Heme Oxygenase-1 Mediates the Anti-Inflammatory Effects of Acute Alcohol on IL-10 Induction Involving p38 MAPK Activation in Monocytes

Alcohol use is associated with increased susceptibility to infections with bacterial and viral pathogens (1). Monocytes/macrophages play an important role in the immune response, because they phagocytose, digest, and process pathogens, present Ags to naive T cells, and release inflammatory and anti-inflammatory cytokines. We have previously shown that acute ethanol exposure inhibits TNF-α and increases IL-10 production after LPS stimulation in human monocytes (2, 3). It has recently been demonstrated that monocytes are the principal source of heme oxygenase-1 (HO-1)³ within PBMC (4). Heme oxygenase (HO) is the rate-limiting enzyme in heme catabolism, leading to the generation of biliverdin, free iron, and carbon monoxide (CO) (5, 6, 7). Three HO isoforms (HO-1, HO-2, and HO-3) that catalyze this reaction have been identified. HO-1 (32 kDa) is inducible, and HO-2 (36 kDa) is constitutively synthesized and exists primarily in the brain and testis, whereas HO-3, a recently cloned gene product (33 kDa), is less well characterized (7, 8). The importance of HO-1 in physiological and pathological states is underlined by the versatility of HO-1 inducers and the cytoprotective effects attributed to heme oxygenase products in conditions that are associated with moderate or severe cellular stress, such as inflammation, endotoxemia, ischemia, and radiation (8). These stress signals also activate the nuclear regulatory factor κB (NF-κB), which is involved in regulation of inflammatory pathways (9, 10). However, acute alcohol treatment can inhibit both inflammatory cytokine induction and NF-κB activation in monocytes (1, 2, 11, 12). Recent research suggests that CO, a product of HO-1 metabolism, augments IL-10 release and inhibits TNF-α production after stimulation with LPS in RAW264.7 macrophages (13), which are similar to those effects displayed after acute ethanol treatment in human monocytes (2, 3). Despite convincing data indicating the cytoprotective function of HO-1 in oxidative stress, the mechanisms by which HO-1 serves as a cytoprotectant remain poorly understood. LPS, a potent HO-1 activator, has a central role in alcohol-induced liver damage via activation of Kupffer cells, the resident liver macrophages (14). LPS binds to CD14 and TLR4 at the cell surface (15) and then activates the MAPK pathways, including p38, p42/p44 ERK, and JNK (15, 16, 17). Otterbein et al. (13) found involvement of the MKK-3/p38 pathway in the effects observed with CO in macrophages, whereas Gong et al. (18) reported a connection between ethanol and the ERK pathway in the activation of HO-1 in HepG2 (E47) cells. Considering that LPS-induced inflammatory pathways are modulated by ethanol, we were interested in investigating the role of p38 in the activation of HO-1 by ethanol in human monocytes. The gene encoding HO-1 contains binding sites for NF-κB and AP-1 in its promoter/enhancer regions (8). Lee et al. (19) presented in vitro and in vivo data suggesting that AP-1 activation may represent one mechanism mediating hyperoxia-induced HO-1 gene transcription in the lung. Here, we demonstrate that acute ethanol augments IL-10 via HO-1 activation in monocytic cells and that TNF-α inhibition by ethanol involves HO-1. We also report that LPS stimulation in combination with ethanol leads to differential activation of the redox-sensitive transcription factors NF-κB and AP-1 and to p38 MAPK phosphorylation, which plays a critical role in mediating the effects of alcohol on HO-1 activation and IL-10 induction.

