Role of Capsule and Suilysin in Mucosal Infection of Complement-Deficient Mice with Streptococcus suis

Abstract

Virulent Streptococcus suis serotype 2 strains are invasive extracellular bacteria causing septicemia and meningitis in piglets and humans. One objective of this study was to elucidate the function of complement in innate immune defense against S. suis. Experimental infection of wild-type (WT) and C3^−/− mice demonstrated for the first time that the complement system protects naive mice against invasive mucosal S. suis infection. S. suis WT but not an unencapsulated mutant caused mortality associated with meningitis and other pathologies in C3^−/− mice. The capsule contributed also substantially to colonization of the upper respiratory tract. Experimental infection of C3^−/− mice with a suilysin mutant indicated that suilysin expression facilitated an early disease onset and the pathogenesis of meningitis. Flow cytometric analysis revealed C3 antigen deposition on the surface of ca. 40% of S. suis WT bacteria after opsonization with naive WT mouse serum, although to a significantly lower intensity than on the unencapsulated mutant. Ex vivo multiplication in murine WT and C3^−/− blood depended on capsule but not suilysin expression. Interestingly, S. suis invasion of inner organs was also detectable in C5aR^−/− mice, suggesting that chemotaxis and activation of immune cells via the anaphylatoxin receptor C5aR is, in addition to opsonization, a further important function of the complement system in defense against mucosal S. suis infection. In conclusion, we unequivocally demonstrate here the importance of complement against mucosal S. suis serotype 2 infection and that the capsule of this pathogen is also involved in escape from complement-independent immunity.

INTRODUCTION

Invasive Streptococcus suis diseases such as septicemia and meningitis account for major economic losses in the swine industry. Furthermore, S. suis is an important zoonotic pathogen, causing mainly meningitis in adult humans (1–3). S. suis is a very diverse organism, and different serotypes are responsible for morbidity in piglets. Serotype 2 strains are important worldwide for diseases in piglets and are by far the leading serotype associated with zoonotic S. suis cases (4).

The complement system is involved in innate and adaptive immune responses. It might be activated by three different routes: the classical, the alternative, and the mannose-binding lectin pathways. All three pathways lead to the formation of C3 convertases (C3Bp or C4b2a) cleaving C3 into the anaphylatoxin C3a and the most important opsonin, C3b. In addition to opsonization of bacteria with C3b, the formation of important cytokine-like peptides, in particular C3a and C5a, is a main function of the complement system. C5aR is the main receptor for C5a and is highly expressed by cells of myeloid origin (5). It should be noted that the impact of the complement system on host defense against S. suis has not been investigated in vivo. Some virulence factors are known to be involved in resistance of serotype 2 strains against killing by neutrophils in the presence of complete serum (6, 7), but only a few studies have specifically addressed putative complement evasion strategies of S. suis.

The polysaccharide capsule of S. suis is an essential virulence factor (8, 9). It protects S. suis against killing by macrophages and neutrophils in vitro. The capsule of serotype 2 strains contains sialic acid (10, 11). Surface-associated sialic acid interferes with the activation of the alternative complement cascade by increasing the affinity constant of C3b to the complement inhibitor factor H (12, 13). However, S. suis serotype 2 expresses also a cell wall-anchored factor H-binding protein (Fhb) which has been shown to be crucial for virulence in piglets (14). As an anti-human factor H serum had no effect on the C3b deposition on the surface of the isogenic ΔFhb mutant, it was concluded that Fhb is the only factor of S. suis binding human factor H (14).

Suilysin is a cholesterol-dependent pore-forming cytolysin expressed by many virulent S. suis strains (15, 16). Intraperitoneal infections of mice indicated that suilysin expression is essential for virulence of S. suis in mice (17). However, a sly knockout mutant was not attenuated in virulence in experimental infection of piglets (18). Interestingly, suilysin expression contributes to resistance against killing by porcine neutrophils and dendritic cells in the presence of active but not inactive serum (19, 20). This finding led to the speculation that suilysin leads to reduced complement deposition on the bacterial surface, as has been demonstrated for pneumolysin (21).

Mice have frequently been used as a model to study the pathogenesis of S. suis diseases using intraperitoneal application (22–24). A mucosal mouse model for S. suis meningitis has not until now been available. We recently described an intranasal colonization model for S. suis serotype 2 in C57BL/6J mice (25). Colonization of mucosal surfaces is regarded to be the first step in the pathogenesis of S. suis diseases in piglets. Therefore, it was reasonable to assume that innate immune defense mechanisms prevented further progress of invasion in the murine colonization model of a virulent serotype 2 strain. In the present study, we investigated the working hypothesis that complement might be crucial for protection against S. suis invasion. Furthermore, we used complement-deficient mice to study the impact of the capsule and suilysin on the evasion of complement-independent host defense.

MATERIALS AND METHODS

Materials and reagents.

Unless stated otherwise, all materials and reagents were purchased from Sigma (Munich, Germany).

Bacterial strains and growth conditions.

S. suis wild-type (WT) strain 10, kindly provided by H. Smith (Lelystad, Netherlands), is a virulent serotype 2 strain that expresses suilysin, muramidase-released protein, extracellular factor, opacity factor of S. suis, and the immunoglobulin M-degrading enzyme of S. suis (IdeSsuis) (8, 26, 27). It has been used by different groups successfully for experimental intranasal infections of piglets (8, 26) and recently by our group in intranasal infection of mice (25). The unencapsulated isogenic mutant 10cpsΔEF and the suilysin mutant 10Δsly were generated in previous studies by insertion of spectinomycin and erythromycin resistance gene cassettes into genes involved in capsule biosynthesis (cpsE and cpsF) and the suilysin gene of strain 10, respectively (8, 28).

Streptococci were grown on Columbia agar supplemented with 7% sheep blood (Oxoid, Wesel, Germany). For detection of the unencapsulated mutant 10cpsΔEF in the competition experiment, Columbia agar supplemented with 7% sheep blood and 100 μg of spectinomycin/ml were used. Cultivation was conducted overnight under aerobic conditions at 37°C.

Experimental intranasal infection of mice.

