|Home | About | Journals | Submit | Contact Us | Français|
The present study addressed the modulatory role of CCR4 in TLR9-mediated innate immunity and explored the underlying molecular mechanisms. Our results demonstrated that CCR4-deficient mice were resistant to both septic peritonitis induced by cecal ligation and puncture (CLP) and CpG DNA/D-galactosamine-induced shock. In bone marrow-derived macrophages (BMMΦs) from CLP-treated CCR4-deficient mice, TLR9-mediated pathways of MAPK/AP-1, PI3K/Akt, and IκB kinase (IKK) /NF-κB were impaired compared to WT cells. While TLR9 expression was not altered, the intensity of internalized CpG DNA was increased in CCR4-deficient macrophages when compared to WT macrophages. Pharmacological inhibitor studies revealed that impaired activation of JNK, PI3K/Akt, and/or IKK/NF-κB could be responsible for decreased proinflammatory cytokine expression in CCR4-deficient macrophages. Interestingly, the CCR4-deficient BMMΦs exhibited an alternatively activated (M2) phenotype and the impaired TLR9-mediated signal transduction responses in CCR4-deficient cells were similar to the signaling responses observed in WT BMMΦs skewed to an alternatively activated phenotype. These results indicated that macrophages deficient in CCR4 impart a regulatory influence on TLR9-mediated innate immunity.
The innate immune system is the first line of defense for protecting the host against invading microbial pathogens and involves a number of host defense systems, including chemokine-directed movement of inflammatory cells and pathogen detection via pattern recognition receptors (PRRs) [1-3]. Chemokines bind to chemokine receptors, which are cell-surface G protein-coupled receptors (GPCRs), leading to subsequent inflammatory signaling events involving various second messengers [4, 5]. Chemokine receptor ligation ultimately leads to cell signaling events via PI3K, MAPK/AP-1, and the NF-κB pathways [5-7]. Toll-like receptors (TLRs) are major PRRs that have evolved to recognize pathogen-associated molecular patterns (PAMPs), which are conserved molecules of invading pathogens. After ligand binding, the majority of TLRs recruit the adaptor protein MyD88, which together forms a complex with interleukin-1 receptor associated kinase (IRAK) 4, IRAK1, and TNF receptor-associated factor 6 (TRAF6). TRAF6 subsequently activates TGF-β-activated kinase 1 (TAK1), which in turn activates MAPK/ AP-1 and the IκB kinase (IKK)/ NF-κB signaling pathways, resulting in a rapid innate immune response [1, 3]. Clearly, similar signaling pathways are evoked by these two disparate groups of receptors, but the manner in which chemokine receptors and TLRs interact during innate immunity is poorly understood.
CC chemokine receptor 4 (CCR4), a specific receptor for CCL17 and CCL22, is generally recognized as a receptor expressed on T helper 2 (Th2) cells [8-10]. However, increasing evidence has confirmed that CCR4 is expressed not only in Th2 cells but also in CD4+CD25+ T regulatory cells [11, 12], NK cells , platelets , DCs , and macrophages [16, 17]. Additionally, CCR4 appears to be a key receptor in the innate immune response as CCR4−/− mice are more resistant to LPS-induced sepsis [16, 18]. However, the cellular signal transduction events accounting for the protection and resistance in CCR4−/− mice have not been fully elucidated.
TLR9 selectively binds unmethylated CpG DNA motifs, which are common in the prokaryote genomes of most bacteria and DNA viruses and possess strong immunostimulatory activities [1, 19-21]. Unlike most PRRs, TLR9 is mainly located in the endosome and cells must internalize CpG DNA into acidified endosomes in order to bind/activate TLR9. Further, IFN-gamma-inducible protein 10 (IP-10 or CXCL10), a type 1 (Th1) chemokine, can also be produced by CpG DNA stimulation [22, 23]. TLR9 activation via CpG DNA ligand has been reported to play a pivotal role in the pathophysiology of sepsis  and severely injured patients with sepsis showed higher surface expression of TLR9 in B lymphocytes compared to healthy controls . In a corollary animal model of experimental sepsis, elevated TLR9 expression in NKT cells and macrophages was associated with organ injury .