RPMI 1640 was from Invitrogen Life Technologies, and FBS was from HyClone. SB203580 was from Sigma-Aldrich, and LPS (Escherichia coli strain 0111:B4) was from Difco Laboratories. HO-1 inhibitors zinc protoporphyrin IX (ZnPP), chromium protoporphyrin IX (CrPP), tin mesoporphyrin IX (SnMP), were from Frontier Scientific; HO-1 activators cobalt protoporphyrin (CoPP) was from Frontier Scientific, and cobalt chloride (CoCl₂) was from Sigma-Aldrich. HO-1 activators and inhibitors were dissolved in 10 mM DMSO. For each monocyte stimulation in the presence of HO-1 inhibitors, a DMSO control was included and the presence of DMSO resulted in no change in TNF-α or IL-10 production (data not shown). NF-κB and AP-1 oligonucleotides were from Promega. Anti-HO-1 Ab was from Stressgen Biotechnologies, the PhosphoPlus p38 kit was from Cell Signaling (NE Biolabs), and polyclonal anti-rabbit secondary Ab was from Santa Cruz Biotechnology.

The study was approved by the Institutional Human Subjects Committee, and informed consent was obtained from each donor. Peripheral blood was taken by venipuncture from healthy volunteers using 10 U/ml heparin as an anticoagulant.

Monocytes from human peripheral blood were isolated by adherence from Ficoll-Hypaque mononuclear cell preparation as previously described (2, 11, 20). Adherent monocytes were cultured overnight in RPMI 1640 supplemented with 10% FBS at 37°C, and then stimulated with LPS with and without 25 mM ethanol and HO-1 inhibitors, HO-1 activators, or SB203580, a p38 MAPK inhibitor, as indicated in the figures. The 25 mM ethanol in vitro concentration is equivalent to a 0.1 mg/dl in vivo blood ethanol level. Cell-free supernatants were collected after stimulation as indicated for each experiment. Cells were either directly lysed off the plates for RNA extraction or Western blot experiments, or scraped off for enzyme activity assays and nuclear extraction. Where indicated, the RAW 264.7 mouse macrophage cell line (a generous gift from Dr. Bowie (Trinity College, Dublin, Ireland) maintained in DMEM with 10% FBS) was used as a surrogate for monocytes/macrophages.

Adult (7- to 10-wk-old) female C57BL/6 mice were purchased from The Jackson Laboratory. Animals were maintained in a temperature- and light-controlled animal facility at the University of Massachusetts Medical School on standard laboratory chow diet and water ad libitum. All animals received care in compliance with institutional requirements and this study underwent full Institutional Animal Care and Use Committee review at the University of Massachusetts Medical School according to National Institutes of Health guidelines. Mice were injected i.p. with LPS (E. coli 0111:B4; Sigma-Aldrich; 0.5 μg/g body weight in 200 μl saline) or saline control. CoPP and ZnPP (each 50 μM/kg body weight in 300 μl saline) were administered i.p., 2 h before ethanol injection. Acute ethanol (20% v/v in saline; 200 μl) was administered i.p. either as a single challenge or 1 h before LPS administration as a combined stimulation. This dose of ethanol resulted in comparable maximum blood alcohol levels (150–200 mg/dl) at 2 h (data not shown). Spleens and serum were collected promptly 90 min after LPS challenge. Splenocytes were isolated by homogenization through a 70-μm pore size cell strainer (BD Falcon), depleted of RBC by osmotic lysis, and cultured for 2 h in RPMI 1640 supplemented with 2-ME (50 μM), penicillin, gentamicin, MEM amino acids, sodium pyruvate, and 10% FBS. Nonadherent cells were washed with warm RPMI 1640, while adherent splenic macrophages were cultured and analyzed for cytokine production in ELISA.

Nuclear and cytoplasmic extracts were obtained from cells with or without stimulation by the modified method of Dignam, as previously described (11, 21). Protein content was determined in all extracts using the Dye Reagent Assay from Bio-Rad. An equal amount (5 μg) of protein from each sample and a consensus NF-κB or AP-1 oligonucleotide (Promega) were used for electromobility gel shift assays (EMSA) as previously described (11, 20). The gels were dried and exposed to autoradiography. The densitometric analysis was performed using GDS-800 system and Labwork 4.0 program (UVP BioImaging Systems).

