|Home | About | Journals | Submit | Contact Us | Français|
C5a peptidase, also called SCPA (surface-bound C5a peptidase), is a surface-bound protein on group A streptococci (GAS), etiologic agents for a variety of human diseases including pharyngitis, impetigo, toxic shock, and necrotizing fasciitis, as well as the postinfection sequelae rheumatic fever and rheumatic heart disease. This protein is highly conserved among different serotypes and is also expressed in human isolates of group B, C, and G streptococci. Human tonsils are the primary reservoirs for GAS, maintaining endemic disease across the globe. We recently reported that GAS preferentially target nasal mucosa-associated lymphoid tissue (NALT) in mice, a tissue functionally analogous to human tonsils. Experiments using a C5a peptidase loss-of-function mutant and an intranasal infection model showed that this protease is required for efficient colonization of NALT. An effective vaccine should prevent infection of this secondary lymphoid tissue; therefore, the potential of anti-SCPA antibodies to protect against streptococcal infection of NALT was investigated. Experiments showed that GAS colonization of NALT was significantly reduced following intranasal immunization of mice with recombinant SCPA protein administered alone or with cholera toxin, whereas a high degree of GAS colonization of NALT was observed in control mice immunized with phosphate-buffered saline only. Moreover, administration of anti-SCPA serum by the intranasal route protected mice against streptococcal infection. These results suggest that intranasal immunization with SCPA would prevent colonization and infection of human tonsils, thereby eliminating potential reservoirs that maintain endemic disease.
The only known reservoir for group A streptococcus (GAS) is humans, where more than 150 M genotypes have evolved the potential to cause diseases ranging from relatively mild skin disease to infections with high mortality such as necrotizing fasciitis. This very common pathogen is the cause for millions of prescriptions of antibiotics the world over and the single most important cause of heart valve disease in most resource-poor countries (5, 17). A common scenario is the following. A child aged 4 to 12 years is treated with one of several antibiotics, and treatment brings relief from symptoms of GAS pharyngitis. Nearly one-third of such children, however, continue to shed streptococci, and a significant number develop recurrent disease caused by the same strain (15, 22, 29, 33). This cycle may be repeated several times during the “strep throat season” in temperate climates. Considerable evidence points to tonsils as the primary reservoir for recurrent infections, a reservoir that maintains this bacterial pathogen in human populations (15, 29).
Early efforts to use M protein purified from streptococci in vaccines were curtailed following a small study in which vaccination was associated with a considerable increase in acute rheumatic fever (14, 27). However, the global rise in severe streptococcal infections, such as toxic shock and necrotizing fasciitis, in the 1990s prompted both commercial and public health interests to restart development of vaccines for prevention of GAS infections and their complications. The antiphagocytic M protein is now the target of vaccine development by three research groups (11, 13, 16). The conserved P145 peptide epitope within the C repeats and proximal to the cell wall is the focus of Good et al. (16). Fischetti and colleagues have centered their efforts on the entire C repeat region, which they express on the surface of Streptococcus gordonii, a bacterium thought to harmlessly colonize the oral-nasal mucosa (13). On the other hand, Dale and his collaborators have constructed recombinant subunit peptides composed of the N-terminal type-specific sequences of 26 different M proteins, those thought to be most common to strains endemic in the United States (19). Other potential vaccine candidates have been identified but are at earlier stages of development (23, 24, 37).
Studies from our laboratory have demonstrated that more than 70% of adults have measurable secretory immunoglobulin A (IgA) in their saliva, directed at the surface-bound C5a peptidase (SCPA) (28). In contrast, only 1 in 10 children under the age of 10 years were observed to have anti-SCPA antibodies in their saliva. These findings suggested that anti-SCPA antibodies could account for the lower incidence of disease and carriage of GAS in adults relative to children and prompted us to perform vaccine studies with mice. Those experiments demonstrated that intranasal administration of recombinant SCPA induced serotype-independent protection against infection in an intranasal murine model (21). Further studies showed that parenteral vaccination with the recombinant, genetically inactive SCP proteins (1) SCPAw (from GAS) or SCPBw (from Streptococcus agalactiae) combined with alum and monophosphoryl lipid A (MPL) adjuvants protected mice from oral-nasal infection by M49 and M1 serotypes of GAS (9). Moreover, immunization with either SCPAw or SCPB protein prevented pneumonia induced by either GAS or S. agalactiae (6, 9). Cross-species protection was expected, because the two proteins are 98% identical in amino acid sequence (8).
