In Vivo Ethanol Exposure Down-Regulates TLR2-, TLR4-, and TLR9-Mediated Macrophage Inflammatory Response by Limiting p38 and ERK1/2 Activati
http://www.100md.com
免疫学杂志 2005年第1期
Abstract
Ethanol is known to increase susceptibility to infections, in part, by suppressing macrophage function. Through TLRs, macrophages recognize pathogens and initiate inflammatory responses. In this study, we investigated the effect of acute ethanol exposure on murine macrophage activation mediated via TLR2, TLR4, and TLR9. Specifically, the study focused on the proinflammatory cytokines IL-6 and TNF- and activation of p38 and ERK1/2 MAPKs after a single in vivo exposure to physiologically relevant level of ethanol followed by ex vivo stimulation with specific TLR ligands. Acute ethanol treatment inhibited IL-6 and TNF- synthesis and impaired p38 and ERK1/2 activation induced by TLR2, TLR4, and TLR9 ligands. We also addressed the question of whether ethanol treatment modified activities of serine/threonine-specific, tyrosine-specific phosphatases, and MAPK phosphatase type 1. Inhibitors of three families of protein phosphatases did not restore ethanol-impaired proinflammatory cytokine production nor p38 and ERK1/2 activation. However, inhibitors of serine/threonine protein phosphatase type 1 and type 2A significantly increased IL-6 and TNF- levels, and prolonged activation of p38 and ERK1/2 when triggered by TLR4 and TLR9 ligands. In contrast, with TLR2 ligand stimulation, TNF- production was reduced, whereas IL-6 levels, and p38 and ERK1/2 activation were not affected. In conclusion, acute ethanol exposure impaired macrophage responsiveness to multiple TLR agonists by inhibiting IL-6 and TNF- production. Mechanism responsible for ethanol-induced suppression involved inhibition of p38 and ERK1/2 activation. Furthermore, different TLR ligands stimulated IL-6 and TNF- production via signaling pathways, which showed unique characteristics.
Introduction
Numerous consequences of ethanol consumption include its effects on immune system. Both acute and chronic ethanol exposure can modify immune responses to bacterial and viral pathogens (1, 2). Acute ethanol increases the risk of infectious complications in trauma and burn patients (3, 4), and chronic ethanol exposure leads to a higher incidence of infectious diseases, bacterial pneumonia in particular (5, 6). Acute ethanol has been shown to down-regulate LPS-induced TNF- and IL-1 production by murine macrophages (7) and by human blood monocytes (8, 9), as well as IL-6 production by murine macrophages (10).
LPS activates immune responses via interactions with TLR4 (11). At present eleven TLRs have been identified in mammals, and macrophages express mRNA for most of them (11, 12). TLRs participate in recognition of invading pathogens and initiate inflammatory responses, including proinflammatory cytokines production. Molecular mechanisms underlying the inflammatory responses stimulated via TLRs involve activation of intracellular signaling pathways that include NF-B and MAPKs (13).
MAPKs are a prominent group of serine/threonine protein kinases that in mammalian cells consist of three families: p38 MAPK, ERK, and the JNK. Mammalian ERK1 and ERK2 (ERK1/2) MAPKs predominantly mediate mitogenic and cellular differentiation signals; p38 and JNK MAPKs are mainly activated by exposure of cells to stress signals (14, 15). In general, MAPKs mediate broad range of physiological processes, including varies aspects of immune responses (16). Numerous reports indicate that MAPK signaling pathways are affected by ethanol in a manner that depends on the organ or cell type, the duration of ethanol administration (acute vs chronic), and the type of stimulatory agents (17). A study of the effect of ethanol exposure on MAPK activity in monocyte/macrophages showed that LPS-induced p38 MAPK activation was inhibited in human blood mononuclear cells cultured in the presence of ethanol (18). In addition, we recently reported that LPS-induced activation of p38 and ERK1/2 was down-regulated in macrophages obtained from mice that received a single dose of ethanol. LPS-stimulated IL-6 production was impaired in the presence of SB202190 and PD98059, inhibitors of p38 and ERK1/2, respectively, indicating the involvement of p38 and ERK1/2 activation in IL-6 production by these cells (10).
Final results of MAPK-mediated processes depend on a length and a degree of activation of the MAPKs. Because MAPKs are activated by a double phosphorylation of a relevant threonine and tyrosine residues, a removal of a single phosphate from phosphothreonine or phosphotyrosine deactivates the enzymes. Thus, the extent and the level of MAPK activation are tightly regulated by intracellular protein phosphatases (19). Ample evidence supports the involvement of phosphoserine/phosphothreonine phosphatases (e.g., protein phosphatase type 1 (PP1)3 and PP2A) (20, 21, 22), phosphotyrosine phosphatases (e.g., hemopoietic protein tyrosine phosphatase/leukocyte protein tyrosine phosphatase, HePTP/LC-PTP) (23), and dual specificity threonine/tyrosine MAPK phosphatases (e.g., MAPK phosphatase (MKP) type 1) (19, 24) in regulation of MAPK signaling. It has been proposed that ethanol may up-regulate phosphatase activity and interfere with signaling phosphorylation relay because a study showed that alcohol-impaired IFN- production was restored by vanadate, the inhibitor of phosphotyrosine phosphatase (25).
This report investigates mechanisms by which acute ethanol suppresses TLR-mediated production of proinflammatory cytokines. We demonstrated that TNF- and IL-6 production, as well as p38 and ERK1/2 activation, induced by LPS, zymosan, and CpG (ligands of TLR4, TLR2, and TLR9, respectively) were impaired in macrophages from ethanol-treated mice. In addition, we showed that inhibitors of protein phosphatases did not restore ethanol-suppressed cytokine production or MAPK activation. Moreover, ethanol administration did not affect MKP-1 levels in murine macrophages. Thus, acute ethanol had a broad suppressive effect on TLR-mediated inflammatory responses. Finally, ethanol inhibition of p38 and ERK1/2 activation triggered by TLR agonists involved mechanism(s) other than the up-regulation of protein phosphatase activity.
Materials and Methods
Animals
Nine 10-wk-old, 20–24 g male C57BL/6 mice (Harlan Breeders) were used in all experiments. Mice were acclimated for 1 wk upon arrival at the animal facilities of Loyola University Medical Center (Maywood, IL). The studies described were performed in accordance with the guidelines established by the Loyola University Chicago Institutional Animal Care and Use Committee.
Ethanol administration
Mice were randomly divided into two groups. One group, the control group, was injected i.p. with 400 μl of saline. The second group, the experimental group, was given a single i.p. injection of 2.9 g/kg body weight of ethanol (400 μl of 20% v/v ethanol in saline). As previously described (26), this dose of ethanol resulted in blood ethanol levels of 300 mg/100 ml (65 mM) at 30 min after administration. At this time point, mice were lethargic and displayed poor balance and motor coordination. Three hours after exposure to ethanol, the behavior of ethanol-treated mice returned to normal, although ethanol was still present in their circulation at 56 ± 14 mg/100 ml (12 mM). Ethanol levels were measured in blood plasma with NAD-alcohol dehydrogenase assay (Sigma-Aldrich), as previously described (26). Previous studies demonstrated that i.p. administration of ethanol did not result in the local inflammation because the percentages of macrophages and neutrophils, collected by peritoneal lavage 3 h after the treatment, were similar between control and ethanol-treated mice (10).
Cell isolation and culture
Mice were sacrificed 3 h after ethanol or saline exposure. Purified splenic macrophages were obtained from spleen cell suspensions by plastic adherence as previously described (10). Briefly, spleen cells were plated in RPMI 1640 medium supplemented with 10% FBS, 2 mM glutamine, penicillin (100 U/ml) and streptomycin (100 μg/ml) (Invitrogen Life Technologies). After a 2-h incubation, nonadherent cells were removed by washing three times with warm medium. This method resulted in a cell preparation, which was >98% positive for Mac-3 and di-I-acetylated low-density lipoprotein uptake, as shown previously by this laboratory (27). Purified macrophages were cultured for 16 h in RPMI 1640 supplemented with 2% FBS for 16 h with or without TLR ligands. In the experiments in which protein phosphatase-specific inhibitors were used, the inhibitors were present throughout the 16-h culture period.
