Streptococcus pneumoniae is a common cause of septicemia in the immunocompetent host. To establish infection, S. pneumoniae has to overcome host innate immune responses, one component of which is the complement system. Using isogenic bacterial mutant strains and complement-deficient immune naive mice, we show that the S. pneumoniae virulence factor pneumolysin prevents complement deposition on S. pneumoniae, mainly through effects on the classical pathway. In addition, using a double pspA/ply mutant strain we demonstrate that pneumolysin and the S. pneumoniae surface protein PspA act in concert to affect both classical and alternative complement pathway activity. As a result, the virulence of the pspA/ply strain in models of both systemic and pulmonary infection is greatly attenuated in wild-type mice but not complement deficient mice. The sensitivity of the pspA/ply strain to complement was exploited to demonstrate that although early innate immunity to S. pneumoniae during pulmonary infection is partially complement-dependent, the main effect of complement is to prevent spread of S. pneumoniae from the lungs to the blood. These data suggest that inhibition of complement deposition on S. pneumoniae by pneumolysin and PspA is essential for S. pneumoniae to successfully cause septicemia. Targeting mechanisms of complement inhibition could be an effective therapeutic strategy for patients with septicemia due to S. pneumoniae or other bacterial pathogens.

Streptococcus pneumoniae is one of a limited number of bacterial pathogens that commonly cause septicemia in a previously healthy host. Even if treated with appropriate antibiotics S. pneumoniae septicemia has a mortality of 20–25%, and identifying the mechanisms allowing S. pneumoniae survival and growth within the systemic circulation is a high priority. To establish invasive infection, bacterial pathogens have to evade important components of the innate immune response, including the complement system. Complement proteins form three enzyme cascades (termed the classical, mannose-binding lectin (MBL) 3 and alternative pathways), activation of which results in opsonization of pathogens by C3b deposition on their surface, lysis of Gram-negative bacteria by the membrane attack complex, and the release of proinflammatory signals such as C3a and C5a (1). Both clinical and laboratory studies provide strong evidence that the complement system is important for innate immunity to S. pneumoniae (2, 3, 4, 5, 6, 7). S. pneumoniae infections are common in patients with deficiencies of complement components (2, 5), and in mouse models of infection loss of complement activity results in a rapidly progressive septicemia and impaired cellular immune responses (7). Innate immunity to S. pneumoniae is critically dependent on both the classical and alternative pathways, whereas the MBL pathway seems to be only of minor importance (7, 8).

The central importance of complement for innate immunity suggests that inhibition of complement activity may be important for S. pneumoniae virulence. Several S. pneumoniae proteins interact with complement components in vitro, including the cell surface proteins PspA and CbpA, and the cytotoxin pneumolysin (9). Experiments with an isogenic pspA capsular serotype 3 S. pneumoniae strain have shown that PspA reduces complement deposition on S. pneumoniae by inhibiting both alternative and classical pathways by an unknown mechanism (8, 10, 11, 12). In addition the reduced virulence of this strain was fully restored in factor B-deficient mice, suggesting that inhibition of alternative pathway activity is one mechanism by which PspA aids S. pneumoniae virulence (10). Pneumolysin is an intracellular protein but is released into the extracellular environment during infection, and is essential for the full virulence of S. pneumoniae (13, 14, 15, 16). In vitro pneumolysin activates the classical pathway by binding to C1q, either directly due to its similarity to C-reactive protein or indirectly by binding to the Fc region of host IgG (17, 18, 19), and this could potentially affect complement-dependent immunity against S. pneumoniae (15). However, pneumolysin has pleiotropic effects that may influence virulence, including impairment of leukocyte function, inhibition of mucociliary clearance, and promotion of inflammation (15, 20, 21, 22), and the importance of the interaction with C1q for disease pathogenesis is unclear. Experiments using S. pneumoniae strains expressing different pneumolysin derivatives have shown distinct effects on infection for the pneumolysin cytotoxic and complement activation domains (23, 24), but the relationship between pneumolysin structure and its in vivo functions is complex and varies in different models of infection (16, 25, 26). Whether pneumolysin prevents complement deposition on S. pneumoniae and therefore inhibits complement-mediated immunity is not known.

We have used isogenic bacterial mutant strains and mouse models of infection in wild-type and complement-deficient mice to investigate the ability of pneumolysin to prevent complement deposition on S. pneumoniae and whether this property affects disease pathogenesis. As pneumolysin and PspA may both affect complement activity, we have also investigated the effects of both proteins in combination on complement-mediated innate immunity to S. pneumoniae. The results demonstrate that the pspA/ply strain is highly sensitive to complement, and we have exploited this property to identify when and at what anatomical sites complement assists innate immunity to S. pneumoniae.

The capsular serotype 2 strain, D39, and its previously described isogenic derivatives containing a deletion affecting the pneumolysin gene (originally called ΔPly and deficient in both the cytolytic and complement activities of pneumolysin; termed ply in this article) and/or a disruption affecting the PspA gene (pspA) (13, 16, 27) were used for this study. Additional experiments were performed using a capsular serotype 23 strain, Io11697 (kind gift from B. Spratt, Imperial College, London, U.K.) and its isogenic ply and pspA derivatives (constructed by transformation with genomic DNA from the corresponding D39 mutant strains using standard protocols) (28). To limit the effects of phase variation on infection experiments, S. pneumoniae capsular serotype 2 strains were passaged through mice by i.p. inoculation and recovery of bacteria from the spleen. Bacteria were cultured at 37°C in 95% air/5% CO2 on Columbia agar containing 5% horse blood or in Todd-Hewitt broth supplemented with 0.5% yeast extract (THY), supplemented when necessary with erythromycin 0.2 μg/ml. Strains were stored at −80°C as single-use aliquots of THY broth culture (OD580 = 0.3 − 0.4) in 10% glycerol.