For experimental S. suis infection and as source of murine blood, the following mice strains with a C57BL/6J background were used at an age of 5 weeks: C57BL/6J WT mice (Charles River WIGA, Sulzfeld, Germany), C3-deficient (C3^−/−) mice (B6.129S4–C3^tm1Crr/J) (29), and C5aR-deficient (C5aR^−/−) mice (B6.129S4–C5ar1^tm1Cge/J) (30). Intranasal infections were conducted as described previously (25), except that a lower dose was used (see below). Briefly, mice were predisposed to infection through application of 5 μl of 1% acetic acid (pH 4.0) in each nostril 1 h prior intranasal infection (conducted in anesthesia via inhalation of isoflurane (IsoFlo, Albrecht, Germany). After controlled recovery, mice were infected using additional anesthesia. Three differently designed infection experiments with C3^−/− and WT mice were conducted. First, C3^−/− and WT mice were infected with 2 × 10⁸ CFU of either S. suis WT (strain 10), 10cpsΔEF, or 10Δsly strain grown to late exponential growth phase (optical density at 600 nm [OD₆₀₀] of 0.8) for the determination of Kaplan-Meier diagrams. The inoculum was applied in two drops of 12.5 μl placed in front of the nostrils. Mice were monitored for 20 days if they were not killed earlier in the case of severe clinical signs. Monoinfections were repeated once using, each time, three mice for infection with S. suis 10cpsΔEF and four mice for infection with S. suis WT (strain 10) and 10Δsly. Second, C3^−/− and WT mice were infected with S. suis WT (strain 10) and 10Δsly as described above but sacrificed 4 days postinfection (dpi) for comparative histological and bacteriological screenings. Experiments were repeated once, including 5 (4) mice in each infection group in the first (second) trial. Furthermore, a competitive infection experiment was conducted with C3^−/− and WT mice. For this, suspensions of S. suis WT (strain 10) and 10cpsΔEF at 10⁸ CFU/ml for each strain were mixed, and two drops of 12.5 μl of this mixture were placed in front of the nostrils. Inoculum concentrations were verified by plating 10-fold serial dilutions. All mice of the competitive infection experiment were sacrificed 2 dpi for determination of bacteriology and pathohistology of inner organs. A further competitive infection experiment was carried out the same way but with C5aR^−/− mice instead of C3^−/− mice. In the competitive infection experiments, the following numbers of mice per group were used in the first (second) trials: 5 (4) in WT versus C3^−/− and 4 (3) in WT versus C5aR^−/− mice.

The animal experiments of the present study were approved by the Committee on Animal Experiments of the Lower Saxonian State Office for Consumer Protection and Food Safety (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, LAVES; permits 33.14-42502-04-12/0742 and 33.12-42502-04-13/1214). The present study was performed in strict accordance with the principles and recommendations outlined in the European Convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes (European Treaty Series, no. 123 [http://www.conventions.coe.int/Treaty/en/Treaties/Html/123.htm]) and the German Animal Protection Law (Tierschutzgesetz).

Clinical score.

The health status of the mice was examined twice daily using a clinical score sheet (25), including weight development, clinical signs of general sickness (rough coat, rapid breathing, and dehydration), and clinical signs indicating meningitis (apathy, apraxia) or septicemia (swollen eyes, depression). A cumulative score of 3 to 4 indicated mild clinical signs, a score of 5 to 6 indicated moderate clinical signs, and a score >6 indicated severe clinical signs with specific regard to neural failure. Mice with a cumulative score of ≥3 were classified as diseased (calculation of morbidity). In the case of severe weight loss (>20%) and/or enduring severe clinical signs, mice were euthanized for reasons of animal welfare by inhalation of CO₂ and blood removal immediately after CO₂ inhalation.

Histological screening.

Necropsy and sampling of organs for bacteriological and histological investigations were conducted exactly as described previously (25). Fibrinosuppurative and purulent lesions were scored as originally described for piglets (26) and subsequently modified for mice (25). The group score ω was calculated by dividing the sum of the highest scores of each animal for any of the investigated organs through the number of animals. Rhinitis was not included.

Immunohistochemistry.

Formalin-fixed and paraffin-embedded nose and brain tissue was evaluated for the presence of streptococcal antigen by immunohistochemistry using the avidin-biotin complex (ABC) method. Sections were deparaffinized in descending series of ethanol and pretreated with sodium-citrate buffer (10 mM, pH 6.0) in the microwave (800 W, 20 min). Primary antibodies consist of rabbit anti-S. suis polyclonal antibodies (diluted 1:2,000). Binding of secondary goat anti-rabbit antibodies and the formation of the ABC were visualized by a chromogen reaction using 3.3-diaminobenzidine-tetrachloride (Vector Laboratories, Burlingame, CA) (31). For negative controls, primary antibodies were replaced by rabbit preimmune serum.

Reisolation of S. suis strains from tissue and TNL.

Determination of the specific bacterial load of each organ and tracheonasal lavage (TNL) was conducted as described previously (25). Briefly, tissue suspensions in phosphate-buffered saline (PBS; pH 7.4) were homogenized, serially diluted, and plated on blood agar plates (in the case of competitive infection experiments with or without spectinomycin). The detection limits for the specific bacterial loads were 1 CFU/mg of organ and 30 CFU/ml of TNL. The competitive index (CI) was calculated by dividing the specific content of the test strain (10cpsΔEF) to the specific content of the WT strain (strain 10) using spectinomycin resistance of strain 10cpsΔEF for differentiation. Typical alpha-hemolytic streptococci were profiled in a S. suis multiplex PCR for the detection of mrp, epf, sly, arcA, gdh, cps1, cps2, cps7, and cps9 (32). Isolates from mice challenged with 10Δsly were additionally investigated in a sly-specific PCR according to a described protocol (25). Isolates from mice challenged with 10cpsΔEF were differentiated in a cps2E-specific PCR with the primer pair cps2Efor (TTTCGCACTTTCAAGACGTG) and cps2Erev (GGACGGGTACCGACTAGACTC) at a final concentration of 0.5 μM.

Flow cytometric analysis of C3 antigen (C3, C3b, and C3c) deposition on the bacterial surface.

Bacteria were grown to late logarithmic growth phase (OD₆₀₀ 0.8), harvested by centrifugation, washed with PBS (pH 7.4), and resuspended in 10 μl of mouse serum (either serum from WT mice or serum from C3^−/− mice) and incubated for 45 min at 37°C. Bacteria were centrifuged, washed with PBS, and incubated with an fluorescein isothiocyanate-conjugated antibody directed against C3c (Dako) at a 1:225 dilution in PBS for 1 h at room temperature. Bacteria were washed with PBS and then resuspended in PBS with 0.375% formaldehyde for further flow cytometric analysis.

Fluorescent bacteria were measured using BD Accuri (Becton Dickinson [BD], 488-nm solid state). Further analysis was performed with BD Accuri C6 software. For each determination, 10,000 events were acquired, and initial analysis of bacterial cells was carried out by dot plot analysis (forward scatter versus sideward scatter) to define the cell population of interest. Subsequently, fluorescent bacteria were detected at channel FL-1. The fluorescence index was calculated by multiplying the percentage of bacteria positive for C3 by the mean fluorescence intensity of the whole bacterial population (33).

