Aldose Reductase Mediates the Lipopolysaccharide-induced Release of Inflammatory Mediators in RAW264.7 Murine Macrophages*

Abnormal production of inflammatory cytokines and chemokines is a key feature of bacterial endotoxin, lipopolysaccharide (LPS)-induced inflammation, and cytotoxicity; however, the mechanisms regulating production of inflammatory markers remain unclear. Herein, we show that inhibition of the aldehyde-metabolizing enzyme aldose reductase (AR; AKR1B3) modulates NF-κB-dependent activation of inflammatory cytokines and chemokines in mouse serum, liver, heart, and spleen. Pharmacological inhibition or small interfering RNA ablation of AR prevented the biosynthesis of tumor necrosis factor-α, interleukin 1β, interleukin-6, macrophage-chemoattractant protein-1, and cyclooxygenase-2 and prostaglandin E2 in LPS-activated RAW264.7 murine macrophages. The AR inhibition or ablation significantly attenuated LPS-induced activation of protein kinase C (PKC) and phospholipase C (PLC), nuclear translocation of NF-κB, and phosphorylation and proteolytic degradation of IκBα in macrophages. Furthermore, treatment of macrophages with 4-hydroxy-trans-2-nonenal (HNE), and cell-permeable esters of glutathionyl-4-hydroxynonanal (GS-HNE) and glutathionyl-1,4-dihydroxynonane (GS-DHN) activated NF-κB and PLC/PKC. Pharmacological inhibition or antisense ablation of AR that catalyzes the reduction of GS-HNE to GS-DHN prevented PLC, PKC, IKKα/β, and NF-κB activation caused by HNE and GS-HNE, but not by GS-DHN, suggesting that reduced GS-lipid aldehydes catalyzed by AR propagate LPS-induced production of inflammatory markers. Collectively, these data provide evidence that inhibition of AR may be a significant therapeutic approach in preventing bacterial endotoxin-induced sepsis and tissue damage.

Bacterial lipopolysaccharide (LPS), 3 a proinflammatory endotoxin, is a component of the outer envelope of all Gram-negative bacteria (1). When Gram-negative bacteria multiply in the host, LPS is released into the circulation, where it is recognized by a variety of circulating cell types, triggering the induction of NF-B-dependent proinflammatory cytokines such as tumor necrosis factor-␣ (TNF-␣), interleukin (IL)-1, prostaglandins, and nitric oxide (1)(2)(3). Acting in an autocrine and paracrine manner, cytokines and chemokines induce and amplify the host response to bacterial infection (4,5). However, excessive cytokine release can have deleterious consequences. For example, septic shock triggered by LPS causes production of reactive oxygen species (ROS) and multiorgan dysfunction (2), with myocardial dysfunction being the major cause of morbidity and mortality (6). It has been shown that TNF-␣ is the earliest cytokine produced in large amounts in response to LPS and that it is the major cause of most of the effects of LPS (7,8). These studies are supported by the anti-TNF therapy that provides protection against the LPS-induced cytotoxicity (9,10). Recent studies have shown that the heart produces large amounts of cytokines, especially TNF-␣, IL-1␤, interferon-␥, macrophagechemoattractant protein-1 (MCP-1), and cyclooxygenase-2 (Cox-2) during septic shock and related pathologies (11,12). In addition, transgenic mice overexpressing TNF-␣ readily develop myocardial dysfunction (13). Furthermore, antioxidants such as N-acetylcysteine and butylated hydroxytoluene have been shown to attenuate LPS-induced activation of NF-B expression of inflammatory cytokines and endotoxemia, indicating that ROS are the obligatory mediators of LPS signaling (14 -16). Although these studies have provided possible therapeutic strategies in preventing LPS-induced toxicity, the mechanisms responsible for LPS-induced multiorgan failure remain poorly understood.