C3 deposition assays were performed as previously described using equal numbers of bacteria for each S. pneumoniae strain compared (7). A purified r43-kDa N-terminal truncated component of PspA, and full-length pneumolysin (modified by Trp433 → Phe substitution to have minimal cytotoxic activity but retaining complement-binding activity) and PsaA (a negative control protein) were obtained by overexpression of previously described plasmids in Escherichia coli (29). Protein concentrations were calculated using the BCA Protein Assay kit (Pierce). For experiments assessing protein inhibition of C3 deposition on S. pneumoniae, serum was incubated with protein for 30 min before addition of the bacteria. The data are presented as a percentage of the value for the wild-type strain and represent means and SD for three to four reactions using different serum (pooled from three to five mice or human volunteers) and bacterial stocks for each strain. Human serum was obtained from healthy subjects unvaccinated against S. pneumoniae. Experiments with the serotype 2 strains used neat mouse serum and a 1 in 5 dilution of human serum, and with the serotype 23 strains a 1 in 5 dilution of mouse serum. Data were analyzed using absolute values and two-tailed t tests.

Infection experiments conformed to institutional and government guidelines for working with animals. Wild-type C57BL/6 mice were obtained from a regulated breeding establishment and complement-deficient mice (C3−/−, C1qa−/−, and Bf−/−) supplied by the investigators (M. Botto) (30, 31, 32). Mixed infection experiments used mice aged from 8 to 16 wk with no prior exposure to S. pneumoniae (7). Mixtures of the two bacterial strains being compared were inoculated in a 1:1 ratio, target organs recovered at fixed time points and homogenized (spleens and lungs) then, along with the inoculum, diluted and plated onto plain and antibiotic medium to determine the ratios of the two strains using antibiotic sensitivities. The relative virulence of the strains is presented as a competitive index (CI), defined as the ratio of the test strain (single or double mutant strain) to the reference strain (wild-type or a single mutant strain) recovered from the mice divided by the ratio of the test strain to the reference strain in the inoculum (28). A CI of <1.0 indicates that the test strain is reduced in virulence compared with the reference strain. The lower the CI the greater the reduction in virulence. Mice were inoculated by the i.p. route (a model of septicemia) with a total of 5 × 104 CFU and the CI calculated for bacteria recovered from the spleen after 24 h, or the intranasal route (under halothane anesthesia, a model of pneumonia) with a total of 5 × 106 CFU and the CIs calculated for bacteria recovered from bronchoalveolar fluid (BAL), lung, blood, and spleen after 4 or 24 h (7, 28). Mixed infection experiments were in general repeated twice using three to six mice in each group, and CIs compared using two-tailed t tests.

Whether pneumolysin prevents complement deposition on S. pneumoniae was analyzed by measuring C3 deposition in human serum on capsular serotype 2 wild-type and plyS. pneumoniae strains using a flow cytometry assay. The effect of PspA alone and in combination with pneumolysin was assessed using pspA and double-mutant pspA/ply strains. The proportion of bacteria positive for C3 was increased on the ply strain compared with the wild-type strain, but the intensity of C3 deposition was not affected (Fig. 1, AC). In contrast, on the pspA strain the intensity of C3 deposition was increased but the proportion of bacteria positive for C3 was less affected compared with the wild-type strain. The pattern of C3 deposition on the double-mutant pspA/ply strain was a combination of the different patterns observed with the individual pspA and ply strains, with both an increased intensity of C3 deposition and an increased proportion of bacteria positive for C3 compared with the wild-type strain. These results demonstrate that pneumolysin does prevent C3 deposition on S. pneumoniae, and that the combination of both pneumolysin and PspA are additive and highly effective in inhibiting complement activity.

FIGURE 1.

C3 deposition on the pspA, ply, and pspA/plyS. pneumoniae strains expressed as a percentage of the results for the corresponding wild-type strain. Error bars represent SDs. AC, Results using human serum and capsular serotype 2 strains, SDs. A, Proportion of bacteria positive for C3. Values of p: ply vs wild type = 0.0006; pspA vs wild type = 0.01; and pspA/ply vs wild type = 0.0005. B, Geometric mean fluorescence intensity (MFI) of C3 deposition. Values of p: ply vs wild type = 0.17; pspA vs wild type = 0.006; and pspA/ply vs wild type = 0.0002. C, Examples of flow cytometry histograms of C3 deposition on the D39 wild-type, ply, pspA, and pspA/ply strains in human serum. DF, Results using wild-type mouse serum and capsular serotype 2 strains. D, Proportion of bacteria positive for C3. Values of p: ply or pspA/ply vs wild type <0.0001; pspA vs wild type = 0.10; and pspA/ply vs ply or pspA < 0.001. E, Geometric MFI of C3 deposition. Values of p: ply or pspA/ply vs wild type <0.006; pspA vs wild type = 0.10; and pspA/ply vs ply or pspA < 0.001. F, Examples of flow cytometry histograms of C3 deposition in serum from wild-type mice on the D39 wild-type, pspA, ply, and pspA/ply strains. GI, Results using wild-type mouse serum and capsular serotype 23 strains. G, Proportion of bacteria positive for C3. Values of p: ply vs wild type = 0.015; pspA vs wild type = 0.002; and pspA/ply vs wild type = 0.003. H, Geometric MFI of C3 deposition. Values of p: ply vs wild type = 0.13; pspA vs wild type = 0.022; and pspA/ply vs wild type = 0.05. I, Examples of flow cytometry histograms of C3 deposition in serum from wild-type mice on the Io11697 wild-type, pspA, ply, and pspA/ply strains. For C, F, and I, The line marked pbs represents the negative control (wild-type bacteria incubated in pbs).

FIGURE 1.

