Pneumolysin, PspA, and PspC Contribute to Pneumococcal Evasion of Early Innate Immune Responses during Bacteremia in Mice

Abstract

The pneumococcal virulence factors include capsule, PspA, PspC, and Ply. Cytometric analysis demonstrated that the greatest levels of C3 deposition were on a Δply PspA⁻ PspC⁻ mutant. Also, Ply, PspA, and PspC expression resulted in C3 degradation in vitro and in vivo. Finally, blood clearance assays demonstrated that there was enhanced clearance of Δply PspA⁻ PspC⁻ pneumococci compared to the clearance of nonencapsulated pneumococci.

Streptococcus pneumoniae possesses virulence factors that function in evasion of complement (3, 27-29). The capsular polysaccharide (CPS) is considered a key factor in complement resistance (13, 21), and the interaction of pneumococci with complement varies according to the CPS type (1, 15, 19). Ply can activate the classical pathway, diverting complement activation (26, 29). Pneumococcal surface protein A (PspA) can also interfere with complement deposition, blocking recruitment of alternative pathway (AP) proteins (4, 24). Pneumococcal surface protein C (PspC; also called CbpA and SpsA) interacts with factor H (7, 22). Factor H regulates the AP by serving as a cofactor during factor I-mediated cleavage of C3b to iC3b (10, 11, 16, 25). Studies have demonstrated that more C3b is deposited on nonencapsulated pneumococci (30) and on PspA⁻ or Ply⁻ strains (24, 31). PspC mutants are less able to inhibit AP activation (17) and have reduced virulence (9). We investigated C3 deposition, complement inactivation, and blood clearance of pneumococci in the absence of Ply, PspA, and PspC.

Pneumococcal strains, growth conditions, and CPS determination.

The pneumococci used are listed in Table 1 and include R36A, D39, and isogenic mutants of D39. LM91, TRE108, and TRE121 are insertion-duplication mutants, and ΔPLY2 and ΔPAC (generated for this study by deleting ply of TRE121) were generated by allelic replacement (29). Bacteria were grown to mid-log phase as described previously (22). When necessary, erythromycin (0.5 μg/ml), tetracycline (15 μg/ml), and trimethoprim (50 μg/ml) were added to media.

TABLE 1.

Pneumococcal strains used in complement deposition and virulence studies

Straina	Phenotype	Antibiotic resistance
D39	Wild type	None
R36A	Nonencapsulated	None
LM91	PspA−	Eryr
TRE108	PspC−	Eryr
ΔPLY2	Ply−	Tmpr
TRE121	PspC− PspA−	Eryr Tetr
ΔPAC	PspC− PspA− Ply−	Eryr Tetr Tmpr

^aWith the exception of ΔPAC, the mutant strains have been described previously. LM91 has been described by McDaniel et al. (20), TRE108 has been described by Brooks-Walter et al. (5), ΔPLY2 has been described by Thornton and McDaniel (29), and TRE121 has been described by Balachandran et al. (2). R36A is a derivative of D39, and all mutant strains are isogenic mutants of D39.

To investigate the combined role of Ply, PspA, and PspC in C3 deposition, we generated ΔPAC. Experiments with ΔPAC were repeated using independent clones from different transformations to ensure that the results were not due to inadvertent mutations. Growth curves demonstrated that growth of D39 and growth of ΔPAC were similar (not shown). The amount of CPS was determined by an enzyme-linked immunosorbent assay as previously described (8), using an anti-CPS type 2 monoclonal antibody, monoclonal antibody 2G1 (provided by M. H. Nahm). D39 and mutants of this strain had similar levels of CPS. The endpoint titers for D39, ΔPAC, and TRE121 were 1,160 ± 655, 1,350 ± 540, and 1,340 ± 530 (means ± standard errors of the means), respectively; these values were significantly different (P = 0.03) from the value for R36A (≤10). CPS of pneumococci was also detected by flow cytometry (fluorescence-activated cell sorting; FACScan cytometer; Becton Dickinson) as described previously (22) using monoclonal antibody 2G1, which demonstrated that CPS was present on viable D39 and ΔPAC (not shown). The capsules of D39 and ΔPAC were visualized following staining with Alcian Blue 8GX (Sigma) by transmission electron microscopy as described previously (14) (Fig. 1).

