Moderate alcohol induces stress proteins HSF1 and hsp70 and inhibits pro-inflammatory cytokines resulting in endotoxin tolerance

Abstract

Binge or moderate alcohol exposure impairs host defense and increases susceptibility to infection due to compromised innate immune responses. However there is a lack of consensus on the molecular mechanism by which alcohol mediates this immunosuppression. Here, we show that cellular stress proteins HSF1 and hsp70 play a mechanistic role in alcohol-mediated inhibition of the TLR4/MyD88 pathway. Alcohol exposure induced transcription factor HSF1 mRNA expression and DNA binding activity in primary human monocytes and murine macrophages. Furthermore HSF1 target gene hsp70 mRNA and protein are upregulated by alcohol in monocytes. In vitro pre-exposure of moderate alcohol reduced subsequent LPS-induced NF-κB promoter activity and downstream TNFα, IL-6 and IL-1β production, exhibiting endotoxin tolerance. Mechanistic analysis demonstrates that alcohol induced HSF1 binds to the TNFα promoter in macrophages at early timepoints exerting transrepression and decreased TNFα expression. Furthermore, association of hsp70 with NF-κB subunit p50 in alcoholtreated macrophages correlates with reduced NF-κB activation at later timepoints. Hsp70 overexpression in macrophages was sufficient to block LPS-induced NF-κB promoter activity suggesting alcoholmediated immunosuppression by hsp70. The direct crosstalk of hsp70 and HSF1 was further confirmed by the loss of alcohol-mediated endotoxin tolerance in hsp70 and HSF1 silenced macrophages. Altogether, our data suggest that alcohol-mediated activation of HSF1 and induction of hsp70 inhibits TLR4-MyD88 signaling and are required for alcohol-induced endotoxin tolerance. Using stress proteins as direct drug targets would be clinically relevant in alcohol abuse patients and may serve to provide a better understanding of alcohol-mediated immunosuppression.

INTRODUCTION

Monocytes and macrophages present the first line of host defense and are pivotal for activation of innate immune responses (1). During infection, pathogen recognition by Toll-like receptors (TLRs) and downstream TLR signaling play an important role in activation and function of innate immune cells. Upon stimulation with TLR4 ligand, lipopolysaccharide (LPS) or endotoxin, downstream MyD88 dependent signaling results in activation of NF-κB-mediated transcription of pro-inflammatory cytokines such as TNFα, IL-1β and IL-6 (2). These innate immune responses are compromised by moderate or binge alcohol exposure increasing host susceptibility to infection, illustrating clinical relevance (3, 4). For example, binge alcohol exposed macrophages and monocytes showed reduced antigen-presenting function, decreased production of pro-inflammatory cytokines TNFα, IL-6, IL-8 and IL-12 and inhibition of NF-κB activation (5-9). Binge alcohol is also shown to induce negative regulators of TLR4 such as interleukin-1 receptor associated kinase-M (IRAK-M) (10) and B-cell lymphoma 3-encoded protein (Bcl-3) (11) likely contributing to endotoxin tolerance (12, 13). However, alcohol induced cellular stress responses in innate immune cells and their contribution to immune suppression by alcohol has received no attention. Here, we identify novel mechanisms of alcohol-mediated endotoxin tolerance, linking cellular stress pathways to innate immune signaling.

Alcohol-mediated induction of cellular stress, particularly oxidative stress by reactive oxygen species (ROS), has been previously demonstrated in heart, lung and liver tissues (14-16). Cellular stress proteins can crosstalk with immune signaling pathways and influence immune responses (17). Amongst various cellular stress proteins, heat shock proteins are likely candidates for mediating alcohol-induced immune tolerance based on their widely characterized anti-inflammatory role in innate immune signaling. For instance, heat shock protein 70 (hsp70, also known as hsp72)³, induced by heat shock, inhibits LPS-induced production of cytokines such as TNFα by macrophages (18, 19). Expression of hsp70 is regulated by its transcription factor heat shock factor 1 (HSF1)⁴ (20), which can also exert anti-inflammatory effects independent of hsp70 (21-23). These immunosuppressive roles of stress-induced HSF1 and hsp70 lend support to our hypothesis that alcohol causes endotoxin tolerance and inhibition of downstream signaling via induction of hsp70 and/or HSF1. Identifying this alcohol-mediated mechanism by which stress proteins crosstalk with inflammatory responses may provide novel mechanisms and identify therapeutic targets for restoration of normal immune functions in binge-drinking patients and also trauma patients who abuse alcohol.

Previous studies from our laboratory show that moderate alcohol activates HSF1 and induces hsp70 in murine macrophages (24). The aim of this study was to delineate the mechanistic role of hsp70 and HSF1 in alcohol-induced endotoxin tolerance. Here, using an in vitro model of alcohol pre-exposure, we provide evidence that alcohol mediates activation of HSF1 and induction of hsp70 in human monocytes and murine macrophages. Alcohol-induced hsp70 bound to p50 subunit of NF-κB and HSF1 directly interacted with cytokine gene promoter exerting negative effects on TLR4-induced pro-inflammatory cytokine production. HSF1 and hsp70 were not only required but sufficient for alcohol-mediated induction of endotoxin tolerance. Collectively, these results uncover a crosstalk mechanism between alcohol induced stress proteins hsp70 and HSF1 and TLR4 signaling molecules resulting in endotoxin tolerance.

MATERIALS AND METHODS

Human subjects and cell lines

Healthy individuals aged 18 to 60, both females and males with no previous alcohol abuse history who consumed less than 6 drinks/week were recruited in the study and all abstained from alcohol for 48-72 hours prior to study. This study was reviewed and approved by University of Massachusetts Institutional Review Board and Department of Defense Human Research Protection Office at the Clinical Trials Unit at University of Massachusetts Medical Center. Mononuclear cells were isolated using Ficoll gradient from peripheral blood collected from the healthy human donors. Human monocytes from peripheral blood mononuclear cells were isolated by selective adherence as previously described (25). The adherent monocytes were cultured in Iscove’s Modified Dulbecco’s Medium (Invitrogen Life Technologies) with 10% FBS (HyClone).

RAW 264.7 murine macrophages were purchased from American Type Culture Collection and maintained in Dulbecco’s Modified Eagle Medium (Invitrogen Life Technologies) containing 10% FBS (HyClone).

Alcohol exposures and cell stimulations

Monocytes or macrophages plated at 1×10⁶ cells/ml were stimulated with LPS (Sigma-Aldrich) derived from Escherichia coli (100 ng/ml) and ethanol at 25 mM or 50 mM concentrations as indicated in the figure legends. Notably, we exposed human monocytes or macrophages to alcohol prior to subsequent endotoxin stimuli to test whether alcohol can tolerize innate immune cells. This alcohol “pre-exposure” closely resembles the in vivo situation where increased correlation has been demonstrated between high blood alcohol content and subsequent infectious complication in trauma patients (26). The 25 mM in vitro alcohol concentration approximates a blood alcohol level of 0.1 g/dl, which is above the legal limit of 0.08 g/dl blood alcohol concentration. Cell viability was not significantly affected by LPS or alcohol treatments (6). For positive induction of HSF1 and hsp70, cells were heat-shocked at 42°C for 45 minutes. Cells treated with low doses of LPS (10 ng/ml) for 24 hours prior to LPS stimulation served as a positive control for endotoxin tolerance.

