Colon ascendens stent peritonitis
Myeloid differentiation factor 88
Chemokine receptors are important for recruiting leukocytes to sites of infection and may contribute to immune cell activation. The present study investigated the role of the chemokine receptor CCR2 in polymicrobial septic peritonitis. The results showed that peritoneal production of the CCR2 ligands CCL2 and CCL12 in septic mice was largely independent of the common Toll-like receptor signaling adaptor MyD88. Antibody blockade of CCR2 reduced the recruitment of macrophages and neutrophils to the infected peritoneal cavities of both wild-type and MyD88-deficient mice, suggesting that CCR2 engagement contributes to the MyD88-independent cellular response against polymicrobial septic peritonitis. Notably, administration of blocking CCR2 antibodies markedly increased local and systemic IL-10 levels in septic wild-type mice, whereas IL-10 was not detected in MyD88-deficient mice irrespective of whether CCR2 was blocked or not. Inhibition of CCR2 directly augmented Toll-like receptor-induced IL-10, but not TNF and IL-6, production of macrophages in vitro. Concomitant with enhanced IL-10 production, CCR2 blockade caused impaired bacterial clearance and aggravated kidney injury in wild-type, but not MyD88-null mice. These results indicate that CCR2 engagement modulates the innate immune response to polymicrobial septic peritonitis by both MyD88-dependent and -independent processes and suggest that a major function of CCR2 in sepsis is to attenuate IL-10 production and IL-10-mediated suppression of host defense.
Activation of Toll-like receptor (TLR) signaling by conserved molecular structures of microbial pathogens is crucial for the induction of innate immune responses to infection 1–3. Myeloid differentiation factor 88 (MyD88) represents a signaling adaptor that is shared by the TLR family and mediates inflammatory gene expression via activation of NF-κB and mitogen-activated protein kinases 4. In addition to the common MyD88 adaptor, TLR engage signaling proteins in a receptor-specific fashion. The Toll/IL-1 receptor (TIR)-domain-associated protein (TIRAP) selectively mediates signaling of TLR2 and TLR4 through the MyD88-dependent pathway 5, 6, whereas the adaptor protein TIR-domain-containing adaptor inducing IFN-β (TRIF) promotes MyD88-independent signaling of TLR3 and TLR4 7, 8. TRIF signaling induces gene expression both by type I IFN-dependent and -independent mechanisms 9, 10.
TLR signaling may not only contribute to pathogen defense, but may also lead to immune pathology associated with certain types of infection. Previous work showed that genetic deficiency of MyD88 protected mice from the lethal effects of polymicrobial septic peritonitis and that the hyperinflammatory response to sepsis was strongly attenuated in the absence of MyD88 11. In contrast, recruitment of neutrophils to the infected peritoneal cavity and bacterial clearance were not altered in MyD88-deficient mice, suggesting that the MyD88-dependent signaling pathway contributes to the immune pathology of sepsis but is dispensable for anti-bacterial defense. Further analysis of the protective immune response to septic peritonitis in MyD88-deficient mice revealed that cytokine expression in spleen and production of the chemokine CCL2 in liver and lung occurred normally.
CCL2 is produced by a variety of cell types in response to inflammatory stimuli and chemoattracts and activates monocytes, T cells, NK cells and basophils 12–17. Elevated levels of CCL2 have been detected in plasma of sepsis patients, after administration of LPS to human volunteers, and in animal models of endotoxemia or septic peritonitis 11, 18–22. Moreover, neutralization of CCL2 aggravated the lethal effects of LPS administration or cecal ligation and puncture in experimental animals 18, 19. CCL2 exerts its biological effects by binding to the chemokine receptor CCR2 that also serves as a receptor for the other CCL2 subfamily members 23–26. Analysis of CCR2-deficient mice revealed a crucial role of this chemokine receptor for pathogen defense and the induction of Th1-type immune responses. Thus, genetic ablation of CCR2 caused an increased susceptibility to Listeria monocytogenes, Leishmania major and Cryptococcus neoformans infections and was associated with an impaired production of IFN-γ and/or an increase in IL-4 and IL-5. In addition to the effects of the CCR2 null mutation in infection models, CCR2-deficient mice failed to recruit macrophages in a model of sterile peritoneal inflammation induced by thioglycollate 27–31. Antibody blockade of CCR2 also prevented macrophage accumulation in thioglycollate peritonitis 32, but did not inhibit infiltration of inflamed joints by monocytes during progression of collagen-induced arthritis suggesting that under certain conditions of inflammation recruitment of macrophages may be independent of CCR2 33.
