Advances in Alcoholic Liver Diseasedownloads.hindawi.com/journals/specialissues/682927.pdf ·...

International Journal of Hepatology

Guest Editors: Natalia Osna, Kusum Kharbanda, Laura Schrum, and Angela Dolganiuc

Advances in Alcoholic Liver Disease

International Journal of Hepatology


Guest Editors: Natalia Osna, Kusum Kharbanda, Laura Schrum,and Angela Dolganiuc

Copyright © 2012 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “International Journal of Hepatology.” All articles are open access articles distributed under theCreative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided theoriginal work is properly cited.

Editorial Board

Chul Ahn, USAAntonio Ascione, ItalyMatthias J. Bahr, GermanySimon Bramhall, UKMaria Buti, SpainUmberto Cillo, ItalyHeather Francis, USAHikaru Fujioka, JapanJunji Furuse, Japan

Matthias Glanemann, GermanyShannon Glaser, USAFredric D. Gordon, USAClaus Hellerbrand, GermanyMasahiko Hirota, JapanPaloma Jara, SpainRoberto Lupi, ItalyShigeru Marubashi, JapanKojiro Michitaka, Japan

Daisuke Morioka, JapanGuy W. Neff, USALun-Xiu Qin, ChinaMiguel A. Serra, SpainPierluigi Toniutto, ItalyTakuji Torimura, JapanRoberto I. Troisi, BelgiumDirk Uhlmann, GermanyYo-ichi Yamashita, Japan

Contents

Advances in Alcoholic Liver Disease, Natalia Osna, Kusum Kharbanda, Laura Schrum,and Angela DolganiucVolume 2012, Article ID 563018, 1 page

Lipid Droplet Accumulation and Impaired Fat Efflux in Polarized Hepatic Cells: Consequences ofEthanol Metabolism, Benita L. McVicker, Karuna Rasineni, Dean J. Tuma, Mark A. McNiven,and Carol A. CaseyVolume 2012, Article ID 978136, 8 pages

Oxidative Stress and Inflammation: Essential Partners in Alcoholic Liver Disease, Aditya Ambade andPranoti MandrekarVolume 2012, Article ID 853175, 9 pages

Markers of Inflammation and Fibrosis in Alcoholic Hepatitis and Viral Hepatitis C, Manuela G. Neuman,Hemda Schmilovitz-Weiss, Nir Hilzenrat, Marc Bourliere, Patrick Marcellin, Cristhian Trepo, Tony Mazulli,George Moussa, Ankit Patel, Asad A. Baig, and Lawrence CohenVolume 2012, Article ID 231210, 10 pages

Cyanamide Potentiates the Ethanol-Induced Impairment of Receptor-Mediated Endocytosis ina Recombinant Hepatic Cell Line Expressing Alcohol Dehydrogenase Activity, Dahn L. Clemens,Dean J. Tuma, and Carol A. CaseyVolume 2012, Article ID 954157, 5 pages

MicroRNA Signature in Alcoholic Liver Disease, Shashi Bala and Gyongyi SzaboVolume 2012, Article ID 498232, 6 pages

Betaine Treatment Attenuates Chronic Ethanol-Induced Hepatic Steatosis and Alterations to theMitochondrial Respiratory Chain Proteome, Kusum K. Kharbanda, Sandra L. Todero, Adrienne L. King,Natalia A. Osna, Benita L. McVicker, Dean J. Tuma, James L. Wisecarver, and Shannon M. BaileyVolume 2012, Article ID 962183, 10 pages

Dose-Dependent Change in Elimination Kinetics of Ethanol due to Shift of Dominant MetabolizingEnzyme from ADH 1 (Class I) to ADH 3 (Class III) in Mouse, Takeshi Haseba, Kouji Kameyama,Keiko Mashimo, and Youkichi OhnoVolume 2012, Article ID 408190, 8 pages

Autologous Bone Marrow Stem Cells in the Treatment of Chronic Liver Disease, Madhava Pai,Duncan Spalding, Feng Xi, and Nagy HabibVolume 2012, Article ID 307165, 7 pages

Alcohol Activates TGF-Beta but Inhibits BMP Receptor-Mediated Smad Signaling and Smad4 Binding toHepcidin Promoter in the Liver, Lisa Nicole Gerjevic, Na Liu, Sizhao Lu, and Duygu Dee Harrison-FindikVolume 2012, Article ID 459278, 11 pages

CYP2E1 Sensitizes the Liver to LPS- and TNF α-Induced Toxicity via Elevated Oxidative and NitrosativeStress and Activation of ASK-1 and JNK Mitogen-Activated Kinases, Arthur I. Cederbaum, Lili Yang,Xiaodong Wang, and Defeng WuVolume 2012, Article ID 582790, 19 pages

Role of Adaptive Immunity in Alcoholic Liver Disease, Emanuele AlbanoVolume 2012, Article ID 893026, 7 pages

Aberrant Hepatic Methionine Metabolism and Gene Methylation in the Pathogenesis and Treatment ofAlcoholic Steatohepatitis, Charles H. Halsted and Valentina MediciVolume 2012, Article ID 959746, 7 pages

Hindawi Publishing CorporationInternational Journal of HepatologyVolume 2012, Article ID 563018, 1 pagedoi:10.1155/2012/563018

Editorial


Natalia Osna,1 Kusum Kharbanda,2 Laura Schrum,3 and Angela Dolganiuc4

1 Department Internal Medicine, Omaha VA Medical Center, University of Nebraska Medical Center and Liver Study Unit,Omaha, NE 68198, USA

2 Liver Study Unit, Omaha VA Medical Center, Omaha, NE 68105, USA3 Department of Internal Medicine, The Liver-Biliary-Pancreatic Center, Cannon Research Center, Carolinas Medical Center,Charlotte, NC 28203, USA

4 Department of Medicine, University of Florida, Gainesville, FL 32611, USA

Correspondence should be addressed to Natalia Osna, [email protected]

Received 23 January 2012; Accepted 23 January 2012

Copyright © 2012 Natalia Osna et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This special issue reflects on multiple factors/mechanismsinvolved in the pathogenesis of ALD. Alcoholic liver injuryis known to cause a broad range of liver abnormalities.Alcohol is primarily metabolized in the hepatocyte leadingto increased secretion of inflammatory mediators, which, inturn, activate and/or influence the response of the nonparen-chymal cells (NPCs) (hepatic stellate cells, Kupffer cells, andsinusoidal endothelial cells) and subsequently control thedegree of liver injury.

This issue overviews general aspects of ALD, suchas oxidative stress and inflammation (A. Ambade and P.Mandrekar) as well as the molecular aspects of these events,including the role of ethanol-metabolizing enzymes, ADH(T. Haseba et al.), and CYP2E1 (A. I. Cederbaum et al.) inthe development of alcohol-induced liver injury. In addition,signaling mechanisms induced by alcohol are examined inthe paper by L. N. Gerjevic et al. As a separate pathogenicaspect, the role of microRNA in ALD is analyzed in the paperby S. Bala and G. Szabo.

The consequences of alcohol-related liver injury, suchas impairment of receptor-mediated endocytosis and lipiddroplet accumulation, are presented in experimental in vitrostudies of C. A. Casey et al. and B. McVicker et al., respec-tively.

One of the mechanisms that affects various liver celltypes and affect disease progression is impairment of methy-lation reactions. In our special issue, the role of impairedmethylation in pathogenesis of steatohepatitis as well astreatment modalities with promethylating agent, betaine, isdiscussed in the paper by C. H. Halsted and V. Medici and K.Kharbanda et al.

Liver also serves as an immune organ and accommodatesa wide variety of cells, including immune cells. The latterconsists of dendritic cells (DCs), natural killer (NK) cells, andlymphocytes, which are present in normal livers. Selectiverecruitment and retention of certain immune populationsoccurs during diverse liver diseases, and these cells playa critical role in the development and resolution of liverinflammation, remodeling, and destruction and actively par-ticipate in immune defense. The role of adaptive immunityin ALD development is overviewed by E. Albano. Also, therole of stem cells in ALD treatment is discussed in the paperby M. Pai et al. Finally, to underline the role of alcohol inprogression of chronic infections (HCV), we included thepaper by M. Neuman et al. which reflects on the markers ofinflammation and fibrosis in alcoholic hepatitis and hepatitisC.

Each paper received external blind review in addition toour reviews.

This special issue covers exciting new areas of ALDpathogenesis and treatment and is strongly recommendedfor the clinicians and basic scientists involved in alcohol re-search.

Natalia OsnaKusum Kharbanda

Laura SchrumAngela Dolganiuc

Hindawi Publishing CorporationInternational Journal of HepatologyVolume 2012, Article ID 978136, 8 pagesdoi:10.1155/2012/978136

Research Article

Lipid Droplet Accumulation and Impaired Fat Efflux inPolarized Hepatic Cells: Consequences of Ethanol Metabolism

Benita L. McVicker,1, 2 Karuna Rasineni,1, 2 Dean J. Tuma,1, 2

Mark A. McNiven,3 and Carol A. Casey1, 2

1 Liver Study Unit-Research Service (151), VA Nebraska-Western Iowa Health Care System, 4101 Woolworth Avenue, Omaha,NE 68105, USA

2 Departments of Internal Medicine and Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha,NE 68198, USA

3 Department of Biochemistry and Molecular Biology, Mayo Clinic College of Medicine, Rochester, MN 55905, USA

Correspondence should be addressed to Carol A. Casey, [email protected]

Received 12 September 2011; Revised 22 November 2011; Accepted 8 December 2011

Academic Editor: Angela Dolganiuc

Copyright © 2012 Benita L. McVicker et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Steatosis, an early manifestation in alcoholic liver disease, is associated with the accumulation of hepatocellular lipid droplets(LDs). However, the role ethanol metabolism has in LD formation and turnover remains undefined. Here, we assessed LDdynamics following ethanol and oleic acid treatment to ethanol-metabolizing WIF-B cells (a hybrid of human fibroblasts (WI38) and Fao rat hepatoma cells). An OA dose-dependent increase in triglyceride and stained lipids was identified which doubled(P < 0.05) in the presence of ethanol. This effect was blunted with the inclusion of an alcohol metabolism inhibitor. The ethanol/OA combination also induced adipophilin, LD coat protein involved in the attenuation of lipolysis. Additionally, ethanol treatmentresulted in a significant reduction in lipid efflux. These data demonstrate that the metabolism of ethanol in hepatic cells is relatedto LD accumulation, impaired fat efflux, and enhancements in LD-associated proteins. These alterations in LD dynamics maycontribute to ethanol-mediated defects in hepatocellular LD regulation and the formation of steatosis.

1. Introduction

Alcohol abuse and alcoholic liver disease (ALD) are majorhealth problems both in the USA and worldwide. The mostprevalent manifestations of ALD are the presence of fattyliver (hepatic steatosis), alcoholic hepatitis, and cirrhosis. Ofthese manifestations, it is noted that hepatic steatosis is areversible early stage of ALD whose presence has been relatedto the liver’s enhanced sensitivity to damaging triggers suchas oxidative stress and endotoxins [1, 2]. Thus, the aberrantcontent of lipids in hepatocytes can act as a key “first hit”in the progression of ALD making lipid accumulation aprime target for therapeutic intervention. Remarkably, littleis known about the regulatory mechanisms involved in theaccumulation of intracellular lipids which are stored indynamic organelles called lipid droplets (LDs). Furthermore,it is unclear how LD formation, degradation (lipolysis), or

export is affected, particularly in the hepatocyte by theadverse effects of alcohol exposure and the metabolism ofethanol.

The organelle identified as having a central role in theaccumulation of lipids in hepatocytes is the LD. LDs areintracellular stores of neutral lipids, predominately choles-terol esters, and triglycerides that are bound by a phospho-lipid monolayer [3]. Long considered to be inert, LDs haverecently attracted great interest as dynamic structures at thehub of lipid and energy metabolism. In general, LDs arethought to originate from the endoplasmic reticulum fromwhere they are trafficked through the cytoplasm, interactingwith various organelles and transporting lipids as the energyneeds of the cell dictate [4, 5]. In the healthy liver, LDs in hep-atocytes play a crucial role in the packaging and distributionof lipids as lipoproteins [6]. However, in disease states suchas ALD, the accumulation of LDs (hepatic steatosis) is likely

2 International Journal of Hepatology

due to disruption in those packaging and distribution rolesas alcohol exposure induces impairments in LD formation,degradation, and/or export processes. Ultimately, excessiveLD accumulation occurs in hepatocytes which can lead tolipotoxicity with consequences of inflammation and subse-quent cell death. However, little is known about how ethanolexposure/metabolism alters the regulation of LD accumula-tion and degradative processes in the hepatocyte.

The liver and, to a lesser extent, the gastrointestinal tractare the main sites of alcohol metabolism. Within the liver,there are two main pathways of alcohol metabolism, alcoholdehydrogenase (ADH) and cytochrome P-450 2E1 (CYP2E1)[7–9]. Relating alcohol metabolism and generated toxicproducts (i.e., acetaldehyde) to mechanisms by which etha-nol causes fatty liver appears to be complex. Historically,it has been proposed that reducing equivalents generatedduring ethanol oxidation inhibit steps of the tricarboxylicacid cycle and oxidation, thereby inhibiting fatty acid oxida-tion [10, 11]. Another proposed mechanism involved in thedevelopment of alcoholic fatty liver is enhanced lipogenesiswith support of several studies demonstrating significantincreases observed in hepatic lipogenesis following chronicethanol administration [12–14]. Alternatively, it has beenshown that the persistence of fatty liver may involve theinhibition of lipoprotein export mechanisms, possibly viaformation of acetaldehyde protein adducts and associatedalterations to the microtubule network in the cell [15,16]. Thus, the production of toxic metabolites of ethanolis thought to play a significant role in altering the traf-ficking and utilization of LDs in hepatocytes. Indeed, itnow appears that similar to previously identified ethanol-mediated impairments to hepatocellular receptors, ligands,and endocytic processes [17–19], LDs may be regulated byinteractions involving classical trafficking pathways and aretherefore highly susceptible to damaging alterations inducedby ethanol metabolites. To better define the role of ethanolmetabolism in hepatic steatosis, we analyzed LD accumu-lation and impaired fat efflux in hepatoma cell cultures(WIF-B cells), a well-established in vitro model for studyingthe consequences of ethanol metabolism on hepatocellulartrafficking events.

WIF-B cells are differentiated cells of hepatic origin thatare a hybrid clone of human fibroblasts (WI38) crossed withrat hepatoma cells [20]. WIF-B cells exhibit long-term viabil-ity in culture, develop a hepatocellular-polarized phenotype,and express human genes coding for liver-specific proteins(e.g., albumin and fibrinogen) [21, 22]. The WIF-B cellshave been shown to adequately mimic in vivo hepatocellularfunctions such as polarity, protein secretion, and transport[23–25]. Additionally, our laboratory has demonstrated thatWIF-B cells are an ideal in vitro model for studying theeffect of ethanol on cellular processes as the cells were foundto exhibit ADH and CYP2E1 activities allowing for theefficient metabolism of ethanol [26]. In our previous work,we characterized several ethanol-mediated cellular defectsin the alcohol-treated WIF-B cells which linked ethanolmetabolism to apoptotic-inducing pathways and with thepotential involvement of altered targeting of proteins [27].In this study, alcohol-treated WIF-B cells were used to char-

acterize the role ethanol metabolism has in the generation,accumulation, efflux, and lipolysis of LDs.

2. Materials and Methods

2.1. Materials. F-12 Coon’s modified culture medium, 4-methylpyrazole (4MP), Oil Red O (ORO), oleic acid (OA),and fatty acid free bovine serum albumin (BSA) wereobtained from Sigma Chemical Co. (St. Louis, MO). Heat-inactivated fetalplex was obtained from Gemini Bio-Pro-ducts (Woodland, CA). BODIPY 493/503 was purchasedfrom Invitrogen (Carlsbad, California). Buffered formalinand isopropanol were obtained from Fisher Scientific (Pitts-burgh, PA). UltraCruz mounting media containing 4,6-diamidino-2-phenylindole (DAPI) was obtained from SantaCruz Biotechnology, Inc (Santa Cruz, CA). All other materi-als were reagent grade.

2.2. WIF-B Cell Culture and Treatment. WIF-B cells werecultured in F-12 media containing 3.5% heat-inactivatedfetalplex in a 7% CO2 atmosphere as described previously[27]. Briefly, the cells were seeded on sterilized glass cover-slips or directly in tissue culture dishes and cultured for 6days to obtain a maximal-polarized phenotype prior to thevarious treatments (ethanol ± OA). It has previously beendetermined that OA, a long chain free fatty acid, significantlyinduces LD formation in hepatocytes [28]. For fatty acidtreatment to WIF-B cells, OA was conjugated with BSA(1.5%) in serum-free F-12 media prior to addition to thecell cultures. In general, confluent and polarized WIF-Bcultures were treated with media (serum-free F-12 with 1.5%BSA) with and without OA (100 μM–1000 μM), 25–50 mMethanol and/or 0.25 mM 4MP, an inhibitor of alcohol dehy-drogenase. The cells were plated on coverslips and stainedwith BODIPY for microscopic analysis of LDs. In othercell cultures, LD formation in WIF-B cells was quantifiedfollowing extraction and spectrophotometric detection ofORO from the stained culture dishes.

2.3. Triglyceride Extraction and Analysis. Extraction of tri-glycerides was performed using the Folch method [29]with slight modifications. Briefly, after exposure to 25 mMethanol for 48 hours, WIF-B cells (approximately 3 ×106 cells/60 mm dish) were rinsed with PBS, harvested byscraping, and the pellet reconstituted in PBS. An aliquot wassaved for protein/DNA determination with the remainingextracted with the addition of chloroform/methanol (2 : 1)followed by vortexing for 20–30 seconds. The sample wasfiltered over Whatman number 1 filter paper with a furtherrinse with 1 mL chloroform/methanol. The final volume ofchloroform/methanol was recorded. Aliquots (1 mL) weremade and the samples were dried completely using a Cen-trivac. Following drying, the triglycerides were hydrolyzedby the addition of 95% Ethanol and 8.0 M KOH at 65◦C for20 minutes. The triglyceride content was determined usingTriglyceride Reagent (Thermo Scientific) as directed by themanufacturer with detection made by spectrophotometricanalysis (Beckman DU640).

International Journal of Hepatology 3

2.4. Oil Red O Staining. ORO staining was performed aspreviously described [30] with minor modifications. Inbrief, WIF-B cultures were fixed in 10% buffered formalin,incubated with 60% isopropanol, and stained of 10 minuteswith ORO solution (0.21% dye in 100% isopropanol).Following staining, the cultures were washed five times withsterile water, the dye is extracted using isopropanol, andthe concentration of ORO in the extract was measuredcolorimetrically (500 nm). Results were expressed as OD/mgprotein or DNA. In several experiments, images of OROstained cells were obtained prior to extraction using anOlympus IX70 microscope in combination with a MicroFiredigital camera (Image Processing Solutions, North Reading,MA) at ×100 and ×200 magnifications.

2.5. Oil Red O-Based Quantification of Fat Efflux. For rapidand convenient quantification of fat efflux from treated cells,WIF-B cultures were rinsed twice (PBS) and the mediachanged to oleate-free F-12 media with or without 25 mMethanol. The plates were sealed and allowed to incubate foran additional 24 hr followed by staining with ORO for quan-tification of LD accumulation. The efflux of fat from the cellswas determined by comparing the amount of ORO taken upby the cells before and after removal of oleate from the media.

2.6. Fluorescence Microscopy. To measure LD content byBODIPY staining, WIF-B cells that were cultured on glasscoverslips were subjected to the various treatments, fixed for20 min with formaldehyde and briefly (2 min) permeabilizedwith D-PBS + 0.1% Triton X-100. Following incubationwith 5 μg/mL BODIPY 493/503 in PBS, the coverslips werewashed and mounted on glass slides using DAPI-containingUltraCruz mounting medium. Cells were viewed with aNikon ECLIPSE 80i Microscope equipped with a Nikon DS-Qi1Mc digital camera (Boyce Scientific, Inc., Gray Summit,MO). Images were processed using NIS-Elements ImagingSoftware.

2.7. Western Blot Analysis. Cell protein was obtained fromWIF-B cells by homogenization in 0.25 M sucrose in 5 mMTris-HCl, pH 7.5 containing protease inhibitor cocktail(Sigma, St. Louis, MO). Cell protein was resolved on 12%reduced gels by SDS-PAGE and transferred onto nitro-cellulose membranes. The blots were blocked for 1 hr inOdyssey blocking buffer (LI-COR Biosciences, Lincoln, NE)at room temperature and subsequently probed overnight at4◦C with primary antibodies, mouse antiadipophilin/ADRP(Fitzgerald, Acton, MA) at 1/500 dilution, and rabbit anti-ratGAPDH (Santa Cruz, Santa Cruz, CA) at a 1/5000 dilution.The blots were then incubated with secondary antibodies(IRDye680 goat anti-rabbit IgG and IRDye 800CW goat anti-mouse IgG) (LI-COR Biosciences, Lincoln, NE) at 1/10,000dilution. Following washing, the blots were scanned andquantified using the Odyssey Infrared Imager (LI-CORBiosciences, Lincoln, NE).

2.8. Real-Time Polymerase Chain Reaction (PCR). RNA wasisolated from the WIF-B cells using a PureLink RNA Mini

Kit (Invitrogen, Carlsbad, CA) according to the man-ufacturer’s instructions. The concentration and quality(260/280 ratio) of the RNA was determined by a NanoDropSpectrophotometer (NanaoDrop Technologies, Wilmington,DE). Real-time PCR reactions were performed using Tagmangene expression assay for rat adipophilin (Cat numberRN01472318 m1) and rat actin (Cat number 4352931E) pur-chased from Applied Biosystems, Carlsbad, CA. Detectionwas performed using a 7500 Real Time PCR System (AppliedBiosystems). The delta-delta Ct method was used to deter-mine the fold change using actin for normalization.

2.9. Statistical Analysis. Results are expressed as mean ±SEM. Comparison of paired values was performed using theStudents t- test with values P < 0.05 being considered signif-icant. Comparisons among groups of data were made usingone-way ANOVA with Tukey’s post hoc test; P < 0.05 wasconsidered significant.

3. Results

3.1. Effects of Fatty Acids and Ethanol on Lipid Accumulationin WIF-B Cells. It is known that alcoholic fatty liver is anearly consequence of alcohol consumption. Central to thiscondition is the accumulation of lipids such as cholesterolesters and triglycerides that are packaged in lipid droplet(LD) organelles in hepatocytes [31–33]. However, the roleethanol metabolism has in steatosis and LD dynamicsremains to be clarified. Here we analyzed fat accumulation inethanol-metabolizing hepatoma hybrid (WIF-B) cells tobetter define steatosis at the cellular level. Initially, we inves-tigated what effect ethanol treatment with or withoutthe addition of exogenous free fatty acids would haveon hepatocellular triglyceride levels. Specifically, WIF-Bcultures were treated up to 48 hours with ethanol and/oroleic acid (OA), a monosaturated omega-9 fatty acid thathas previously been identified as having a role in hepaticsteatosis [28]. As expected, cellular triglyceride levels werefound to be increased in a dose-dependent manner withOA treatment (Figure 1). Also, the addition of ethanol intothe cell cultures resulted in a 2- to 3-fold enhancementin triglyceride levels over the concentration range of OAtreatment (Figure 1). This noted enhancement of cellulartriglycerides in the ethanol and OA-treated cells correlatedwith the accumulation of neutral lipids packaged intocytoplasmic LDs. This was demonstrated using an ORO-based colorimetric quantitative assay which detected theconcentration-dependent elevations in vesicular lipid con-tent following oleate treatment that doubled in the presenceof ethanol (Figure 2). Microscopic analysis subsequent toBODIPY staining of the cytoplasmic LDs paralleled theobserved OA and ethanol-induced increases in triglycerideand ORO-stained lipids (Figure 3).

3.2. Enhanced Fat Accumulation in WIF-B Cultures RequiresEthanol Metabolism. The role ethanol metabolism by alcoholdehydrogenase (ADH) has in the formation of LDs was testedby including a specific ADH inhibitor, 4-methyl pyrazole(4-MP) in the OA/ethanol-treated cultures. In WIF-B cells


0

10

20

30

40

50

60

70

80

90

0 160 330 660 1000

Oleic acid concentration (μM)

ControlEthanol

Trig

lyce

ride

s (n

mol

)/m

illio

n c

ells

Figure 1: Ethanol-induced triglyceride accumulation in oleate-treated WIF-B cells. Cells were treated for 48 hr with or without25 mM ethanol in presence of different concentrations of oleic acid.Values are means ± SEM (N = 4). ∗Significantly different (P <0.05) from control group.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 200 400 600 800 1000

Abs

orba

nce

(OD

500)


ControlEthanol

Figure 2: Oil Red O-based quantification of ethanol and oleicacid-induced fat accumulation in hepatoma hybrid cells. WIF-Bcells were treated for 48 hr with or without 25 mM ethanol inpresence of different concentrations of oleic acid and lipid contentquantified with Oil Red staining. Values are means ± SEM (N = 4).∗Significantly different (P < 0.05) from control group.

treated with oleic acid alone (control cells), the presence of 4-MP had no effect on cellular levels of neutral lipids detectedby ORO staining (Figure 4(a)). However, in ethanol-treatedcells, the inclusion of 4-MP reduced the accumulationof neutral lipids to control levels (Figure 4(b)). Theseresults demonstrate that the metabolism of ethanol andthe subsequent formation of reactive metabolites is associ-ated with lipid droplet accumulation in the hepatoma cells.

3.3. Ethanol-Induced Effects Associated with Cellular LipidRetention Include Alterations in Lipolysis, Lipid Efflux, andCell Survival Mechanisms. As just described, the combinedeffects of ethanol and OA to WIF-B cells result in the sub-stantial retention of lipids in the form of LDs. It is predictedthat the ethanol-mediated lipid retention is, in part, due toimpaired degradation of the accumulating LDs. In supportof this prediction, we found that the combined treatmentof ethanol and OA significantly elevated the expression of aLD-associated protein, adipophilin (Figure 5). Adipophilin,otherwise known as adipose differentiation-related protein(ADRP), is a well-characterized LD protein that is knownto be involved in LD homeostasis particularly by playinga role in the attenuation of lipolysis [5, 34]. In additionto alterations observed in lipolytic mechanisms, it was alsoobserved that ethanol and OA treatment resulted in asignificant reduction in lipid efflux from the cells. The datain Figure 6 reflect this finding as significantly less of theaccumulated lipid was released from ethanol-treated cellsfollowing starvation compared to those treated with OAalone. And finally, to determine if ethanol-induced steatosiscould result in cell injury, we measured apoptosis in theethanol- and OA-treated WIF-B cells. It was determinedthat the induction of hepatocellular apoptosis correlated theobserved increase in LD accumulation. Specifically, in OA-and ethanol-treated cells where we observed the highestincrease in LD accumulation, the activity of a key executionerenzyme of programmed cell death mechanisms (caspase-3) was significantly enhanced (2 to 3-fold, P < 0.05) inthe presence of ethanol compared to OA alone-treated cells(Figure 7).

4. Discussion

The accumulation of lipids in the liver is an early pathologicalstage in the development of alcoholic liver disease (ALD) thatoccurs in most individuals that chronically consume alcohol[35, 36]. Furthermore, ethanol-induced fatty infiltration hasbeen suggested to sensitize the liver to damaging risk factorssuch as oxidative stress and prodeath signaling mechanisms.However, the mechanisms involved in fatty liver diseaserepresented as hepatocellular steatosis, and particularly theaccumulation and/or regulation of lipid droplets (LDs) inhepatocytes, remains to be elucidated.

It is known that the accumulation of excessive lipid inhepatocytes can be related to alterations in mechanismsinvolving the uptake, synthesis, and esterification of free fattyacids [1, 37]. Also, the damaging effects of ethanol have beenimplicated in impairments in lipid degradation (lipolysis) aswell as secretory mechanisms. Indeed, ethanol consumptionhas been linked to altered triglyceride and phospholipidsynthesis, impairments in fatty acid oxidation, and thesecretion of very low-density lipoproteins (VLDLs) [12–14,38]. However, details remain to be determined concerningLD formation, accumulation, and lipolysis in ethanol-damaged hepatocytes. To date work has been completeddescribing LD accumulation in human and animal modelsalong with the role of ADRP in LD maturation [39–41].Also, a recent study has defined specific ethanol-mediated


+750 μM oleic acid

Con

trol

Eth

anol

(-) oleic acid

Figure 3: Lipid droplet accumulation in ethanol and oleic acid-treated cultures. WIF-B cells were treated without (control) or with 25 mMethanol in the presence of 750 uM oleic acid followed by immunohistochemical analysis of the presence of lipid droplets by BODIPY staining.