Epidemiological data suggest that GAS has a strong tropism for human tonsils (29), a tropism that we confirmed in an intranasal murine model of infection (30). Nasal mucosa-associated lymphoid tissue (NALT) is the primary streptococcal target following intranasal inoculation of mice (30). Because persistence of GAS in tonsils following antibiotic therapy is thought to be the primary source of this pathogen, the optimal vaccine should prevent or eliminate colonization of that tissue. This in turn should reduce the reservoir and provide for maximum herd immunity in the overall population. To test this possibility, experiments were designed to evaluate whether intranasal inoculation of SCPAw either alone or with cholera toxin (CTX), a known mucosal adjuvant, can prevent infection and colonization of NALT. Experiments were performed using a bioluminescent M49 strain, Xen-20, which was generated by introducing a modified Photorhabdus luminescens lux cassette into the chromosome (30); thus, persistence of streptococci in NALT could be monitored by optical biophotonic imaging from the noses of live mice or by viable counts of streptococci in dissected NALT. Experiments showed that intranasal immunization with SCPAw protein, either alone or with cholera toxin, significantly enhanced clearance of streptococci from NALT following intranasal infection with strain Xen-20, whereas NALTs of mice vaccinated with phosphate-buffered saline (PBS) and subsequently infected with streptococci became heavily colonized. Intranasal administration of hyperimmune mouse serum also significantly reduced colonization of NALT following intranasal infection.
SCPAw, a truncated form of the C5a peptidase with replacement of two catalytic residues, was constructed as described previously (1, 9). SCPAw lacks signal sequence, propeptide, cell wall spanning, and peptidoglycan anchor domains. Aspartic acid (D130) and serine (S512) residues were replaced with alanine. The protein was expressed in Escherichia coli strain BLR and purified by methods described previously (1).
Streptococci were grown in THY medium (Todd-Hewitt broth supplemented with 0.5% yeast extract) (Difco Laboratories, Detroit, MI) or on solid media containing Difco blood agar base and sheep blood at 37°C in 5% CO2. Strain 90-226 (serotype M1) was originally isolated from the blood of a septic patient (12). Strain Xen-20 is a bioluminescent strain made by introducing a modified lux operon onto the chromosome of the clinical isolate GAS 591 (serotype M49) as described previously (30). Biochemical profiles, in vitro growth rates, and capacity to colonize NALT in mice were indistinguishable from those of the parental strain (30).
The 90-226 scpA(D130A) loss-of-function mutant (SCPA−) was created by gene replacement (4) using plasmid construct pPE7. Plasmid pPE7 was made by subcloning a 1.9-kb BamHI/ClaI scpA1 fragment from pGEX-4T-1 carrying a scpA1 gene with a single base substitution at codon 130 into the BamHI and ClaI restriction sites of pGhost5 (36). A single-base-pair change in codon 130 (GAT to GCT) generated by site-directed mutagenesis altered the putative active-site aspartic acid to alanine, thus inactivating protease activity (36). Plasmid pPE7 was introduced into strain 90-226 emm1.0::ova (31), and transformants were selected on erythromycin at 1.0 μg/ml. After temperature shifts, transformants were screened for erythromycin sensitivity. These colonies were either wild type or a recombinant strain containing the site-directed mutation in scpA1. The DNA region including the mutation site was amplified and digested with Fnu4HI for diagnosis. In addition, the PCR product was sequenced to confirm the mutation. The resulting strain, 90-226 scpA(D130A), and the SCPA D130A protein expressed in E. coli lack C5a peptidase activity (36).
Adult female BALB/c mice (6 to 7 weeks old) were purchased from the National Cancer Institute (Frederick, MD) and used at the age of 7 to 10 weeks. Mice infected with GAS were housed in biosafety level-2 facilities. Mice were anesthetized with an isoflurane-oxygen mixture for 2 min and then immunized intranasally with 20 μg of SCPAw without or with 2 μg of CTX (List Biological Laboratories, Inc.) in a total volume of 15 μl. Control mice were given 20 μg of tetanus toxoid (TT) plus 2 μg of CTX or PBS alone. Mice were immunized at weekly intervals and challenged with GAS 10 days after the last boost.
