| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Neurology, Klinikum Grosshadern, Ludwig Maximilians-University, Munich, Germany;
Max von Pettenkofer-Institute for Hygiene and Microbiology, Ludwig Maximilians-University, Munich, Germany; and
Rheumatology Section, Division of Medicine, Faculty of Medicine, Imperial College, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
10% for all invasive diseases, but it is significantly higher for pneumococcal meningitis than for pneumonia or bacteremia, reaching up to 34% (3, 4). The higher mortality of meningitis is linked to the inefficiency of the host immune response within the CNS to control and overcome S. pneumoniae infection ("local immunodeficiency" within the CNS). Although protective anti-capsular Abs can be produced following immunization, the innate immune response is clearly important in controlling pneumococcal infection in the nonimmune host (5, 6). Innate resistance to systemic infection with S. pneumoniae (outside the immunoprivileged CNS) has been studied in mice. Naturally occurring anti-phosphocholine Abs and C-reactive protein (CRP)3 can protect mice from lethal S. pneumoniae infection (7, 8). This protective effect of anti-phosphocholine Abs and CRP requires an intact complement system (9, 10). The complement system consists of some 30 fluid-phase and cell-membrane proteins and can be activated essentially by three distinct routes, the classical, alternative, and mannose-binding lectin pathways (for review, see Refs. 11, 12, 13, 14). In general, the classical pathway is activated by binding of the C1 complex (which consists of C1q and two molecules of both C1r and C1s) to Abs (or CRP) ligated to an Ag on the bacterial surface whereas the alternative and the lectin pathway can be activated directly by bacterial cell surface components. All three pathways converge at the point of the C protein C3. In mice, administration of cobra venom factor (CVF), a convertase analog that depletes C3, leads to impaired opsonization and clearance of S. pneumoniae (15). Furthermore, C3-deficient mice also have impaired clearance of S. pneumoniae from the blood (16). By using mice with genetic deficiencies of specific complement components, Brown et al. (10) provided evidence that the classical pathway is the dominant pathway for activation of the complement system during innate immunity to S. pneumoniae, loss of which results in severe septicemia and the rapid development of fatal disease.
Within the CSF, essential C proteins like C3 are almost totally absent under physiological conditions (17, 18, 19). During the course of bacterial meningitis, C protein concentrations can increase in the CSF, but remain below blood levels (18, 20). Nonetheless, the rise in C protein seems to be essential for limiting pneumococcal outgrowth within the CSF. In a rabbit meningitis model, complement depletion by CVF before intracisternal challenge with pneumococci was associated with enhanced bacterial outgrowth within the subarachnoid space and a >100-fold decrease in LD50 (21). In patients with bacterial meningitis, a favorable outcome was correlated with higher levels of the C proteins C4 and C3 (22).
Beside its crucial role in the killing of invading microorganisms, C proteins or protein fragments that are generated during complement activation are potential regulators of the host immune response (12, 23). In pneumococcal meningitis, the inflammatory host response has been indicated to cause damage to the brain, thus accounting for the outcome of meningitis (for review, Refs. 24 and 25). Current data on the role of the complement system in the innate immune response to S. pneumoniae infection are limited, but also conflicting. Although Ernst et al. (26) reported that C5-derived chemotactic activity accounts for neutrophil accumulation observed in a rabbit model of pneumococcal meningitis, Tuomanen et al. (21) demonstrated that complement depletion did not affect the magnitude of CSF leukocytosis in the same model.