C3 deposition on the pspA, ply, and pspA/plyS. pneumoniae strains expressed as a percentage of the results for the corresponding wild-type strain. Error bars represent SDs. AC, Results using human serum and capsular serotype 2 strains, SDs. A, Proportion of bacteria positive for C3. Values of p: ply vs wild type = 0.0006; pspA vs wild type = 0.01; and pspA/ply vs wild type = 0.0005. B, Geometric mean fluorescence intensity (MFI) of C3 deposition. Values of p: ply vs wild type = 0.17; pspA vs wild type = 0.006; and pspA/ply vs wild type = 0.0002. C, Examples of flow cytometry histograms of C3 deposition on the D39 wild-type, ply, pspA, and pspA/ply strains in human serum. DF, Results using wild-type mouse serum and capsular serotype 2 strains. D, Proportion of bacteria positive for C3. Values of p: ply or pspA/ply vs wild type <0.0001; pspA vs wild type = 0.10; and pspA/ply vs ply or pspA < 0.001. E, Geometric MFI of C3 deposition. Values of p: ply or pspA/ply vs wild type <0.006; pspA vs wild type = 0.10; and pspA/ply vs ply or pspA < 0.001. F, Examples of flow cytometry histograms of C3 deposition in serum from wild-type mice on the D39 wild-type, pspA, ply, and pspA/ply strains. GI, Results using wild-type mouse serum and capsular serotype 23 strains. G, Proportion of bacteria positive for C3. Values of p: ply vs wild type = 0.015; pspA vs wild type = 0.002; and pspA/ply vs wild type = 0.003. H, Geometric MFI of C3 deposition. Values of p: ply vs wild type = 0.13; pspA vs wild type = 0.022; and pspA/ply vs wild type = 0.05. I, Examples of flow cytometry histograms of C3 deposition in serum from wild-type mice on the Io11697 wild-type, pspA, ply, and pspA/ply strains. For C, F, and I, The line marked pbs represents the negative control (wild-type bacteria incubated in pbs).

Close modal

Whether pneumolysin and PspA affects either the alternative or classical pathways was assessed using serum from wild-type, C1qa−/−, and Bf−/− mice for C3 deposition assays on wild-type, pspA, ply, and pspA/plyS. pneumoniae strains. Although C3 deposition on the mutant S. pneumoniae strains was increased in serum from wild-type mice, the patterns of C3 deposition differed compared with those in human serum (Fig. 1, DF). Both the proportion of bacteria positive for C3 and the intensity of C3 deposition were increased for the ply strain, but the increase in C3 deposition on the pspA strain was not statistically significant when compared with the wild-type strain. Furthermore, in mouse serum the increase in C3 deposition on the pspA/ply strain compared with the pspA and ply strains was more marked than in human serum. Similar experiments were also performed with a capsular serotype 23 strain. There was increased C3 deposition on the capsular serotype 23 ply and pspA strain compared with the wild-type strain, and C3 deposition was strikingly increased on the capsular serotype 23 pspA/ply strain compared with the pspA and ply strains (Fig. 1, GI). Hence, the combination of PspA and pneumolysin strongly inhibit C3 deposition on both capsular serotype 2 and 23 strains, although the effect was markedly stronger for the serotype 23 strain suggesting there is variation in the efficacy of these proteins between S. pneumoniae strains.

In serum from C1qa−/− mice (in which only alternative pathway-dependent C3 deposition on S. pneumoniae can occur) there was a significant increase in both the proportion of bacteria positive for C3 and intensity of C3 deposition on the serotype 2 pspA strain compared with the wild-type strain, whereas there was only a small increase in the intensity of C3 deposition on the ply strain with no significant increase in the proportion of bacteria positive for C3 (Fig. 2, A and B). In contrast, in serum from Bf−/− mice (in which only classical pathway-dependent C3 deposition on S. pneumoniae can occur) there was a significant increase in the proportion of bacteria positive for C3 on the ply strain compared with the wild-type strain (Fig. 2,C). There was also an increase in the proportion of bacteria positive for C3 on the pspA strain, but this did not reach statistical significance. Due to loss of alternative pathway-mediated amplification of C3 deposition, the intensity of C3 deposition on bacteria in serum from Bf−/− mice was too low for differences between mutant strains to become apparent (Fig. 2 D). These results indicate that pneumolysin prevents complement deposition on S. pneumoniae mainly through effects on the classical pathway, and that PspA mainly inhibits alternative pathway activity although some effect on the classical pathway may also occur.

FIGURE 2.

C3 deposition in mouse serum on the capsular serotype 2 pspA, ply, and pspA/plyS. pneumoniae strains expressed as a percentage of the results for the D39 wild-type strain. A, Proportion of bacteria positive for C3 in serum from C1qa−/− mice. Values of p: pspA vs wild type = 0.014; pspA vs pspA/ply and ply vs wild type >0.12; and pspA/ply vs wild type = 0.001. B, Geometric MFI of C3 deposition in serum from C1qa−/− mice. Values of p: ply vs wild type = 0.025; pspA vs wild type = 0.014; pspA/ply vs wild type = 0.008; and pspA/ply vs ply = 0.02. C, Proportion of bacteria positive for C3 in serum from Bf−/− mice. Values of p: ply vs wild type = 0.043; pspA vs wild type = 0.11; pspA/ply vs wild type <0.001; pspA/ply vs pspA = 0.037; and ply vs pspA/ply = 0.058. D, Geometric MFI of C3 deposition in serum from Bf−/− mice (no statistics performed).

FIGURE 2.

C3 deposition in mouse serum on the capsular serotype 2 pspA, ply, and pspA/plyS. pneumoniae strains expressed as a percentage of the results for the D39 wild-type strain. A, Proportion of bacteria positive for C3 in serum from C1qa−/− mice. Values of p: pspA vs wild type = 0.014; pspA vs pspA/ply and ply vs wild type >0.12; and pspA/ply vs wild type = 0.001. B, Geometric MFI of C3 deposition in serum from C1qa−/− mice. Values of p: ply vs wild type = 0.025; pspA vs wild type = 0.014; pspA/ply vs wild type = 0.008; and pspA/ply vs ply = 0.02. C, Proportion of bacteria positive for C3 in serum from Bf−/− mice. Values of p: ply vs wild type = 0.043; pspA vs wild type = 0.11; pspA/ply vs wild type <0.001; pspA/ply vs pspA = 0.037; and ply vs pspA/ply = 0.058. D, Geometric MFI of C3 deposition in serum from Bf−/− mice (no statistics performed).

