Akt Cys-310-targeted Inhibition by Hydroxylated Benzene Derivatives Is Tightly Linked to Their Immunosuppressive Effects*

The hydroxylated benzene metabolite hydroquinone (HQ) is mainly generated from benzene, an important industrial chemical, and is also a common dietary component. Although numerous reports have addressed the tumorigenesis-inducing effects of HQ, few papers have explored its molecular regulatory mechanism in immunological responses. In this study we characterized Akt (protein kinase B)-targeted regulation by HQ and its derivatives, in suppressing inflammatory responses using cellular, molecular, biochemical, and immunopharmacological approaches. HQ down-regulated inflammatory responses such as NO production, surface levels of pattern recognition receptors, and cytokine gene expression with IC50 values that ranged from 5 to 10 μm. HQ inhibition was mediated by blocking NF-κB activation via suppression of its translocation pathway, which is composed of Akt, IκBα kinase β, and IκBα. Of the targets in this pathway, HQ directly targeted and bound to the sulfhydryl group of Cys-310 of Akt and sequentially interrupted the phosphorylation of both Thr-308 and Ser-473 by mediation of β-mercaptoethanol, according to the liquid chromatography/mass spectroscopy analysis of the interaction of HQ with an Akt-derived peptide. Therefore, our data suggest that Akt and its target site Cys-310 can be considered as a prime molecular target of HQ-mediated immunosuppression and for novel anti-Akt-targeted immunosuppressive drugs.

Inflammation is one of the innate immunity responses and a multiple step process that is mediated by activated inflammatory or immune cells. In particular, macrophages that differentiate from monocytes have a critical role in managing many different immunopathological events in inflammation, such as the overproduction of inflammatory mediators (nitric oxide (NO) and prostaglandin E 2 ), which are generated by activated inducible nitric oxide synthase and cyclooxygenase-2, and various chemokines and cytokines, such as tumor necrosis factor-␣ (TNF-␣), 3 interleukin (IL)-1␤, IL-6, and IL-12 (1,2). The pro-inflammatory events are mainly initiated by the activation of surface receptors (pattern recognition receptors), such as the Toll-like receptors (TLR-4) or TLR-2 after binding to their ligands, including lipopolysaccharide (LPS) from microorganisms (3). During the activation of pattern recognition receptors, a complicated intracellular signaling machinery, including mitogen-activated protein kinases (MAPKs), serine/threonine protein kinases, and non-receptor type tyrosine kinases, are upregulated to trigger inflammatory gene expression via activation of transcription factors, such as nuclear factor (NF)-B and activator protein-1 (AP-1) (4,5). Primary (peritoneal and bone marrow (BM)-derived) macrophages and cancerous macrophage-like cells (e.g. RAW264.7 and J774 cells) induced by LPS are now regarded as a useful in vitro model for evaluating the potency of anti-inflammatory drugs and for exploring their anti-inflammatory mechanisms because of their ability to display similar inflammatory states (6).
Hydroquinone (benzene-1,4-diol; HQ) is a benzene metabolite mainly found in cigarette smoke, coffee, industrial chemicals, and petroleum byproducts and is a ubiquitous environmental pollutant. In addition, cigarette smoke-mediated allergy symptoms are caused by HQ, which acts as a strong hapten (7), and previous findings that it blocked interferon-␥ production in Th1 cells (8), enhanced IL-4 production in CD4ϩ T cells (9), increased immunoglobulin E levels in antigen-primed mice (9), and blocked IL-12 production via suppression of NF-B binding activity (7), inhibition of lymphocyte proliferation (10), and suppression of macrophage-mediated phagocytosis (11) strongly suggest that it acts as an immunomodulator rather than as a simple toxic chemical. In particular, strong decreases in production of cytokines (IL-1␤, IL-2, IL-6, IL-10, and TNF-␣) and inflammatory mediators (NO and prostaglandin E 2 ) by HQ (12) have led to the hypothesis that this compound can be applied like a strong anti-inflammatory drug with immunosuppressive properties. Currently, the molecular inhibitory mechanisms of benzene metabolites and their molecular targets in the modulation of immune responses remain largely uncharacterized. In the present study we investigated the molecular effect of HQ on the modulation of inflammatory processes mediated by macrophages.

Animals
ICR and C57BL/6 male mice (6 -8 weeks old, 17-21 g) were obtained from Daehan Biolink (Chungbuk, Korea) and maintained in plastic cages under conventional conditions. Water and pelleted diets (Samyang, Daejeon, Korea) were supplied ad libitum. Studies were performed in accordance with guidelines established by the Kangwon National University Institutional Animal Care and Use Committee.

Construction of Expression Vectors
GFP-fused wild type (WT) Akt construct (GFP-Akt-WT) was prepared by amplification using a typical culture method with competent E. coli (DH5␣). pcDNA-HA, pcDNA-HA-tagged Akt, pcDNA-HA-tagged PDK1, and pcDNA-Myc-tagged PDK1 constructs were used as reported previously (16). The mutant at Cys-310 (GFP-Akt-C310A) was created by using the QuikChange sitedirected mutagenesis kit (Stratagene) with the primers 5Ј-ACC ATG AAG ACC TTT GCC GGC ACA CCT GAG TAC (forward) and GTA CTC AGG TGT GCC GGC AAA GGT CTT CAT GGT (reverse), as described by the manufacturer with pCMV5 GFP-Akt-WT as template. All constructs were confirmed by automated DNA sequencing. Sequences of the mutagenic oligonucleotides are available upon request.

