35606-35613.

    12. Farley, M. M. 2001. Group B streptococcal disease in nonpregnant adults. Clin. Infect. Dis. 33:556-561.

    13. Fink, D. L., B. A. Green, and J. W. St. Geme III. 2002. The Haemophilus influenzae Hap autotransporter binds to fibronectin, laminin, and collagen IV. Infect. Immun. 70:4902-4907.

    14. Green, R. J., R. A. Frazier, K. M. Shakesheff, M. C. Davies, C. J. Roberts, and S. J. Tendler. 2000. Surface plasmon resonance analysis of dynamic biological interactions with biomaterials. Biomaterials 21:1823-1835.

    15. Kreikemeyer, B., S. Oehmcke, M. Nakata, R. Hoffrogge, and A. Podbielski. 2004. Streptococcus pyogenes fibronectin-binding protein F2: expression profile, binding characteristics, and impact on eukaryotic cell interactions. J. Biol. Chem. 279:15850-15859.

    16. Laarmann, S., D. Cutter, T. Juehne, S. J. Barenkamp, and J. W. St. Geme. 2002. The Haemophilus influenzae Hia autotransporter harbours two adhesive pockets that reside in the passenger domain and recognize the same host cell receptor. Mol. Microbiol. 46:731-743.

    17. Lowrance, J. H., D. L. Hasty, and W. A. Simpson. 1988. Adherence of Streptococcus sanguis to conformationally specific determinants in fibronectin. Infect. Immun. 56:2279-2285.

    18. Massey, R. C., S. R. Dissanayeke, B. Cameron, D. Ferguson, T. J. Foster, and S. J. Peacock. 2002. Functional blocking of Staphylococcus aureus adhesins following growth in ex vivo media. Infect. Immun. 70:5339-5345.

    19. McCafferty, M. H., M. Lepow, T. M. Saba, E. Cho, H. Meuwissen, J. White, and S. F. Zuckerbrod. 1983. Normal fibronectin levels as a function of age in the pediatric population. Pediatr. Res. 17:482-485.

    20. McElroy, M. C., D. J. Cain, C. Tyrrell, T. J. Foster, and C. Haslett. 2002. Increased virulence of a fibronectin-binding protein mutant of Staphylococcus aureus in a rat model of pneumonia. Infect. Immun. 70:3865-3873.

    21. Menendez, R., V. Marco, V. Rubio, and A. Sole. 1995. Quantification of fibronectin in bronchoalveolar lavage fluid and in lung parenchyma: differences between smokers and non-smokers. Respiration 62:312-316.

    22. Myszka, D. G., and T. A. Morton. 1998. CLAMP: a biosensor kinetic data analysis program. Trends Biochem. Sci. 23:149-150.

    23. Nakamura, T., A. Amano, I. Nakagawa, and S. Hamada. 1999. Specific interactions between Porphyromonas gingivalis fimbriae and human extracellular matrix proteins. FEMS Microbiol. Lett. 175:267-272.

    24. Narasimhan, C., and C. S. Lai. 1989. Conformational changes of plasma fibronectin detected upon adsorption to solid substrates: a spin-label study. Biochemistry 28:5041-5046.

    25. Schrag, S. J., S. Zywicki, M. M. Farley, A. L. Reingold, L. H. Harrison, L. B. Lefkowitz, J. L. Hadler, R. Danila, P. R. Cieslak, and A. Schuchat. 2000. Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis. N. Engl. J. Med. 342:15-20.

    26. Signas, C., G. Raucci, K. Jonsson, P. E. Lindgren, G. M. Anantharamaiah, M. Hook, and M. Lindberg. 1989. Nucleotide sequence of the gene for a fibronectin-binding protein from Staphylococcus aureus: use of this peptide sequence in the synthesis of biologically active peptides. Proc. Natl. Acad. Sci. USA 86:699-703.

    27. Spellerberg, B., E. Rozdzinski, S. Martin, J. Weber-Heynemann, N. Schnitzler, R. Lutticken, and A. Podbielski. 1999. Lmb, a protein with similarities to the LraI adhesin family, mediates attachment of Streptococcus agalactiae to human laminin. Infect. Immun. 67:871-878.

    28. Tamura, G. S., J. M. Kuypers, S. Smith, H. Raff, and C. E. Rubens. 1994. Adherence of group B streptococci to cultured epithelial cells: roles of environmental factors and bacterial surface components. Infect. Immun. 62:2450-2458.

    29. Tamura, G. S., and A. Nittayajarn. 2000. Group B streptococci and other gram-positive cocci bind to cytokeratin 8. Infect. Immun. 68:2129-2134.

    30. Tamura, G. S., A. Nittayajarn, and D. L. Schoentag. 2002. A glutamine transport gene, glnQ, is required for fibronectin adherence and virulence of group B streptococci. Infect. Immun. 70:2877-2885.

    31. Tamura, G. S., and C. E. Rubens. 1995. Group B streptococci adhere to a variant of fibronectin attached to a solid phase. Mol. Microbiol. 15:581-589.

    32. Ugarova, T. P., C. Zamarron, Y. Veklich, R. D. Bowditch, M. H. Ginsberg, J. W. Weisel, and E. F. Plow. 1995. Conformational transitions in the cell binding domain of fibronectin. Biochemistry 34:4457-4466.

    33. van der Flier, M., N. Chhun, T. M. Wizemann, J. Min, J. B. McCarthy, and E. I. Tuomanen. 1995. Adherence of Streptococcus pneumoniae to immobilized fibronectin. Infect. Immun. 63:4317-4322.

    34. Yang, K. D., N. H. Augustine, L. A. Gonzalez, J. F. Bohnsack, and H. R. Hill. 1988. Effects of fibronectin on the interaction of polymorphonuclear leukocytes with unopsonized and antibody-opsonized bacteria. J. Infect. Dis. 158:823-830.

    35. Zhou, X. Z., J. M. Whiteley, J. A. Hoch, and K. I. Varughese. 1997. Purification and preliminary crystallographic studies on the sporulation response regulatory phosphotransferase protein, Spo0B, from Bacillus subtilis. Proteins 27:597-600.(Glen S. Tamura, James R. Hull, Michael D)