Bacteria were grown to exponential phase in THY and then suspended in fresh THY to a final concentration of 0.5 × 108 to 2 × 108 CFU per 15 μl. GAS were introduced intranasally into mice as previously described (30). The total volume of the inoculum was 15 μl (7.5 μl per nostril). The volume of inoculum and the dose of GAS used had previously been shown to avoid direct introduction of bacteria into lungs and to successfully colonize NALTs of inoculated mice. Mice were imaged for 5 min using Xenogen's IVIS charge-coupled device (CCD) camera system (Xenogen Corporation, Alameda, Calif.) at 24 and 48 h postinoculation (in some cases at 2 h post-intranasal GAS challenge). The total photon emission from NALT was quantified using the LivingImage software package (Xenogen Corporation, Alameda, Calif.) as previously described (30). At the end of the experiment, mice were sacrificed, and NALTs were aseptically removed, homogenized in Hanks balanced salt solution, and plated on blood agar plates after serial dilution to determine CFU.
Saliva was collected from each mouse 1 week after the last immunization, following intraperitoneal administration of 130 μl of 0.025% pilocarpine in PBS. After the addition of 2 μl of 50 mM phenylmethylsulfonyl fluoride protease inhibitor, saliva samples were immediately frozen on dry ice and stored at −20°C. Blood was collected by puncturing the retro-orbital plexus by using microhematocrit capillary tubes or by puncturing the heart. Blood was allowed to clot at room temperature, clots were removed by microcentrifugation, and serum was then stored at −20°C.
An enzyme-linked immunosorbent assay (ELISA) was used to measure specific antibody in immune and nonimmune mouse sera and saliva. Purified SCPAw (2 μg/ml) in 0.05 M sodium carbonate buffer (pH 9.6) was bound to flat-bottom microtiter wells (PolySorp Surface; Nalge Nunc International, Denmark) overnight at 4°C (100 μl/well). Excess peptide was removed, and wells were washed four times with PBS (pH 7.4) containing 0.1% Tween 20 (PBS-TW). Mouse sera and saliva were assayed using twofold dilutions of a 1:200 or 1:20 dilution, respectively, in PBS-TW. Diluted sera or saliva was added to wells (100 μl/well) and incubated for 1 h at 37°C. Unbound antibody was removed, and wells were washed four times with PBS-TW. Alkaline phosphatase-conjugated goat anti-mouse IgG (1:5,000) or IgA (1:10,000) diluted in PBS-TW was added to wells (100 μl/well) to detect specific serum or salivary antibody, respectively. Plates were incubated for 1 h at 37°C. After removal of unbound secondary antibody and washing, 100 μl of disodium 4-nitrophenyl phosphate (0.5 mg/ml) in 0.05 M sodium carbonate buffer (pH 9.6)-10 mM MgCl2 was added to each well, and the plates were incubated for 20 min at 37°C. Antibody titers were defined as the reciprocal of the highest dilution of samples which yielded anoptical density at 405 nm of more than 3 standard deviations above the meanoptical density of control samples obtained from mice immunized with PBSalone.
For passive protection experiments, immune sera collected from mice immunized three times weekly with SCPAw/CTX, SCPAw, or PBS were pooled groupwise after titers were determined. Sera with titers ranging above 50,000 were pooled for SCPAw/CTX- and SCPAw-immunized sera. Undiluted pooled immune sera or control sera (from mice immunized with PBS) were administered to anesthetized mice by intranasal instillation (20 μl per mouse). Two hours after the serum transfer, mice were challenged intranasally with bioluminescent GAS preincubated with the corresponding serum for 1 h at 4°C, as described above. The mice were imaged and sacrificed 24 h later. NALT was removed and homogenates prepared as described above. Bacterial loads in mice were assessed by measuring total photons emitted for 5 min from live animals or by counting CFU in NALT homogenates. Anti-SCPB rabbit serum was obtained from a rabbit immunized with SCPB protein, as described in reference 9. Control serum was obtained from a rabbit immunized with TT.