Our recent studies show remarkable and unexpected metabolic regulation of TNF-␣ signaling by the enzyme aldose reductase (AR; AKR1B1 in human, AKR1B4 in rat, and AKR1B3 in mouse), a member of aldo-keto reductase superfamily (17)(18)(19). AR reduces one of the most abundant and toxic lipid aldehydes, 4-hydroxy-trans-2-nonenol (HNE), to 1, 4dihydroxynonene (DHN) and its glutathione conjugate, GS-HNE, to GS-DHN (20,21). We have demonstrated that AR plays a pivotal role in the proliferation of vascular smooth mus-cle cells, apoptosis of vascular endothelial cells and restenosis of rat carotid artery (17,18,22). Inhibition of AR significantly decreases neointima formation in balloon-injured rat carotid arteries, and also diminishes the in situ activation of NF-B during restenosis (23). Our recent observations show that AR mediates the mitogenic and cytotoxic signals of cytokines and growth factors (17)(18)(19). Inhibition or ablation of AR attenuates TNF-␣and growth factor-induced IB␣ phosphorylation and degradation, activation of NF-B and PKC, proliferation of vascular smooth muscle cells, and also apoptosis of vascular endothelial cells and human lens epithelial cells (17)(18)(19). Similarly, ablation of AR attenuated the TNF-␣-induced expression of adhesion molecules, intercellular adhesion molecule and vascular cell adhesion molecule, and monocyte adhesion to vascular endothelial cells (18). However, the involvement of AR in inflammatory signals induced by LPS is not known. Therefore, we have investigated the effect of pharmacological inhibition or RNA interference ablation of AR on LPS-induced expression of various cytokines and chemokines and other inflammatory markers such as Cox-2 and prostaglandin E 2 (PGE 2 ) and examined the possible mechanism of AR mediation in LPS toxicity.
Our studies indicate possible therapeutic application of AR inhibitors as anti-inflammatory drugs to treat septic shock induced by Gram-negative bacterial infections.

EXPERIMENTAL PROCEDURES
Materials-Dulbecco's modified Eagle's medium, phosphate-buffered saline, penicillin/streptomycin solution, trypsin, and fetal bovine serum were purchased from Invitrogen. Sorbinil and zopolrestat were gifts from Pfizer, and tolrestat was obtained from American Home Products. Normal or phospho-specific antibodies against PLC-␤3, PLC-␥1, IKK, and IB␣ were obtained from Cell Signaling Inc. Mouse anti-rabbit glyceraldehyde-3-phosphate dehydrogenase antibodies were obtained from Research Diagnostics Inc. Protein-HNE antibodies, cyclooxygenase (Cox) activity assay and PGE 2 assay kits were obtained from Cayman Chemical Co. The colorimetric non-radioactive NF-B p65 Transcription Factor Assay kit was obtained from Chemicon Laboratories. NF-B SEAP reporter vector and control vector were obtained from Clontech Laboratories. Lipopolysaccharide (Escherichia coli) and the reagents used in Western blot analysis were obtained from Sigma. All other reagents used were of analytical grade.
Cell Culture and Animals-The Balb/c mice (25-30 g) were obtained from Taconic Laboratories and housed in pathogenfree conditions with free access to food and water at the institutional animal care facility. The RAW264.7 macrophage cell lines obtained from ATCC were grown in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum.
RNA Interference Ablation of AR in Macrophages-The ablation of AR mRNA was essentially carried out as described earlier (24). Briefly, RAW264.7 cells were incubated with serum-free medium containing the AR-siRNA (AATCGGTG-TCTCCAACTTCAA) or scrambled siRNA (AAAATCTCCC-TAAATCATACA; control) to a final concentration of 100 nM and the RNAiFect TM transfection reagent (Qiagen). After 15 min of incubation at 25°C, the medium was aspirated and replaced with fresh Dulbecco's modified Eagle's medium con-taining 10% serum. The cells were cultured for 48 h at 37°C, and AR expression was determined by measuring AR protein by Western blot analysis using anti-AR antibodies and by measuring AR activity in the total cell lysates (24).
Determination of Cytokine Levels-The mice were preinjected with sorbinil (25 mg/kg body weight, via the intraperitoneal route) or carrier for 24 h followed by LPS (4 g/kg body weight) injection. At different time intervals the animals were killed and blood and heart tissues were collected. The RAW264.7 cells were preincubated with 10 M sorbinil, tolrestat, or zopolrestat for 24 h followed by incubation with 1 g/ml LPS. The cytokines (TNF-␣, IL-6, IL-12, and interferon-␥) and chemokine (MCP-1) levels were measured in mice serum and homogenates of heart, liver, and spleen and also in the culture medium of RAW264.7 cells using BD Biosciences Mouse Inflammation Cytometric Bead Array Kits according to the manufacturer's instructions by fluorescence-activated cell sorter automation.