To characterize the role of the complement system in pneumococcal meningitis, we used gene-targeted mice either lacking functional C3, affecting all complement activation pathways, or C1q, affecting only the classical pathway. Thereby, we investigated 1) pneumococcal outgrowth within the CNS and blood, 2) the inflammatory host response to S. pneumoniae infection within the CNS, 3) the development of CNS complications, and 4) short-term outcome of the disease.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
A well-characterized mouse model of pneumococcal meningitis was used in this study (19, 27). Briefly, meningitis was induced by transcutaneous injection of 15 µl of a bacterial suspension containing 107 CFU/ml S. pneumoniae type 3 into the cisterna magna under short-term anesthesia with halothane. Mice were weighed, put into cages, and allowed to awaken. Twenty-four hours after infection, mice were evaluated clinically as described previously (19). Thereafter, the body temperature was measured via a rectal probe, mice were reweighed, and anesthetized with ketamine/xylazine. Subsequently, a catheter was inserted into the cisterna magna to measure intracranial pressure (ICP) and to determine CSF leukocyte counts. Then, blood samples were taken by transcardial puncture. After deep anesthesia with ketamine/xylazine, mice were perfused transcardially with 15 ml of ice-cold PBS containing 10 U/ml heparin. The brains were removed and rapidly frozen.
Determination of the blood-brain barrier (BBB) integrity
To assess BBB integrity, mouse brain homogenates were examined for infiltration by albumin, an abundant serum protein that is normally excluded from the brain by the intact BBB, using ELISA as described previously (19).
Determination of bacterial titers in blood and organs
Cerebella were dissected and homogenized in sterile saline. Blood samples and cerebellar homogenates were diluted serially in sterile saline, plated on blood agar plates, and cultured for 24 h at 37°C with 5% CO2.
mRNA isolation and RT-PCR analysis
Total RNA was extracted from frozen brain sections with TRIzol reagent (Invitrogen Life Technologies) and reverse transcribed using Superscript II (Invitrogen Life Technologies). The cDNA was amplified by PCR with gene-specific primers (MWG Biotech) of the following sequences: C3 sense, 5'-CACCGCCAAGAATCGCTAC-3'; C3 antisense, 5'-GATC AGGTGTTTCAGCCGC-3'; C1q sense, 5'-TCTCAGCCATTCGGCAGAAC-3'; C1q antisense, 5'-TAACACCTGGAAGAGCCCCTT-3'; C1ra sense, 5'-CTTCCGCTACATCACCAC-3'; C1ra antisense, 5'-GCTAACTTATCTTCTGTTA-3'; complement receptor-related protein y (Crry) sense, 5'-CATCACAGCTTCCTTCTGCC-3'; Crry antisense; 5'-ATCGTTGCTGGTACAGTATA-3'; C3aR sense, 5'-GCTGTGGTCACTGTCTTTTTTATCTG-3'; C3aR antisense, 5'-AAGGAACTTTCTGGATCAGTAATCAATAG-3'. Mouse 
-actin was coamplified as an internal control using the following primer sequences: sense 5'-GGACTCCTATGTGGGTGACGAGG-3' and antisense 5'-GGGAGAGCATAGCCCTCGTAGAT-3'. Linearity of DNA amplification was determined for each primer set in experiments establishing the PCR procedures in terms of cDNA amounts and cycle number applied. PCR products were separated on a 1.5% agarose gel, stained with ethidium bromide, visualized by UV illumination, and photographed. Densitometry was performed on the negative image and the relative absorbances of C3, C1q, C1ra, Crry, and C3aR were normalized by relation to absorbance of -actin RT-PCR products.
Gene array analysis
The relative mRNA expression of cytokines, chemokines, and some related inflammatory genes was analyzed with a pathway-specific oligonucleotide microarray. The mouse inflammatory cytokines and receptor Oligo GEArray (OMM-011; Superarray) contain 112 inflammatory cytokine and receptor genes and different housekeeping genes. Detailed information about this oligonucleotide array including description of gene probes, experimental protocol, and data analysis method can be obtained at the supplier’s website (www.superarray.com). In brief, total RNA was extracted from frozen brain sections containing the lateral ventricles and hippocampal tissue with TRIzol reagent (Invitrogen Life Technologies) and quantified by densitometry. Then, equal amounts of total RNA were pooled from eight mice in each experimental group and analyzed twice (except for the C3-deficient control group which had four mice and was analyzed once). For analysis, 3 µg of total RNA was reverse transcribed. Then, biotin-labeled cRNA was synthesized from cDNA with the use of the TrueLabeling-AMP Linear RNA Amplification kit (Superarray). cRNA was purified using the ArrayGrade cRNA Cleanup kit (Superarray), quantified, and hybridized overnight to inflammatory cytokine and receptor gene-specific probes that were spotted on the GEArray membranes (9 µg of cRNA/membrane). After incubation with streptavidin-AP conjugate, the array image was developed with CPD-Star chemiluminescent substrate (chemiluminescent detection kit; Superarray) and recorded with x-ray film. The image was scanned into raw data and analyzed by TINA 2.08e software (Raytest). The signal from the expression of each gene on the array was normalized to the signal derived from the internal GADPH standard on the same membrane and expressed as percentage expression of GADPH. For the sake of specificity, the sensitivity of the test was limited by ignoring relative expression levels <5% of positive controls. Gene probes were excluded from analysis if their relative expression levels differed by >2-fold within an experimental group. This left 98 gene probes for analysis. Genes with >2-fold differences in their mean expression level between experimental groups were considered as differently expressed.