0

0.2

0.4

0.6

0.8

1

1.2

0 200 750

Abs

orba

nce

(OD

500)


Control

− 4MP

+4MP

(a)

0

0.2

0.4

0.6

0.8

1

1.2

0 200 750

Abs

orba

nce

(OD

500)


− 4MP+4MP

Ethanol

(b)

Figure 4: Ethanol and oleic acid induced fat accumulation requires ethanol metabolism. Oil Red O staining after 48 hours in the presenceof increasing concentrations of oleic acid either with or without the addition of 0.25 mM 4-methylpyrazole (4MP) to inhibit ethanolmetabolism. Results were from the ORO-based detection of neutral lipids in (a) oleate alone-treated (control) cells and (b) WIF-B culturestreated with both oleic acid and 25 mM ethanol. Values are means ± SEM (N = 4). ∗Significantly different (P < 0.05) from control group.

alterations in LD protein properties that are related to LDformation in steatosis [42]. Here, we contribute to the studyof ethanol-induced LD accumulation by demonstratingthat as a consequence of ethanol metabolism in polarizedhepatoma cells, LD enhancement is related to changes inlipid efflux and ultimately cell survival. Importantly, theseeffects were demonstrated using WIF-B cells which are awell-characterized model to study hepatocyte protein traf-

ficking machinery as well as the biological basis of ethanol-induced fatty liver. The WIF-B cells are natural ethanol-metabolizing cells and have been shown to accumulatetriglycerides when exposed to ethanol in a manner similarto that observed in ethanol-fed animals. Thus, the WIF-Bcells are an ideal model system to decipher consequencesof ethanol-mediated enhancements in cellular triglyceridelevels and related storage in lipid droplet organelles. Indeed,


O E25 + ONone E25

Adipophilin

GAPDH

(a)

0

1000

2000

3000

4000

5000

6000

7000

8000

C E25 O

#

E25 + O

Adi

poph

ilin

pro

tein

(co

ntr

ol %

)

(b)

C E25 O E25 + O0

100200300400500600700800900

1000 #

Adi

poph

ilin

mR

NA

(co

ntr

ol %

)

(c)

Figure 5: Adipophilin (ADRP) protein content and mRNA expres-sion in WIF-B cells. WIF-B cells were cultured for 48 hours in theabsence (control media, C) or presence of 25 mM ethanol (E25)with or without the presence of 0.5 mM oleic acid (O and E25+ O). (a) Representative Western blot indicating the presence ofADRP and glyceraldehyde-3-phosphate dehydrogenase (GAPDH)as the loading control. (b) Relative content of adipophilin proteinexpressed as percent of control from four independent experiments.(c) Adipophilin mRNA expression detected in the WIF-B cells.∗Significantly different (P < 0.05) from control group and E25group. #Significantly different from all other groups (C, E25, andO).

we observed a dose-dependent increase in LD accumulationin ethanol- and oleic-acid-treated WIF-B cells shown by thequantitative assessment of cellular triglycerides and stainingof neutral lipids. Additionally, this observed ethanol-inducedlipid retention in WIF-B cells was found to be related tochanges in LD protein dynamics and cellular lipid efflux. Itis known that hepatocytes can take up long chain fatty acidssuch as oleic acid which can then be esterified to neutrallipids (cholesterol esters and triglycerides) and packagedinto and stored as phospholipid-covered LD organelles [3].Under normal or fasting conditions, hepatocytes effectivelymetabolize and degrade the stored LDs. However, underhepatocellular damaging conditions (ethanol exposure), the

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

InitialPost 24 hr (no ethanol)

Post 24 hr + ethanolPost 24 hr + ethanol + 4MP

#

OD

500

nm

C +750 μM oleic

Figure 6: Fat efflux from hepatic cells is impaired by ethanoltreatment. WIF-B cell cultures were “loaded” with fat for 48 hours,washed, and then reintroduced into fresh media. Oil Red O stainingwas measured before and after a 24-hour “washout” period. Dataare expressed as fat content pre- and post-washout as intensities ofOil Red O for quantification. Values are means ± SEM (N = 4).∗Significantly different (P < 0.05) from control (initial fat load)group. #Significantly different from all other groups.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Cas

pase

3 a

ctiv

ity

(Nm

oles

/mg

prot

ein

)

C E25 O E25 + O

Figure 7: Ethanol and oleate-induced apoptosis in polarizedhepatoma cultures. WIF-B cells were cultured for 48 hours in theabsence (control media, C) or presence of 25 mM ethanol (E25)with or without the presence of 0.5 mM oleic acid (O and E25+ O). Apoptotic cell death was measured by spectrofluorometricdetection of caspase-3 activity. Results are expressed as nanomolesof detected fluorometric substrate (AMC) cleaved and releasedby active caspase-3 enzyme per mg protein for four independentexperiments. ∗Significantly different (P < 0.05) from control group.

storage of fatty acids is enhanced leading to the accumulationof LDs. The increase in cytoplasmic LDs was thoughtto be the result of ethanol-induced alterations in lipidmetabolism. The work presented here shows that enhancedtriglycerides and LD formation are also a consequence of theethanol metabolism in the cell that results in changes to LDproperties. To aid in our understanding of this observed hep-atocellular LD retention, future work using the WIF-B cellmodel may contribute to the correlation of ethanol-mediatedchanges to hepatocyte membrane trafficking processes andthe attenuation of LD disassembly.


The study of LD biology is an emerging area of investiga-tion fueled by the knowledge that the dysregulation of neu-tral lipid stores is linked to a variety of disease states includingalcoholic liver disease as we have described here. It is alsonot surprising that in those disease states, cellular LDs arethought of as active organelles whose composition, biogene-sis, trafficking, and storage/degradation have been found tobe complex involving a varied proteome and genome. Inthis study, we reported the effect ethanol metabolism has onone of the abundant and well-studied LD proteins, ADRP.Because of the availability of antibodies to LD proteins andprotocols to identify LDs (e.g., ORO-based assays and BOD-IPY staining), we were able to show that ADRP protein levelsand hence LDs were significantly increased when treated withethanol, oleic acid, or both. The increase in ADRP appearsto be partially controlled at the level of transcription asADRP mRNA expression was increased as well. Therefore,since ADRP is well known for its role in lipid homeostasisparticularly by attenuating lipolysis, our data demonstratethat the ethanol-amplified LD accumulation in hepaticcells involves impairments in the breakdown of lipids asa consequence of alcohol metabolism. We were also ableto show that the metabolism of ethanol was necessary foralcohol-mediated effects on fat efflux. Particularly, the effluxof lipid stores in oleate-loaded cells was found to be signifi-cantly decreased when the cells were starved in the presenceof ethanol. Moreover, the ethanol-mediated reduction inlipid efflux was abrogated in the presence of the alcohol dehy-drogenase inhibitor, 4-MP. And finally, we demonstrated thatthe ethanol-mediated enhanced presence of LD organellesmay be involved in the promotion of hepatocyte toxicityas the lack of proper fatty acid mobilization/secretion maylead to cell death signaling. Indeed, a measure of hepaticapoptosis was found to be increased in WIF-B cultures thatwere treated with ethanol and oleic acid, a finding thatcorrelated to the observed accumulation of LDs under thesame conditions. Ongoing work is aimed at delineatingcontributing mechanisms involved in the observed LDaccumulation associated with the metabolism of ethanol. Inparticular, it is of interest to determine the role of ethanol-mediated alterations in LD-vesicle trafficking machinery ofthe hepatocyte as it is likely that the regulation of LDsinvolves modulation of vesicle-based degradative processes.

In summary, our findings suggest that the metabolism ofethanol in hepatocytes is directly related to impaired lipidlipolysis and fat efflux that contribute to increased hepaticLD accumulation. Moreover, we have identified that hepaticcell death may be a potential consequence of LD dysregu-lation and accumulation due to ethanol’s alterations in LDproperties. In addition to yielding therapeutic leads for alco-holic liver disease, the knowledge gained here may possiblybe extended to other diseases that involve lipid accumulation,including nonalcoholic fatty liver disease, atherosclerosis,diabetes, and cancer. These conditions affect millions ofAmericans and are major health concerns. Future workexamining how ethanol impacts LD vesiculation, trafficking,and targeting within the hepatocyte will likely contribute tothe study of LD biology and our understanding of fatty liverdisease.

Abbreviations

ALD: alcoholic liver diseaseLDs: lipid dropletsWIF-B: hybrid of human fibroblast (WI

38) and rat hepatoma (Fao) cells4-MP: 4-methylpyrazoleADRP: adipophilinADH: alcohol dehydrogenaseCYP2E: cytochrome P-450 2E1.

Acknowledgment

This paper was supported by NIH/NIAAA funding, 5RCIAA019032, 1RO1 AA020735-01 (Drs. C. A. Casey and M. A.McNiven, Multiple PI awards) and K01AA015577 (Dr. BenitaMcVicker) and the Department of Veterans Affairs.

References

[1] Y. Lu, J. Zhuge, X. Wang, J. Bai, and A. I. Cederbaum, “Cy-tochrome P450 2E1 contributes to ethanol-induced fatty liverin mice,” Hepatology, vol. 47, no. 5, pp. 1483–1494, 2008.

[2] S. Yang, H. Lin, and A. M. Diehl, “Fatty liver vulnerabilityto endotoxin-induced damage despite NF-κB induction andinhibited caspase 3 activation,” American Journal of Physiology,vol. 281, no. 2, pp. G382–G392, 2001.

[3] S. Martin and R. G. Parton, “Lipid droplets: a unified view ofa dynamic organelle,” Nature Reviews Molecular Cell Biology,vol. 7, no. 5, pp. 373–378, 2006.

[4] H. Weidberg, E. Shvets, and Z. Elazar, “Lipophagy: selectivecatabolism designed for lipids,” Developmental Cell, vol. 16,no. 5, pp. 628–630, 2009.

[5] R. Zimmermann, A. Lass, G. Haemmerle, and R. Zechner,“Fate of fat: the role of adipose triglyceride lipase in lipolysis,”Biochimica et Biophysica Acta, vol. 1791, no. 6, pp. 494–500,2009.

[6] G. F. Gibbons, K. Islam, and R. J. Pease, “Mobilisation of tri-acylglycerol stores,” Biochimica et Biophysica Acta, vol. 1483,no. 1, pp. 37–57, 2000.

[7] K. E. Crow, N. W. Cornell, and R. L. Veech, “The rate of etha-nol metabolism in isolated rat hepatocytes,” Alcoholism: Clini-cal and Experimental Research, vol. 1, no. 1, pp. 43–50, 1977.

[8] R. B. Jennett, D. J. Tuma, and M. F. Sorrell, “Effect of ethanoland its metabolites on microtubule formation,” Pharmacology,vol. 21, no. 5, pp. 363–368, 1980.

[9] M. X. Yang and A. I. Cederbaum, “Glycerol increases contentand activity of human cytochrome P-4502E1 in a transducedHepG2 cell line by protein stabilization,” Alcoholism: Clinicaland Experimental Research, vol. 21, no. 2, pp. 340–347, 1997.

[10] A. I. Cederbaum, C. S. Lieber, D. S. Beattie, and E. Rubin, “Ef-fect of chronic ethanol ingestion on fatty acid oxidation byhepatic mitochondria,” Journal of Biological Chemistry, vol.250, no. 13, pp. 5122–5129, 1975.

[11] C. S. Lieber and M. Savolainen, “Ethanol and lipids,” Alco-holism: Clinical and Experimental Research, vol. 8, no. 4, pp.409–423, 1984.

[12] M. Guzman and J. Castro, “Zonal heterogeneity of the effectsof chronic ethanol feeding on hepatic fatty acid metabolism,”Hepatology, vol. 12, no. 5, pp. 1098–1105, 1990.

[13] C. Ji, C. Chan, and N. Kaplowitz, “Predominant role ofsterol response element binding proteins (SREBP) lipogenic


pathways in hepatic steatosis in the murine intragastric etha-nol feeding model,” Journal of Hepatology, vol. 45, no. 5, pp.717–724, 2006.

[14] H. Tsukamoto, G. Lew, and E. C. Larkin, “Hepatic origin oftriglycerides in fatty livers produced by the continuous intra-gastric infusion of an ethanol diet,” Lipids, vol. 19, no. 6, pp.419–422, 1984.

[15] K. Kervinen, S. Horkko, W. F. Beltz, and Y. A. Kesaniemi,“Modification of VLDL apoprotein B by acetaldehyde altersapoprotein B metabolism,” Alcohol, vol. 12, no. 3, pp. 189–194,1995.

[16] D. J. Tuma, S. L. Smith, and M. F. Sorrell, “Acetaldehyde andmicrotubules,” Annals of the New York Academy of Sciences, vol.625, pp. 786–792, 1991.

[17] C. A. Casey, S. L. Kragskow, M. F. Sorrell, and D. J. Tuma,“Chronic ethanol administration impairs the binding andendocytosis of asialo-orosomucoid in isolated hepatocytes,”Journal of Biological Chemistry, vol. 262, no. 6, pp. 2704–2710,1987.

[18] B. L. McVicker and C. A. Casey, “Effects of ethanol on re-ceptor-mediated endocytosis in the liver,” Alcohol, vol. 19, no.3, pp. 255–260, 1999.

[19] B. L. Tworek, D. J. Tuma, and C. A. Casey, “Decreased bindingof asialoglycoproteins to hepatocytes from ethanol-fed rats:consequence of both impaired synthesis and inactivation ofthe asialoglycoprotein receptor,” Journal of Biological Chem-istry, vol. 271, no. 5, pp. 2531–2538, 1996.

[20] M. R. Shanks, D. Cassio, O. Lecoq, and A. L. Hubbard, “Animproved polarized rat hepatoma hybrid cell line. Generationand comparison with its hepatoma relatives and hepatocytesin vivo,” Journal of Cell Science, vol. 107, no. 4, pp. 813–825,1994.

[21] V. Bender, P. Bravo, C. Decaens, and D. Cassio, “The structuraland functional polarity of the hepatic human/rat hybrid WIF-B is a stable and dominant trait,” Hepatology, vol. 30, no. 4, pp.1002–1010, 1999.

[22] C. Decaens, P. Rodriguez, C. Bouchaud, and D. Cassio, “Estab-lishment of hepatic cell polarity in the rat hepatoma-humanfibroblast hybrid WIF-B9: a biphasic phenomenon going froma simple epithelial polarized phenotype to an hepatic polarizedone,” Journal of Cell Science, vol. 109, no. 6, pp. 1623–1635,1996.

[23] G. Ihrke, G. V. Martin, M. R. Shanks, M. Schrader, T. A.Schroer, and A. L. Hubbard, “Apical plasma membrane pro-teins and endolyn-78 travel through a subapical compartmentin polarized WIF-B hepatocytes,” Journal of Cell Biology, vol.141, no. 1, pp. 115–133, 1998.

[24] R. A. Joseph, B. D. Shepard, G. T. Kannarkat, T. M. Rutledge,D. J. Tuma, and P. L. Tuma, “Microtubule acetylation andstability may explain alcohol-induced alterations in hepaticprotein trafficking,” Hepatology, vol. 47, no. 5, pp. 1745–1753,2008.

[25] B. D. Shepard, R. A. Joseph, G. T. Kannarkat, T. M. Rutledge,D. J. Tuma, and P. L. Tuma, “Alcohol-induced alterations inhepatic microtubule dynamics can be explained by impairedhistone deacetylase 6 function,” Hepatology, vol. 48, no. 5, pp.1671–1679, 2008.

[26] C. S. Schaffert, S. L. Todero, B. L. McVicker, P. L. Tuma, M. F.Sorrell, and D. J. Tuma, “WIF-B cells as a model for alcohol-induced hepatocyte injury,” Biochemical Pharmacology, vol.67, no. 11, pp. 2167–2174, 2004.

[27] B. L. McVicker, D. J. Tuma, J. L. Kubik, P. L. Tuma, andC. A. Casey, “Ethanol-induced apoptosis in polarized hepatic

cells possibly through regulation of the Fas pathway,” Alco-holismClinical and Experimental Research, vol. 30, no. 11, pp.1906–1915, 2006.

[28] Y. Fujimoto, J. Onoduka, K. J. Homma et al., “Long-chainfatty acids induce lipid droplet formation in a cultured humanhepatocyte in a manner dependent of acyl-CoA synthetase,”Biological and Pharmaceutical Bulletin, vol. 29, no. 11, pp.2174–2180, 2006.

[29] J. Folch, M. Lees, and G. H. Sloane Stanley, “A simple methodfor the isolation and purification of total lipides from animaltissues,” The Journal of Biological Chemistry, vol. 226, no. 1, pp.497–509, 1957.

[30] W. Cui, S. L. Chen, and K. Q. Hu, “Quantification and mech-anisms of oleic acid-induced steatosis in HepG2 cells,” Ameri-can Journal of Translational Research, vol. 2, no. 1, pp. 95–104,2010.

[31] F. F. Chao, D. L. Stiers, and J. A. Ontko, “Hepatocellular tri-glyceride synthesis and transfer to lipid droplets and nascentvery low density lipoproteins,” Journal of Lipid Research, vol.27, no. 11, pp. 1174–1181, 1986.

[32] K. M. Mak, C. Ren, A. Ponomarenko, Q. Cao, and C. S. Lieber,“Adipose differentiation-related protein is a reliable lipiddroplet marker in alcoholic fatty liver of rats,” Alcoholism:Clinical and Experimental Research, vol. 32, no. 4, pp. 683–689,2008.

[33] T. J. Peters and S. R. Cairns, “Analysis and subcellular localiza-tion of lipid in alcoholic liver disease,” Alcohol, vol. 2, no. 3, pp.447–451, 1985.

[34] P. E. Bickel, J. T. Tansey, and M. A. Welte, “PAT proteins, anancient family of lipid droplet proteins that regulate cellularlipid stores,” Biochimica et Biophysica Acta, vol. 1791, no. 6,pp. 419–440, 2009.

[35] E. Albano, “Alcohol, oxidative stress and free radical damage,”Proceedings of the Nutrition Society, vol. 65, no. 3, pp. 278–290,2006.

[36] C. P. Day and O. F. W. James, “Steatohepatitis: a tale of two’Hits’?” Gastroenterology, vol. 114, no. 4, pp. 842–845, 1998.

[37] M. You, M. Fischer, M. A. Deeg, and D. W. Crabb, “Ethanolinduces fatty acid synthesis pathways by activation of sterolregulatory element-binding protein (SREBP),” Journal ofBiological Chemistry, vol. 277, no. 32, pp. 29342–29347, 2002.

[38] T. M. Donohue Jr., “Alcohol-induced steatosis in liver cells,”World Journal of Gastroenterology, vol. 13, no. 37, pp. 4974–4978, 2007.

[39] B. H. J. Chang, L. Li, A. Paul et al., “Protection against fattyliver but normal adipogenesis in mice lacking adipose dif-ferentiation-related protein,” Molecular and Cellular Biology,vol. 26, no. 3, pp. 1063–1076, 2006.

[40] Y. Imai, G. M. Varela, M. B. Jackson, M. J. Graham, R. M.Crooke, and R. S. Ahima, “Reduction of hepatosteatosis andlipid levels by an adipose differentiation-related protein anti-sense oligonucleotide,” Gastroenterology, vol. 132, no. 5, pp.1947–1954, 2007.

[41] B. K. Straub, P. Stoeffel, H. Heid, R. Zimbelmann, and P.Schirmacher, “Differential pattern of lipid droplet-associatedproteins and de novo perilipin expression in hepatocytesteatogenesis,” Hepatology, vol. 47, no. 6, pp. 1936–1946, 2008.

[42] D. J. Orlicky, J. R. Roede, E. Bales et al., “Chronic ethanol con-sumption in mice alters hepatocyte lipid droplet properties,”Alcoholism: Clinical and Experimental Research, vol. 35, no. 6,pp. 1020–1033, 2011.


Review Article

Oxidative Stress and Inflammation: Essential Partners inAlcoholic Liver Disease

Aditya Ambade and Pranoti Mandrekar

Department of Medicine, University of Massachusetts Medical School, 364 Plantation Street, Worcester, MA 01605, USA

Correspondence should be addressed to Pranoti Mandrekar, [email protected]

Received 13 September 2011; Revised 18 December 2011; Accepted 19 December 2011

Academic Editor: Laura Schrum

Copyright © 2012 A. Ambade and P. Mandrekar. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Alcoholic liver disease (ALD) is a multifaceted disease that is characterized by hepatic steatosis or fat deposition and hepatitis orinflammation. Over the past decade, multiple lines of evidence have emerged on the mechanisms associated with ALD. The keymechanisms identified so far are sensitization to gut-derived endotoxin/lipopolysaccharide resulting in proinflammatory cytokineproduction and cellular stress due to oxidative processes, contributing to the development and progression of disease. Whileoxidative stress and inflammatory responses are studied independently in ALD, mechanisms linking these two processes play amajor role in pathogenesis of disease. Here we review major players of oxidative stress and inflammation and highlight signalingintermediates regulated by oxidative stress that provokes proinflammatory responses in alcoholic liver disease.

1. Introduction

The pathogenesis of alcoholic liver disease (ALD) is a conse-quence of chronic alcohol abuse and approximately 44% ofthe 26,000 deaths from cirrhosis are due to ALD in the UnitedStates [1]. Alcoholic hepatitis, the clinical presentation ofALD, remains to be a common life threatening cause ofliver failure, especially when it is severe. Chronic alcoholconsumption has long been associated with progressiveliver disease from steatosis to inflammation, developmentof hepatic cirrhosis, and the subsequent increased risk ofhepatocellular carcinoma. Several studies have attempted toidentify the molecular pathways, direct or indirect, affectedby alcohol exposure in the liver. These pathways rangefrom oxidative stress, metabolism-related effects, and inflam-mation to apoptosis. Induction of oxidative stress andactivation of the inflammatory cascade are identified as keyelements in the pathophysiology of ALD [2]. While theseintracellular mechanisms affected by alcohol are studiedexclusively, the interplay of signaling molecules betweenpathways leading to alcoholic liver disease has received lessattention. Unraveling these interactions of oxidative stressmediators and inflammatory signaling in the liver will aidin identification of new integrative approaches as it relatesto alcoholic liver injury and provide potential new directions

to develop therapeutic target intervention. The goal of thisconcise paper is to first review alcohol-induced reactiveoxygen species and oxidative stress generated by alcoholmetabolism, endoplasmic reticulum stress, mitochondrialROS in the liver, protein adduct formation, and autophagyand chaperone function and then to describe stress-mediatedactivation of receptors, kinases, and transcription factorsresulting in proinflammatory signaling in ALD.

2. Classical Mechanisms ofAlcoholic Liver Disease

Research done, so far, on the effects of pathophysiologicalmechanisms of alcoholic liver disease suggests the involve-ment of two main liver cell types, resident macrophages, orKupffer cells and hepatocytes. The role of gut-derived endo-toxin and liver macrophage activation is clearly established inALD by Thurman and colleagues [2]. The deleterious effectsof alcohol, attributed to its metabolism, primarily occurin hepatocytes [3]. Alcohol metabolism pathway includinginduction of cytochrome P450 2E1 [3] results in adductformation and generation of reactive oxygen radicals respec-tively creating an oxidative microenvironment and damagein the liver [2]. In the currently accepted model of ALD,chronic alcohol induces oxidative stress and sensitization


to endotoxin, which activates the CD14/TLR4 pathway anddownstream signaling resulting in proinflammatory cytokineproduction [4]. The proinflammatory cytokines, particularlyTNFα, then provoke hepatocellular injury and death byextrinsic, via TNFR1 [5] and intrinsic death pathways [6]leading to ALD. While the role of oxidative stress andmacrophage activation, the two main pathophysiologicalprocesses affected in ALD, were studied independently inthe past, recent studies suggest that these pathways areinterconnected in ALD improving our understanding of thedisease.

3. Reactive Oxygen Species (ROS) and Alcohol

While activation of inflammatory responses are centralto alcoholic liver injury, excessive generation of reactiveoxygen species plays an equally significant role in alcohol-induced cellular damage [7]. Alcohol-induced liver disease isassociated with a state of “oxidative stress”. The metabolismof alcohol by alcohol dehydrogenase [ADH] leads to for-mation of acetaldehyde. Further, the acetaldehyde is metab-olized to acetate by acetaldehyde dehydrogenase [ALDH].Acetaldehyde, a reactive intermediate has an ability to formadducts with DNA [8, 9]. Whether acute or chronic, alco-hol metabolism increases production of acetaldehyde andenhances formation of DNA adducts leading to tissue injury.On the other hand, metabolism of alcohol via cytochromeP4502E1 induces production of reactive oxygen specieswhich facilitates adduct formation, activates stress proteins,induces endoplasmic reticulum stress, and affects lysosomalfunction and autophagy leading to mitochondrial injury andhepatocellular death.

3.1. Alcohol Metabolism and ROS. Ethanol is primarily meta-bolized in the liver by oxidative enzymatic pathways. Theclassical pathway of alcohol metabolism involves enzymaticbreakdown of alcohol by the enzyme, alcohol dehydrogenase(ADH) and its subsequent conversion to acetaldehyde andformation of acetate. ADH is predominantly expressed inliver [10] but other tissues like gastric mucosa express ADHand contribute to metabolism of alcohol [10]. Aldehydedehydrogenase (ALDH) contributes to oxidation of aldehydeintermediates resulting in acetate which is unstable andbreaks down to water and carbon dioxide. The secondmajor pathway for ethanol degradation is the microsomalsystem catalyzed by cytochrome P450 enzymes. The 2E1isoform of the cytochrome P450 (CYP2E1) system is inducedduring chronic alcohol consumption. Activation of CYP2E1leads to ROS generation and highly reactive free radicalsincluding superoxide anions and hydroxyl radicals resultingin oxidative stress and cell death [11]. The role of CYP2E1in hepatocyte injury has been elucidated using HEPG2cells overexpressing CYP2E1 [12], CYP2E1 knockout mice,and transgenic mice [13]. Increased oxidative stress frominduction of CYP2E1 in vivo sensitizes hepatocytes toLPS and TNFα toxicity [14] and CYP2E1 knock-in miceshowed elevated hepatic steatosis and liver injury afteralcohol feeding [13]. On the other hand, CYP2E1 knockoutmice showed decreased oxidant stress, upregulation of

PPARα and were protective to alcohol-induced liver injury.Peroxynitrite, activation of p38 and JNK MAP kinases,and mitochondrial dysfunction are downstream mediatorsof the CYP2E1-LPS/TNF potentiated hepatotoxicity [15].Oxidation of ethanol by alcohol dehydrogenase and sub-sequent metabolism of acetaldehyde results in increasedNADH/NAD+ ratio in the cytoplasm and mitochondria[16]. The increase in NADH results in inhibition of mito-chondrial β-oxidation and accumulation of intracellularlipids [17]. Alcohol/CYP2E1-mediated ROS has the potentialto peroxidize lipids and inhibit mitochondrial and peroxiso-mal β-oxidation enzymes such as acyl-CoA dehydrogenases,carnitine palmitoyl transferase-1 (CPT-1), and peroxisomalproliferator-inducing pathways, respectively [18]. This dis-ruption leads to increased fatty acids, substrates of theseenzymes, and their accumulation resulting in developmentof hepatic steatosis. Oxidative stress and ROS generation dueto alcohol metabolism not only increase accumulation oflipids in hepatocytes but also sensitize the liver to subsequentinsults by cytokines.

3.2. Mitochondria and Oxidative Stress. In mitochondria,ROSs are generated as undesirable side products of theoxidative energy metabolism. An excessive and/or sustainedincrease in ROS production has been implicated in thepathogenesis of ALD, ischemia/reperfusion injury, and otherdiseases [19]. Oxidative stress induced by alcohol is closelyassociated with alterations in mitochondrial function result-ing in cellular death. Hepatic mitochondria either acutelyor chronically exposed to ethanol generate increased levelsof reactive oxygen species (ROS) [20]. The induction ofmitochondrial dysfunction is also linked to the metabolismof alcohol by CYP2E1 and increased oxidative stress [11].Primary hepatocytes and rat hepatoma cells when treatedwith ethanol led to an increase in ROS/RNS and loss ofmitochondrial function due to damaged mitochondrial DNAand ribosomes and subsequent inhibition of mitochondrialprotein synthesis [21, 22]. Studies have shown that alcohol-induced ROS generation leads to alteration in mitochondrialmembrane permeability and transition potential that inturn initiates the release of proapoptotic factors such ascytochrome c [21]. Transition of mitochondrial permeabilityresults in increased caspase-3 activation in hepatocytes andthis depends on p38 MAPK activation but is independentof caspase-8 [5]. Various studies show that decreased ATPsynthesis accompanied by reduced mitochondrial proteinsynthesis, inhibition of the oxidative phosphorylation system(OxPhos), and damage to mitochondrial DNA leads todysfunctional mitochondria and oxidative stress in alcoholicliver disease [23]. Peroxisome proliferator activated receptorgamma (PPARγ)-coactivator 1 alpha (PGC-1α), a transcrip-tion coactivator involved in mitochondrial biogenesis, isinvolved in defenses against ROS by inducing many ROS-mediated detoxifying enzymes. PGC-1 gene expression waslower in hepatic tissues of rats exposed to ethanol [24].In vitro exposure of hepatoma cells to 500 mM ethanolsignificantly decreased hepatic SIRT-1; PGC-1α leads toROS-induced mitochondrial and cellular injury [25]. Certainsirtuins, a family of protein deacetylases, were found to


regulate glucose and fat metabolism in mammals [26, 27]and to enhance mitochondrial biogenesis in liver and musclethrough PGC-1α and to influence cell survival [28].

Recent studies used an antioxidant peptide targetedto mitochondria to show that altered ROS metabolismfacilitates enhanced expression of HIF-1alpha [29], which, inturn, increases TNF-alpha secretion. These findings providein vivo evidence for the action of mitochondrial ROS onHIF-1alpha activity and demonstrate that changes in mito-chondrial function within physiologically tolerable limits canmodulate the immune response [29]. These studies suggestthat alcohol-induced mitochondrial stress pathways set thestage for proinflammatory cytokine-induced cell death andliver injury.