Cox Activity Assay-Macrophages (2 ϫ 10 5 cells/well in 6-well plates) were growth arrested in serum-free medium with or without AR inhibitors for 24 h followed by incubation with 1 g/ml LPS for another 24 h. Similarly AR siRNA or control siRNA-transfected macrophages were serum-starved for 24 h followed by incubation with 1 g/ml LPS for another 24 h. The macrophages were homogenized in cold buffer containing 0.1 M Tris-HCl, pH 7.8, and 1 mM EDTA and Cox activity was measured in 96-well plates according to the manufacturer's instructions (Cayman Chemical Co.). Briefly, 10 l of standard/ sample were incubated in the presence of arachidonic acid and colorimetric substrate, N,N,N,N-tetramethyl-p-phenylenediamine, in a total reaction volume of 210 l. The Cox peroxidase activity was measured colorimetrically by monitoring the appearance of oxidized N,N,N,N-tetramethyl-p-phenylenediamine at 590 nm using an ELISA reader (Packard).
Determination of NF-B Activation-The cytosolic as well as nuclear extracts were prepared as described before (17). The NF-B activity was determined using the calorimetric non-radioactive NF-B p65 Transcription Factor Assay kit (Chemicon Internationals) as per the supplier's instructions. Briefly, a double-stranded biotinylated oligonucleotide containing the consensus sequence for NF-B binding (5Ј-GGGACTTTCC-3Ј) was mixed with the nuclear extract assay buffer provided. After incubation, the mixture (probe ϩ extract ϩ buffer) was transferred to the streptavidin-coated ELISA plate. The ELISA reactions were developed, stopped, and read at 450 nm using an ELISA plate reader. For each experiment, triplicate samples were measured for statistical significance.
NF-B-dependent Reporter Gene Assay-The effect of AR inhibition/ablation on NF-B-dependent reporter gene transcription induced by LPS was analyzed by secretary alkaline phosphatase (SEAP) assay. Briefly, macrophage (5 ϫ 10 5 cells/ well) were plated in 6-well plates and transiently transfected with pNF-B-SEAP plasmid (250 ng) or control plasmid pTAL-SEAP for 6 h using the Lipofectamine Plus reagent. Subsequently, macrophages were treated with AR inhibitors and then stimulated with 1 g/ml LPS. Similarly, in other experiments simultaneous transfections were carried out with both AR siRNA as well as pNF-B-SEAP plasmid followed by incubation with 1 g/ml LPS. The cell culture medium was harvested after 48 h of LPS treatment and analyzed for SEAP activity according to the protocol essentially as described by the manufacturer (Clontech Labs) using a Packard microplate reader.
Determination of PKC Activity-The membrane-bound total PKC activity was measured by using the Promega SignaTECT TM PKC assay system according to the manufacturer's instructions and as described earlier (17). Briefly, aliquots of the reaction mixture (25 mM Tris-HCl, pH 7.5, 1.6 mg/ml phosphatidylserine, 0.16 mg/ml diacylglycerol, and 50 mM MgCl 2 ) were mixed with [␥-32 P]ATP (3,000 Ci/mmol, 10 Ci/l) and incubated at 30°C for 10 min. The extent of phosphorylation was detected by measuring radioactivity retained on the filter paper by using a scintillation counter.
Western Blot Analysis-An equal amount of macrophage cell extracts were separated on 12% SDS-PAGE, electroblotted on nitrocellulose membranes, and probed with specific antibodies against AR, Cox-1, Cox-2, PLC␤3, PLC␥1, IB-␣, IKK ␣, ␤, ␥, and HNE. Antibody binding was detected by enhanced picochemiluminescence (Pierce). Immunopositive bands were quantified using Kodak Image station 2000R loaded with Kodak one-dimensional image analysis software and the average changes in -fold intensities were calculated.