RNA analysis and qPCR

Total RNA was isolated from the monocytes/macrophages using the RNeasy Mini column purification kit (Qiagen) according to manufacturer’s instructions. RNA was quantified by spectrophotometric analysis and quality of RNA was verified by measurement of OD 260/280 ratio. cDNA was synthesized using the Reverse Transcription system (Promega) according to the manufacturer’s instructions. The reaction mixture for qPCR contained 6.25 μL iTAQ universal SYBR Green PCR master mix (Bio-Rad Laboratories), 0.25 μM each forward and reverse primers, and 2.5 μL cDNA (corresponding to 25 ng RNA) for a total reaction volume of 12.5 μL. Real-time quantitative PCR was performed using the CFX96 real-time detection system (Bio-Rad Laboratories). The ΔCt of the gene was normalized to the ΔCt of housekeeping gene 18S and fold change of mRNA was expressed relative to untreated cells. Human TNFα, IL-6, IL-1β primers and mouse TNFα, IL-6, IL-1β, hsp70, HSF1 and 18S primer pairs were synthesized by IDT Inc. and enumerated in Table I. Human hsp70 and HSF1 primer pairs were purchased from SABiosciences.

Table I.

Sequences of primers used for qPCR

Gene	Primer Sequence
hTNFα	Fwd, 5’-ATCTTCTCGAACCCCGAGTGA-3’
	Rev, 5’-CGGTTCAGCCACTGGAGCT-3’
hIL-6	Fwd, 5’-CGAGCCCACCGGGAACGAAA-3’
	Rev, 5’-GGACCGAAGGCGCTTGTGGAG-3’
hIL-1β	Fwd, 5’-CACGCTCCGGGACTCACAGC-3’
	Rev, 5’-GGAGAACACCACTTGTTGCTCCA-3’
mTNFα	Fwd, 5’-GAAGTTCCCAAATGGCCTCC-3’
	Rev, 5’-GTGAGGGTCTGGGCCATAGA-3’
mIL-6	Fwd, 5’ ACAACCACGGCCTTCCCTACTT-3’
	Rev, 5’-CACGATTTCCCAGAGAACATGTG-3’
mIL-1β	Fwd, 5’-CAGGCAGGCAGTATCACTCA-3’
	Rev, 5’-TGTCCTCATCCTGGAAGGTC-3’
mHsp70	Fwd, 5’-AACTACAAGGGCGAGAACCGGTC-3’
	Rev, 5’-GATGATCCGCAGCACGTTCAGA-3’
mHSF1	Fwd, 5’-ACTCCAACCTGGACAACCTG-3’
	Rev, 5’-GGAGGCTCTTGTGGAGACAG-3’
18S	Fwd, 5’-GTAACCCGTTGAACCCCATT-3’
	Rev, 5’-CCATCCAATCGGTAGTAGCG-3’

Enzyme-linked immunosorbent assay (ELISA)

Cell-free supernatants were collected from human monocyte or RAW macrophage cultures and analyzed for human TNFα (Thermoscientific), IL-6 and IL-1β (BD Biosciences) or murine TNFα (BD Biosciences), IL-6 (Biolegend) and IL-1β (R&D Systems) according to the manufacturer’s instructions.

Nuclear lysates prepared from RAW macrophage (25 μg) were assayed for p-S326 HSF1 and total HSF1 using ELISA kits from Enzo Life Sciences according to manufacturer’s instructions.

Preparation of nuclear, cytoplasmic and whole cell extracts

Nuclear and cytoplasmic extracts from macrophages were prepared as previously described (27). Briefly, at the end of the stimulation period, cells collected in ice-cold PBS were resuspended in cold hypotonic buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 1 mM PMSF, and 10 μg/ml protease inhibitors such as aprotinin, antipain, and leupeptin) (Sigma-Aldrich) and incubated on ice for 20 min. Cells were then lysed with 0.6% Nonidet P-40 by vortexing for 20 seconds. The lysate was centrifuged at 12,000 g for 1 minute to pellet the nuclei, and the supernatant was stored at −80°C as the cytoplasmic extract. The nuclear pellet was then resuspended in ice-cold buffer B (20 mM HEPES (pH 7.9), 400 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM PMSF, and 20% glycerol). All tubes were kept on a shaker at 4°C for 30 min. The lysate was then centrifuged at 12,000 g for 10 min, and the supernatant was stored at −80°C as the nuclear extract.

Monocytes/macrophages were resuspended in lysis buffer (10% glycerol, 1% Triton, 20 mM Tris, pH7.6, 150mM NaCl, 25 mM β-glycerophosphate, 50 mM NaF, 1 mM Na-orthovanadate, 1 mM EDTA and 1 mM DTT with protease inhibitors) and incubated for 20 minutes on ice. The lysates were then centrifuged at 12,000 g for 10 min and the supernatant was stored as whole cell lysates.

Protein content was determined in the whole cell, cytoplasmic or nuclear extracts by the Bio-Rad protein assay dye reagent (Bio-Rad Laboratories).

Immunoprecipitation

For immunoprecipitations, samples were precleared with 50 μl of TrueBlot anti-rabbit IgG immunoprecipitation beads (eBioscience) for 1 h, and the precleared samples were incubated with 5 μg of anti-p50 Ab (Santa Cruz Biotech) overnight at 4°C. The next day, 50 μl of the TrueBlot anti-rabbit IgG immunoprecipitation beads were added to each sample for 1 h. The beads were washed three times with lysing buffer (50 mM Tris, pH 8, 150 mM NaCl, 1% NP-40, 1 mM EDTA, 1 mM PMSF, 1 mM NaF, 1 mM Na-orthovanadate with protease inhibitors) and then eluted with sample buffer. These immunoprecipitated samples were then used for immunoblotting.

Immunoblotting (Western blotting)

Heat denatured whole cell, cytoplasmic or nuclear lysates (5-25 μg) were loaded into each well, separated on 10% SDS-polyacrylamide gel and electroblotted onto nitrocellulose membranes. Nonspecific binding was blocked by incubation of the membranes in TBS/5% nonfat dried milk/0.1% Tween 20 or TBS/5% BSA/0.1% Tween 20 followed by the antibodies indicated in each experiment. The mouse antibodies against hsp70 (HSPA1A), hsp90β (HSP90AB1; constitutive form) and β-actin were purchased from StressMarq Biosciences and Abcam respectively. The rabbit antibodies against p65, p50 were from Cell Signaling Technology and against TBP-1 from Santa Cruz Biotechnology. The antibodies were detected using HRP-conjugated anti-mouse and anti-rabbit secondary antibodies (Santa Cruz Biotechnology and Abcam) and chemiluminescence assay reagents from Bio-Rad. The immunoreactive bands were quantified by densitometric analysis using an UVP System (Bio-Rad Laboratories).