3.3. Protein Adducts and Lysosomal Dysfunction. Alcoholmetabolism and oxidative stress result in the formation ofreactive aldehydes such as acetaldehyde, malondialdehyde(MDA), and 4-hydroxy-2-nonenal (HNE) that can bind toproteins to form adducts [30]. In vivo models of chronicalcohol consumption have shown that acetaldehyde, MDA,and HNE adduct formation are increased in various organsincluding the liver [30]. A strong corelation between 4-HNE adducts and expression of CYP2E1 in patients withALD was recently shown [31]. Acetaldehyde and MDA reactwith proteins synergistically to form hybrid protein adductscalled malondialdehyde-acetaldehyde (MAA) adducts [32].Recognition of MAA-adducts by Kupffer cells, endothelial,and stellate cells via the scavenger receptor resulted in upreg-ulation of cytokine and chemokine production and increasedexpression of adhesion molecules [32]. Circulating antibod-ies to MAA-adducts were detected in patients with alcoholichepatitis and cirrhosis and correlated with the severityof liver injury [33]. Chronic alcohol feeding also inducesformation of gamma-ketoaldehyde protein adducts in mouselivers [34]. These adducts are formed in a TNFR1/CYP2E1dependent, but cyclooxygenase-independent manner inmouse liver [34]. Existence of protein adducts during chronicalcohol consumption and their identification in animalmodels has been challenging, limiting investigation of theirprecise role in ALD.

Increased ROS and lipid peroxidation rate in microsomaland lysosomal membranes with a simultaneous decreasein the levels of glutathione sulfhydryls and glutathione-S-transferase activity was observed during alcohol exposure[35]. Elevation of cathepsin B in hepatic cytosol fractions,indicating lysosomal leakage, was reported in ethanol-fedrats [36]. Lysosomal leakage was increased in alcohol-fedmice deficient in superoxide dismutase (SOD) indicating thatoxidative stress correlated with loss of lysosomal functionincreased hepatic fat and inflammatory cell infiltration[37]. The exact mechanisms responsible for ethanol-inducedchanges in lysosomal function are not clear but there isevidence of enhanced lysosomal membrane fragility, whichcould result from either altered lipid peroxidation, oxidativestress, or both [38]. More recently, degradation of a cell’sown cytosolic components in the lysosomes as a protectivemechanism against the damaging effects of oxidative stresshas been described and is termed autophagy [38]. Alcoholic

liver injury is associated with decreased autophagy result-ing in accumulation of damaged proteins and liver celldeath [38]. Recent studies show that macro-, micro- andchaperone-mediated autophagy is linked to innate andadaptive immune responses [39]. While autophagy acts asan effector and regulator of pattern recognition receptorsincluding TLR4 signaling in macrophages, loss or defectiveautophagy results in accumulation of cytosolic componentsand chronic inflammatory responses [40]. How loss ofautophagy after chronic alcohol consumption contributes toproinflammatory responses in alcoholic liver disease remainsto be investigated.

3.4. Endoplasmic Reticulum (ER) Stress. The unfolded pro-tein response (UPR) is a protective response of the cell alsoreferred to as the ER stress response during pathologicalconditions. In alcoholic liver disease, increased expressionof glucose regulatory protein (GRP)78, GRP94, CHOP,and caspase-12 indicated a UPR/ER stress response [41].Upregulation and activation of ER-localized transcriptionfactors such as SREBP-1c and SREBP-2 were associatedwith increased lipid accumulation and induction of fattyliver during chronic alcohol exposure [42]. Another impor-tant inducer of ER stress, homocysteine, was increased inalcoholic human subjects leading to hyperhomocysteinemia,also observed in alcohol feeding rodent models [43]. Therole of ER stress in triglyceride accumulation and fattyliver comes from studies showing that betaine increases anenzyme, betaine homocysteine methyltransferase (BHMT)and reduces homocysteine levels to inhibit lipid accumula-tion [43]. Recent studies suggest that ER/UPR stress path-ways intersect with innate immune signaling determining theduration and intensity of inflammatory response [44]. Addi-tional mechanistic studies to link ER/UPR stress and innateimmune responses as a pathophysiological contributor inALD are warranted.

3.5. Alcohol, Stress, and Molecular Chaperones. Stress or heatshock proteins (hsps) are ubiquitous and highly conser-ved proteins, functioning as molecular chaperones, whoseexpression is induced by oxidative stress stimuli and inresponse to accumulation of unfolded cellular proteins.Oxidative stress induces heat shock proteins via activationof the heat shock transcription factor (HSF) [45]. MaleWistar rats fed with acute as well as chronic alcohol showedinduction of hsp70 in the various regions of the brain andthe liver [46, 47]. However, the intensity of induction ofhsp70 in the liver, the principal organ of ethanol oxidation,was much lower than the hippocampus or striatal areasof the brain [47]. Hsp90 levels, on the other hand, wereincreased in cultured rat hepatocytes exposed to acutealcohol [47, 48]. Acute and chronic alcohol treatment ofmonocytes/macrophages showed alterations in hsp70 andhsp90 mRNA and protein levels based on the length ofalcohol exposure [49]. Acute alcohol induces HSF andhsp70, whereas chronic alcohol induces hsp90 but not hsp70protein, through activation of HSF [49]. Hsp90 functions asa molecular chaperone controlling activity of various kinasesand signaling molecules of the LPS signaling pathway such


as CD14 [50], IKK [51], and IRAK [52]. Comprehensivestudies on the effect of acute and chronic alcohol exposureon chaperone function of hsps in inflammatory responses inthe alcoholic liver could provide novel mechanistic insightsin ALD.

4. Inflammatory Response and ALD

Extensive studies over the past two decades have identifiedthe importance of macrophage activation in the liver bygut-derived endotoxin after prolonged alcohol consump-tion [2]. Central to this activation is the sensitization ofmacrophages due to alcohol exposure and is associated withmechanisms ranging from upregulation and engagementof surface receptors on innate immune cells, intracellularkinases and transcription factors contributing to inductionof proinflammatory cytokines.

4.1. Pattern Recognition Receptors, Alcohol, and Immune Cells.Pattern recognition receptors (PRRs) are expressed on livernonparenchymal and parenchymal cells and function as sen-sors of microbial danger signals enabling the vertebrate hostto initiate an immune response. The complexity of cellularexpression of PRRs in the liver provides unique aspects topathogen recognition and tissue damage in the liver [53].Toll-like receptors (TLRs) that are membrane associatedor endosomal recognize distinct microbial components andactivate different signaling pathways by selective utilizationof adaptor molecules [54]. TLRs such as TLR4 and TLR2that detect PAMPs like LPS and lipoproteins, respectively,are located on the cell surface whereas; TLRs such as TLR3,TLR7, and TLR9 that detect viral RNA and DNA are locatedin the endosome [55]. The pivotal role of TLR4 as well asother TLRs has been extensively studied in alcoholic tissueinjury [56–59].

The interaction of oxidative stress and TLR signaling isemerging. TLR4 is capable of inducing ROS leading to oxida-tive stress [59–61]. Kupffer cells or hepatic macrophages pro-duce reactive oxygen species (ROS) in response to antigenicstimuli and chronic alcohol exposure as well as endotoxin[62, 63]. Alcohol-induced sensitization of macrophages toLPS has been attributed to ROS production [59, 60, 64].Previous studies from Nagy and colleagues [64, 65] also showthat chronic ethanol feeding increases the sensitivity of Kupf-fer cells to lipopolysaccharide (LPS), leading to increasedtumor necrosis factor alpha (TNFα) expression. NADPHoxidase and ROS generation exhibit direct interaction withthe TLR4 receptor and activation of down-stream kinasesand transcription factors [61]. Studies by Gustot et al. [59]show that oxidative stress regulates TLR 2, 4, 6, and 9mRNA expression in alcoholic liver. Thus, it appears thatTLR mRNA, protein expression, and immune signaling canbe strongly influenced by oxidative stress in ALD makingthese two events dependent on each other and not mutuallyexclusive. Besides ROS, TLRs also mediate responses to hostmolecules including intracellular mediators [66]. Amongstthe well-characterized DAMPs, high-mobility group box1 (HMGB1), S100 proteins, hyaluronan, and heat shockprotein 60 (hsp60) are known to be recognized by TLR2 and

TLR4 [66, 67]. In addition, necrotic or apoptotic cells are alsorecognized as DAMPs by TLRs [67]. In alcoholic liver injury,apoptotic bodies, generated due to alcohol-induced oxidativestress, could be recognized by DAMPs [68] and contribute toinflammatory responses in the liver.

Activation of TLR4 recruits IRAK-1 to the TLR4 complexvia interaction with MyD88 and IRAK-4 [69]. The role ofMyD88, the common TLR4 adaptor molecule, was evaluatedin a mouse model of alcoholic liver injury [60]. Thesestudies showed that MyD88 knockout mice were highlysusceptible to alcohol-induced fatty liver [60]. While alcoholfeeding in TLR4 deficient mice prevented activation ofNADPH oxidase, alcohol-fed MyD88 deficient mice showedhigh NADPH oxidase activity and increased oxidative stressresulting in liver injury [60].

Increasing evidence suggests that downstream signalingcomponents activated by TLRs as well as cytokines andchemokines produced can be regulated by oxidative stresspathways. These interactions of stress pathways leading toinflammation could contribute largely to initiation andperpetuation of alcohol-related injury in the liver. The cross-talk of stress regulated intracellular molecules with TLRs,intracellular kinases and transcription factors resulting inalterations in cytokines/chemokines in ALD are of greatimportance.

4.2. MAPKs and IKKs. LPS/TLR4-induced ROS activation[61] plays an important role in activation of downstreamsignaling molecules such as IRAK1/4, TRAF6 leading toactivation of MAP kinases and NFκB during chronicalcohol exposure [69]. Mitogen-activated protein kinase[MAPK] signaling cascade plays an essential role in severalcellular processes including proliferation, differentiation,and apoptosis. Acute alcohol exposure results in activa-tion of baseline p42/44 MAPK in hepatocytes [70] whilechronic alcohol exposure causes potentiation of endotoxin-stimulated p42/44 MAPK, and p38 MAPK signaling inKupffer cells leading to increased synthesis of TNFα [71,72]. LPS stimulation of Kupffer cells in vitro exposed tochronic alcohol in vivo exhibited increased p38 activity anddecreased JNK activity [71, 73]. Inhibition of p38 activationimpaired alcohol-mediated stabilization of TNFα mRNAlikely via interaction with tristetraprolin (TTP) [74]. On theother hand, ERK1/2 inhibition did not alter TNFα mRNAstability but affected mRNA transcription in chronic alcohol-exposed macrophages via Egr-1 binding to the promoter[75]. Whether alcohol-induced ROS plays a role in MAPKactivation in ALD is not yet determined.

TLR4-induced MyD88-dependent and independentpathways lead to IKK kinase activation resulting in proin-flammatory cytokine production [71]. Oxidative stress andROS-mediated molecular chaperones such as hsp90 areshown to facilitate IKK kinase activity and downstreamNFκB nuclear activation [51]. Studies show that chronicalcohol-induced NFκB activation in macrophages is due toincreased hsp90 resulting in elevated IKK kinase activity[49]. Inhibition of hsp90 in chronic alcohol-exposed macro-phages resulted in decreased IKK kinase activity and NFκB


binding suggesting a cross-talk between cellular stress andinflammatory pathways [49].

4.3. Transcription Factors in ALD. The transcription fac-tor NFκB is a ubiquitous transcription factor that canbe activated by a large number of extracellular stimulisuch as cytokines, chemokines, growth factors, and bac-terial or viral products [76]. NFκB activation triggers theinduction of inflammatory genes and plays an importantrole in initiation and progression of alcoholic liver dis-ease [69, 77]. While TLR-mediated activation of NFκB iswell established, ROS-induced activation of NFκB occursbut remains poorly understood. Chronic alcohol exposureinduces LPS/TLR4-mediated NFκB activation in humanmonocytes and macrophages contributing to productionof proinflammatory cytokine, TNFα [77]. Whether ROSmediates activation of NFκB directly during ALD is unclear.TLR4-induced MyD88-independent signaling leads to acti-vation of IKKε and interferon regulatory factor 3 (IRF3) anddownstream Type I IFN activation [78, 79]. Previous studiesshow that ROS mediates LPS-induced IRF3 activation [80].Investigators found that IRF3 binds to the TNFα promoterin macrophages after chronic alcohol administration [81]and induces TNFα production. Whether alcohol-inducedROS mediates IRF3 induction to increase proinflammatorycytokines and liver injury needs further investigation.

Alcohol-mediated fatty liver injury is associated withincreased expression of genes regulating fatty acid synthesisand suppression of genes involved in fatty acid oxidation[82]. Transcription factors like SREBP and PPARα play apivotal role in fatty acid metabolism and rodent modelsas well as in vitro treatment studies with alcohol showdownregulation of PPARα mRNA [83]. Further, DNA-binding activity of PPARα is significantly reduced resultingin decreased expression of target genes involved in fatty acidmetabolism after alcohol exposure [83]. Decreased PPARαactivity was accompanied by increased oxidative stress in theliver resulting in increased sensitization of TNFα-inducedliver injury [83].

Another transcription factor, STAT3, in alcoholic liverinjury was recently investigated in hepatocyte-specific STAT3knockout (H-STAT3KO) mice and macrophage/neutrophil-specific STAT3 KO (M/N-STAT3KO) and endothelial STAT3mice [84]. Compared with wild-type mice, Kupffer cells fromalcohol-fed hepatocyte-specific STAT3KO mice producedsimilar amounts of ROS and hepatic proinflammatorycytokines compared to control mice [85]. On the other hand,Kupffer cells from M/N-STAT3KO mice produced higherROS and TNFα compared with wild-type controls. Theseresults suggest that STAT3 in hepatocytes promotes ROSproduction and inflammation whereas myeloid cell STAT3reduces ROS and hepatic inflammation during alcoholic liverinjury [85]. Thus, STAT3 may regulate hepatic inflammatorycytokines via ROS production.

5. Stress and Immune Signaling:How Are They Linked in ALD?

Cellular stress responses during alcohol exposure includeoxidative stress due to metabolism of alcohol in the liver,

ER stress, mitochondrial imbalance, heat shock proteininduction, and inflammatory processes. Numerous mousemodels have been used to study the discrete role of eachof the stress responses in alcohol-mediated liver pathology.Yet, accumulating evidence suggests that these pathwayscannot be regarded separately and are tightly interrelated.Similar to other inflammatory diseases [86], alcoholic liverdisease is multifactorial and it is important to take intoaccount interactions between various cellular responsesfor a better understanding of the pathogenesis of ALD.Based on studies so far, a clear relationship betweenoxidative stress and inflammation is emerging in ALD. Itis increasingly apparent that in addition to gut-derivedendotoxin, alcohol-induced upregulation of oxidative stressmediators plays a major part in activation of receptors,intracellular kinases, and transcription factors in innateimmune cells (Figure 1). Pathways described above thatare interrelated in ALD include ROS-mediated activationof TLR4 in macrophages, mitochondrial ROS regulationof transcriptional activators such as PGC-1α and HIF-1αpromoting TNFα induction, ROS and autophagy associatedenhancement of proinflammatory cytokine production [86],ER stress-associated innate immune cell activation, hsp-mediated activation of proinflammatory signaling kinases,and finally direct activation of transcription factors such asNFκB and STAT3 by ROS. Alcoholic liver disease exhibitsenhanced inflammatory responses and exaggerated TNFαproduction leading to liver injury. While TNFRI knockoutmice are protected from alcohol-induced liver injury [87],alcohol-induced ROS production was unaffected in TNFRIknockout mice indicating that ROS predominantly servesas a redox signal for proinflammatory cytokine productionand may not be a direct toxicant to hepatocytes [87]. Thesestudies argue against the direct role of ROS or oxidative stressin alcoholic liver injury and in fact support the notion thatoxidative stress/ROS primarily affects and is indispensable toproinflammatory activation and cytokine induction in ALDcreating a vicious cycle of the two pathways (Figure 1). Thus,attempts to further clarify the importance of oxidative stressand its cross-talk with inflammatory pathways will providean insight into pathogenesis of ALD and open avenues fornovel therapeutic targets.

6. Conclusion

This paper clearly implicates the role of oxidative stressin proinflammatory signaling and macrophage activationduring liver injury providing a feed-forward mechanismin ALD. Therefore, targeting redox-sensitive inflammatorypathways and transcription factors offers great promise fortreatment of ALD. Investigation of agents that interfere withoxidative stress mediators directly hampering inflammatorycytokine production is needed. Whether these agents willthen alleviate alcoholic liver disease in patients should betested.

Acknowledgments

This work was supported by the PHS Grant # AA017986 (toPM) and AA017545 (to PM) from the National Institute of


Hepatocytes

Transcription factors

Receptors and kinases

Kupffer cells

Oxidative stress InflammatoryresponseER stress

MitochondrialROS

Autophagy

Hsps

Kupffer cells

DCs Neutrophils

Cytokines andchemokines

ALD

AP-1

STAT3

ADH/CYP2E1

TLRs

MAPKs

PI3K/AKT

IKK

NF B

HIF-1

PPAR

Figure 1: Oxidative stress and inflammation: interacting mechanisms in ALD. The development of alcoholic liver injury is a complexprocess involving oxidative stress microenvironment in the liver contributed by hepatocytes and macrophages. In addition to the activationof macrophages by gut-derived endotoxin, cellular stress responses contribute to proinflammatory cytokine production creating a tightlyinterrelated network in ALD.

Alcohol Abuse and Alcoholism and its contents are the soleresponsibility of the authors and do not necessarily representthe views of the NIAAA.

References

[1] M. Amini and B. A. Runyon, “Alcoholic hepatitis 2010: aclinician’s guide to diagnosis and therapy,” World Journal ofGastroenterology, vol. 16, no. 39, pp. 4905–4912, 2010.

[2] R. G. Thurman, B. U. Bradford, Y. Iimuro et al., “Mechanismsof alcohol-induced hepatotoxicity: studies in rats,” Frontiers inBioscience, vol. 4, pp. e42–e46, 1999.

[3] Y. Lu, J. Zhuge, X. Wang, J. Bai, and A. I. Cederbaum,“Cytochrome P450 2E1 contributes to ethanol-induced fattyliver in mice,” Hepatology, vol. 47, no. 5, pp. 1483–1494, 2008.

[4] M. D. Wheeler, “Endotoxin and kupffer cell activation inalcoholic liver disease,” Alcohol Research and Health, vol. 27,no. 4, pp. 300–306, 2003.

[5] J. G. Pastorino, N. Shulga, and J. B. Hoek, “TNF-α-inducedcell death in ethanol-exposed cells depends on p38 MAPKsignaling but is independent of Bid and caspase-8,” AmericanJournal of Physiology, vol. 285, no. 3, pp. G503–G516, 2003.

[6] J. G. Pastorino and J. B. Hoek, “Ethanol potentiates tumornecrosis factor-α cytotoxicity in hepatoma cells and primaryrat hepatocytes by promoting induction of the mitochondrialpermeability transition,” Hepatology, vol. 31, no. 5, pp. 1141–1152, 2000.

[7] T. Zima and M. Kalousova, “Oxidative stress and signaltransduction pathways in alcoholic liver disease,” Alcoholism:Clinical and Experimental Research, vol. 29, no. 11, pp. 110S–115S, 2005.

[8] S. Balbo, M. Hashibe, S. Gundy et al., “N2-ethyldeoxyguano-sine as a potential biomarker for assessing effects of alcohol

consumption on DNA,” Cancer Epidemiology Biomarkers andPrevention, vol. 17, no. 11, pp. 3026–3032, 2008.

[9] P. J. Brooks and J. A. Theruvathu, “DNA adducts fromacetaldehyde: implications for alcohol-related carcinogenesis,”Alcohol, vol. 35, no. 3, pp. 187–193, 2005.

[10] C. S. Lieber, “Alcohol: its metabolism and interaction withnutrients,” Annual Review of Nutrition, vol. 20, pp. 395–430,2000.

[11] Y. Lu and A. I. Cederbaum, “CYP2E1 and oxidative liver injuryby alcohol,” Free Radical Biology and Medicine, vol. 44, no. 5,pp. 723–738, 2008.

[12] A. I. Cederbaum, D. Wu, M. Mari, and J. Bai, “CYP2E1-dependent toxicity and oxidative stress in HepG2 cells,” FreeRadical Biology and Medicine, vol. 31, no. 12, pp. 1539–1543,2001.

[13] Y. Lu, D. Wu, X. Wang, S. C. Ward, and A. I. Cederbaum,“Chronic alcohol-induced liver injury and oxidant stressare decreased in cytochrome P4502E1 knockout mice andrestored in humanized cytochrome P4502E1 knock-in mice,”Free Radical Biology and Medicine, vol. 49, no. 9, pp. 1406–1416, 2010.

[14] A. I. Cederbaum, L. Yang, X. Wang, and D. Wu, “CYP2E1sensitizes the liver to LPS- and TNF α-induced toxicity viaelevated oxidative and nitrosative stress and activation of ASK-1 and JNK mitogen-activated kinases,” International Journal ofHepatology, vol. 2012, Article ID 582790, 19 pages, 2012.

[15] A. Cahill, C. C. Cunningham, M. Adachi et al., “Effects ofalcohol and oxidative stress on liver pathology: the role ofthe mitochondrion,” Alcoholism: Clinical and ExperimentalResearch, vol. 26, no. 6, pp. 907–915, 2002.

[16] S. K. Mantena, A. L. King, K. K. Andringa, A. Landar,V. Darley-Usmar, and S. M. Bailey, “Novel interactions ofmitochondria and reactive oxygen/nitrogen species in alcohol


mediated liver disease,” World Journal of Gastroenterology, vol.13, no. 37, pp. 4967–4973, 2007.

[17] D. Wu and A. I. Cederbaum, “Alcohol, oxidative stress, andfree radical damage,” Alcohol Research and Health, vol. 27, no.4, pp. 277–284, 2003.

[18] J. K. Reddy and M. S. Rao, “Lipid metabolism and liverinflammation. II. Fatty liver disease and fatty acid oxidation,”American Journal of Physiology, vol. 290, no. 5, pp. G852–G858, 2006.

[19] W. Droge, “Free radicals in the physiological control of cellfunction,” Physiological Reviews, vol. 82, no. 1, pp. 47–95,2002.

[20] C. C. Cunningham and S. M. Bailey, “Ethanol consumptionand liver mitochondria function,” Biological Signals andReceptors, vol. 10, no. 3-4, pp. 271–282, 2001.

[21] B. R. Zelickson, G. A. Benavides, M. S. Johnson et al., “Nitricoxide and hypoxia exacerbate alcohol-induced mitochondrialdysfunction in hepatocytes,” Biochimica et Biophysica Acta, vol.1807, no. 12, pp. 1573–1582, 2011.

[22] J. B. Hoek, A. Cahill, and J. G. Pastorino, “Alcohol andmitochondria: a dysfunctional relationship,” Gastroenterology,vol. 122, no. 7, pp. 2049–2063, 2002.

[23] K. K. Andringa, A. L. King, H. B. Eccleston et al., “Analysis ofthe liver mitochondrial proteome in response to ethanol andS-adenosylmethionine treatments: novel molecular targets ofdisease and hepatoprotection,” American Journal of Physiology,vol. 298, no. 5, pp. G732–G745, 2010.

[24] W. W. Chaung, A. Jacob, Y. Ji, and P. Wang, “Suppressionof PGC-1alpha by ethanol: implications of its role in alcoholinduced liver injury,” International Journal of Clinical andExperimental Medicine, vol. 1, no. 2, pp. 161–170, 2008.

[25] C. S. Lieber, M. A. Leo, X. Wang, and L. M. DeCarli,“Effect of chronic alcohol consumption on Hepatic SIRT1and PGC-1α in rats,” Biochemical and Biophysical ResearchCommunications, vol. 370, no. 1, pp. 44–48, 2008.

[26] F. Picard and L. Guarente, “Molecular links between aging andadipose tissue,” International Journal of Obesity, vol. 29, no. 1,pp. S36–S39, 2005.

[27] J. T. Rodgers, C. Lerin, W. Haas, S. P. Gygi, B. M. Spiegelman,and P. Puigserver, “Nutrient control of glucose homeostasisthrough a complex of PGC-1α and SIRT1,” Nature, vol. 434,no. 7029, pp. 113–118, 2005.

[28] W. Droge and H. M. Schipper, “Oxidative stress and aberrantsignaling in aging and cognitive decline,” Aging Cell, vol. 6, no.3, pp. 361–370, 2007.

[29] D. Wang, D. Malo, and S. Hekimi, “Elevated mito-chondrial reactive oxygen species generation affects theimmune response via hypoxia-inducible factor-1α in long-lived Mclk1+/− mouse mutants,” Journal of Immunology, vol.184, no. 2, pp. 582–590, 2010.

[30] O. Niemela, S. Parkkila, S. Yla-Herttuala et al., “Covalentprotein adducts in the liver as a result of ethanol metabolismand lipid peroxidation,” Laboratory Investigation, vol. 70, no.4, pp. 537–546, 1994.

[31] Y. Wang, G. Millonig, J. Nair et al., “Ethanol-inducedcytochrome P4502E1 causes carcinogenic etheno-DNA lesionsin alcoholic liver disease,” Hepatology, vol. 50, no. 2, pp. 453–461, 2009.

[32] G. M. Thiele, M. J. Duryee, M. S. Willis et al.,“Malondialdehyde-acetaldehyde (MAA) modified proteinsinduce pro-inflammatory and pro-fibrotic responses by liverendothelial cells,” Comparative Hepatology, vol. 3, supplement1, p. S25, 2004.

[33] D. J. Tuma, “Role of malondialdehyde-acetaldehyde adducts inliver injury,” Free Radical Biology and Medicine, vol. 32, no. 4,pp. 303–308, 2002.

[34] S. Roychowdhury, M. R. McMullen, M. T. Pritchard, W. Li, R.G. Salomon, and L. E. Nagy, “Formation of γ-ketoaldehyde-protein adducts during ethanol-induced liver injury in mice,”Free Radical Biology and Medicine, vol. 47, no. 11, pp. 1526–1538, 2009.

[35] J. Jajte, J. Stetkiewicz, and T. Wronska-Nofer, “Combinedexposure to m-xylene and ethanol: oxidative stress in therat liver,” International Journal of Occupational Medicine andEnvironmental Health, vol. 16, no. 4, pp. 345–350, 2003.

[36] T. M. Donohue, T. V. Curry-McCoy, A. A. Nanji et al.,“Lysosomal leakage and lack of adaptation of hepatoprotectiveenzyme contribute to enhanced susceptibility to ethanol-induced liver injury in female rats,” Alcoholism: Clinical andExperimental Research, vol. 31, no. 11, pp. 1944–1952, 2007.

[37] T. V. Curry-Mccoy, N. A. Osna, A. A. Nanji, and T. M.Donohue, “Chronic ethanol consumption results in atypicalliver injury in copper/zinc superoxide dismutase deficientmice,” Alcoholism: Clinical and Experimental Research, vol. 34,no. 2, pp. 251–261, 2010.

[38] T. M. Donohue, “Autophagy and ethanol-induced liver in-jury,” World Journal of Gastroenterology, vol. 15, no. 10, pp.1178–1185, 2009.

[39] T. Saitoh and S. Akira, “Regulation of innate immuneresponses by autophagy-related proteins,” Journal of CellBiology, vol. 189, no. 6, pp. 925–935, 2010.

[40] Y. Xu, C. Jagannath, X. D. Liu, A. Sharafkhaneh, K. E.Kolodziejska, and N. T. Eissa, “Toll-like receptor 4 is a sensorfor autophagy associated with innate immunity,” Immunity,vol. 27, no. 1, pp. 135–144, 2007.

[41] N. Kaplowitz and C. Ji, “Unfolding new mechanisms ofalcoholic liver disease in the endoplasmic reticulum,” Journalof Gastroenterology and Hepatology, vol. 21, supplement 3, pp.S7–S9, 2006.

[42] C. Ji, C. Chan, and N. Kaplowitz, “Predominant role ofsterol response element binding proteins (SREBP) lipogenicpathways in hepatic steatosis in the murine intragastricethanol feeding model,” Journal of Hepatology, vol. 45, no. 5,pp. 717–724, 2006.

[43] C. Ji and N. Kaplowitz, “Betaine decreases hyperhomocys-teinemia, endoplasmic reticulum stress, and liver injury inalcohol-fed mice,” Gastroenterology, vol. 124, no. 5, pp. 1488–1499, 2003.

[44] F. Martinon and L. H. Glimcher, “Regulation of innate immu-nity by signaling pathways emerging from the endoplasmicreticulum,” Current Opinion in Immunology, vol. 23, no. 1, pp.35–40, 2011.

[45] T. Finkel and N. J. Holbrook, “Oxidants, oxidative stress andthe biology of ageing,” Nature, vol. 408, no. 6809, pp. 239–247,2000.

[46] V. Calabrese, M. Renis, A. Calderone, A. Russo, M. L. Barcel-lona, and V. Rizza, “Stress proteins and SH-groups in oxidant-induced cell damage after acute ethanol administration in rat,”Free Radical Biology and Medicine, vol. 20, no. 3, pp. 391–397,1996.