Preparation of Peritoneal Macrophages and BM-derived Macrophages
Peritoneal macrophages were obtained as reported previously (17). BM-derived macrophages were also prepared from BM-derived cells after treating granulocyte-macrophage colony-stimulating factor and IL-4, as reported previously (18).

Cell Culture
Peritoneal macrophage, BM-derived macrophages, RAW264.7, wild type HEK293, stable cell lines expressing the Myc-tagged PDK1 WT, derived from HEK 293 cells (19), and A21 cells were maintained in RPMI1640 or Dulbecco's modified Eagle's medium supplemented with 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal bovine serum in the presence or absence of G418. Cells were grown at 37°C with 5% CO 2 .

Determination of NO Production
RAW264.7 cells (1 ϫ 10 6 cells/ml) under normal culture conditions were preincubated with each compound for 30 min and continuously activated with LPS (2 g/ml) for 24 h. The nitrite in culture supernatants was also measured by a Griess assay as reported previously (17).

Flow Cytometric Analysis
Expression of RAW264.7 cell surface adhesion molecules under LPS treatment was determined by flow cytometric analysis as reported previously (21). Stained cells were analyzed with a FACScan cytometer (BD Biosciences).

Luciferase Reporter Gene Activity Assay
HEK293 cells (1 ϫ 10 6 cells/ml) were transfected with 1 g of plasmids containing NF-B-Luc or AP-1-Luc as well as ␤-galactosidase using the calcium phosphate method in a 12-well plate according to the manufacturer's protocol. The cells were used for experiments 48 h after transfection. Luciferase assays were performed using the Luciferase Assay System (Promega) as reported previously (23).
Nuclear lysates were prepared with a three-step procedure (24). After treatment, cells were collected with a rubber policeman, washed with 1ϫ PBS, and lysed in 500 l of lysis buffer containing 50 mM KCl, 0.5% Nonidet P-40, 25 mM HEPES, pH 7.8, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 20 g/ml aprotinin, and 100 M DTT on ice for 4 min. Cell lysates were then centrifuged at 14,000 rpm for 1 min in a microcentrifuge. In the second step, the pellet (the nuclei fraction) was washed once in washing buffer, which was the same as the lysis buffer without Nonidet P-40. In the final step nuclei were treated with an extraction buffer containing 500 mM KCl, 10% glycerol, and several other reagents as in the lysis buffer. The nuclei/extraction buffer mixture was frozen at Ϫ80°C and then thawed on ice and centrifuged at 14,000 rpm for 5 min. Supernatant was collected as nuclear extract.
Whole cell or nuclear lysates were then analyzed by immunoblotting. Proteins were separated on 10% SDS-polyacrylamide gels and transferred by electroblotting to polyvinylidene difluoride membrane. Membranes were blocked for 60 min in Tris-buffered saline containing 3% bovine serum albumin, 20 mM NaF, 2 mM EDTA, and 0.2% Tween 20 at room temperature. The membrane was incubated for 60 min with specific primary antibody at 4°C, washed 3 times with the same buffer, and incubated for an additional 60 min with horseradish peroxidase-conjugated secondary antibody. The total and phosphorylated levels of MEK, ERK, p38, JNK, IB␣, IKK␤, p85, PDK1, Akt, MBP, ␥-tubulin, and ␤-actin were visualized using the ECL system (Amersham Biosciences).

Confocal Microscopy
RAW264.7 cells (1 ϫ 10 4 cells) were plated in 12-well plates containing sterile coverslips and grown at 37°C for 24 h. The medium was then replaced with serum-free media, and the cells were allowed to grow for another 24 h before treatment. Cells were treated with HQ for 30 min followed by stimulation with LPS (2 g/ml) for 1 h. After treatment, the cells were washed twice with PBS prewarmed to 37°C and fixed onto the coverslips by incubation in 3.7% formaldehyde for 10 min. Cells were then washed 3 times with PBS and permeabilized by incubation in 100% methanol for 6 min at Ϫ20°C. The coverslips were blocked in 1% bovine serum albumin for 1 h at room temperature with shaking. Antibody to the NF-B p65 subunit (1:50) was added to the 1% bovine serum albumin solution and incubated for 1 h with shaking at room temperature. For nuclear staining, Hoechst solution (Sigma) was added at a final concentration of 0.5 mg/ml and incubated for 1 h in the dark. Coverslips were then washed 3 times each with PBS. Alexa 488-conjugated secondary antibody (1:100) in 1% bovine serum albumin was then added and incubated for 1 h with shaking at room temperature. Coverslips were washed three times with PBS and mounted onto slides using Fluorescent mounting medium (DakoCytomation). The nuclear translocation of p65 was imaged by laser-scanning confocal microscopy on a Zeiss LSM 510 META confocal microscope equipped with a Zeiss 37 1C incubation system. Images were analyzed using the Zeiss LSM Image Examiner.