TLR ligands included TLR4 ligand: LPS from Escherichia coli O111:B4 (100 ng/ml; Sigma-Aldrich); TLR2 ligand: zymosan (10 μg/ml; InvivoGen); TLR9 ligand: CpG-DNA (5 μM, ODN1826; InvivoGen). Zymosan and CpG-DNA were endotoxin-free (2.5 x 10–4 EU/mg) as certified by manufacturer.
Inhibitors included inhibitors of serine/threonine-specific protein phosphatases PP1 and PP2A, cantharidin and microcystin LR (Calbiochem-Novabiochem); an inhibitor of tyrosine-specific phosphatases, including MKPs; sodium orthovanadate (Calbiochem-Novabiochem); and an inhibitor of MKP-1, triptolide (Biomol). Cantharidin and microcystin LR, and triptolide were dissolved in DMSO and further diluted in cell culture medium. Sodium orthovanadate was dissolved in cell culture medium. The final concentrations of inhibitors were based on comparisons with concentrations reported in literature: cantharidin, 5 μM (28), microcystin LR, 5 nM (29, 30), and triptolide, 0.5 μM (31). Sodium orthovanadate was used at a final concentration of 1 mM (32, 33). Appropriately diluted DMSO vehicle controls were included. After 16 h culture, 95–98% cells were viable, as confirmed by trypan blue exclusion. Culture supernatants were harvested and frozen at –80°C before assessment of IL-6 and TNF- content.
Measurement of IL-6 and TNF-
The concentrations of IL-6 and TNF- in splenic macrophage supernatants were measured by commercially available ELISA kits (Endogen and BD Pharmingen, respectively), according to the manufacturer instructions, as previously described (10).
Cell extracts preparation
For the measurement of the magnitude of MAPK phosphorylation, mice were sacrificed 3 h after ethanol or saline administration. Adherence-purified splenic macrophages and were preincubated for 30 min at 37°C with inhibitors, before the addition of TLR ligands. Cells were subsequently cultured for 15–90 min in the presence or absence of 100 ng/ml LPS, 10 μg/ml zymosan, or 5 μM CpG. At the end of stimulation, the cells were washed twice with PBS and lysed as previously described (10). In brief, cell lysates were centrifuged at 12,000 x g for 10 min at 4°C. Collected supernatants were frozen at –80°C. Protein content was assessed by the Lowry method using a commercially available kit (Sigma-Aldrich).
Western blot analysis
Western blot analysis was performed as previously described (10). Abs against phospho-p38 MAPK (Thr180/Tyr182) and phospho-p44/42 (phospho-ERK1/2) MAPK (Thr202/Tyr204) (Cell Signaling Technology, Beverly, MA) were diluted 1/1000. Anti-MPK-1 Ab (Santa Cruz Biotechnology) was diluted 1/400. For the control of equal protein loading, Abs against nonphosphorylated p38 MAPK (diluted 1/1000; Cell Signaling Technology), and to GAPDH (diluted 1/4000; Novus Biologicals) were used. HRP-conjugated sheep anti-rabbit secondary Ab (Sigma-Aldrich) was diluted 1/5000, and an HRP-conjugated rabbit anti-mouse secondary Ab (diluted 1/10000; Novus Biologicals) was used to control an equal protein loading. Western blots were quantified by densitometric analysis using Ambis Optical Imaging System (AMBIS).
Statistical analysis
Data are expressed as mean ± SEM of each group. Data were analyzed by t test or ANOVA, followed by post hoc testing with Fisher’s protected least significant difference test. A value of p < 0.05 was considered significant.
Results
Acute ethanol treatment inhibits IL-6 and TNF- production triggered by TLR2, TLR4, and TLR9 agonists
We showed previously that a single dose of ethanol had a suppressive effect on IL-6 production by murine macrophages up to 24 h after the treatment (10). As a stimulus, we used E. coli-derived LPS, which triggers inflammatory responses through interactions with TLR4. To determine whether ethanol’s anti-inflammatory effect was general and not unique to the nature of the stimulant, we evaluated ligands of other TLRs. Thus, in addition to LPS, we used zymosan and CpG-DNA, agonists of TLR2 and TLR9, respectively. Mice were given a single dose of ethanol (2.9 g/kg) or saline and were sacrificed 3 h later. Macrophages obtained from ethanol- or saline-treated mice were cultured in the presence of LPS, zymosan, or CpG, and the levels of IL-6 and TNF- were measured in cell culture supernatants. All three TLR ligands, LPS, zymosan, and CpG, were strong inducers of proinflammatory cytokines, and this production was inhibited by acute ethanol exposure (Fig. 1). In three experiments, ethanol significantly (p < 0.05) reduced LPS-, zymosan-, and CpG-stimulated production of IL-6 by 50–60% and TNF- by 70–90%. IL-6 and TNF- were not detected in the media from nonstimulated cultures (data not shown).
FIGURE 1. Ethanol-induced suppression of IL-6 and TNF- production. Macrophages were obtained from mice 3 h after administration of saline or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 16 h. Cell culture supernatants were assayed for IL-6 (A) and TNF- (B) by ELISA. Data were expressed as the mean cytokine concentrations (picograms per milliliter) ± SEM. Minimum of six animals per group was analyzed and t test was applied to compare control and ethanol groups. *, p < 0.05 from control group. IL-6 and TNF- were not detected in the supernatants from unstimulated cultures (data not shown).
TLR4, TLR2, and TLR9 represent different subfamilies of TLRs (11, 34). Moreover, TLR4 and TLR2 are expressed on the cell surface (11, 35), whereas TLR9 is localized in intracellular compartments, mostly in endoplasmic reticulum (36). The fact that ethanol inhibited macrophage responses mediated by all these TLRs suggests that acute ethanol exerts a broadly suppressive effect on inflammatory response.
Acute ethanol treatment down-regulates activation of p38 and ERK1/2 MAPK induced by TLR2, TLR4, and TLR9 agonists
Interactions of TLRs with their agonists trigger intracellular signaling events, which involve activation of MAPKs (37). Moreover, we previously showed that acute ethanol inhibited LPS-triggered activation of p38 and ERK1/2 MAPKs (10). Therefore, we examined in this report whether ethanol affected activation of p38 and ERK1/2 MAPKs, induced by additional TLR agonists. Macrophages isolated from mice after ethanol or saline administration were stimulated with LPS, zymosan, or CpG. Phosphorylation levels of p38 and ERK1/2 were analyzed by Western blot. The three TLR agonists activated p38 and ERK1/2 in macrophages from both treatment groups (Fig. 2A). However, ethanol exposure resulted in a significant reduction (p < 0.05, 45–70% of the saline control cell levels) of p38 and ERK1/2 phosphorylation induced by LPS, as expected, and by zymosan and CpG (Fig. 2, B and C).
FIGURE 2. Ethanol-induced inhibition of p38 and ERK1/2 MAPKs activation. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG for 30 min. Cell extracts were prepared and analyzed by Western blot with Abs to phospho-p38 and phospho-ERK1/2. To show the equal protein loading, blots were redeveloped with anti-GAPDH Ab. Representative Western blot (A) is shown; OD of four blots from at least six mice per group, developed with Abs against phospho-p38 (B) or phospho-ERK1/2 (C) was performed. The t test was applied to compare control and ethanol groups. *, p < 0.05 from controls.
Therefore, ethanol-induced inhibition of p38 and ERK1/2 MAPKs paralleled down-regulation of proinflammatory cytokine production (as shown in Fig. 1). These results suggest that the mechanism of ethanol’s suppression of inflammatory responses in macrophages may involve the impairment of intracellular signaling processes.