Cell-free culture supernatants were tested for TNF-α (BD Biosciences) and IL-10 (eBioscience) in specific ELISA according to the manufacturer’s instructions.

Cells were stimulated as indicated in the figures, washed with PBS, and directly lysed in the plates with the lysis buffer containing 10 mM Tris (pH 7.4), 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 20 mM Na₄P₂O₇, 2 mM Na₃VO₄, 1% Triton X-100, 10% glycerol, 0.1% SDS, 0.5% deoxycholate, 1 mM PMSF, and protease inhibitor mixture (Roche). The lysate was vortexed at 10-min intervals and centrifuged after 30 min at 13,000 rpm for 10 min at 4°C. Thirty micrograms of protein was used with the Western immunoblotting protocol according to the p38 MAPK kit from Cell Signaling. For HO-1 detection, 50 μg of whole-cell extracts were processed similar to above and separated by SDS-PAGE, as we previously described (11). The membranes were incubated with anti-HO-1 Ab, followed by secondary HRP-labeled Abs. The HO-1-specific bands were visualized using ECL method (Pierce) and analyzed as described for EMSA using UVP system.

Total RNA was isolated at times indicated using the RNeasy kit from Qiagen according to the manufacturer’s instructions. Reverse transcriptions were performed using the First Strand cDNA Synthesis kit (Promega) according to the manufacturer’s instructions. One microgram of total RNA was transcribed to cDNA in a 20-μl reaction volume. For transcript quantification purposes by real-time PCR, the SYBR Green Mix containing Thermo-Start DNA Polymerase was used according to the manufacturer’s instructions (Eurogentech). Primers for HO-1 (forward, 5′-GCT TGT TGC GCT CAT TCT CC; and reverse, 5′-GCC ACC AAG GAG GTA CAC ATA) were from IDT. The 18S primers were purchased from Ambion. The PCR using 1 μl of cDNA was conducted in iCycler Thermal Cycler (Bio-Rad). A hot-start phase was applied at 95°C for 10 min for all primers. cDNA was amplified for 45 cycles (HO-1) at 95°C for 10 s, 60°C for 10 s, and 72°C for 25 s. At each cycle, accumulation of PCR products was detected by monitoring the increase in fluorescence by dsDNA-binding SYBR Green. After the PCR was performed, a dissociation/melting curve was constructed in the range of 55–95°C. Data were analyzed using the Bio-Rad ICycler software and comparative Ct method with the following formula: ΔCt = Ct (target, HO-1) − Ct (normalizer, 18S). Fold increase in the expression of HO-1 mRNA in experimental groups compared with medium control was calculated as 2^−(ΔΔCt).

HO-1 activity was measured as described by Tenhunen et al. (22) and modified by Balla et al. (23). Briefly, equal numbers of cells were stimulated as indicated, washed twice with PBS at room temperature, and then harvested from the plates in cold harvest buffer (0.1 M KPO₄, 20% glycerol, 1 mM EDTA (pH 7.4)), and then transferred into a 12 × 75-mm culture tube and sonicated for 3 s on ice. Protein concentration was determined and equal amounts of protein were subsequently used for each experimental group. The sonicated protein lysate was added to a NADPH-generating system containing 530 μl of K₂HPO₄ (0.1 M), 50 μl of Desferal (65 mg/ml), 50 μl of NAPDH (4 mM), and 0.8 mg of protein of rat liver cytosol prepared from 105,000 × g supernatant fraction as a source of biliverdin reductase, as previously reported by Tenhunen et al. and Balla et al. (22, 23). The samples were placed into a shaking water bath for 5 min. Hemin (1 mg) was added to the test tubes, and the reaction was conducted at room temperature for 10 min in the dark and terminated by adding 20 μl of p-hydroxymercuribenzoic acid (50 mM) to each tube. Hemin alone was included as reference. The absorption was measured at 470 nm using a quartz cuvette. The standard EmM of 1 M bilirubin (Bi) measured at 470 nm is 66 (22, 23). The HO-1 activity is calculated using the following formula, as previously reported (22, 23): picomolar Bi/minutes × milligrams of protein = (OD₄₇₀ × 1.010 final volume of the probe × 1000 pmol)/(0.066 OD/nanomoles of bilirubin/milliliter × 0.2-ml volume × 10 min × milligrams of protein/milliliter × 1 nM).