Immunoprecipitation and in Vitro Kinase Assay for PDK1 and Akt
HEK 293 cells were placed on ice and extracted with lysis buffer containing 50 mM Tris-HCl, pH 7.5, 1% (v/v) Nonidet P-40, 120 mM NaCl, 25 mM sodium fluoride, 40 mM ␤-glycerol phosphate, 0.1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, and 1 mM benzamidine. Lysates were centrifuged for 15 min at 12,000 ϫ g, and Myc-PDK1 or HA-Akt protein was immunoprecipitated from 500 g of cell-free extracts with anti-Myc 9E10 monoclonal antibody or anti-HA 12CA5 monoclonal antibody immobilized on protein G-Sepharose (Amersham Biosciences). The immune complexes were washed once with lysis buffer containing 0.5 M NaCl followed by lysis buffer and finally with kinase assay buffer (50 mM Tris-HCl, pH 7.5, 0.1% (v/v) 2-mercaptoethanol (mer-EtOH)). In vitro kinase assays were performed for 30 min (Akt) or 60 min (PDK1) at 30°C in a 50-l reaction volume containing 30 l of immunoprecipitates in kinase buffer, 100 M Suntide (RRK-DGATMKTFCGTPE), or 30 M Crosstide (GRPRTSSFAEG) as substrate, 10 mM MgCl 2 , 1 M protein kinase A inhibitor peptide (Alexis), and 100 M [␥-32 P]ATP (1000 -2000 cpm/ pmol; Amersham Biosciences). Reactions were stopped by adding EDTA to a final concentration of 50 mM and processed as described previously (16). Protein concentrations were determined by the method of Bradford (Bio-Rad) using bovine serum albumin as a standard.

Kinase Assay with Purified Enzymes
A kinase assay was performed by kinase profiler service from Millipore (Billerica MA). In a final reaction volume of 25 l, human purified enzymes (Akt1, p70S6K, protein kinase A, protein kinase C␣, or serum and glucocorticoid-inducible kinase) (1-5 milliunits) was incubated with the reaction buffer in the absence of mer-EtOH. The reaction was initiated by the addition of MgATP. After incubation for 40 min at room temperature, the reaction was stopped by the addition of 5 ml of a 3% phosphoric acid solution. 10 l of the reaction was then spotted onto a P30 filtermat and washed 3 times for 5 min in 75 mM phosphoric acid and once in methanol before drying and scintillation counting.

Characterization of Hydroquinone-Suntide Adduct
HPLC Analysis-A reverse phase HPLC system was used for the analysis of HQ as reported previously (25). The HPLC system (Waters) consisted of a pump (Waters TM 600 Controller), a UV-visible spectrophotometric detector (Waters TM 486 Tunable Absorbance Detector), an autosampler (Waters TM 717 plus Autosample), a degasser (Waters TM In-line Degasser), a reverse phase column (Luna 5-m C 18 analytical column (150 ϫ 4.6 mm)) and an integrator (Borwin 1.20 software). For isocratic analysis, 1% acetic acid in H 2 O 2 was used as the mobile phase. The flow rate of mobile phase was 1 ml/min, and the column eluate was monitored by a UV detector set at 290 nm.
Matrix-assisted Laser Desorption Ionization Time-of-flight Mass Spectrometry-␣-Cyano-4-hydroxycinnamic acid (20 mg) (Bruker Daltonics, Bremen, Germany) was dissolved in 1 ml of acetone:ethanol (1: 2, v/v), and 0.5 l of the matrix solution was mixed with an equivalent volume of sample. Analysis was performed using an Ultraflex TOF/TOF system (Bruker Daltonics). The Ultraflex TOF/TOF system was operated in positive ion reflect mode. Each spectrum was the cumulative average of 250 -450 laser shots. Mass spectra were first calibrated in the closed external mode using the peptide calibration standard II (Bruker Daltonics), sometimes using the internal statistical mode to achieve maximum calibration mass accuracy. Standard settings included the following: mass values, MH ϩ (monoisotopic); mass tolerance, varied between 75 and 100 ppm.
Molecular Modeling Study-To understand the binding mode of action of a cysteine-HQ adduct (mer-EtOH-hydroquine (MHQ)), a molecular modeling study in the active site using Sybyl Version 8.02 (Tripos Associates) operating under Red Hat Linux 4.0 on an IBM computer (Intel Pentium 4, 2.8 GHz CPU, 1 GB memory) and the x-ray crystallographic structure of an activated Akt ternary complex with glycogen synthase kinase and AMP-PNP (26) has been carried out. The structure of the MHQ was drawn into the Sybyl package with standard bond lengths and angles and was minimized using the conjugate gradient method. The Gasteiger-Huckel charge, with a distance-dependent dielectric function, was applied for the minimization process. The PDB code 1O6L structure from the Protein Data Bank was chosen, and the structure was polished following the structure preparation tool in Sybyl. After this, MHQ was merged to Cys-310 for having S-S bonding, and the consequent complex was minimized by subset minimization tool for producing the conformationally stable complex structure.