Macrophages can be differentially activated to express a specific protein phenotype with specialized functionalities in response to exposure to microenvironmental signals such as cytokines and microbial products. In the same fashion of the Th1/Th2 paradigm, the polarized macrophages can be divided into at least two groups; classically activated (or M1) and alternatively activated (or M2) macrophages [27-29]. Classical activation (M1) is mediated by the priming stimulus IFN-γ either alone or in concert with microbial trigger such as LPS. One of the hallmarks of M1 activation is the generation of NO by iNOS . On the other hand, alternative activation (M2) is generally mediated by Th2 cytokines such as IL-4 and/or IL-13. Found in inflammatory zone-1 (FIZZ1), Arginase-1, and Mannose receptor are major markers of alternative macrophage activation. M2 macrophages have been shown to be correlated with resolution of excessive inflammation, wound repair, and promoting tumor progression, while M1 macrophages appeared to be significantly involved in pathogen defense . We previously reported that peritoneal macrophages from CCR4−/− mice possess features of an M2 phenotype. However, intracellular activation circuits that M2 macrophages rely upon for signal transduction are not well understood.
In the present study, we explored the modulation of TLR9-mediated innate immunity in CCR4−/− mice and investigated the molecular mechanisms of this regulation. Our studies demonstrate that CpG DNA activation of bone marrow-derived macrophages (BMMΦs) recovered from CCR4−/− septic mice possess a dramatically altered signal transduction profile, as compared to their WT counterpart. Further, the signal transduction pathways that are altered in the CpG DNA challenged BMMΦs recovered from the CCR4−/− mice are nearly identical to the signaling modifications observed in alternatively activated BMMΦs derived from WT mice.
To specifically examine the role of CCR4 on TLR9-mediated innate immunity in vivo, we compared the susceptibility of WT and CCR4−/− mice to cecal ligation and puncture (CLP)-induced sepsis and to CpG/D-galactosamine (D-GalN) induced acute systemic inflammatory response. As shown in Figure 1A, all WT mice died within 3 days after the initiation of polymicrobial sepsis induced by CLP, whereas 80% of CCR4−/− mice were alive after 7 days and were long-term survivors. In addition, all WT mice challenged intraperitoneally with CpG DNA (20 nmol) and D-GalN (20 mg) died within 24 h, whereas only one of seven CCR4−/− mice injected with CpG/D-GalN died (86% survival) (Fig. 1B). We also investigated the serum level of proinflammatory cytokines TNF-α and IL-6 after CpG/D-GalN stimulation and found that challenged CCR4−/− mice displayed significant suppression in serum levels of both TNF-α and IL-6 (Fig. 1C). Our results indicate that in the absence of CCR4 a protective effect on CLP-induced sepsis and CpG/D-GalN-induced shock was induced in vivo and correlated with the suppression of circulating proinflammatory cytokines.
Macrophages have been shown to play a pivotal role during the evolution of a septic response  and studies have identified that severe sepsis induces a shift in these cells to an alternatively activated (M2) macrophage that possesses a dramatically changed phenotype [16, 31]. Interestingly, CCR4−/− macrophages express an alternatively activated phenotype , which formed the basis to assess the protein expression system subsequent to TLR9 activation in BMMΦs from CLP-treated CCR4−/− mice. We assessed the signal transduction pathways that potentially could be altered in these cells and initially focused on MAPK/AP-1 and PI3K/Akt pathways because of their crucial role in the induction of cytokines involved in severe acute inflammation [32, 33]. As compared to the 0 time-point, the levels of phospho (p)-JNK, p-p38 MAPK, and p-ERK were increased in both WT and CCR4−/− BMMΦs at 0.5 h after CpG DNA challenge. However, the levels of these proteins were significantly lower in CCR4−/− BMMΦs compared to WT BMMΦs (Fig. 2A and B). These differences were more pronounced in BMMΦs from septic mice than from naïve mice (data not shown). To exclude the TLR9 independent effects of CpG DNA, BMMΦs from CLP-treated WT andCCR4−/− mice were stimulated with control non-CpG DNA for 30 min. The levels of p-JNK, p-p38 MAPK, and p-ERK were not dramatically changed in non-CpG DNA-treated group as compared with non-stimulated group (Fig. 2C). The expression of AP-1, a downstream transcription factor of JNK, p38 MAPK and ERK, was consistently lower in CCR4−/− BMMΦs than in WT BMMΦs (Fig. 2D) (mean relative intensities at 1 h were 1.1 and 2.4, respectively). Interestingly, it has been previously shown that the upregulated expression of p-Akt in neutrophils from septic patients can correlate with the severity of sepsis-induced acute lung injury . Here we show that p-Akt was up-regulated after CpG DNA challenge and was significantly higher in BMMΦs from CLP-treated WT mice compared to CLP-treated CCR4−/− mice (Fig. 2E and F). Our results indicate that both MAPK/AP-1 and PI3K/Akt signaling pathways were suppressed in CpG DNA challenge BMMΦs recovered from CCR4−/− CLP-treated mice, as compared to WT mice.