Chromatin immunoprecipitation (ChIP)⁵

Chromatin immunoprecipitation was carried out using the Chromatrap ChIP assay kit according to the manufacturer’s instruction. Briefly, 1×10⁶ RAW macrophages were treated with LPS or ethanol for indicated times, fixed with 1% of formaldehyde and resuspended in hypotonic buffer before subjecting to chromatin shearing by sonication for 6 cycles of 10 seconds at 80% amplitude. Shearing efficiency was confirmed by agarose gel electrophoresis. After setting aside input chromatin, an aliquot of chromatin samples were incubated on a shaker at 4°C for 3 hours with antibodies against HSF1 (Enzo life sciences) or IgG control antibodies as negative control in low salt buffer. The samples were then precipitated by binding to protein A-agarose columns. After reversing the chromatin cross links at the final step, PCR was performed with purified chromatin DNA (immunoprecipitated and input) using TNFα and hsp70 gene promoter specific primers (TNFα forward, 5’-AGCGAGGACAGCAAGGGA-3’; reverse, 5’-TCTTTTCTGGAGGGAGTGTGG-3’; hsp70 forward, 5’- AACTCCGATTACTCAAGGGAGGC-3’; reverse, 5’-GATTCTGAGTAGCTGTCAGCG-3’). The PCR products were run on 1% agarose gel and bands were quantified using UVP System based densitometry analysis software (Bio-Rad Laboratories). HSF1-ChIP PCR data was analyzed relative to input and expressed as percentage of input.

Transient transfection of plasmids and siRNA

The hsp70-CMV5 overexpression (pCMV5-hsp70) and 4x NF-κB tandem promoter driven-luciferase reporter (p(κB)₄-luc) plasmids were kind gifts from Dr. R. Morimoto (Northwestern University, Chicago, IL, USA) and Dr. N. Mackman (Scripps Research Institute, La Jolla, CA, USA) respectively (28, 29). RAW macrophages were transfected with 2 μg of plasmid at 1:3 DNA:lipofectamine ratio in Opti-MEM for 6 hours using Lipofectamine 2000 (Invitrogen). For knockdown experiments, RAW macrophages were transfected with 10 nM hsp70 siRNA (from Invitrogen Stealth Library) or HSF1 siRNA (30) (synthesized by Invitrogen) or negative control in Opti-MEM for 24 hours using Lipofectamine 2000 as previously shown (31). siRNA knockdown efficiency was confirmed by qPCR for hsp70 or HSF1. Twenty-four hours after transfection, cells were treated with ethanol or LPS for appropriate experiments.

Luciferase activity assay

RAW macrophages were transfected with 0.5 μg p(κB)₄-luc alone or in combination with varying amounts of pCMV5-hsp70 using transfection reagent Lipofectamine 2000 (Invitrogen) at a DNA:lipofectamine ratio of 1:3. Twenty four hours after transfection, cells were treated with ethanol or LPS for indicated times and luciferase activity was assessed with Dual Luciferase reporter assay (Promega), according to the manufacturer’s instructions. Firefly and Renilla luciferase activities were determined using a Promega Glomax 96 Luminometer. NF-κB promoter-driven transcriptional activity, as detected by Firefly luciferase activity, was normalized with Renilla luciferase activity.

Electrophoretic mobility shift assay (EMSA)

A double-stranded HSE (5′GCCTCGAATGTTCGCGAAGTT3′) consensus sequence was used for EMSA (24). End-labeling was accomplished by treatment with T4 polynucleotide kinase in the presence of γ³²P-ATP (Dupont-NEN). Labeled oligonucleotide was purified on a polyacrylamide copolymer column (Bio-Rad Laboratories). Nuclear protein (5 μg) was added to a binding reaction mixture containing 20 mM HEPES (pH 7.9), 50 mM KCl, 0.1 mM EDTA, 1 mM DTT, 5% glycerol, 200 μg/ml BSA, 2 μg polydeoxyinosinic:polydeoxycytidylic acid. Samples were incubated at room temperature for 20 min followed by incubation with 50,000 cpm γ³²P-labeled HSE oligonucleotide for 10 minutes. For the cold competition reaction, a 20-fold excess of a specific, unlabeled, double-stranded probe was added to the reaction mixture 30 minutes before adding the labeled oligonucleotide. All reactions were run on a 4% polyacrylamide gel, and the dried gel was exposed to X-ray film at −80°C.

Statistical analysis

Results are presented as mean ± SD. Student’s t test was used to determine the statistical significance of differences between samples. Values of p < 0.05 were considered to represent statistical significant differences.

RESULTS

Pretreatment of human monocytes with a single dose of alcohol induces endotoxin tolerance and inhibits pro-inflammatory cytokine production

Several studies have shown that binge or short-term alcohol exposure (upto 24 hours) can suppress innate immune responses (3, 4). In this study, we wanted to first establish the in vitro model of alcohol pre-exposure followed by subsequent endotoxin stimulation in monocytes. We conducted a comprehensive analysis of the effect of multiple in vitro alcohol pre-exposure timepoints on expression of TLR4-induced MyD88 dependent pro-inflammatory cytokines TNFα, IL-6 and IL-1β in primary human monocytes. Figure 1 shows that treatment of monocytes with 100 ng/ml LPS for 2 hours caused a significant induction of TNFα (Figure 1A), IL-1β (Figure 1B) and IL-6 (Figure 1C) mRNA expression, respectively. However, this induction was significantly decreased in monocytes pre-exposed to 25 mM alcohol for 3-24 hours before LPS stimulation. It should be noted that we used a physiologically relevant concentration of 25 mM alcohol (approximates 0.1 g/dl blood alcohol concentration) and alcohol alone did not induce any of these cytokines in monocytes. Alcohol pre-exposure and LPS treatment also did not affect cell viability (data not shown). Monocytes pretreated with low doses of LPS for 24 hours prior to 100 ng/ml LPS stimulation showed inhibition of cytokine mRNA and served as a positive control for LPS/endotoxin tolerance (32). ELISA data corroborated these changes in cytokine mRNA levels. We observed inhibition of LPS-induced TNFα (Figure 2A), IL-1β (Figure 2B) and IL-6 (Figure 2C) in 3-24 hours alcohol pre-exposure in human monocytes consistent with the mRNA results. Notably, the inhibitory effect of alcohol on LPS-induced IL-1β observed even at one hour after alcohol pre-exposure was statistically significant despite no change in IL-1β mRNA, suggesting that alcohol likely also regulates IL-1β expression at the post-transcriptional level, in addition to inhibition of mRNA transcription. Thus our results show that pre-exposure of human monocytes to physiologically relevant concentrations of alcohol caused inhibition of LPS/TLR4 stimulated pro-inflammatory cytokine production at the mRNA and protein levels.

Figure 1. Pre-exposure of human monocytes to alcohol results in decreased levels of TLR4- induced pro-inflammatory cytokine mRNA.