In Vivo Inflammatory Models
Septic Shock Models-C57BL/6 male mice were orally pretreated with HQ (100 mg/kg), JS-III-49 (25 and 50 mg/kg), and indomethacin (1 mg/kg) 4 times. Thirty min after the final treatment, D-GalN (600 mg/kg) and LPS (50 g/kg) were intraperitoneally co-injected, as reported previously (27). The effect of HQ or its derivative on septic shock-induced mouse lethality was calculated from the number of survival mice after 7 days of observations.
Arachidonic Acid-induced Mouse Ear Edema-ICR mice (n ϭ 7) were orally pretreated with JS-III-49 (25 and 50 mg/kg) and indomethacin (1 mg/kg) 4 times. After the final treatment, AA (2% (w/v)) was applied to the ear of the mouse (25 l/ear), as described previously (28). The thickness of the edema was measured with a constant-pressure-thickness gauge 4 h after AA treatment. To evaluate curative efficacy, the inhibitory effect of ear edema by testing inhibitors was calculated as reported previously.
Determination of Serum TNF-␣ Level-Ninety min after intraperitoneal injection of D-GalN (600 mg/kg) and LPS (50 g/kg), blood was collected, and serum samples were used to measure TNF-␣ levels by means of an enzyme-linked immunosorbent assay kit.
Histological Analysis of the Liver-Tissue samples taken from the liver of the mice at 8 h after challenge with LPS and D-GalN were fixed with 10% formalin in PBS and then embedded in paraffin. Approximately 4-m-thin tissue sections were stained with hematoxylin and eosin for histopathological examination as reported previously (29).
Statistical Analysis-A Student's t test and one-way analysis of variance were used to determine the statistical significance of differences between values for the various experimental and control groups. Data are expressed as the means Ϯ S.E., and the results were obtained from at least three independent experiments performed in triplicate. p values of 0.05 or less were considered as statistically significant.