We next focused on IKK/NF-κB pathway, another important pathway leading to cytokine expression during acute inflammation . In this set of studies, we found that IκB-α expression levels were not dramatically altered in either macrophage population after CpG DNA stimulation. However, the expressions of p-IKKα/β, p-IκB-α, and p-NF-κB (p65) were observed in WT and CCR4−/− BMMΦs following CpG DNA stimulation and these levels were significantly lower in CCR4−/− BMMΦs (Fig. 3A and B). Consistent with these latter data, the NF-κB-DNA binding activities were increased in both WT and CCR4−/− BMMΦs and the binding activity was lower in CCR4−/− BMMΦs from CLP-treated mice compared to WT BMMΦs from CLP-treated mice (Fig. 3C) (mean relative intensities at 1 h were 2.1 and 3.8, respectively). Together, these results suggest that the IKK/NF-κB pathway is impaired in macrophages from CLP-treated CCR4−/− mice.
Following CpG DNA recognition, TLR9 recruits the intracellular adapter molecule MyD88 which ultimately leads to the activation of MAPK/AP-1 and IKK/NF-κB pathways . Since we showed TLR9-mediated MAPK/AP-1 and IKK/NF-κB pathways were impaired in BMMΦs from CLP-treated CCR4−/− mice, we hypothesized that signaling molecules upstream of MAPK/AP-1 and IKK/NF-κB pathways might also be impaired in BMMΦs from these mice, including TLR9 itself. To investigate this hypothesis, MyD88-dependent signaling molecules in BMMΦs from CLP-treated WT and CCR4−/− mice were examined after stimulation with CpG DNA (Fig. 4). While the basal level of MyD88 in BMMΦs from CCR4−/− mice was lower than that from WT mice, stimulation with CpG DNA had no dramatic effect on subsequent MyD88 levels in either CCR4−/− or WT BMMΦs (Fig. 4A). Interestingly, the expression levels of IRAK1 and TAK1, but not of TRAF6, TRAF2 or Toll/IL-1R domain-containing adaptor-inducing IFN-beta (TRIF), were decreased in CLP-treated CCR4−/− BMMΦs after 1 h of CpG DNA stimulation (Fig. 4A and B). We next examined the extracellular and intracellular expression of TLR9 by flow cytometry (Fig. 4C). The mean fluorescent intensity of TLR9 was significantly increased 24 h after CpG DNA stimulation in both WT and CCR4−/− BMMΦs. However, the expression levels were not significantly different between WT and CCR4−/− BMMΦs. These results suggest that the CpG DNA mediated MyD88-dependent signaling molecules IRAK1 and TAK1 are decreased in CCR4−/− BMMΦs from CLP-treated mice compared to CLP-treated WT mice and this response is not due to altered expression of TLR9.
In order to determine whether the impairment of TLR9-mediated MAPK/AP-1, IKK/NF-κB, and PI3K/Akt pathways in CCR4−/− BMMΦs was due to decreased CpG DNA internalization, WT and CCR4−/− BMMΦs were stimulated with FITC-labeled CpG oligonucleotide (ODN), and the internalized FITC-CpG ODN was quantified (Fig. 4D). Interestingly, the level of internalized FITC-CpG ODN was higher in CCR4−/− BMMΦs than in WT BMMΦs, suggesting that the defects in TLR9-mediated signaling pathways in CCR4−/− mice was not due to inadequate amounts of CpG DNA internalization.
As TLR9-mediated signaling pathways were impaired in CCR4−/− BMMΦs (Fig. 2 and and3),3), we hypothesized that TLR9-mediated expression of proinflammatory cytokines should be altered in CCR4−/− BMMΦs. To examine the role of CCR4 on CpG DNA-induced cytokine/chemokine expression, BMMΦs from CLP-treated WT and CCR4−/− mice were stimulated by CpG DNA for 6 and 24 h. The level of IL-12 p70 at 6 h was lower in CCR4−/− BMMΦs when compared to WT BMMΦs (Fig. 5). In addition, concentrations of TNF-α and CCL3, two other proinflammatory cytokines, were also suppressed in CCR4−/− BMMΦs at 24 h when compared to levels in WT BMMΦs. The level of CXCL10, a Th1 cytokine, at 24 h was suppressed in CCR4−/− BMMΦs. And the level of CCL22, one of the CCR4 ligands, at 24 h was also suppressed in CCR4−/− BMMΦs (Fig. 5) as compared with WT BMMΦs, but CCL17, the other CCR4 ligand, was not detectable (data not shown).