Protein array analysis
Mice brains were screened for 62 cytokines/chemokines using a mouse-specific Cytokine Ab Array (Array 3.1; Ray Biotech). Detailed information about this cytokine Ab array including Ab list, sensitivity data, and experimental protocol can be obtained at the supplier’s website (www. raybiotech.com). Briefly, 30-µm-thick cryosections containing the lateral ventricles and hippocampal tissue were homogenized in lysis buffer (10 mM HEPES (pH 7.9), 10 mM KCl, 1.5 mM MgCl2, and a mixture of protease inhibitors, including PMSF, aprotinin, leupeptin, and pepstatin A), then centrifuged at 12,000 x g for 15 min at 4°C. Protein concentrations were determined in supernatants using the Nanoquant assay (Carl Roth) and equal protein amounts of eight separate brain extracts per group were pooled (except for the C3-deficient control group which had four mice). A total of 2000 µg of pooled protein was applied per array membrane and handled according to the manufacturer’s instructions. Cytokine-antibody complexes on membranes were detected by ECL and recorded with x-ray films. The images were scanned into raw data and analyzed by TINA 2.08e software (Raytest). To normalize the results, the ODs of each spot were then expressed as percentage of the average optical densities of the six positive controls contained on each membrane. For the sake of specificity, the sensitivity of the test was limited by ignoring relative expression levels <5% of positive controls. Proteins with >2-fold differences in their expression level between experimental groups were considered as differently expressed.
In addition, murine serum samples were profiled using the mouse-specific cytokine Ab array. Briefly, protein concentrations were measured by Nanoquant assay and equal protein amounts of eight serum samples per group were pooled. A total of 2000 µg of pooled protein was applied per array membrane and handled according to the manufacturer’s instructions.
Immunoassays for IL-1, IL-12, MIP-1
, and L-selectin
Immunoreactive IL-1, IL-12, MIP-1 (murine CCL9/10), and L-selectin were determined using commercially available ELISA kits (Quantikine Assay kits; R&D Systems). Briefly, frozen brain sections were homogenized in lysis buffer, then centrifuged at 12,000 rpm for 15 min at 4°C, and 50 µl of the supernatant was used for each determination. Additionally, the protein concentration of the supernatant was measured using the Nanoquant assay (Carl Roth). Immunoreactive IL-1, IL-12, MIP-1, and L-selectin concentrations were expressed as picograms per milligram of brain protein.
Experimental groups in the mouse model
The following experimental groups were investigated: 1) wild-type (wt) mice injected intracisternally with 15 µl of PBS (n = 9); 2) wt mice injected intracisternally with S. pneumoniae (n = 13); 3) C1q-deficient mice injected intracisternally with S. pneumoniae (n = 14); 4) C3-deficient mice injected intracisternally with S. pneumoniae (n = 13). To reconstitute complement-mediated opsonic activity in the CSF of C3-deficient mice, additional mice of this strain (n = 4) were injected with 25 µl of serum obtained from a healthy wt mice 1 h before infection; C3-deficient mice (n = 4) injected with 25 µl of serum obtained from a healthy C3-deficient mouse served as controls. Additionally, brain samples obtained from uninfected C3-deficient controls (n = 4) were used for gene and protein array analyses. Gene-targeted mice lacking expression of C1q were generated by Botto et al. (28), as described previously, and then backcrossed for 10 generations onto a C57BL/6 background. The C3-deficient mouse strain was originally produced by Wessels et al. (29), obtained from The Jackson Laboratory (strain B6.129S4-C3tm1Crr/J), and also bred back against the C57BL/6 background for 10 generations. wt C57BL/6 mice were purchased from Charles River Laboratories. Age- and sex-matched groups of wt, C1q-deficient, and C3-deficient mice were used for the experiments. All the experiments were approved by the Government of Upper Bavaria.