Close modal

In contrast to the results with serum from wild-type mice, the increase in complement deposition on the pspA/ply strain in serum from C1qa−/− mice was not statistically significantly different to the pspA strain, and in serum from Bf−/− mice was no greater than the additive effects of the single pspA and ply mutations (Fig. 2, AC). Hence the marked increase in C3 deposition on the pspA/ply strain in wild-type serum requires intact classical and alternative pathways.

To investigate further whether PspA and pneumolysin in combination are more effective at preventing C3 deposition on S. pneumoniae than each protein alone, C3 deposition assays were performed with the serotype 2 pspA/ply strain in the presence of a truncated purified rPspA and a noncytotoxic derivative of pneumolysin (29). Addition of PsaA, a S. pneumoniae lipoprotein with no known interaction with complement, was used as a negative control. Small doses of protein (10 μg/ml, data not shown) and addition of PspA or pneumolysin in combination with PsaA had no consistent statistically significant effect on complement deposition on S. pneumoniae. However, in both mouse and human serum addition of 100 μg/ml PspA in combination with 100 μg/ml of pneumolysin reduced the intensity of C3 deposition on the serotype 2 pspA/ply strain (Fig. 3), providing further evidence that the proteins in combination effect complement activity to a stronger degree than individual proteins.

FIGURE 3.

Relative geometric MFI of C3 deposition in pooled human (A) or wild-type mouse (B) serum on the capsular serotype 2 pspA/ply strain in the presence of 200 μg/ml purified protein. The results are expressed as a percentage of the results in the absence of additional protein. Error bars represent SDs. Which proteins are present is written beneath each column. Values of p for: A, PspA + Ply vs PsaA + Ply = 0.025; PspA + Ply vs PsaA + PspA = 0.017; B, PspA + Ply vs PsaA + Ply = 0.031; PspA + Ply vs PsaA + PspA = 0.007.

FIGURE 3.

Relative geometric MFI of C3 deposition in pooled human (A) or wild-type mouse (B) serum on the capsular serotype 2 pspA/ply strain in the presence of 200 μg/ml purified protein. The results are expressed as a percentage of the results in the absence of additional protein. Error bars represent SDs. Which proteins are present is written beneath each column. Values of p for: A, PspA + Ply vs PsaA + Ply = 0.025; PspA + Ply vs PsaA + PspA = 0.017; B, PspA + Ply vs PsaA + Ply = 0.031; PspA + Ply vs PsaA + PspA = 0.007.

Close modal

To investigate whether the effects of pneumolysin on complement deposition impair innate immune responses to S. pneumoniae, infection experiments with the serotype 2 ply strains were performed in wild-type and complement-deficient mice. If pneumolysin prevents complement-dependent immunity, then the ply strain will have reduced virulence in wild-type but not complement-deficient mice. However, using bacterial CFU in target organs or the rate of disease development to compare the virulence of different bacterial strains given as single strain inocula to wild-type and complement-deficient mice will be confounded by the rapid replication of all S. pneumoniae strains in complement-deficient mice (7, 8, 10). Hence infection with a mutant bacterial strain attenuated in virulence will progress faster in complement-deficient mice than in wild-type mice independent of whether the bacterial mutation affects a protein that interacts with complement. To avoid this problem and to provide a specific assessment of the interaction of pneumolysin with complement during infection, the relative virulence of wild-type and ply bacterial strains were compared using mixed infections within individual wild-type and complement-deficient mice and calculating a CI. As PspA influences virulence in mice through inhibition of the alternative pathway (8, 10), the pspA strain was used to validate the ability of mixed infection experiments and CIs to assess the specific interaction of bacterial proteins and the host immune system.

After i.p. inoculation the pspA strain was greatly reduced in virulence in wild-type mice, with a CI of 0.019. However, in keeping with previous reports (8, 10, 11) and demonstrating the utility of the mixed infection approach, the CI of the pspA strain was increased in C3/−, Bf−/−, and, to a lesser extent, in C1qa−/− mice (Table I), indicating that during infection PspA affects alternative and classical pathway activity. The ply strain had a less marked reduction in virulence in wild-type mice compared with the pspA strain with a CI of 0.29 (Table I), perhaps because pneumolysin is released into the extracellular environment and in mixed infections the ply strain could be partially complemented by the ply+ wild-type strain. Nevertheless, the CI of the ply strain was increased over 2-fold in C3−/− and C1qa−/− mice (CIs of 0.77 and 0.58, respectively), but not Bf−/− mice (CI of 0.30) (Table I). These results mirror the results of the C3 deposition assays and demonstrate that a major role for pneumolysin during systemic infection is prevention of immunity mediated by the classical but not alternative pathway.

Table I.

CI ± SD for serotype 2 pspA, ply, and pspA/ply mutant strains recovered from spleens of wild-type and complement-deficient mice 24 h after i.p. inoculationa