FIG. 1.

Detection of type 2 CPS: electron micrographs of D39 (A), ΔPAC (B), and R36A (C) stained with Alcian Blue to visualize pneumococcal CPS (indicated by arrows).

C3 deposition on the pneumococcal surface.

C3 deposition assays were performed as described previously (24), except that pneumococci (10⁵ CFU/ml) were incubated with normal human serum (NHS) (CompTech) as a complement source. C3 was then detected using goat anti-human C3-biotinylated (1:100 dilution) (CompTech) and strepavidin-conjugated Alexa Fluor 488 (Molecular Probes). Fluorescence-activated cell sorting was used to calculate the percentage of C3-positive bacteria and mean fluorescence intensities (MFI). The results demonstrated that the percentages of ΔPAC, TRE121, and single mutants that were C3 positive were consistently greater than the percentage of D39 that was C3 positive (Fig. 2A). The MFI of D39, LM91, TRE108, ΔPLY2, TRE121, R36A, and ΔPAC were 7.8 ± 4, 57.7 ± 7, 71.3 ± 8, 75.7 ± 5, 102 ± 16.1, 138 ± 39.5, and 468.4 ± 50, respectively. Histograms demonstrated that more C3 was present on ΔPAC (Fig. 2B). These data support the findings of other studies that demonstrated the synergistic roles of Ply and PspA in complement inhibition (31) and suggest that in the absence of virulence proteins, pneumococci are more vulnerable to complement than a nonencapsulated strain is.

FIG. 2.

Detection of C3 deposition on the surface of pneumococci using flow cytometry analysis. (A) Percentages of the populations that were positive for C3 deposition. Student's t test was used to compare C3 deposition data, and a P value of <0.05 was considered significant. The P values for comparisons with D39 were as follows: R36A, P = 0.0001; ΔPAC, P = 0.0002; TRE121, P = 0.006; LM91, P = 0.013; TRE108, P = 0.02; and ΔPLY2, P = 0.01. (B) Representative histograms for C3 deposition on D39 (MFI, 7.8 ± 4.0 [mean ± standard error for three or more independent experiments]), R36A (138 ± 39.5), TRE121 (102 ± 16.1), and ΔPAC (468.4 ± 50). The fluorescence intensity of C3 detected on encapsulated ΔPAC was significantly greater (P = 0.007) than that detected on nonencapsulated R36A.

In vitro and in vivo C3 processing by pneumococci expressing Ply, PspA, and PspC.

We first investigated the role of Ply, PspA, and PspC in complement inactivation using in vitro C3 degradation assays. Western analysis was performed as described previously to detect iC3b (6, 23), except that pneumococci (10⁵ CFU/ml) were incubated with 20% NHS for 30 min at 37°C. C3 fragments in supernatants were analyzed by Western blotting using an anti-human iC3b monoclonal antibody (Quidel). This antibody detects generation of iC3b, which serves as an indicator of AP activity. The remaining incubation procedures were performed as described previously (6, 23). Degradation of C3b to iC3b resulted in the appearance of 75- and 40-kDa fragments. With D39 and R36A we detected degradation of human C3b (109 kDa) to iC3b, as indicated by the appearance of 75- and 40-kDa bands. TRE121 exhibited some ability to degrade C3b, whereas ΔPAC was unable to cleave C3b as efficiently as the other strains (Fig. 3A). These results suggest that Ply, PspA, and PspC function cooperatively to degrade C3b on the pneumococcal surface.

FIG. 3.

Western blot analysis of pneumococci following complement inactivation assays. (A) Following incubation with NHS, activated human C3b (represented by the band at 109 kDa) and its cleaved fragments, iC3b (represented by bands at 75 and 40 kDa), were detected using an anti-iC3b monoclonal antibody. D39 (lane 1), R36A (lane 2), and TRE121 (lane 3) exhibited enhanced cleavage of C3b to iC3b compared to the cleavage in ΔPAC (lane 4). (B) Following challenge of mice with 10⁸ CFU of a strain, blood was collected at 5 min (lanes 1), 10 min (lanes 2), 20 min (lanes 3), and 30 min (lanes 4). C3 processing was detected using a mouse C3 antiserum. C3 (represented by bands at approximately 120 to 170 kDa) was detected in naïve mouse blood and in blood by 5 min after infection with ΔPAC, R36A, and D39. Degradation fragments of iC3b (corresponding to bands at approximately 70 and 43 kDa) were detected in D39- and R36A-infected mice by 10 min (lanes 2 to 4).