[47] V. Calabrese, M. Renis, A. Calderone et al., “Stress proteins andSH-groups in oxidant-induced cellular injury after chronicethanol administration in rat,” Free Radical Biology andMedicine, vol. 24, no. 7-8, pp. 1159–1167, 1998.

[48] S. Ikeyama, K. Kusumoto, H. Miyake, K. Rokutan, andS. Tashiro, “A non-toxic heat shock protein 70 inducer,


geranylgeranylacetone, suppresses apoptosis of cultured rathepatocytes caused by hydrogen peroxide and ethanol,” Jour-nal of Hepatology, vol. 35, no. 1, pp. 53–61, 2001.

[49] P. Mandrekar, D. Catalano, V. Jeliazkova, and K. Kodys,“Alcohol exposure regulates heat shock transcription factorbinding and heat shock proteins 70 and 90 in monocytes andmacrophages: implication for TNF-α regulation,” Journal ofLeukocyte Biology, vol. 84, no. 5, pp. 1335–1345, 2008.

[50] V. L. Vega and A. De Maio, “Geldanamycin treatmentameliorates the response to LPS in murine macrophages bydecreasing CD14 surface expression,” Molecular Biology of theCell, vol. 14, no. 2, pp. 764–773, 2003.

[51] M. Broemer, D. Krappmann, and C. Scheidereit, “Require-ment of Hsp90 activity for IκB kinase (IKK) biosynthesis andfor constitutive and inducible IKK and NF-κB activation,”Oncogene, vol. 23, no. 31, pp. 5378–5386, 2004.

[52] D. De Nardo, P. Masendycz, S. Ho et al., “A central role for theHsp90·Cdc37 molecular chaperone module in interleukin-1 receptor-associated-kinase-dependent signaling by Toll-likereceptors,” Journal of Biological Chemistry, vol. 280, no. 11, pp.9813–9822, 2005.

[53] G. Szabo, A. Dolganiuc, and P. Mandrekar, “Pattern recog-nition receptors: a contemporary view on liver diseases,”Hepatology, vol. 44, no. 2, pp. 287–298, 2006.

[54] K. Takeda and S. Akira, “Toll-like receptors in innate immu-nity,” International Immunology, vol. 17, no. 1, pp. 1–14, 2005.

[55] S. Uematsu and S. Akira, “Toll-like receptors and type IInterferons,” Journal of Biological Chemistry, vol. 282, no. 21,pp. 15319–15324, 2007.

[56] T. Uesugi, M. Froh, G. E. Arteel, B. U. Bradford, and R. G.Thurman, “Toll-like receptor 4 is involved in the mechanismof early alcohol-induced liver injury in mice,” Hepatology, vol.34, no. 1, pp. 101–108, 2001.

[57] S. Yamashina, Y. Takei, K. Ikejima, N. Enomoto, T. Kitamura,and N. Sato, “Ethanol-induced sensitization to endotoxin inKupffer cells is dependent upon oxidative stress,” Alcoholism:Clinical and Experimental Research, vol. 29, no. 12, pp. 246S–250S, 2005.

[58] P. Mandal, P. H. Park, M. R. McMullen, B. T. Pratt, and L.E. Nagy, “The anti-inflammatory effects of adiponectin aremediated via a heme oxygenase-1-dependent pathway in ratkupffer cells,” Hepatology, vol. 51, no. 4, pp. 1420–1429, 2010.

[59] T. Gustot, A. Lemmers, C. Moreno et al., “Differential liversensitization to toll-like receptor pathways in mice withalcoholic fatty liver,” Hepatology, vol. 43, no. 5, pp. 989–1000,2006.

[60] I. Hritz, P. Mandrekar, A. Velayudham et al., “The critical roleof toll-like receptor (TLR) 4 in alcoholic liver disease is inde-pendent of the common TLR adapter MyD88,” Hepatology,vol. 48, no. 4, pp. 1224–1231, 2008.

[61] H. S. Park, H. Y. Jung, E. Y. Park, J. Kim, W. J. Lee, and Y. S.Bae, “Cutting edge: direct interaction of TLR4 with NAD(P)Hoxidase 4 isozyme is essential for lipopolysaccharide-inducedproduction of reactive oxygen species and activation of NF-κB,” Journal of Immunology, vol. 173, no. 6, pp. 3589–3593,2004.

[62] M. Yin, B. U. Bradford, M. D. Wheeler et al., “Reduced earlyalcohol-induced liver injury in CD14-deficient mice,” Journalof Immunology, vol. 166, no. 7, pp. 4737–4742, 2001.

[63] H. Kono, I. Rusyn, M. Yin et al., “NADPH oxidase-derivedfree radicals are key oxidants in alcohol-induced liver disease,”Journal of Clinical Investigation, vol. 106, no. 7, pp. 867–872,2000.

[64] V. Thakur, M. T. Pritchard, M. R. McMullen, Q. Wang, andL. E. Nagy, “Chronic ethanol feeding increases activation ofNADPH oxidase by lipopolysaccharide in rat Kupffer cells:role of increased reactive oxygen in LPS-stimulated ERK1/2activation and TNF-α production,” Journal of LeukocyteBiology, vol. 79, no. 6, pp. 1348–1356, 2006.

[65] L. E. Nagy, “Recent insights into the role of the innateimmune system in the development of alcoholic liver disease,”Experimental Biology and Medicine, vol. 228, no. 8, pp. 882–890, 2003.

[66] M. T. Lotze, H. J. Zeh, A. Rubartelli et al., “The grateful dead:damage-associated molecular pattern molecules and reduc-tion/oxidation regulate immunity,” Immunological Reviews,vol. 220, no. 1, pp. 60–81, 2007.

[67] J. A. Sloane, D. Blitz, Z. Margolin, and T. Vartanian, “Aclear and present danger: endogenous ligands of Toll-likereceptors,” NeuroMolecular Medicine, vol. 12, no. 2, pp. 149–163, 2010.

[68] C. A. Casey, S. M. L. Lee, R. Aziz-Seible, and B. L. McVicker,“Impaired receptor-mediated endocytosis: its role in alcohol-induced apoptosis,” Journal of Gastroenterology and Hepatol-ogy, vol. 23, supplement 1, pp. S46–S49, 2008.

[69] P. Mandrekar and G. Szabo, “Signalling pathways in alcohol-induced liver inflammation,” Journal of Hepatology, vol. 50, no.6, pp. 1258–1266, 2009.

[70] A. R. Aroor, Y. J. Lee, and S. D. Shukla, “Activation of MEK1/2 and p42/44 MAPK by angiotensin II in hepatocyte nucleusand their potentiation by ethanol,” Alcohol, vol. 43, no. 4, pp.315–322, 2009.

[71] R. Kishore, M. R. McMullen, and L. E. Nagy, “Stabilizationof tumor necrosis factor α mRNA by chronic ethanol: role ofA+U-rich elements and p38 mitogen-activated protein kinasesignaling pathway,” Journal of Biological Chemistry, vol. 276,no. 45, pp. 41930–41937, 2001.

[72] L. Shi, R. Kishore, M. R. McMullen, and L. E. Nagy, “Chronicethanol increases lipopolysaccharide-stimulated Egr-1 expres-sion in RAW 264.7 macrophages. Contribution to enhancedtumor necrosis factor α production,” Journal of BiologicalChemistry, vol. 277, no. 17, pp. 14777–14785, 2002.

[73] A. Koteish, S. Yang, H. Lin, X. Huang, and A. M. Diehl,“Chronic ethanol exposure potentiates lipopolysaccharideliver injury despite inhibiting Jun N-terminal kinase andcaspase 3 activation,” Journal of Biological Chemistry, vol. 277,no. 15, pp. 13037–13044, 2002.

[74] K. R. Mahtani, M. Brook, J. L. E. Dean, G. Sully, J. Saklatvala,and A. R. Clark, “Mitogen-activated protein kinase p38controls the expression and posttranslational modificationof tristetraprolin, a regulator of tumor necrosis factor alphamRNA stability,” Molecular and Cellular Biology, vol. 21, no.19, pp. 6461–6469, 2001.

[75] R. A. J. Kishore, J. R. Hill, M. R. Mcmullen, J. Frenkel, andL. E. Nagy, “ERK1/2 and Egr-1 contribute to increased TNF-αproduction in rat Kupffer cells after chronic ethanol feeding,”American Journal of Physiology, vol. 282, no. 1, pp. G6–G15,2002.

[76] S. Ghosh, M. J. May, and E. B. Kopp, “NF-κB and rel proteins:evolutionarily conserved mediators of immune responses,”Annual Review of Immunology, vol. 16, pp. 225–260, 1998.

[77] P. Mandrekar, S. Bala, D. Catalano, K. Kodys, and G.Szabo, “The opposite effects of acute and chronic alcohol onlipopolysaccharide-induced inflammation are linked to IRAK-M in human monocytes,” Journal of Immunology, vol. 183, no.2, pp. 1320–1327, 2009.


[78] T. Kawai, O. Takeuchi, T. Fujita et al., “Lipopolysaccharidestimulates the MyaD88-independent pathway and results inactivation of IFN-regulatory factor 3 and the expression ofa subset of lipopolysaccharide-inducible genes,” Journal ofImmunology, vol. 167, no. 10, pp. 5887–5894, 2001.

[79] K. A. Fitzgerald, D. C. Rowe, B. J. Barnes et al., “LPS-TLR4signaling to IRF-3/7 and NF-κB involves the toll adaptersTRAM and TRIF,” Journal of Experimental Medicine, vol. 198,no. 7, pp. 1043–1055, 2003.

[80] E. Chiang, O. Dang, K. Anderson, A. Matsuzawa, H. Ichijo,and M. David, “Cutting edge: apoptosis-regulating signalkinase 1 is required for reactive oxygen species-mediatedactivation of IFN regulatory factor 3 by lipopolysaccharide,”Journal of Immunology, vol. 176, no. 10, pp. 5720–5724, 2006.

[81] X. J. Zhao, Q. Dong, J. Bindas et al., “TRIF and IRF-3binding to the TNF promoter results in macrophage TNFdysregulation and steatosis induced by chronic ethanol,”Journal of Immunology, vol. 181, no. 5, pp. 3049–3056, 2008.

[82] D. W. Crabb and S. Liangpunsakul, “Alcohol and lipidmetabolism,” Journal of Gastroenterology and Hepatology, vol.21, supplement 3, pp. S56–S60, 2006.

[83] M. Fischer, M. You, M. Matsumoto, and D. W. Crabb, “Per-oxisome proliferator-activated receptor α (PPARα) agonisttreatment reverses PPARα dysfunction and abnormalities inhepatic lipid metabolism in ethanol-fed mice,” Journal ofBiological Chemistry, vol. 278, no. 30, pp. 27997–28004, 2003.

[84] A. M. Miller, H. Wang, O. Park et al., “Anti-inflammatory andanti-apoptotic roles of endothelial cell STAT3 in alcoholic liverinjury,” Alcoholism: Clinical and Experimental Research, vol.34, no. 4, pp. 719–725, 2010.

[85] N. Horiguchi, L. Wang, P. Mukhopadhyay et al., “Celltype-dependent pro- and anti-inflammatory role of signaltransducer and activator of transcription 3 in alcoholic liverinjury,” Gastroenterology, vol. 134, no. 4, pp. 1148–1158, 2008.

[86] E. Naik and V. M. Dixit, “Mitochondrial reactive oxygenspecies drive proinflammatory cytokine production,” Journalof Experimental Medicine, vol. 208, no. 3, pp. 417–420, 2011.

[87] M. Yin, E. Gabele, M. D. Wheeler et al., “Alcohol-inducedfree radicals in mice: direct toxicants or signaling molecules?”Hepatology, vol. 34, no. 5, pp. 935–942, 2001.


Research Article

Markers of Inflammation and Fibrosis in AlcoholicHepatitis and Viral Hepatitis C

Manuela G. Neuman,1, 2 Hemda Schmilovitz-Weiss,3 Nir Hilzenrat,4 Marc Bourliere,5

Patrick Marcellin,6 Cristhian Trepo,7 Tony Mazulli,8 George Moussa,8 Ankit Patel,1

Asad A. Baig,1 and Lawrence Cohen9

1 In Vitro Drug Safety and Biotechnology, MaRS Discovery Centre, 101 College Street, South Tower, Toronto, ON, Canada M5G 1L72 Department of Pharmacology and Toxicology and Institute of Drug Research, Faculty of Medicine, University of Toronto,Toronto, ON, Canada

3 Gastroenterology Unit, Gastroenterology Division, Rabin Medical Center, Hasharon-Golda Campus, 7 Keren Kayemet St.,Petach-Tiqwa 49372, Israel

4 Division of Gastroenterology, Department of Medicine, Hepatitis SMBD—Jewish General Hospital Montreal, McGillUniversity School of Medicine, 3755 Chemin de la Cote-St-Catherine, Montreal, QC, Canada H3T 1E2

5 Department of Hepato-Gastroenterology, Saint-Joseph Hospital, Marseille, France6 Service Hepato-Gastroenterology, Hopital Beaujon, Clichy, France7 Hopital de la Croix Rousse, 103, Grande Rue de la Croix-Rousse, 69317, Lyon Cedex 04, France8 Department of Microbiology, Mount Sinai Hospital and Department of Laboratory Medicine, University of Toronto,550 University Av., Toronto, ON, Canada M5G 1L7

9 Division of Gastroenterology, Sunnybrook Health Sciences Centre, Department of Medicine, University of Toronto,2075 Bayview Ave, Toronto, ON, Canada M4N 3M5

Correspondence should be addressed to Manuela G. Neuman, m [email protected]

Received 31 July 2011; Revised 31 October 2011; Accepted 23 November 2011

Academic Editor: Kusum Kharbanda

Copyright © 2012 Manuela G. Neuman et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

High levels of profibrinogenic cytokine transforming factor beta (TGF-β), metalloprotease (MMP2), and tissue inhibitor of matrixmetalloprotease 1 (TIMP1) contribute to fibrogenesis in hepatitis C virus (HCV) infection and in alcohol-induced liver disease(ALD). The aim of our study was to correlate noninvasive serum markers in ALD and HCV patients with various degrees ofinflammation and fibrosis in their biopsies. Methods. Serum cytokines levels in HCV-infected individuals in the presence or absenceof ALD were measured. Student’s-t-test with Bonferroni correction determined the significance between the groups. Results. Bothtumor-necrosis-factor- (TNF)-α and TGF-β levels increased significantly with the severity of inflammation and fibrosis. TGF-βlevels increased significantly in ALD patients versus the HCV patients. Proinflammatory cytokines’ responses to viral and/or toxicinjury differed with the severity of liver inflammation. A combination of these markers was useful in predicting and diagnosingthe stages of inflammation and fibrosis in HCV and ALD. Conclusion. Therapeutic monitoring of TGF-β and metalloproteasesprovides important insights into fibrosis.

1. Introduction

Alcohol-induced liver disease (ALD) encompasses a spec-trum of hepatic injury, ranging from simple steatosis tocirrhosis. Alcohol ethylic (ethanol), a hepatotoxin, producesthe oldest form of liver injury known to humankind [1–4].In addition, ethanol-inducible cytochrome P-450 (CYP2E1)

increases vulnerability of the heavy drinker to commonlyprescribed drugs [5–7]. Moreover, dysregulated cytokines,including TNF-α and downstream cytokines, play a pivotalrole in the pathophysiology of ALD [8, 9]. In addition,Th1 cells produce interleukin (IL)-2, interferon (IFN)-γ,and TNF-α, that promote inflammation and cell-mediatedimmunity in an attempt to control infection [10–12]. Th2


cells produce IL-10 [10] and IL-4, inducing fibrogenesis[13–17]. In HCV, an impaired HCV-specific CD4+ T-cellresponse can lead to persistence of the virus characterizedby inflammation [18–22]. The expression of suppressor ofcytokine signaling (SOCS) proteins permits the host to betterrespond to therapies [23].

The initial liver histological changes are characterized byaccumulation of inflammatory cells and matrix depositionaround the portal vein [24–26]. In liver disease, includingALD and HCV, liver fibrosis is defined as the abnormal accu-mulation of extracellular matrix (ECM) [26–33]. In all cases,inflammation plays a crucial role [24, 34]. TGF-β mediatesthe effects through signal mothers against decapentaplegic(Smad) proteins [35, 36].

In response to liver injury neutrophils migrate tothe site of infection through chemokines. Distinct pat-terns of expression of each chemokine were noted onKupffer cells (IL-8) (CXC) [4, 36], sinusoidal endothelialcells macrophage inflammatory protein 1 (MIP-1) (CC)[37], hepatocytes (CXC chemokines, IL-8) [38], lympho-cytes (MIP-1), (CX3C) [39], hepatic stellate cells (HSCs){monocyte chemoattractant protein 1 (MCP-1) [40–42],regulated upon activation of normal T-cell expressed andpresumably secreted (RANTES) (CC)} [42]. Other biomark-ers that are attributed to HSC activity include levels of TIMPs[43, 44]. MMP-1 levels, on the other hand, significantlydecreased during fibrogenesis [45]. Alcohol consumption inHCV-infected people is known to cause accelerated progres-sion of liver fibrosis, a higher frequency of cirrhosis, and anincreased incidence of HCC [46–48]. Genes associated withfibrosis/cell adhesion/ECM were not specific to alcohol andhave been reported in HCV-induced liver cirrhosis [49].

Previously, we reported that sera levels of RANTES, TNF-α, IL-6, IL-8, and IL-12 as well as TGF-β in HCV-infectedindividuals are higher as compared to healthy controls [34].A strong correlation was observed between the degree ofinflammation, as shown by the histological activity index,and TNF-α levels, thus indicating the possibility of usingTNF-α levels as a marker of the degree of liver inflammation.Furthermore, TGF-β levels were significantly higher amongthose with moderate and mild fibrosis (F2-F3) regardlessof the inflammation, suggesting the role TGF-β in HCVpatients takes place mostly in earlier stages of the diseasebefore cirrhosis is well established [34].

The aim of our present study was to correlate noninvasiveserum markers: cytokines, chemokines in ALD, and chronicHCV patients with various degrees of inflammation andfibrosis in their biopsies.

2. Methods

2.1. Patients. We studied 260 AH : 140 (mild histologicalactivity index (HAI), 60, high HAI, and 60 cirrhotics; AH,comorbidity, 60 HCV (30 cirrhotics). From 1180 HCV (thatdeclared not drinking) : 170—low fibrosis—HAI; 450—mildfibrosis—low HAI; 440—moderate fibrosis—HAI; 120—high fibrosis—high HAI. ALD was considered as resultingfrom long-term heavy drinking (over 80 mg/alcohol/day).

The patient population was 98% Caucasian, treated inCanada, France, and Israel.

2.1.1. Chemical Measurements. The laboratory services ateach of the participating sites performed routine blood testsincluding alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, albumin, and platelet counts.

2.1.2. Liver Biopsy. After submitting an informed consentdocument, all patients underwent a percutaneous liverbiopsy to ascertain the diagnosis and their stage of liverinjury. Biopsy specimens were fixed, paraffin-embedded, andstained with haematoxylin and eosin, Masson’s trichrome,and Sirius red. All specimens were examined and gradedby the pathologists of the specific medical center. A fibrosisscore from 0 to 4 and an inflammatory score from 0 to16 were adopted according to the severity and the extentof damage. Liver fibrosis was evaluated according to theMETAVIR scoring system. Fibrosis (F) was staged on a scaleof 0 to 4: F0, no fibrosis; F1, portal fibrosis without septa; F2,few septa; F3, numerous septa without cirrhosis; F4, cirrhosis[32].

2.1.3. Inclusion/Exclusion Criteria. Each treating physicianestablished a diagnosis based upon the following criteria:clinical presentation, a history of excessive alcohol con-sumption, and exclusion of other etiology, elevated livertransaminases, neutrophil counts, serum bilirubin, andimpaired coagulation. In addition, diagnosis was reachedusing appropriate virological and histological criteria.

HCV-infected patients tested positive for antibody toHCV on third-generation enzyme-linked immunosorbent-assay (ELISA) or recombinant immunoblot assay (AbbottLaboratories, Chicago, IL, USA). After that the viral load wasmeasured. The patients had persistently elevated serum ALTlevels for more than 6 months and no evidence of infectionwith hepatitis B virus (absence of detectable hepatitis Bsurface antigen). Additionally, there was no presence ofantihuman immunodeficiency virus antibodies. Also othercauses of chronic liver disease (hepatotoxic drugs, autoim-mune chronic hepatitis, hemochromatosis, Wilson’s disease,and -1 antitrypsin deficiency) and a history of decom-pensated cirrhosis (ascites, bleeding esophageal varices, orhepatic encephalopathy) were excluded. Moreover, none ofthese patients had received immunomodulatory or antiviraltherapy. Liver histology showed lesions characteristic ofchronic hepatitis.

All the patients described in the study had been treatedfor their disease in the specific medical facility in his (her)own country. The Ethics Committees of the specific MedicalCentre approved this study, which is in concordance withthe ethical guidelines of the 1975 Declaration of Helsinkifor research involving human participants. Informed consentwas obtained from all participating patients.

2.1.4. Characteristics Studied. The following characteristicswere compared between the groups: sex, age, duration ofHCV infection, or alcohol use (80 g or more/day for at least


Table 1: Baseline characteristics ALD patients.

Characteristics F2 (n = 140) F3 (n = 60) F4 (n = 60)

Age (years) 27 ± 8 46 ± 12 64 ± 5∗

Sex (F/M) 24/116 8/52 0/60

Alcohol consumption (years) 10 ± 4 20 ± 6 28 ± 16

ALT (U/L) 46.0± 3.0 65.0± 11.0∗∗ 57.0± 12.0∗∗

AST (U/L) 45.5± 4.5 80.5± 15.0∗∗ 77.0± 10.0∗∗

Bilirubin (mg/dL) 1.2± 0.5 1.96± 1.20 5.80± 1.50

Values represent mean ± S.D. ∗P < 0.05 higher than F2 and F3; ∗∗P < 0.001 higher, when compared to F2.

Table 2: Baseline characteristics: HCV-infected individuals.

Characteristics F0-F1 (n = 170) F2 (n = 450) F3 (n = 440) F4 (n = 120)

Age (years) 39 ± 16 47 ± 6 46 ± 7 54 ± 10∗

Sex (F/M) 48/122 177/273 132/308 50/70

HCV infection (years) 15 ± 6 18 ± 7 22 ± 8 27 ± 12

ALT (U/L) 70.0± 20.0 79.0± 10.0 99.0± 5.0∗∗ 86.0± 20.0

AST (U/L) 32.0± 4.0 42.0± 8.0 40.5± 15.0 50.8± 5.0

Bilirubin (mg/dL) 0.26± 0.5 3.6± 2.5 5.5± 3.50 8.5± 5.5

Values represent means ± S.D; ∗P < 0.05 higher when compared to F0-F1, F2, and F3; ∗∗P < 0.001 higher when compared to F0 or F2.

1 year), liver histology, HCV genotype, and viral load. Tables1, 2, and 3 record the characteristics of the patients studied.

2.1.5. Laboratory Methods. We measured serum: IL-6, IL-8, TNF-α, TGF-β, RANTES, Fas-ligand (FAS-L), hyaluronicacid (HA), TIMP, and apoptosome (M30). We also per-formed the correlation between these cytokines, chemokines,and apoptosis markers with the degree of inflammation andfibrosis as shown by light microscopy (LM) as well as withother biochemical parameters such as ALT, AST, bilirubin,albumin, ferritin as well as HCV genotype, and viral load.Patient serum specimens were kept at 4◦C immediately aftercollection, centrifuged, aliquoted for each measurement, andfrozen at −80◦C within 2 h of being drawn. This providesthe optimal conditions for reliable results [11]. CytoscreenImmunoassay kits (BioSource International, Camarillo, CA,USA) for human IL-6, IL-8, TNF-α, TGF-β, RANTES,TIMP quantified the cytokines as described previously[23].

2.2. Fas/sFasL Measurements. Cytoscreen, ImmunoassayKits, Human Fas (BioSource International, Camarillo, CA,USA), and soluble FasL (Bender MedSystems, Vienna,Austria) were used for the quantitative determination ofFas/sFasL in serum as previously described [50]. The cor-relation coefficient was linear (Fas r = 0.996; FasL r =0.998) in a concentration range between 0.23 and 15 ng/mLfor Fas and between 0.16 and 10 ng/mL for FasL. Thesamples having higher concentrations were diluted. Eachspecimen was analyzed in triplicate with a sensitivity andspecificity of 96% and 92%, respectively. We used standardsand reference reagents available from the National Institutefor Biological Standards and Controls (NIBSC, Herts., UK).These methods are standardized in our laboratory accordingto the procedures described [11, 50, 51].

2.3. Apoptosis Measurements. Apoptosis was measured usingthe M30-Apoptosense ELISA (Bender MedSystems, Vienna,Austria) using the manufacturer instruction. It is a solid-phase, two-site immunosorbent assay. The absorbance wasmeasured in a microplate reader at 450 nm. The corre-lation coefficient was linear (r = 0.995) in a concen-tration range between 50 and 1000 U/mL. The sampleshaving higher concentrations were diluted. Each speci-men was analyzed in triplicate with a sensitivity of 95%and specificity of 90%. We used standards and refer-ence reagents available from Bender MedSystems (ViennaAustria).

2.4. Detection of Serum Hepatitis C Virus RNA. For quan-titation of HCV RNA in human we used “AMPLICORHCV MONITOR test (Roche Diagnostic, PQ, Canada andNeuilly, France)”. The test is specially designed for assessingviral load with the low linear sensitivity 6 × 102. Theprocedure is based on five major steps required by thespecimen preparation, reverse transcription of target RNA togenerate complementary DNA (cDNA), PCR amplificationof target cDNA using HCV-specific complementary primers,hybridization of the amplified DNA to oligonucleotideprobes specific to the target, and detection of the probebind amplifier. In 100 patients we have evaluated theperformance of newly developed automated real-time PCRassay, the COBAS Ampliprep/COBAS TaqMan (CAP/CTM)with AMPLICOR HCV MONITOR test (Roche Diagnostic,PQ, Canada) COBAS as previously described [52]. Theoverall concordance for negative/positive results was 100%for HCV. All assays were equally able to quantify HCV ingenotype 1. The results indicate that the real-time PCRassay covers better viral dynamic. Serum HCV RNA detec-tion and HCV genotyping were performed for diagnosticpurposes.


Table 3: Baseline characteristics ALD infected with HCV.

Characteristics F2 (n = 10) F3 (n = 40) F4 (n = 30)

Age (years) 50 ± 15 66 ± 27 64 ± 32

Sex (F/M) 0/10 2/38 5/25

HCV infection (years) 22 ± 7 25 ± 15 35 ± 25

ALT (U/L) 34.5± 10.5 60.5± 14.0∗ 126.0± 42.0∗∗

AST (U/L) 70.0± 25.0 154.0± 45.0 250.0± 75.0

Bilirubin (mg/dL) 0.36± 0.5 5.5± 2.50 12.80± 8.00

Values represent means ± S.D; ∗P < 0.05 higher when compared to F2; ∗∗P < 0.001 higher when compared to F2 or F3.

2.5. Genotyping of Hepatitis C Virus. HCV genotyping wasperformed at initiation of treatment in the 5′ untranslatedregion of the HCV genome, using reverse hybridization withthe line probe assay (InGeN, Rungis, France). The HCVline-probe assay contains 15 probe lines, allowing identifi-cation of HCV types 1 to 5 as well as their subtypes a and b[23].

2.6. Statistical Analysis. We calculated the statistical signifi-cance of parameters by using SPSS 9.0 for windows (SPSSInc., Chicago, Illinois, USA). Normality of the data was testedby means of Shapiro and Wilk’s W-test. Most of the datawas in a normal distribution. All statistical significance wasassessed at the 0.05 levels. Baseline data were descriptivelysummarized. To test differences between groups, we com-pared mean and standard deviation (S.D.) of each parameterusing either parametric or nonparametric tests. Differencesamong groups were determined by the use of confidenceintervals and analysis of variance. The 2 test or Fisher’sexact test was used to compare frequency data betweengroups. The initial histological lesions were evaluated bythe non-parametric rank correlation for each parameter.Mann-Whitney and Wilcoxon rank-sum tests were usedto compare values of continuous variables. Correlationsbetween variables were analyzed calculating the Spearmanrank correlation coefficient. To determine the independentprognostic value of the selected characteristics, a logisticregression model was used.

3. Results

3.1. Patient Characteristics. Baseline demographics and dis-ease characteristics were comparable across groups (HCV.HCV-ALD, and ALD), as shown in Tables 1, 2, and 3,confirmed by the Kruskal-Wallis test. All subjects hadclinically compensated but active liver disease based onevaluations of liver biopsies and ALT and AST levels. A num-ber of laboratory abnormalities, including elevated serumaminotransferases, were seen in patients with alcoholic liverinjury, in the presence or absence of HCV. In Tables 1 and 3we reported a specific feature used to diagnose ALD patients:the AST/ALT ratio which was higher than 1.2. Ratios ofAST/ALT greater than 1 are considered suggestive of alcohol-induced etiology.

Serum AST/ALT was typically elevated 1.2–2 times inALD individuals as well as in HCV-ALD patients. In alcoholic

hepatitis, the levels of AST and ALT were 2–6 times higherthan the upper limits of normal levels in healthy individuals.