Mixed Infection Comparison StrainsMouse Strain
 Wild type C3−/− Bf−/− C1qa−/− 
pspA vs D39 0.019 ± 0.01 (n = 10) 0.16 ± 0.05b (n = 10) 0.16 ± 0.08b (n = 10) 0.11 ± 0.06b (n = 7) 
ply vs D39 0.29 ± 0.13 (n = 8) 0.77 ± 0.22c (n = 8) 0.30 ± 0.09d (n = 6) 0.58 ± 0.12c (n = 5) 
pspA/ply vs D39 0.004 ± 0.003 (n = 5) 0.11 ± 0.06b (n = 5) nd nd 
pspA/ply vs ply 0.005 ± 0.007 (n = 8) 1.11 ± 0.24b (n = 7) 0.13 ± 0.05e (n = 4) 0.33 ± 0.19e (n = 5) 
Mixed Infection Comparison StrainsMouse Strain
 Wild type C3−/− Bf−/− C1qa−/− 
pspA vs D39 0.019 ± 0.01 (n = 10) 0.16 ± 0.05b (n = 10) 0.16 ± 0.08b (n = 10) 0.11 ± 0.06b (n = 7) 
ply vs D39 0.29 ± 0.13 (n = 8) 0.77 ± 0.22c (n = 8) 0.30 ± 0.09d (n = 6) 0.58 ± 0.12c (n = 5) 
pspA/ply vs D39 0.004 ± 0.003 (n = 5) 0.11 ± 0.06b (n = 5) nd nd 
pspA/ply vs ply 0.005 ± 0.007 (n = 8) 1.11 ± 0.24b (n = 7) 0.13 ± 0.05e (n = 4) 0.33 ± 0.19e (n = 5) 
a

nd, not done. Versus the CI for the same mixed infection in wild-type mice:

b

p < 0.011;

c

p < 0.002;

d

p = 0.12;

e

p < 0.02.

Whether the high levels of complement deposition on the pspA/ply strain results in a synergistic decrease in virulence was also investigated using mixed infections. Mixed infection of the pspA/ply strain vs the wild-type D39 strain in wild-type mice gave a very low CI of 0.004, whereas the CI in C3−/− mice was 27-fold higher at 0.11 (Table I), indicating that loss of both PspA and pneumolysin results in a strain which is greatly attenuated in virulence in the presence of an intact complement system. To confirm that loss of both PspA and pneumolysin results in a synergistic reduction in virulence, mixed infections of the pspA/ply strain vs the ply strain were performed. As the ply mutation is present in both strains, if the effects on virulence of mutations in pspA and ply were independent of each other the CI for the pspA/ply vs ply strains should be similar to the CI for the pspA vs wild-type strain (28, 33). However, after i.p. inoculation in wild-type mice the CI for the pspA/ply strain vs the ply strain was nearly four times lower than the CI for the pspA strain vs the wild-type D39 strain (0.005 vs 0.019, p = 0.006) (Table I), demonstrating that loss of both pneumolysin and PspA has synergistic effects on virulence.

Mixed infections of the pspA/ply strain vs the ply strain were also performed in C3−/−, C1qa−/−, and Bf−/− mice (Table I). Strikingly, in C3−/− mice the CI was 1.1, showing that the reduced virulence of the pspA/ply strain was not apparent in mice with no effective complement activity. The CIs of the pspA/ply strain vs the ply strain in C1qa−/− and Bf−/− mice were also increased but only to 0.33 and 0.13, respectively, demonstrating that both classical and alternative pathway activity are required for the full loss of virulence of the pspA/ply strain to be apparent. These results are compatible with the C3 deposition data, and confirm that inhibition of complement deposition on S. pneumoniae by the combination of PspA and pneumolysin is essential for full virulence during systemic infection.

As the CI for the pspA/ply strain vs ply strain mixed infection was largely restored in C3−/− mice, mixed infections comparing the CI for this combination of strains in different target organs and at different time points between wild-type and C3−/− mice can be used as a biological indicator for the effects of complement-mediated immunity. After intranasal inoculation in wild-type and C3−/− mice, CIs for the pspA/ply strain vs ply strain mixed infection were obtained in BAL and lung at 4 and 24 h and in blood and spleen at 24 h (the first time point at which systemic infection consistently develops in wild-type mice in this model). At 4 h the CIs in BAL and lung in wild-type mice had fallen to 0.36 and 0.26, respectively (Fig. 4,A). However, these CIs were close to 1 in C3−/− mice, demonstrating that the reduced virulence of the pspA/ply strain compared with the ply strain was complement-dependent and that complement has a role in early innate immunity in the lung. By 24 h this effect was not noticeable, with similar CIs in BAL and lung between C3−/− and wild-type mice (Fig. 4,B). The CIs in blood and spleen for C3−/− mice at 24 h were also similar to those for BAL and lung, with the pspA/ply strain representing over one in five of the colonies recovered from all four sites (Fig. 4,B), demonstrating that in the absence of complement the pspA/ply can readily spread from the lungs to the blood. In marked contrast, in wild-type mice no pspA/ply colonies were recovered from culture of the blood and spleen at 24 h despite the recovery of large numbers of the ply strain, giving a CI of <0.0004 (Fig. 4 B). To ensure that the differences in CI between wild-type and C3−/− mice within the blood and spleen represent failure of the pspA/ply strain to survive in the systemic circulation rather than an inability to cross from the lung or peritoneum into the blood, mixed infections comparing the pspA/ply strain vs the ply strain 6 h after i.v. inoculation were performed. Later time points were not possible as the numbers of bacteria recovered from wild-type mice were too low. The CIs for bacteria recovered from the spleen were reduced in wild-type mice (0.37 ± 0.18) but remained close to 1 in C3−/− mice (0.81 ± 0.04, p = 0.0016 compared with the CI for wild-type mice), confirming that the pspA/ply strain has a particular problem with surviving within the blood. These results demonstrate that complement activity is essential for preventing spread of S. pneumoniae from the pulmonary compartment to the systemic circulation, and that the combination of pneumolysin and PspA is required for S. pneumoniae to overcome this aspect of innate immunity.

FIGURE 4.

CIs for the pspA/ply strain compared with the ply strain in BAL, lungs, blood, or spleen at 4 (A) and 24 (B) h after intranasal inoculation of wild-type (gray shading) or C3−/− (diagonal shading) mice.

FIGURE 4.

CIs for the pspA/ply strain compared with the ply strain in BAL, lungs, blood, or spleen at 4 (A) and 24 (B) h after intranasal inoculation of wild-type (gray shading) or C3−/− (diagonal shading) mice.