Bactericidal assay.

Killing of S. suis in blood was investigated in freshly drawn heparinized blood from either WT or C3^−/− mice. Sampling of blood was conducted according to the recommendations of the German Society for Laboratory Animal Science (Gesellschaft für Versuchstierkunde) and the German Veterinary Association for the Protection of Animals (Tierärztliche Vereinigung für Tierschutz e. V.) (http://www.gv-solas.de). Stocks of frozen bacterial suspensions including 15% glycerol were thawed prior to the bactericidal assay, and 1.5 × 10³ CFU were mixed with 50 μl of heparinized blood and 10 μl of either RPMI (Gibco/Invitrogen) or porcine hyperimmune serum raised against S. suis WT (strain 10). Incubation was conducted at 37°C for 2 h in a horizontal shaker. The multiplication factor (MF) represents the ratio of CFU after the 2 h of incubation to the CFU at time zero.

ELISA for detection of the anaphylatoxin C3a.

EDTA plasma was collected by cardiac puncture from S. suis WT (strain 10)- and 10Δsly-infected WT and C3^−/− mice sacrificed 4 dpi and, for comparison, from noninfected WT and C3^−/− mice. For the detection of C3a/C3a-desArg in EDTA plasma by enzyme-linked immunosorbent assay (ELISA), purified rat anti-mouse C3a antibody (clone I87-1162; BD), and biotinylated rat anti-mouse C3a (clone I87-419; BD) antibody were used as described elsewhere (34). As a standard sample, zymosan-activated EGTA plasma was included in each run. In addition, purified murine C3a (BD Pharmingen, Heidelberg, Germany) was used as a standard to calculate actual peptide concentrations. The specificity of the C3a/C3a-desArg ELISA was confirmed using zymosan-activated EGTA plasma from C3^−/− mice as a negative control.

Statistical analysis.

If not stated otherwise, experiments were performed at least five times, and results were expressed as means and standard deviations. Statistical analysis of Kaplan-Meier diagrams was conducted with the log-rank test. Differences in bacterial loads of inner organs and pathohistological, as well as clinical scores, were analyzed with the Mann-Whitney U-test. The prevalences of meningitis were compared using the Fisher exact test. Differences in flow cytometric data and multiplication in murine blood were analyzed with the Student t test. P values of <0.05 were considered significant.

RESULTS

Role of complement in protective innate immunity of mice against S. suis invasion.

Kaplan-Meier diagrams for mortality of WT and C3^−/− mice infected intranasally with S. suis serotype 2 (WT strain 10) were determined to elucidate the impact of complement on host defense against S. suis (Fig. 1A). For this, we used the protocol for a recently described intranasal colonization model (25). In agreement with our previous results, none of the WT mice infected with S. suis WT in the present study died or developed severe signs of disease (Fig. 1A; see also Fig. S1 and S2A and B in the supplemental material). In contrast, all 8 C3^−/− mice died or had to be killed within 5 days after infection because of severe clinical signs, such as weight loss greater than 20% (4 of 8) and central nervous dysfunction and lethargy (8 of 8; Fig. 1A; see also Fig. S2A and B in the supplemental material).

FIG 1.

S. suis serotype 2 causes meningitis and mortality in complement-deficient (C3^−/−) but not in wild-type (WT) mice after mucosal infection. (A) Kaplan-Meier diagram for mortality of WT and C3^−/− mice intranasally infected with 2 × 10⁸ CFU of S. suis WT (strain 10) as described in Materials and Methods. ***, Significant difference between WT and C3^−/− mice (P < 0.001; log-rank test). (B) Immunolabeling of extracellular leukocyte-associated bacteria in the meninges of a C3^−/− mouse infected with S. suis WT (strain 10) using an antiserum raised against S. suis serotype 2. Scale bar, 200 μm.

Mortality in C3^−/− mice was associated with severe lesions in numerous tissues, including multifocal suppurative meningitis and pneumonia in 4 of 8 animals (Table 1). Accordingly, the group pathological score ω (ranging from 0 to 5) was high in C3^−/− mice infected with the S. suis WT for determination of the Kaplan-Meier survival diagram (ω = 3.1, Table 1). The S. suis challenge strain was detectable in the lungs, spleens, hearts, kidneys, and brains of all C3^−/− mice. The specific bacterial load of these organs ranged between 10³ and 10⁵ CFU/mg tissue (see Fig. S3A in the supplemental material). Furthermore, S. suis antigen was detectable by immunohistochemistry inside immune cells and extracellularly in meningitis lesions of C3^−/− mice, demonstrating unequivocally that these lesions were caused by S. suis bacteria which had invaded the central nervous system (Fig. 1B and, for comparison, see Fig. S4A in the supplemental material). In contrast, S. suis was not recorded in any inner organ of the surviving WT mice sacrificed 20 dpi with the exception of one isolate from the kidney (see Fig. S3A in the supplemental material). Furthermore, S. suis WT was not detectable in the TNL fluid of these WT mice, suggesting that these mice had substantially reduced or even eliminated S. suis from the respiratory mucosa within 20 days after infection (see Fig. S3A in the supplemental material).

TABLE 1.

Histological scoring of fibrinosuppurative and purulent necrotizing lesions of WT and C3^−/− mice infected intranasally with S. suis WT (strain 10), the isogenic suilysin mutant (10Δsly), and the isogenic unencapsulated mutant (10cpsΔEF) for determination of Kaplan-Meier diagrams

S. suis strain	dpia	Mouse		No. of mice/total no. of mice per scoreb												ωc
		No.	Strain	Nose (rhinitis)			Spleen, kidney, liver, heart [splenitis, (peri-)nephritis (peri-)hepatitis, epi-/myo-/endocarditis]			Lung (pneumonia, pleuritis)			Brain and spinal cord (meningitis, encephalitis)
				4	2	1	4	2	1	4	2	1	5	3	1
WT (strain 10)	20	8	WT	1/8	2/8	2/8	0/8	3/8	1/8	0/8	1/8	0/8	0/8	0/8	0/8	1.0
	3. 4. 5	8	C3−/−	4/8	3/8	1/8	4/8	1/8	0/8	4/8	0/8	3/8	4/8	0/8	1/8	3.1
10Δsly	15, 20	8	WT	0/8	4/8	1/8	0/8	1/8	3/8	0/8	0/8	1/8	0/8	0/8	0/8	0.6
	1, 5, 7, 20	8	C3−/−	0/8	2/8	1/8	6/8*	1/8*	0/8	3/8	2/8	2/8	2/8	0/8	0/8	3.6
10cpsΔEF	20	6	WT	0/6	0/6	2/6	0/6	0/6	3/6	0/6	0/6	0/6	0/6	0/6	0/6	1.0
	20	6	C3−/−	0/6	0/6	0/6	0/6	2/6	1/6	0/6	0/6	0/6	0/6	0/6	1/6	0.8

^adpi, days postinfection on which mice were sacrificed for reasons of humanity or because of the end of the experiment (20 dpi).