The statistical significance of the data was calculated using Wilcoxon signed-rank nonparametric method, ANOVA, and Student’s t tests, as indicated. A value of p < 0.05 was considered significant.

Based on observations that both ethanol treatment and the HO-1 product, carbon monoxide, augment IL-10 production in monocytes (3, 13), we sought to evaluate the role of HO-1 in ethanol-mediated regulation of IL-10 production. We first evaluated the influence of an HO-1 inhibitor, ZnPP, on IL-10 release in normal human monocytes after LPS or ethanol plus LPS stimulation. Maximal IL-10 production was seen at 40 h after stimulation. Consistent with our previous observation (3), ethanol augmented IL-10 induction in LPS-stimulated monocytes (p < 0.01) (Fig. 1,A). LPS- and LPS-plus-ethanol-induced IL-10 release were significantly abrogated in the presence of the HO-1 inhibitor, ZnPP (p < 0.01) (Fig. 1,A). The inhibitory effect of ZnPP on monocyte IL-10 production was dose dependent (Fig. 1,B). Inhibition of ethanol-induced IL-10 production was not limited to ZnPP because comparable doses of other well-established HO-1 inhibitors, including ZnPP, tin protoporphyrin IX (SnPP), SnMP, and chromium mesoporphyrin (CrMP) also prevented ethanol-mediated augmentation of LPS-induced monocyte IL-10 production (Fig. 1,C). These results suggested that HO-1 was involved in regulation of IL-10 by ethanol in monocytes. Consistent with the hypothesis that HO-1 is involved in IL-10 induction, CoPP, an activator of HO-1 enzyme activity, resulted in IL-10 induction in medium or ethanol-treated monocytes (Fig. 1,D). Similar to the effect of ethanol, CoPP significantly up-regulated LPS-induced monocyte IL-10 production (p < 0.01). The lack of further augmentation of IL-10 production in the presence of CoPP in ethanol-plus-LPS-treated monocytes likely indicated a maximal effect on the IL-10 induction by ethanol plus LPS through pathways involving up-regulation and function of HO-1 (Fig. 1 D).

To evaluate the biological effect of ethanol on HO-1, next, we tested the effect of ethanol on HO-1 activity in monocytes. HO-1 activity was significantly induced by LPS (p < 0.01) and it was further up-regulated in the presence of ethanol (p < 0.01) (Fig. 2,A). The HO-1 inhibitor, ZnPP, significantly reduced HO-1 activity both in LPS- (p < 0.04) and in LPS-plus-ethanol-stimulated cells (p < 0.01). The HO-1 activator, CoPP, induced HO-1 activity, which was augmented by LPS but not by additional ethanol stimulation. We also evaluated whether the ethanol-induced augmentation of HO-1 activity correlated with changes in HO-1 RNA and protein levels. Increased mRNA levels of HO-1 were induced at 4 h (Fig. 2,B) and 18 h (data not shown) after LPS stimulation and even greater induction was seen at both time points in LPS-plus-ethanol-stimulated cells. These LPS- and ethanol-induced increases in HO-1 mRNA were also present at the levels of HO-1 protein (Fig. 2,C). Previous reports suggested that the AP-1 nuclear regulatory factor plays a central role in HO-1 induction (8, 19, 24). We found that LPS, an inducer of HO-1, resulted in nuclear translocation and DNA binding of AP-1 in monocytes (Fig. 3,A). Although ethanol alone did not activate AP-1, LPS-induced AP-1 binding was augmented by ethanol treatment (p < 0.05) (Fig. 3, A and B). These results suggested that augmentation of AP-1 activation by acute ethanol might contribute to increased HO-1 induction.