Close modal

Both pneumolysin and PspA are well-recognized virulence factors for the important human pathogen S. pneumoniae. However, the multiple biological properties of pneumolysin and PspA have made evaluation of how these proteins assist S. pneumoniae disease pathogenesis difficult to assess. In particular, it has not been clear whether the in vitro observation that pneumolysin interacts with C1q protects S. pneumoniae against complement-dependent immunity during infection. Experiments with S. pneumoniae strains expressing pneumolysin derivatives with intact cytotoxic but impaired C1q-binding properties have given conflicting results about the importance of the complement-binding domain, with some studies suggesting that this domain aids the progress of infection in mouse models and others suggesting that it has little influence (14, 23, 24, 25). In this paper we have used bacterial strains deficient in pneumolysin and/or PspA and genetically engineered complement-deficient mice to characterize the consequences of the interactions of these proteins with the complement system during infection.

We have shown that in serum obtained from unvaccinated humans and from mice with no prior exposure to S. pneumoniae C3 deposition is increased on ply strains. Similar results were obtained for a serotype 23 strain in mouse serum, demonstrating that the increased deposition of C3 on ply strains is not just a property of serotype 2 strains. Although the human serum used for these experiments may contain low levels of specific Ab due to prior S. pneumoniae infection, the mouse serum does not (7), and, broadly, our results indicate that pneumolysin is important for overcoming complement-dependent innate immunity. We have previously shown that during innate immunity the classical pathway strongly affects the proportion of bacteria positive for C3 whereas the alternative pathway mainly affects the intensity of C3 deposition on S. pneumoniae (7). In human serum there is an increased proportion of the ply strain positive for C3 but little effect on the intensity of C3 deposition, a pattern that suggests pneumolysin prevents classical pathway-dependent C3 deposition on S. pneumoniae. This was confirmed by experiments using serum from mice with genetic deficiencies in the classical or alternative pathways, and would be predicted from the in vitro observation that pneumolysin binds C1q (17, 18). Furthermore, mixed infection experiments in wild-type and complement-deficient mice demonstrated that the effects of pneumolysin on complement assist S. pneumoniae infection only in mice with an intact classical pathway. These data are the first direct evidence that pneumolysin affects classical pathway-mediated complement immunity. The results of the C3 deposition assays and mixed infections for the pspA strain are in keeping with previous results (10, 12), and demonstrate that PspA assists virulence by inhibiting both alternative and, to a lesser extent, classical pathway activity. The CIs for both the ply and pspA strains were not restored completely in C3−/−-deficient mice, suggesting that the respective cytotoxic and lactoferrin-binding properties of these proteins also contribute to virulence in these models of infection (15, 20, 21, 22, 34, 35). The relatively preserved virulence of the plystrain when assessed using a mixed infection model is probably due to extracellular complementation of the ply strain by pneumolysin released from the wild-type strain (23).

PspA is highly expressed on the surface of S. pneumoniae during infection (36) and therefore ideally situated to prevent complement binding to S. pneumoniae, although the exact mechanism by which it prevents alternative pathway activity is not known. Although pneumolysin is located in the cytoplasm of S. pneumoniae, it is steadily released into the culture medium during growth in vitro, either through a basal level of autolysis (37) or via a yet to be characterized export mechanism (38). Free pneumolysin has also been detected in vivo during experimental infections (39). By binding C1q, pneumolysin, rather than inhibit the classical pathway, may actually increase complement activity, and how it is able to prevent complement deposition on S. pneumoniae is unclear at present. Complement deposition is highly localized to molecules within the immediate vicinity of the activating stimulus (1), and release of pneumolysin into extracellular fluid could divert complement activity away from the bacterial surface, as well as consume the available C3 (15). Uncontrolled complement activation distant to S. pneumoniae may prevent effective phagocytosis of the bacteria and aid bacterial growth by increasing vascular permeability, thereby allowing an influx of nutrient-containing serum to the site of infection. In addition, the increase in complement activity may damage host cells and so promote virulence. Further research is necessary to characterize in detail the mechanisms by which the interaction of pneumolysin with the classical pathway results in reduced C3 deposition on S. pneumoniae and increased virulence.

Research into bacterial pathogenesis classically involves assessing the phenotypes of mutant strains containing deletions of a single virulence gene. However this approach will not identify important biological effects caused by the interactions between different virulence factors. As both pneumolysin and PspA affect complement activity, we have investigated whether pneumolysin and PspA in combination results in additional inhibition of complement activity compared with the individual proteins. In human or mouse serum C3 deposition was markedly increased on the pspA/ply strains compared with the single-mutant and wild-type strains. Exogenous pneumolysin and PspA in combination reversed this effect, although a truncated form of PspA was used for these experiments and may have only been partially effective. As previously demonstrated (27) the double pspA/ply mutant strain was markedly attenuated in virulence compared with the single mutant and wild-type bacterial strains in wild-type mice, and we have now shown that this attenuation in virulence is not seen in complement-deficient mice. Therefore, the synergistic effect on virulence of PspA and pneumolysin is mainly due to their combined inhibition of complement-dependent immunity. The interaction of PspA, pneumolysin, and other S. pneumoniae proteins that affect complement such as CbpA (40) requires further investigation.

Complement components are found in high concentration within serum, but are also found in bronchial secretions and inflammatory exudates induced by infection. Hence, there are several points during S. pneumoniae infection when complement-mediated immunity may be important. Using the sensitivity of the pspA/ply strain to complement as an indicator for when complement-mediated immunity is active, we have demonstrated that early in infection there is significant complement activity in the lungs against S. pneumoniae, and presumably other bacteria. After 24 h there was no detectable effect of complement in the pulmonary compartment, perhaps suggesting that within the lungs complement aids the early rapid clearance of bacteria but other immune mechanisms are dominant at later time points. However, the most striking effect of complement was preventing the spread of S. pneumoniae from the lungs to the blood, with pspA/ply bacteria only able to successfully invade the systemic circulation in C3−/− mice. These data combined with the results of the C3 deposition assays suggest that prevention of both classical and alternative pathway-mediated C3 deposition on S. pneumoniae by pneumolysin and PspA greatly facilitates the development of S. pneumoniae septicemia. Other pathogens which commonly cause septicemia such as Staphylococcus aureus and Streptococcus pyogenes also have multiple mechanisms affecting complement activity (9), and inhibition of complement may be an essential step for many bacterial pathogens to be able to cause septicemia in a healthy host. Preventing the bacterial mechanisms that inhibit complement activity could be a potential novel therapy for patients with septicemia.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was sponsored by the Wellcome Trust (Grant 066335).