We demonstrated above that impaired TLR9-mediated signaling of the MAPK/AP-1, PI3K/Akt, and IKK/NF-κB pathways was accompanied by suppressed cytokine expression in CCR4−/− macrophages. Next we determined if any of these pathways contribute to the decreased cytokine expression in CLP-treated CCR4−/− macrophages. BMMΦs from CLP-treated WT mice were preincubated with the relevant pharmacological inhibitors, stimulated with CpG DNA, and the subsequent levels of TNF-α, CCL3, and IL-12 p70 were measured. Inhibitors of Akt, JNK, and IKK, but not p38 MAPK suppressed the expression of TNF-α, CCL3, and IL-12 p70. By pretreatment with ERK inhibitor, the expression level of TNF-α was decreased, while CCL3 was not significantly changed, and IL-12 p70 was increased (Fig. 6A-C). These results suggest that p38 MAPK and ERK do not significantly contribute to impaired CCL3 and IL-12 p70 expression, and p38 MAPK does not contribute to impaired TNF-α expression. On the other hand, Akt, JNK, and/or IKK/NF-κB may be responsible for the impaired cytokine expression observed in CCR4−/− macrophages.
We demonstrated that PI3K/Akt appears to be one of the important pathways in TLR9-mediated signaling, however, little is known about the relationship between PI3K/Akt and other signaling pathways. To investigate the relationship between PI3K/Akt, MAPK, and IKK/NF-κB in macrophages, BMMΦs from CLP-treated WT mice were preincubated with medium (containing control levels of DMSO) or with the pharmacological inhibitor of Akt (LY294002), stimulated with CpG DNA, and the level of p-JNK, p-38 MAPK, p-ERK, p-IKKα/β, and p-IκB-α were measured. We first confirmed that p-Akt was suppressed by the Akt inhibitor (Fig. 6D) and the expression level of p-IκB-α, but not p-IKKα/β, was suppressed in LY294002-treated macrophages (Fig. 6E). On the other hand, the expression levels of p-JNK and p-p38 MAPK in LY294002-treated macrophages were comparable to the control level (Fig. 6F). The p-ERK level in the LY294002-treated macrophages was slightly higher or comparable to control level. These results suggest that PI3K/Akt pathway is an upstream positive regulator of IκB-α/NF-κB but not IKK, JNK, p38 MAPK, or ERK in macrophages from CLP-treated mice.
We previously showed that peritoneal macrophages from CCR4−/− mice exhibited a feature of alternatively activated (or M2) protein phenotype, but did not understand the mechanism (16). To determine if the impaired TLR9-mediated signaling response in CLP-treated CCR4−/− macrophages is due to skewing toward an M2 phenotype, we examined whether BMMΦs from CLP-treated CCR4−/− mice express an alternatively activated phenotype. An assessment of the M2 protein markers, FIZZ1, Arginase-1, and Mannose receptor, were all markedly higher in CCR4−/− BMMΦs, as compared to WT BMMΦs (Fig. 7A). Previous study has shown that CCL17 and IL-10 inhibit classically activated macrophages generation from resident macrophages stimulated with CpG DNA . To examine whether the cocktail of IL-4, IL-13, IL-10, and CCL17 can skew WT macrophages to M2 phenotype more efficiently compared with the cocktail of IL-4 and IL-13, BMMΦs from CLP-treated mice were incubated with IL-4 and IL-13, or IL-4, IL-13, IL-10, and CCL17, and the mRNA levels of M2 markers were measured. The mRNA levels of FIZZ1, Arginase-1, and Mannose receptor, were significantly higher in IL-4, IL-13, IL-10, and CCL17-treated BMMΦs than in IL-4 and IL-13-treated BMMΦs (Fig. 7B). Next, BMMΦs from WT CLP-treated mice were incubated with or without IL-4, IL-13, IL-10, and CCL17, and TLR9-mediated signaling responses were examined. The levels of p-JNK, p-p38 MAPK, p-ERK, p-Akt, p-IκB-α, p-NF-κB (p65), and p-IKKα/β were lower in M2-skewed BMMΦs compared to non-treated BMMΦs, and the levels of IRAK1 and TAK1 were totally suppressed (Fig. 7C). On the other hand, MyD88 was slightly increased. These results suggest that the impaired-TLR9-mediated signaling response in BMMΦs from CCR4−/− mice is, at least partially, because of their M2 phenotype.