A two-sample Student t test was used to test the differences between experimental groups. Resulting P values of <0.05 were taken to indicate statistically significant differences.
Our previous study showed that streptococci have a strong tropism for NALT following intranasal infection (30), and we observed very few polymorphonuclear leukocytes in NALT that was heavily infected with strain 90-226 (unpublished data). We postulated that mutational elimination of SCPA function would permit more-rapid recruitment to the mucosal surface and/or to NALT and thus prevent colonization of that tissue. The contribution of SCPA to NALT infection was assessed by comparing the persistence of a SCPA− mutant and the parent strain 90-226 in NALT, lymph nodes, and spleen following intranasal infection (Table (Table1).1). The 90-226 scpA(D130A) loss-of-function mutant (SCPA−) was generated by an allelic replacement of the gene, as described in Materials and Methods. This strain has a single amino acid altered at the active site of the SCPA protein (aspartic acid to alanine), which results in loss of C5ase activity. Six groups of three mice each were infected intranasally with either wild-type 90-226 or SCPA− streptococci. Mice were sacrificed 1, 3, and 5 days after inoculation with bacteria. At 24 h, homogenized NALTs from mice inoculated with the SCPA− strain contained five times fewer viable streptococci than those inoculated with wild-type 90-226 but still retained considerable amounts of streptococci. By days 3 and 5, those inoculated with SCPA− streptococci contained 641- and 100-fold fewer streptococci, respectively, and some mice had completely cleared bacteria from NALT. Lymph nodes did not become infected with streptococci until day 3, and SCPA− streptococci were completely eliminated from this tissue by day 5. In two mice wild-type 90-226 streptococci went beyond NALT, ultimately infecting the spleen. In contrast, none of the mice infected with the SCPA− bacteria harbored streptococci in their spleens.
Mice were immunized intranasally either with the SCPAw protein, with or without the adjuvant CTX, or with PBS alone and were then challenged intranasally with the bioluminescent GAS strain Xen-20 (serotype M49) (30). Bioluminescence was barely detectable in the nose regions of mice 2 h after challenge; however, by 24 h postchallenge, bioluminescence was clearly evident in the noses of all mice except those immunized with SCPAw mixed with the mucosal adjuvant CTX (Fig. (Fig.1A),1A), indicating that the bacteria failed to colonize these mice. Surprisingly, mice immunized with SCPAw alone appeared to emit levels of bioluminescence similar to those of PBS controls. However, further analysis of photon emission from a defined region of the nose showed that the mean photon counts from mice vaccinated with SCPAw alone were also significantly lower than those of mice immunized with PBS (Fig. (Fig.1B)1B) (P < 0.01). Most mice immunized with SCPAw and CTX showed bioluminescence below measurable levels (Fig. (Fig.1A),1A), and as expected, differences in photon counts between mice vaccinated with SCPAw plus adjuvant versus no adjuvant were also statistically significant (Fig. (Fig.1B)1B) (P < 0.02).
We previously reported that NALT is the primary site of GAS colonization after intranasal infection (30). Therefore, the numbers of viable streptococci in NALT were assessed in order to confirm bioluminescence data and to confirm that bacteria were located primarily in NALT. NALT was removed from mice at 24 h after challenge, and the number of CFU in this tissue was measured by plating diluted homogenates on blood agar plates. The number of viable streptococci in the homogenized NALT was significantly lower in mice immunized with SCPAw/CTX than in those immunized with either SCPAw only or PBS, and NALTs from mice that received PBS had significantly more CFU than those from either group vaccinated with SCPAw (Fig. (Fig.1C).1C). These results are consistent with those from bioluminescence assays. Experiments were repeated with the nonbioluminescent strain 90-226 (serotype M1) and with TT as an unrelated control antigen. As expected, NALTs from mice immunized with SCPAw/CTX were significantly protected from GAS colonization compared to mice that received PBS or the control antigen TT/CTX (Fig. (Fig.2).2). This not only confirmed that SCPAw vaccination prevents colonization of mice when administered intranasally with an adjuvant but also reaffirmed that antigen-induced protection is serotype independent.