^bScores of 4 and 5 indicate moderate to severe diffuse or multifocal fibrinosuppurative or purulent necrotizing inflammations of the indicated tissue. Scores of 2 and 3 indicate mild focal fibrinosuppurative or purulent necrotizing inflammation of the respective organs. Individual single perivascular immune cells received a score of 1. *, Values associated with infiltration of pectoral and abdominal cavities with neutrophilic granulocytes and fibrin in two cases.

^cω = Σ score_max/n_animals (25). Rhinitis was not included in the score ω.

The pathological and bacteriological comparisons in the described experiment were hampered by the fact that all WT mice were sacrificed 20 dpi, substantially later than S. suis WT-infected C3^−/− mice, which died within the first 5 days after experimental infection. Therefore, an additional experiment was conducted to compare histopathological lesions and bacterial loads of inner organs of WT and C3^−/− mice at the same time point, namely, 4 dpi. None of the WT mice had severe lesions of inner organs, whereas 6 of 9 C3^−/− mice showed multifocal suppurative meningitis and other pathologies by 4 dpi. The pathological group score ω was significantly higher in S. suis WT-infected C3^−/− mice at 4 dpi in comparison to the respective WT mice (ω = 4.0 versus ω = 1.1, respectively [P = 0.0017]; Table 2). These results demonstrated that most of the C3^−/− mice, in contrast to the WT mice, developed severe lesions of the inner organs, including the brain, within 4 days of mucosal S. suis infection.

TABLE 2.

Scoring of fibrinosuppurative and purulent necrotizing lesions of WT and C3^−/− mice 4 days after intranasal infection with S. suis WT (strain 10, mrp⁺ epf⁺ sly⁺ cps2⁺) and its suilysin mutant (10Δsly)

S. suis strain	dpia	Mouse		No. of mice/total no. of mice per scoreb												ωc
				Nose (rhinitis)			Spleen, kidney, liver, heart [splenitis, (peri-)nephritis (peri-)hepatitis, epi-/myo-/endocarditis]			Lung (pneumonia, pleuritis)			Brain and spinal cord (meningitis, encephalitis)
		No.	Strain	4	2	1	4	2	1	4	2	1	5	3	1
WT (strain 10)	4	9	WT	5/9	3/9	1/9	0/9	5/9	0/9	0/9	1/9	0/9	0/9	0/9	0/9	1.1
	4	9	C3−/−	5/9	2/9	0/9	3/9	6/9	0/9	1/9	2/9	1/9	6/9	0/9	0/9	4.0*
10Δsly	4	9	WT	2/9	4/9	0/9	1/9	3/9	0/9	0/9	0/9	0/9	0/9	0/9	0/9	1.9
	4	9	C3−/−	5/9	1/9	0/9	4/9	4/9	0/9	0/9	3/9	0/9	0/9	1/9	0/9	2.3†

^aDays postinfection, on which mice were sacrificed.

^bScores of 4 and 5 indicate moderate to severe diffuse or multifocal fibrinosuppurative or purulent necrotizing inflammations of the indicated tissue. Scores of 2 and 3 indicate mild focal fibrinosuppurative or purulent necrotizing inflammation of the respective organs. Individual single perivascular immune cells received a score of 1.

^cω = Σ score_max/n_animals (25). Rhinitis was not included in the score ω. Differences between WT and C3^−/− mice infected with S. suis WT (strain 10) and differences between C3^−/− mice infected either with S. suis WT (strain10) or 10Δsly were significant (*, P < 0.01; †, P < 0.05).

The S. suis WT challenge strain was detectable at 4 dpi in 8 of 9 C3^−/− mice in the spleen and the liver, as well as in the heart, and in 7 of 9 C3^−/− mice also in the lung and the brain. The specific bacterial load of these organs ranged between 10¹ and 10⁵ CFU/mg of tissue (Fig. 2). In contrast, S. suis was less often recorded in inner organs of WT mice by 4 dpi. None of the S. suis WT-infected WT mice tested positive for the challenge strain in more than one inner organ, and in the majority of these WT mice (5 of 9) S. suis was not detected at all in any inner organ. However, the specific bacterial load of S. suis WT in TNL was higher in WT mice compared to C3^−/− mice at 4 dpi (Fig. 2), indicating that complement deficiency does not promote initial mucosal colonization of S. suis.

FIG 2.

Specific bacterial loads of tracheonasal lavage (TNL) and indicated inner organs of WT (●) and C3^−/− mice (○) at 4 dpi with 2 × 10⁸ CFU of S. suis WT (strain 10 [A]) or its isogenic suilysin mutant (10Δsly [B]). Medians are indicated by horizontal bars. Significant differences between WT and C3^−/− mice are indicated (*, P < 0.05; **, P < 0.01; ***, P < 0.001), as are significant differences in the bacterial loads of S. suis WT (strain 10) and 10Δsly (+, P < 0.05 [Mann-Whitney U-test]).

In summary, these results demonstrated that a functional complement system is crucial for protection against invasive fatal S. suis infection in an intranasal mouse model. Comparative intranasal infection of C3^−/− mice with S. suis serotype 2 and respective isogenic mutants might be used to identify factors involved in complement-independent virulence mechanisms, since S. suis WT-infected C3^−/− mice developed fatal disease associated with meningitis and other pathologies.

Immunohistochemical detection of S. suis in the nasal cavity of WT-infected C3^−/− mice.

Since the upper respiratory tract is the port of entry for S. suis in our mouse model and also in field infections of piglets, we investigated the localization of S. suis in the nasal cavity by immunohistochemistry in diseased C3^−/− mice infected with S. suis WT. S. suis antigen was mainly detectable in the mucus of the nasal cavity (Fig. 3A and B and, for comparison, see Fig. S4B in the supplemental material). Parts of the mucus displayed S. suis antigen extracellularly and within leukocytes (Fig. 3A). Other regions of the mucus showed immunolabeling of S. suis bacteria without adjacent infiltrating immune cells (Fig. 3B). In addition, S. suis bacteria and leukocytes were detected in necrotizing lesions of the nasal epithelium (Fig. 3C and, for comparison, see Fig. S4C in the supplemental material). Bacterial antigen was not found within or adherent to epithelial cells. In conclusion, immunohistochemical investigation of diseased C3^−/− mice indicated that S. suis is in these animals either embedded in the mucus or associated with necrotic lesions in the nasal cavity.