Protein phosphatase-specific inhibitors up-regulate p38 activation
Duration of MAPK phosphorylation is a crucial determinant of the physiological outcome of processes mediated by MAPKs (19). Therefore, protein phosphatases that can dephosphorylate threonine, tyrosine, or both amino acid residues may be important regulators of MAPKs activity. The mechanism of ethanol’s inhibition of the immune response could involve up-regulation of phosphatase activities by ethanol (25). Thus, we examined whether extending MAPKs’ activation by inhibition of protein phosphatases, would restore ethanol-impaired activation of MAPKs and subsequent cytokine production. We used cantharidin and microcystin LR, inhibitors of serine/threonine PP1 and PP2A, and sodium orthovanadate, an inhibitor of tyrosine phosphatases and MKPs. In addition, triptolide, an anti-inflammatory compound that was shown to inhibit LPS-induction of MKP-1, (31) was examined. Evaluation of these inhibitors on LPS-induced p38 activation was assessed by Western blot analysis (Fig. 3). Maximum MAPKs phosphorylation in macrophages from both ethanol and control mice occurred after 30 min stimulation. At 60 min, phosphorylation diminished, although still present, and at 90 min could no longer be detected. Therefore, to assess the effect of phosphatase inhibitors on duration of MAPK activation, we stimulated macrophages with LPS for 90 min. At this time, as expected, phosphorylated p38 had returned to baseline levels and was no longer detected. In contrast, phospho-p38 was detected in the cells stimulated in the presence of phosphatase inhibitors. The strongest p38 activation levels were associated with serine/threonine phosphatase inhibitors, weaker with tyrosine phosphatase inhibitor, and weakest with triptolide, which indicates that at the concentrations used, inhibitors of serine/threonine phosphatases most dramatically extended duration of p38 activity. This result could be explained by the fact that serine/threonine phosphatases target both MAPK and MAPK kinases, therefore they have the capacity to activate the MAPK pathway at two regulation levels. However, phosphatase inhibitors did not eliminate the suppressive effect of ethanol treatment, and activation of p38 in macrophages from ethanol-treated mice, in the presence of inhibitors, was still impaired when compared with the corresponding control cells.
FIGURE 3. Protein phosphatase inhibitors increased LPS-induced p38 phosphorylation, but did not eliminate ethanol’s suppressive effect. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol. Cells were then stimulated with LPS (100 ng/ml) for 90 min with or without cantharidin and microcystin LR, vanadate, and triptolide. Cell extracts were prepared and analyzed by Western blot with Abs against phospho-p38. To show the equal protein loading, blots were redeveloped with Ab to nonphosphorylated p38. Representative blot of two experiments with six animals per group is shown.
Acute ethanol treatment does not affect LPS-, zymosan-, and CpG-induced MKP-1 levels
To determine whether up-regulation of dephosphorylation of MAPKs could be responsible for the ethanol dependent effects on cell signaling, we also evaluated the effect of ethanol on the levels of MKP-1, which specifically dephosphorylates MAPKs (19). Because MAPK activators also induce MKP-1, we stimulated macrophages from both treatment groups with LPS, zymosan, or CpG and assessed MKP-1 levels by Western blot analysis (Fig. 4). MKP-1 synthesis was induced by the three TLR ligands and was detected after 30 and 60 min stimulation; however, its levels did not differ between treatment groups. Thus, acute ethanol administration had no effect on the levels of MPK-1 synthesis, which indicates that ethanol-induced immunosuppression could not be explained by changes in MPK-1 activity. However, it is possible that other members of MKP family may be affected by ethanol.
FIGURE 4. MKP-1 synthesis levels were not affected by ethanol treatment. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 30 and 60 min. Cell extracts were prepared and analyzed by Western blot with Ab against MKP-1.
Inhibitors of PP1 and PP2A increase IL-6 and TNF- production stimulated by LPS and CpG, but not zymosan
In Fig. 3 we showed that inhibition of tyrosine phosphatases increased phosphorylation of p38. However, this treatment did not eliminate the ethanol-induced suppression of p38 possibly because inflammatory responses in macrophages involve additional signaling molecules, including NF-B (38). We went on to examine the effect of protein phosphatase inhibitors on IL-6 and TNF- synthesis, to determine whether these inhibitors could restore ethanol-impaired cytokine production to control levels. IL-6 and TNF- were not detected in supernatants from the cultures that contained sodium orthovanadate and triptolide despite the increase in p38 activity (data not shown). These results could be a consequence of the inhibitory action of suppressor of cytokine signaling proteins, which can be induced by TLR ligands (39, 40). Alternatively, it is possible that vanadate impaired the activity of Src homology protein-2 tyrosine phosphatase, which is involved in cytoplasmic signaling triggered by various cell surface receptors (41). This enzyme is important in IL-1-induced production of IL-6 via the pathway that includes NF-B, but not MAPKs (42). Triptolide could inhibit IL-6 and TNF- production through its effect on NF-B because it was shown that NF-B transcriptional activation was suppressed by triptolide (43, 44). In contrast, inhibition of PP1 and PP2A by cantharidin and microcystin LR up-regulated IL-6 and TNF- synthesis when triggered by LPS and CpG (p 0.01) (Fig. 5). Although PP1 and PP2A inhibitors restored ethanol-impaired IL-6 and TNF- production to control levels (i.e., cytokine levels from control cells stimulated without inhibitors), when LPS and CpG were used as inducers, a difference between the cytokine levels from ethanol vs saline treatment groups was not eliminated. In contrast, stimulation with zymosan in the presence of phosphatase inhibitors lead to further reduction of TNF- production, whereas IL-6 levels were not affected. Therefore, even though all three: LPS, zymosan, and CpG triggered IL-6 and TNF- production, the regulation of signaling pathways initiated by these TLR agonists differs. For example PP2A and PP1 may be more important in regulation of signaling initiated by LPS and CpG than by zymosan.
FIGURE 5. PP1 and PP2A inhibitors increase IL-6 and TNF- production triggered by LPS and CpG, but not zymosan. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 16 h in the presence or absence of cantharidin (5 μM) and microcystin LR (5 nM). Cell culture supernatants were assayed for IL-6 (A) and TNF- (B) by ELISA. Data were expressed as the mean cytokine concentrations (picograms per milliliter) ± SEM. Minimum of six animals per group was analyzed. ANOVA followed by Fisher’s LSD test was applied to compare values between the groups. *, p < 0.05 from control; #, p < 0.05 from cultures without inhibitors. IL-6 and TNF- were not detected in cell supernatants from unstimulated cultures (data not shown).
Inhibitors of PP1 and PP2A augment p38 and ERK1/2 activation triggered by LPS and CpG, but not zymosan
Because inhibitors of PP1 and PP2A affected IL-6 and TNF- production, we examined whether cantharidin and microcystin LR could also affect p38 and ERK1/2 activation in macrophages treated with LPS, zymosan, or CpG. The cells were pretreated with the inhibitors and stimulated for 60 min (Fig. 6). At 60 min, p38 and ERK1/2 phosphorylation levels were on the decline, but were still detectable. Moreover, the inhibitory effect of ethanol exposure on MAPKs activation was still noticeable. The presence of inhibitors resulted in an increase in p38 and ERK1/2 phosphorylation levels in LPS and CpG-stimulated cells; in contrast, zymosan-induced p38 and ERK1/2 activation were not affected. These results parallel the effect of PP1 and PP2A inhibitors on TLR ligand-induced cytokine production. Therefore, indicating a possible link between an increase in LPS- and CpG-induced IL-6 and TNF- production and the up-regulation of p38 and ERK1/2. Importantly, although PP1 and PP2A inhibitors did not increase zymosan-induced proinflammatory cytokine production, they also did not affect zymosan-stimulated p38 and ERK1/2 phosphorylation. These results further implicate p38 and ERK1/2 in the induction of IL-6 and TNF- synthesis and indicate that ethanol’s anti-inflammatory action could be explained by its down-regulation of p38 and ERK1/2 activation.
FIGURE 6. PP1 and PP2A inhibitors augmented LPS- and CpG-induced but not zymosan-induced p38 and ERK1/2 activation and did not eliminate ethanol’s suppressive effect. Macrophages were isolated from mice 3 h after administration of saline (control) or ethanol. Cells were then stimulated with LPS (100 ng/ml) for 60 min with or without cantharidin and microcystin LR. Cell extracts were prepared and analyzed by Western blot with Abs against phospho-p38 and phospho-ERK1/2. To show the equal protein loading, blots were redeveloped with anti-GAPDH Ab. Representative Western blot (A) and OD (as compared with stimulated control cells without inhibitors) of four blots from six mice per group, developed with Abs against phospho-p38 (B) or phospho-ERK1/2 (C) are shown. The t test was applied to compare treatment groups. *, p < 0.05 from cells stimulated without inhibitors.
Discussion
In this study, we report that acute ethanol treatment resulted in a broad inhibition of macrophage inflammatory responses, as exemplified by reduced production of IL-6 and TNF- by macrophages stimulated with agonists of various TLR subfamilies. We also show that the impairment of proinflammatory cytokine production was accompanied by a down-regulation of p38 and ERK1/2 MAPKs phosphorylation in murine macrophages. These results indicate that the mechanism of ethanol-induced suppression of innate immune responses may involve the impairment of MAPKs activation.