The present study therefore investigated the role of CCR2 in polymicrobial septic peritonitis. We show that chemokine ligands for CCR2 are produced normally in MyD88-null mice and are involved in MyD88-independent phagocyte recruitment to the septic focus. In contrast, CCR2 blockade augmented IL-10 production, impaired bacterial clearance and aggravated kidney damage in wild-type, but not MyD88-deficient mice. Together with in vitro analyses of macrophage IL-10 production these results support the concept that a major function of CCR2 in sepsis is to counteract IL-10 production and immunosuppression.
2.1 Peritoneal expression of CCR2-ligands in septic peritonitis is MyD88 independent
Chemokines orchestrate the migration of leukocytes during inflammation via binding to specific chemokine receptors and therefore are crucial for generating immune responses at the site of an infectious challenge. Because Toll-like receptors are potent inducers of chemokine production during infection, we investigated the levels of various chemokines in peritoneal lavage fluid of both wild-type and MyD88-deficient mice after induction of polymicrobial septic peritonitis using the CASP model. Peritoneal levels of CXCL1 and CXCL2, which both attract neutrophils, were severely diminished in MyD88-deficient mice 12 h after induction of peritonitis (Fig. 1). In contrast, the CCR2 ligands CCL2 and CCL12 were expressed at substantial levels in both experimental groups (Fig. 1). Whereas CCL2 levels exhibited a partial, but significant, reduction in MyD88-deficient mice as compared to wild-type controls, production of CCL12 was similar in both groups. These results on the regulation of local chemokine production are consistent with our previous analyses, showing MyD88-independent expression of CCL2 in peripheral organs such as liver, lung, and spleen 11, which represent secondary sites of infection during septic peritonitis. In addition, we demonstrate that CCL12 is regulated in a MyD88-independent manner.
2.2 Peritoneal local leukocyte accumulation is CCR2 dependent
To analyze the role of CCR2 ligands during the course of polymicrobial septic peritonitis we blocked the CCR2 activity using mAb MC21 32. The accumulation of infiltrating macrophages (Mac-1high Gr-1neg) and neutrophils (Gr-1high Mac-1high) in the peritoneal cavity was examined 12 h after induction of septic peritonitis. Absolute cell numbers of both macrophages and neutrophils were strongly diminished in mice treated with the mAb against CCR2 but not isotype-matched control mAb MC67 (Fig. 2). The inhibitory effect of CCR2 blockade on macrophage and neutrophil recruitment to the infected peritoneal cavity was also demonstrated in MyD88-deficient mice (Fig. 2). Consistent with our previous data 11, the results in Fig. 2 also confirm that cell recruitment to the infected peritoneal cavity was independent of MyD88. Thus, recruitment and accumulation of innate immune cells involved in host defense mechanisms are largely dependent on CCR2. The contribution of CCR2 to the local innate cellular immune response, however, is MyD88 independent.
2.3 IL-10 production during septic peritonitis is modulated by CCR2
Resident peritoneal cells and infiltrating leukocytes release various cytokines and chemokines during septic peritonitis thereby exhibiting a feedback regulatory activity during the local immune response to infection. We therefore examined the effects of CCR2 blockade on cytokine production in the peritoneal cavity. In wild-type mice treated with the control mAb MC67 or injected with PBS, septic peritonitis induced high levels of IL-10 (Fig. 3A), whereas TNF was not detectable (data not shown). Notably, antibody-mediated blockade of CCR2 substantially amplified peritoneal IL-10 levels in these mice. In MyD88-deficient mice, however, we could not observe any induction of IL-10 in response to septic peritonitis and treatment with the CCR2-blocking antibody did not increase IL-10 levels (Fig. 3A). To further investigate whether CCR2 blockade would also affect the systemic production of IL-10, serum cytokine levels were determined. The results in Fig. 3B show that systemic IL-10 production was enhanced by about fourfold in septic wild-type mice treated with the anti-CCR2 as compared to the MC67 antibody or PBS controls. In septic MyD88-deficient mice, however, serum IL-10 levels were similar to those of non-septic mice and were not altered by the CCR2 blockade. These results therefore reveal an important role of CCR2 for regulating IL-10 production during septic peritonitis. Moreover, this regulatory effect of CCR2 is not observed in the absence of MyD88.