FIG 3.

S. suis antigen in the nasal cavity of diseased C3^−/− mice was detected via immunohistochemistry mainly in association with mucus but focally also in necrotic lesions. (A) The mucus was infiltrated with leukocytes displaying immunohistochemically labeled S. suis antigen intracellularly. Scale bar, 200 μm (inset scale bar, 50 μm). (B) Immunolabeling of S. suis bacteria entrapped in the mucus without infiltrating immune cells. Scale bar, 200 μm. (C) Necrotic lesions were observed with numerous immunohistochemically labeled S. suis bacteria and leukocytes positive for S. suis antigen. No obvious differences were observed in this immunohistochemical investigation between S. suis WT (strain 10)- and 10Δsly-infected C3^−/− mice (panels A and B are from a WT [strain 10]-infected mouse, and panel C is from a 10Δsly-infected C3^−/− mouse, but very similar findings were observed in both groups).

Role of the capsule of S. suis in virulence and short-term colonization in C3^−/− mice.

Since the capsule of S. suis is an important virulence factor protecting this pathogen against phagocytosis, we investigated the virulence of an unencapsulated mutant of S. suis serotype 2 in C3^−/− mice. In contrast to C3^−/− mice infected with WT strain 10, none of the C3^−/− mice infected with the isogenic unencapsulated mutant 10cpsΔEF died or developed severe clinical signs (Fig. 4; see also Fig. S2E and F in the supplemental material). Only one mouse of this group had a clinical score of >3, indicating disease. Differences in morbidity and mortality of C3^−/− mice infected with WT strain 10 and the unencapsulated mutant were significant (P = 0.01 and P = 0.002, respectively). Severe or moderate fibrinosuppurative lesions were not detectable in the organs of C3^−/− mice infected with the unencapsulated mutant, leading to a low pathological group score (ω = 0.8; Table 1).

FIG 4.

Kaplan-Meier diagram for mortality of C3^−/− mice intranasally infected with 2 × 10⁸ CFU of S. suis WT (strain 10), the isogenic suilysin mutant (10Δsly), and the unencapsulated mutant (10cpsΔEF) as described in Materials and Methods. Significant differences between mice infected with S. suis WT, 10Δsly, and 10cpsΔEF are indicated (*, P < 0.05; **, P < 0.01 [log-rank test]).

Invasion and pathology of host tissue was also investigated in a further experiment, including the killing of all animals by 2 dpi. This experiment was conducted as a competitive infection of WT and C3^−/− mice with S. suis WT and the unencapsulated mutant. S. suis WT was detectable in all investigated inner organs of C3^−/− mice (except for the brain of one mouse) with a mean bacterial load above 10³ CFU/mg of tissue (Fig. 5A). In contrast, the inner organs of WT mice were mostly negative for S. suis and in the few positive cases were infected with <10² CFU/mg of tissue (Fig. 5A). The unencapsulated mutant was isolated only from the lung of one C3^−/− mouse but not in any other inner organ. However, the unencapsulated mutant was present in the TNL of 7 of 9 C3^−/− mice, although with a significantly reduced bacterial content compared to S. suis WT (5 × 10² ± 9 × 10² CFU/ml versus 6 × 10⁵ ± 6 × 10⁵ CFU/ml, respectively [P = 0.0007]; Fig. 5). The competitive indices for S. suis 10cpsΔEF versus WT in TNL were 0.00006 ± 0.00034 and 0.0018 ± 0.000174 for WT and C3^−/− mice, respectively. Differences in these indices for WT and C3^−/− mice were statistically significant (P = 0.027). Pathohistological screening of the coinfected C3^−/− mice but not the coinfected WT mice often revealed moderate or severe suppurative lesions in the lung, spleen, heart, and liver (Table 3). Lesions in the brain were not recorded (in accordance with the early time point of sacrifice). The pathological group score ω of C3^−/− mice obtained at 2 dpi already a value of 3.3, which was substantially higher than the score ω for infected WT mice (ω = 1.4, P = 0.059).

FIG 5.

Reisolation of S. suis WT (strain 10 [A]) and the unencapsulated mutant (10cpsΔEF [B]) after competitive intranasal infection of WT (●) and C3^−/− (○) mice at 2 dpi. Each mouse was infected with 10⁸ CFU each of S. suis WT (strain 10) and 10cpsΔEF (ratio of 1:1). The specific bacterial loads in the TNL and in the indicated inner organs were determined using the spectinomycin-resistant phenotype of 10cpsΔEF, as described in Materials and Methods. For each location, one reisolated colony was investigated in a multiplex PCR for profiling of S. suis and in a cps-specific PCR (32). Medians are indicated by horizontal bars. 10cpsΔEF was not detectable in inner organs except for the lungs of one C3^−/− mouse. Significant differences between WT and C3^−/− mice are indicated (*, P < 0.05; **, P < 0.01; ***, P < 0.001), as are significant differences in the specific bacterial loads of S. suis WT (strain 10) and 10cpsΔEF (+, P < 0.05; ++, P < 0.01; +++, P < 0.001 [Mann-Whitney U-test]).

TABLE 3.

Histological scoring of fibrinosuppurative and purulent lesions of WT, C3^−/−, and C5aR^−/− mice competitively infected intranasally with S. suis WT (strain 10) and its unencapsulated mutant (10cpsΔEF)

S. suis straina	dpib	Mouse		No. of mice/total no. of mice per scorec												ωd
		No.	Strain	Nose (rhinitis)			Spleen, kidney, liver, heart [splenitis, (peri-)nephritis (peri-)hepatitis, epi-/myo-/endocarditis]			Lung (pneumonia, pleuritis)			Brain and spinal cord (meningitis, encephalitis)
				4	2	1	4	2	1	4	2	1	5	3	1
WT (strain 10) + 10cpsΔEF (set 1)	2	9	WT	6/9	3/9	0/9	0/9	2/9	4/9	1/9	1/9	3/9	0/9	0/9	0/9	1.4
	2	9	C3−/−	4/9	0/9	0/9	3/9	6/9	0/9	6/9	0/9	1/9	0/9	0/9	0/9	3.3
WT (strain 10) + 10cpsΔEF (set 2)	2	7	WT	3/7	1/7	1/7	0/7	1/7	3/7	0/7	1/7	0/7	0/7	0/7	0/7	0.9
	2	7	C5aR−/−	6/7	1/7	0/7	0/7	3/7	4/7	0/7	1/7	0/7	0/7	0/7	0/7	1.4

^aSet 1, competitive infection of WT and C3^−/− mice; set 2, competitive infection of WT and C5aR^−/− mice.