(A-C) Adherence isolated human monocytes were treated with 100 ng/ml LPS or 25 mM alcohol (Et) alone for 2 hours or pre-exposed to 25 mM alcohol for 1, 3, 5, 7 or 24 hours followed by LPS stimulation for 2 hours. Levels of TNFα (A), IL-1β (B) and IL-6 (C) mRNA were analyzed by qPCR. Fold change in expression of genes was calculated with respect to untreated. LPS tolerance control: 10 ng/ml LPS for 24 hrs followed by LPS (100 ng/ml). Data summarize mean ± SD of 13 independent experiments. (* p<0.05, ** p<0.01, *** p<0.005, ns not significant)

Figure 2. Pre-exposure of human monocytes to alcohol inhibits TLR4-induced pro-inflammatory cytokine protein levels.

(A-C) Adherence isolated human monocytes were treated with 100 ng/ml LPS or 25 mM alcohol (Et) alone or pre-exposed to 25 mM alcohol for 1, 3, 5, 7 or 24 hours followed by LPS stimulation for 18-22 hours (overnight). Secreted TNFα (A), IL-1β (B) and IL-6 (C) protein in overnight culture supernatants was analyzed by ELISA. LPS tolerance control: 10 ng/ml LPS for 24 hrs followed by LPS (100 ng/ml). Data summarize mean ± SD of 13 independent experiments. (* p<0.05, ** p<0.01, *** p<0.005, ns not significant)

Alcohol induces HSF1 expression and activation in monocytes and macrophages

The role of cellular stress proteins in TLR4 signaling and tolerance is reported (17). We have previously shown that alcohol and LPS added together activated HSF1 in human monocytes (24). To delineate the role of stress-induced HSF1 in alcohol-mediated immune suppression, we first determined the effect of alcohol pretreatment for 2-24 hours on expression and activity of HSF1 in human monocytes and RAW264.7 murine macrophages. Alcohol exposure significantly induced HSF1 mRNA in monocytes (Figure 3A) and RAW macrophages (Figure 3B) at 14-24 hours compared to untreated cells. Upregulation of HSF1 mRNA in RAW macrophages occurred at 50 mM alcohol concentration (Figure 3B). This increase in HSF1 mRNA was not affected by subsequent LPS stimulation (Figure 3C). Next, RAW macrophages were pre-exposed to alcohol followed by LPS to determine HSF1 DNA binding activity measured by EMSA. As shown in Figure 3D, 50 mM alcohol alone for one hour induced higher HSF1 DNA-binding activity compared to untreated cells, consistent with our previously published results in human monocytes (24). Cells pretreated with 50 mM alcohol for one hour prior to LPS stimulation also showed comparable levels of HSF1 DNA-binding activity as cells treated with alcohol alone but higher than LPS stimulated macrophages, indicating subsequent LPS stimulation did not further augment alcohol-induced HSF1 activity. Since chaperone protein, hsp90β or HSP90AB1, is known to sequester HSF1 in the cytoplasm (33), we examined cytoplasmic hsp90β and observed no significant change in alcohol treated cells compared to untreated cells suggesting that alcohol-induced HSF1 activation is independent of hsp90β levels (Figure 3E). HSF1 undergoes post-translational modifications such as activating Ser326 phosphorylation by heat-shock induced mammalian target of rapamycin (mTOR)⁶ kinase which affects DNA binding activity (34). To determine whether alcohol regulates HSF1 DNA binding activity via phosphorylation, we analyzed HSF1 phosphorylation at Ser326 in alcohol-treated cells using a plate-based ELISA. While heat shocked cells showed significant induction of phospho-S326 HSF1 as expected, we did not observe phospho-S326 HSF1 levels in the presence of alcohol alone or with alcohol pretreatment followed by LPS stimulation (Figure 3F). This suggests that alcohol regulates HSF1 activity independent of activating phosphorylation of S326 on HSF1 or hsp90β levels.

Figure 3. Alcohol induces HSF1 expression and activation in human monocytes and RAW macrophages.

(A) Human monocytes isolated by adherence were exposed to 25 mM alcohol (Et) for 2-24 hours and total RNA collected at different timepoints were analyzed for HSF1 mRNA by qPCR. Graph depicts mean ± SD of 7 independent experiments. (B) RAW macrophages were exposed to 25 or 50 mM alcohol (Et) for 2-24 hours and total RNA collected at different timepoints were analyzed for HSF1 mRNA by qPCR. Data summarize mean ± SD of 3 independent experiments. (C) Adherence isolated human monocytes were treated with 100 ng/ml LPS, 25 mM alcohol (Et) alone or pre-exposed to 25 mM alcohol followed by LPS stimulation for indicated timepoints. HS denotes heat shocked control: 42°C for 45 minutes. Total RNA was subjected to HSF1 mRNA determination by qPCR. Data summarizes mean ± SD of 7 independent experiments. (D-F) RAW macrophages were stimulated with 100 ng/ml LPS for 30 minutes, 25 or 50 mM alcohol (Et) alone for 1 or 24 hours or pre-exposed to 25 or 50 mM alcohol for 1 or 24 hours followed by LPS stimulation for 30 minutes. HS denotes heat shocked control: 42°C for 45 minutes. (D) HSF1 DNA-binding activity was detected in nuclear extracts by EMSA using a ³²P-labeled, double-stranded HSE oligonucleotide. A representative experiment is depicted and graph summarizes mean ± SD of 4 independent experiments. A 20-fold excess of unlabeled oligonucleotide added to the heat shocked sample to confirm specificity of HSF1 binding was included as competition control (Comp). (E) Expression of hsp90β in the cytoplasmic extracts was assayed by western blotting. A representative experiment is depicted and graph summarizes mean ± SD of 4 independent experiments. (F) Phosphorylated HSF1 (phosphoserine 326) was detected in nuclear lysates by ELISA and normalized to total HSF1. Graph depicts mean ± SD of 4 independent experiments. (* p<0.05, ** p<0.01, ns not significant)

Alcohol induced hsp70 expression is mediated by HSF1 in monocytes and macrophages

Stress mediated activation of HSF1 induces target genes such as hsp70 (20). Here, we examined the effect of alcohol on HSF1 target gene hsp70 expression. Human monocytes treated with 25 mM alcohol showed significant upregulation of hsp70 mRNA at 14-24 hours compared to untreated cells (Figure 4A). As shown in Figure 4B, upregulation of hsp70 mRNA observed in cells pretreated with alcohol for 24 hours was not further augmented by subsequent LPS stimulation but was significantly higher than LPS-treated cells. Heat-shocked monocytes showed significantly high expression of hsp70 mRNA and served as a positive control (Figure 4B). Next, Figure 4C illustrates that monocytes exposed to alcohol for 24 hours exhibited significant induction of hsp70 protein as determined by immunoblotting. Pre-exposure of alcohol followed by LPS did not further increase hsp70 in monocytes compared to alcohol alone, indicating that upregulation of hsp70 is mediated solely by alcohol. Therefore, similar to heat stress, alcohol can upregulate hsp70 expression both at the mRNA and protein level in human monocytes independent of TLR4 stimulation.