To further address the question as to whether engagement of CCR2 may directly modulate IL-10 production, we analyzed the IL-10 levels in supernatants of anti-CCR2 antibody-treated bone marrow-derived macrophages in vitro after stimulation with TLR-ligands. As shown in Fig. 4 macrophages produced high levels of CCL2 upon exposure to LPS, Pam3Cys and CpG-DNA, which are known to engage TLR4, TLR2 or TLR9, respectively. Consistent with the in vivo data obtained in the CASP model of septic peritonitis, IL-10 production of LPS-, Pam3Cys-, and CpG-DNA-stimulated macrophages was significantly increased by treatment with the CCR2 mAb (Fig. 4). Enhancement of IL-10 release by CCR2 blockade was observed for various concentrations of TLR agonists. In contrast, production of TNF, IL-6, and CCL2 by LPS-, Pam3Cys-, and CpG-DNA-stimulated macrophages was not significantly altered by CCR2 blockade. These results therefore suggest that CCR2 directly and specifically down-regulates the TLR-induced expression of IL-10 by innate immune cells.
2.4 CCR2 blockade impairs anti-bacterial defense
Inhibition of CCR2 leads to increased IL-10 levels and reduced peritoneal phagocyte recruitment during septic peritonitis. Because these effects may alter host defense against infection, we investigated the bacterial load in mice treated with anti-CCR2 mAb MC21. The results in Fig. 5A demonstrate that CCR2 blockade significantly increased bacterial numbers in the peritoneal cavities of wild-type mice 12 h after induction of septic peritonitis when compared to control antibody treatment. In contrast, inhibition of CCR2 had no effect on the bacterial counts determined in the peritoneal cavities of MyD88-deficient mice (Fig. 5A). It should also be noted that bacterial numbers of control antibody-treated wild-type and MyD88-deficient mice did not differ significantly confirming our previous findings 11. In a complementary set of experiments, the effects of CCR2 blockade on bacterial expansion in kidneys of septic mice were examined. The results in Fig. 5B demonstrate that bacterial numbers in kidneys were significantly elevated in mice treated with anti-CCR2 mAb as compared to control antibody-treated mice. These results suggest that inhibition of CCR2 impairs anti-bacterial defense in septic peritonitis and that this effect is only observed in the presence of MyD88.
2.5 Differential effects of CCR2 blockade and MyD88 deficiency on acute renal failure
Acute renal failure is a hallmark of severe sepsis induced by the CASP model of polymicrobial peritonitis 34. To investigate whether CCR2 blockade may also influence the development of organ injury during septic peritonitis, we determined serum creatinine levels, which serve as an indicator of acute renal failure. In wild-type mice, septic peritonitis was found to substantially increase serum creatinine levels, which were further elevated by blockade of CCR2 (Fig. 6A). In contrast, we could not detect an increase of serum creatinine in septic as compared to untreated MyD88-deficient mice (Fig. 6A). Furthermore, blockade of CCR2 did not elevate serum creatinine levels in MyD88-deficient mice (Fig. 6A). These results suggest that engagement of CCR2 may attenuate acute renal failure. Moreover, induction of acute renal failure and the adverse effects of CCR2 blockade were found to be MyD88 dependent.
CCL2 and CCL12, which are ligands for the chemokine receptor CCR2, are released at high levels during human and experimental sepsis 18–22. Moreover, previous investigations in a murine model of acute septic peritonitis showed CCL2 production in liver and lung to occur in a MyD88-independent fashion 11. The present study examined the role of CCR2 in sepsis using antibody-mediated receptor blockade. The results demonstrate that anti-CCR2 treatment of mice exhibits marked adverse effects on the innate immune response to polymicrobial septic peritonitis and aggravates sepsis severity. Specifically, CCR2 blockade decreased the recruitment of phagocytes to the infectious focus, impaired local bacterial clearance, increased local and systemic levels of the immunosuppressive cytokine IL-10, and augmented kidney injury. We also demonstrate that these effects were mediated by both MyD88-dependent and -independent mechanisms. Considered together with previous work showing that neutralization of the CCR2 ligand CCL2 had detrimental effects in murine endotoxemia and in a model of acute septic peritonitis 18, 19, these studies provide strong evidence for an important protective role of CCR2 engagement in sepsis.