Increasing evidence suggest that chemokine receptors and TLRs work in a synergistic manner during innate immune responses. Previously, we and others have examined this linkage in CCR4−/− mice, which are resistant to LPS (TLR4 ligand) and Pam3Cys (TLR2 ligand) challenge [16, 18]. In our previous study, peritoneal macrophages from CCR4−/− mice exhibited an alternatively activated (M2) phenotype, and TLR4 and TLR2-mediated NF-κB signaling pathway were suppressed whereas p-p38 MAPK and p-JNK were conversely increased in peritoneal CCR4−/− macrophages . Since TLR9 plays a key role in sepsis [24-26], we hypothesized that CCR4 could also modulate TLR9-mediated innate immune response in experimental sepsis and shock. In our study, we showed CCR4−/− mice were highly resistant to CLP-induced sepsis, which seems to be inconsistent with our previous study that administration of CCL22, one of the CCR4 ligands, is protective and inhibition of CCL22 is detrimental to CLP-induced sepsis . However, previous studies have shown that both CCL22 and CCL17 may bind to different chemokine receptors in addition to CCR4 [36, 37]. Indeed, CCR8 gene level was 4.5 fold higher in CCR4−/− BMMΦs as compared with the level in WT BMMΦs (data not shown). Therefore CCR4 deficiency does not necessarily result in the inhibition of CCL22 or CCL17 bioactivity. We also confirmed that CCR4−/− mice were resistant to CpG/D-GalN-induced shock likely via the suppression of certain proinflammatory cytokines, which contribute to the pathophysiology of acute systemic inflammatory response. However the mechanisms that direct cytokine regulation in this system are not well understood.
We observed that TLR9-mediated cytokine expression and lethality in CCR4−/− mice were modulated in our experimental models of severe systemic inflammation, which led us to hypothesize that TLR9-mediated intracellular signaling could be altered in cells recovered from CCR4−/− mice subsequent to experimental sepsis. In contrast to dendritic cells, which are well studied in TLR9 signaling [2, 3, 20], we focused on signaling pathways in macrophages, key cells in sepsis-induced pathology . Since ligand binding to TLR9 results in activation of MAPK/AP-1 and IKK/NF-κB signaling pathways , we initially focused on these intracellular activation events. MAPK pathways are highly conserved signaling pathways which regulate various cellular functions . The MAPK signaling pathways can be activated by many stimuli, such as an environmental stress, cytokines, and TLR ligands [2, 38, 39]. In this study, activation of JNK, p38 MAPK, and ERK were all significantly impaired in CLP-treated CCR4−/− macrophages after CpG DNA stimulation. Activity of AP-1, a downstream transcription factor of JNK, p38 MAPK, and ERK, was consistently suppressed in CCR4−/− macrophages. These results suggest that the absence of CCR4 attenuates TLR9-mediated MAPK/AP-1 signaling, which would help control the excessive inflammatory response which accompanies sepsis.
In our study, both the expression levels of p-IKKα/β, p-IκB-α, and p-NF-κB (p65) and the NF-κB DNA binding activity were consistently suppressed in BMMΦs from CLP-treated CCR4−/− mice compared to CLP-treated WT mice. Thus, we would expect to find less IκB-α in the BMMΦs from WT mice than from CCR4−/− mice, but instead we found comparable levels. The reason for this result is unclear, but possibly the peak of IκB-α degradation occurs at time points we did not measure or other IκB proteins such as IκB-β, IκB-ε, or IκBNS may participate in regulating NF-κB activity. Our results clearly suggest that the absence of CCR4 attenuates TLR9-mediated IKK/NF-κB signaling, which would also help control the excessive proinflammatory response which accompanies early sepsis.
PI3Ks are lipid kinases, which phosphorylate phosphoinositides, are key enzymes involved in a number of important signaling events . Class I PI3K catalyzes the production of phosphatidylinositol3, 4, 5-triphosphate by phosphorylating phosphatidylinositol-4, 5-bisphosphate, resulting in activating downstream targets including Akt. The PI3K/Akt pathway is thought to participate in the TLR9 signaling pathway, as well as in TLR2 and TLR4 signaling pathways, and can act as either a positive or negative regulator of TLR signaling [40-44]. Akt has also been shown to play a key role in sepsis . In our investigation we observed that the level of p-Akt was suppressed in BMMΦs from CLP-treated CCR4−/− mice compared with WT mice. These results indicate that the absence of CCR4 attenuates TLR9-mediated PI3K/Akt signaling pathway during experimental sepsis.
After stimulation with CpG DNA, TLR9 recruits the adapter molecule MyD88 resulting in the activation of the MAPK/AP-1 and IKK/NF-κB signaling pathways. Since our results showed that both MAPK/AP-1 and IKK/NF-κB pathways were suppressed after CpG DNA stimulation, we asked whether upstream TLR9-mediated MyD88-dependent signaling molecules and TLR9 itself, were also modulated. Indeed we found that the expression of MyD88-dependent signaling molecules IRAK1 and TAK1 were decreased in macrophages from CCR4−/− mice. On the other hand, we showed the intensity of TLR9 expression was comparable between WT and CCR4−/− BMMΦs after CpG DNA challenge and further, that the intensity of internalized CpG DNA increased in CCR4−/− macrophages compared to WT macrophages. These results indicate that impaired TLR9-mediated signaling pathways such as MAPK/AP-1, PI3K/Akt, and IKK/NF-κB pathways are not due to decreases in TLR9 expression nor decreases in internalization of CpG DNA.