Figure 4. Alcohol induces HSF1-mediated hsp70 expression in human monocytes and RAW macrophages.

(A) Human monocytes isolated by adherence were exposed to 25 mM alcohol (Et) for 2-24 hours and total RNA collected at different timepoints were analyzed for hsp70 mRNA by qPCR. Graph depicts mean ± SD of 7 independent experiments. (B and C) Adherence isolated human monocytes were treated with 100 ng/ml LPS, 25 mM alcohol (Et) alone or pre-exposed to 25 mM alcohol followed by LPS stimulation for indicated timepoints. HS denotes heat shocked control: 42°C for 45 minutes. (B) Total RNA was subjected to hsp70 mRNA determination by qPCR. Data summarizes mean ± SD of 7 independent experiments. (C) hsp70 protein was detected in whole cell lysates by western blotting. The densitometry graph represents quantitation of bands seen in the gel and depicts mean ± SD (n=9). Representative gels are shown with hsp70 (top) and loading control, β-actin (bottom). (D) RAW macrophages were stimulated with 100 ng/ml LPS or 50 mM alcohol (Et) for 1 hour or pre-exposed to alcohol for 1 hour followed by LPS stimulation for 1 hour. HS denotes heat shocked control: 42°C for 45 minutes. Chromatin immunoprecipitation assay was performed using anti-HSF1 antibody and semi-quantitative PCR was carried out using hsp70 promoter specific primers. A representative gel picture is shown above the densitometry graph, which represents quantitation of bands seen in the gel (n=4). Input DNA is shown to ensure equal amount of the sheared DNA. (* p<0.05, ** p<0.01, *** p<0.005, ns not significant)

Previous studies in our laboratory have shown that alcohol exposure upregulates HSF1 directed hsp70 promoter driven transcriptional activity in macrophages (24). As illustrated in Figure 4D, here we show using chromatin immunoprecipitation (ChIP) analysis that binding of HSF1 to the hsp70 promoter in RAW macrophages is significantly increased upon treatment with alcohol alone for one hour. Interestingly one hour alcohol pre-exposure followed by LPS treatment for an hour reduced HSF1 binding to the hsp70 promoter which was still higher than in LPS stimulated cells alone. Therefore, alcohol-mediated HSF1 activation mediates induction of hsp70 by direct binding to its promoter region in monocytes and macrophages.

Alcohol induced HSF1 and hsp70 mediates LPS-induced NF-κB inhibition in macrophages

Previous studies demonstrate decreased DNA-binding activity and reduced nuclear translocation of NF-κB in cells exposed to alcohol and LPS together (27). To implicate alcohol-mediated anti-inflammatory proteins HSF1 and hsp70 in TLR4-mediated NF-κB activation, we tested whether alcohol pre-exposure affects nuclear translocation of LPS-induced p65 and p50 NF-κB subunits in RAW macrophages. Nuclear translocation of NF-κB subunits p65, containing the transactivation domain, and p50, for DNA binding, is necessary for induction of expression of pro-inflammatory cytokines such as TNFα, IL-6 and IL-1β (2). As expected, LPS stimulation increased nuclear levels of p65 in macrophages (Figure 5A). Macrophages pre-exposed to alcohol for 24 hours, but not one hour, had significantly reduced LPS stimulated nuclear p65 compared to cells stimulated with LPS alone. Furthermore, nuclear expression of the p50 subunit of NF-κB was upregulated in LPS stimulated macrophages which was significantly inhibited by 24 hours of alcohol pre-exposure (Figure 5B). Cytoplasmic p65 and p50 in alcohol-treated treated cells remained unchanged (Supplementary Figure 1). Next, we tested whether pre-exposure to alcohol has any effect on LPS-induced NF-κB promoter activity. The upregulation of LPS-induced NF-κB promoter-driven luciferase is significantly blocked by 24 hours of alcohol pre-exposure (Figure 5C), further establishing that alcohol pre-exposure mediates inhibition of TLR4/MyD88/NF-κB signaling.

Figure 5. Alcohol inhibits TLR4-stimulated NF-κB activation via hsp70 and HSF1.

(A-B) RAW macrophages were stimulated with 100 ng/ml LPS for 30 minutes, 50mM alcohol (Et) alone for 1 or 24 hours or pre-exposed to alcohol followed by LPS stimulation for 30 minutes. NF-κB subunits p65 (A) and p50 (B) were detected in nuclear lysates by western blotting. The densitometry graph represents quantitation of bands seen in the gel and depicts mean ± SD of 4 independent experiments. Representative gels are shown with p65 (top) or p50 (top) and loading control TATA-binding protein 1 (TBP-1) (bottom). (C) RAW macrophages were transiently transfected with pNF-κBluc at 1:3 DNA to lipofectamine ratio. 24 hours after transfection, macrophages were treated with 100 ng/ml LPS for 6 hours, 50 mM alcohol (Et) alone for 24 hours or pre-exposed to alcohol for 1 or 24 hours followed by LPS stimulation for 6 hours and luciferase activity was measured in the cell lysates. Graph depicts mean ± SD (n=3). (D-E) RAW macrophages were stimulated with 100 ng/ml LPS for 1 hour or 50mM alcohol (Et) for 1 or 24 hours or pre-exposed to alcohol for 1 or 24 hours followed by LPS stimulation for 1 hr. HS denotes heat shocked control: 42°C for 45 minutes. (D) Whole cell lysates were used for immunoprecipitation with anti-p50 antibody and levels of hsp70 and p50 in immunoprecipitated samples were analyzed by immunoblotting. The densitometry graph represents quantitation of bands seen in the gel and depicts mean ± SD of 3 independent experiments. Representative gels are shown with hsp70 (top) and p50 (bottom). (E) Chromatin immunoprecipitation assay was performed using anti-HSF1 antibody and semi-quantitative PCR was carried out using TNFα promoter specific primers. A representative gel picture is shown above. The densitometry graph represents quantitation of bands seen in the gel (n=4). Input DNA is shown to ensure equal amount of the sheared DNA. (* p<0.05, ** p<0.01, *** p<0.005)

Heat shock-induced hsp70 has been shown to interact with NF-κB subunit p50 in human lymphoma cells (35). To determine whether alcohol induced hsp70 plays a role in NF-κB inhibition, we examined hsp70-p50 interaction during alcohol exposure of macrophages. Alcohol-induced hsp70 showed significantly increased association with p50 subunit after 24 hours of alcohol alone or alcohol pre-exposure followed by LPS as examined by immunoprecipitation experiments (Figure 5D). As expected, this association was not observed in the macrophages stimulated with LPS alone. Thus, alcohol induced hsp70 likely contributes to decreased nuclear translocation and activation of NF-κB downstream of TLR4 via interaction with NF-κB subunit p50.