Analysis of peritoneal chemokine levels of septic mice suggested that production of the CCR2 ligands CCL2 and CCL12 was largely independent of the common TLR signaling adapter protein MyD88. In contrast, local CXCL1 and CXCL2 release was almost completely abolished in the absence of MyD88. These results are consistent with our previous work showing that CCL2 levels in liver and lung of septic mice are not significantly affected by MyD88 deficiency, even though those of CXCL1 and CXCL2 were substantially reduced 11. It, therefore, appears that the release of CCR2 ligands is part of the MyD88-independent innate immune response to acute septic peritonitis. In contrast to CCL2 and CCL12 protein production, however, the biological activity of these chemokines may variably depend on additional stimuli that engage the MyD88 signaling pathway. Thus, antibody blockade of CCR2 inhibited peritoneal phagocyte recruitment to a similar extent in wild-type and MyD88-deficient mice, whereas increased IL-10 levels and the adverse effects of anti-CCR2 treatment on bacterial clearance and kidney injury were only observed in wild-type mice.
Immunosuppression is considered a central pathogenic event in sepsis 35. Clinical investigations showed that high levels of IL-10 are associated with fatal outcome of sepsis 36, 37 and neutralization of IL-10 during established experimental sepsis exerts beneficial effects 38, 39. It was therefore important to note that CCR2 blockade markedly elevated both local and systemic IL-10 levels during septic peritonitis. These findings are consistent with previous work showing that neutralization of CCL2 augmented IL-10 concentrations in serum of LPS-treated mice and in kidneys of mice subjected to cecal ligation and puncture 18, 40. Down-regulation of IL-10 may therefore represent an important function of the CCR2 pathway in sepsis. Our experiments show that peritoneal IL-10 levels were increased in mice treated with the CCR2 antibody event though the total number of peritoneal macrophages, which are a likely source of IL-10, was diminished by two- to threefold. However, experiments with macrophages in vitro have shown that, depending on the concentration of the Toll-like receptor stimulus, CCR2 blockade may increase IL-10 production by as much as four- to tenfold. It is therefore quite conceivable that the enhanced IL-10 release at the single-cell level may be sufficient to compensate the reduction of total cell number. In addition, high systemic IL-10 levels in CCR2 antibody-treated mice may also contribute to IL-10 levels at local sites like peritoneal cavity. Previous work has shown that systemic IL-10 levels in the CASP model of septic peritonitis are mostly derived from hepatic Kupffer cells 41, suggesting that this major population of tissue macrophages may also be involved in the effects of CCR2 blockade.
In contrast to IL-10 production in response to septic peritonitis in vivo or Toll-like receptor engagement in vitro, lymph node cells from CCL2-deficient mice immunized with ovalbumin were impaired in their capacity to produce IL-10 following in vitro restimulation with antigen 42. It should be considered, however, that under these experimental conditions IL-10 was produced by T lymphocytes, whereas in LPS-treated mice and mice subjected to septic peritonits IL-10 is predominantly released by macrophages. Moreover, we have shown that CCR2 blockade increased TLR-stimulated IL-10 production of macrophages in vitro. It is therefore conceivable that the effects of CCR2 engagement on cytokine production are cell-type dependent.
CCR2 engagement may influence IL-10 levels in vivo by altering the absolute numbers and relative proportions of inflammatory cell subsets at sites of infection due to chemotactic activity or by directly modulating IL-10 production of immune cells that are exposed to microbes and their products. Our results show that inhibition of CCR2 in vitro augmented IL-10 production of macrophages that were stimulated with Toll-like receptor agonists thereby favoring the latter possibility. Furthermore, our data that anti-CCR2 treatment enhanced IL-10 levels in wild-type but not MyD88-deficient mice are also consistent with the notion that CCR2 down-regulates Toll-like receptor-induced IL-10 production. In accordance, investigations using human monocytes have shown that exogenous CCL2 and C5a can inhibit IL-12 production, but only CCL2 exerted its effects in an IL-10-dependent fashion, suggesting that CCL2 enhances IL-10 activity 43. Considered together, these results therefore provide evidence that CCR2 represents an important regulator of IL-10 release during sepsis.