Statistical analysis
The principal statistical test used for comparison of infected wt mice with wt controls was the unpaired Student t test. The principal statistical test used for comparisons between infected wt, infected C1q-deficient, and infected C3-deficient mice was one-way ANOVA and Scheffe’s test. Differences were considered significant at p < 0.05. Data are expressed as mean ± SD.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
In brain homogenates from PBS-injected control mice, constitutive expression of some factors of the complement system like C1q, Crry (data not shown), and C3aR were detected by RT-PCR, while other complement components like C1r and C3 were absent from healthy brains. Pneumococcal infection of the CNS caused an up-regulation or induction of all investigated components of the complement system, as observed in brains obtained from wt mice 24 h after challenge (Fig. 1).
|
To test whether the complement system limits pneumococcal outgrowth within the immunoprivileged CNS, we used mice deficient in C3, the central complement component, or C1q, an integral part of the C1 complex that triggers the activation of the classical pathway.
In cerebellar homogenates, the number of S. pneumoniae was almost 20-fold higher in C3-deficient mice than in wt mice. In C1q-deficient mice, cerebellar titers were increased 12-fold compared with wt mice, but were only about half as high as those seen in C3-deficient mice (Fig. 2a). Intrathecal reconstitution with complement-sufficient wt serum (but not C3-deficient serum) in C3-deficient mice completely restored the, albeit limited, ability of mice to combat pneumococcal infection of the CSF.
|
Effect of C1q and C3 deficiency on the inflammatory host response
Because components of the complement system can modulate inflammatory responses, we further characterized the host immune response within mouse brains. Determinations of CSF leukocyte counts revealed a marked accumulation of leukocytes within the subarachnoid space in complement-sufficient wt mice following intracisternal pneumococcal infection. Both, infected C1q-deficient and infected C3-deficient mice had significantly lower CSF leukocyte counts than infected wt mice (Fig. 2c). Intrathecal substitution of complement-sufficient wt serum (but not C3-deficient serum) in C3-deficient mice led to reversal of CSF leukocyte counts close to the levels of infected wt mice.
The altered recruitment of inflammatory cells observed in the CSF of complement-deficient mice prompted us to analyze the expression of cytokines associated with inflammatory responses. First, we used a pathway-specific, mouse oligonucleotide microarray that allows simultaneous screening for 112 inflammatory cytokines and receptors. Fourteen gene probes were discarded from further analysis due to >2-fold differences in expression levels within an experimental group. Among the remaining gene probes, no differences were found in the expression levels between wt and C3-deficient control brains. Pneumococcal infection was associated with an induction or up-regulation of 15 inflammatory cytokines and receptors. Among them were the proinflammatory cytokines and receptors IL-1, IL-6, CXCL1, CXCL2, CXCL9, CXCL10, CCL2, and C3 (Fig. 3). The expression of all these meningitis-induced inflammatory mediators was blunted in brains from infected C3-deficient mice.
|
, IL-6, and IL-12, the CXC chemokines KC (murine CXCL1), MIP-2 (murine CXCL2), and CXCL16, the CC chemokines CCL2, MIP-1, CCL12, and the hemopoietin G-CSF (Fig. 4). With the exceptions of KC, CCL12, and G-CSF, the expression level of all these cytokines was >2-fold lower in brains of infected C3-deficient mice than infected wt mice. The differences in cytokine expression between infected C1q-deficient mice and infected wt mice, however, were all below the 2-fold difference level. Moreover, the cytokine protein expression pattern did not differ between wt and C3-deficient control brains.