Similar to anti-inflammatory properties of hsp70, HSF1 can inhibit pro-inflammatory cytokines such as TNFα by binding to the promoter region and acting as a transcriptional repressor (36). To investigate whether alcohol affects HSF1 binding to pro-inflammatory TNFα gene promoter, we performed ChIP analysis. We observed significantly increased HSF1 binding to TNFα promoter in macrophages treated with alcohol alone for one hour (Figure 5E). Although we also observed HSF1 binding to the TNFα promoter upon LPS stimulation alone, this DNA binding activity is probably non-repressive and inconsequential since HSF1 has been shown to be transiently inactivated via phosphorylation upon treatment with LPS alone (21). Alcohol pretreatment for one hour prior to LPS stimulation exhibited some reduction of HSF1 occupancy of TNFα promoter compared to LPS, however it remained comparable to alcohol alone. This indicates alcohol-induced HSF1-mediated direct repression of TNFα transcription could be a mechanism of alcohol induced tolerance in monocytes/macrophages. Collectively alcohol-induced hsp70 and HSF1 exert inhibitory effects on TLR4 signaling resulting in decreased NF-κB activation.

Overexpression of hsp70 is sufficient to induce endotoxin tolerance in macrophages

Our results so far have shown that binge alcohol induces hsp70 which plays an important role in NF-κB inhibition and alcohol-mediated endotoxin tolerance in monocytes/macrophages. Here we determined if expression of hsp70 would be sufficient to mimic alcohol-mediated tolerance and inhibition of TLR4 signaling. Transient transfection of RAW macrophages with pCMV5-hsp70 plasmid showed a significant increase in hsp70 expression as illustrated by western blotting (Figure 6A). Maximal expression of hsp70 was observed at ratio of 1:3 DNA to lipofectamine and was even higher than levels of hsp70 in heat shocked cells (Figure 6A). To investigate the effect of hsp70 on NF-κB activation, we transiently co-transfected RAW cells with NF-κB-promoter driven luciferase plasmid (p(κB)₄-luc) and increasing amounts of pCMV5-hsp70 while maintaining the optimal total DNA to lipofectamine ratio of 1:3. As illustrated in Supplementary Figure 2, co-transfection with both plasmids did not interfere with hsp70 expression and yielded macrophages with optimal overexpression of hsp70. Figure 6B shows that macrophages transfected with p(κB)₄-luc and stimulated with LPS for 6 hours significantly increased NF-κB promoter-driven luciferase activity. On the other hand, overexpression of hsp70 dose-dependently decreased NF-κB-driven luciferase activity. Therefore, hsp70 overexpression can inhibit NF-κB promoter activation downstream of TLR4 in a manner similar to macrophages exposed to alcohol (Figure 5C).

Figure 6. Overexpression of hsp70 is sufficient to mimic alcohol-mediated induction of TLR4 tolerance.

(A) RAW macrophages were transiently transfected with pCMV5-hsp70 using 1:1, 1:2 and 1:3 ratios of 2 μg DNA to lipofectamine. 24 hours after transfection, whole cell lysate was collected and analyzed for hsp70 protein by western blotting. The densitometry graph represents quantitation of bands seen in the gel and depicts mean ± SD (n=5). Representative gels are shown with hsp70 (top) and loading control, β-actin (bottom). UT denotes untransfected and HS denotes heat shocked control: 42°C for 45 minutes. (B) RAW macrophages were transiently transfected with 0.5, 1 or 1.5 μg pCMV5-hsp70 in combination with 0.5 μg p(κB)₄-luc at 1:3 DNA to lipofectamine ratio, maintaining total plasmid DNA at 2 μg. 24 hours after transfection, macrophages were stimulated with 100 ng/ml LPS for 6 hours and luciferase activity was measured in the cell lysates. Graph depicts mean ± SD (n=3). (C-E) RAW macrophages were transiently transfected with 2 μg pCMV5-hsp70 using 1:3 ratio of DNA to lipofectamine. 24 hours after transfection, macrophages were stimulated with 100 ng/ml LPS for 6 hours and culture supernatants were assayed for secreted TNFα (C), IL-1β (D) and IL-6 (E) by ELISA. Graph depicts mean ± SD (n=4). (* p<0.05, ** p<0.01, *** p<0.005)

To determine whether hsp70 is sufficient to mimic the inhibitory effect of alcohol exposure on TLR4-mediated downstream cytokine production, we examined the effect of hsp70 overexpression on LPS-stimulated pro-inflammatory cytokines. As shown in Figure 6, untransfected RAW cells exhibit significant induction of TNFα (Figure 6C), IL-1β (Figure 6D) and IL-6 (Figure 6E) expression in response to LPS, while overexpression of hsp70 decreased cytokine production in LPS-stimulated transfected macrophages. Taken together our results here show that pro-inflammatory cytokine production inversely correlated with hsp70 expression in macrophages suggesting that hsp70 is sufficient to induce NF-κB inhibition and decrease pro-inflammatory cytokines demonstrated in alcohol-mediated endotoxin tolerance.

Knockdown of hsp70 and HSF1 reverses alcohol-induced endotoxin tolerance in macrophages

To determine if HSF1 and hsp70 are required for alcohol induced TLR4 tolerance, we employed siRNA to inhibit expression of these proteins in alcohol-treated RAW macrophages. As illustrated in Figure 7A, we achieved 80% knockdown of HSF1 in untreated and heat-shocked RAW macrophages. Pre-exposure of control transfected RAW cells to alcohol for 24 hours resulted in significant inhibition of LPS-induced TNFα production (Figure 7B). However silencing of HSF1 by siRNA in alcohol-exposed RAW macrophages prevented downregulation of LPS-induced TNFα by alcohol, indicating HSF1 plays an important role in alcohol-mediated endotoxin tolerance in macrophages. Furthermore we wanted to ascertain the role of hsp70 in alcohol-mediated inhibition of pro-inflammatory cytokine production and endotoxin tolerance. To test this, we inhibited expression of hsp70 in RAW macrophages by transient transfection with siRNA and confirmed 90% knockdown in heat-shocked RAW macrophages (Figure 7C). As observed in Figure 7D, hsp70-deficient RAW macrophages pre-exposed to alcohol for 24 hours followed by LPS stimulation prevented downregulation of TNFα production compared to alcohol-exposed macrophages, indicating importance of hsp70 in alcohol-mediated endotoxin tolerance. These results show that HSF1 and hsp70 are both required for alcohol-induced endotoxin tolerance in monocytes/macrophages.

Figure 7. HSF1 and hsp70 are required for alcohol-mediated TLR4 tolerance.