Both HO-1 activation and acute ethanol inhibit TNF-α production (2, 11, 13). Thus, we examined whether HO-1 was involved in attenuation of LPS-induced TNF-α production by acute alcohol treatment. Consistent with our previous reports (2, 11), we found that acute ethanol down-regulated LPS-induced TNF-α production in human monocytes (p < 0.01) (Fig. 4). Inhibition of HO-1 activity with ZnPP significantly augmented TNF-α release both in LPS- and LPS-plus-ethanol-treated monocytes (Fig. 4), suggesting a role for HO-1 in the down-regulation of TNF-α production by ethanol. Higher TNF-α induction by LPS in the presence of HO-1 inhibition also suggested a role for HO-1 in TNF-α regulation. ZnPP alone had no effect on TNF-α production (data not shown).

To analyze the differential influence of alcohol on TNF-α and IL-10 production, we performed time course experiments (Fig. 5, A and B). We found low cytokine levels in the medium and alcohol-alone controls at any of the analyzed time points. There was minimal TNF-α in culture supernatants at early time points (1.5–3 h) and no difference between LPS- and LPS-plus-EtOH-stimulated samples (Fig. 5,A). Alcohol inhibited LPS-induced TNF-α starting at the peak of induction (9 h) and the effect was steady until 48 h, although the TNF-α levels were stepwise diminished in time. IL-10 showed a different kinetics (Fig. 5,B). The LPS and LPS plus EtOH induced low levels of IL-10 at 1.5–9 h. IL-10 production peaked at 24 h and was maintained at similar levels at 48 h. There was a significant increase of IL-10 in culture supernatants of LPS-plus-EtOH- compared with LPS-stimulated cells at 24 and 48 h. To identify signaling events that may contribute to ethanol-induced regulation of HO-1, we investigated p38 MAPK, a common element in regulation of HO-1, IL-10, and TNF-α production (25, 26, 27, 28, 29). The p38 MAPK inhibitor, SB203580, had no significant effect on monocytes treated with culture medium or ethanol (Fig. 5, C and D). We found significant inhibition of both TNF-α (Fig. 5,C) and IL-10 (D) production in LPS-treated monocytes in the presence of SB203580. Furthermore, alcohol-induced modulation of cytokine production was prevented in the presence of the p38 inhibitor, suggesting that alcohol may also act through p38 MAPK to regulate both TNF-α and IL-10 production. We next evaluated the effect of alcohol on p38 MAPK phosphorylation. Peak activation of the phosphorylated p38 protein was between 5 and 15 min after LPS stimulation (data not shown); thus, we tested phospho-p38 levels after 15 min stimulation. Alcohol alone induced no increase in p38 phosphorylation, but it significantly augmented LPS-induced phospho-p38 levels (p < 0.05) (Fig. 6,A). The MAPK inhibitor, SB203580, prevented p38 phosphorylation in LPS- and in LPS-plus-alcohol-treated monocytes (Fig. 6,A). This is in agreement with the previously published data that, whereas SB203580 does not act on the upstream MEKs, the p38 MAPK can be phosphorylated by TAB1-mediated autophopsphorylation and SB203580 can inhibit this process (〈www.biosource.com〉; product no. PHZ1253 (SB203580; 4-(4-Fluorophenyl)-2-(4-methylsulfinylphenyl)5-(4-pyridyl)1H-imidazole) analysis sheet) (30). Total p38 protein levels remained comparable between the different stimulation groups (Fig. 6,A). Average densitometry of five experiments showed the same stimulation pattern (Fig. 6,B). In contrast to p38 MAPK activation, there was no increase in phosphorylated ERK1/2 MAPK by ethanol in LPS-stimulated cells (data not shown). Next, we wished to assess whether p38 MAPK activation was involved in HO-1 activation by acute alcohol treatment. Monocytes were treated with 25 mM ethanol in the presence or absence of SB203580, a p38 MAPK inhibitor, and HO-1 activity was measured as described in Materials and Methods. Alcohol-induced HO-1 activity was substantially reduced in the presence of SB203580, suggesting the involvement of p38 MAPK kinase activity in regulation of HO-1 biological activity (Fig. 6 C). Cobalt chloride (CoCl₂), an HO-1 activator used as a positive control, induced HO-1 activity.