Effect of Hydroquinone on LPS-induced Inflammatory
Responses-We previously reported that HQ displays an immunosuppressive effect on various immune responses. Before starting the molecular mechanistic study on HQ-mediated effects, the effects induced by LPS in macrophages were first examined during HQ treatment. Three different types of macrophages, cancerous macrophage-like RAW264.7 cells and two primary (peritoneal and BM-derived) macrophages, were employed. As Fig. 1A shows, HQ suppressed the production of NO in a dose-dependent manner like several kinase inhibitors, including piceatannol, PP2, and LY294002 (data not shown). Furthermore, HQ also remarkably diminished the enhanced expression of surface molecules, such as TLR4 and CD69, in response to LPS (Fig. 1B), as in the case of BAY 11-7082, a NF-B inhibitor (data not shown). Although it has been already published that the inhibitory effect of HQ occurs at the transcriptional level, inhibitory patterns at the mRNA level and the treatment time dependence were simultaneously confirmed. In agreement with previous results (22), Fig. 1C shows that HQ was capable of reducing not only the mRNA levels of proinflammatory genes but also early exposure before LPS led to maximum suppression. These results, therefore, suggest that early events triggered by LPS exposure could be the target of HQ inhibition.
Effect of Hydroquinone on LPS-induced NF-B Activation-We carefully further characterized the inhibitory effect of HQ on NF-B activation. Even though it is generally well accepted that HQ acts as a NF-B inhibitor (7, 30), whether HQ directly or indirectly blocks this transcription factor is still controversial. To address this question, NF-B translocation signals were mainly investigated in this study. As Fig. 2A shows, this compound selectively diminished the luciferase activity mediated by NF-B but not AP-1. Furthermore, Fig. 2, B-D, clearly supports that HQ suppressed the upstream signaling events for NF-B (p65) translocation. Thus, phosphorylation of IKK␤ and IB␣, an important process for NF-B (p50/p65) translocation, was clearly suppressed in response to HQ exposure (Fig. 2, B and C). In the case of curcumin treatment, it also more strongly blocked IKK␤ than IKK␣ (Fig. 2C). The final step of the IKK/ IB␣ signaling cascade, nuclear translocation of NF-B (p65), was also diminished 1 h after LPS administration by HQ treatment in a dose-dependent manner according to immunoblotting (Fig. 2D) and confocal (Fig. 2E) analyses. Therefore, these results indicate that the molecular target of HQ could be one of upstream signaling enzymes that regulate NF-B translocation.
Effect of Hydroquinone on the Upstream Signaling Enzymes for Nuclear NF-B Translocation-The major components of the NF-B activation pathway are now well known and consist of Akt/PDK1 and phosphatidylinositol 3-kinase as well as proteintyrosine kinases, such as Src, Syk, and Jak2 (31). Therefore, the potential inhibitory target of HQ was first investigated by exploring the NF-B signaling pathway. Interestingly, our data implied that Akt could be a potential target of HQ. Thus, Fig. 3A shows that inhibition of phosphatidylinositol 3-kinase/p85 and PDK1 phosphorylation was not observed. Instead, this compound clearly blocked the phosphorylation of Akt and showed two different inhibitory patterns; phosphorylation on Thr-308 was inhibited at 2 and 5 min and Ser-473 at 5-15 min (Fig. 3A). On the other hand, this compound more strongly suppressed Akt phosphorylation when treated for 30 min in RAW264.7 cells (Fig. 3B, left panel). A similar effect was also obtained with peritoneal macrophages (Fig. 3B,  right panel), suggesting that HQ inhibits Akt phosphorylation in both cancerous and primary cell states. Meanwhile, MAPK phosphorylation, which is required for AP-1 translocation, was not suppressed by HQ (Fig. 3C), in agreement with the AP-1-mediated luciferase activity assay ( Fig. 2A).
Effect of Hydroquinone on Akt Kinase Activity-To better characterize the Akt inhibitory mechanism, Akt kinase assay conditions were introduced. First, using specific peptide substrates, Suntide (14) and Crosstide (15), which originate from the amino acid sequences of Akt or its substrate (glycogen synthase kinase 3) (Fig. 4A), and immunoprecipitated Akt or PDK1, the direct inhibitory effect of HQ on Akt kinase activity was carefully investigated. Similarly to the aforementioned data, HQ at 100 M strongly suppressed Akt kinase activity by up to 95% with the Crosstide substrate peptide (Fig. 4B). The Akt phosphorylation induced by insulin in 2A1 cells (HEK293 cells that stably express HA-PKB) was also remarkably diminished by HQ, implying that the inhibition seemed to be generally observable during Akt activation conditions (Fig. 4C). A particularly interesting finding was obtained in the PDK1 kinase assays with two different substrates. Namely, although phosphorylation of MBP by PDK1 was not affected by HQ, phosphorylation of Suntide was suppressed (Fig. 4D). Consistently, the activity of immunoprecipitated PDK1 was not altered when HQ was directly treated to the RAW264.7 cells (Fig. 4E), suggesting that PDK1 is not the direct target, but HQ-mediated inhibition could be relevant to the sequence of the Akt-derived peptide Suntide (Fig. 4A). Effect of Thiolation on Hydroquinone-mediated Inhibition of NO Production-It has been reported that HQ and its chemical derivatives require hydroxyl groups in the para form for their maximum anti-inflammatory activities (32). Whether this structural feature is simply important for chemical reactions that generate or neutralize toxic radicals or specific interaction with target molecules that mediate its variable biological effects is still not fully understood. In this study we first attempted to identify the biological role of HQ in terms of its chemical properties using several anti-oxidants or thiol compounds. Fig. 5A, left panel, clearly shows that HQ inhibition of NO production could be due to electrophilic reactivity leading to thiolation. Thus, combined treatment of thiol compounds, such as L-cysteine, N-acetyl-L-cysteine, and DTT, abrogated HQ inhibition, whereas ␣-tocopherol at 12.5 M, which exhibits strong antioxidative activity (data not shown), did not. The exact same pattern was also seen for NO production from BM-derived macrophages (Fig. 5A, right panel). Meanwhile, the abrogating effect of L-cysteine was only found when it was administered at the same time or before HQ treatment (Fig. 5B), suggesting that thiolation needs to precede HQ. On the other hand, the inhibitory mechanism of the para-type HQ derivative tBHQ on NO production (Fig. 5C, left panel) seems to be the same as HQ because pretreatment with L-cysteine before tBHQ also abolished its ability to suppress NO production (Fig. 5C, right panel). Therefore, these results suggest that the thiolation-inducing activity of HQ could be a major inhibitory mechanism.
Effect of Thiolation on Hydroquinone-mediated Inhibition of Akt Kinase Activity-To obtain direct evidence linking thiolation to the regulation of Akt kinase activity, several molecular studies, including Akt kinase assays and immunoblotting anal-FIGURE 2. The effect of HQ on NF-B activation. A, HEK293 cells transfected with plasmid constructs containing NF-B-Luc or AP-1-Luc (1 g/ml each) as well as ␤-galactosidase were treated with HQ in the presence or absence of phorbol 12-myristate 13-acetate (PMA, 0.1 M) for 12 h. Luciferase activity was determined by a luminometer as described under "Experimental Procedures." B and C, RAW264.7 cells (5 ϫ 10 6 cells/ml) were incubated with HQ, curcumin, or U0126 in the presence or absence of LPS (2 g/ml) for 30 min. The levels of phospho-IB␣ or IKK(␣/␤) from total lysates were analyzed by immunoblotting. D, RAW264.7 cells (5 ϫ 10 6 cells/ml) were incubated with HQ in the presence or absence of LPS (2 g/ml) for 30 min. Then the nuclear fraction was obtained from the cells, as described under "Experimental Procedures". The nuclear levels of p65 and lamin A/C were determined by immunoblotting analysis. E, confocal analysis of the p65 level in nuclei of LPS-activated RAW264.7 cells in the presence or absence of HQ was performed as described under "Experimental Procedures." Data in A represent the mean Ϯ S.E. of three independent experiments performed in triplicate. The results (B-E) show one representative experiment of three. p Ͻ 0.05 (*) and p Ͻ 0.01 (**) compared with the control group. ysis, were executed. As shown in Fig. 6A, L-cysteine treatment also clearly abrogated HQ-mediated inhibition of Akt phosphorylation in response to LPS at 5 and 30 min (Fig. 6A). The remarkably decreased level of Akt phosphorylation induced by insulin stimulation was also restored by co-treatment with DTT (Fig. 6B). Most of all, DTT was able to attenuate HQmediated inhibition of both Akt kinase activity with the Crosstide substrate (Fig. 6C) and PDK1 activity with the Suntide substrate (Fig. 6D). Therefore, our results suggest that the HQinduced thiolation reaction could play a critical role in modulation of Akt phosphorylation and kinase activity.
Effect of ␤-Mercaptoethanol on Formation of the Hydroquinone-Suntide Adduct and Regulation of Kinase Activity-Because HQ could be thiolated with a sulfhydryl compound, we moved to a kinase assay system to determine why phosphorylation of the Suntide peptide but not MBP was selectively blocked by HQ in the PDK1 kinase assay (Fig. 4D). In particular, because Suntide has a cysteine residue (Fig. 4A) that is con-served in proteins with an activation loop, such as protein kinase C␣, protein kinase C-related kinase (PRK), and serum and glucocorticoid-inducible kinase, and important for threonine phosphorylation (14), we focused our experiments on the potential role of the sulfhydryl group of that cysteine. Interestingly, Fig. 7A indicates that HQ strongly blocked the kinase activity of Akt in a dose-dependent manner in the normal kinase buffer system with mer-EtOH. However, without mer-EtOH, none of the kinases was significantly suppressed by HQ, suggesting that mer-EtOH might play a critical role in inhibition by HQ. To address this possibility, chemical conversion was examined using HPLC and MS. Surprisingly, an adduct formed with Suntide, mer-EtOH, and HQ was identified by LC-MS-MS analysis (Fig. 7B), whereas adduct formation between HQ and thiol compounds was not observed in PBS conditions according to the HPLC analysis (Fig. 7C). Therefore, these results suggest that HQ blocks the phosphorylation of Thr-308 through chemical modification of Cys-310 by the MHQ generated with electrophilic agents, such as mer-EtOH (Fig. 7D).