CCR2 blockade inhibited the recruitment of macrophage-like cells (Mac-1+Gr-1-) as well as neutrophils (Gr-1hiMac-1hi) to the infected peritoneal cavity. Whereas CCR2 is known to be a potent mediator of mononuclear phagocyte trafficking to inflamed tissues, CCR2 is not expressed on neutrophils and CCR2 ligands are not chemotactic for neutrophils 27–29, 31, 32, 44. Using mouse neutrophils we could confirm the absence of CCR2 expression and the lack of CCL2 chemotactic activity on these cells (data not shown). However, CCR2 may influence neutrophil recruitment by indirect mechanisms, because endogenous CCL2 was previously shown to attract neutrophils via stimulation of LTB4 production by macrophages 19. It is therefore conceivable that the reduced neutrophil recruitment in anti-CCR2-treated mice resulted from the diminished CCR2-dependent production of LTB4 or other as yet unidentified mediators, while the effects of the anti-CCR2 antibody on macrophage attraction are most likely related to a direct inhibition of chemotactic activity.
Cell recruitment in MyD88-deficient mice was not impaired although peritoneal levels of CXCL1, CXCL2 and, to a lower extent, CCL2 were significantly decreased. These findings are in accordance with our previous work 11. It should be noted, however, that despite this reduction in chemokine production substantial amounts of CCL2 and CCL12 and residual amounts of CXCL1 and CXCL2 were present in the peritoneal cavities of MyD88-null mice. It is therefore conceivable that the amounts of chemokines produced in MyD88 deficient mice are sufficient for cell recruitment. In fact, it is well known from in vitro chemotaxis experiments that chemokines exhibit a bell-shaped dose-response curve with very high amounts of chemokines resulting even in reduced cell migration. In addition, it is possible that distinct classes of chemoattractants such as complement products or lipid mediators contribute to leukocyte recruitment in septic peritonitis. Consistent with this possibility, we have shown in additional experiments that the combined blockade of CCR2 as well as CXCL1 and CXCL2 is more efficient in blocking peritoneal cell recruitment, but even under these conditions blockade was not complete (C.F. and B.H., unpublished data).
However, chemoattractants are not the only mechanism determining the extent of cell accumulation at sites of infection. Thus, even in case of reduced chemotactic activity in peritoneal cavities of MyD88-deficient mice attenuation of leukocyte apoptosis might compensate for diminished cell recruitment. Accordingly, TLR- and MyD88-mediated signals are well known to trigger apoptosis of leukocytes and other cell types 45–50. Moreover, our previous work 51 has shown that attenuation of TLR responsiveness through induction of in vivo LPS tolerance results in diminished neutrophil apoptosis and enhanced peritoneal neutrophil accumulation during septic peritonitis. Finally, endothelial TLR were shown to influence neutrophil sequestration 52, suggesting that also altered functions of MyD88-deficient endothelial cells during leukocyte adhesion and transmigration might play a role.
The results of the present study also revealed the CCR2 blockade markedly impaired bacterial clearance in septic peritonitis. In accordance with a protective role of CCR2 in host defense, CCR2-deficient mice were found to be more susceptible to infection with various pathogens including Listeria monocytogenes, Leishmania major, and Cryptococcus neoformans28, 30, 31. Furthermore, neutralization of CCL2 enhanced bacterial expansion in the peritoneal cavity in a model of acute septic peritonitis 19. A possible explanation for these findings may be provided by the critical role of the CCR2 pathway for the recruitment of phagocytic cells. Accordingly, phagocytes have previously been shown to be crucial for the immune defense against a septic challenge both in humans and experimental animals 19, 53–55. Our results demonstrate, however, that the effects of CCR2 blockade on peritoneal immune cell recruitment were MyD88 independent, whereas impairment of bacterial clearance and aggravation of kidney injury were both MyD88 dependent. In contrast, CCR2 blockade elevated local and systemic IL-10 levels only in the presence of MyD88. These findings suggest that the effects of CCR2 engagement on the up-regulation of IL-10 production rather than on the inhibition of cell recruitment may have a major and predominant influence on anti-bacterial defense. Consistent with this notion, increased susceptibility of CCR2-null mice to infection was also associated with decreased Th1 responses 28, 30, 31. It should be noted, however, that these observations due not exclude a role for cell recruitment, because it is possible that the numbers of phagocytes attracted in the absence of active CCR2 may be sufficient for normal bacterial clearance unless they are deactivated by IL-10.