Determination of ROS-The serum-starved macrophages (1.5 ϫ 10 4 cells/well in a 24-well plate) without or with 10 M sorbinil or tolrestat were treated with the ROS-sensitive fluorophore 2Ј,7Ј-dichlorofluorescein diacetate for 30 min. Subsequently, macrophage were exposed to LPS (1 g/ml) for 60 min and fluorescence was measured with a CytoFluorII fluorescence plate reader (PerSeptive Biosystems, Inc., Framingham, MA) at excitation of 485 nm and emission of 528 nm.
Determination of Intracellular Lipid Peroxidation-The lipid peroxidation was determined by measuring total ␣,␤-unsaturated aldehyde (26) levels in RAW264.7 cells (1 ϫ 10 7 cells/well) treated without or with LPS in the absence and presence of sorbinil. The aldehydes were quantified colorimetrically using a lipid peroxidation kit (Bioxytech LPO-586 TM ) obtained from Oxford Biomedical Research, Oxford, MI, as per the supplier's instructions. Briefly, the determination is based on the reaction of the chromogenic reagent, methanesulfonic acid, with ␣,␤unsaturated aldehydes such as HNE at 45°C. One molecule of aldehyde reacts with two molecules of reagent to yield a stable chromophore with maximal absorbance at 586 nm.
Preparation of GS-aldehyde Esters-HNE was synthesized as described previously (20). The conjugate of glutathione ethyl ester with HNE (GS-HNE-ester) was prepared as described (25). Briefly, 1 mol of [4-3 H]HNE (55,000 cpm/nmol) was incubated with 5 mol of GSH ethyl ester in 0.1 M potassium phosphate, pH 7.0, for 1 h at room temperature. The reaction was monitored by following the decrease in absorbance at 224 nm. The GS-HNE-ester was purified by reverse phase HPLC. The reduced form of the esterified glutathione-HNE conjugate (GS-DHN-ester) was prepared by incubating 100 nmol of GS-HNE-ester with 300 nmol of NADPH and 100 g of AR in 0.1 M potassium phosphate, pH 6.0, for 3 h at 37°C. The reaction was monitored by following the consumption of NADPH at 340 nm. The GS-DHN-ester was separated from GS-HNE-ester by reverse phase HPLC using a Varian reverse phase ODS C 18 column pre-equilibrated with 0.1% aqueous trifluoroacetic acid. The compounds were eluted using a gradient consisting of solvent A (0.1% aqueous trifluoroacetic acid) and solvent B (100% acetonitrile) at a flow rate of 1 ml/min. The gradient was established such that solvent B reached 24% in 20 min, 26% in 30 min, and was held at this value for 10 min. In the next 10 min solvent B reached 60%, and in an additional 5 min it reached 100% where it was held for 10 min. Chemical identities of the GS-HNE and GS-DHN-esters were established by electrospray ionization mass spectrometry (ESI/MS) as described before (25). Statistical Analysis-Data are presented as mean Ϯ S.E. and the p values were determined using the unpaired Student's t test.
Inhibition of AR Prevents LPS-induced Cytokine Production in RAW264.7 Macrophages-To examine the mechanism of AR-mediated regulation of LPS-induced production of cytokines and chemokines in mice, and to exclude the possible nonspecific inhibition of other enzymes by sorbinil, we systematically examined the effect of pharmacological inhibition of AR with three structurally different inhibitors, sorbinil, tolrestat, and zopolrestat, and also by ablating AR message by RNA interference. We first determined the effect of AR inhibition/ablation on LPS-induced production of cytokines and chemokines in RAW 264.7 macrophages. As shown in Fig. 1, A-D (left panels), incubation of RAW264.7 macrophages with LPS for 16 h caused 28-, 11-, 50-, and 5-fold increases of TNF-␣, IL-1␤, IL-6, and MCP-1 levels, respectively, in culture medium and AR inhibitors prevented the increase by 80 -90%. AR inhibitors had no effect on the basal levels of inflammatory cytokines. These results thus suggest that AR inhibition prevents LPS-induced inflammatory marker production in RAW macrophages. The levels of interferon-␥ were not detectable in the macrophages even with LPS treatment.