Table 2 presents the characteristic analysis of HCV-patients. Consistent with previous studies and with patientpopulations in Europe and North America, the majority ofHCV-subjects had genotype 1 and 98% of all the patientswere Caucasians.

The viral load was not significantly different between theHCV (7.3± 6.4× 106) and ALD-HCV (6.8± 5.8× 106).

In patients with mild fibrosis (F3), there was a dramaticincrease in the HAI score as compared to those withmoderate fibrosis (F2; P < 0.001). Additionally, as thefibrosis progressed to high levels seen in cirrhotic patients(F4), the HAI significantly increased as compared to patientswith mild fibrosis (F3; P < 0.05).

In Figures 1(a) and 1(b), in the upper panel, the ALTvalues were plotted versus the HAI. As can be seen, thelevel of ALT was higher when compared to the normallevel. However, the ALT was not significantly different inALD patients or HCV patients in spite of the different HAIlevels. Notably, there was a significant increase in TNF-αlevels as the HAI increased (Figure 1(a) (HCV patients) andFigure 1(b) (ALD patients)). Moreover, the TNF-α levelsin ALD patients were significantly higher (P < 0.05) whencompared with TNF-α levels at the same HAI in HCVinfected individuals.

Regardless the presence or absence of HCV-infectionin the ALD patients, there were no significant differencesbetween the levels of TNF-α. A positive correlation wasfound between the degree of inflammation and TNF-α levels(r = 0.90, P < 0.001) in all patients. Also, there was asignificant increase in the levels of IL-8 and RANTES atmoderate HAI as compared to patients presenting lowerHAI (r = 0.89; P < 0.001), followed by a continuedincrease in IL-8 levels and more profoundly RANTESlevels (r = 0.92, P < 0.001). This pattern was observedin all ALD, HCV, and ALD-HCV patients. There wereno significant differences between the levels of IL-8 andRANTES between the 3 types of patients. Finally, the levelsof IL 6 between these individuals did not differ and werenot significantly distinct when compared with the level ofHAI.

As seen in the upper panel of Figure 1, ALT values do notdiffer with increased severity of inflammation. In contrast,as seen in the lower panel of the graph, TNF-α values differsignificantly with the increased severity of inflammation (P <0.05) up until the moderate inflammatory response. There


Minimal Mild Moderate High

0

4

850

100

150

200

250

300

350

400


0

4

8

50100150200250300350400

(a)


0

4

850

100

150

200

250

300

350

400

HA

I ve

rsu

s A

LT (

IU)


0

4

8

100

200

300

400

500

600

HA

I ve

rsu

s T

NF

(pg/

mL)

(b)

Figure 1: (a) Correlation between the ALT ± SD measured in U/L (upper panel) (triangles) and HAI (columns: minimal, mild, moderate,and high inflammation by histology) in HCV. The lower panel presents the correlation between TNF-α ± SD ((pg/mL) rhomboid) andhistological activity index (HAI-columns) in the same set of patients. (b) Correlation between the ALT (U/L ± SD (triangles)) and HAI(solid black columns) is presented in the upper panel graph. TNF-α ((pg/mL ± SD) rhomboid) is correlated in the lower panel with HAI(solid black columns HAI-0, mild, moderate, high) in patients with ALD.

is no significant difference in the levels of TNF -α betweenmoderate and high inflammatory index.

ALT (IU/L) values do not differ with the increased sever-ity of inflammation neither in HCV Figure 1(a), nor in ALDindividuals (Figure 1(b)). On the contrary, TNF-α valuesincrease significantly with the increased severity of inflam-mation (P < 0.05) as described by a histologic inflammatoryresponse. Both in HCV and in ALD the TNF-α graduallyincreases, paralleling the increase in HAI until the moderateHAI. There is a significant difference between the levels ofTNF-α in high HAI versus the moderate HAI in patientswith ALD (P < 0.05). This is an exclusive feature for theALD individuals since in HCV-infected individuals there isa plateau of TNF-α (Figure 1(a)), between the patients withmoderate and high HAI.

Additionally, apoptosis (M-30) increased dramaticallyin the moderate and high HAI. When correlating IL-8and TNF-α levels with HCV genotype, no significance wasobserved. There was no correlation between the levels ofALT and AST with the severity of inflammation. There wasa significant increase in the level of FAS-L when comparingpatients with mild HAI to those with minimal HAI (r = 0.80,P < 0.05). The level of FAS-L was maintained in patients withmoderate HAI, whereas in patients with high HAI a majorincrease was seen (P < 0.001). Also in ALD-HCV significantincrease in FAS levels was observed in patients with moderate

HAI as compared to those with minimal and mild HAI(r = 0.78, P < 0.05). This level was maintained in patientswith high HAI. The same pattern was seen with M-30.A good correlation was observed between the apoptosomelevels and TNF-α in patients presenting the same degree ofinflammation (r = 0.82).

In patients with higher HAI, the levels of apoptosisand proinflammatory cytokines were significantly higher(P < 0.001) in patients with ALD when compared withpatients with HCV.

The levels of TGF-β increased with the severity of liverfibrosis in HCV (Figure 2). The patients with ALD and HCVhad comparable levels of TGF-β in the stages of fibrosis F2-F3. The ALD patients with F4 had levels of TGF-β two timesas high as the patients with HCV. There is no statisticaldifference between the TGF-β levels of in HCV and ALDpatients at F0–F3. At the highest fibrosis level (F4) there isa statistical difference (P < 0.05) in TGFβ between HCV andALD patients.

Hyaluronic acid is a glycosaminoglycan synthesized byHSCs. HA is eliminated from the circulation by sinusoidalendothelial cells. In our patients HA levels followed the samepattern as TGF-β in HCV, increasing significantly from 54±16 (ng/mL) (F0-1) to 122 ± 32 (F2) (P < 0.05), 207 ±45 (F3) (P < 0.05); 567 ± 85 (F4) (P < 0.05). In ALDpatients, the levels of HA (ng/mL) also increased significantly


F2 F3 F4

01234

20

40

60

80

100

Fibr

osis

inde

x/T

GF-

(ng/

mL)

F0-1

Figure 2: Correlation between the levels of TGF-β (ng/mL) in patients with HCV (triangle) and ALD (circle) and fibrosis score (solid blackcolumns: fibrosis 0-1 (F0-1); fibrosis 2 (F2), fibrosis 3 (F3), cirrhosis (F4)).

versus the previous fibrosis stage 82±16 (F2), 128±25 (F3)(P < 0.05); 497 ± 53 (F4) (P < 0.001). No statisticallysignificant difference in HA was observed between the HCV-ALD patients and the patients with non-HCV-infected ALD.Their values were 65±35 (F2); 120±44 (F3); 360±125 (F4).

In HCV-infected individuals, it is important to observethe increase of TIMP1 values (ng/mL) (F0-1) 65 ±12 to 126 ± 12 (F2); 147 ± 45 (F3); 672 ± 58 (F4) (P <0.001 versus the previous fibrosis stages). In ALD subjects,the TIMP1 values increased from 105 ± 65 (F2), 205 ±45 (F3), 872 ± 83 (F4) (P < 0.001 versus the previousfibrosis stages). The levels of TIMP1 (ng/mL) in ALD-HCVindividuals were significantly different when compared withHCV individuals at the same HAI (P < 0.001) but notsignificantly different from the ALD patients.

Patients with moderate fibrosis presented a dramaticincrease in the HAI as compared to those with mild fibrosis(P < 0.001). In addition, as the fibrosis progressed tohigher levels in the cirrhotic patients, the HAI significantlyincreased as compared to patients with moderate fibrosis(P < 0.05). The levels of apoptosome (M30) were withinthe normal ranges in all patients that presented minimalinflammation. This level was maintained in patients withmoderate HAI. There is a sequential increase in apoptosomein ALD patients. However, major increase was seen in ALDpatients (P < 0.001) versus the moderate HAI. Additionally,in HCV individuals, a significant increase of M30 wasobserved (P < 0.05). The same pattern was seen in thelevels of M30 in ALD-HCV (P < 0.05) increased versus themoderate HAI.

4. Discussion

The pattern of alcohol consumption is expressed and regu-lated differently in diverse geographical regions. Consideringthe fact that our population was recruited from Europe andNorth America we defined high risk drinking on the average80 g alcohol intake/day.

Over the past fifteen years our laboratory has been focus-ing on the alterations of the cytokine network in patientswith HCV. We have reported a correlation between TNFα-as a mediator of inflammation and the degree of HAI.In addition, in our previous studies with noncirrhotics,we have established that TGFβ- levels reflect the histologyfibrosis score [9]. Thus, measuring TNFα-and TGFβ-levelsand correlating them with HAI and fibrosis scores shouldbring vital information in assessing and monitoring thenatural history of the disease in patients with HCV. In thepresent study we aimed to evaluate the cytokine-chemokinenetwork in ALD in the presence or absence of HCVinfection. Moreover, we aimed to correlate serum cytokineand chemokine levels, with the severity of the disease andto compare the same parameters in ALD patients in thepresence of absence of HCV infection.

Liver fibrosis, regardless of its cause, is characterizedby excess deposition of collagens and ECM proteins in theparenchyma [26]. The accumulation of interstitial collagensand ECM proteins results from an imbalance between theirproduction and degradation by the MMPs. Fibrosis is nowincreasingly used as an end point for clinical trials in liverdisease, as stated by the international fibrosis group [53, 54].

The tissue inhibitors of metalloproteinases designatedTIMP-1, TIMP-2, and TIMP-3 regulate the extracellularactivity of their MMPs, which become activated whenreleased from the cells. TIMPs overexpression correlateswith decreased activity of ECM removing MMPs. Currentevidence indicates that TIMP-1 is the most importantendogenous inhibitor of interstitial collagenase, whereasTIMP-2 is particularly important in the inhibition of MMP-2[55, 56].

TIMP has also emerged as a molecule with dual action.It inhibits the activity of MMPs and prevents stellate cellsapoptosis [57].

Another component of the ECM is HA, synthesized byHSCs and eliminated from the circulation by sinusoidalendothelial cells equipped with a specific receptor [58–60].


Levels of this ECM protein and other parameters of liverdamage have been proposed to assess the progression offibrosis in ALD [61, 62], as well as in patients with HCV [63].Therefore the importance of the present work lies in definingthese biomarkers and their role in ALD.

An additional aim of the present study was to evaluateseveral markers of fibrosis in a cohort of subjects with aprimary etiology HCV and ALD in the presence or absenceof HCV infection. The overall goal of this paper was toanalyze in HCV and ALD possible markers of inflammationand fibrosis with focus on the ones that correlate bestwith the morphologic grade of inflammation and fibrosis.Moreover, special focus has been placed on three markers(TGF-β, TIMP1, and HA). Other scientists [64–66] alsoaimed at establishing tools needed to facilitate diagnosisand appropriate treatment by using circulating markersto replace some of the repeated liver biopsies currentlyrequired in the management of many patients with liverdisease.

We report here for the first time detailed analyses ofmarkers of liver inflammation and fibrosis in a subgroup ofHCV and alcoholic patients whose liver biopsy was available.Although the liver biopsy has limitations, the gold standardof determining and scoring fibrosis is its histological assess-ment. [67]. Accordingly, a number of circulating markers ofhepatic inflammation and fibrogenesis have been proposedas an alternative to this procedure. Therefore, in additionto morphologic stages of fibrosis, serum markers of fibrosishave also been considered to predict the vulnerability forthe future development and progression of the fibrosis. Inchronic HCV patients the combination of several tests inaddition to a liver biopsy improved the diagnostic scores[68]. In addition, the concept of “rate of fibrosis progression”has been the search of noninvasive surrogate measures ofliver fibrosis [69]. In ALD, Lieber’s team has shown thatchronic alcohol consumption may affect the levels of someof these fibrosis markers [70–72]. In addition, the roleof nutrition in ALD had been emphasized [73]. In viewof the present work, we consider promising candidates asnoninvasive fibrosis markers: TIMP1, HA, TGF-β. However,the greatest clinical usefulness of this test resided in its abilityto exclude cirrhosis. Better predictors of inflammation andfibrosis can be obtained by combining panels of noninvasivebiomarkers in which the accuracy has been analyzed bya series of algorithms [74, 75]. A combination of newspecific biomarkers like the ones we described in thispaper with the well-established laboratory changes suchas AST/ALT ratio [76] might help in distinguishing theetiology of hepatitis. Studies demonstrated that the presenceof HCV and drinking habits are cofactors of risk for alcohol-induced liver damage [77]. Moreover, practice guidelines intreating alcoholic liver disease recognized the importanceof biomarkers in identifying the seriousness of the damage[78, 79]. In addition, the genetic polymorphism in ALDwas evaluated [80]. In a genome-wide association studyperformed in 61089 individuals, there were identified lociassociated with high plasma liver enzyme in individuals withliver disease. Genes of inflammation and immunity (CD276,CDH6, GCKR, HNF1A, HPR, ITGA1, RORA, and STAT4)

have been shown to be expressed in liver disease patients[81].

Our study is important, therefore, as it examines thecorrelation between liver inflammation and serum inflam-matory markers. Serum biomarkers play a substantial role inidentifying and scoring the severity of liver injury induced byalcohol and its comorbidity with HCV. These findings needto be further investigated in patient samples with biopsy-based assessment of inflammation, apoptosis, and fibrosis aswe evaluate in the present cohort.

Since 98% of the entire patient population, originat-ing from Canada, France, and Israel, was Caucasian, ourstudy lacked diversity. Though limiting upon first glance,this aspect can actually be useful in that this populationmight represent a more homogeneous group regarding theirgenetic, immune, and environmental setting. This pointwas clearly demonstrated also by the recent study thatinvestigated the role of Genotype PNPLA3 rs738409(GG)which is associated with alcoholic liver cirrhosis in alcoholicCaucasians of German ancestry [82]. In addition, excellentcorrelation was found between TGF-β and procollagen IIIN-peptide (PIIINP) and peptide in patients with alcoholic liverdisease [83]. New approaches using transient elastography(Fibroscan) may add additional information to evaluate thedegree of fibrosis in hepatitis C and alcohol patients [84, 85].

In summary, all of the inflammatory biomarkers wereelevated both in ALD and in HCV, but some of thebiomarkers that were effective for HCV did not performas well, in a subgroup with alcoholic fibrosis as primaryetiology. It should be noted that these results are based on arelatively small cohort of ALD when compared with the verylarge cohort of HCV.

Many individual clinical and laboratory features, alongwith specific histological characteristics, have also been testedas measures of disease prognosis.

Cytokines play a specific role in determining the ALDand HCV progression of liver disease. There is significantimbalance in cytokine milieu that leads to the impairment ofthe healing process in both ALD and HCV. This paper bringsevidence suggesting that dysregulated cytokines, includingTNF-α and downstream cytokines, play a pivotal role inthe pathophysiology of ALD, HCV, and their comorbidity.Environmental factors such as the continued use of heavyalcohol can accelerate the progression of ALD disease byactivating HSC and enhancing the profibrinogenic cytokineexpression in the liver. However, the progression of fibrosisis highly variable, with some individuals progressing rapidlyover time whereas other individuals have stable liver disease.

Abbreviations

ALD: Alcohol liver diseaseAH: Alcoholic hepatitisCCR/CXCR–CC/CXC: Chemokine receptorCD: Cluster of differentiationCXC: Chemokines presenting an amino

acid between the two N-terminalcysteine residues

ECM: Extracellular matrix


ELISA: Enzyme-linked immunosorbent assayHA: Hyaluronic acidHAI: Hepatic activity indexHCV: Hepatitis C virusHSC: Hepatic stellate cellsIFN: InterferonIL: InterleukinMCP-1: Monocyte chemo-attractant protein 1MHC: Major histocompatible complexMIP-1: Macrophage inflammatory protein 1MMP: Matrix metalloproteinasesRANTES: Regulated upon activation, normal T-cell

expressed, and presumably secretedSD: Standard deviationSmad: Signal mothers against decapentaplegicSOCS: Suppressor of cytokine signalingSTAT: Signal transducer and activator of

transcriptionTIMP: Tissue inhibitor of metalloproteinasesTh: T helperTreg: T regulatoryTGF-β: Transforming growth factor betaTIMP: Tissue inhibitors of metalloproteinasesTNF-α: Tumour necrosis factor-alpha

Acknowledgment

The work was funded by a grant from In Vitro Drug Safetyand Biotechnology, Toronto, ON, Canada.

References

[1] C. S. Lieber, “Metabolism of ethanol and associated hepato-toxicity,” Drug and Alcohol Review, vol. 10, no. 3, pp. 175–202,1991.

[2] N. Kaplowitz and H. Tsukamoto, “Oxidative stress and liverdisease,” Progress in Liver Diseases, vol. 14, pp. 131–159, 1996.

[3] C. S. Lieber, “New concepts of the pathogenesis of alcoholicliver disease lead to novel treatments,” Current Gastroenterol-ogy Reports, vol. 6, no. 1, pp. 60–65, 2004.

[4] M. G. Neuman, “Apoptosis in diseases of the liver,” CriticalReviews in Clinical Laboratory Sciences, vol. 38, no. 2, pp. 109–166, 2001.

[5] H. J. Zimmerman and W. C. Maddrey, “Acetaminophen(paracetamol) hepatotoxicity with regular intake of alcohol:analysis of instances of therapeutic misadventure,” Hepatology,vol. 22, no. 3, pp. 767–773, 1995.

[6] M. G. Neuman, R. G. Cameron, J. A. Haber, G. G. Katz, I.M. Malkiewicz, and N. H. Shear, “Inducers of cytochromeP450 2E1 enhance methotrexate-induced hepatocytotoxicity,”Clinical Biochemistry, vol. 32, no. 7, pp. 519–536, 1999.

[7] C. Rigamonti, M. Vidali, M. F. Donato et al., “Serumautoantibodies against cytochrome P450 2E1 (CYP2E1) pre-dict severity of necroinflammation of recurrent Hepatitis C,”American Journal of Transplantation, vol. 9, no. 3, pp. 601–609,2009.

[8] C. J. McClain and D. A. Cohen, “Increased tumor necrosisfactor production by monocytes in alcoholic hepatitis,” Hep-atology, vol. 9, no. 3, pp. 349–351, 1989.

[9] M. G. Neuman, “Cytokines in Alcoholic Liver,” AlcoholResearch and Health, vol. 27, pp. 313–322, 2003.

[10] A. Gramenzi, P. Andreone, E. Loggi et al., “Cytokine profileof peripheral blood mononuclear cells from patients withdifferent outcomes of hepatitis C virus infection,” Journal ofViral Hepatitis, vol. 12, no. 5, pp. 525–530, 2005.

[11] F. Farinati, R. Cardin, M. Bortolami, M. Guido, and M. Rugge,“Oxidative damage, pro-inflammatory cytokines, TGF-α andc-myc in chronic HCV-related hepatitis and cirrhosis,” WorldJournal of Gastroenterology, vol. 12, no. 13, pp. 2065–2069,2006.

[12] A. R. N. Zekri, M. S. El-din Ashour, A. Hassan, H. M. A. El-Din, A. M. R. El-Shehaby, and M. A. Abu-Shady, “Cytokineprofile in Egyptian hepatitis C virus genotype-4 in relation toliver disease progression,” World Journal of Gastroenterology,vol. 11, no. 42, pp. 6624–6630, 2005.

[13] D. T. Y. Lau, B. A. Luxon, S. Y. Xiao, M. R. Beard, and S.M. Lemon, “Intrahepatic gene expression profiles and alpha-smooth muscle actin patterns in hepatitis C virus inducedfibrosis,” Hepatology, vol. 42, no. 2, pp. 273–281, 2005.

[14] R. Thimme, V. Lohmann, and F. Weber, “A target on the move:innate and adaptive immune escape strategies of hepatitis Cvirus,” Antiviral Research, vol. 69, no. 3, pp. 129–141, 2006.

[15] K. Ando, L. G. Guidotti, S. Wirth et al., “Class I-restrictedcytotoxic T lymphocytes are directly cytopathic for their targetcells in vivo,” Journal of Immunology, vol. 152, no. 7, pp. 3245–3253, 1994.

[16] L. G. Guidotti and F. V. Chisari, “To kill or to cure: optionsin host defense against viral infection,” Current Opinion inImmunology, vol. 8, no. 4, pp. 478–483, 1996.

[17] L. G. Guidotti and F. V. Chisari, “Noncytolytic control ofviral infections by the innate and adaptive immune response,”Annual Review of Immunology, vol. 19, pp. 65–91, 2001.

[18] G. Missale, R. Bertoni, V. Lamonaca et al., “Different clinicalbehaviors of acute hepatitis C virus infection are associatedwith different vigor of the anti-viral cell-mediated immuneresponse,” Journal of Clinical Investigation, vol. 98, no. 3, pp.706–714, 1996.

[19] A. Cerny and F. V. Chisari, “Pathogenesis of chronic hepatitisC: immunological features of hepatic injury and viral persis-tence,” Hepatology, vol. 30, no. 3, pp. 595–601, 1999.

[20] A. Ulsenheimer, J. T. Gerlach, N. H. Gruener et al., “Detectionof functionally altered hepatitis C virus-specific CD4+ T cellsin acute and chronic hepatitis C,” Hepatology, vol. 37, no. 5,pp. 1189–1198, 2003.

[21] N. Semmo, C. L. Day, S. M. Ward et al., “Preferential loss ofIL-2-secreting CD4+ T helper cells in chronic HCV infection,”Hepatology, vol. 41, no. 5, pp. 1019–1028, 2005.

[22] C. Prezzi, M. A. Casciaro, V. Francavilla et al., “Virus-specificCD8+ T cells with type 1 or type 2 cytokine profile arerelated to different disease activity in chronic hepatitis C virusinfection,” European Journal of Immunology, vol. 31, no. 3, pp.894–906, 2001.

[23] M. G. Neuman, J. P. Benhamou, P. Marcellin et al., “Cytokine-chemokine and apoptotic signatures in patients with hepatitisC,” Translational Research, vol. 149, no. 3, pp. 126–136, 2007.

[24] R. Bataller and D. A. Brenner, “Liver fibrosis,” Journal ofClinical Investigation, vol. 115, no. 2, pp. 209–218, 2005.

[25] G. Ramadori and B. Saile, “Portal tract fibrogenesis in theliver,” Laboratory Investigation, vol. 84, no. 2, pp. 153–159,2004.

[26] R. G. Cameron and M. G. Neuman, “Novel morphologicfindings in alcoholic liver disease,” Clinical Biochemistry, vol.32, no. 7, pp. 579–584, 1999.


[27] J. H. Lefkowitch, “Morphology of alcoholic liver disease,”Clinics in Liver Disease, vol. 9, no. 1, pp. 37–53, 2005.

[28] M. R. Teli, C. P. Day, A. D. Burt, M. K. Bennett, and O. F. W.James, “Determinants of progression to cirrhosis or fibrosis inpure alcoholic fatty liver,” The Lancet, vol. 346, no. 8981, pp.987–990, 1995.

[29] T. M. Worner and C. S. Lieber, “Perivenular fibrosis as pre-cursor lesion of cirrhosis,” Journal of the American MedicalAssociation, vol. 254, no. 5, pp. 627–630, 1985.

[30] K. G. Ishak, H. J. Zimmerman, and M. B. Ray, “Alcoholicliver disease: pathologic, pathogenetic and clinical aspects,”Alcoholism, vol. 15, no. 1, pp. 45–66, 1991.

[31] P. Christoffersen and K. Nielsen, “Histological changes inhuman liver biopsies from chronic alcoholics,” Acta Patholog-ica et Microbiologica Scandinavica A, vol. 80, no. 4, pp. 557–565, 1972.

[32] P. Bedossa and T. Poynard, “An algorithm for the gradingof activity in chronic hepatitis C The METAVIR Cooperativestudy group,” Hepatology, vol. 24, no. 2, pp. 289–293, 1996.

[33] C. Guyot, S. Lepreux, C. Combe et al., “Hepatic fibrosis andcirrhosis: the (myo) fibroblastic cell subpopulations involved,”International Journal of Biochemistry and Cell Biology, vol. 38,no. 2, pp. 135–151, 2006.

[34] M. G. Neuman, “Markers of fibrosis in liver disease,” Roma-nian Journal of Hepatology, vol. 2, pp. 56–58, 2006.

[35] S. Benlloch, B. Beltran, R. Moreno, and M. Berenguer,“Fibrogenesis and liver transplantation,” Gastroenterologia yHepatologia, vol. 26, no. 6, pp. 381–395, 2003.

[36] J. Massague and D. Wotton, “Transcriptional control by theTGF-β/Smad signaling system,” EMBO Journal, vol. 19, no. 8,pp. 1745–1754, 2000.

[37] Y. S. Huang, C. Y. Chan, J. C. Wu, C. H. Pai, Y. Chao, and S. D.Lee, “Serum levels of interleukin-8 in alcoholic liver disease:relationship with disease stage, biochemical parameters andsurvival,” Journal of Hepatology, vol. 24, no. 4, pp. 377–384,1996.

[38] D. H. Adams, S. Hubscher, J. Fear, J. Johnston, S. Shaw, andS. Afford, “Hepatic expression of macrophage inflammatoryprotein-1α and macrophage inflammatory protein-1β afterliver transplantation,” Transplantation, vol. 61, no. 5, pp. 817–825, 1996.

[39] R. F. Schwabe, R. Bataller, and D. A. Brenner, “Humanhepatic stellate cells express CCR5 and RANTES to induceproliferation and migration,” American Journal of Physiology,vol. 285, no. 5, pp. G949–G958, 2003.

[40] M. Salmi, D. Adams, and S. Jalkanen, “Cell adhesion andmigration IV. Lymphocyte trafficking in the intestine andliver,” American Journal of Physiology, vol. 274, no. 1, pp. G1–G6, 1998.

[41] G. Ramadori and B. Saile, “Inflammation, damage repair,immune cells, and liver fibrosis: specific or nonspecific, this isthe question,” Gastroenterology, vol. 127, no. 3, pp. 997–1000,2004.

[42] P. Bioulac-Sage, J. Kuiper, T. J. C. Van Berkel, and C.Balabaud, “Lymphocyte and macrophage populations in theliver,” Hepato-Gastroenterology, vol. 43, no. 7, pp. 4–14, 1996.

[43] R. Lichtinghagen, D. Michels, C. I. Haberkorn et al., “Matrixmetalloproteinase (MMP)-2, MMP-7, and tissue inhibitor ofmetalloproteinase-1 are closely related to the fibroproliferativeprocess in the liver during chronic hepatitis C,” Journal ofHepatology, vol. 34, no. 2, pp. 239–247, 2001.

[44] V. Leroy, F. Monier, S. Bottari et al., “Circulating matrixmetalloproteinases 1, 2, 9 and their inhibitors TIMP-1 andTIMP-2 as serum markers of liver fibrosis in patients with

chronic hepatitis C: comparison with PIIINP and hyaluronicacid,” American Journal of Gastroenterology, vol. 99, no. 2, pp.271–279, 2004.

[45] O. Nunez, A. Fernandez-Martınez, P. L. Majano et al.,“Increased intrahepatic cyclooxygenase 2, matrix metallo-proteinase 2, and matrix metalloproteinase 9 expression isassociated with progressive liver disease in chronic hepatitis Cvirus infection: role of viral core and NS5A proteins,” Gut, vol.53, no. 11, pp. 1665–1672, 2004.

[46] T. R. Morgan, D. Brenner, J. Everhart et al., “Hepatitis C andalcohol: fundamental and translational research directions,”Alcoholism, vol. 27, no. 4, pp. 726–731, 2003.

[47] M. G. Neuman, D. A. Brenner, B. Rehermann et al., “Mecha-nisms of alcoholic liver disease: cytokines,” Alcoholism, vol. 25,no. 5, pp. 251S–253S, 2001.

[48] C. L. Nevins, H. Malaty, M. E. Velez, and B. S. Anand,“Interaction of alcohol and hepatitis C virus infection onseverity of liver disease,” Digestive Diseases and Sciences, vol.44, no. 6, pp. 1236–1242, 1999.

[49] D. Seth, M. D. Gorrell, S. Cordoba, G. W. McCaughan, andP. S. Haber, “Intrahepatic gene expression in human alcoholichepatitis,” Journal of Hepatology, vol. 45, no. 2, pp. 306–320,2006.

[50] M. G. Neuman, “Cytokines and inflamed liver,” Clinical andInvestigative Medicine, vol. 23, pp. 1–6, 1998.

[51] M. G. Neuman, N. H. Shear, S. Bellentani, and C. Tiribelli,“Role of cytokines in ethanol-induced cytotoxicity in vitro inHep G2 cells,” Gastroenterology, vol. 117, no. 7, pp. 157–166,1998.

[52] D. Wolff and A. Gerritzen, “Comparison of the Roche COBASAmplicor Monitor, Roche COBAS Ampliprep/COBASTaqman and Abbott RealTime Test assays for quantificationof hepatitis C virus and HIV RNA,” Clinical Chemistry andLaboratory Medicine, vol. 45, no. 7, pp. 917–922, 2007.

[53] J. G. McHutchison, L. M. Blatt, M. De Medina et al.,“Measurement of serum hyaluronic acid in patients withchronic hepatitis C and its relationship to liver histology.Consensus Inferon study group,” Journal of Gastroenterologyand Hepatology, vol. 15, no. 8, pp. 945–951, 2000.

[54] J. McHutchison, T. Poynard, and N. Afdhal, “Fibrosis as anend point for clinical trials in liver disease: a report of theinternational fibrosis group,” Clinical Gastroenterology andHepatology, vol. 4, no. 10, pp. 1214–1220, 2006.