|
, a key proinflammatory cytokine in pneumococcal meningitis, as well as for IL-12p70, MIP-1, and L-selectin, three immune factors whose expression pattern in murine pneumococcal meningitis was, to our knowledge, still undefined. Consistent with the array data, the meningitis-induced increase in IL-1, IL-12, MIP-1, and L-selectin was abrogated in infected C3-deficient mice, but not in C1q-deficient mice (Fig. 5).
|
, CXCL9, CXCL13, and CCL12 were >2-fold higher in blood of infected C3-deficient mice than wt mice. Infected C1q-deficient mice showed a substantial increase in blood levels of IL-6, IFN-, and CXCL13; the expression level of the other cytokines was similar to that in infected wt mice. Thus, the cytokine expression profile in the blood contrasts substantially to that observed in the brain.
|
Because CNS complications are mainly due to the strong host immune response (for review, Refs. 24 and 25), we investigated the impact of both complement deficiencies on two important CNS complications, namely BBB breakdown and rise in ICP.
Pneumococcal infection of the CNS induced a substantial increase in ICP within 24 h. Mice with a targeted disruption of the C3 gene (but not the C1qa gene) had significantly lower ICP values than infected wt mice (Fig. 7a). Because vasogenic edema is the predominant cause of the meningitis-induced increase in ICP, we investigated the integrity of the BBB by measuring brain albumin concentrations. In complement-sufficient, PBS-injected control mice, only small amounts of albumin were detectable in the brain. Intracisternal injection of pneumococci caused a marked extravasation of albumin into the brain (Fig. 7b). In brains of infected C3-deficient (but not C1q-deficient) mice, however, the immunoreactivity for albumin was reduced nearly to levels observed in PBS-injected control mice.
|
Twenty-four hours after pneumococcal infection, all wt mice had meningitis, as evidenced by positive bacterial cultures of the cerebellum and CSF pleocytosis, and clinically, all mice were lethargic and hypothermic, lost weight, and showed impaired motor activity and function (Table I). None of the wt mice died within 24 h after pneumococcal infection. In C3-deficient mice, however, meningitis was associated with substantial mortality. Similarly, C1q-deficient mice had an increased mortality due to pneumococcal meningitis (Table I). Compared with infected wt mice, C3- and C1q-deficient mice also showed a more pronounced reduction in body temperature. In addition, abnormal breathing (notably gasping and audible crackles) was observed more frequently in infected C3- and C1q-deficient mice than in wt mice (Table I).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
30–35% (34, 35). In experimental infection models, intracranial inoculation of 104 CFU/ml S. pneumoniae type 3 in C57BL/6 mice resulted in the death of all infected mice, whereas only one-quarter of mice died after intratracheal injection of an identical dose of the same strain (36, 37). The bleak prognosis is closely related to the immune milieu of the CNS. The number of macrophages and dendritic cells is low in the subarachnoid space, the infection site in meningitis. Pattern recognition receptors and many other immune factors are constitutively expressed at low (or even undetectable) levels. Additionally, the blood-brain/CSF barrier secludes the subarachnoid space from circulating blood opsonins (for review, see Refs. 38 and 39). This immune compromise is an adaptation that restricts inflammatory processes which might impair the function of vital tissues like the CNS (39). However, many neuropathologic changes result in the induction of an immune response in the CNS (38). The CNS-typical immune reactivity is reflected in the expression of complement components. Within the CNS, most complement components are virtually absent under physiological conditions (40) (our data). Here, we demonstrated an induction/up-regulation of the CNS expression of C1q, C1r, C3, and C3aR upon pneumococcal infection. Our data are compatible with previous studies that described enhanced C3 and factor B mRNA expression in brains of mice with Listeria monocytogenes meningoencephalitis (41) and increased CSF concentrations of C3, C4, and factor B in patients with bacterial meningitis (18, 22). Using mice deficient in the complement components C1q or C3, we also investigated the functional role of complement expression in murine pneumococcal meningitis. The major findings of these investigations were that 1) the complement system limits pneumococcal outgrowth in the CNS, although it is unable to eradicate the pathogen; 2) the complement system is a central regulator of the innate immune response to pneumococcal CNS infection; and 3) the suppression of the inflammatory reaction in the CNS in C3-deficient mice is associated with reduced CNS complications, but surprisingly results in decreased survival, presumably due to worse systemic complications.