We showed previously that subcutaneous injection of mice with SCPAw mixed with alum and MPL or intranasal administration of SCPAw without adjuvant produced a strong serum antibody response (9). To assess the induction of both systemic and local mucosal immune responses following intranasal immunization, SCPAw-specific IgG in sera and IgA in saliva from the same mice were measured by ELISA. The data represent antibody titers in sera (Fig. (Fig.3A)3A) and in saliva (Fig. (Fig.3B)3B) collected from mice 1 week after final immunizations. All mice immunized with SCPAw produced a serum anti-SCPAw IgG response; however, the mean antibody titer in mice immunized with SCPAw/CTX was higher. Titers of specific anti-SCPAw IgG in sera from these mice ranged from 50,000 to 100,000, with an average of 88,000. The average anti-SCPAw IgG titer for mice immunized with SCPAw alone was 48,000. The difference in antibody titers between these two groups of mice is significant. As expected, none of the control PBS-immunized mice had measurable anti-SCPAw antibody in their sera.
SCPAw-specific IgA salivary responses were detected in all mice immunized with SCPAw/CTX. Nine out of 10 mice immunized with SCPAw without an adjuvant had measurable levels of IgA in their saliva. Average IgA titers in saliva were 600 and 100 in mice immunized with SCPAw/CTX and SCPAw alone, respectively. As expected, none of the control mice had salivary IgA directed at SCPA. One mouse that had no salivary IgA response, and a low level of specific IgG in serum, had heavy colonization in NALT at 24 h postchallenge. Thus, inclusion of CTX with the SCPAw antigen significantly enhanced both the serum IgG and salivary IgA responses.
Vaccination with SCPA was postulated to induce antibody that would neutralize protease destruction of C5a and/or opsonic antibody, thereby enhancing recruitment of phagocytes and enabling clearance of offending streptococci before stable colonization could occur. Our previous reports did not define the mechanism or assess the importance of antibody and cellular immunity in protection. Therefore, the potential of immune sera from vaccinated mice to prevent infection of NALT was investigated. Pooled sera obtained from mice immunized with SCPAw were passively transferred intranasally to naïve mice in order to directly determine the protective efficacy of anti-SCPAw antibody. Sera were pooled from mice known to have high titers of SCPA-specific antibody. Two hours later, recipient mice were challenged with the bioluminescent strain Xen-20, which had been preincubated with the corresponding pooled sera as described in Materials and Methods. Mice that received control sera (from PBS-vaccinated mice) developed high levels of GAS infection in NALT by 24 h after challenge, as indicated by the strong luminescence emanating from their nose regions (Fig. (Fig.4).4). However, recipients of immune sera, either from SCPAw- or from SCPAw/CTX-immunized mice, emitted little or no detectable light from the nose area. This finding indicates that sera from SCPAw-immunized mice can passively prevent NALT infection and is consistent with experiments that demonstrated efficient clearance of GAS from NALTs of mice actively immunized with SCPAw (Fig. (Fig.1).1). It should be noted that sera from mice immunized with SCPAw/CTX protected mice more efficiently than sera from mice vaccinated with SCPA alone. The latter had a lower titer of anti-SCPA IgG, again suggesting a direct correlation between antibody titer and protection. Similar results were obtained in separate experiments where mice were challenged with the M1 strain 90-226. Pooled sera from SCPAw- or SCPAw/CTX-immunized mice significantly reduced numbers of CFU in NALTs of recipient mice relative to control sera from mice immunized with PBS (Fig. (Fig.55).
Serum from a rabbit immunized with recombinant SCPB (originally derived from group B streptococci) was previously shown to neutralize the C5ase activity associated with both purified recombinant SCPA protein and whole streptococci (9). Moreover, immunization with recombinant SCPB protein provided protection against both GAS and group B streptococcal lung infections. These experiments and the fact that SCPB and SCPA are 98% identical in amino acid sequence suggested that either protein could be used in a vaccine for prevention of GAS infections. This possibility was further tested with the passive protection model used above. Serum from a rabbit that had been immunized by subcutaneous injection with SCPBw protein mixed with alum and MPL (9) was administered intranasally prior to intranasal challenge of recipient mice with GAS strain 90-226 that had been preincubated with the same serum. The results shown in Fig. Fig.66 demonstrate that intranasal administration of anti-SCPB sera significantly impeded infection and colonization of NALT following intranasal challenge with this type M1 streptococcus. The difference in CFU in NALT between control mice, which had received sera from a rabbit immunized with TT, and mice given rabbit SCPB-immune sera was significant at both 4 and 24 h after bacterial challenge. These data confirm that antibodies against C5a peptidase protein from group B streptococci will also provide protection against group A streptococcal infection.