We have previously reported that the anti-inflammatory effects of acute ethanol involve inhibition of NF-κB nuclear translocation and DNA binding in LPS-stimulated monocytes (11, 31). Previous studies showed that the anti-inflammatory effect of HO-1 involved inhibition of NF-κB activation (13). To investigate whether the anti-inflammatory effect of acute alcohol on NF-κB could be mediated through HO-1, we preincubated human monocytes with the HO-1 activator, CoCl₂, or with the HO-1 inhibitors, SnMP or ZnPP, for 30 min, and then stimulated the cells with LPS and/or ethanol. Evaluation of nuclear proteins in the EMSA revealed NF-κB activation in LPS-stimulated monocytes (p < 0.005), which was significantly down-regulated by acute ethanol treatment (p < 0.01) (Fig. 7, A and B). Pretreatment of monocytes with the HO-1 activator, CoCl₂, also inhibited LPS-induced NF-κB activation (p < 0.05), suggesting that HO-1 activation inhibits NF-κB. To our surprise, inhibition of HO-1 activity with either SnMP or ZnPP failed to prevent ethanol-induced inhibition of NF-κB activation. Further investigations revealed that overall HO-1 activity at 1 h was lower compared with similar in vitro stimulation for 40 h (Fig. 2,A), and there was no difference in HO-1 activity between LPS- and LPS-plus-EtOH-stimulated samples at 1 h, when NF-κB activity was inhibited by alcohol (Fig. 7 C). Based on these results, we concluded ethanol influence on LPS-mediated cell activation at this early time point (1 h) might be HO-1 independent.

To validate the biological significance of our results with in vitro ethanol treatment in human monocytes, we tested the effects of in vivo administration of acute alcohol on LPS-induced cytokine responses in C57BL/6 mice. First, we tested IL-10 production in splenic macrophages collected after in vivo administration of LPS, ethanol, or their combination. LPS, a potent macrophage activator, induced significant IL-10 production (Fig. 8,A). IL-10 production in splenic macrophages was significantly higher in mice that received acute alcohol in combination with LPS. Administration of the HO-1 inhibitor, ZnPP, down-regulated IL-10 production induced by ethanol plus LPS, suggesting a role for HO-1 in induction of IL-10. Next, we tested serum cytokine levels in mice after in vivo administration of alcohol and/or LPS. A single dose of acute alcohol administration significantly augmented LPS-induced serum IL-10 levels, but alcohol alone did not induce serum IL-10 increase (Fig. 8,B). Administration of the HO-1 activator, CoPP, failed to further increase the LPS-plus-ethanol-induced serum IL-10 levels in mice (Fig. 8,B). This was similar to our findings in human monocytes (Fig. 1,D). Identical with its effects in human monocytes in vitro, acute ethanol administration in vivo attenuated LPS-induced TNF-α serum levels in mice (Fig. 8,A). Alcohol-induced suppression of serum TNF-α levels remained unchanged in mice treated with the HO-1 activator, CoPP (Fig. 8 C). These results suggested that the in vivo regulation of IL-10 and TNF-α production by acute alcohol involved HO-1.