Several years ago, Tuomanen et al. (21) provided the first evidence for a functional role of the complement system in limiting pneumococcal outgrowth within the CNS. In rabbits depleted of C3 by CVF, intracisternal inoculation of S. pneumoniae resulted in higher bacterial titers than in complement-sufficient control animals (21). In agreement with these data, we found higher numbers of S. pneumoniae within the CNS in C3-deficient mice than in wt mice. In addition, we showed that 1) C1q deficiency led to higher bacterial titers in the CNS, although it does not reach the levels seen in C3-deficient mice; and 2) intrathecal serum substitution restores bacterial killing in C3-deficient mice to wt levels, although it does not enable the host to impede pneumococcal outgrowth. First, our data are consistent with recent observations in other infection models, in which S. pneumoniae was inoculated either by the intranasal or i.p. route, thus causing pneumonia and septicemia. In these models, C1q-deficient mice (the classical pathway is affected), factor B-deficient mice (the alternative pathway is affected), and C3-deficient mice (all three pathways are affected) had an increased susceptibility to S. pneumoniae infection (10). Additionally, C1q-deficient and factor B-deficient mice were reported to be less susceptible to pneumococcal infection than C3-deficient mice (10). Combined, these findings demonstrate that the classical and the alternative pathways are both required for efficient opsonophagocytosis of S. pneumoniae (in nonimmunized hosts). The latter pathway amplifies pneumococcal opsonization by C3, which the former pathway initiated (42, 43). Second, the results of our reconstitution experiments, when combined with the finding that pneumococcal infection induces complement expression in the CNS, show that the inability of the host immune response to overcome S. pneumoniae infection within the CSF (44) does not seem to be due to the lack of C proteins. At least three other mechanisms can be proposed for mediating the "local immunodeficiency" within the subarachnoid space: 1) the scarcity of cellular components of the innate immunity, 2) the absence of pattern recognition receptors, and 3) the active suppression of host defense mechanisms (for review, see Refs. 38 , 39 , and 45). A potential immunosuppressive factor is the anti-inflammatory cytokine TGF, which is present in the CSF (46) at concentrations that can deactivate innate immune cells (47) and is further increased during bacterial meningitis (48). Accordingly, we recently found that mice lacking the TGFRII on neutrophils and macrophages, thus making them insensitive to active TGF, showed a 140-fold reduction in cerebellar bacterial titers compared with that of their wt littermates which was paralleled by a 2- to 3-fold increase in CSF pleocytosis (49).
All in all, our data suggest that the absence of complement components in the CSF during early infection allows invading S. pneumoniae to multiply nearly as efficiently as it can in vitro, reaching high titers up to 109 CFU/ml (44). During the course of infection, when complement factors are either expressed inside the CNS or gain access to the CSF through the disrupted BBB, other factors like the presence of immunosuppressive cytokines seem to be responsible for the insufficient host defense against pneumococcal infection of the CNS.