Role of Akt Cys-310 in the Phosphorylation of Akt and Hydroquinone-mediated Inhibition of Akt Kinase Activity-To
address the role of Cys-310 in Thr-308 phosphorylation and as a HQ binding site, a molecular biological method was employed. We first prepared a mutant Akt in which Cys-310 was substituted with alanine (Akt-C310A; Fig. 8A) and overexpressed it in HEK293 cells. The phosphorylation patterns of the mutant Akt were examined and compared with WT. As  Fig. 8B displays, phosphorylation of Thr-308 was severely deficient in the mutant Akt. Similarly, Ser-473 phosphoryla-tion was also dramatically altered regardless of the 3-fold increase in expression (Fig. 8B). More intriguingly, HQ suppression of Akt kinase activity was strongly seen in WT conditions, whereas the mutant activity was attenuated by only 50% after HQ treatment (Fig. 8C). Therefore, these results suggest that Cys-310 could play a central role in modulation of Akt phosphorylation and serve as a major binding site for HQ. To support this possibility, a docking model has been speculated using previous structural data (26). Thus, the hydroxyl group of MHQ seemed to form H-bonds with the Glu-298 and to hydrophobically interact with His-196 (Fig.  8D), which is considered an essential role to stabilizing the  complex in the active site. These chemical interactions presumably looked to make an environment to disturb the molecular interaction between PDK1 and its substrate amino acid residue Thr-308. From this modeling study, therefore, we propose that the MHQ bonded with Cys-310 is able to work as a Thr-308 phosphorylation blocker in Akt.
Effect of Novel Hydroquinone Derivatives on PDK1 Kinase Activity, NO Production, Septic Shock, and Ear Edema Models-Because it has also been suggested that Cys-310 of Akt is a potentially important site for anti-cancer drug development (33,34), we further characterized the possibility that HQ could be chemically modified to be pharmacologically improved. Indeed, several derivatives (Fig. 9A) strongly blocked the phosphorylation of Suntide by the PDK1 kinase (Fig. 9B) and consequently LPS-induced NO production (Fig. 9C) with IC 50 values that ranged from 2 to 25 M (Fig. 9D) without altering normal cell viability, except for one compound (JS-III-73) (Fig. 9E). In particular, these derivatives also worked well in vivo. When it was orally administered, one of these compounds (JS-III-49 at 50 and 100 mg/kg) as well as HQ (100 mg/kg) strongly suppressed lethality induced by the combined treatment of D-GalN and LPS by up to 70% (Fig. 10A). The serum level of TNF-␣, a major cause leading to higher lethality (35), enhanced by LPS/ D-GalN treatment, was also remarkably suppressed by this drug as did prednisolone, a steroid drug (36) (Fig. 10B). In agreement, liver damage induced by LPS/D-GalN treatment was revealed to be protected by JS-III-49 administration (Fig. 10C). Moreover, JS-III-49 at 50 mg/kg also significantly ameliorated ear edema induced by AA by 55% (Fig. 10D). Although indomethacin exhibited a stronger anti-edema effect than JS-III-49 (Fig. 10D), this standard drug reduced normal body weight unlike HQ derivative (data not shown). Therefore, these results FIGURE 6. The effect of thiol compounds on HQ-mediated inhibition of Akt phosphorylation and kinase activity. A, RAW264.7 cells (5 ϫ 10 6 cells/ml) pretreated with L-cysteine were incubated with HQ in the presence or absence of LPS (2 g/ml) for 5 and 30 min. The phosphorylated or total levels of Akt from total lysates were analyzed by immunoblotting. B, 2A1 cells (5 ϫ 10 6 cells/ml) pretreated with DTT were stimulated with HQ in the presence or absence of insulin (100 M). After immunoblotting, total or phosphoprotein levels of Akt were identified with the total protein-or phospho-specific antibodies. C, kinase activity of immunoprecipitated Akt prepared from 2A1 cells (5 ϫ 10 6 cells/ml) stimulated with insulin (100 M) was measured with Crosstide (30 M) after incubation with DTT and HQ for 10 min. D, HQ and DTT were directly incubated with immunoprecipitated PDK1 for 10 min. Kinase activity was measured with Suntide (100 M). Data (C and D) represent the mean Ϯ S.E. of three independent experiments performed in triplicate. The results (A and B) show one representative experiment of three. p Ͻ 0.05 (*) and p Ͻ 0.01 (**) compared with the control group. p Ͻ 0.05 (#) and p Ͻ 0.01 (##) compared with the stimulus ϩ HQ group with the same concentration.
suggest that HQ can be further developed chemically to generate a more powerful Akt Cys-310 targeted immunosuppressive agent for use in vivo.