To examine ethanol’s effect on inflammatory responses generated by diverse stimuli, we chose LPS, zymosan, and CpG, ligands of TLR4, TLR2, and TLR9, respectively. Based on the comparisons of the amino acid sequences and genomic structure, TLR4, TLR2, and TLR9 represent major subfamilies of TLRs (11, 34). We did not examine stimulation via TLR3, which constitutes another TLR subfamily, because TLR3 is not present on macrophages, but only on mature dendritic cells (45). However, it was reported that TLR3 agonist-induced degradation of IL-1R-associated kinase 1, phosphorylation of p38, IL-6, and IL-12 production were impaired in peritoneal macrophages by acute ethanol treatment (46, 47).
Immunomodulatory effects of ethanol are complex, and depend on the duration of ethanol exposure (acute vs chronic) and the presence and characteristics of additional stimuli. Acute ethanol is linked to the inhibition of inflammatory responses as exemplified by the impairment of TNF-, IL-1 and IL-6 production (7, 8, 9, 10, 48, 49). This effect could be rooted in the altered properties of cellular membranes caused by ethanol. Specifically, acute ethanol exposure results in an increase in cell membrane fluidity (50), which could directly affect lipid rafts stability. These highly ordered semisolid plasma membrane subdomains contain a variety of membrane-associated proteins, many of them involved in cell signaling (51, 52). Therefore, ethanol-induced increase in cell membrane fluidity could hinder the formation of lipid rafts or alter the composition of proteins within those structures and result in the impairment of intracellular signaling. Not surprisingly, ample evidence indicates that ethanol affects signal transduction processes, among them MAPK pathway (17), which is also activated in macrophages following stimulation of TLRs.
We treated macrophages with ligands of TLR2, TLR4, and TLR9. Whereas viral and bacterial CpG-DNA motifs are the only known agonists of TLR9, TLR2, and TLR4 are activated by a broad collection of pathogen-associated molecular patterns (PAMPs). Diversity among TLR2 agonists could possibly be explained by the propensity of TLR2 to recognize PAMPs by forming heterodimers with TLR1 or TLR6, which could add to a variety of ligand binding sites. However, TLR4 is not known to associate with other TLRs, yet it is involved in signal transduction induced by LPS, taxol or respiratory syncytial virus coat protein, which are not known to have any structural similarities. Thus, it was hypothesized that in a specific ligand recognition TLRs may act not as receptors, but rather as integrators of cellular signaling (53). Therefore, TLRs may not directly bind its agonists, but rather play a role in stabilization of signaling complexes that contain a variety of pattern recognition receptors and other proteins, only some of them involved in active signaling (54, 55).
Even with the question of the exact nature of interactions between TLRs and their agonists not yet fully elucidated, it is clear that any changes in the physical or chemical state of the cell membrane would affect cell processes that involve membrane-associated proteins. Therefore, ethanol-induced increase in the cell membrane fluidity could explain the mechanism underlying ethanol’s broadly suppressive effect on IL-6 and TNF- proinflammatory cytokines production, and p38 and ERK1/2 activation. It is expected that such changes in membrane properties would not be permanent. Consistent with this concept, we showed that suppression of inflammatory responses was transient (10). Induction of cytokine production and activation of MAPKs were attenuated at 3 and 24 h and returned to control levels 48 h after a single dose of ethanol. Interestingly, it was reported that changes in membrane fluidity caused by ethanol were responsible for the impairment of TNF- production (56). These authors showed that in human monocytes/macrophage cell cultures, acute ethanol-induced TNF- suppression occurred at the posttranscriptional level and could be attributed to a decrease in membrane-based processing of TNF- by TNF--converting enzyme.
Although acute ethanol is linked to the inhibition of inflammatory responses, chronic ethanol intake is associated with their up-regulation (57, 58, 59). This apparent contradiction could also originate from ethanol-induced modifications in cell membrane properties. In chronic exposure, abnormal, slowly degraded phospholipids (e.g., phosphatidylethanol) are formed (60), moreover the fluidizing effects of ethanol are counterbalanced by changes in membrane lipids toward more saturated fatty acids that help to stabilize the membranes in ethanol-exposed cells. Therefore, acute and chronic ethanol exposure could differ in their effects on membrane protein interactions and intracellular signaling.
We showed recently that acute ethanol administration impaired LPS-induced p38 and ERK1/2 activation (10). In this study, we demonstrated that activation of p38 and ERK1/2 by multiple TLR ligands was similarly affected by ethanol. MAPK activity may be regulated at many points within the signaling pathways, including duration of MAPK phosphorylation. Because three families of protein phosphatases: serine/threonine, tyrosine, and MAPK phosphatases are involved in down-regulation of MAPKs, we examined whether ethanol affected their activity. Our data reveal that phosphatase inhibitors could not restore ethanol-impaired IL-6 and TNF- production and p38 and ERK1/2 phosphorylation levels. In addition, the levels of MKP-1 (the enzyme inducible by activators of MAPK pathways) in macrophages from saline- and ethanol-treated mice were not different. Therefore, ethanol’s inhibition of proinflammatory cytokine production could not be explained by its effect on protein phosphatases involved in control of activation of MAPKs. Interestingly, with agonists of TLR4, TLR2, and TLR9 being equally effective in the stimulation of inflammatory responses, the phosphatase-dependent regulation of these pathways showed differences. Specifically, serine/threonine phosphatase inhibitors up-regulated LPS and CpG, but not zymosan-stimulated p38 and ERK1/2 phosphorylation, as well as IL-6 and TNF- production. These results indicate differences in the regulation of signaling pathways activated by the TLR2, TLR4, and TLR9. Several adaptor molecules involved in TLR signaling have been identified (61). Because inhibitors of serine/threonine phosphatases similarly modified TLR4- and TLR9-induced pathways, and MyD88 is one of the adaptors used by TLR2 and TLR4, and the only adaptor used by TLR9, a difference in regulation of signaling may occur down stream of MyD88 activation. Alternatively, differences between TLR-mediated signaling could result from the participation of additional cell surface receptors, which could bind PAMPs and contribute to inflammatory responses. For instance, dectin-1 expressed on macrophages and dendritic cells, is a phagocytic receptor for -glucan-containing particles, including zymosan (62). The intracellular portion of dectin-1 contains an ITAM-like signaling domain (63), and concomitant engagement of dectin-1 and TLR2 was shown to augment inflammatory responses stimulated by zymosan (64). Inhibition of tyrosine phosphatases by vanadate would be expected to enhance activation of dectin-1, which could contribute to a diverse regulation of inflammatory responses induced with zymozan or LPS. However, in our hands, vanadate inhibited IL-6 and TNF- production stimulated with both PAMPs. Contribution of dectin-1 to a signaling through TLR2 and its regulation can be further assessed by the use of lipopeptide PAM3CysSerLys4, which is a TLR2 ligand that does not bind to dectin-1. In future studies, we will investigate whether acute ethanol and protein phosphatases differently affect inflammatory responses induced by PAM3CysSerLys4 and zymosan.
In conclusion, the data presented in this study support the notion that ethanol has wide-ranging suppressive effects on inflammatory immune responses, and that the mechanisms underlying these effects involve TLR-initiated signal transduction pathways, which include MAPK activation. Considering multitude and complexity of consequences associated with ethanol consumption, there is a great need to investigate the mechanisms of its action. Understanding the mechanisms of ethanol-induced effects would allow to device treatments that offset the negative consequences of ethanol in our society.
Acknowledgments
We thank Luis Ramirez for technical assistance, and Eric Boehmer and Dr. Douglas E. Faunce for helpful discussions.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants R01AA12034, T32AA13527, and 1 F31 AA015019-01, and by a Ralph and Marion C. Falk Foundation, Illinois Excellence in Academic Medicine grant.
2 Address correspondence and reprint requests to Dr. Elizabeth J. Kovacs, Departments of Cell Biology, Neurobiology, and Anatomy and Surgery, Burn and Shock Trauma Institute, Building 110, Room 4237, Loyola University Medical Center, 2160 South First Avenue, Maywood, IL 60153. E-mail address: ekovacs@lumc.edu
3 Abbreviations used in this paper: PP1, serine/threonine-specific protein phosphatase type 1; PP2A, protein phosphatase type 2A; PAMP, pathogen-associated molecular pattern; MKP, MAPK phosphatase.