Considered together, the results of the present report suggest that CCR2 not only contributes to the immune response against a septic challenge by attracting innate effector cells, but that a major function of CCR2 in sepsis is to attenuate IL-10 production and IL-10-mediated suppression of host defense.
4 Materials and methods
4.1 Mouse strains and colon ascendens stent peritonitis (CASP) model of polymicrobial peritonitis
MyD88-deficient mice backcrossed at least eight times to the C57BL/6 background were kindly provided by Dr. S. Akira (Osaka, Japan). Control C57BL/6 mice were purchased from Harlan Winkelmann (Borchem, Germany). Mice at 8–12 weeks of age were used for all experiments. The CASP procedure used for induction of septic peritonitis was described in detail previously 56. Rat anti-mCCR2 mAb (MC21) and isotype-matched control mAb (MC67) were administered at 300 µg per mouse and were injected i.v. immediately after surgery. mAb MC21 does not trigger internalization of CCR2, it is not chemotactic, and does not induce calcium flux (32 and data not shown). Thus, mAb MC21 lacks agonistic activity for CCR2. Injection of MC21 antibody also does not lead to the depletion of circulating monocytes and neutrophils or hepatic Kupffer cells (data not shown).
4.2 Bacterial counts and peritoneal neutrophil accumulation
Peritoneal lavage fluid and kidneys were collected from mice sacrificed before (0 h) or 12 h after CASP. Serial dilutions of lavage fluid or kidney homogenates were plated on blood agar plates. CFU were counted after incubation at 37°C for 24 h and calculated per whole cavity. In addition, peritoneal lavage cells were counted and differentiated by staining with Ab against Mac-1 (M1/70) and Ly-6/Gr-1 (RB6–8C5) using appropriate isotype matched controls (all antibodies BD PharMingen, San Diego, CA).
4.3 Local and systemic cytokine and chemokine production and serum creatinine determination
Peripheral organs were snap-frozen in liquid nitrogen and homogenized after thawing in 1 ml PBS containing complete protease inhibitors (Roche Diagnostics, Mannheim, Germany). Insoluble material was removed from organ extracts by centrifugation (6,000×g for 20 min at 4°C). Immune mediator concentrations in organ extracts, serum peritoneal lavage fluid, and culture supernatants were measured by EILSA specific for CXCL1, CXCL2, CCL2, CCL12, or IL-10 (all from R&D Systems, Minneapolis, MN). Serum creatinine levels were measured by standardized protocols at the Institute of Clinical Chemistry, Technische Universität München, Munich, Germany.
4.4 Bone marrow-derived macrophages
For differentiation of bone marrow-derived macrophages, 6–10-week-old mice were used. Femurs of mice were flushed with PBS, and erythrocytes were lysed by treatment with ammonium chloride. Bone marrow cells were plated in suspension-culture petri dishes (Greiner, Frickenhausen, Germany) and cultured in RPMI 1640 supplemented with 10% fetal calf serum and 6.25 µg/ml recombinant mouse M-CSF (R&D Systems, Minneapolis, MN). Cultures were supplied with fresh medium containing M-CSF every 3–4 days. Cells were harvested and used at day 10. The purity of the macrophage cell population was assessed by FACS-Analysis (FACSCalibur flow cytometer, CellQuest software, BD Biosciences, San Diego, CA) using Mac-1 (CD11b; BD PharMingen) and F4/80 antibodies (Serotec, Düsseldorf, Germany) and was >80% in all experiments. Macrophages (1×106) were stimulated with ultra pure LPS from Salmonella minnessota R595 (Quadratech, Surrey, UK) or Pam3Cys (EMC Microcollections, Tübingen, Germany) at different concentrations in the presence of anti-CCR2 mAb or control mAb (5 µg/ml each). Supernatants were collected after stimulation for 20 h and IL-10 was determined by ELISA.
4.5 Statistical analysis
Statistical analysis of the data was performed using the Mann-Whitney U-test or the Student's t-test where appropriate. All data are presented as mean ± SEM. The level of significance was p<0.05.
We thank Dr. S. Akira (Osaka University, Japan) for providing MyD88-deficient mice for use in these studies and Dr. P. Luppa for help with creatinine determinations. This work was supported by the Deutsche Forschungsgemeinschaft through SFB 576 project A7 and the Kommission für Klinische Forschung, Klinikum rechts der Isar.