3

Abbreviations used in this paper: MBL, mannose-binding lectin; CI, competitive index; BAL, bronchoalveolar lavage.

1
Walport, M. J..
2001
. Complement. First of two parts.
N. Engl. J. Med.
344
:
1058
-1066.
2
Alper, C. A., N. Abramson, R. B. Johnston, Jr, J. H. Jandl, F. S. Rosen.
1970
. Increased susceptibility to infection associated with abnormalities of complement-mediated functions and of the third component of complement (C3).
N. Engl. J. Med.
282
:
350
-354.
3
Gross, G. N., S. R. Rehm, A. K. Pierce.
1978
. The effect of complement depletion on lung clearance of bacteria.
J. Clin. Invest.
62
:
373
-378.
4
Winkelstein, J. A..
1981
. The role of complement in the host’s defense against Streptococcus pneumoniae.
Rev. Infect. Dis.
3
:
289
-298.
5
Sampson, H. A., A. M. Walchner, P. J. Baker.
1982
. Recurrent pyogenic infections in individuals with absence of the second component of complement.
J. Clin. Immunol.
2
:
39
-45.
6
Mold, C., B. Rodic-Polic, T. W. Du Clos.
2002
. Protection from Streptococcus pneumoniae infection by C-reactive protein and natural antibody requires complement but not Fcγ receptors.
J. Immunol.
168
:
6375
-6381.
7
Brown, J. S., T. Hussell, S. M. Gilliland, D. W. Holden, J. C. Paton, M. R. Ehrenstein, M. J. Walport, M. Botto.
2002
. The classical pathway is the dominant complement pathway required for innate immunity to Streptococcus pneumoniae infection in mice.
Proc. Natl. Acad. Sci. USA
99
:
16969
-16974.
8
Ren, B., M. A. McCrory, C. Pass, D. C. Bullard, C. M. Ballantyne, Y. Xu, D. E. Briles, A. J. Szalai.
2004
. The virulence function of Streptococcus pneumoniae surface protein A involves inhibition of complement activation and impairment of complement receptor-mediated protection.
J. Immunol.
173
:
7506
-7512.
9
Jarva, H., T. S. Jokiranta, R. Wurzner, S. Meri.
2003
. Complement resistance mechanisms of streptococci.
Mol. Immunol.
40
:
95
-107.
10
Tu, A. H., R. L. Fulgham, M. A. McCrory, D. E. Briles, A. J. Szalai.
1999
. Pneumococcal surface protein A inhibits complement activation by Streptococcus pneumoniae.
Infect. Immun.
67
:
4720
-4724.
11
Ren, B., A. J. Szalai, O. Thomas, S. K. Hollingshead, D. E. Briles.
2003
. Both family 1 and family 2 PspA proteins can inhibit complement deposition and confer virulence to a capsular serotype 3 strain of Streptococcus pneumoniae.
Infect. Immun.
71
:
75
-85.
12
Ren, B., A. J. Szalai, S. K. Hollingshead, D. E. Briles.
2004
. Effects of PspA and antibodies to PspA on activation and deposition of complement on the pneumococcal surface.
Infect. Immun.
72
:
114
-122.
13
Berry, A. M., J. Yother, D. E. Briles, D. Hansman, J. C. Paton.
1989
. Reduced virulence of a defined pneumolysin-negative mutant of Streptococcus pneumoniae.
Infect. Immun.
57
:
2037
-2042.
14
Berry, A. M., J. E. Alexander, T. J. Mitchell, P. W. Andrew, D. Hansman, J. C. Paton.
1995
. Effect of defined point mutations in the pneumolysin gene on the virulence of Streptococcus pneumoniae.
Infect. Immun.
63
:
1969
-1974.
15
Paton, J. C..
1996
. The contribution of pneumolysin to the pathogenicity of Streptococcus pneumoniae.
Trends Microbiol.
4
:
103
-106.
16
Berry, A. M., A. D. Ogunniyi, D. C. Miller, J. C. Paton.
1999
. Comparative virulence of Streptococcus pneumoniae strains with insertion-duplication, point, and deletion mutations in the pneumolysin gene.
Infect. Immun.
67
:
981
-985.
17
Paton, J. C., B. Rowan-Kelly, A. Ferrante.
1984
. Activation of human complement by the pneumococcal toxin pneumolysin.
Infect. Immun.
43
:
1085
-1087.
18
Mitchell, T. J., P. W. Andrew, F. K. Saunders, A. N. Smith, G. J. Boulnois.
1991
. Complement activation and antibody binding by pneumolysin via a region of the toxin homologous to a human acute-phase protein.
Mol. Microbiol.
5
:
1883
-1888.
19
Rossjohn, J., R. J. Gilbert, D. Crane, P. J. Morgan, T. J. Mitchell, A. J. Rowe, P. W. Andrew, J. C. Paton, R. K. Tweten, M. W. Parker.
1998
. The molecular mechanism of pneumolysin, a virulence factor from Streptococcus pneumoniae.
J. Mol. Biol.
284
:
449
-461.
20
Paton, J. C., A. Ferrante.
1983
. Inhibition of human polymorphonuclear leukocyte respiratory burst, bactericidal activity, and migration by pneumolysin.
Infect. Immun.
41
:
1212
-1216.
21
Feldman, C., N. C. Munro, P. K. Jeffery, T. J. Mitchell, P. W. Andrew, G. J. Boulnois, D. Guerreiro, J. A. Rohde, H. C. Todd, P. J. Cole, et al
1991
. Pneumolysin induces the salient histologic features of pneumococcal infection in the rat lung in vivo.
Am. J. Respir. Cell Mol. Biol.
5
:
416
-423.
22