DISCUSSION
HQ is representative of the toxic benzene-type compounds found in cigarette smoke, coffee, and numerous industrial products from petroleum companies. Although the compound has been identified as a toxic molecule, the toxicological mechanisms of HQ in various cellular and immune responses are not fully understood. Recently, we proposed that HQ could be used as an anti-inflammatory drug because it diminished the release of inflammatory mediators (21). However, most data seem to suggest that this compound possesses potent immunosuppressive activity. For example, HQ was reported to suppress the production or expression of various cytokines, such as TNF-␣ (also in Fig. 1C), interferon-␥, IL-1␤, IL-3, IL-6 (also in Fig. 1C), IL-10, and IL-12 in macrophage-like RAW264.7 cells and in lymphocytes under stimulation with LPS and keyhole limpet hemocyanin (7,8,12,30). In addition to its inhibitory effects on the production of various inflammatory mediators such as NO (also in Fig. 1A) and prostaglandin E 2 , HQ also negatively modulated receptor-mediated phagocytosis (12), lymphocyte proliferation (8,12), up-regulation of pattern recognition receptor levels (Fig. 1B), and enhanced differentiation during the Th2 response (9). These results led us to explore the molecular aspects of HQ inhibition and to focus on a common pathway required for cytokine production, phagocytosis, and inflammatory mediator release. The main target that is consistent with these observations is NF-B, a common pathway in the proinflammatory response. Numerous papers have also suggested it is involved in HQ-mediated inhibition of a variety of cellular responses (7,21,30,37). Even though recent articles suggested that HQ was capable of preventing NF-B from binding to its DNA promoter site (30,32), our data instead strongly implied that HQ suppressed the translocation of NF-B. Ma et al. (30) reported that HQ did not inhibit LPS-induced activation of IKK activity, degradation of IB␣, or translocation of activated NF-B into the nucleus, but HQ did block the formation of NF-B-DNA complexes. In contrast, we found that both phosphorylation of IB␣ and IKK␤, but not IKK␣, was completely blocked, and the level of p65 translocation was dose-dependently inhibited by HQ treatment. In addition, HQ-mediated inhibition of p65 nuclear translocation was also confirmed by confocal microscopic analysis (Fig. 2E). As Fig. 2C shows, the fact that phosphorylation of IKK␣ was clearly seen during HQ treatment seems to explain why IKK kinase activity was not blocked by HQ during previous studies (30). A signaling cascade that links IKK␤ and IB␣ to the translocation of NF-B has also been characterized by numerous findings (38,39).
The next question raised in this study is what is the target of HQ that mediates the block in NF-B translocation? Our data strongly indicate that Akt is the only protein involved in the immunosuppressive effects of HQ. The phosphorylation patterns of Akt (Fig. 3), its kinase activity (Fig. 4), and adduct formation between HQ and Akt-derived peptide fragments (Fig. 7) are strong examples of its involvement. Akt has been recently proposed to be an important component in immune responses in a large number of studies. Akt-specific small interfering RNA strongly blocked IB␣ phosphorylation and NF-B activation (17). Moreover, IKK␤ has been identified as a substrate of Akt, according to direct binding and phosphorylation assays (39 -41). The involvement of Akt in the inflammatory pathway has also been demonstrated using a variety of cell types (including dendritic cells, monocytes, and macrophages) (42,43) and by measuring several inflammatory parameters (including NO and TNF-␣) (44,45). LPS treatment increased the activity of phosphatidylinositol 3-kinase, PDK1, and Akt, as assessed by their phosphorylation levels and kinase activ- ities (22,46). Consequently, strong inhibitors of these pathways, such as LY294002, wortmannin, and Akt inhibitors, were clearly shown to block inflammatory events mediated by LPS (47,48). Nonetheless, Akt inhibitors are mostly developed as anti-cancer drugs because the predominant role of Akt has been more clearly demonstrated in cancer cell survival and proliferation phenomena (49,50). However, the accumulating evidence should motivate more researchers to develop promising anti-inflammatory drugs targeted to Akt.
The number of patents and papers on the Akt inhibitor development rapidly grows as more information on the functional role of Akt is attained. So far, two approaches have been generally used to develop Akt inhibitors. The first approach involves screening direct binding inhibitors targeted to the ATP binding site of Akt. Example compounds include GSK690693 (51), 6-phenylpurines (52), and conjugates of oligoarginine peptides with adenine, adenosine, adenosine-5Јcarboxylic acid, and 5-isoquinolinesulfonic acid (53). Although these compounds displayed strong inhibitory effects, other side effects would be expected due to non-selectivity because most kinases share a conserved amino acid sequence in the ATP binding site. Indeed, AGC family kinases, which include cAMPdependent protein kinase A, cGMP-dependent kinase, and protein kinase C, have been reported to have well conserved sequences in the ATP binding motif (54). The other way focuses on finding compounds that bind to selective allosteric sites on Akt. Indeed, most Akt inhibitor development trials have concentrated on this approach because stronger selectivity can be obtained. 