Alcohol consumption is associated with alterations in immune responses including monocyte dysfunctions mostly due to ethanol-induced modifications in cytokine production and Ag presentation capacity (1, 32). Although acute and chronic alcohol use differently affect proinflammatory cytokine production, both have immunosuppressive effects (20, 31, 32, 33, 34). The modulatory potential of acute alcohol on cytokine expression includes attenuation of TNF-α and augmentation of IL-10 production in monocytes (2, 3, 11). Ethanol generates oxidative stress, which plays an important role in immune cell activation as well as in alcoholic liver injury (14, 35, 36). HO-1 is an important antioxidant molecule and its up-regulation is a critical cytoprotective mechanism activated during cellular stress (8, 37, 38, 39). In this report, we demonstrate that, in monocytes, modulation of IL-10 and TNF-α production by acute alcohol treatment involves HO-1 activation. Our data provide evidence for involvement of HO-1 in acute alcohol-induced up-regulation of IL-10 in LPS-stimulated monocytes. We found that HO-1 inhibitors prevented and HO-1 activators reproduced the effects of acute ethanol on monocyte IL-10 production. We further identified that the mechanism by which ethanol affected HO-1 activation and IL-10 production involved activation of the p38 MAPK (Fig. 9).

The IL-10 promoter is regulated through the p38 MAPK pathway in human macrophages (40). Our results demonstrated that acute ethanol treatment increased the levels of phosphorylated p38 in the same LPS-stimulated human monocytes that showed increased IL-10 and reduced TNF-α production. These effects of acute ethanol in our experiments were similar to the effects of CO, a product of HO-1 activation on monocytes (13). Consistent with the previously described role of p38 MAPK activation in IL-10 induction, inhibition of p38 activity with SB203580 prevented ethanol-induced augmentation of monocyte IL-10 production (28, 40). Activation of p38 MAPK has been shown to induce HO-1 expression in various cell types including macrophages (13), hepatocytes (24), pulmonary epithelial cells (41), and vascular cells (42). Our studies revealed that augmentation of p38 MAPK phosphorylation contributed to alcohol-induced HO-1 activation in LPS-stimulated monocytes. Recent studies demonstrated that transcriptional regulation of HO-1 gene expression involves the MAP kinases JNK and p38 as well as AP-1 activation in rat hepatocytes (24). We found that acute ethanol treatment increased multiple signaling elements in HO-1 gene activation including p38 phosphorylation and nuclear binding of the AP-1 transcription factor. It is important to note that HO-1 and IL-10 each can perpetuate the anti-inflammatory effects (38). The HO-1 product, CO, augments IL-10 production and IL-10 induces expression of HO-1 (13, 27, 38). Moreover, both HO-1-induced IL-10 production and IL-10-induced HO-1 activation are dependent on p38 MAPK activation (13, 26, 27, 38, 39). Thus, by increasing p38 MAPK phosphorylation, acute ethanol likely contributes to regulation of both IL-10 and HO-1 production in monocytes and thereby amplifies the anti-inflammatory effects of acute alcohol.

Our results show that acute ethanol significantly augmented HO-1 enzyme activity. The mechanism by which HO-1 mediates cytoprotective functions is yet to be fully explored. However, the three major catalytic products, CO, ferritin, and bilirubin, may represent potential targets. CO up-regulates IL-10 and down-regulates TNF-α in activated murine macrophages (13). Our results showed that acute alcohol had similar effects on LPS-stimulated monocytes and, furthermore, inhibition of HO-1 activity could prevent the modulatory effects of acute alcohol on IL-10 and TNF-α production. Most important, we found that acute ethanol increased LPS-induced HO-1 enzyme activity. Of the three catalytic products of HO-1, both CO and bilirubin are potential candidates for mediation of the anti-inflammatory effects of alcohol. Carbon monoxide and ethanol share several similar effects on inflammation, immunoregulation, and monocytes (3, 13). In addition to acute ethanol and CO inhibiting TNF-α and inducing IL-10, both CO and ethanol can suppress T cell proliferation and inhibit IL-2 production (43, 44, 45). Thus, the HO-1-mediated immunomodulatory effects of acute alcohol may expand beyond its anti-inflammatory effects. Indeed, a recent report found that HO-1 expression inhibited dendritic cell maturation but conserved IL-10 expression (46). This is similar to the alcohol-induced inhibition of monocyte-derived dendritic cell differentiation and increased IL-10 production we reported earlier (42). Bilirubin, another product of HO-1 enzyme activity, has recently been shown to exert anti-inflammatory properties and to inhibit NF-κB activation in T lymphocytes (46, 47).