Received for publication July 27, 2004. Accepted for publication October 8, 2004.(Joanna Goral, and Elizabe)
Ethanol is known to increase susceptibility to infections, in part, by suppressing macrophage function. Through TLRs, macrophages recognize pathogens and initiate inflammatory responses. In this study, we investigated the effect of acute ethanol exposure on murine macrophage activation mediated via TLR2, TLR4, and TLR9. Specifically, the study focused on the proinflammatory cytokines IL-6 and TNF- and activation of p38 and ERK1/2 MAPKs after a single in vivo exposure to physiologically relevant level of ethanol followed by ex vivo stimulation with specific TLR ligands. Acute ethanol treatment inhibited IL-6 and TNF- synthesis and impaired p38 and ERK1/2 activation induced by TLR2, TLR4, and TLR9 ligands. We also addressed the question of whether ethanol treatment modified activities of serine/threonine-specific, tyrosine-specific phosphatases, and MAPK phosphatase type 1. Inhibitors of three families of protein phosphatases did not restore ethanol-impaired proinflammatory cytokine production nor p38 and ERK1/2 activation. However, inhibitors of serine/threonine protein phosphatase type 1 and type 2A significantly increased IL-6 and TNF- levels, and prolonged activation of p38 and ERK1/2 when triggered by TLR4 and TLR9 ligands. In contrast, with TLR2 ligand stimulation, TNF- production was reduced, whereas IL-6 levels, and p38 and ERK1/2 activation were not affected. In conclusion, acute ethanol exposure impaired macrophage responsiveness to multiple TLR agonists by inhibiting IL-6 and TNF- production. Mechanism responsible for ethanol-induced suppression involved inhibition of p38 and ERK1/2 activation. Furthermore, different TLR ligands stimulated IL-6 and TNF- production via signaling pathways, which showed unique characteristics.
Introduction
Numerous consequences of ethanol consumption include its effects on immune system. Both acute and chronic ethanol exposure can modify immune responses to bacterial and viral pathogens (1, 2). Acute ethanol increases the risk of infectious complications in trauma and burn patients (3, 4), and chronic ethanol exposure leads to a higher incidence of infectious diseases, bacterial pneumonia in particular (5, 6). Acute ethanol has been shown to down-regulate LPS-induced TNF- and IL-1 production by murine macrophages (7) and by human blood monocytes (8, 9), as well as IL-6 production by murine macrophages (10).
LPS activates immune responses via interactions with TLR4 (11). At present eleven TLRs have been identified in mammals, and macrophages express mRNA for most of them (11, 12). TLRs participate in recognition of invading pathogens and initiate inflammatory responses, including proinflammatory cytokines production. Molecular mechanisms underlying the inflammatory responses stimulated via TLRs involve activation of intracellular signaling pathways that include NF-B and MAPKs (13).
MAPKs are a prominent group of serine/threonine protein kinases that in mammalian cells consist of three families: p38 MAPK, ERK, and the JNK. Mammalian ERK1 and ERK2 (ERK1/2) MAPKs predominantly mediate mitogenic and cellular differentiation signals; p38 and JNK MAPKs are mainly activated by exposure of cells to stress signals (14, 15). In general, MAPKs mediate broad range of physiological processes, including varies aspects of immune responses (16). Numerous reports indicate that MAPK signaling pathways are affected by ethanol in a manner that depends on the organ or cell type, the duration of ethanol administration (acute vs chronic), and the type of stimulatory agents (17). A study of the effect of ethanol exposure on MAPK activity in monocyte/macrophages showed that LPS-induced p38 MAPK activation was inhibited in human blood mononuclear cells cultured in the presence of ethanol (18). In addition, we recently reported that LPS-induced activation of p38 and ERK1/2 was down-regulated in macrophages obtained from mice that received a single dose of ethanol. LPS-stimulated IL-6 production was impaired in the presence of SB202190 and PD98059, inhibitors of p38 and ERK1/2, respectively, indicating the involvement of p38 and ERK1/2 activation in IL-6 production by these cells (10).
Final results of MAPK-mediated processes depend on a length and a degree of activation of the MAPKs. Because MAPKs are activated by a double phosphorylation of a relevant threonine and tyrosine residues, a removal of a single phosphate from phosphothreonine or phosphotyrosine deactivates the enzymes. Thus, the extent and the level of MAPK activation are tightly regulated by intracellular protein phosphatases (19). Ample evidence supports the involvement of phosphoserine/phosphothreonine phosphatases (e.g., protein phosphatase type 1 (PP1)3 and PP2A) (20, 21, 22), phosphotyrosine phosphatases (e.g., hemopoietic protein tyrosine phosphatase/leukocyte protein tyrosine phosphatase, HePTP/LC-PTP) (23), and dual specificity threonine/tyrosine MAPK phosphatases (e.g., MAPK phosphatase (MKP) type 1) (19, 24) in regulation of MAPK signaling. It has been proposed that ethanol may up-regulate phosphatase activity and interfere with signaling phosphorylation relay because a study showed that alcohol-impaired IFN- production was restored by vanadate, the inhibitor of phosphotyrosine phosphatase (25).
This report investigates mechanisms by which acute ethanol suppresses TLR-mediated production of proinflammatory cytokines. We demonstrated that TNF- and IL-6 production, as well as p38 and ERK1/2 activation, induced by LPS, zymosan, and CpG (ligands of TLR4, TLR2, and TLR9, respectively) were impaired in macrophages from ethanol-treated mice. In addition, we showed that inhibitors of protein phosphatases did not restore ethanol-suppressed cytokine production or MAPK activation. Moreover, ethanol administration did not affect MKP-1 levels in murine macrophages. Thus, acute ethanol had a broad suppressive effect on TLR-mediated inflammatory responses. Finally, ethanol inhibition of p38 and ERK1/2 activation triggered by TLR agonists involved mechanism(s) other than the up-regulation of protein phosphatase activity.
Materials and Methods
Animals
Nine 10-wk-old, 20–24 g male C57BL/6 mice (Harlan Breeders) were used in all experiments. Mice were acclimated for 1 wk upon arrival at the animal facilities of Loyola University Medical Center (Maywood, IL). The studies described were performed in accordance with the guidelines established by the Loyola University Chicago Institutional Animal Care and Use Committee.
Ethanol administration
Mice were randomly divided into two groups. One group, the control group, was injected i.p. with 400 μl of saline. The second group, the experimental group, was given a single i.p. injection of 2.9 g/kg body weight of ethanol (400 μl of 20% v/v ethanol in saline). As previously described (26), this dose of ethanol resulted in blood ethanol levels of 300 mg/100 ml (65 mM) at 30 min after administration. At this time point, mice were lethargic and displayed poor balance and motor coordination. Three hours after exposure to ethanol, the behavior of ethanol-treated mice returned to normal, although ethanol was still present in their circulation at 56 ± 14 mg/100 ml (12 mM). Ethanol levels were measured in blood plasma with NAD-alcohol dehydrogenase assay (Sigma-Aldrich), as previously described (26). Previous studies demonstrated that i.p. administration of ethanol did not result in the local inflammation because the percentages of macrophages and neutrophils, collected by peritoneal lavage 3 h after the treatment, were similar between control and ethanol-treated mice (10).
Cell isolation and culture
Mice were sacrificed 3 h after ethanol or saline exposure. Purified splenic macrophages were obtained from spleen cell suspensions by plastic adherence as previously described (10). Briefly, spleen cells were plated in RPMI 1640 medium supplemented with 10% FBS, 2 mM glutamine, penicillin (100 U/ml) and streptomycin (100 μg/ml) (Invitrogen Life Technologies). After a 2-h incubation, nonadherent cells were removed by washing three times with warm medium. This method resulted in a cell preparation, which was >98% positive for Mac-3 and di-I-acetylated low-density lipoprotein uptake, as shown previously by this laboratory (27). Purified macrophages were cultured for 16 h in RPMI 1640 supplemented with 2% FBS for 16 h with or without TLR ligands. In the experiments in which protein phosphatase-specific inhibitors were used, the inhibitors were present throughout the 16-h culture period.
TLR ligands included TLR4 ligand: LPS from Escherichia coli O111:B4 (100 ng/ml; Sigma-Aldrich); TLR2 ligand: zymosan (10 μg/ml; InvivoGen); TLR9 ligand: CpG-DNA (5 μM, ODN1826; InvivoGen). Zymosan and CpG-DNA were endotoxin-free (2.5 x 10–4 EU/mg) as certified by manufacturer.