In this study, we also demonstrated that the complement system is a key regulator of the innate immune response to S. pneumoniae infection of the CNS. Genetic depletion of C3 led to an attenuated influx of leukocytes into the CSF. Because the influx of leukocytes requires the antecedent local expression of cytokines and chemokines, we profiled their brain expression by gene and protein arrays. Gene array analysis identified 15 cytokine mRNAs up-regulated in infected wt brains, but not C3-deficient brains. They included the cytokines IL-1 and IL-6 as well as the chemokines CXCL1, CXCL2, and CCL2, all of which have been implicated recently as key immune factors in pneumococcal meningitis (for review, see Refs. 25 and 50). Comparable results were obtained when mouse brain protein extracts were analyzed by ELISA and/or protein array. The expression levels for IL-1, IL-6, IL-12, MIP-2 (CXCL2), CXL16, CCL2, MIP-1, and L-selectin were significantly lower in infected C3-deficient mice than in infected wt mice. However, similar amounts of G-CSF and CCL12 were found in brain homogenates obtained from both mouse strains. Possible explanations for the elevated G-CSF and CCL12 levels in brains from infected C3-deficient mice may be their influx from blood to brain, longer CNS residence times than other cytokines, or a combination of both. In infected C3-deficient mice, we observed a nearly unaltered BBB permeability for albumin, suggesting that the BBB was maintained. Therefore, to exhibit increased brain levels, bloodborne cytokines must be able to pass the intact BBB. For several cytokines, including IL-6 and IFN-, saturable transport systems were found in normal brain (51). Chemokines like CXCL1, CCL3, and CCL4 were also reported to cross the intact BBB, either by passive diffusion or by reversal trapping in the cerebral vasculature (52, 53). Our protein array analyses revealed increased levels of G-CSF and CCL-12 in serum and brain from C3-deficient mice, whereas IL-6, IFN-, or CCL-2 were nearly absent in the brain, but highly up-regulated in serum. Although data on the CNS residence time of individual cytokines are very limited, there is evidence that IL-6 is rapidly enzymatically degraded in the brain (54), whereas G-CSF remains intact for many hours (55). Thus, we speculate that the increased G-CSF (and perhaps also CCL12) levels observed in brains of infected C3-deficient mice are due to a clear influx from the blood and a sustained CNS residence time.
All in all, the results of our gene and protein expression analysis revealed a profound suppression of cytokine and chemokine expression in mice lacking the central complement component C3. These data fit to the markedly attenuated CSF pleocytosis observed in C3-deficient mice, which is in agreement with recent studies that demonstrated that 1) the recombinant complement fragments C5a and C5b-9, when injected into the CSF, can induce infiltration of neutrophils into the subarachnoid space (56, 57) and 2) Abs against human C5a markedly attenuated the chemotactic activity of the CSF in experimental pneumococcal meningitis (26). However, our findings contrast with a previous study by Tuomanen et al. (21) who showed that complement depletion by CVF resulted in a slight delay in CSF pleocytosis but had no effect on the magnitude of leukocyte response. The absence of an effect of CVF on CSF leukocyte numbers may be due to a loss of effect of CVF treatment as a result of both, clearance of CVF from the body (58) and up-regulation of complement synthesis within the brain. Possibly, the results by Tuomanen et al. (21) may also be influenced by yet undefined, complement-independent effects of CVF as previously noted by Libert et al. (59).
According to the concept that, in pneumococcal meningitis, the development of CNS complications occurs as a consequence of the intense host immune response (25), we found that the reduced immune response in infected C3-deficient mice was associated with attenuated CNS complications. Nevertheless, C3 deficiency (and also C1q deficiency) was accompanied with a worsening of short-term outcome; more than one-third of infected C3-deficient (and also C1q-deficient) mice died within 24 h after intracisternal pneumococcal challenge, whereas none of the infected wt mice died within this observation period. Because C3-deficient (and also C1q-deficient) mice displayed significantly higher bacterial numbers in the blood stream, it is conceivable that the more severe bacteremia contributes to the more adverse outcome observed in these mice. This suggestion is supported by the following observations: 1) hypothermia, an indicator of severe septicemia in mice, was more pronounced in C3-deficient (and also C1q-deficient) mice than in wt mice; 2) C3-deficient (and also C1q-deficient) mice displayed more frequent breathing problems than wt mice; 3) C3-deficient (and also C1q-deficient mice) had higher blood levels of IFN-, IL-6, CCL12, and CXCL13 than wt mice (this is indicative of a more pronounced systemic immune response); and 4) intracranial complications, the most important determinants for an unfavorable clinical outcome, were reduced in C3-deficient mice (and also, albeit to a lesser degree, in C1q-deficient mice), compared with those observed in infected wt mice. Moreover, we recently observed that MyD88-deficient mice had a similar phenotype to that of C3-deficient mice. Infected MyD88-deficient mice displayed an attenuated immune response and less pronounced intracranial complications, but had a higher mortality rate that was associated with more severe bacteremia, more pronounced hypothermia, more frequent breathing problems, and more pronounced secondary pneumonia (19), when compared with infected wt mice. Because MyD88 deficiency was associated with a substantial suppression of the brain expression of complement factors, it is conceivable that the absence of complement factors is responsible for the clinical phenotype observed in both mouse strains.