[55] W. G. Stetler-Stevenson, L. A. Liotta, and D. E. Kleiner,“Extracellular matrix 6: role of matrix metalloproteinases intumor invasion and metastasis,” FASEB Journal, vol. 7, no. 15,pp. 1434–1441, 1993.

[56] J. Li, A. S. Rosman, M. A. Leo, Y. Nagai, and C. S. Lieber,“Tissue inhibitor of metalloproteinase is increased in theserum of precirrhotic and cirrhotic alcoholic patients and canserve as a marker of fibrosis,” Hepatology, vol. 19, no. 6, pp.1418–1423, 1994.

[57] F. R. Murphy, R. Issa, X. Zhou et al., “Inhibition of apop-tosis of activated hepatic stellate cells by tissue inhibitorof metalloproteinase-1 is mediated via effects on matrixmetalloproteinase inhibition. Implications for reversibility ofliver fibrosis,” Journal of Biological Chemistry, vol. 277, no. 13,pp. 11069–11076, 2002.

[58] U. Lindqvist, “Is serum hyaluronan a helpful tool in themanagement of patients with liver diseases?” Journal ofInternal Medicine, vol. 242, no. 1, pp. 67–71, 1997.

[59] F. Oberti, E. Valsesia, C. Pilette et al., “Noninvasive diagnosisof hepatic fibrosis or cirrhosis,” Gastroenterology, vol. 113, no.5, pp. 1609–1616, 1997.


[60] M. G. Phillips, V. R. Preedy, and R. D. Hughes, “Assessmentof prognosis in alcoholic liver disease: can serum hyaluronatereplace liver biopsy?” European Journal of Gastroenterology andHepatology, vol. 15, no. 9, pp. 941–944, 2003.

[61] F. Stickel, G. Poeschl, D. Schuppan et al., “Serum hyaluronatecorrelates with histological progression in alcoholic liverdisease,” European Journal of Gastroenterology and Hepatology,vol. 15, no. 9, pp. 945–950, 2003.

[62] S. Naveau, B. Raynard, V. Ratziu et al., “Biomarkers for theprediction of liver fibrosis in patients with chronic alcoholicliver disease,” Clinical Gastroenterology and Hepatology, vol. 3,no. 2, pp. 167–174, 2005.

[63] L. A. Adams, M. Bulsara, E. Rossi et al., “Hepascore: an accu-rate validated predictor of liver fibrosis in chronic hepatitis Cinfection,” Clinical Chemistry, vol. 51, no. 10, pp. 1867–1873,2005.

[64] W. M. C. Rosenberg, “Rating fibrosis progression in chronicliver diseases,” Journal of Hepatology, vol. 38, no. 3, pp. 357–360, 2003.

[65] W. M. C. Rosenberg, M. Voelker, R. Thiel et al., “Serummarkers detect the presence of liver fibrosis: a cohort study,”Gastroenterology, vol. 127, no. 6, pp. 1704–1713, 2004.

[66] G. Sebastiani and A. Alberti, “Non invasive fibrosis biomarkersreduce but not substitute the need for liver biopsy,” WorldJournal of Gastroenterology, vol. 12, no. 23, pp. 3682–3694,2006.

[67] X. Forns and R. Bataller, “Can we identify liver fibrosisin HCV-infected patients without a liver biopsy?” CurrentHepatitis Reports, vol. 2, pp. 145–151, 2003.

[68] M. Bourliere, G. Penaranda, C. Renou et al., “Validation andcomparison of indexes for fibrosis and cirrhosis predictionin chronic hepatitis C patients: proposal for a pragmaticapproach classification without liver biopsies,” Journal of ViralHepatitis, vol. 13, no. 10, pp. 659–670, 2006.

[69] A. K. Burroughs and E. Cholongitas, “Non-invasive tests forliver fibrosis: encouraging or discouraging results?” Journal ofHepatology, vol. 46, no. 5, pp. 780–781, 2007.

[70] C. S. Lieber, D. G. Weiss, T. R. Morgan, and F. Paronetto,“Aspartate aminotransferase to platelet ratio index in patientswith alcoholic liver fibrosis,” American Journal of Gastroen-terology, vol. 101, no. 7, pp. 1500–1508, 2006.

[71] M. Nakano, T. M. Worner, and C. S. Lieber, “Perivenularfibrosis in alcoholic liver injury: ultrastructure and histologicprogression,” Gastroenterology, vol. 83, no. 4, pp. 777–785,1982.

[72] Y. Ponomarenko, M. A. Leo, W. Kroll, and C. S. Lieber, “Effectsof alcohol consumption on eight circulating markers of liverfibrosis,” Alcohol and Alcoholism, vol. 37, no. 3, pp. 252–255,2002.

[73] F. Stickel, B. Hoehn, D. Schuppan, and H. K. Seitz, “Reviewarticle: nutritional therapy in alcoholic liver disease,” Alimen-tary Pharmacology and Therapeutics, vol. 18, no. 4, pp. 357–373, 2003.

[74] M. G. Neuman, J. P. Benhamou, M. Bourliere et al., “Serumtumour necrosis factor-α and transforming growth factor-βlevels in chronic hepatitis C patients are immunomodulatedby therapy,” Cytokine, vol. 17, no. 2, pp. 108–117, 2002.

[75] M. G. Neuman, J. P. Benhamou, I. M. Malkiewicz et al.,“Kinetics of serum cytokines reflect changes in the severityof chronic hepatitis C presenting minimal fibrosis,” Journal ofViral Hepatitis, vol. 9, no. 2, pp. 134–140, 2001.

[76] J. A. Cohen and M. M. Kaplan, “The SGOT/SGPT ratio. Anindicator of alcoholic liver disease,” Digestive Diseases andSciences, vol. 24, no. 11, pp. 835–838, 1979.

[77] S. Bellentani, G. Saccoccio, G. Costa et al., “Drinking habitsas cofactors of risk for alcohol induced liver damage. TheDionysos study group,” Gut, vol. 41, no. 6, pp. 845–858, 1997.

[78] R. S. Oshea, S. Dasarathy, A. J. McCullough et al., “AASLDpractice guidelines on alcoholic liver disease,” Hepatology, vol.51, no. 1, pp. 307–328, 2010.

[79] F. Stickel and H. K. Seitz, “Alcoholic steatohepatitis,” BestPractice and Research, vol. 24, no. 5, pp. 683–693, 2010.

[80] F. Stickel and C. H. Osterreicher, “The role of genetic poly-morphisms in alcoholic liver disease,” Alcohol and Alcoholism,vol. 41, no. 3, pp. 209–224, 2006.

[81] J. C. Chambers, W. Zhang, J. Sehmi et al., “Genome-wideassociation study identifies loci influencing concentrations ofliver enzymes in plasma,” Nature Genetics, vol. 43, no. 11, pp.1131–1138, 2011.

[82] F. Stickel, S. Buch, K. Lau et al., “Genetic variation in thePNPLA3 gene is associated with alcoholic liver injury incaucasians,” Hepatology, vol. 53, no. 1, pp. 86–95, 2011.

[83] F. Stickel, R. Urbaschek, D. Schuppan et al., “Serum collagentype VI and XIV and hyaluronic acid as early indicators foraltered connective tissue turnover in alcoholic liver disease,”Digestive Diseases and Sciences, vol. 46, no. 9, pp. 2025–2032,2001.

[84] L. Castera, J. Vergniol, J. Foucher et al., “Prospective com-parison of transient elastography, Fibrotest, APRI, and liverbiopsy for the assessment of fibrosis in chronic hepatitis C,”Gastroenterology, vol. 128, no. 2, pp. 343–350, 2005.

[85] S. Mueller, G. Millonig, L. Sarovska et al., “Increased liverstiffness in alcoholic liver disease: differentiating fibrosis fromsteatohepatitis,” World Journal of Gastroenterology, vol. 16, no.8, pp. 966–972, 2010.


Research Article

Cyanamide Potentiates the Ethanol-Induced Impairment ofReceptor-Mediated Endocytosis in a Recombinant Hepatic CellLine Expressing Alcohol Dehydrogenase Activity

Dahn L. Clemens,1, 2 Dean J. Tuma,1, 2 and Carol A. Casey1, 2

1 Liver Study Unit, Department of Veterans’ Affairs, Omaha, NE 68105, USA2 Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE 68198, USA

Correspondence should be addressed to Carol A. Casey, [email protected]

Received 30 July 2011; Accepted 16 November 2011


Copyright © 2012 Dahn L. Clemens et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Ethanol administration has been shown to alter receptor-mediated endocytosis in the liver. We have developed a recombinanthepatic cell line stably transfected with murine alcohol dehydrogenase cDNA to serve as an in vitro model to investigate theseethanol-induced impairments. In the present study, transfected cells were maintained in the absence or presence of 25 mMethanol for 7 days, and alterations in endocytosis by the asialoglycoprotein receptor were determined. The role of acetaldehydein this dysfunction was also examined by inclusion of the aldehyde dehydrogenase inhibitor, cyanamide. Our results showed thatethanol metabolism impaired internalization of asialoorosomucoid, a ligand for the asialoglycoprotein receptor. The addition ofcyanamide potentiated the ethanol-induced defect in internalization and also impaired degradation of the ligand in the presenceof ethanol. These results indicate that the ethanol-induced impairment in endocytosis is exacerbated by the inhibition of aldehydedehydrogenase, suggesting the involvement of acetaldehyde in this dysfunction.

1. Introduction

Previous work from our laboratories has shown that etha-nol administration impairs multiple aspects of the processof receptor-mediated endocytosis (RME) in isolated hepato-cytes [1–8]. Decreased binding, internalization, and degrada-tion of two ligands known to be internalized by RME, asial-oorosomucoid (ASOR) and epidermal growth factor (EGF),have been described. These impairments appear to be due tothe metabolism, and not to the acute presence, of ethanol.However, the direct involvement of ethanol metabolism inthese impairments has yet to be demonstrated. In order tostudy these impairments in more detail, we have developed acell culture system. Using the differentiated hepatoblastomacell line Hep G2, which does not efficiently metabolizeethanol, we have established a stably transfected cell linethat expresses alcohol dehydrogenase activity [9]. Hep G2cells were chosen since previous work has shown thatthe cells actively bind, internalize, and degrade ASOR byRME in a similar fashion to that which is seen with rat

hepatocytes [10]. These cells, designated HAD (for havingalcohol dehydrogenase), have previously been shown tometabolize ethanol efficiently and produce physiologicalamounts of acetaldehyde. In addition, impaired binding ofthe ligand, ASOR, to the asialoglycoprotein receptor has beendemonstrated in these cells after chronic incubation withethanol [11]. The impaired ligand binding was alleviatedby the addition of pyrazole (an alcohol dehydrogenase in-hibitor). These results indicated that the impairment inbinding of ASOR was dependent on the oxidation of ethanol,providing very strong evidence for a requirement of ethanolmetabolism in this dysfunction. In the current study, we havefurther characterized the impaired RME by investigating theeffects of 7 days of ethanol administration on internalization,degradation, and binding of ASOR in HAD cells. We alsoincubated cells in the presence of ethanol and cyanamide (analdehyde dehydrogenase inhibitor), to increase steady-statelevels of acetaldehyde in the cultures in order to examine theinvolvement of acetaldehyde in the impairments.



2.1. Establishment of Recombinant HAD Cells. HepG2 cellswere stably transfected with pIC-14, an eukaryotic expres-sion plasmid containing a cDNA copy of the murine alcoholdehydrogenase gene Adh-1 as described in a previous paper[9].

2.2. Cells and Culture Conditions. Hep G2 cells [12] were cul-tured in Dulbecco’s modified Eagle’s medium (DMEM) con-taining high glucose supplemented with 2 mM L-glutamine,10% fetal bovine serum, and 50 ug/mL gentamicin. Recom-binant HAD cells were cultured in the same medium con-taining 200 ug/mL hygromycin B. All cells were cultured at37◦C for 7 days in a humidified environment containing5% CO2. During ethanol and acetaldehyde metabolismstudies, the growth media were supplemented with 25 mMethanol or 0.1 mM cyanamide (or both) and were culture insealed 25 cm2 flasks to minimize evaporation of ethanol andacetaldehyde.

2.3. Binding of ASOR and Antiasialoglycoprotein Receptor(ASGP-R) Antibody to Intact Cells. Cells were removed fromthe plates with Eagle’s medium containing 2 mM EDTA andwashed twice with Eagle’s/0.1%BSA before binding exper-iments were initiated. (A) Ligand binding was performedfor 3 hours at 4◦C in the presence of 2 ug/mL 125I-ASOR.After the 3 hours, cells were washed 4-5 times with Eagle’smedium, and amounts of radioactive ligand bound weredetermined. Nonspecific binding (less than 10% of specificbinding) was determined in the presence of 100-fold excessunlabeled ligand. (B) Antibody binding was also performedat 4◦C, but in this case, the cells were incubated with primaryantibody (1 : 100 final dilution) for 1 hour, washed, andthen incubated with 125I-protein A for 1 hour [8]. Cellswere then washed, and radioactivity associated with the cellswas determined. Nonspecific binding was determined in thepresence of nonimmune serum.

2.4. Internalization and Degradation of ASOR. Internaliza-tion and degradation of 125I-ASOR was determined over atime course of 5 hours. 125I-ASOR was added to cell culturesat a final concentration of 2 ug/mL, and at the indicatedtimes, the flasks were placed on ice and the cells removedfrom the plate by the addition of 2 mM EDTA to the wells.Degradation products in the media were determined bythe presence of acid-soluble radioactivity, while internalizedligand was represented by radioactivity in the cell pellet.

2.5. General. Results are expressed as fmoles ASOR bound,internalized or degraded per million cells. A normalizedvalue of 11 ug DNA per million cells was used for thesecalculations [9]. Statistical analysis was determined using theStudent’s t test. A probability of 0.05 or less was consideredsignificant.

3. Results

Initially, we examined the ability of the ligand, ASOR, to bindto the ASGP-R after culturing HAD cells in the presence of25 mM ethanol for 7 days either in the presence or absenceof the aldehyde dehydrogenase inhibitor, cyanamide. Thesedata are shown in Figure 1. The addition of ethanol alone tothe growth media did not result in changes in ASOR bind-ing. In the presence of cyanamide, which increases acetalde-hyde concentrations to uM levels in these cells [9], the bind-ing was significantly impaired by 55% (Figure 1(a)). Thesedata show that there was no measurable impairment inligand binding in the presence of ethanol alone in this seriesof experiments, but that during ethanol metabolism and in-hibition of aldehyde dehydrogenase activity ligand bindingwas significantly decreased. Monitoring acetaldehyde levelsrevealed that acetaldehyde levels in cells cultured in the pre-sence of ethanol and cyanamide peaked at approximately 150uM at 36 hours and then gradually declined.

We next examined the ability of a polyclonal antibodyagainst human ASGP-R to bind to the cells under these ex-perimental conditions. No differences in antibody bindingwere seen in the HAD cell populations, regardless of treat-ment conditions (Figure 1(b)). Thus, the levels of the recep-tor protein are unchanged in these cultures. The data presen-ted here, along with that from Figure 1(a), indicate aninactivation of the receptor as a result of acetaldehyde-in-duced inhibition of ligand binding in the ethanol-treatedcultures containing cyanamide (a powerful inhibitor of alde-hyde dehydrogenase). This inactivation phenomenon is sim-ilar to that observed in rat hepatocytes after ethanol admin-istration for 1-2 weeks [8].

We also examined the effects of ethanol oxidation on theability of the asialoglycoprotein receptor to internalize lig-and. The results indicated that internalization of ASOR wassignificantly impaired early in the time course of internal-ization (the first 150 min, Figure 2(a)), and that the addi-tion of cyanamide further potentiated this impairment(Figure 2(b)). Since cyanamide alone (in the absence ofethanol) did not alter internalization of ASOR, the resultsimplicate a role for acetaldehyde in the impaired internaliza-tion in these cells.

We also investigated the degradation of ASOR in thesecell populations. In the presence of ethanol alone, degrada-tion of ligand was unaltered over the 150 min time courseof the experiment (Figure 3(a)). However, in the presence ofcyanamide in the culture medium, the degradation was de-creased over the 5-hour course (Figure 3(b)). Again, thesedata implicate the direct involvement of acetaldehyde in thesedysfunctions and further support the suggestion that aceta-ldehyde mediates many of the impairments associated withethanol-induced hepatic cytotoxicity.

4. Discussion

We have previously reported the establishment of a recom-binant hepatic cell line (HAD) that stably expresses alcoholdehydrogenase and oxidizes ethanol to acetaldehyde. Duringethanol metabolism, ligand binding to the ASGP-R was


0

5

10

15

20

25

30

fmol

es/m

illio

n c

ells

+cyanamide − cyanamide

ControlEthanol

(a)

0

50

100

150

200

250

300

fmol

es/m

illio

n c

ells

+cyanamide − cyanamide

ControlEthanol

(b)

Figure 1: (a) Effects of ethanol (25 mM) and cyanamide (0.1 mM) on binding of 125I-ASOR to HAD cells treated for 7 days. HAD cells weregrown to confluence and treated as described in Section 2. Ligand binding was performed at 4◦C for 3 hours and non-specific binding wasdetermined in the presence of 100-fold excess unlabeled ligand. Data are presented as means ± SEM for 6–10 sets of cells. Legends: control:HAD cells cultured in the absence of added ethanol; ethanol: HAD cells cultured in the presence of 25 mM ethanol. Values significantlydifferent from controls are indicated, ∗P < 0.05. (b) Effects of ethanol administration (25 mM) and cyanamide (0.1 mM) on binding ofanti-ASGP-R antibody to HAD cells. Cell cultures were obtained from the same conditions as in Figure 1(a). Binding of antibody wasdetermined by the ability of cells to bind a polyclonal antihuman ASGP-receptor antibody, followed by radiolabeled detection of 125I-proteinA. Data are presented as means± SEM for 6–10 experiments. Legends: control: HAD cells cultured in the absence of added ethanol; ethanol:HAD cells cultured in the presence of 25 mM ethanol.

0

20

40

60

80

100

120

140

160

180

0 30 60 120 180 240 300

Time (min)

ControlEthanol

fmol

es/m

illio

n c

ells

(a)

0 30 60 120 180 240 300

Time (min)

ControlEthanol

0

50

100

150

200

250

300

350

400

fmol

es/m

illio

n c

ells

(b)

Figure 2: Kinetics of internalization (a) and degradation (b) of 125I-ASOR in HAD cells after 7 days of treatment with 25 mM ethanol. Cellswere treated with (�) or without (�) ethanol for 7 days as previously described. At the end of the treatment period, cells were incubated withEagles medium containing 1% BSA and 2 ug per mL iodinated ASOR. At the indicated times, duplicate plates were removed from the 37◦Cincubator, the cells were removed from the plates with 2 mM EDTA, and internalization (pellet associated, a) and degradation (acid-solublemedia radioactivity, b) were determined. Data are presented as fmoles internalized or degraded per million cells and are means ± SEM for6–10 experiments. Values significantly different from controls are indicated, ∗P < 0.05.


0 30 60 120 180 240 300

Time (min)

0

50

100

150

200

250

ControlEthanol

fmol

es/m

illio

n c

ells

(a)

0 30 60 120 180 240 300

Time (min)

0

50

100

150

200

250

300

350

400

450

ControlEthanol

fmol

es/m

illio

n c

ells

(b)

Figure 3: Effects of cyanamide on kinetics of internalization (a) and degradation (b) of 125I-ASOR in HAD cell cultures grown in the presence(�) or absence (��) of 25 mM ethanol. Internalization (a) and degradation (b) of ASOR were determined as described in Figure 3. Data arepresented as fmoles ASOR degraded or internalized per million cells and are means ± SEM for 4-5 separate experiments. Values significantlydifferent from controls are indicated, ∗P < 0.05.

shown to be decreased, and the impairment was alleviatedupon the addition of an alcohol dehydrogenase inhibitor,pyrazole. These results suggested that ethanol metabolismwas necessary for the observed impairments in ligandbinding. Data presented in this study demonstrate that after 7days of ethanol exposure in HAD cells, ligand internalizationis also impaired. These impairments were most apparentin the early times of internalization of the ligand. Uponaddition of cyanamide, an aldehyde dehydrogenase inhibitorknown to increase acetaldehyde levels in cells actively metab-olizing ethanol, the impairments in internalization of ASORwere potentiated dramatically. In addition, degradation ofASOR and binding to its receptor were both significantlydecreased when the cells were incubated with ethanol andcyanamide, but not with ethanol alone. The presence ofcyanamide alone, in the absence of ethanol, did not alterany aspect of endocytosis tested. Additionally, we havedemonstrated that higher levels of acetaldehyde increase theseverity of other ethanol-metabolism-mediated impairments[13]. Taken together these data from the present study as wellas previous data implicate an important role for acetaldehydein the ethanol-induced impairments in endocytosis in HADcells.

These findings are important and relevant to our under-standing and examination of ethanol-induced alterations inprotein trafficking in the liver. It has been suggested thatthe production of acetaldehyde may be responsible for someof the hepatic impairments attributed to the oxidation ofethanol [14, 15], and the formation of acetaldehyde-proteinadducts could be the mechanism by which alcohol ultimately

damages cells [16, 17]. Although acetaldehyde is stronglyimplicated as a mediator of the ethanol-induced dysfunctionin hepatic protein secretion, the role of acetaldehyde in im-paired RME and signal transduction has not been es-tablished. In addition, this has been difficult to examinemechanistically in a culture system, since the ability of hepat-ocytes to efficiently metabolize ethanol, as well as manyother liver-specific functions, are rapidly lost in culture [18].The development of a hepatic cell line which is capable ofoxidizing ethanol, along with results presented in this studyshowing a direct involvement of acetaldehyde in ethanol-impaired RME in these cells, allows a differentiation betweenthe hepatotoxic effects of ethanol itself and its oxidation pro-ducts. This is the first report for a direct effect of acetaldehydeon a continuous endocytic process, such as internalizationand degradation of ligands by RME. The results of these stud-ies should aid in an examination of the hypothesis that theproduction of acetaldehyde and its presence in the cell couldlead to the formation of acetaldehyde-protein adducts andthrough their accumulation eventually cause hepatic dys-function.

In conclusion, these studies show a direct involvementfor ethanol metabolism and subsequent acetaldehyde pro-duction during the impaired receptor-mediated endocytosiswhich occurs after ethanol administration. Use of the HADcells allows us to have the ability to examine the mechani-sm(s) which are responsible for the ethanol-induced impair-ments in liver cell function, which we feel are related to theprogression of alcoholic liver injury. Future work with theHAD cells will aid in the elucidation of ethanol’s effects on


altered endocytosis, signalling events, and protein traffickingevents in the liver. In addition, we can now examine the roleof acetaldehyde in the impairments.

Acknowledgments

This work was supported by Grant no. AA07846 from theNational Institute on Alcohol Abuse and Alcoholism and agrant from the University of Nebraska Medical Center andthe Veterans Administration.

References

[1] C. A. Casey, S. L. Kragskow, M. F. Sorrell, and D. J. Tuma,“Chronic ethanol administration impairs the binding andendocytosis of asialo-orosomucoid in isolated hepatocytes,”Journal of Biological Chemistry, vol. 262, no. 6, pp. 2704–2710,1987.

[2] C. A. Casey, S. L. Kragskow, M. F. Sorrell, and D. J. Tuma,“Ethanol-induced impairments in receptor-mediated endocy-tosis of asialoorosomucoid in isolated rat hepatocytes: timecourse of impairments and recovery after ethanol withdrawal,”Alcoholism, vol. 13, no. 2, pp. 258–263, 1989.

[3] C. A. Casey, S. L. Kragskow, M. F. Sorrell, and D. J. Tuma,“Effect of chronic ethanol administration on total asialogly-coprotein receptor content and intracellular processing ofasialoorosomucoid in isolated rat hepatocytes,” Biochimica etBiophysica Acta, vol. 1052, no. 1, pp. 1–8, 1990.

[4] C. A. Casey and D. J. Tuma, “Receptors and endocytosis,” inMolecular and Cell Biology of the Liver, A. V. LeBouton, Ed.,chapter 5, pp. 117–141, CRC, Ann Arobr, Mich, USA, 1993.

[5] C. A. Casey, S. L. Kragskow, M. F. Sorrell, and D. J. Tu-ma, “Zonal differences in ethanol-induced impairments inreceptor-mediated endocytosis of asialoglycoproteins in iso-lated rat hepatocytes,” Hepatology, vol. 13, no. 2, pp. 260–266,1991.

[6] T. M. McCashland, D. J. Tuma, M. F. Sorrell, and C. A. Casey,“Zonal differences in ethanol-induced impairments in hepaticreceptor binding,” Alcohol, vol. 10, no. 6, pp. 549–554, 1993.

[7] D. D. Dalke, M. F. Sorrell, C. A. Casey, and D. J. Tuma,“Chronic ethanol administration impairs receptor-mediatedendocytosis of epidermal growth factor by rat hepatocytes,”Hepatology, vol. 12, no. 5, pp. 1085–1091, 1990.

[8] B. L. Tworek, D. J. Tuma, and C. A. Casey, “Decreased bindingof asialoglycoproteins to hepatocytes from ethanol-fed rats:consequence of both impaired synthesis and inactivation ofthe asialoglycoprotein receptor,” Journal of Biological Chem-istry, vol. 271, no. 5, pp. 2531–2538, 1996.

[9] D. L. Clemens, C. M. Halgard, R. R. Miles, M. F. Sorrell,and D. J. Tuma, “Establishment of a recombinant hepaticcell line stably expressing alcohol dehydrogenase,” Archives ofBiochemistry and Biophysics, vol. 321, no. 2, pp. 311–318, 1995.

[10] R. J. Fallon and A. L. Schwartz, “Asialoglycoprotein receptorphosphorylation and receptor-mediated endocytosis in hep-atoma cells. Effect of phorbol esters,” Journal of BiologicalChemistry, vol. 263, no. 26, pp. 13159–13166, 1988.

[11] D. L. Clemens, C. M. Halgard, J. R. Cole, R. M. Miles, M. F.Sorrell, and D. J. Tuma, “Impairment of the asialoglycoproteinreceptor by ethanol oxidation,” Biochemical Pharmacology, vol.52, no. 10, pp. 1499–1505, 1996.

[12] B. B. Knowles, C. C. Howe, and D. P. Aden, “Human hepato-cellular carcinoma cell lines secrete the major plasma proteinsand hepatitis B surface antigen,” Science, vol. 209, no. 4455, pp.497–499, 1980.

[13] D. L. Clemens, A. Forman, T. R. Jerrells, M. F. Sorrell, andD. J. Tuma, “Relationship between acetaldehyde levels and cellsurvival in ethanol-metabolizing hepatoma cells,” Hepatology,vol. 35, no. 5, pp. 1196–1204, 2002.

[14] R. B. Jennett, D. J. Tuma, and M.F. Sorrell, “Effects of aceta-ldehyde on hepatic proteins,” in Progress in Liver Disease, H.Popper and F. Schaffner, Eds., pp. 325–333, Saunders, Pa, USA,1990.

[15] C. S. Lieber, “Metabolic effects of acetaldehyde,” BiochemicalSociety Transactions, vol. 16, no. 3, pp. 241–247, 1988.

[16] M. F. Sorrell and D. J. Tuma, “Hypothesis: alcoholic liver injuryand the covalent binding of acetaldehyde,” Alcoholism, vol. 9,no. 4, pp. 306–309, 1985.

[17] D. J. Tuma, T. Hoffman, and M. F. Sorrell, “The chemistry ofacetaldehyde-protein adducts,” Alcohol and Alcoholism, vol. 1,pp. 271–276, 1991.

[18] S. Zakhari, “Overview: how is alcohol metabolized by thebody?” Alcohol Research and Health, vol. 29, no. 4, pp. 245–254, 2006.


Review Article

MicroRNA Signature in Alcoholic Liver Disease

Shashi Bala and Gyongyi Szabo

Department of Medicine, University of Massachusetts Medical School, Worcester, MA 01605, USA

Correspondence should be addressed to Gyongyi Szabo, [email protected]

Received 31 August 2011; Accepted 30 September 2011


Copyright © 2012 S. Bala and G. Szabo. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Alcoholic liver disease (ALD) is a major global health problem. Chronic alcohol use results in inflammation and fatty liver, and insome cases, it leads to fibrosis and cirrhosis or hepatocellular carcinoma. Increased proinflammatory cytokines, particularly TNFalpha, play a central role in the pathogenesis of ALD. TNF alpha is tightly regulated at transcriptional and posttranscriptional levels.Recently, microRNAs (miRNAs) have been shown to modulate gene functions. The role of miRNAs in ALD is getting attention,and recent studies suggest that alcohol modulates miRNAs. Recently, we showed that alcohol induces miR-155 expression bothin vitro (RAW 264.7 macrophage) and in vivo (Kupffer cells, KCs of alcohol-fed mice). Induction of miR-155 contributed toincreased TNF alpha production and to the sensitization of KCs to produce more TNF alpha in response to LPS. In this paper, wesummarize the current knowledge of miRNAs in ALD and also report increased expression of miR-155 and miR-132 in the totalliver as well as in isolated hepatocytes and KCs of alcohol-fed mice. Our novel finding of the alcohol-induced increase of miRNAsin hepatocytes and KCs after alcohol feeding provides further insight into the evolving knowledge regarding the role of miRNAs inALD.