Even though sorbinil, tolrestat, and zopolrestat are specific inhibitors of AR, their nonspecific effects cannot be rigorously ignored. Hence, we investigated the effect of genetic ablation of AR on LPS-induced production of cytokines and chemokines. As shown in Fig. 2, A and B, transfection of AR siRNA decreased Ͼ95% the AR protein levels as well as AR activity in RAW264.7 cells, but transfection with control siRNA did not affect the levels of AR protein as well as activity in macrophages, suggesting specific inhibition of the AR message by AR siRNA. Incubation of macrophages with LPS for 16 h caused significant 24-, TABLE 1 Effect

of AR inhibitor on LPS-induced cytokine and chemokine production in mice serum
The Balb/c mice (25 g) were preinjected intraperitoneally with sorbinil (25 mg/kg body weight/day) for 2 days or carrier followed by single intraperitoneal injection of LPS (4 g/kg body weight). Sorbinil injections were given at every 24 h after LPS injection. The mice were killed at 1, 3, and 7 days after LPS injection and serum was prepared from the blood. The cytokines/chemokines levels were measured in the serum by using BD Biosciences Mouse Inflammation Cytometric bead array kit. All data are expressed as mean Ϯ S.E. 11-, 55-, and 4-fold increases in TNF-␣, IL-1␤, IL-6, and MCP-1 levels, respectively, in the culture medium (Fig. 1, A-D, right  panels). Transfection of cells with AR-siRNA but not control siRNA significantly (80 -90%) prevented an LPS-induced increase in the levels of TNF-␣, IL-1␤, IL-6, and MCP-1. However, AR ablation alone did not change the basal levels of these inflammatory cytokines and chemokines in macrophages. To confirm results obtained with the bead array system, we have performed reverse transcriptase-PCR analysis. Incubation of macrophages with LPS for 4 h caused significant 3.5-, 6-, 5-, and 7-fold increases in TNF-␣, IL-6, IL-1␤, and MCP-1 mRNA levels, respectively (Fig. 3). Inhibition of AR with three structurally distinct inhibitors significantly (80 -90%) prevented LPS-induced increases in the mRNA levels of TNF-␣, IL-6, IL-1␤, and MCP-1 suggesting the involvement of AR in the LPS-mediated increase in inflammatory signals. AR Inhibition/Ablation Attenuates LPS-induced Cox-2 Expression and PGE 2 Production in RAW264.7 Cells-LPS caused a 17-fold increase in the biosynthesis of PGE 2 as com-

TABLE 2 Effect of AR inhibitor on LPS-induced cytokine and chemokine production in mice heart, spleen, and liver tissue homogenates
The Balb/c mice (25 g) were preinjected intraperitoneally with sorbinil (25 mg/kg body weight/day) for 2 days or carrier followed by a single intraperitoneal injection of LPS (4 g/kg body weight). Sorbinil injections were given at every 24 h after LPS injection. The mice were killed at 1, 3, and 7 days after LPS injection and liver, heart, and spleen tissues were dissected out and (10%) homogenates were prepared in phosphate-buffered saline. The cytokines/chemokines levels were measured in the homogenates by using BD Biosciences Mouse Inflammation Cytometric bead array kit. All data are expressed as mean Ϯ S.E. pared with untreated cells. However, when the macrophages were challenged with LPS in the presence of AR inhibitors or AR ablation only a 4-fold increase in the biosynthesis of PGE 2 was observed. AR inhibitors alone did not alter the basal biosynthesis of PGE 2 (Fig. 4, A and B). Because the biosynthesis of PGE 2 is catalyzed by Cox-1 and Cox-2 enzymes, we next measured the effect of AR inhibition and ablation on LPSinduced Cox activities. LPS increased Cox activity by 9-fold and inhibition or ablation of AR significantly (Ͼ80%) prevented the LPS-induced increase of Cox activity (Fig. 5, A  and B). Because the activity of Cox is contributed by Cox-1 (constitutive) and Cox-2 (inducible), we next measured the effect of AR inhibition on LPS-induced expression of these proteins in macrophages. As shown in Fig. 5C, the levels of Cox-1 were not affected by LPS but Cox-2 protein increased by ϳ3-fold (Fig. 5D). The LPS-induced increase in Cox-2 protein was abolished by AR inhibition.