Inhibitors included inhibitors of serine/threonine-specific protein phosphatases PP1 and PP2A, cantharidin and microcystin LR (Calbiochem-Novabiochem); an inhibitor of tyrosine-specific phosphatases, including MKPs; sodium orthovanadate (Calbiochem-Novabiochem); and an inhibitor of MKP-1, triptolide (Biomol). Cantharidin and microcystin LR, and triptolide were dissolved in DMSO and further diluted in cell culture medium. Sodium orthovanadate was dissolved in cell culture medium. The final concentrations of inhibitors were based on comparisons with concentrations reported in literature: cantharidin, 5 μM (28), microcystin LR, 5 nM (29, 30), and triptolide, 0.5 μM (31). Sodium orthovanadate was used at a final concentration of 1 mM (32, 33). Appropriately diluted DMSO vehicle controls were included. After 16 h culture, 95–98% cells were viable, as confirmed by trypan blue exclusion. Culture supernatants were harvested and frozen at –80°C before assessment of IL-6 and TNF- content.
Measurement of IL-6 and TNF-
The concentrations of IL-6 and TNF- in splenic macrophage supernatants were measured by commercially available ELISA kits (Endogen and BD Pharmingen, respectively), according to the manufacturer instructions, as previously described (10).
Cell extracts preparation
For the measurement of the magnitude of MAPK phosphorylation, mice were sacrificed 3 h after ethanol or saline administration. Adherence-purified splenic macrophages and were preincubated for 30 min at 37°C with inhibitors, before the addition of TLR ligands. Cells were subsequently cultured for 15–90 min in the presence or absence of 100 ng/ml LPS, 10 μg/ml zymosan, or 5 μM CpG. At the end of stimulation, the cells were washed twice with PBS and lysed as previously described (10). In brief, cell lysates were centrifuged at 12,000 x g for 10 min at 4°C. Collected supernatants were frozen at –80°C. Protein content was assessed by the Lowry method using a commercially available kit (Sigma-Aldrich).
Western blot analysis
Western blot analysis was performed as previously described (10). Abs against phospho-p38 MAPK (Thr180/Tyr182) and phospho-p44/42 (phospho-ERK1/2) MAPK (Thr202/Tyr204) (Cell Signaling Technology, Beverly, MA) were diluted 1/1000. Anti-MPK-1 Ab (Santa Cruz Biotechnology) was diluted 1/400. For the control of equal protein loading, Abs against nonphosphorylated p38 MAPK (diluted 1/1000; Cell Signaling Technology), and to GAPDH (diluted 1/4000; Novus Biologicals) were used. HRP-conjugated sheep anti-rabbit secondary Ab (Sigma-Aldrich) was diluted 1/5000, and an HRP-conjugated rabbit anti-mouse secondary Ab (diluted 1/10000; Novus Biologicals) was used to control an equal protein loading. Western blots were quantified by densitometric analysis using Ambis Optical Imaging System (AMBIS).
Statistical analysis
Data are expressed as mean ± SEM of each group. Data were analyzed by t test or ANOVA, followed by post hoc testing with Fisher’s protected least significant difference test. A value of p < 0.05 was considered significant.
Results
Acute ethanol treatment inhibits IL-6 and TNF- production triggered by TLR2, TLR4, and TLR9 agonists
We showed previously that a single dose of ethanol had a suppressive effect on IL-6 production by murine macrophages up to 24 h after the treatment (10). As a stimulus, we used E. coli-derived LPS, which triggers inflammatory responses through interactions with TLR4. To determine whether ethanol’s anti-inflammatory effect was general and not unique to the nature of the stimulant, we evaluated ligands of other TLRs. Thus, in addition to LPS, we used zymosan and CpG-DNA, agonists of TLR2 and TLR9, respectively. Mice were given a single dose of ethanol (2.9 g/kg) or saline and were sacrificed 3 h later. Macrophages obtained from ethanol- or saline-treated mice were cultured in the presence of LPS, zymosan, or CpG, and the levels of IL-6 and TNF- were measured in cell culture supernatants. All three TLR ligands, LPS, zymosan, and CpG, were strong inducers of proinflammatory cytokines, and this production was inhibited by acute ethanol exposure (Fig. 1). In three experiments, ethanol significantly (p < 0.05) reduced LPS-, zymosan-, and CpG-stimulated production of IL-6 by 50–60% and TNF- by 70–90%. IL-6 and TNF- were not detected in the media from nonstimulated cultures (data not shown).
FIGURE 1. Ethanol-induced suppression of IL-6 and TNF- production. Macrophages were obtained from mice 3 h after administration of saline or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 16 h. Cell culture supernatants were assayed for IL-6 (A) and TNF- (B) by ELISA. Data were expressed as the mean cytokine concentrations (picograms per milliliter) ± SEM. Minimum of six animals per group was analyzed and t test was applied to compare control and ethanol groups. *, p < 0.05 from control group. IL-6 and TNF- were not detected in the supernatants from unstimulated cultures (data not shown).
TLR4, TLR2, and TLR9 represent different subfamilies of TLRs (11, 34). Moreover, TLR4 and TLR2 are expressed on the cell surface (11, 35), whereas TLR9 is localized in intracellular compartments, mostly in endoplasmic reticulum (36). The fact that ethanol inhibited macrophage responses mediated by all these TLRs suggests that acute ethanol exerts a broadly suppressive effect on inflammatory response.
Acute ethanol treatment down-regulates activation of p38 and ERK1/2 MAPK induced by TLR2, TLR4, and TLR9 agonists
Interactions of TLRs with their agonists trigger intracellular signaling events, which involve activation of MAPKs (37). Moreover, we previously showed that acute ethanol inhibited LPS-triggered activation of p38 and ERK1/2 MAPKs (10). Therefore, we examined in this report whether ethanol affected activation of p38 and ERK1/2 MAPKs, induced by additional TLR agonists. Macrophages isolated from mice after ethanol or saline administration were stimulated with LPS, zymosan, or CpG. Phosphorylation levels of p38 and ERK1/2 were analyzed by Western blot. The three TLR agonists activated p38 and ERK1/2 in macrophages from both treatment groups (Fig. 2A). However, ethanol exposure resulted in a significant reduction (p < 0.05, 45–70% of the saline control cell levels) of p38 and ERK1/2 phosphorylation induced by LPS, as expected, and by zymosan and CpG (Fig. 2, B and C).
FIGURE 2. Ethanol-induced inhibition of p38 and ERK1/2 MAPKs activation. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG for 30 min. Cell extracts were prepared and analyzed by Western blot with Abs to phospho-p38 and phospho-ERK1/2. To show the equal protein loading, blots were redeveloped with anti-GAPDH Ab. Representative Western blot (A) is shown; OD of four blots from at least six mice per group, developed with Abs against phospho-p38 (B) or phospho-ERK1/2 (C) was performed. The t test was applied to compare control and ethanol groups. *, p < 0.05 from controls.
Therefore, ethanol-induced inhibition of p38 and ERK1/2 MAPKs paralleled down-regulation of proinflammatory cytokine production (as shown in Fig. 1). These results suggest that the mechanism of ethanol’s suppression of inflammatory responses in macrophages may involve the impairment of intracellular signaling processes.
Protein phosphatase-specific inhibitors up-regulate p38 activation
Duration of MAPK phosphorylation is a crucial determinant of the physiological outcome of processes mediated by MAPKs (19). Therefore, protein phosphatases that can dephosphorylate threonine, tyrosine, or both amino acid residues may be important regulators of MAPKs activity. The mechanism of ethanol’s inhibition of the immune response could involve up-regulation of phosphatase activities by ethanol (25). Thus, we examined whether extending MAPKs’ activation by inhibition of protein phosphatases, would restore ethanol-impaired activation of MAPKs and subsequent cytokine production. We used cantharidin and microcystin LR, inhibitors of serine/threonine PP1 and PP2A, and sodium orthovanadate, an inhibitor of tyrosine phosphatases and MKPs. In addition, triptolide, an anti-inflammatory compound that was shown to inhibit LPS-induction of MKP-1, (31) was examined. Evaluation of these inhibitors on LPS-induced p38 activation was assessed by Western blot analysis (Fig. 3). Maximum MAPKs phosphorylation in macrophages from both ethanol and control mice occurred after 30 min stimulation. At 60 min, phosphorylation diminished, although still present, and at 90 min could no longer be detected. Therefore, to assess the effect of phosphatase inhibitors on duration of MAPK activation, we stimulated macrophages with LPS for 90 min. At this time, as expected, phosphorylated p38 had returned to baseline levels and was no longer detected. In contrast, phospho-p38 was detected in the cells stimulated in the presence of phosphatase inhibitors. The strongest p38 activation levels were associated with serine/threonine phosphatase inhibitors, weaker with tyrosine phosphatase inhibitor, and weakest with triptolide, which indicates that at the concentrations used, inhibitors of serine/threonine phosphatases most dramatically extended duration of p38 activity. This result could be explained by the fact that serine/threonine phosphatases target both MAPK and MAPK kinases, therefore they have the capacity to activate the MAPK pathway at two regulation levels. However, phosphatase inhibitors did not eliminate the suppressive effect of ethanol treatment, and activation of p38 in macrophages from ethanol-treated mice, in the presence of inhibitors, was still impaired when compared with the corresponding control cells.