1. Introduction

MicroRNAs (miRNAs) are 20–22 nucleotides long noncod-ing RNAs that were first described in 1993 [1]. MiRNAsplay a central role in diverse cellular processes includingdevelopment, immunity, cell-cycle control, metabolism, viralor bacterial disease, stem-cell differentiation, and oncogen-esis [2–4]. In general, miRNAs are transcribed from RNApolymerase II or III in the nucleus and transported to thecytoplasm, where they are processed into mature miRNAs[4]. Mature miRNAs can target hundreds of genes by eitherbinding to the 3′ or 5′ untranslated (UTR) region of mRNA[4]. Emerging evidence suggests that miRNAs not onlytarget mRNAs but also they are capable of modulatingtranscription and methylation processes [5–7]. Moreover,not only the sense strand (miRNA) of mature miRNAmodulates gene function, but also the antisense strand (starform; ∗) plays an important role in the miRNA regulatorynetwork [8]. However, the biological significance of theantisense strand (star form) is largely unknown but is slowlygetting attention. In a short time, miRNA research hasreceived tremendous attention due to their fine-tuning roles

in almost all biological pathways. Moreover, disease-specifictissue miRNA signatures have been identified in variousetiologies such as hepatocellular carcinoma (HCC), hepatitisC virus (HCV), hepatitis B virus (HBV), cardiac disease,neuroinflammation, rheumatic arthritis (RA), and variouscancers [3, 9–14]. In this paper, we highlight the emergingroles of miRNAs in alcoholic liver disease.

1.1. MiRNA in Innate Immune Response. Innate immunityis the first line of host defense against foreign pathogensand also in response to damaged self (endogenous dangersignals). Toll-like receptors (TLRs) are the most widely stud-ied danger signal sensors. MiRNAs have been implicated invarious immune responses and are believed to be essentialregulators of these processes [15]. The number of miRNAsinvolved in immune responses is growing, and among them,miR-155, -146a, -125b, -132, -9, -212 and -181, are thekey players and are elegantly reviewed in [16, 17]. Theinflammation-related miRNAs deserve attention in ALD, asthe activation of the innate immune system is a hallmark ofalcoholic steatohepatitis.


2. MiRNA in Alcoholic Liver Disease

2.1. Alcoholic Liver Disease (ALD). Alcoholic liver disease(ALD) is a global health-related problem, which contributessignificantly to liver-related mortality. Increased inflamma-tion and fat accumulation are the hallmarks of ALD. Theprogression of ALD involves a complex network of signalingmolecules and chronic alcohol abuse in some cases leadsto liver cirrhosis [18]. Alcohol alone or its metabolites (ac-etaldehyde) act on multiple signaling pathways and resultin increased intestine permeability and ROS generation[19, 20]. Increased gut permeability is associated withtranslocation of bacteria and bacterial products into thelumen of the intestine, which results in the imbalance ofintestine homeostasis [20]. LPS is a major component ofa Gram-negative bacterial cell wall, and it is detoxified inthe liver via both parenchymal and nonparenchymal cells[21, 22]. It is believed that increased LPS in the circulationdisrupts the liver homeostasis, resulting in Kupffer cell(KC; liver macrophages) activation. Upon activation, KCsproduce TNF alpha, which then induces the activation ofother signaling cascades to amplify the inflammation. TNFalpha-induced inflammation is more prevalent in alcoholichepatitis [23]. The role of the LPS/TLR4 axis has beenappreciated in ALD, since TLR4 KO mice have been shownto be protected from liver damage in a mouse model of ALD[24].

2.2. MiRNA Profiling in the Livers of Alcohol-Fed Mice.Alcohol has been shown to modulate the epigenetic factorsin various organs including liver and brain and was reviewedrecently [25]. As alcohol exerts epigenetic effects, it is con-ceivable that alcohol might target miRNAs to regulate genefunctions. To date, there are very few studies related tothe roles of miRNAs in ALD. Previously, our laboratorydemonstrated the differential expression of some miRNAsin the livers of alcohol-fed mice by microarray analysis[26]. MiR-27b, miR-214, miR-199a-3p, miR-182, miR-183,miR-200a, and miR-322 were found to be downregulated,whereas miR-705 and miR-1224 were increased after 4 weeksof alcohol feeding in mice [26]. However, the physiologicalrelevance of these miRNAs in ALD has yet to be determined.

2.3. The Role of miRNA in Alcohol-Induced Intestinal Perme-ability. Alcohol and its metabolites are known to increaseintestinal permeability [27]. In the past, various signalingmolecules and transcription factors were reported to beinvolved in alcohol-mediated intestinal permeability. Re-cently, miR-212 has been identified as a new player andimplicated in alcohol-induced intestinal permeability, whereit targets a major tight junction protein, Zonula occludens 1(ZO-1) [28]. ZO-1 plays an essential role in the regulation ofintestinal permeability. Induction of miR-212 and decreasein ZO-1 protein were observed both in colon biopsy samplesfrom patients with ALD and in alcohol-treated CaCO-2 cells[28].

Recently, miR-29a and miR-122a were reported to mod-ulate intestinal membrane permeability. MiR-29a regulatesintestinal membrane permeability in patients with irritable

bowel syndrome (IBS) [29]. Increased expression of miR-29a was found in blood microvesicles, small bowel, andcolon tissues of IBS patients. MiR-29a targets the glutaminesynthetase gene (GLUL), which in turn regulates intestinalmembrane permeability [29]. MiR-122a was found to targetoccludin (a transmembrane tight junction protein) bothin Caco-2 cells and mice enterocytes and hence plays animportant role in regulating intestinal permeability [30]. Thepotential role of these miRNAs in alcohol-induced intestinalpermeability is yet to be determined.

2.4. Potential Role of miRNA in Alcohol-Mediated OxidativeStress. Not only does increased circulating endotoxin playa crucial role in ALD, but also increased reactive oxygenspecies (ROS and oxidative stress) production contributes tothe pathogenesis of ALD [20, 27]. In agreement with this,NADPH oxidase is reported as a major source of oxidantsin ALD, and mice deficient in this oxidase (p47phox−/−) wereprotected from early alcohol-induced liver injury [31]. Ingeneral, the potential role of miRNAs in oxidative stress-mediated etiologies is emerging. In line with this, miR-27a∗, miR-27b∗, miR-29b∗, miR-24-2∗, and miR-21∗ werereported to be differentially expressed in response to H2O2-induced oxidative stress in RAW 264.7 macrophage [32].Furthermore, overexpression of miR-27b∗ suppressed LPS-induced activation of NF-κB [32]. These data suggest thatmacrophage function can be regulated by oxidative stress-responsive miRNAs via modulating the NF-κB pathway [32].Interestingly, in this study, only the star form of maturemiRNA was found to play a role in H2O2-induced oxidativestress. To our knowledge, there are no studies indicating thedirect involvement of miRNAs in alcohol-induced oxidativestress.

Alcohol has been shown to downregulate miR-199 inrat liver sinusoidal endothelial cells (LSECs) and human en-dothelial cells [33]. Decreased miR-199 was associated withincreased mRNA expression of endothelin-1 (ET-1) and hyp-oxia-inducible factor-1α (HIF-1α) [33]. Authors concludedthat alcohol-induced ET-1 likely contributes to inflammationin patients with cirrhosis.

2.5. Role of miRNA-155 in Kupffer Cell Activation in AlcoholicLiver Disease. Previous studies from our laboratory andothers have shown increased TNF alpha in vivo and invitro alcohol models [34, 35]. Moreover, increased TNFalpha was found in patients with alcoholic hepatitis [23].Previously, alcohol has been shown to regulate TNF alphamRNA stability in RAW 264.7 macrophage and KCs [35].Furthermore, miRNAs modulate gene expression both atposttranscriptional and translational levels [3, 5, 6]. Recently,we have shown that prolonged alcohol exposure inducesmiR-155 in RAW264.7 macrophage and KCs [34]. However,no changes were found in the expression of miR-146a and-125b, which are also involved in immune responses. Thisobservation suggests that alcohol particularly targets miR-155 [34]. We further showed that miR-155 expression corre-lates with TNF alpha levels. Functionally, miR-155 regulatesTNF alpha mRNA stability, and thereby contributes toincreased TNF alpha in KCs of alcohol-fed mice. TNF alpha


is tightly regulated both at transcriptional and posttranscrip-tional levels. Posttranscriptional regulation of TNF alphaincludes mRNA stability, polyadenylation, and translationalinitiation, which target adenine and uridine- (AU-) richelements (AREs) in its 3UTR [36]. Trans-acting factors suchas HuR, tristetraprolin (TTP), and TIA-1, bind to 3UTR ofTNF alpha and modulate its expression either by stabilizingor destabilizing its transcripts [36]. Interestingly, chronic al-cohol has been shown to increase TNF alpha mRNA stabilityvia HuR protein [35, 37, 38]. As our results indicatedthat miR-155 enhances TNF alpha transcript stability, it isplausible to suggest that miR-155 might be exerting its effectvia targeting HuR or other trans-acting factors. Hence, ourstudy directly or indirectly implicates the role of miR-155 inTNF alpha regulation. Given the fact that one miRNA canregulate multiple genes of a pathway; it is likely that miR-155might regulate other genes, which are directly or indirectlyinvolved in TNF alpha regulation.

2.6. Induction of miRNA in the Livers of Alcohol-Fed Mice.As described above, previously, we found induction of miR-155 in alcohol-exposed RAW264.7 macrophage and KCsof alcohol-fed animals [34]. Here, we examined the liverexpression of other miRNA including miR-132, -125b, and-146a which also regulate various immune responses. MiR-132 was shown to play a role in neuroinflammation [39]and also regulates innate antiviral immunity, where it targetsthe p300 transcriptional coactivator [40]. However, the roleof miR-132 in the alcohol-mediated TLR response is yetto be elucidated; therefore, in this study, we determinedthe effect of alcohol on miR-132. MiR-146a is linked withendotoxin tolerance, where it regulates IRAK-1 and TRAF-6 and acts as a negative regulator of TLR4 signaling [41],whereas miR-125b limits TNF alpha production in RAW264.7 macrophage [42].

Among the miRNAs tested, we found significant induc-tion of miR-132 in the livers of alcohol-fed mice (Figure 1).As expected, a significant increase in miR-155 was alsoobserved in the livers of alcohol-fed mice (Figure 1), andthis result was consistent with our previous report [21].Contrary to this observation, no significant changes wereobserved in miR-125b and -146a expression (Figure 1). Themice used in the study were housed and cared for as peranimal protocols approved by the Institutional Animal Useand Care Committee of the University of MassachusettsMedical School.

2.7. Increase in miR-155 in Hepatocytes of Alcohol-Fed Mice.Because of the crucial role of miR-155 in inflammation,most studies are focused on its role in immune cells suchas monocytes, macrophages, dendritic cells, and T cells (re-viewed in [16, 17, 41]). Induction of miR-155 has beenreported in various inflammatory-related diseases such asRA, neuro- or autoimmune-inflammation and various can-cers [43–46]. The concept of cell-specific effects of miRNAis emerging, and reports showing the role of immune-related miRNAs in hepatocytes are sparse. Recent studiessuggest that hepatocytes do respond to LPS-induced TLRsignaling and express functional NOD1 and 2 receptors [47].

0

1

2

3

0.022

0.034

Total liver miRNAs

Fold

ch

ange

/Sn

oRN

A 2

02

miR-125b

Pair-fedEthanol-fed

miR-155 miR-132 miR-146a

Figure 1: Increased expression of miR-155 and miR-132 in thelivers of alcohol-fed mice. C57BL/6 eight-week-old female mice(n = 6/group) were fed with Lieber-Decarli diet either containing5% alcohol (ethanol-fed) or isocaloric liquid diet (pair-fed) for 4weeks. After 4 weeks, the livers were isolated and stored in RNAlater (Qiagen) for RNA analysis at −80◦C. Total RNA from liverswas isolated using the miRNeasy kit (Qiagen), and quantificationof miRNAs was carried out by TaqMan miRNA assays (AppliedBiosystems). The data were normalized to SnoRNA202 (endoge-nous control) and shown as fold change over the pair-fed controlgroup. Data represent mean values ± S.E.M. Statistical significancewas determined using T-test (two-tailed).

Therefore, next, we examined the effect of alcohol on miRNAexpression in hepatocytes, which are the predominant liver-cell population.

Hepatocytes were isolated from mice fed with Lieber-DeCalri diet either with 5% alcohol (ethanol-fed) or isoca-loric diet (pair-fed) for 4 weeks [34]. Hepatocytes isolationwas performed with the method described earlier by ourgroup [24]. Total RNA was isolated with the miRNeasy kit(Qiagen) and subjected to miRNA analysis as describedpreviously [34]. Briefly, TaqMan miRNA assay (Applied Bi-osystems) was used, and snoRNA202 was used as internalcontrol to normalize the technical variations between thesamples. Fold change was calculated in comparison to pair-fed mice.

Interestingly, we found increased expression of miR-155in hepatocytes of alcohol-fed mice compared to pair-fed mice(Figure 2). To our best knowledge, this is the first studyreporting the induction of miR-155 in hepatocytes afteralcohol feeding. No obvious changes were observed in miR-146a expression and there was minimal increase in miR-132expression in hepatocytes of alcohol-fed mice (Figure 2). Incontrast, expression of miR-125b was found to be downreg-ulated in hepatocytes after alcohol feeding (Figure 2). Thephysiological relevance of alcohol-induced miR-155 in hepa-tocytes is the subject of ongoing investigation.

In diet-induced (methyl-deficient diet) nonalcoholicsteaohepatitis, increased miR-155 was associated with de-creased levels of C/EBP-β and SOCS1 proteins [48]. In vitro


0

1

2

3

0.05

0.02

Hepatocytes miRNAs

Fold

ch

ange

/Sn

oRN

A 2

02

miR-125bmiR-155 miR-132 miR-146a

Pair-fedEthanol-fed

Figure 2: Induction of miR-155 in hepatocytes of alcohol-fed mice.Hepatocytes were isolated from the mice (n = 5-6) after 4 weeksof feeding. Briefly, the livers were perfused, followed by digestion.Parenchymal cells (hepatocytes) were collected after low-speedcentrifugation, washed, and were plated onto 6-well collagen-coatedplates (BD-Biosciences). After 3 h, floating cells were removed,and adherent cells were washed twice with PBS and lysed inQIAzole (Qiagen). Total RNA was isolated and miRNA expressionwas quantified by TaqMan miRNA assays (Applied Biosystems).The data were normalized to SnoRNA202 (endogenous control)and shown as the fold change over the pair-fed control group.Data represent mean values ± S.E.M. Statistical significance wasdetermined using non-parametric Mann-Whitney test.

overexpression of miR-155 in mouse primary hepatocytesresulted in a decreased level of C/EBP-β and SOCS1 proteins[48]. Both C/EBP-β and SOCS1 are tumor suppressors [49]and often downregulated in hepatocellular carcinomas andhepatoblastomas [50]. The relevance of C/EBP-β and SOCS1could be translated to ALD due to the fact that SOCS1 notonly acts as a tumor suppressor, but also is a negativeregulator of LPS signaling, while chronic alcohol use resultsin increased inflammatory cytokine production. Moreover,SOCS1 is also an important mediator of cellular oxidativestress [51].

C/EBP-β regulates several hepatic genes such as catalaseand methionine adenosyltransferase 1a that are involved inregulation of oxidative stress [52, 53]. As oxidative stressplays an essential role in the pathogenesis of ALD, it is reason-able to argue that increased miR-155 observed in hepatocytesof alcohol-fed mice might contribute to increased oxidativestress via the regulation of genes involved in oxidative stresspathways. However, further studies are needed to prove thisspeculation in ALD.

More recently, miR-155 has been demonstrated to playa role in antiviral immunity against HBV infection inhuman hepatoma cells, HepG2 [54]. Overexpression of miR-155 resulted in decreased SOCS1, which enhances STAT 1and 3 phosphorylation [54]. Increased expression of sev-eral interferon-inducible antiviral genes was observed afterectopic expression of miR-155 in human hepatoma cells [54].

0

1

2

3

0.0004

0.006

Kupffer cells miRNAs

Fold

ch

ange

/Sn

oRN

A 2

02

Pair-fedEthanol-fed

miR-155 miR-132

Figure 3: Upregulation of miR-155 and -132 in Kupffer cells ofalcohol-fed mice. Kupffer cells were isolated either from pair-fedor alcohol-fed mice (n = 12–14/group, cells from two mice werepooled) after 4 weeks of feeding. Cells were plated and after 2 hof incubation nonadherent cells were removed. Fresh medium wasadded to the adherent, and they were rested overnight. Next day,cells were washed with PBS and lysed with QIAzole (Qiagen). TotalRNA isolated using the miRNeasy kit was used quantify miRNA asdescribed above. Data represent mean values ± S.E.M. Statisticalsignificance was determined using T-test (two-tailed).

It was concluded that miR-155 acts as a positive regulator ofJAK/STAT signaling via targeting SOCS1 and hence increasesthe expression of some IFN-inducible antiviral genes such asISG15 and MxA [54].

In this study, we also found decreased miR-125b inhepatocytes of alcohol-fed mice (Figure 2). Downregulationof miR-125b has been reported in HCC and is associated withincreased placenta growth factor (PIGF) [55]. However, thephysiological relevance of miR-125b downregulation in ALDhas yet to be explored.

2.8. Increase of miR-155 and miR-132 in Kupffer Cells ofAlcohol-Fed Mice. Next, we examined the expression ofmiRNAs in KCs of alcohol-fed mice. KCs were isolated asdescribed earlier [34]. We focused our study on miR-155 andmiR-132, as the expression of these miRNAs was increasedin the livers after alcohol feeding (Figure 1). Interestingly,we not only observed a significant increase in miR-155,which was consistent with our previous report [34], butalso induction of miR-132 in the KCs of alcohol-fed mice(Figure 3).

Induction of miR-132 in alcohol-fed mice is interesting,as the role of this miRNA in innate immunity is notmuch appreciated. Recently, miR-132 was shown to play arole in the innate viral response, where it targets p300, atranscriptional coactivator [40]. Chronic alcohol use predis-poses individuals to infections (bacterial and viral), and theinduction of miR-132 after alcohol feeding is of great interest.We also found a modest increase of miR-132 in hepatocytesof alcohol-fed mice. The physiological relevance of miR-132


in the alcohol-mediated immune response is currently underinvestigation.

Collectively, our results indicate an increase in miR-155in different cell populations of the liver after alcohol feeding.It is most likely that the induction of miR-155 we observed inhepatocytes of alcohol-fed mice is involved in both LPS andoxidative stress signaling pathways, and hence contributes tothe progression of ALD.

3. Conclusions

Our understanding of the role of miRNAs in liver diseaseis expanding, and the current studies suggest that miRNAsplay a crucial role in alcoholic liver disease. However, in-depth understanding of association of miRNAs with theircell-specific roles in ALD is yet to be explored. We notonly showed induction of miR-155, but also miR-132 in thelivers and KCs of alcohol-fed mice. Furthermore, an increasein miR-155 was also observed in hepatocytes of alcohol-fed mice. The cell-specific effect of these miRNAs in ALDdeserves further investigation.

Acknowledgments

This work was supported by NIAAA Grant no. AA011576 (G.Szabo). The authors thank Shiv Mundkur for his technicalhelp.

References

[1] R. C. Lee, R. L. Feinbaum, and V. Ambros, “The C. elegansheterochronic gene lin-4 encodes small RNAs with antisensecomplementarity to lin-14,” Cell, vol. 75, no. 5, pp. 843–854,1993.

[2] V. Ambros, “The functions of animal microRNAs,” Nature,vol. 431, no. 7006, pp. 350–355, 2004.

[3] S. Bala, M. Marcos, and G. Szabo, “Emerging role of microR-NAs in liver diseases,” World Journal of Gastroenterology, vol.15, no. 45, pp. 5633–5640, 2009.

[4] D. P. Bartel, “MicroRNAs: target recognition and regulatoryfunctions,” Cell, vol. 136, no. 2, pp. 215–233, 2009.

[5] F. Sato, S. Tsuchiya, S. J. Meltzer, and K. Shimizu, “MicroRNAsand epigenetics,” FEBS Journal, vol. 10, pp. 1598–1609, 2011.

[6] L. Sinkkonen, T. Hugenschmidt, P. Berninger et al., “MicroR-NAs control de novo DNA methylation through regulationof transcriptional repressors in mouse embryonic stem cells,”Nature Structural and Molecular Biology, vol. 15, no. 3, pp.259–267, 2008.

[7] R. Garzon, S. Liu, M. Fabbri et al., “MicroRNA-29b inducesglobal DNA hypomethylation and tumor suppressor gene re-expression in acute myeloid leukemia by targeting directlyDNMT3A and 3B and indirectly DNMT1,” Blood, vol. 113, no.25, pp. 6411–6418, 2009.

[8] H. Zhou, X. Huang, H. Cui et al., “MiR-155 and its star-formpartner miR-155∗ cooperatively regulate type I interferonproduction by human plasmacytoid dendritic cells,” Blood,vol. 116, no. 26, pp. 5885–5894, 2010.

[9] K. Z. Qu, K. Zhang, H. Li, N. H. Afdhal, and M. Albitar,“Circulating MicroRNAs as biomarkers for hepatocellularcarcinoma,” Journal of Clinical Gastroenterology, vol. 45, no.4, pp. 355–360, 2011.

[10] S. Ura, M. Honda, T. Yamashita et al., “Differential microRNAexpression between hepatitis B and hepatitis C leading diseaseprogression to hepatocellular carcinoma,” Hepatology, vol. 49,no. 4, pp. 1098–1112, 2009.

[11] A. Budhu, H. L. Jia, M. Forgues et al., “Identification ofmetastasis-related microRNAs in hepatocellular carcinoma,”Hepatology, vol. 47, no. 3, pp. 897–907, 2008.

[12] T. Nakasa, Y. Nagata, K. Yamasaki, and M. Ochi, “A mini-review: MicroRNA in arthritis,” Physiological Genomics, vol.43, no. 10, pp. 566–570, 2011.

[13] O. C. Maes, H. M. Chertkow, E. Wang, and H. M. SChipper,“MicroRNA: implications for Alzheimer disease and otherhuman CNS disorders,” Current Genomics, vol. 10, no. 3, pp.154–168, 2009.

[14] J. Jiang, Y. Gusev, I. Aderca et al., “Association of microRNAexpression in hepatocellular carcinomas with hepatitis infec-tion, cirrhosis, and patient survival,” Clinical Cancer Research,vol. 14, no. 2, pp. 419–427, 2008.

[15] D. Baltimore, M. P. Boldin, R. M. O’Connell, D. S. Rao, andK. D. Taganov, “MicroRNAs: new regulators of immune celldevelopment and function,” Nature Immunology, vol. 9, no. 8,pp. 839–845, 2008.

[16] R. M. O’Connell, D. S. Rao, A. A. Chaudhuri, and D. Balti-more, “Physiological and pathological roles for microRNAs inthe immune system,” Nature Reviews Immunology, vol. 10, no.2, pp. 111–122, 2010.

[17] D. T. Gracias and P. D. Katsikis, “MicroRNAs: key componentsof immune regulation,” Advances in Experimental Medicineand Biology, pp. 15–26, 2012.

[18] B. Gao, E. Seki, D. Brenner, S. L. Friedman, J. Cohen, and L. E.Nagy, “Innate immunity and alcoholic liver disease,” AmericanJournal of Physiology, vol. 300, pp. G516–G525, 2011.

[19] R. K. Rao, “Acetaldehyde-induced barrier disruption and par-acellular permeability in caco-2 cell monolayer,” Methods inMolecular Biology, vol. 447, pp. 171–183, 2008.

[20] G. Szabo and S. Bala, “Alcoholic liver disease and the gut-liveraxis,” World Journal of Gastroenterology, vol. 16, no. 11, pp.1321–1329, 2010.

[21] G. Szabo, S. Bala, J. Petrasek, and A. Gattu, “Gut-liver axis andsensing microbes,” Digestive Diseases, vol. 28, no. 6, pp. 737–744, 2010.

[22] M. J. Scott and T. R. Billiar, “β2-integrin-induced p38MAPK activation is a key mediator in the CD14/TLR4/MD2-dependent uptake of lipopolysaccharide by hepatocytes,” Jour-nal of Biological Chemistry, vol. 283, no. 43, pp. 29433–29446,2008.

[23] D. B. Hill, S. Barve, S. Joshi-Barve, and C. McClain, “Increasedmonocyte nuclear factor-κB activation and tumor necrosisfactor production in alcoholic hepatitis,” Journal of Laboratoryand Clinical Medicine, vol. 135, no. 5, pp. 387–395, 2000.

[24] I. Hritz, P. Mandrekar, A. Velayudham et al., “The critical roleof toll-like receptor (TLR) 4 in alcoholic liver disease is inde-pendent of the common TLR adapter MyD88,” Hepatology,vol. 48, no. 4, pp. 1224–1231, 2008.

[25] P. Mandrekar, “Epigenetic regulation in alcoholic liver dis-ease,” World Journal of Gastroenterology, vol. 17, no. 20, pp.2456–2464, 2011.

[26] A. Dolganiuc, J. Petrasek, K. Kodys et al., “MicroRNA ex-pression profile in lieber-decarli diet-induced alcoholic andmethionine choline deficient diet-induced nonalcoholic ste-atohepatitis models in mice,” Alcoholism, vol. 33, no. 10, pp.1704–1710, 2009.

[27] A. Keshavarzian, A. Farhadi, C. B. Forsyth et al., “Evidence thatchronic alcohol exposure promotes intestinal oxidative stress,


intestinal hyperpermeability and endotoxemia prior to devel-opment of alcoholic steatohepatitis in rats,” Journal of He-patology, vol. 50, no. 3, pp. 538–547, 2009.

[28] Y. Tang, A. Banan, C. B. Forsyth et al., “Effect of alcohol onmiR-212 expression in intestinal epithelial cells and its poten-tial role in alcoholic liver disease,” Alcoholism, vol. 32, no. 2,pp. 355–364, 2008.

[29] Q. Zhou, W. W. Souba, C. M. Croce, and G. N. Verne, “Micro-RNA-29a regulates intestinal membrane permeability in pa-tients with irritable bowel syndrome,” Gut, vol. 59, no. 6, pp.775–784, 2010.

[30] D. Ye, S. Guo, R. Al-Sadi, and T. Y. Ma, “MicroRNA regulationof intestinal epithelial tight junction permeability,” Gastroen-terology, vol. 141, no. 4, pp. 1323–1333, 2011.

[31] H. Kono, I. Rusyn, M. Yin et al., “NADPH oxidase-derivedfree radicals are key oxidants in alcohol-induced liver disease,”Journal of Clinical Investigation, vol. 106, no. 7, pp. 867–872,2000.

[32] S. Thulasingam, C. Massilamany, A. Gangaplara et al., “miR-27b∗, an oxidative stress-responsive microRNA modulatesnuclear factor-kB pathway in RAW 264.7 cells,” Molecular andCellular Biochemistry, vol. 352, no. 1-2, pp. 181–188, 2011.

[33] S. Yeligar, H. Tsukamoto, and V. K. Kalra, “Ethanol-inducedexpression of ET-1 and ET-BR in liver sinusoidal endothelialcells and human endothelial cells involves hypoxia-induciblefactor-1α and microRNA-199,” Journal of Immunology, vol.183, no. 8, pp. 5232–5243, 2009.

[34] S. Bala, M. Marcos, K. Kodys et al., “Up-regulation of micro-RNA-155 in macrophages contributes to increased TumorNecrosis Factor α (TNFα) production via increased mRNAhalf-life in alcoholic liver disease,” Journal of Biological Chem-istry, vol. 286, no. 2, pp. 1436–1444, 2011.

[35] R. Kishore, M. R. McMullen, E. Cocuzzi, and L. E. Nagy,“Lipopolysaccharide-mediated signal transduction: stabiliza-tion of TNF-alpha mRNA contributes to increased lipopoly-saccharide-stimulated TNF-alpha production by Kupffer cellsafter chronic ethanol feeding,” Comparative Hepatology, vol.14, no. 3, supplement 1, p. S31, 2004.

[36] M. Piecyk, S. Wax, A. R. P. Beck et al., “TIA-1 is a translationalsilencer that selectively regulates the expression of TNF-α,”EMBO Journal, vol. 19, no. 15, pp. 4154–4163, 2000.

[37] R. Kishore, M. R. McMullen, and L. E. Nagy, “Stabilization oftumor necrosis factor α mRNA by chronic ethanol: role of A+ U-rich elements and p38 mitogen-activated protein kinasesignaling pathway,” Journal of Biological Chemistry, vol. 276,no. 45, pp. 41930–41937, 2001.

[38] M. R. McMullen, E. Cocuzzi, M. Hatzoglou, and L. E. Nagy,“Chronic ethanol exposure increases the binding of HuR tothe TNFα 3′-untranslated region in macrophages,” Journal ofBiological Chemistry, vol. 278, no. 40, pp. 38333–38341, 2003.

[39] I. Shaked, A. Meerson, Y. Wolf et al., “MicroRNA-132 poten-tiates cholinergic anti-inflammatory signaling by targetingacetylcholinesterase,” Immunity, vol. 31, no. 6, pp. 965–973,2009.