Though deaths from group A streptococcal infections in the United States and Europe decreased significantly in the first half of the 20th century, pharyngitis continues to trouble families with young children, is a primary cause of missed school days, and is an important source of medical costs. Antibiotics have failed to reduce the incidence of serious sequelae such as rheumatic fever and rheumatic heart disease in resource-poor countries around the world and to prevent the emergence of toxic shock, necrotizing fasciitis, and pockets of rheumatic fever in the United States and other developed countries. In fact, 30 to 40% of children treated with antibiotics continue to shed streptococci, and many have recurrent disease. These considerations have once again prompted both commercial and public health agencies to develop vaccines that will prevent streptococcal disease. Vaccine development is aimed at eliminating colonization of susceptible individuals and at diminishing the reservoir which maintains the organism in human populations. Several reports indicate that as many as 40% of asymptomatic schoolchildren harbor streptococci in their throats (10), and persistently infected tonsils have been suggested to be an ongoing source of streptococci for those who suffer with recurrent tonsillitis (15, 29). Therefore, a vaccine will likely need to reduce the incidence of colonization and prevent initial tonsil infections in children in order to effectively reduce the incidence of disease in a given population.
Surface macromolecules, including M protein (3, 13, 16, 19, 25), C5a peptidase (9, 21), SfbI (18), group A carbohydrate (37), and lipoproteins of unknown function (26), have been targeted as vaccine candidates. Our laboratory focused on recombinant forms of the C5a peptidase because this protein is located on the surface, is highly conserved in all serotypes, and has not been associated with autoimmune reactions. Group B streptococci, which express a nearly identical protein, are not associated with autoimmune sequelae, nor do hyperimmune rodent sera contain tissue-reactive antibodies (unpublished results). Moreover, a vaccine containing this antigen has the potential to reduce the incidence of infection by other beta-hemolytic streptococcal species, such as human isolates of group C and G streptococci that express on their surfaces C5a peptidases nearly identical in amino acid sequence to SCPA. Earlier studies from our laboratory showed that immunization with the truncated peptide ΔSCPA49, a 2.9-kb fragment of scpA from a serotype M49 strain, protected mice against intranasal challenge with other streptococcal serotypes (21). In addition, intranasal immunization with this peptide evoked both systemic and mucosal antibodies in mice that blocked pharyngeal colonization (21). Recently, our laboratory reported that subcutaneous immunization of mice with SCPAw mixed with adjuvants elicited strong serum antibody responses and protected mice from nasopharyngeal colonization (9). Neither of these studies, however, defined the mechanism of protection or focused on NALT infections.
Although murine infection is an imperfect model of human pharyngitis or tonsillitis, numerous studies have used murine infection models to assess protection after vaccination (3, 9, 13). We recently showed that GAS has a strong tropism for NALT, suggesting that bacteria first enter this tissue via M cells that are scattered along the nasal epithelium and that they form microcolonies within NALT by 24 h after intranasal inoculation. Functionally, this model has many characteristics associated with colonization and infection of human tonsils (30). This tropism was confirmed by the experiments presented here, which also showed that streptococci with a loss of C5ase activity were less able to infect and colonize NALT. Although SCPA− streptococci were able to penetrate NALT, albeit in smaller numbers, they were cleared more rapidly than wild-type bacteria. This suggests that SCPA may not only delay clearance of streptococci from the mucosal surface but may also function to delay clearance of streptococci that have crossed the epithelial barrier and are present within NALT. This is consistent with previous observations that SCPA slowed the clearance of streptococci in a subdermal mouse infection model (20).