FIGURE 3. Protein phosphatase inhibitors increased LPS-induced p38 phosphorylation, but did not eliminate ethanol’s suppressive effect. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol. Cells were then stimulated with LPS (100 ng/ml) for 90 min with or without cantharidin and microcystin LR, vanadate, and triptolide. Cell extracts were prepared and analyzed by Western blot with Abs against phospho-p38. To show the equal protein loading, blots were redeveloped with Ab to nonphosphorylated p38. Representative blot of two experiments with six animals per group is shown.
Acute ethanol treatment does not affect LPS-, zymosan-, and CpG-induced MKP-1 levels
To determine whether up-regulation of dephosphorylation of MAPKs could be responsible for the ethanol dependent effects on cell signaling, we also evaluated the effect of ethanol on the levels of MKP-1, which specifically dephosphorylates MAPKs (19). Because MAPK activators also induce MKP-1, we stimulated macrophages from both treatment groups with LPS, zymosan, or CpG and assessed MKP-1 levels by Western blot analysis (Fig. 4). MKP-1 synthesis was induced by the three TLR ligands and was detected after 30 and 60 min stimulation; however, its levels did not differ between treatment groups. Thus, acute ethanol administration had no effect on the levels of MPK-1 synthesis, which indicates that ethanol-induced immunosuppression could not be explained by changes in MPK-1 activity. However, it is possible that other members of MKP family may be affected by ethanol.
FIGURE 4. MKP-1 synthesis levels were not affected by ethanol treatment. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 30 and 60 min. Cell extracts were prepared and analyzed by Western blot with Ab against MKP-1.
Inhibitors of PP1 and PP2A increase IL-6 and TNF- production stimulated by LPS and CpG, but not zymosan
In Fig. 3 we showed that inhibition of tyrosine phosphatases increased phosphorylation of p38. However, this treatment did not eliminate the ethanol-induced suppression of p38 possibly because inflammatory responses in macrophages involve additional signaling molecules, including NF-B (38). We went on to examine the effect of protein phosphatase inhibitors on IL-6 and TNF- synthesis, to determine whether these inhibitors could restore ethanol-impaired cytokine production to control levels. IL-6 and TNF- were not detected in supernatants from the cultures that contained sodium orthovanadate and triptolide despite the increase in p38 activity (data not shown). These results could be a consequence of the inhibitory action of suppressor of cytokine signaling proteins, which can be induced by TLR ligands (39, 40). Alternatively, it is possible that vanadate impaired the activity of Src homology protein-2 tyrosine phosphatase, which is involved in cytoplasmic signaling triggered by various cell surface receptors (41). This enzyme is important in IL-1-induced production of IL-6 via the pathway that includes NF-B, but not MAPKs (42). Triptolide could inhibit IL-6 and TNF- production through its effect on NF-B because it was shown that NF-B transcriptional activation was suppressed by triptolide (43, 44). In contrast, inhibition of PP1 and PP2A by cantharidin and microcystin LR up-regulated IL-6 and TNF- synthesis when triggered by LPS and CpG (p 0.01) (Fig. 5). Although PP1 and PP2A inhibitors restored ethanol-impaired IL-6 and TNF- production to control levels (i.e., cytokine levels from control cells stimulated without inhibitors), when LPS and CpG were used as inducers, a difference between the cytokine levels from ethanol vs saline treatment groups was not eliminated. In contrast, stimulation with zymosan in the presence of phosphatase inhibitors lead to further reduction of TNF- production, whereas IL-6 levels were not affected. Therefore, even though all three: LPS, zymosan, and CpG triggered IL-6 and TNF- production, the regulation of signaling pathways initiated by these TLR agonists differs. For example PP2A and PP1 may be more important in regulation of signaling initiated by LPS and CpG than by zymosan.
FIGURE 5. PP1 and PP2A inhibitors increase IL-6 and TNF- production triggered by LPS and CpG, but not zymosan. Macrophages were obtained from mice 3 h after administration of saline (control) or ethanol and stimulated with LPS (100 ng/ml), zymosan (10 μg/ml), or CpG (5 μM) for 16 h in the presence or absence of cantharidin (5 μM) and microcystin LR (5 nM). Cell culture supernatants were assayed for IL-6 (A) and TNF- (B) by ELISA. Data were expressed as the mean cytokine concentrations (picograms per milliliter) ± SEM. Minimum of six animals per group was analyzed. ANOVA followed by Fisher’s LSD test was applied to compare values between the groups. *, p < 0.05 from control; #, p < 0.05 from cultures without inhibitors. IL-6 and TNF- were not detected in cell supernatants from unstimulated cultures (data not shown).
Inhibitors of PP1 and PP2A augment p38 and ERK1/2 activation triggered by LPS and CpG, but not zymosan
Because inhibitors of PP1 and PP2A affected IL-6 and TNF- production, we examined whether cantharidin and microcystin LR could also affect p38 and ERK1/2 activation in macrophages treated with LPS, zymosan, or CpG. The cells were pretreated with the inhibitors and stimulated for 60 min (Fig. 6). At 60 min, p38 and ERK1/2 phosphorylation levels were on the decline, but were still detectable. Moreover, the inhibitory effect of ethanol exposure on MAPKs activation was still noticeable. The presence of inhibitors resulted in an increase in p38 and ERK1/2 phosphorylation levels in LPS and CpG-stimulated cells; in contrast, zymosan-induced p38 and ERK1/2 activation were not affected. These results parallel the effect of PP1 and PP2A inhibitors on TLR ligand-induced cytokine production. Therefore, indicating a possible link between an increase in LPS- and CpG-induced IL-6 and TNF- production and the up-regulation of p38 and ERK1/2. Importantly, although PP1 and PP2A inhibitors did not increase zymosan-induced proinflammatory cytokine production, they also did not affect zymosan-stimulated p38 and ERK1/2 phosphorylation. These results further implicate p38 and ERK1/2 in the induction of IL-6 and TNF- synthesis and indicate that ethanol’s anti-inflammatory action could be explained by its down-regulation of p38 and ERK1/2 activation.
FIGURE 6. PP1 and PP2A inhibitors augmented LPS- and CpG-induced but not zymosan-induced p38 and ERK1/2 activation and did not eliminate ethanol’s suppressive effect. Macrophages were isolated from mice 3 h after administration of saline (control) or ethanol. Cells were then stimulated with LPS (100 ng/ml) for 60 min with or without cantharidin and microcystin LR. Cell extracts were prepared and analyzed by Western blot with Abs against phospho-p38 and phospho-ERK1/2. To show the equal protein loading, blots were redeveloped with anti-GAPDH Ab. Representative Western blot (A) and OD (as compared with stimulated control cells without inhibitors) of four blots from six mice per group, developed with Abs against phospho-p38 (B) or phospho-ERK1/2 (C) are shown. The t test was applied to compare treatment groups. *, p < 0.05 from cells stimulated without inhibitors.
Discussion
In this study, we report that acute ethanol treatment resulted in a broad inhibition of macrophage inflammatory responses, as exemplified by reduced production of IL-6 and TNF- by macrophages stimulated with agonists of various TLR subfamilies. We also show that the impairment of proinflammatory cytokine production was accompanied by a down-regulation of p38 and ERK1/2 MAPKs phosphorylation in murine macrophages. These results indicate that the mechanism of ethanol-induced suppression of innate immune responses may involve the impairment of MAPKs activation.