AR Inhibition/Ablation Prevents LPS-induced Activation of NF-B in
Macrophages-Activation of the redox-sensitive transcription factor, NF-B, transcribes the genes necessary for induction of inflammatory Cox-2, cytokines, and chemokines. We therefore examined the effect of inhibition and ablation of AR on LPS-induced NF-B activation in RAW264.7 cells. Within 2 h of LPS addition to macrophages ϳ10-fold activation of NF-B was observed and the increase was significantly attenuated by AR inhibitors as well as siRNA ablation of AR (Fig. 6, A  and B). However, AR inhibitors as well as AR ablation had no effect on basal NF-B activity in macrophages. The inhibitory effect of AR on NF-B activity was further analyzed in macrophages transfected with the NF-B SEAP reporter plasmid by monitoring SEAP activity in response to LPS challenge. AR  Cells were growth-arrested in Dulbecco's modified Eagle's medium containing 0.1% serum with or without the indicated AR inhibitors and challenged with LPS (1 g/ml). The total RNA was isolated at 4 h and reverse transcriptase-PCR analysis was carried out using specific primers for the indicated cytokines as described under "Experimental Procedures." Equal amounts of PCR products were electrophoresed with 2% agarose, 1 ϫ TAE gels containing ethidium bromide. Reverse transcriptase-PCR analysis with GADPH served as control. inhibition or ablation also decreased the LPS-dependent activation of NF-B SEAP activity in macrophages transfected with pNF-B-SEAP plasmid but not in control pTALSEAP plasmid (Fig. 6, C and D). Modulation of NF-B by AR inhibitors was further supported by immunoblot analysis of p65 in cytoplasmic and nuclear extracts of LPS and LPS ϩ sorbinil-treated macrophages. As shown in Fig. 7A, LPS caused translocation of p65 from the cytoplasm to the nuclei within 10 min of LPS challenge. The LPS-induced nuclear localization of p65 was prevented by pretreating the cells with sorbinil, suggesting that AR inhibition acts upstream to nuclear translocation of NF-B. This conclusion is further supported by modulation of LPSinduced phosphorylation as well as degradation of IB-␣ by AR inhibition (Fig. 7B).
Inhibition of IKK Activities by Aldose Reductase Inhibitor-Macrophages were stimulated with LPS in the presence or absence of the AR inhibitor, sorbinil and whole cell extracts were tested for the phosphorylation status of IKKs using phosphospecific antibodies for IKK, which recognize phosphorylation of both ␣ and ␤ forms of IKK. LPS significantly increased the phosphorylation of IKK␣/␤ and AR inhibition prevented it (Fig. 7C). However, LPS alone or LPS ϩ sorbinil did not alter the expression of IKK ␣, ␤, and ␥ isozymes (Fig. 7D). These observations suggest that AR inhibition prevents NF-B DNA binding as well as its transcriptional activity by inhibiting phospho-   rylation of IKK␣/␤ activities, which indicates the possible involvement of upstream kinases in the activation of IKK.
The Inhibition/Ablation of AR Prevents LPS-induced PKC Activity in RAW264.7 Macrophages-Inhibition or ablation of AR inhibited the LPS-induced phosphorylation of total membrane-bound PKC (Fig. 8, A and B). Because PLC is upstream to PKC, we examined the effect of AR inhibition on LPS-induced PLC activity. The maximum activities of PLC-␤3 and -␥1 were observed at 5 and 20 min, respectively, after incubation of macrophages with LPS, which were inhibited by AR inhibition (Fig.  8, C and D). These observations indicate that the inhibitory mechanism of AR inhibitors on NF-B activation may be via the PLC-PKC pathway.
Inhibition of AR Prevents LPS-induced ROS Production-Because LPS is known to increase ROS, we next examined the effect of AR inhibition on LPS-induced generation of ROS. As shown in Fig. 9A, treatment of RAW264.7 cells with LPS caused a significant increase in ROS levels in 60 min and AR inhibition prevented it. Because ROS cause peroxidation of membrane lipids resulting in the formation of lipid aldehydes such as HNE, which could readily conjugate with glutathione (GSH) and both HNE and GS-HNE can be reduced by AR, we investigated the effect of AR inhibition on LPS-induced lipid peroxidation caused by LPS in macrophages. As expected, LPS increased the levels of ␣,␤-unsaturated aldehydes and protein-HNE adducts by nearly 3-fold within 6 h (Fig. 9, B and C) and inhibition of AR slightly increased the levels of ␣,␤-unsaturated aldehydes and protein-HNE adducts.