^bdpi, days postinfection, on which mice were sacrificed.

^cScores of 4 and 5 indicate moderate to severe diffuse or multifocal fibrinosuppurative or purulent necrotizing inflammations of the indicated tissue. Scores of 2 and 3 indicate mild focal fibrinosuppurative or purulent necrotizing inflammation of the respective organs. Individual single perivascular immune cells received a score of 1.

^dω = Σ score_max/n_animals (25). Rhinitis was not included in the score ω.

In conclusion, different infection experiments demonstrated that the capsule of S. suis is crucial for virulence in complement-deficient mice, indicating that the capsule is involved also in evasion of complement-independent innate immunity. Furthermore, a competitive infection experiment showed that the unencapsulated mutant is severely attenuated in short-term colonization of the upper respiratory tract in WT and to a lesser extent also in C3^−/− mice in this model.

Role of suilysin in infection of C3^−/− mice.

We also investigated whether suilysin was crucial for complement-independent host-pathogen interaction, leading to the devastating S. suis infection in C3^−/− mice. As shown in Fig. 4 and Table 1, experimental infection of C3^−/− mice with the isogenic suilysin mutant 10Δsly resulted in 88% mortality and severe or moderate fibrinosuppurative lesions in numerous inner organs. Accordingly, the pathohistological group score ω was high in C3^−/− mice infected with 10Δsly for the determination of the Kaplan-Meier survival diagram (ω = 3.6; Table 1). However, death occurred significantly later in 10Δsly-infected C3^−/− mice compared to S. suis WT-infected C3^−/− mice (on average, 2.8 days [P = 0.0367]; Fig. 4). Comparative histopathological screening of mice at 4 dpi revealed that the lack of suilysin led to a significant lower histopathological group score at this time point (ω = 2.3 for 10Δsly versus ω = 4.0 for WT [P = 0.03]; Table 2). The most striking difference between S. suis WT and 10Δsly-infected C3^−/− mice at 4 dpi was the significantly higher prevalence of moderate or severe meningitis in the former (6 of 9 and 0 of 9, respectively [P = 0.009]; Table 2). Differences between the two groups of mice were less distinct with regard to specific bacterial loads of different inner organs and TNL. The bacterial loads of kidney and heart were even significantly higher in 10Δsly-infected C3^−/− mice at 4 dpi (P = 0.046 and P = 0.015, respectively; Fig. 2). Although the median of the bacterial loads of brain tissue was substantially higher in WT-infected compared to 10Δsly-infected C3^−/− mice (1,301 ± 1,337 and 23 ± 1,091 CFU/mg of brain, respectively), the differences were not significant (P = 0.198). In conclusion, suilysin is not crucial for complement-independent host-pathogen interaction, leading to invasion of host tissue and fatal disease in C3^−/− mice, but contributes to an early onset and faster progression of pathology in particular suppurative meningitis.

Impact of suilysin and capsule expression on complement deposition on the bacterial surface in murine serum.

To better understand why S. suis serotype 2 infection of WT but not of C3^−/− mice was restricted to the nasal cavity, complement deposition on the bacterial surface was investigated in vitro. For this, S. suis strains were analyzed by flow cytometry with an antibody against C3 after incubation in murine sera. Importantly, >40% of S. suis WT bacteria were positive for C3 (most likely C3b/iC3b) deposition after incubation in WT serum (Fig. 6A). Incubation of bacteria in serum of C3^−/− mice was conducted as a control for this flow cytometric analysis and confirmed the specificity. The percentage of bacteria positive for C3 antigen were significantly increased in the unencapsulated mutant 10cpsΔEF and in the suilysin mutant 10Δsly compared to the WT (Fig. 6A). Furthermore, the unencapsulated mutant showed a very distinct phenotype with regard to the intensity of C3 antigen deposition on the bacterial surface, since the mean fluorescence intensity measured for the unencapsulated mutant in this assay was 5-fold higher than the values obtained for S. suis WT and the suilysin mutant (Fig. 6B). No significant differences in the intensity of C3 antigen deposition were found between the S. suis WT strain and the suilysin mutant. In conclusion, murine C3 antigen (most likely C3b/iC3b) is deposited on the surface of encapsulated S. suis, although the intensity is significantly lower than on the surface of an unencapsulated isogenic mutant.

FIG 6.

C3 deposition on the surfaces of S. suis WT (strain 10), the unencapsulated mutant (10cpsΔEF), and the suilysin mutant (10Δsly) in the presence of naive WT serum. As a control, bacteria were incubated with the sera of C3^−/− mice. Bacteria were incubated for 45 min in sera and analyzed by flow cytometry as described in Materials and Methods. (A) Proportion of bacteria positive for C3 after incubation for 45 min. (B) Intensity of C3 deposition on the surfaces of S. suis WT (strain 10), 10Δsly, and 10cpsΔEF strains in sera from WT mice and, as a control, C3^−/− mice. The means and standard deviations of five independent experiments are indicated. Significant differences are indicated (*, P < 0.05 [Student t test]).

Impact of suilysin and capsule expression on survival of S. suis in murine blood of WT and C3^−/− mice.

Since C3 antigen deposition was clearly detectable on the surfaces of S. suis WT bacteria, we comparatively evaluated survival of S. suis WT in blood of WT and C3^−/− mice. The MF of S. suis WT was significantly higher in C3^−/− blood than in WT blood (Fig. 7, P = 0.028). S. suis WT and also 10Δsly showed very high mean MF values of 63.7 ± 24 and 68.7 ± 12, respectively, in C3^−/− blood. Differences in multiplication between S. suis WT and the 10Δsly mutant were neither recorded for C3^−/− nor for WT blood. Addition of hyperimmune sera confirmed that murine blood cells were a priori capable of eliminating S. suis in WT and C3^−/− blood in the presence of specific immunoglobulins under these conditions. Interestingly, the unencapsulated mutant showed some multiplication in C3^−/− but not in WT blood in the absence of specific immunoglobulin. The MF of the unencapsulated mutant 10cpsΔEF (35.8 ± 10) was significantly lower than the MF of S. suis WT and 10Δsly in C3^−/− blood (P = 0.044 and 0.002, respectively) and comparable to the MF of S. suis WT and 10Δsly in WT blood. In summary, multiplication of S. suis in murine blood ex vivo is complement-restricted and depends on capsule but not suilysin expression in complement-deficient and WT blood.