(A and C) RAW macrophages were transfected with 10 nM siRNA targeting HSF1 (A) or hsp70 (B) or negative control (Neg ctrl) siRNA. Total RNA was subjected to HSF1 or hsp70 mRNA determination by qPCR and percent knockdown in unstimulated and heat shocked cells (42°C for 45 min) was calculated with respect to untransfected for each condition. Graph depicts mean ± SD (n=3). (B and D) RAW macrophages were transfected with 10 nM siRNA targeting HSF1 (B) or hsp70 (D) or negative control (Neg ctrl) siRNA. 24 hours after transfection, macrophages were stimulated with 100 ng/ml LPS for 2 hours, 50 mM alcohol (Et) alone for 24 hours or pre-exposed to alcohol followed by LPS stimulation for indicated times. TNFα mRNA was analyzed by qPCR. Graph depicts mean ± SD of 3 independent experiments. (* p<0.05, ns not significant)

DISCUSSION

Although the suppressive effect of short-term or binge alcohol exposure on innate immune activation has been documented (3, 4, 37), the underlying mechanism of alcohol-mediated endotoxin tolerance is not completely understood. Here we delineate novel mechanistic roles of stress proteins hsp70 and HSF1 in suppression of the TLR4/MyD88 pathway in an in vitro setting of short-term alcohol pre-exposure of human peripheral blood monocytes and murine RAW macrophages. To our knowledge, this is the first report describing the crosstalk mechanisms between alcohol-mediated stress proteins and their relevance in suppression of TLR4/MyD88 pro-inflammatory cytokine responses. Previous studies have implicated kinases such as interleukin-1 receptor associated kinase-M (IRAK-M) and protein kinase C (PKC) and other negative regulators including heme-oxygenase 1 (HO-1) and suppressors of cytokine signaling (SOCS) 1 and SOCS3 in alcohol-mediated anti-inflammatory effects on TLR or cytokine-stimulated cells (10, 38-40). The inhibition of TLR4 responses by alcohol has also been attributed to alcohol-induced perturbation of TLR4-CD14 association in lipid rafts (41). Recently alcohol-induced B-cell lymphoma 3-encoded protein (Bcl-3), an inhibitor of NF-κB signaling, has been reported to contribute to alcohol-mediated inhibition of LPS-induced TNFα production (11). However a lack of consensus on the mechanisms by which alcohol pre-exposure of immune cells leads to immunosuppression demands further investigation. Our group has previously shown that binge alcohol exposure induces cellular stress proteins hsp70 and HSF1 in macrophages (24). The crosstalk between stress pathways and immune signaling has been recently recognized (17, 42). Whether alcohol induced stress pathways have any effect in macrophage/monocyte inflammatory responses remains unclear. Here, we demonstrate that alcohol-mediated endotoxin tolerance correlated with alcohol-induced increase in HSF1 activity and hsp70 expression strongly supporting our hypothesis of a role for cellular stress proteins in alcohol-mediated endotoxin tolerance. Both HSF1 and hsp70 exert anti-inflammatory effects and are required for alcohol-mediated endotoxin tolerance. We showed that interaction between alcohol-induced hsp70 and NF-κB subunit p50 and alcohol-mediated HSF1 binding to TNFα promoter result in negative regulation of TLR4 signaling (Figure 8).

Figure 8. Alcohol-induced stress proteins HSF1 and hsp70 play an important role in alcohol-induced TLR4-MyD88 tolerance.

LPS stimulation induces downstream TLR4-MyD88 signaling resulting in NF-κB activation and production of pro-inflammatory cytokines. However, alcohol pre-exposure induced HSF1 and hsp70 (highlighted in gray) directly interact with subsequent LPS stimulated immune signaling molecules resulting in inhibition of NF-κB activation and pro-inflammatory cytokine production causing endotoxin tolerance.

Our results provide evidence that pre-exposure of human monocytes and macrophages to moderate physiological concentrations of alcohol (based on NIAAA/NIH standards: 25 mM approximates 0.1 g/dl blood alcohol concentration, higher than the legal limit 0.08 g/dl which is reached by 4-5 drinks in 2 hours) inhibits pro-inflammatory cytokine production and nuclear translocation of NF-κB downstream of TLR4 stimulation. A comprehensive timecourse analysis of 1-24 hours revealed a time-dependent alcohol-mediated decrease in LPS-induced IL-6, IL-1β and TNFα with maximal inhibition observed at 24 hours of alcohol pretreatment in human monocytes. Our observations are in agreement with Bala et al. (11) who reported inhibition of LPS-induced TNFα production by monocytes and macrophages after 18 hours pre-exposure of binge alcohol. This anti-inflammatory effect of binge alcohol on monocyte/macrophage function is in contrast to the effects observed after chronic alcohol exposure, which sensitized macrophages to augment cytokine production in response to LPS/TLR4 signaling (10, 43, 44). Here we also demonstrate that moderate alcohol inhibits IL-1β and IL-6 mRNA and protein, resulting in endotoxin tolerance. While alcohol exposure alone did not affect NF-κB activity, our results show an inhibitory effect of alcohol pre-exposure on subsequent LPS-induced NF-κB nuclear translocation and promoter-driven luciferase activity at the 24 hours pretreatment timepoint in macrophages. This was consistent with the reported decrease in NF-κB DNA-binding activity observed in 16 hours alcohol pretreated LPS-stimulated macrophages (11). Thus, our detailed analysis illustrates that binge alcohol-mediated endotoxin tolerance is a result of inhibition of TLR4/MyD88 dependent NF-κB activation and pro-inflammatory cytokine production in monocytes and macrophages.

Previous studies show that alcohol induces activation of HSF1 and induces hsp70 in monocytes and macrophages (24). Here we postulated that stress proteins HSF1 and hsp70 play a mechanistic role in alcohol-mediated endotoxin tolerance. HSF1 mRNA expression is constitutive in all cells while its activation is regulated at the post-translational level by cellular stress signals (20). Interestingly, our results show induction of HSF1 mRNA expression by alcohol exposure even in cells subsequently stimulated by LPS indicating alcohol-mediated regulation of HSF1 at the transcriptional level. Transcription factor C/EBPβ has been implicated in induction of HSF1 expression in colonic epithelial cells (45) and upregulation of C/EBPβ protein in response to chronic alcohol exposure has been demonstrated in mouse liver and hepatic nuclear extracts (46, 47). Therefore it is plausible that HSF1 mRNA induction by alcohol may be regulated by C/EBPβ in monocytes/macrophages. Based on the UCSC Genomic Browser Database, putative binding sites for other transcription factors such as Forkhead box protein O1 (FOXO1), Early growth response protein 1 (EGR-1) and Specificity protein 1 (SP-1) have also been reported in the HSF1 promoter region. Alcohol treatment enhances activation of SP-1 and EGR-1 (48, 49). Further investigation will be required to identify the regulatory mechanisms responsible for alcohol-induced HSF1 mRNA expression.