Reduced Glutathione-Aldehyde Conjugates Catalyzed by AR Propagate LPS Signals in Macrophages-Because the host response to LPS is known to be mediated by ROS (14 -16), we asked if AR-mediated reduction of toxic lipid aldehydes and their glutathione conjugates such as GS-DHN could activate   10A and 11A). Inhibition or ablation of AR significantly blunted the effects of HNE/GS-HNE on IKK-␣/␤ phosphorylation and NF-B activation but had no effect on the ability of GS-DHN, the already reduced form of GS-HNE, to activate NF-B (Figs. 10A and 11A). This suggested that GS-DHN is sufficient for NF-B activation and may be involved in IKK-␣/␤ phosphorylation.
To determine whether GS-DHN serves as a cellular sensor of ROS-induced insults, we examined its effects on the phosphorylation events upstream of IKK/NF-B activation in RAW264.7 macrophages. After GS-DHN challenge, the activity of PKC was increased by ϳ2.5-fold within 60 min (Fig. 10B). GS-DHN also induced the phosphorylation of PLC-␤3 and PLC-␥1 (Fig. 11, B and C), which activated PKC, but had no effect on total PLC protein (not shown). As expected, HNE and GS-HNE had similar effects on the phosphorylation of the kinases upstream of NF-B (Figs. 10B and 11, A-C). However, pharmacologic inhibition or siRNA-mediated ablation of AR significantly decreased the HNE-and GS-HNE-induced phosphorylation of PLC, PKC, and IKK, but had no effect on GS-DHN-initiated phosphorylation of PLC and its downstream kinases. These findings suggest that glutathione-lipid alcohol (such as GS-DHN) formed by the reduction of glutathione-lipid aldehyde (such as GS-HNE) catalyzed by AR could be an obligatory mediator of LPS-induced inflammation.

DISCUSSION
To the best of our knowledge, the present study is the first to demonstrate that inhibition of AR significantly improves LPSinduced cytotoxicity leading to formation of inflammatory cytokines. LPS, an endotoxin found in the outer membrane of Gram-negative bacteria, is a major trigger of septic shock (1-3), which is, in part, a consequence of the hosts response to overwhelming bacterial infection. Sepsis is characterized by microvascular thrombosis, decreased organ perfusion, and organ ischemia, which leads to multiorgan dysfunction and death (2,5,26). Innate immune cells such as macrophages recognize the presence of invading bacteria and initiate the host response by releasing cytokines and chemokines (6). When cytokines in the inflammatory cells are present in effective concentrations, pathogens are removed without adverse consequences; however, excessive amounts of inflammatory cytokines lead to septic shock and death (2,27,28). A decrease in cardiac muscle contractility is the major cause of mortality and morbidity in sepsis (8,9). Despite substantial advances in antimicrobial therapy, the mortality in severe sepsis continues to be ϳ40%, reflecting the limited therapeutic options (29).
The mechanisms underlying the multiorgan failure, especially heart failure that occurs during septic shock have been the subject of intense investigation. It is well known that activation of redox-sensitive transcription factors such as NF-B and AP-1 are involved in the pathologies associated with LPS-induced sepsis (30 -32). However, the precise mechanisms of LPS signaling leading to the activation of these transcription factors are not known. Even though multiple studies suggest that antioxidant therapy prevents LPS-induced septic shock (14 -16), the mechanisms by which ROS mediate inflammatory signals leading to NF-B activation are not clearly understood. Because lipid peroxidation initiated by ROS generates toxic aldehydes that conjugate with glutathione and can be reduced by AR (33), we systematically investigated the effect of inhibiting LPS-induced cytotoxicity without or with inhibition or ablation of AR in macrophages. Inhibition of AR prevented the LPS-induced activation of NF-B and release of proinflammatory cytokines, indicating that such inhibitors could be used therapeutically to treat bacterial sepsis. Similar results with AR ablation indicate that pharmacological inhibition is equally efficient.