FIG 7.

Survival of S. suis WT (strain 10), 10cpsΔEF, and 10Δsly in murine blood ex vivo (bactericidal assay). The bacterial strains were incubated at 37°C for 2 h in blood of WT and C3^−/− mice as indicated. The multiplication factor (MF) represents the ratio of CFU after the 2 h of incubation to the CFU at time zero. Porcine hyperimmune serum was added as indicated to demonstrate that efficient killing of S. suis by protective immunity is a priori possible under these conditions. Means are indicated by horizontal bars. Significant differences are indicated(*, P < 0.05; **, P < 0.01; ***, P < 0.001 [Student t test]).

Role of anaphylatoxins in innate immunity of mice against S. suis invasion.

Differences in clinics, pathology, and bacteriology between infected C3^−/− and WT mice might not only result from differences in opsonization of S. suis but also be a consequence of deficient anaphylatoxin formation in C3^−/− mice. Therefore, formation of the anaphylatoxin C3a in S. suis-infected mice and the role of the anaphylatoxin receptor C5aR in innate immunity against S. suis were also investigated. We focused on detection of the anaphylatoxin C3a/C3a-desArg in the blood of S. suis-infected mice, since C5a des-arg is rapidly bound to C5aR and C5L2 and thereby removed from the circulation (both anaphylatoxins, C3a and C5a, are metabolized within seconds in vivo to their des-arginated forms). At 4 days after intranasal infection with S. suis WT and 10Δsly, the plasma concentrations of murine C3a in WT mice were 2,617.6 ± 1,195 ng/ml and 2,910.5 ± 1,269 ng/ml, respectively, values significantly higher than in uninfected control mice (940.0 ± 736 ng of mC3a/ml [P = 0.0035 and 0.0061], respectively; see Fig. S5 in the supplemental material). As expected, C3a was not detectable in plasma of C3^−/− mice (see Fig. S5 in the supplemental material). These results showed that the anaphylatoxin C3a is formed systemically upon infection of WT mice with either S. suis WT or 10Δsly, although moderate or severe inflammations of inner organs were not recorded in these mice (with one exception; Table 2).

C5a mediates chemotaxis of granulocytes and activation of immune cells. In order to elucidate the impact of C5a-mediated signaling on host defense against S. suis, C5aR^−/− mice were infected in our intranasal model. This was conducted as a competitive infection similar to that of C3^−/− and WT mice with S. suis WT and 10cpsΔEF, including the sacrifice of all animals at 2 dpi. Similar to the previous experiment, S. suis was only detectable in inner organs of 2 of 7 WT mice. The numbers of S. suis WT bacteria were significantly higher in C5aR^−/− mice than in WT mice, but in contrast to C3^−/− mice with mean values of <10³ CFU/mg of tissue (Fig. 8A). The specific bacterial loads in the spleens, livers, hearts, and lungs were significantly lower in C5aR^−/− mice than in C3^−/− mice, as determined in our previous experiment. Accordingly, moderate or severe fibrinosuppurative or purulent lesions were not recorded in the inner organs of C5aR^−/− mice in contrast to C3^−/− mice (Table 3). Neutrophilic accumulation of the splenic red pulp was detected more often in C5aR^−/− mice than in WT mice, but differences in the pathological score reflecting fibrinosuppurative lesions of inner organs were not significant. Moderate multifocal suppurative rhinitis was a typical finding of coinfected WT, C3^−/−, and C5aR^−/− mice sacrificed at 2 dpi (Table 3). The unencapsulated mutant was not detectable in inner organs of C5aR^−/− mice except for the brain of one mouse. The load of S. suis WT bacteria in the TNL showed no differences between C5aR^−/− and WT mice (Fig. 8A). However, the unencapsulated mutant was detectable in TNL of 4 of 7 C5aR^−/− mice but only in the TNL of 1 of 7 WT mice (Fig. 8B).

FIG 8.

Reisolation of S. suis WT (strain 10 [A]) and the isogenic unencapsulated mutant (10cpsΔEF [B]) after competitive intranasal infection of WT (●) and C5aR^−/− (△) mice. Each mouse was infected with 10⁸ CFU each of S. suis WT (strain 10) and 10cpsΔEF (ratio of 1:1). The specific bacterial loads in the TNL and in the indicated inner organs were determined by using the spectinomycin-resistant phenotype of 10cpsΔEF, as described in Materials and Methods. For each location, a reisolated colony was investigated in a multiplex PCR for profiling of S. suis and in a cps-specific PCR (32). 10cpsΔEF was not detectable in inner organs except for the brain of one C5aR^−/− mouse. Medians are indicated by horizontal bars. Significant differences between WT and C5aR^−/− mice are indicated (***, P < 0.001), as are the significant differences in the specific bacterial loads of S. suis WT and 10cpsΔEF (+, P < 0.05; ++, P < 0.01; +++, P < 0.001). Significant differences between C3^−/− (Fig. 5) and C5aR^−/− mice are also indicated (#, P < 0.05; ##, P < 0.01; ###, P < 0.001). Significance was determined using the Mann-Whitney U-test.

In conclusion, complement is activated systemically in WT mice infected with S. suis intranasally, and C5aR is involved in protective innate immunity of these mice preventing high bacterial loads of different inner organs at 2 dpi after mucosal infection. However, comparison of the bacteriological and histological results of C3^−/− and C5aR^−/− infection shows that C5aR signaling is only one of the functions of the complement system involved in protective immunity against mucosal S. suis infection.

DISCUSSION

Infection of piglets with S. suis might lead to very different states of the host: the piglet might either eliminate this pathogen without showing signs of disease, become an inapparent carrier, or develop different types of diseases such as acute meningitis and chronic endocarditis. Only mucosal animal models make it possible to investigate all of these different putative outcomes and to identify the factors which determine whether initial colonization is restricted or the first step toward invasive infection. In the present study, we used an intranasal mouse model to investigate for the first time the impact of the complement system on the pathogenesis of S. suis diseases in vivo. The results unequivocally demonstrate that the complement system protects mice against invasive mucosal infection of S. suis. Although similar findings have been obtained for S. pneumoniae (35, 36), the protective role was surprising as S. suis serotype 2 was described to be resistant against complement-dependent killing by dendritic cells and neutrophils in the absence of specific antibodies (19, 20). Furthermore, Brazeau et al. (37) found that inactivation of serum had no effect on phagocytosis of S. suis by murine macrophages. However, the protective complement-mediated immunity of WT mice against S. suis is in agreement with our in vitro findings showing that C3 antigen is deposited on the bacterial surface of S. suis WT and that multiplication of S. suis is significantly higher in the blood of complement-deficient mice.