Macrophages can be phenotypically polarized and classified into at least two groups, classically activated (or M1) and alternatively activated (or M2) macrophages, that are defined by the proteins they express . FIZZ1, Arginase-1, and Mannose receptor are major markers of alternative macrophage activation, while iNOS is a widely used marker of classically activated macrophages. Alternatively activated macrophages have been shown to be correlated with resolution of excessive inflammation, tumor development, and wound repair, while classically activated macrophages appeared to be significantly involved in pathogen defense . Our results indicate that BMMΦs from CLP-treated CCR4−/− mice exhibit an alternatively activated M2 phenotype, which, at least in part, lead to impaired TLR9-mediated signaling response, contributing to resolution of excessive inflammation. We also confirmed that BMMΦs recovered from CLP-treated WT mice in which CCR4 was neutralized by CCR4 antibody during macrophage maturation have features of M2-polarized phenotype (data not shown). In addition, the signal transduction pathway proteins altered in the CCR4−/− macrophages were similar to those altered in WT macrophages treated with M2 skewing cytokines (IL-4, IL-13, IL-10, and CCL17). The reason we used CCL17 plus other Th2 cytokines for M2 skewing was based on previous study demonstrating that CCL17 and IL-10 inhibit classically activated macrophages generation from resident macrophages stimulated with CpG DNA . However, we did not see any dramatic differences between IL-4/IL-13/IL-10-treated and IL-4/IL-13/IL-10/CCL17-treated WT BMMΦs in TLR9-mediated signaling resposes (data not shown).
We also assessed the expression of inflammatory cytokines TNF-α, CCL3, and IL-12 p70, as their expression can be controlled through TLR9-mediated MAPK/AP-1 and IKK/NF-κB pathway [2, 3]. The level of these inflammatory cytokines was significantly suppressed after CpG stimulation in BMMΦs from CLP-treated CCR4−/− mice. Therefore, we tried to determine which pathway was responsible for this impaired expression and found that the pharmacological inhibition of Akt, JNK, or IKK resulted in decreased expression of TNF-α, CCL3, and IL-12 p70. Although our results do not evaluate how much each pathway contributes to the suppressed cytokine expression, Akt, JNK, and/or IKK/NF-κB pathway are all candidates for driving the observed cytokine suppression in CCR4−/− macrophages. Studies were also initiated to explore the relationship among the PI3K/Akt, MAPK/AP-1 and IKK/NF-κB signaling pathways and subsequent cytokine expression via the use of signaling pathway inhibitors. These investigations showed that the inhibition of Akt led to decreased expression of p-IκB-α, but not p-JNK, p-p38 MAPK, and p-IKKα/β. These results suggest that Akt is an upstream positive signaling molecule of IκB-α/NF-κB, but not IKKα/β, JNK, and p38 MAPK.
To summarize the salient findings of this research, we have shown that CCR4−/− mice were highly resistant to CLP-treated sepsis and CpG DNA-induced shock in vivo, which was accompanied by suppressed proinflammatory cytokine expression. We also demonstrated that TLR9-mediated MAPK/AP-1, PI3K/Akt, and IKK/NF-κB signaling pathways were impaired, and that the expression of several TLR9-mediated MyD88-dependent molecules were suppressed in macrophages from CLP-treated CCR4−/− mice. Interestingly, these signaling pathways impairments were not due to decreases in TLR9 expression or in the internalization of CpG DNA. Alterations in these signaling molecules subsequently regulated the expression of macrophage-derived inflammatory cytokines (TNF-α, CCL3, IL-12 p70, CXCL10, and CCL22) from CLP-treated CCR4−/− mice. A potential mechanism that may account for the alterations in TLR9 signal transduction in the CCR4−/− mice is that regulators of G-protein signaling (RGS) may be altered in the CCR4−/− mice. The biologic activity of RGS proteins have previously been shown to reach beyond only regulating GPCRs and can interact with other intracellular activation systems [45, 46]. One potential mechanistic explanation for our data could relate to alterations in normal RGS activity in the CCR4−/− mice, which would influence TLR9 signaling. This concept is worthy of further exploration.