Rodents have been used extensively as an experimental animal model for LPS-induced septic shock (34 -36), where increased levels of serum cytokines, in particular TNF-␣ are observed (7,8). Therefore, we investigated the effect of AR inhibition on LPS-induced serum cytokine and chemokine levels using a mouse model of sepsis. Consistent with the results obtained in cultured macrophages, severely elevated levels of serum cytokines subsequent to LPS challenge were markedly suppressed by AR inhibition in vivo, suggesting an anti-inflammatory role for AR inhibitors in mice. Various reports from experimental models of endotoxin challenge and patients in septic shock have shown multiorgan dysfunction (2,37,38). The major factors that influence multiorgan dysfunction include cytokines such as TNF-␣, IL-1␤, and IL-6, and other inflammatory mediators such as nitric oxide, prostaglandins, Cox-2, and lipid metabolites (11,12). The cytokines are generally considered to be circulating molecules, but tissues such as heart, liver, and spleen also synthesize TNF-␣, IL-1␤, and IL-6 (36,37). In fact, increased local production of cytokines in the heart has been strongly implicated in myocardial dysfunction observed during various pathological conditions (11,12). Increased synthesis and autocrine secretion of inflammatory cytokines may have local consequences on cardiac contractile protein mRNA expression. For example, IL-1␤ activates, whereas IL-6 represses contractile protein genes (27,39). Whether the autocrine production of cytokines, in addition to cytokines produced as a consequence of the systemic inflammatory response, contributes to the multiorgan dysfunction in sepsis is unknown. Because, serum levels of LPS-induced cytokines are inhibited by AR inhibition; we investigated the effect of AR inhibition on cardiac, hepatic, and spleenic levels of cytokines. Indeed, inhibition of AR resulted in the inhibition of cytokines and other inflammatory factors generated locally in the tissues indicating a major role of AR in the regulation of LPS-induced multiorgan dysfunction. Taken together, this study is the first to demonstrate that AR inhibition prevents endotoxin-induced organ dysfunction.
Prevention of LPS-induced increase in the activation of NF-B and activation of protein kinases upstream of NF-B such as IKK and PKC by AR inhibition or ablation suggests that AR inhibition could attenuate LPS-induced inflammation by dis-rupting a signaling cascade upstream of PKC/PLC, which causes synthesis of cytokines and chemokines via NF-B. These results are in agreement with our earlier observations that inhibition of AR prevents cytokine and hyperglycemia-induced NF-B signaling in vascular smooth muscle cells, vascular endothelial cells, and human lens epithelial cells (17-19, 23, 24). Even though AR inhibition attenuated phosphorylation of PLC, it was not clear how AR could mediate ROS signals. Because, we have shown earlier that AR reduces lipid aldehydes and their conjugates with glutathione to corresponding alcohols (20,21), it was possible that the reduced form of lipid aldehydes or their glutathione conjugates act as signaling intermediates. To investigate this possibility, we examined the effect of AR inhibition/ablation on HNE, and cell permeable esters of GS-HNE and GS-DHN-induced cell signaling. Our observations that AR inhibition/ablation prevented HNE-and GS-HNE-ester-induced signals and subsequent activation of NF-B/PKC/IKK/PLC, but GS-DHN-ester signaling was resistant to AR inhibition, suggest a novel role for a reduced glutathione-lipid aldehyde conjugate (such as GS-DHN) as an obligatory mediator of ROS-induced cytotoxicity. Our recent study indicates that GS-HNE and GS-DHN-esters readily enter in the cells where esterases cleave off ester group (25).
In summary, this study shows that AR-catalyzed reduction of lipid aldehydes produced by ROS in response to endotoxin is necessary for activation of the signaling cascade leading to NF-B nuclear translocation and increased production of inflammatory cytokines. The observation that inhibition of AR dramatically attenuates LPS-induced cytokine production in vitro and in vivo suggests that modulation of lipid aldehyde reduction could provide a cellular approach for preventing the maladaptive host response to bacterial infections and other NF-B-dependent cellular events. Specific inhibitors of AR could be used for short term therapy of severe bacterial toxininduced sepsis and associated inflammatory processes.