To demonstrate C3 processing in vivo, we used previously described methods (12), except that naïve CBA/N mice (Jackson Laboratories) were challenged intravenously with 10⁸ CFU of D39, R36A, or ΔPAC suspended in 0.2 ml of lactated Ringer's solution. Blood samples were collected from mice by retroorbital bleeding at 5, 10, 20, and 30 min postinfection. Serum was collected and diluted 1:25 with 2× sodium dodecyl sulfate loading buffer, and 20 μl of each sample was used in a Western analysis as described previously (12) with polyclonal goat anti-mouse C3 (Immunology Consultants Laboratory). Figure 3B shows the C3, C3b, and inactivated fragments in mouse serum at 5, 10, 20, and 30 min after infection with different strains. C3 (bands at approximately 120 to 170 kDa) was detected in naïve mouse serum and in serum 5 min after infection with ΔPAC, R36A, and D39. iC3b was more evident in serum collected after challenge with D39 or R36A, as indicated by the presence of bands at approximately 70 and 43 kDa at 10 min. This pattern of C3 inactivation was not as evident after infection with ΔPAC. These observations supported the results of the in vitro C3 assays and further suggest that expression of Ply, PspA, and PspC, independent of the CPS, accelerates C3 inactivation that could enhance pneumococcal survival.

Pneumococcal clearance during infection.

Clearance assays were performed as described previously (2, 18, 23), and groups of five CBA/N mice per challenge strain were infected intravenously with 2 × 10⁴ CFU of pneumococci (Table 1). Blood was collected at zero time, 10 and 20 min, and 24 h. Pneumococci in blood were enumerated by plating serial dilutions on blood agar. Three independent experiments were performed, and all animal experiments were conducted by following the University of Mississippi Medical Center IACUC guidelines. We also monitored the mortality of mice for up to 21 days. Data from clearance assays demonstrated that mice challenged with D39 had the highest number of pneumococci in their blood at 24 h (Fig. 4). A significant reduction in the number of nonencapsulated pneumococci in the bloodstream occurred after 20 min, whereas the numbers of the combination mutants were significantly reduced by 10 min and at 20 min (Fig. 4). This indicates that in the absence of Ply, PspA, and PspC the type 2 CPS cannot effectively evade early innate responses and suggests that the rapid clearance of TRE121 and ΔPAC could be due to their inability to successfully evade complement deposition.

FIG. 4.

Pneumococcal blood clearance assays. The numbers of CFU detected in the bloodstream following infection with TRE121 and following infection with ΔPAC were significantly reduced by 10 min (P = 0.003 [one asterisk] and P = 0.004 [two asterisks], respectively, for comparisons with R36A). The numbers of nonencapsulated R36A in the blood were significantly reduced by 20 min (P = 0.01 for a comparison with D39). Three independent experiments using groups of five mice per challenge strain were performed, and the mortality of mice was monitored for 21 days after infection. Mice challenged with D39 succumbed to infection by 36 h, and mice challenged with pneumococci lacking only one of the virulence proteins succumbed to infection between 48 and 72 h after infection. The values are the numbers of pneumococci in blood (log CFU/ml; mean ± the standard error of the mean), and Student's t test was used to compare pneumococcal clearance data. A P value of <0.05 was considered significant.

Together, our results obtained using ΔPAC extend previous observations demonstrating that PspC has an additive role in complement deposition and emphasize the importance of Ply, PspA, and PspC in the establishment of pneumococcal disease. Since other workers have used C3-deficient mice to demonstrate that virulence can be restored in PspA, PspA-Ply, and PspC mutants (13, 31), it would be of value to use complement-deficient mice to investigate complement-independent virulence mechanisms employed by ΔPAC. The results of such studies using Ply-, PspA-, and PspC-deficient mutants belonging to different serotypes may identify common complement-dependent and -independent pneumococcal virulence mechanisms.