2,3,5-Trisubstituted pyridine derivatives (55), 2,3-diphenylquinoxaline (56), and aminofurazans (57) are good examples of compounds with promising activity. In view of these, one of the recent interesting findings on the allosteric sites of Akt was the identification of specific amino acid residues, Cys-296 and Cys-310, in the activation (T-) loop of Akt (58). Thus, lactoquinomycin and related pyranonaphthoquinones were found to block Akt kinase activity by irreversibly binding to Cys-296 and Cys-310 in response to suppressing cancer cell proliferation (33,34). Using these compounds, it was shown that those amino acids and the adjacent residues could be another allosteric site in Akt. Similarly, in our study the benzene-type toxic molecule HQ was shown to clearly bind to Cys-310 by LC-MS-MS analyses (Fig. 7B). Furthermore, Akt-C310A, which is deficient in kinase activity as well as its phosphorylation compared with that of Akt-WT (Fig. 8B), was only partially suppressed by HQ, whereas the activity of WT was almost completely suppressed (Fig. 8C). Furthermore, HQ and its structural analogs have been shown to strongly block Akt phosphorylation and kinase activity (Fig. 9B) and NO production (Fig. 9C) without altering cell viability (Fig. 9E). In particular, structural modifications of HQ showed us that the activity of HQ could be improved by this strategy. Indeed, one derivative (JS-III-49) when orally administered at 50 mg/kg strongly suppressed lethality of septic shock (Fig. 10A), enhanced levels of serum TNF-␣ (Fig. 10B), liver damages (Fig.  10C), and ear swelling (Fig. 10D). Therefore, these results suggest that C-310 and the adjacent sequence in Akt could be drug design targets and applicable for development of novel drugs for Akt inhibition as suggested previously (33,34). Indeed, according to a docking model postulated from the structural data base of Akt (26), it is considered that HQ chemically adducted with Cys-310-Mer-EtOH can associate with Glu-298 by hydrogen bonds and with His-195 by hydrophobic interaction (Fig. 8D) to make an environment that interrupts the phosphorylation of Thr-308 in Akt by a structural hindrance.
It has been speculated that the reactivity of the pyranonaphthoquinone core of lactoquinomycin and frenolicin B could cause these compounds to directly bind to the T-loop cysteine(s) and consequently interfere with the catalytic activity of Akt through decreasing phosphorylation of Thr-308 (33). Indeed, phosphorylation of Thr-308 in Suntide by PDK1 was severely reduced when Cys-310 was replaced with Ala (14), indicating the important role of Cys-310 in the initial phosphorylation stage. Interestingly, HQ was not capable of directly binding to the cysteine in Suntide according to the HPLC analysis after incubating HQ and the peptide in PBS buffer (Fig. 7C) even though powerful thiol compounds such as DTT, L-cysteine, and N-acetyl-L-cysteine strikingly abrogated HQ-mediated inhibition of NO production (Fig. 5), Akt phosphorylation (Fig. 6A), and Akt kinase activity (Fig. 6C). However, mer-EtOH enabled HQ to bind to the thiol group of the cysteine in Suntide, generating a new chemical adduct molecule composed of Suntide, mer-EtOH, and HQ, according to the dose-dependent suppression of kinase activity (Fig. 7A) and LC-MS-MS analysis (Fig.  7B). Unlike the peptide analysis conditions, mer-EtOH was not used in the cellular experimental assays. Despite this, HQ still maintained its inhibitory activity, implying that some of cellular molecules with thiol groups may work similarly to mer-EtOH to make the active, conjugated form. Indeed, several conjugates including HQ-glutathione have been reported previously (59 -61), and the mitochondrial enzymes, such as NADP reductase and GSH reductase, which metabolize HQ, have already been well characterized (61,62). Furthermore, it has been shown that the enzyme activity of mitochondrial dehydrogenases was enhanced by HQ by up to 2.2-fold (data not shown). In particular, the fact that these molecules can act as toxins that could contribute to the toxicity of HQ seems to open the possibility that these conjugates might block Akt activity directly in a cellular system. This possibility also leads us to a new hypothesis that these toxic conjugates might be involved in mediating the various toxicological responses of HQ, such as immunosuppression and nephrotoxicity (60). In summary, we found that HQ-mediated down-regulation of inflammatory responses, such as NO production and cytokine gene expression from macrophage-like RAW264.7 cells and primary macrophages, was mediated by blocking NF-B activation via suppression of the translocation path-way composed of Akt, IKK␤, and IB␣. In particular, HQ was able to bind Cys-310 and interrupt the phosphorylation of both Thr-308 and Ser-473 in Akt by thiolation of the sulfhydryl group with mer-EtOH, according to the analysis of the interaction of HQ with Suntide. Therefore, our data suggest FIGURE 10. The effect of HQ derivative on septic shock-induced lethality and AA-induced ear edema formation. A, mice orally treated with HQ or JS-III-49 were intraperitoneally injected with D-GalN (600 mg/kg) and LPS (50 g/kg). Lethality was then observed for 7 days. CMC, sodium carboxylmethylcellulose. B, JS-III-49 and prednisolone were orally administered 30 min, and then D-GalN (600 mg/kg)/LPS (50 g/kg) was peritoneally injected, and blood was collected after 1.5 h. Serum TNF-␣ levels released were assayed by enzyme-linked immunosorbent assay. C, the protection effect of JS-III-49 on liver damage induced by the treatment of D-GalN (600 mg/kg) and LPS (50 g/kg) by was examined by histological analysis of liver sections as described under "Experimental Procedures." PV, portal vein; CV, central vein. D, mice orally treated with JS-III-49 were applied with AA (2%/ear). Ear swelling was then measured with dial thickness gauge after 4 h. Data (B and D) represent the mean Ϯ S.E. of three independent experiments performed in triplicate. The results (A and C) show one representative experiment of two done with seven mice. p Ͻ 0.05 (*) and p Ͻ 0.01 (**) compared with the control group.
that Akt and its target site Cys-310 can be considered a prime target of HQ and could be applied to novel anti-inflammatory drug development.