Alcohol exposure was also demonstrated to induce HSF1 nuclear translocation, phosphorylation and activation in astrocytes and colonic epithelial cells (50, 51). Elevation in hsp70 mRNA and protein levels as well as HSF1 protein expression was also observed in livers of mice subjected to short-term alcohol administration (52). Here we demonstrate that alcohol pre-exposure increased HSF1 DNA binding activity in macrophages, consistent with our previous studies of HSF1 activation in macrophages treated with alcohol and LPS together (24). This data was also corroborated by our ChIP analysis which revealed significantly higher binding of HSF1 to the promoter region of target gene hsp70 in alcohol treated macrophages. Alcohol pre-exposure followed by LPS stimulation also induced HSF1 binding to hsp70 promoter albeit to a lesser extent, possibly due to a concomitant increase in HSF1 binding at the TNFα promoter observed in these alcohol-pretreated LPS-stimulated macrophages. Alcohol-mediated activation of HSF1 may be regulated by nuclear translocation, transactivation of DNA binding activity or post-translational modifications. In an effort to determine the mechanism of alcohol-induced HSF1 activation, we first examined the effect of alcohol exposure on cytoplasmic levels of constitutive hsp90β (HSP90AB1), which can directly bind to and inhibit HSF1 nuclear translocation. While we show that cytoplasmic hsp90β expression was not affected by alcohol pre-exposure, previous studies have shown that hsp90-HSF1 complexes were reduced during alcohol treatment in macrophages (24). It is likely that alcohol pre-exposure dissociates HSF1-hsp90 complexes without affecting hsp90β levels in the cytoplasm. Next we examined the effect of alcohol on levels of phosphoserine326-HSF1 (an activating phosphorylation), which is known to be phosphorylated by mTOR kinase (34). Alcohol has been demonstrated to induce mTOR expression in murine liver and cerebral cortex while other groups have shown alcohol-induced decrease in phosphorylation (activation) of mTOR in mouse myocytes and brain cells (53, 54). Based on these observations, we postulated that alcohol could modulate phosphorylation of HSF1 in alcohol-treated cells. However we did not observe an increase in pS326-HSF1 levels in alcohol treated macrophages indicating that regulation of HSF1 activity does not occur via pS326. Ser230 and Ser320, targeted by Ca²⁺/calmodulin dependent kinase II (CaMKII) and protein kinase A (PKA), are other residues correlated with improved HSF1 DNA binding activity which are activated during the heat shock stress response (55, 56). Alcohol exposure has been shown to enhance nuclear localization of the catalytic subunit of PKA (57) indicating possible regulation of HSF1 phosphorylation as a mechanism of activation. It is noteworthy that alcohol-mediated induction of HSF1 and its target gene hsp70 is consistently maintained over a longer period (upto 24 hours) compared to heat shock stress induction which is quick and transient (58). This suggests that alcohol-induced HSF1 activation follows a mechanism distinct to that induced by heat shock and may thus be regulated by alternate post-translational modifications. Therefore, further investigations will be required to determine the mechanism of alcohol-mediated HSF1 activation.

Our results demonstrate alcohol-mediated induction of HSF1 target gene hsp70 in human monocytes at the mRNA and protein level. The upregulation of hsp70 mRNA at 14-24 hours of alcohol exposure is consistent with the observation of alcohol-mediated endotoxin tolerance at 24 hours of alcohol pretreatment. Our findings here in human monocytes support previously published results from our group showing increased hsp70 levels in RAW macrophages when treated with alcohol and LPS together (24). Importantly, alcohol-mediated upregulation of hsp70 mRNA and protein was not further altered by LPS in human monocytes, suggesting endotoxin stimulation did not affect or enhance induction of hsp70 in alcohol treated monocytes. Our results indicate that alcohol-induced hsp70 likely interacts with the LPS-stimulated TLR4 signaling pathway to exert its anti-inflammatory effects. To address crosstalk mechanisms and direct effects of alcohol induced hsp70 in macrophages, we overexpressed hsp70 and tested its effect on LPS-induced NF-κB signaling. We showed an hsp70 dose-dependent decrease in NF-κB promoter-driven luciferase activity and inhibition of expression of NF-κB dependent pro-inflammatory cytokines TNFα, IL-6 and IL-1β. These results are in agreement with previous reports of inhibition of NF-κB translocation and activation by hsp70 (59) and also corroborated our hypothesis that hsp70 expression is sufficient to mimic alcohol-induced endotoxin tolerance.

The anti-inflammatory properties of stress proteins have been reported previously (17). HSF1 has been demonstrated to exert anti-inflammatory effects by repressing transcription of TLR4 induced genes such as TNFα (36). Here we show HSF1 binding to TNFα promoter by ChIP analysis in alcohol-pretreated macrophages indicating that alcohol-induced HSF1 could directly repress pro-inflammatory cytokines by binding to a specific promoter region, a plausible mechanism of alcohol-mediated endotoxin tolerance. Hsp70 has been shown to directly bind NF-κB subunit p50, resulting in decreased NF-κB activation in lymphoma cells (35). Here, we show that the alcohol-mediated inhibition of NF-κB nuclear translocation could be attributed to the association of subunit p50 with hsp70. The interaction of hsp70 with p50 could hamper nuclear translocation and promoter transactivation by NF-κB, resulting in reduced expression of pro-inflammatory cytokines observed in our studies, thus providing a direct link between alcohol, hsp70 and endotoxin tolerance. Finally our siRNA experiments confirm HSF1 and hsp70 as plausible intermediates in alcohol-mediated endotoxin tolerance. Our results showed that knockdown of HSF1 or hsp70 prevented alcohol-mediated downregulation of TNFα and thus endotoxin tolerance, indicating both of these stress proteins are required for alcohol-mediated tolerance.

Collectively our studies here identify that clinically relevant alcohol pre-exposure induces stress proteins HSF1 and hsp70 which negatively regulates TLR4 mediated pro-inflammatory responses. It should be noted that the quantity and form of alcohol consumed also seems to be a factor in modulation of immune responses. For example, consumption of moderate amounts of red wine containing antioxidants, which would dampen cellular stress responses, did not show an immunosuppressive effect in mice and humans (60, 61). Moderate alcohol-mediated immunosuppression may also have protective effects in a tissue-specific context, for instance, possibly by ‘dampening’ the inflammatory immune responses which contribute to cardiovascular disease (62, 63). Here, we have examined the effect of moderate alcohol exposure on TLR4 responses with a specific emphasis on an alcohol pre-exposure model which has direct physiological/clinical relevance. For example, accident or trauma patients are more susceptible to subsequent viral or bacterial infections due to prior consumption of alcohol that can likely compromise their innate immune responses (26, 64). Our studies here not only test the effect of this physiological relevant alcohol exposure on innate immune function but also recognize novel crosstalk mechanisms affected by alcohol that reduce inflammatory responses. While alcohol induced HSF1 may exert repressive effects on cytokine production at early time points, the interaction of hsp70 and NFκB at later time-points is important in alcohol mediated inhibition of pro-inflammatory cytokines. Identification of the crosstalk between HSF1, hsp70 and TLR4 signaling molecules that play a role in alcohol-induced endotoxin tolerance will provide an insight into novel links between stress proteins and immune signaling pathways and unravel novel therapeutic mechanisms for intervention in alcohol abuse patients. Consequently, the use of drugs directly targeting stress proteins hsp70 or HSF1 in innate immune cells may aid in restoration of normal immune function in alcohol exposed individuals. The ready availability of such drugs for treatment of cancer (65, 66) adds to the appeal of these stress proteins as possible drug targets in treatment of alcohol abuse.