HO-1 activation in hepatocytes and in Kupffer cells, the resident macrophages of the liver, plays a protective role in different forms of acute liver injury. For example, HO-1 activation with hemin or administration of biliverdin, a product of HO-1 enzyme activity, was protective against acetaminophen-induced liver injury (48). HO-1 gene transfer also prevented CD95/FasL-mediated apoptosis in liver allografts (49). The cytoprotective effect of HO-1 activation was also established in liver cirrhosis. In the bile duct ligation model of cirrhosis, HO-1 activation was found in Kupffer cells as well as in hepatocytes in early stages of cirrhosis, whereas it dropped in later stages of cirrhosis, suggesting a protective role for HO-1 (50). HO-1 seems to have a central role in hepatic response to oxidative stress (37). Thus, it is not surprising that HO-1 activation plays a role in intracellular protection during alcohol-induced insults. Our results show that, in human monocytes, which can be precursors of tissue macrophages, HO-1 activation mediates some of the anti-inflammatory effects of acute ethanol treatment. However, the regulation of HO-1 in increased proinflammatory cytokine activation by chronic alcohol is yet to be explored.

Both acute and chronic alcohol administration result in oxidative stress. Alcohol stimulation in vitro as used in our experiments is a model for acute, moderate alcohol consumption. The 25 mM in vitro alcohol used in our experiments is equivalent to 0.1 g/dl blood alcohol concentration reached after consumption of three to four drinks in humans. Moderate alcohol consumption, such as one to two drinks per day, is associated with beneficial effects on cardiovascular disease and overall mortality (51). Although the mechanisms for the cardioprotective effects of alcohol are yet to be elucidated, inflammatory cell involvement in atherosclerosis is a potential target (51, 52). Recent reports showed that HO-1 inhibits atherosclerotic lesion formation in LDL-receptor knockout mice (53). Furthermore, IL-10 is associated with cardioprotective effects (54). Also, the levels of bilirubin in the normal human population correlate with the incidence of atherosclerotic events (55). Multiple studies suggest that atherosclerosis is also attenuated by acute, moderate alcohol use (51). Here, we found that acute alcohol inhibited TNF-α, a key proinflammatory cytokine in macrophages activation in the atherosclerotic plaques (52). Our studies suggested that inhibition of TNF-α production by monocytes involved both HO-1-dependent and -independent mechanisms. Alcohol appeared to inhibit TNF-α induction at different levels. First, we found that NF-κB, a nuclear regulatory factor involved in TNF-α gene activation, was inhibited by acute alcohol independent of HO-1 activation suggesting an early, HO-1-independent effect of alcohol. Second, our data using HO-1 inhibitors demonstrated that HO-1 activation was involved in inhibition of monocyte TNF-α production by acute alcohol. Third, as we previously reported, alcohol-induced IL-10 contributes to down-regulation of the late phase of TNF-α production, and our current findings suggest that this is mediated by alcohol-induced HO-1 activation (3). Thus, the biological significance of the anti-inflammatory effects of acute ethanol on HO-1 and IL-10 induction may reach beyond immunity and deserves further investigations.

The authors have no financial conflict of interest.