In conclusion, the present study demonstrated that 1) the complement system appears to mediate partial killing, but not the clearance, of S. pneumoniae from the CSF; 2) both, the classical and alternative pathways are vital for host defense against S. pneumoniae infection within the CNS; and 3) the complement system plays a critical role in the generation of the host immune response to S. pneumoniae within the CNS. According to this, C3 deficiency led to diminished CNS inflammation, which was paralleled by an attenuation of meningitis-associated CNS complications. Unfortunately, the price to pay for the amelioration of meningitis was a more severe bacteremia, which ultimately resulted in a worsening of this disease.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB576/A5 (to U.K.); SFB576/B8 (to J.H.)). 
2 Address correspondence and reprint requests to Dr. Uwe Koedel, Department of Neurology, Klinikum Grosshadern, Ludwig Maximilians-University, Marchioninistrasse 15, Munich D-81377, Germany. E-mail address: Uwe.Koedel{at}med.uni-muenchen.de
3 Abbreviations used in this paper: CRP, C-reactive protein; CVF, cobra venom factor; ICP, intracranial pressure; BBB, blood-brain barrier; Crry, complement receptor-related protein y; wt, wild type.
Received for publication January 26, 2006. Accepted for publication November 2, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
receptors. J. Immunol. 168: 6375-6381. 
attenuates neutrophil recruitment in the central nervous system during experimental bacterial meningitis. Infect. Immun. 67: 2590-2601. compensates for the impaired host defense of IL-1 type I receptor-deficient mice during pneumococcal pneumonia. J. Immunol. 167: 5240-5246. expression in human leptomeningeal cells and transforming growth factor--like activity in human cerebrospinal fluid. Lab. Invest. 67: 360-368. . Nature 334: 260-262. [Medline]1, tumor necrosis factor-, and interleukin 6 in cerebrospinal fluid: association with clinical outcome for children with bacterial meningitis. Clin. Infect. Dis. 25: 328-329. [Medline]This article has been cited by other articles:
|
|
 |
|
  R. Paul, B. Obermaier, J. Van Ziffle, B. Angele, H.-W. Pfister, C. A. Lowell, and U. Koedel Myeloid Src kinases regulate phagocytosis and oxidative burst in pneumococcal meningitis by activating NADPH oxidase J. Leukoc. Biol., October 1, 2008; 84(4): 1141 - 1150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. R. Quin, Q. C. Moore III, J. A. Thornton, and L. S. McDaniel Peritoneal Challenge Modulates Expression of Pneumococcal Surface Protein C during Bacteremia in Mice Infect. Immun., March 1, 2008; 76(3): 1122 - 1127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Marks, M. Verolin, A. Stensson, and N. Lycke Differential CD28 and Inducible Costimulatory Molecule Signaling Requirements for Protective CD4+ T-Cell-Mediated Immunity against Genital Tract Chlamydia trachomatis Infection Infect. Immun., September 1, 2007; 75(9): 4638 - 4647. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Koedel, U. M. Merbt, C. Schmidt, B. Angele, B. Popp, H. Wagner, H.-W. Pfister, and C. J. Kirschning Acute Brain Injury Triggers MyD88-Dependent, TLR2/4-Independent Inflammatory Responses Am. J. Pathol., July 1, 2007; 171(1): 200 - 213. [Abstract] [Full Text] [PDF]
|
|
| ||||||
for representative examples) or (b) was <5% of the internal GAPDH standard. Fourteen gene probes were discarded from further analysis due to >2-fold differences in expression levels within a group. Among the remaining 98 gene probes, no differences were found in the relative expression levels between wt and C3-deficient control brains. Pneumococcal infection was associated with an induction or up-regulation of 15 inflammatory cytokines and receptors, including the cytokines IL-1