Mouse CpG DNA (ODN) (HC4033) and control non-CpG DNA (HC4034) were purchased from Cell Sciences (Canton, MA). Mouse FITC labeled CpG ODN was purchased from Invivogen (San Diego, CA). D-galactosamine was purchased from Sigma-Aldrich (St. Louis, MO). Antibodies specific for p-SAPK/JNK, p-p38 MAPK (3D7), p38 MAPK, p-p44/42 (or p-ERK), ERK, p-Akt (Ser473), Akt (Ser473), p-IκB-α (Ser32), IκB-α, p-NF-κB (p65) (Ser536), NF-κB (p65), p-IKKα/β, IKKβ (L570), TRAF2, TAK1, and HRP-conjugated rabbit and mouse IgG antibody were purchased from Cell Signaling Technology (Danvers, MA). Antibodies for JNK1 (C17), IRAK1 (H273), and TRAF6 (H274) were obtained from Santa Cruz (Santa Cruz, CA). Antibodies for TRIF and GAPDH were purchased from Abcam (Cambridge, MA). Antibody for MyD88 was purchased from Millipore (Billerica, MA). Antibody for PE-conjugated TLR9 was purchased from Imgenex (San diego, CA). Double-stranded NF-κB, AP-1, and Octomer-1 (Oct-1) consensus ODN probes were purchased from Promega (Madison WI). LY294002, SP600125, SB203580, U0126, and Wedelolactone were purchased from Calbiochem (San Diego, CA). Recombinant mouse IL-4, IL-13, IL-10, and CCL17/TARC were purchased from R&D systems (Minneapolis, MN). .
Female WT C57BL/6 mice (6 to 10 weeks old) were purchased from Taconic (Hudson, NY). Female CCR4−/− mice, also female C57BL/6, were provided by Tularik (Thousand Oaks, CA) and were generated as previously described . Mice were housed under specific pathogen-free conditions, and all animal experiments were approved by the Animal Use Committee at the University of Michigan.
CLP surgery was performed as described . Briefly, mice were anesthetized by i.p. injection of mixture of 2.25 mg of ketamine hydrochloric acid (Abbott Laboratories, Chicago, IL) and 150 μg of xylazine (Lloyd Laboratories, Shenandoah, IA). A 1-cm midline incision was made on the lower abdomen. The exposed cecum was ligated with a 3-0 silk suture, and punctured nine times using a 21-gauge needle. The cecum was returned to the peritoneal cavity, and the incision was closed with surgical staples. All mice received 1 ml of sterile saline s.c. for fluid resuscitation immediately after surgery. For the CpG DNA-induced shock model, mice were injected i.p. with CpG DNA (20 nmol) in the presence of D-GalN (20 mg).
Bone marrow cells were collected from CLP-treated (48 h) WT and CCR4−/− mice by flushing their femurs and tibias with RPMI 1640 (Mediatech, Herndon, VA). Then the cells were cultured with bone marrow medium (20% FCS, 30% L-cell supernatant, L-glutamine, penicillin/streptomycin (P/S) in RPMI 1640). On day 3, fresh bone marrow medium was added. On day 6, the cells were replated in RPMI 1640 medium. After overnight rest, cells were stimulated with CpG DNA for the indicated times. In some experiments, cells were pre-treated with or without IL-4, IL-13, IL-10, and CCL17 for 48 h, and were stimulated with CpG DNA.
The cells were lysed at each time point after CpG DNA challenge in cell lysis buffer (Cell Signaling Technology), kept on ice for 20 min, and centrifuged at 15,000 g for 15 min. The supernatant was collected and stored at −80 °C until use. Total protein concentration of the samples was measured by bicinchroninic acid (BCA) protein assay (Pierce, Rockford, IL). Equal amount (15-30 μg) of cell lysates were fractionated by SDS-PAGE (Nupage; Invitrogen, Carlsbad, CA). Then the proteins were transferred onto nitrocellulose membrane (Invitrogen). After the overnight incubation with appropriate primary antibody, the membrane was counterstained with HRP-conjugated rabbit or mouse IgG antibody, visualized with enhanced chemiluminescence detection reagents (ECL; GE Healthcare, Piscataway, NJ). The images were analyzed using Image J 1.37v (National Institute of Health).
Nuclear protein extracts were prepared using CelLytic™ NuCLEAR™ Extraction Kit (Sigma-Aldrich) according to the manufacture’s instruction. In brief, after CpG DNA stimulation, 1 × 107 cells were collected, washed with Dulbecco's PBS twice, resuspended in 1 × Lysis buffer. After 15 min incubation, a detergent (0.6% IGEPAL) was added, and cells were vortexed for 10 s and centrifuged at 10,000 g for 30 s. The pelleted nuclei were resuspended in extraction buffer and incubated on ice for 30min. Samples were centrifuged at 20,000 g for 5 min, and the supernatants containing nuclear protein were stored at −80 °C until use.