Malley, R., P. Henneke, S. C. Morse, M. J. Cieslewicz, M. Lipsitch, C. M. Thompson, E. Kurt-Jones, J. C. Paton, M. R. Wessels, D. T. Golenbock.
2003
. Recognition of pneumolysin by Toll-like receptor 4 confers resistance to pneumococcal infection.
Proc. Natl. Acad. Sci. USA
100
:
1966
-1971.
23
Rubins, J. B., D. Charboneau, J. C. Paton, T. J. Mitchell, P. W. Andrew, E. N. Janoff.
1995
. Dual function of pneumolysin in the early pathogenesis of murine pneumococcal pneumonia.
J. Clin. Invest.
95
:
142
-150.
24
Rubins, J. B., D. Charboneau, C. Fasching, A. M. Berry, J. C. Paton, J. E. Alexander, P. W. Andrew, T. J. Mitchell, E. N. Janoff.
1996
. Distinct roles for pneumolysin’s cytotoxic and complement activities in the pathogenesis of pneumococcal pneumonia.
Am. J. Respir. Crit. Care Med.
153
:
1339
-1346.
25
Benton, K. A., J. C. Paton, D. E. Briles.
1997
. The hemolytic and complement-activating properties of pneumolysin do not contribute individually to virulence in a pneumococcal bacteremia model.
Microb. Pathog.
23
:
201
-209.
26
Alexander, J. E., A. M. Berry, J. C. Paton, J. B. Rubins, P. W. Andrew, T. J. Mitchell.
1998
. Amino acid changes affecting the activity of pneumolysin alter the behaviour of pneumococci in pneumonia.
Microb. Pathog.
24
:
167
-174.
27
Berry, A. M., J. C. Paton.
2000
. Additive attenuation of virulence of Streptococcus pneumoniae by mutation of the genes encoding pneumolysin and other putative pneumococcal virulence proteins.
Infect. Immun.
68
:
133
-140.
28
Brown, J. S., S. M. Gilliland, D. W. Holden.
2001
. A Streptococcus pneumoniae pathogenicity island encoding an ABC transporter involved in iron uptake and virulence.
Mol. Microbiol.
40
:
572
-585.
29
Ogunniyi, A. D., R. L. Folland, D. E. Briles, S. K. Hollingshead, J. C. Paton.
2000
. Immunization of mice with combinations of pneumococcal virulence proteins elicits enhanced protection against challenge with Streptococcus pneumoniae.
Infect. Immun.
68
:
3028
-3033.
30
Wessels, M. R., P. Butko, M. Ma, H. B. Warren, A. L. Lage, M. C. Carroll.
1995
. Studies of group B streptococcal infection in mice deficient in complement component C3 or C4 demonstrate an essential role for complement in both innate and acquired immunity.
Proc. Natl. Acad. Sci. USA
92
:
11490
-11494.
31
Matsumoto, M., W. Fukuda, A. Circolo, J. Goellner, J. Strauss-Schoenberger, X. Wang, S. Fujita, T. Hidvegi, D. D. Chaplin, H. R. Colten.
1997
. Abrogation of the alternative complement pathway by targeted deletion of murine factor B.
Proc. Natl. Acad. Sci. USA
94
:
8720
-8725.
32
Botto, M., C. Dell’Agnola, A. E. Bygrave, E. M. Thompson, H. T. Cook, F. Petry, M. Loos, P. P. Pandolfi, M. J. Walport.
1998
. Homozygous C1q deficiency causes glomerulonephritis associated with multiple apoptotic bodies.
Nat. Genet.
19
:
56
-59.
33
Beuzon, C. R., D. W. Holden.
2001
. Use of mixed infections with Salmonella strains to study virulence genes and their interactions in vivo.
Microbes Infect.
3
:
1345
-1352.
34
Hammerschmidt, S., G. Bethe, P. H. Remane, G. S. Chhatwal.
1999
. Identification of pneumococcal surface protein A as a lactoferrin-binding protein of Streptococcus pneumoniae.
Infect. Immun.
67
:
1683
-1687.
35
Hakansson, A., H. Roche, S. Mirza, L. S. McDaniel, A. Brooks-Walter, D. E. Briles.
2001
. Characterization of binding of human lactoferrin to pneumococcal surface protein A.
Infect. Immun.
69
:
3372
-3381.
36
Ogunniyi, A. D., P. Giammarinaro, J. C. Paton.
2002
. The genes encoding virulence-associated proteins and the capsule of Streptococcus pneumoniae are upregulated and differentially expressed in vivo.
Microbiology
148
:
2045
-2053.
37
Berry, A. M., R. A. Lock, D. Hansman, J. C. Paton.
1989
. Contribution of autolysin to virulence of Streptococcus pneumoniae.
Infect. Immun.
57
:
2324
-2330.
38
Balachandran, P., S. K. Hollingshead, J. C. Paton, D. E. Briles.
2001
. The autolytic enzyme LytA of Streptococcus pneumoniae is not responsible for releasing pneumolysin.
J. Bacteriol.
183
:
3108
-3116.
39
Spreer, A., H. Kerstan, T. Bottcher, J. Gerber, A. Siemer, G. Zysk, T. J. Mitchell, H. Eiffert, R. Nau.
2003
. Reduced release of pneumolysin by Streptococcus pneumoniae in vitro and in vivo after treatment with nonbacteriolytic antibiotics in comparison to ceftriaxone.
Antimicrob. Agents Chemother.
47
:
2649
-2654.
40
Duthy, T. G., R. J. Ormsby, E. Giannakis, A. D. Ogunniyi, U. H. Stroeher, J. C. Paton, D. L. Gordon.
2002
. The human complement regulator factor H binds pneumococcal surface protein PspC via short consensus repeats 13 to 15.
Infect. Immun.
70
:
5604
-5611.