This study assessed whether intranasal vaccination with SCPAw protein blocks colonization of this human tonsil homologue and also evaluated whether the strong mucosal adjuvant CTX would enhance the secretory immune response and protection when SCPA antigen was delivered intranasally. Intranasal administration of recombinant SCPAw protein, alone or mixed with 2 μg of biologically active CTX, to mice elicited strong serum and salivary antibody responses to SCPAw protein. The potential of CTX to augment the secretory response to antigens is well described (34), so it is not surprising that both IgG and IgA anti-SCPAw titers were significantly higher in mice that received this adjuvant. Mice that received SCPA/CTX also cleared streptococci from NALT more efficiently, whether clearance was assessed by reduced emission of bioluminescence from the nose or by reduced numbers of viable bacteria in homogenized NALT. Protection was observed when mice were challenged with either serotype M49 or serotype M1 strains of streptococci. Hyperimmune rabbit serum was shown to neutralize C5ase activity, and high concentrations of specific IgG correlated with protection (9), prompting us to postulate that antibody contributed substantially to the protective response (9, 21). This was substantiated here using a passive protection assay in which both hyperimmune rabbit and pooled mouse sera were inhaled from the nares by naïve mice. As with active immunization, infection of NALT was significantly reduced by administration of serum containing an anti-SCPAw or anti-SCPBw antibody. Protection was more complete if the sera had higher titers of antibody, i.e., if they were obtained from mice that had been immunized with SCPAw plus adjuvant. Low yields prevented testing of the protective activity of saliva. The experiments presented here did not address the contribution of cellular immunity to a protective response in this model. Nevertheless, these results are consistent with adoptive T-cell experiments in mice, which demonstrated that antigen-specific T cells are first primed and expanded in NALT following intranasal infection by group A streptococci (31). Antibody could function either to speed clearance of streptococci from the mucosal surface before they reach the interior of NALT or to reduce the replication of streptococci by promoting clearance from NALT. Our experiments do not distinguish between these two possibilities.
The mechanisms underlying protective immunity to GAS infection in humans following natural infection are poorly understood, and reliable immune correlates are lacking. Opsonic M type-specific antibody and antibody directed to the C repeats of M protein have been shown to protect against infection in mice (3, 11, 16), and sera from adult volunteers immunized with M24 protein protected mice against intraperitoneal challenge with M24 streptococci (2). In an early clinical study by Fox et al., healthy adults who were immunized subcutaneously with purified M1 protein had significantly lower rates of streptococcal illness following mucosal challenge with live homologous streptococci (14). However, while 13 of 14 exhibited significant increases in type-specific antibody titers, only 5 of the vaccinated individuals in this experiment had opsonic antibodies in their sera (2). Affinity-purified human anti-M6 secretory IgA was not opsonic but protected mice against intranasal challenge. On the other hand, opsonic serum containing anti-M6 IgG failed to protect mice (3).
Less is known about the role of anti-SCPA antibodies in human resistance to streptococcal infection. Immunogenicity of SCPA protein in humans has been reported. O'Connor et al. discovered that most healthy adults have measurable levels of anti-SCPA IgG and IgA in their sera and saliva, respectively, whereas these antibodies were considerably less commonly found in these body fluids from children (28). Anti-SCPA IgG concentrations were found to be significantly higher in convalescent-phase than in acute-phase sera from children recently infected with streptococcal pharyngitis, indicating that SCPA is expressed during infection and is highly immunogenic in humans (35). Antibody directed at the peptidase could act at three levels: it could neutralize protease activity, block uptake of streptococci by M cells or epithelial cells (32), and/or opsonize the bacteria. Consistent with these possibilities are the facts that human sera containing high titers of anti-SCPA IgG are able to inhibit C5ase activity (28), anti-SCPA antibodies can reduce invasion of A549 cells (32), and anti-SCPB can promote phagocytosis of group B streptococci in vitro (7). No antipeptidase antibody has been shown to opsonize GAS, nor has a correlation between anti-SCPA antibodies and resistance to streptococcal disease in humans been made.
This work was supported by National Institutes of Health grant AI20016. We thank Wyeth Research, Pearl River, N.Y., for partial funding of this research and for supplying SCPAw protein.
Editor: V. J. DiRita