Electrophoretic mobility shift assay was performed using the Gel Shift Assay System (Promega) as described previously . 5μg of nuclear protein was incubated with 32P-labeled double-stranded ODN of AP-1, NF-κB, or Oct-1 (an ubiquitous transcription factor used as a loading control) in gel shift binding buffer (Promega), electrophoresed in 6% DNA Retardation Gel (Invitrogen) and visualized by autoradiography. The images were analyzed by Image J 1.37v (National Institute of Health).
For serum sample preparation of ELISA, mice were anesthetized by i.p. injection of mixture of 2.25 mg of ketamine hydrochloric acid and 150 μg of xylazine at 2 h after CpG DNA challenge, heart was exposed, and blood was collected by heart puncture, and was centrifuged at 7,000 rpm for 8 min. The supernatants were stored at −20 °C until use. Supernatants of BMMΦs for ELISA were collected at 6 h and 24 h after CpG DNA challenge and stored at −20 °C until use. Mouse TNF-α, IL-6, CCL3, IL-12 p70, CXCL10, and CCL22 were measured by standardized sandwich ELISA assays. The captured antibodies, detection antibodies, and recombinant cytokines were purchased from R&D Systems (Minneapolis, MN). The limit of ELISA detection for TNF-a, IL-6, and CCL22 were greater than 25, 25, and 10 pg/ml, respectively, and that for other cytokines were consistently greater than 50 pg/ml.
RNA was isolated from BMMΦs by TRIzol according to manufacture's instruction (Invitrogen). Reverse transcription was performed to yield cDNA in a 25-μL reaction mixture containing 1X first strand (Invitrogen), 250 ng of oligo (dT) primer, 1.6mM dNTPs (Invitrogen), 5 U RNase inhibitor (Invitrogen), and 100 U of Moloney murine leukemia virus reverse transcriptase (Invitrogen) at 38°C for 60 min and the reaction was stopped by incubating the cDNA at 94°C for 10 min. Real-time quantitative PCR analysis was performed by 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA). The sequences of the primers for Mannose receptor were 5′-CCATCGAGACTGCTGCTGAG-3′ (forward) and 5′-AGCCCTTGGGTTGAGGATCC-3′ (reverse) (SYBR Green). The sequences of the primers for FIZZ-1 were 5′- TCCAGCTAACTATCCCTCCACTGT-3′ (forward), 5′-GGCCCATCTGTTCATAGTCTTGA-3′ (reverse), and the probe was 6FAM-CGAAGACTCTCTCTTGC-TAMRA. Gene expression assays for Arginase-1 (Mm01190441_g1) were purchased from Applied Biosystems. GAPDH (Applied Biosystems) was used for loading control.
After fixation with 4 % paraformaldehyde (Fisher Scientific), cells were permeabilized by Perm/Wash buffer (BD Biosciences) for 15 min, and then stained with PE-conjugated TLR9 antibody for 20 min. Analysis was performed with Cytomics FC500 Flow Cytometry Systems (Beckman Coulter, Fullerton, CA) and FlowJo 7.1.3 software (Tree Star, Ashland, OR).
Analysis for CpG ODN internalization was performed as described in the literature [44, 49]. Briefly, cells were stimulated with FITC-labeled CpG ODN (1 μM) for 1 h, then washed with PBS, and the cell surface FITC-CpG ODN was quenched by 0.1% trypan blue (Sigma-Aldrich). Internalized FITC-CpG ODN was analyzed with Cytomics FC500 Flow Cytometry Systems and FlowJo 7.1.3 software (Tree Star).
For survival studies, the log-rank test was used. For other data, statistical significance was determined using the unpaired Student's t test or ANOVA followed by Turkey's test for multiple comparisons as appropriate. Values are presented as mean ± SEM. P values less than 0.05 were deemed statistically significant. Calculations were performed using the Prism 4.0 software program for Windows (GraphPad Software, San Diego, CA) or StatView II (Abacus Concepts, Berkeley, CA).
We wish to thank Ms. Pamela M. Lincoln, Ms. Holly L. Evanoff, and Ms. Lisa Riggs for their technical assistance, and Ms. Robin G. Kunkel for her artistic work. And we are indebted to Dr. Judith M. Connett for critical reading of the manuscript.
This work was supported by grants received by Steven L. Kunkel from the National Institutes of Health (HL31327, HL74024, and HL31963). Makoto Ishii is funded by the Kanae Foundation for the promotion of medical science in Japan.
Conflict of interest
The authors have no conflicting financial interests.