To examine ethanol’s effect on inflammatory responses generated by diverse stimuli, we chose LPS, zymosan, and CpG, ligands of TLR4, TLR2, and TLR9, respectively. Based on the comparisons of the amino acid sequences and genomic structure, TLR4, TLR2, and TLR9 represent major subfamilies of TLRs (11, 34). We did not examine stimulation via TLR3, which constitutes another TLR subfamily, because TLR3 is not present on macrophages, but only on mature dendritic cells (45). However, it was reported that TLR3 agonist-induced degradation of IL-1R-associated kinase 1, phosphorylation of p38, IL-6, and IL-12 production were impaired in peritoneal macrophages by acute ethanol treatment (46, 47).
Immunomodulatory effects of ethanol are complex, and depend on the duration of ethanol exposure (acute vs chronic) and the presence and characteristics of additional stimuli. Acute ethanol is linked to the inhibition of inflammatory responses as exemplified by the impairment of TNF-, IL-1 and IL-6 production (7, 8, 9, 10, 48, 49). This effect could be rooted in the altered properties of cellular membranes caused by ethanol. Specifically, acute ethanol exposure results in an increase in cell membrane fluidity (50), which could directly affect lipid rafts stability. These highly ordered semisolid plasma membrane subdomains contain a variety of membrane-associated proteins, many of them involved in cell signaling (51, 52). Therefore, ethanol-induced increase in cell membrane fluidity could hinder the formation of lipid rafts or alter the composition of proteins within those structures and result in the impairment of intracellular signaling. Not surprisingly, ample evidence indicates that ethanol affects signal transduction processes, among them MAPK pathway (17), which is also activated in macrophages following stimulation of TLRs.
We treated macrophages with ligands of TLR2, TLR4, and TLR9. Whereas viral and bacterial CpG-DNA motifs are the only known agonists of TLR9, TLR2, and TLR4 are activated by a broad collection of pathogen-associated molecular patterns (PAMPs). Diversity among TLR2 agonists could possibly be explained by the propensity of TLR2 to recognize PAMPs by forming heterodimers with TLR1 or TLR6, which could add to a variety of ligand binding sites. However, TLR4 is not known to associate with other TLRs, yet it is involved in signal transduction induced by LPS, taxol or respiratory syncytial virus coat protein, which are not known to have any structural similarities. Thus, it was hypothesized that in a specific ligand recognition TLRs may act not as receptors, but rather as integrators of cellular signaling (53). Therefore, TLRs may not directly bind its agonists, but rather play a role in stabilization of signaling complexes that contain a variety of pattern recognition receptors and other proteins, only some of them involved in active signaling (54, 55).
Even with the question of the exact nature of interactions between TLRs and their agonists not yet fully elucidated, it is clear that any changes in the physical or chemical state of the cell membrane would affect cell processes that involve membrane-associated proteins. Therefore, ethanol-induced increase in the cell membrane fluidity could explain the mechanism underlying ethanol’s broadly suppressive effect on IL-6 and TNF- proinflammatory cytokines production, and p38 and ERK1/2 activation. It is expected that such changes in membrane properties would not be permanent. Consistent with this concept, we showed that suppression of inflammatory responses was transient (10). Induction of cytokine production and activation of MAPKs were attenuated at 3 and 24 h and returned to control levels 48 h after a single dose of ethanol. Interestingly, it was reported that changes in membrane fluidity caused by ethanol were responsible for the impairment of TNF- production (56). These authors showed that in human monocytes/macrophage cell cultures, acute ethanol-induced TNF- suppression occurred at the posttranscriptional level and could be attributed to a decrease in membrane-based processing of TNF- by TNF--converting enzyme.
Although acute ethanol is linked to the inhibition of inflammatory responses, chronic ethanol intake is associated with their up-regulation (57, 58, 59). This apparent contradiction could also originate from ethanol-induced modifications in cell membrane properties. In chronic exposure, abnormal, slowly degraded phospholipids (e.g., phosphatidylethanol) are formed (60), moreover the fluidizing effects of ethanol are counterbalanced by changes in membrane lipids toward more saturated fatty acids that help to stabilize the membranes in ethanol-exposed cells. Therefore, acute and chronic ethanol exposure could differ in their effects on membrane protein interactions and intracellular signaling.
We showed recently that acute ethanol administration impaired LPS-induced p38 and ERK1/2 activation (10). In this study, we demonstrated that activation of p38 and ERK1/2 by multiple TLR ligands was similarly affected by ethanol. MAPK activity may be regulated at many points within the signaling pathways, including duration of MAPK phosphorylation. Because three families of protein phosphatases: serine/threonine, tyrosine, and MAPK phosphatases are involved in down-regulation of MAPKs, we examined whether ethanol affected their activity. Our data reveal that phosphatase inhibitors could not restore ethanol-impaired IL-6 and TNF- production and p38 and ERK1/2 phosphorylation levels. In addition, the levels of MKP-1 (the enzyme inducible by activators of MAPK pathways) in macrophages from saline- and ethanol-treated mice were not different. Therefore, ethanol’s inhibition of proinflammatory cytokine production could not be explained by its effect on protein phosphatases involved in control of activation of MAPKs. Interestingly, with agonists of TLR4, TLR2, and TLR9 being equally effective in the stimulation of inflammatory responses, the phosphatase-dependent regulation of these pathways showed differences. Specifically, serine/threonine phosphatase inhibitors up-regulated LPS and CpG, but not zymosan-stimulated p38 and ERK1/2 phosphorylation, as well as IL-6 and TNF- production. These results indicate differences in the regulation of signaling pathways activated by the TLR2, TLR4, and TLR9. Several adaptor molecules involved in TLR signaling have been identified (61). Because inhibitors of serine/threonine phosphatases similarly modified TLR4- and TLR9-induced pathways, and MyD88 is one of the adaptors used by TLR2 and TLR4, and the only adaptor used by TLR9, a difference in regulation of signaling may occur down stream of MyD88 activation. Alternatively, differences between TLR-mediated signaling could result from the participation of additional cell surface receptors, which could bind PAMPs and contribute to inflammatory responses. For instance, dectin-1 expressed on macrophages and dendritic cells, is a phagocytic receptor for -glucan-containing particles, including zymosan (62). The intracellular portion of dectin-1 contains an ITAM-like signaling domain (63), and concomitant engagement of dectin-1 and TLR2 was shown to augment inflammatory responses stimulated by zymosan (64). Inhibition of tyrosine phosphatases by vanadate would be expected to enhance activation of dectin-1, which could contribute to a diverse regulation of inflammatory responses induced with zymozan or LPS. However, in our hands, vanadate inhibited IL-6 and TNF- production stimulated with both PAMPs. Contribution of dectin-1 to a signaling through TLR2 and its regulation can be further assessed by the use of lipopeptide PAM3CysSerLys4, which is a TLR2 ligand that does not bind to dectin-1. In future studies, we will investigate whether acute ethanol and protein phosphatases differently affect inflammatory responses induced by PAM3CysSerLys4 and zymosan.
In conclusion, the data presented in this study support the notion that ethanol has wide-ranging suppressive effects on inflammatory immune responses, and that the mechanisms underlying these effects involve TLR-initiated signal transduction pathways, which include MAPK activation. Considering multitude and complexity of consequences associated with ethanol consumption, there is a great need to investigate the mechanisms of its action. Understanding the mechanisms of ethanol-induced effects would allow to device treatments that offset the negative consequences of ethanol in our society.
Acknowledgments
We thank Luis Ramirez for technical assistance, and Eric Boehmer and Dr. Douglas E. Faunce for helpful discussions.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grants R01AA12034, T32AA13527, and 1 F31 AA015019-01, and by a Ralph and Marion C. Falk Foundation, Illinois Excellence in Academic Medicine grant.
2 Address correspondence and reprint requests to Dr. Elizabeth J. Kovacs, Departments of Cell Biology, Neurobiology, and Anatomy and Surgery, Burn and Shock Trauma Institute, Building 110, Room 4237, Loyola University Medical Center, 2160 South First Avenue, Maywood, IL 60153. E-mail address: ekovacs@lumc.edu
3 Abbreviations used in this paper: PP1, serine/threonine-specific protein phosphatase type 1; PP2A, protein phosphatase type 2A; PAMP, pathogen-associated molecular pattern; MKP, MAPK phosphatase.
Received for publication July 27, 2004. Accepted for publication October 8, 2004.(Joanna Goral, and Elizabe)