The unencapsulated mutant was found to be severely attenuated in the infection of complement-deficient mice. This result is in accordance with previous in vitro findings of different groups showing that unencapsulated S. suis mutants are killed by different immune cells in the presence of inactivated serum (19, 20, 28). Interestingly, it has been shown that the capsule inhibits phagocytosis by macrophages through destabilization of lipid microdomains (38). This complement-independent immune evasion function of the capsule might, at least partially, explain the attenuation of the unencapsulated mutant in complement-deficient mice. Importantly, the results presented here do not exclude that the capsule is also crucial for complement evasion. This is supported by the observations that the intensity of C3b deposition on the bacterial surface is substantially enhanced in the unencapsulated mutant, as shown in the present study by flow cytometric analysis and in a study by Lecours et al. (20) by immunofluorescence microscopy. Furthermore, multiplication of the unencapsulated mutant was found to be higher in C3^−/− blood than in WT blood.

In S. pneumoniae the polysaccharide capsule contributes to colonization of the respiratory tract of WT mice (39). Here, we showed that an unencapsulated S. suis mutant is also severely attenuated in survival in the upper respiratory tract of WT and C3^−/− mice. However, complement depletion by application of cobra venom had no significant effect on colonization of the upper respiratory tract by either WT or unencapsulated pneumococci (39). In contrast, significant differences in competitive indices (S. suis WT versus the unencapsulated mutant) between WT and C3^−/− mice were recorded in the present study, which suggests that a function(s) of the capsule of S. suis in interaction with the complement system is important for survival in the upper respiratory tract in this murine model. An impact of complement on the survival of S. suis in the upper respiratory tract is in accordance with the rhinitis found in S. suis-infected mice at 2 dpi and the association of S. suis with infiltrating leukocytes and focal necrotic lesions in the nasal cavity. Further studies are needed to investigate whether focal necrosis of the respiratory epithelium is crucial for host tissue invasion by S. suis in C3^−/− mice.

Our findings demonstrated that suilysin expression is not crucial for invasive infection of complement-deficient mice leading to fatal diseases. In fact, we observed only 12% survival of C3^−/− mice infected mucosally with 10⁸ CFU of 10Δsly, but mortality occurred significantly later and histopathological lesions were less severe at 4 dpi than in S. suis WT-infected C3^−/− mice. In a previous study suilysin expression was found to be essential for S. suis to kill WT BALB/c mice after intraperitoneal application of 10⁷ and 10⁸ CFU (17). One possible reason for the difference in the clinical outcome of the two studies is that expression of suilysin is crucial for protection against complement-dependent immunity. Accordingly, it has been shown that pneumolysin expression prevents C3 deposition on the surface of pneumococci and is crucial for causing systemic infection in WT mice (21). We also recorded a higher percentage of C3-positive bacteria for the suilysin mutant than for S. suis WT. However, these differences were not as pronounced as one would expect with regard to the substantial differences in mortality between the two studies. Furthermore, we did not observe differences in multiplication of WT S. suis and the isogenic suilysin mutant in WT blood. Most likely, the different application routes used in both studies were important for the impact of suilysin on pathogenesis. Intraperitoneal application of S. suis in mice in a high doses might lead to cytotoxic effects of secreted suilysin. Extensive cytotoxicity might have substantially determined mortality after intraperitoneal application. In the nasal cavity, S. suis is mainly surrounded by mucus at the primary site of infection, as shown via immunohistological investigation. This mucus might be very important for the protection of mucosal surfaces against the cytotoxic effects of suilysin. Since suilysin expression is also not crucial for virulence of S. suis in experimental mucosal infections of piglets (without complement deficiency) (18), the results from the intraperitoneal mouse infection should be interpreted with caution.

We have shown here that suilysin contributes to the pathogenesis of S. suis meningitis, an observation that is in accordance with the findings of a very recent study comparing sequence type 1 and 104 strains (24). Suilysin is known to affect porcine choroid plexus epithelial cells (40), and cytotoxic effects of suilysin have also been reported for brain microvascular endothelial cells (41, 42). Furthermore, suilysin-stimulated release of proinflammatory molecules, such as arachidonic acid (43) and interleukin-8 (44, 45), may play an important role in initiating changes in permeability and adhesion properties of the blood brain and the blood cerebrospinal fluid barriers, promoting immune cells to enter the central nervous system.

Anaphylatoxins might influence pathogenesis in very different ways. Our results show that the anaphylatoxin C3a is formed systemically upon mucosal infection of WT mice with S. suis. Furthermore, invasion of inner organs was significantly increased in C5aR^−/− mice in comparison to WT mice. The latter suggests a protective role of C5a in mucosal infection with S. suis. C5a-mediated chemotaxis and activation of immune cells might contribute to the elimination of this pathogen at an early stage of pathogenesis. The high frequency of moderate multifocal suppurative rhinitis in C5aR^−/− mice is in agreement with the original phenotypic characterization of C5aR^−/− mice, demonstrating a paradoxical increase in the number of neutrophils recovered from the lungs of Pseudomonas aeruginosa-infected C5aR^−/− mice (30). We hypothesize that the infiltrating neutrophils in the nasal cavity were not efficiently killing S. suis in the case of C5aR deficiency. Interestingly, the genome of S. suis serotype 2 includes an open reading frame encoding a putative C5a protease (NCBI reference sequence YP_006074193.1). Our results suggest that despite this putative C5a degradation, sufficient amounts of C5a are functional in S. suis-infected mice to protect mice against S. suis invasion.

The S. suis serotype 2 strain used here causes meningitis and other pathologies in intranasal infection of normal piglets, i.e., without complement deficiency (26, 46, 47). In comparison to piglets, WT mice can be regarded as less susceptible to mucosal S. suis infections, as demonstrated here and in our previous study (25). We speculate that these differences might be related to host-specific complement evasion mechanisms. Accordingly, our laboratory recently identified an IgM protease cleaving porcine but not murine IgM (27). Since IgM is part of the innate immune response and an important activator of the classical complement pathway, we are currently investigating IgM cleavage by S. suis as a putative complement evasion mechanism. Furthermore, it is unknown whether the binding of factor H to the surface of S. suis exhibits differences between host species (14). The human pathogen Neisseria meningitidis only binds human factor H, and transgenic expression of human factor H substantially predisposes rats to bacteremia (48, 49).

In the present study, complement-deficient mice were for the first time used for S. suis research, including a new intranasal animal model for identification of complement-independent virulence mechanisms. We anticipate that this research will substantially contribute to future studies on the pathogenesis of this important mucosal pathogen.