[40] D. Lagos, G. Pollara, S. Henderson et al., “MiR-132 regulatesantiviral innate immunity through suppression of the p300transcriptional co-activator,” Nature Cell Biology, vol. 12, no.5, pp. 513–519, 2010.

[41] M. A. Nahid, M. Satoh, and E. K. L. Chan, “MicroRNA in TLRsignaling and endotoxin tolerance,” Cellular and MolecularImmunology, vol. 8, no. 5, pp. 388–403, 2011.

[42] E. Tili, J. J. Michaille, A. Cimino et al., “Modulation of miR-155 and miR-125b levels following lipopolysaccharide/TNF- αstimulation and their possible roles in regulating the response

to endotoxin shock,” Journal of Immunology, vol. 179, no. 8,pp. 5082–5089, 2007.

[43] M. Kurowska-Stolarska, S. Alivernini, L. E. Ballantine et al.,“MicroRNA-155 as a proinflammatory regulator in clinicaland experimental arthritis,” Proceedings of the National Acade-my of Sciences of the United States of America, vol. 108, no. 27,pp. 11193–11198, 2011.

[44] R. M. O’Connell, D. Kahn, W. S. J. Gibson et al., “MicroRNA-155 promotes autoimmune inflammation by enhancing in-flammatory T cell development,” Immunity, vol. 33, no. 4, pp.607–619, 2010.

[45] E. Tili, J.-J. Michaille, D. Wernicke et al., “Mutator activityinduced by microRNA-155 (miR-155) links inflammation andcancer,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 108, no. 12, pp. 4908–4913, 2011.

[46] Q. S. Xia, Y. Ishigaki, X. Zhao et al., “Human SMG-1 isinvolved in gemcitabine-induced primary microRNA-155/BIC up-regulation in human pancreatic cancer PANC-1 cells,”Pancreas, vol. 40, no. 1, pp. 55–60, 2011.

[47] M. J. Scott, C. Chen, Q. Sun, and T. R. Billiar, “Hepatocytesexpress functional NOD1 and NOD2 receptors: a role forNOD1 in hepatocyte CC and CXC chemokine production,”Journal of Hepatology, vol. 53, no. 4, pp. 693–701, 2010.

[48] I. P. Pogribny, A. Starlard-Davenport, V. P. Tryndyak et al.,“Difference in expression of hepatic microRNAs miR-29c,miR-34a, miR-155, and miR-200b is associated with strain-specific susceptibility to dietary nonalcoholic steatohepatitis inmice,” Laboratory Investigation, vol. 90, no. 10, pp. 1437–1446,2010.

[49] J. Worm, J. Stenvang, A. Petri et al., “Silencing of microRNA-155 in mice during acute inflammatory response leads toderepression of c/ebp Beta and down-regulation of G-CSF,”Nucleic Acids Research, vol. 37, no. 17, pp. 5784–5792, 2009.

[50] M. Tomizawa, H. Horie, H. Yamamoto et al., “Reciprocal ex-pression of CCAAT/enhancer binding proteins alpha and betain hepatoblastomas and its prognostic significance,” Oncologyreports, vol. 17, no. 2, pp. 341–344, 2007.

[51] J. Oh, M. W. Hur, and C. E. Lee, “SOCS1 protects protein tyro-sine phosphatases by thioredoxin upregulation and attenuatesJaks to suppress ROS-mediated apoptosis,” Oncogene, vol. 28,no. 35, pp. 3145–3156, 2009.

[52] M. Taniguchi, M. Hashimoto, N. Hori, and K. Sato, “CCAAT/enhancer binding protein-β (C/EBP-β), a pivotal regulator ofthe TATA-less promoter in the rat catalase gene,” FEBS Letters,vol. 579, no. 25, pp. 5785–5790, 2005.

[53] R. Ikeda, T. Nishida, F. Watanabe et al., “Involvement ofCCAAT/enhancer binding protein-β (C/EBPβ) in epigeneticregulation of mouse methionine adenosyltransferase 1A geneexpression,” International Journal of Biochemistry and CellBiology, vol. 40, no. 9, pp. 1956–1969, 2008.

[54] C. Su, Z. Hou, C. Zhang, Z. Tian, and J. Zhang, “Ectopicexpression of microRNA-155 enhances innate antiviral immu-nity against HBV infection in human hepatoma cells,” VirologyJournal, vol. 8, article 354, 2011.

[55] G. Alpini, S. S. Glaser, J. P. Zhang, H. Francis, Y. Han, and J.Gong, “Regulation of placenta growth factor by microRNA-125b in hepatocellular cancer,” Journal of Hepatology, vol. 55,no. 6, pp. 1339–1345, 2011.


Research Article

Betaine Treatment Attenuates ChronicEthanol-Induced Hepatic Steatosis and Alterations tothe Mitochondrial Respiratory Chain Proteome

Kusum K. Kharbanda,1, 2 Sandra L. Todero,1 Adrienne L. King,3 Natalia A. Osna,2

Benita L. McVicker,2 Dean J. Tuma,2 James L. Wisecarver,4 and Shannon M. Bailey3

1 Research Service-151, Veterans Affairs Nebraska-Western Iowa Health Care System, 4101 Woolworth Avenue, Omaha,NE 68105, USA

2 Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE 68198, USA3 Department of Environmental Health Sciences, University of Alabama at Birmingham, Birmingham, AL 35294, USA4 Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, NE 68198, USA

Correspondence should be addressed to Kusum K. Kharbanda, [email protected]

Received 29 July 2011; Accepted 31 August 2011


Copyright © 2012 Kusum K. Kharbanda et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Introduction. Mitochondrial damage and disruption in oxidative phosphorylation contributes to the pathogenesis of alcoholic liverinjury. Herein, we tested the hypothesis that the hepatoprotective actions of betaine against alcoholic liver injury occur at the levelof the mitochondrial proteome. Methods. Male Wister rats were pair-fed control or ethanol-containing liquid diets supplementedwith or without betaine (10 mg/mL) for 4-5 wks. Liver was examined for triglyceride accumulation, levels of methionine cyclemetabolites, and alterations in mitochondrial proteins. Results. Chronic ethanol ingestion resulted in triglyceride accumulationwhich was attenuated in the ethanol plus betaine group. Blue native gel electrophoresis (BN-PAGE) revealed significant decreases inthe content of the intact oxidative phosphorylation complexes in mitochondria from ethanol-fed animals. The alcohol-dependentloss in many of the low molecular weight oxidative phosphorylation proteins was prevented by betaine supplementation. Thisprotection by betaine was associated with normalization of SAM : S-adenosylhomocysteine (SAH) ratios and the attenuation ofthe ethanol-induced increase in inducible nitric oxide synthase and nitric oxide generation in the liver. Discussion/Conclusion. Insummary, betaine attenuates alcoholic steatosis and alterations to the oxidative phosphorylation system. Therefore, preservationof mitochondrial function may be another key molecular mechanism responsible for betaine hepatoprotection.

1. Introduction

Chronic ethanol exposure has been shown to significantlyalter liver mitochondrial structural and functional integrity.Ethanol consumption alters mitochondrial morphology, in-duces mitochondrial DNA damage, and impairs ribosomalactivity and structure [1–4] resulting in depressed mitochon-drial protein synthesis and associated loss of electron trans-port chain complexes levels and function. It has also beenshown that alcohol exposure increases the sensitivity of livermitochondria to induce mitochondrial permeability transi-tion pore [5] that may be linked to higher cyclophilin D levelsin liver mitochondria [6]. Together, these chronic ethanol-

induced alterations result in depressed respiratory capacityand impaired oxidative phosphorylation, events critical tothe development of alcoholic liver injury [7–10].

In recent years, advancements in proteomic technologieshave facilitated the examination of alcohol-dependent altera-tions to the mitochondrial proteome [11]. Using both con-ventional and blue native (BN)-PAGE proteomics methods,Bailey et al. have reported that ethanol exposure results inthe decrease of both nuclear and mitochondrial encodedgene products of the oxidative phosphorylation system [11].Similar defects in the mitochondrial proteome such as reduc-tions in cytochrome c oxidase subunits and mitochondrialmembrane potential, have also been reported in genetically


altered mice exhibiting deficiency in liver levels of SAM [12],buttressing the concept that SAM plays a critical role inmaintaining proper mitochondrial function.

Several studies including ours have demonstrated thatwhile the alcohol-induced decrease in hepatic SAM levels isdetrimental, it is the decreased hepatocellular SAM : S-ade-nosylhomocysteine (SAH) ratio that adversely affects manycrucial SAM-dependent methylation reactions and the ulti-mate generation of many hallmark features of alcoholic liverdisease [13–16]. We have further shown that the addition ofSAM can normalize alcohol-induced SAM : SAH ratios [17]and preserve mitochondrial respiratory capacity by main-taining the mitochondrial genome and proteome while at-tenuating alcohol-dependent increases in mitochondrial su-peroxide production [3, 18].

Betaine, a methyl donor and another key metabolite ofthe methionine cycle, has been shown to normalize hepato-cellular SAM : SAH ratio, correct defective cellular methyla-tion reactions, and prevent the alcohol-mediated steatosis,apoptosis, and accumulation of damaged proteins [14, 17,19–23]. Based on this, we investigated whether betaine pre-vents alcohol-induced changes to the mitochondrial oxida-tive phosphorylation system in a rat model of chronic alcoholexposure. This assessment was complemented by determina-tions of liver cytochrome P450 2E1 (CYP2E1) protein andactivity, glutathione (GSH) levels, SAM : SAH ratios, NOS2expression, NO generation, and triglyceride levels.


2.1. Diet Formulation. Nutritionally adequate Lieber-DeCarlicontrol and ethanol liquid diets [24] were purchased fromDyets, Inc. (Bethlehem, Pa, USA). The ethanol diet consistedof 18% of total energy as protein, 35% as fat, 11% as carbo-hydrate, and 36% as ethanol. In the control diet, ethanolwas replaced isocalorically with carbohydrate such that bothethanol and control rats consumed identical amounts of allnutrients except carbohydrate.

2.2. Ethanol and Betaine Feeding Procedure. Male Wistarrats (Charles River Laboratories, Wilmington, Mass, USA)weighing 180 to 200 g (approximately 45–48 days old) wereweight-matched and divided into four groups. Group 1 wasfed the control diet. Group 2 was fed the same diet as Group1 except 1% (w/v) betaine was added to the diet. Group 3was fed the ethanol diet, and Group 4 was fed the ethanoldiet containing 1% (w/v) betaine. Rats in groups 1–3 werefed the amount of diet consumed by rats in group 4. Overall,each group consisted of 8 rats fed the appropriate diet for 4-5 weeks. Twenty four hours before sacrifice, the total dailyvolume of the diet was divided with 1/4 given at 8:00 am, 1/4at 12:00 noon, and 1/2 at 4:00 pm. In addition, animals weregiven 1/4 their respective diets 60–90 minutes prior to death.This regimen was followed to minimize differences in feedingpatterns that exists between the groups of rats. The care,use, and procedures performed on these rats were approvedby the Institutional Animal Care and Use Committee at theOmaha Veterans Affairs Medical Center and complied withNIH guidelines.

2.3. Liver Histology and Detection of Lipid Accumulation.Formalin fixed liver tissue was processed for hematoxylin-eosin staining and evaluated for steatosis and inflammation.In addition, fresh frozen liver sections were fixed in 4% w/vparaformaldehyde in 50 mM PIPES, pH 7.0, and the ac-cumulated lipid were visualized by staining with 1 μg/mLBODIPY 493/503 (Invitrogen, Carlsbad, Calif, USA). Afterincubation, slides were washed twice with PBS and mountedwith “Vectashield with DAPI” (Vector laboratories, Burm-ingham, Calif, USA). Images were obtained with a Zeiss510 Meta Confocal Laser Scanning Microscope using an ex-citation wavelength of 488 nm and an emission wavelengthof 505 nm.

2.4. Triglycerides. Total lipids were extracted from the liverto quantify the triglyceride mass using the triglyceride diag-nostics kit (Thermo DMA kit, Thermo Electron ClinicalChemistry, Louisville, Colo, USA) as detailed in our publi-cation [14].

2.5. Mitochondria Isolation. Pieces of fresh liver were homog-enized in cold 5 mmol/L Tris (pH 7.4) containing 0.25 mol/Lsucrose and 1 mmol/L EDTA, and liver mitochondria wereisolated by differential centrifugation techniques [3].

2.6. Blue Native Gel Electrophoresis. Ethanol and/or betaineeffects on the levels of mitochondrial proteins that comprisethe oxidative phosphorylation system were assessed usingBN-PAGE proteomics as detailed [3]. Image analysis on two-dimensional BN-PAGE gels was performed using QuantityOne software (Bio-Rad Laboratories, Hercules, Calif, USA).

2.7. SAM, SAH, and GSH Levels. The perchloric acid extractof total liver was filtered through a 0.22 μm membrane filterand directly subjected to HPLC analysis for the determina-tion of SAM, SAH, and GSH levels, as detailed [14]. GSHlevels in the mitochondrial fractions were also determined.

2.8. CYP2E1 Activity. The activity was measured in liverhomogenates by the formation of 4-nitrocatechol (4-NC),as previously described [25]. CYP2E1 specific activity is ex-pressed as nmol 4-NC produced per hr per mg protein.

2.9. NOS2 Gene Expression. NOS2 gene expression wasdetermined by real-time quantitative PCR. Total hepaticRNA was extracted, treated with DNase I (Invitrogen, Carls-bad, Calif, USA) and used to synthesize cDNA using TaqmanReverse Transcription reaction kits (Applied Biosystems,Foster City, Calif, USA). After amplification, real-time quan-titative PCR was performed using an Applied Biosystems7500 Real-Time PCR System and Taqman Assay-on-Demandgene expression assays for rat NOS2 and β-actin (housekeep-ing gene) according to the manufacturer’s instructions. Thecomparative Ct method was used to determine the relativeconcentration of the RNA transcript and the result expressedas “fold change” relative to the housekeeping gene.

2.10. Western Blotting. Immunoblots were performed byloading equal amounts of homogenate or mitochondrial


proteins onto SDS-PAGE gels. Isolated liver mitochondriawere used to detect mitochondria proteins, while total liverhomogenates were used for examining inducible nitric oxidesynthase (NOS2) and CYP2E1. Levels of NOS2 protein weredetected using a 1 : 1,000 dilution of antibody (BD/Pharm-ingen, San Diego, Calif, USA). CYP2E1 protein was detectedusing a 1 : 2,000 dilution (Calbiochem, Gibbstown, NJ, USA).Cytochrome c oxidase subunit I was detected using 1 : 5,000dilution (Molecular Probes, Eugene, Ore, USA). After mem-branes had been incubated with the appropriate secondaryantibodies, proteins were visualized using standard enhancedchemiluminescence detection methods. The intensities ofimmunoreactive protein bands were quantified using Quan-tity One software (Bio-Rad Laboratories, Hercules, Calif,USA).

2.11. Mitochondrial NO. Levels of nitrates and nitrites (theend product of nitric oxide, NO) were measured in the mito-chondrial fraction using the Griess reaction as detailed [26].

2.12. Statistical Analysis. Data were analyzed by ANOVA,followed by Student’s Newman-Keuls post hoc test. A P value<0.05 was regarded as statistically significant.

3. Results

3.1. Liver Histopathology, Triglycerides, and SAM : SAH Ratios.The histopathological evaluation of livers within each groupwere consistent with our previously published data [14]. Liv-ers from the rats fed ethanol for 4-5 weeks displayed micro-and macrovesicular steatosis; however, no steatosis was ob-served in livers of rats fed the betaine-supplemented ethanoldiet (Figure 1). Indeed, these livers showing similar histologyas the livers of control or the betaine-supplemented controlrats (Figure 1). Visualization of lipid droplets using greenfluorescent BODIPY 493/503 showed considerable accumu-lation of neutral lipids, including esterified cholesterol inethanol-fed rat livers (Figure 1). Minimal lipid accumulationwas observed in the controls, or betaine-supplemented eth-anol fed-rat livers (Figure 1). Biochemical analysis of the livertriglycerides levels corroborated the histopathology and neu-tral fat staining results. A significant attenuation of hepatictriglyceride content was observed in rats fed the betaine-supplemented ethanol diet as compared to the rats fed eth-anol alone (Figure 2).

Ethanol consumption for 4 weeks had no effect on hep-atic SAM levels, but dramatically increased SAH levels [13,14], resulting in a lower SAM : SAH ratio as compared withcontrols (Figure 3). Feeding rats a betaine-supplementedcontrol or ethanol diet increased hepatic SAM levels 3- and 6-fold, respectively (data not shown). SAH levels followed thesame pattern as SAM in these two groups (data not shown).These relative changes in the levels of SAM and SAH inboth betaine-supplemented groups resulted in comparablehepatic SAM : SAH ratios as in the controls. These resultswere similar to our previous observations [14].

3.2. Liver NOS2 and NO Generation. Chronic ethanol con-sumption caused induction of NOS2 both at the gene ex-

pression and the protein level (Figures 4(a) and 4(b), resp.),which is in agreement with other reports [27]. As a conse-quence of NOS2 induction, increased levels of nitrite/nitrate,byproducts of nitric oxide metabolism, were detected inmitochondrial fractions of ethanol-fed rats (Figure 4(c)).Interestingly, the ethanol-dependent increase in hepaticNOS2 and nitrite/nitrate levels were attenuated by supple-mentation of the ethanol diet with betaine (Figures 4(a)–4(c)).

3.3. Liver CYP2E1 Activity and Protein Levels. Chronic eth-anol consumption induced a 10-fold increase in CYP2E1 ac-tivity (Figure 5(a)). This increase in activity was also reflectedby an induction in CYP2E1 protein level (Figure 5(b)). Amarginal, but statistically significant, decrease (∼12%) inboth CYP2E1 protein and activity was observed in livers ofrats fed the betaine-supplemented ethanol diet (Figures 5(a)and 5(b)).

3.4. Mitochondrial GSH Levels. While chronic ethanol con-sumption decreased total liver GSH in the current study(Figure 6(a)), mitochondrial GSH was increased followingchronic ethanol consumption (Figure 6(b)). Betaine treat-ment had no effect on the ethanol-induced decrease in totalliver GSH or increased mitochondrial GSH levels observed(Figures 6(a) and 6(b)).

3.5. Assessment of Oxidative Phosphorylation Proteins. Exam-ination of oxidative phosphorylation proteins by BN-PAGErevealed that betaine prevented an ethanol-induced loss ofoxidative phosphorylation proteins. Representative one- andtwo-dimension BN-PAGE proteomic maps are shown inFigures 7(a) and 7(c). Similar to our previously publisheddata [3], the two-dimensional proteomic maps revealed aloss of respiratory chain proteins following chronic ethanolexposure. This was more apparent for proteins that comprisecytochrome c oxidase (i.e., complex IV) and the NADH de-hydrogenase (i.e., complex I) (Figure 7(b)). Supplementa-tion of the ethanol diet with betaine significantly attenuatedthe ethanol-induced loss in these mitochondrial proteins.The effects of ethanol and betaine were also reflected in West-ern blot analysis of complex IV. Representative data for com-plex IV is shown in Figure 7(d). For example, the ethanol-de-pendent loss in complex IV subunit I was attenuated by theinclusion of betaine in the ethanol diet.

4. Discussion

The role of mitochondrial dysfunction in the pathogenesis ofalcoholic liver disease has long been documented by multiplelaboratories [2–4, 7–10]. Mitochondria are a recognizedsource of reactive oxygen and reactive nitrogen species fol-lowing ethanol consumption and are also a key target of sub-sequent oxidative posttranslational modifications includingcomponents of the oxidative phosphorylation system [10].Recent studies using BN-PAGE proteomics, show that chron-ic ethanol consumption decreases the levels of several nuclearand mitochondrial encoded proteins that comprise the indi-vidual oxidative phosphorylation complexes [3]. We have


Figure 1: Betaine attenuates chronic ethanol-dependent steatosis. Hematoxylin-eosin (A) and BODIPY 493/503-DAPI (B, green BODIPY-labeled lipid droplets, nuclei blue) stained images from representative livers of (a) control; (b) control + betaine; (c) ethanol; (d) ethanol+ betaine rats (n = 8 rats’/group) fed the respective diets for 4-5 weeks. Liver sections of the representative ethanol-fed rat stained withhematoxylin-eosin or BODIPY 493/503-DAPI shows micro- and macrovesicular steatosis (arrows).

Figure 2: Betaine attenuates chronic ethanol-mediated increase inhepatic triglyceride levels. Triglyceride content in the liver lipidextract was quantified using the diagnostics kit (Thermo ElectronClinical Chemistry, Louisville, Colo, USA). Data represent the mean± S.E.M. for n = 8 animals per treatment group. Values not sharinga common letter are statistically different, P < 0.05.

Figure 3: Betaine prevents the chronic ethanol-dependent decreasein the hepatic SAM : SAH ratio. Liver SAM and SAH levels weredetermined by HPLC analysis, and the SAM : SAH ratio wascalculated as previously described [14]. Data represent the mean ±S.E.M. for n = 8 animals per group. Values not sharing a commonletter are statistically different, P < 0.05.


(a) (b)

(c)

Figure 4: Betaine attenuates chronic ethanol-induced increase hepatic NOS2 gene and protein expression and elevates mitochondrialnitrite/nitrate levels. (a) Liver NOS2 mRNA levels were measured by quantitative PCR using the comparative Ct concentration method.The data shown are mean ± SEM of eight determinations from each group. Values not sharing a common subscript letter are statisticallydifferent, P < 0.05. (b) Hepatic NOS2 (130 kDa) protein expression was measured by immunoblotting and normalized to β-actin (45 kDa).Representative immunoblots of NOS2 protein for one pair of untreated and one pair of betaine-treated control and ethanol animals is shown.The bar graph results below represent the mean volume integration units (V.I.U) of NOS2 ± S.E.M. for n = 8 animals per group. Values notsharing a common letter are statistically different, P < 0.05. (c) Total mitochondrial nitrite and nitrate levels were measured by the Griessreaction as detailed [26]. The bar graph results represent the mean ± S.E.M. for n = 8 animals per group. Values not sharing a commonletter are statistically different, P < 0.05.


(a) (b)

Figure 5: Betaine has minimal effect on chronic ethanol-dependent increase in CYP2E1 activity and protein. (a) Hepatic CYP2E1 activitywas determined as detailed in Section 2. Data represent the mean ± S.E.M. for n = 8 animal per groups. Values not sharing a commonletter are statistically different, P < 0.05. (b) CYP2E1 protein levels were measured by immunoblotting. The bar graph results represent themean volume integration units (V.I.U) ± S.E.M. for n = 8 animals per group. Values not sharing a common letter are statistically different,P < 0.05. The top figure shows representative immunoblots of CYP2E1 protein for one pair of untreated and one pair of betaine-treatedcontrol and ethanol animals.

further shown that SAM treatment prevented these losses,particularly, select subunits of Complexes I and IV, whichmay explain, in part, how SAM maintains mitochondrialfunction and protects against the development of alcoholicliver injury [3, 18].

In this current study, we show for the first time thatdietary supplementation with the methionine cycle metabo-lite, betaine, also protects against an ethanol-induced loss inoxidative phosphorylation system proteins. While the exactmechanism for this protection at the organelle level is notknown, we propose that by preventing NOS2 inductionand NO generation, betaine preserves the functioning ofthe electron transport chain, maintains the integrity of theliver, and protects against the development of alcoholic liverinjury. Moreover, we propose that changes are associatedwith the normalization of hepatic SAM : SAH ratio andmaintenance of methylation potential in response to betainesupplementation during chronic ethanol ingestion.

It is notable that NOS2 is absent in healthy, normal liveruntil it is transcriptionally activated by proinflammatorystimuli to produce large amounts of nitric oxide [27, 28].Arguably, the chronic ethanol-mediated decrease in theSAM : SAH ratio in liver could be responsible for the ethanol-dependent increase in NOS2 expression, since studies haveshown that impaired methylation increases NOS2 geneexpression and nitric oxide [29]. Thus, lower hepatocellularSAM : SAH ratios would favor hypomethylation of the NOS2promoter leading to increased NOS2 gene transcription andsubsequent NOS2 protein increase and NO generation as

seen in the current study (Figure 4). Conversely, it has alsobeen shown that hypermethylation of the NOS2 gene pro-moter silences and downregulates NOS2 expression in foamcells [30]. In agreement with these results, we observed thatconcurrent supplementation with betaine in the ethanol dietsuppressed ethanol-induced NOS2 gene and protein expres-sion and NO generation. This finding suggests that betainethrough maintenance of the SAM : SAH ratio and the meth-ylation potential in liver could prevent the upregulation ofNOS2 at the level of transcription.

Studies by Wu and Cederbaum have proposed thatCYP2E1-mediated oxidative stress causes oxidative injury tothe mitochondrion [31]. It has further been suggested thatthe alcohol-dependent increase in CYP2E1 may be partial-ly related to decreased proteolysis of this protein in responseto impaired proteasome function in the alcoholic liver [32].Recently, we reported that the proteasome activity is direct-ly impaired at ratios of SAM : SAH that correspond to thoseobserved in livers of ethanol-fed rats [33]. Indeed, theseresults corroborate previous observations of a negative cor-relation between alcohol-induced increases in liver mRNAand protein levels of CYP2E1 with SAM : SAH ratios [34].It should also be noted that SAM interacts and inhibits thecatalytic activity of CYP2E1 in a reversible and noncom-petitive manner in vitro [35]. However, despite elevatedSAM levels and normal SAM : SAH ratios, CYP2E1 activ-ity was only modestly inhibited (∼12%) in rats fed thebetaine-supplemented ethanol diets (Figures 5(a) and 5(b)).


(a) (b)

Figure 6: Betaine and ethanol increase mitochondrial GSH levels. HPLC was employed to determine GSH levels in (a) liver and the (b)mitochondrial fractions from control, ethanol, control + betaine, or ethanol + betaine groups. Data represent the mean ± S.E.M. for n = 8animals per group. Values not sharing a common letter are statistically different, P < 0.05.

Regardless of elevated CYP2E1 in the livers of betaine-sup-plemented ethanol-fed rat, indices of steatosis nor defects inthe mitochondrial respiratory chain was observed in theseanimals [14, 17, 19–23]. Taken together, these results suggestthat the increased CYP2E1 activity is not sufficient to causeand the onset of liver steatosis or mitochondrial proteomedefects.

Early studies demonstrated that chronic ethanol feedingspecifically caused marginal decrease in total and a markeddecrease in mitochondrial GSH levels compared to controls,which was associated with mitochondrial lipid peroxidationand progression of liver damage [36, 37]. Subsequently,studies showed that feeding SAM attenuated both ethanol-induced depletion of mitochondrial GSH and mitochondrialdysfunction [38]. Contrary to these studies, an alcohol-mediated increase in mitochondrial GSH has also beenreported [3, 39]. Betaine treatment has been shown to restorealcohol-induced hepatic depletion of total GSH levels [40–43]. None of these studies, however, examined mitochondrialGSH status in particular following betaine treatment. In ourstudy, we observe a significant decrease in total liver GSH(Figure 6(a)) but an (albeit small) increase in mitochondrialGSH in response to alcohol (Figure 6(b)) corroborating pre-vious reports [3, 39]. We further observed that betaine treat-ment was unable to correct alcohol-induced changes in liveror mitochondrial GSH levels (Figures 6(a) and 6(b)). Ourresults suggest that neither the ethanol-induced defects northe protective role of betaine on the various parameters ex-amined in this study appears to be mechanistically related tothe status of total or mitochondrial GSH level per se.

The present data reiterates that the normal mitochon-drial proteome and function relies on the maintenance of

methylation reactions. Indeed, supplementation with methyldonors, SAM [3, 4], and betaine (present study) at con-centrations that maintain methylation reactions preservesmitochondrial proteome [3, 4] as well as prevents ethanol-dependent defects in mitochondrial respiration, mitochon-drial ribosome dissociation, increases in mitochondrialsuperoxide production, and mitochondrial DNA damage[3, 4]. While the protective action of SAM and betaine at thelevel of the mitochondrion is now recognized, the mecha-nisms responsible for this protection are not clear. Similarly,how an alteration in the SAM : SAH directly or indirectlyinfluences the composition of the mitochondrial proteomeis undefined. Recent studies by Bailey and colleagues, using acomprehensive proteomics approach, demonstrate multipleethanol and SAM specific alterations in key proteins involvedin the oxidative phosphorylation system, as well as methio-nine, choline, and sulfur metabolism and chaperone systems[18]. Particularly, the SAM-dependent impacts on methio-nine and choline metabolism enzymes may be especiallyimportant, as these effects are predicted to protect methyl-transferase reactions through boosting SAM levels. Simi-larly, the analyses on the oxidative phosphorylation systemprovided further support for an association among SAM,methylation, and respiratory chain maintenance as multiplesubunits of the respiratory complexes were preserved in theSAM-supplemented ethanol group [18]. These findings arefurther validated in the current study, as we observed theability of betaine to attenuate the alcohol-dependent lossin complex I and IV subunits. Methylation reactions havebeen reported to play an important role in the biosynthesisof lipoic acid, ubiquinone and biotin [44]; key cofactors ofmultiple cellular and mitochondrial enzyme systems. Taken


(a) (b)

(c)

(d)

Figure 7: Betaine prevents the chronic ethanol-mediated loss in oxidative phosphorylation proteins. Representative (a) one- and (c) two-dimension blue native (BN)-PAGE proteomic gels of mitochondria isolated from rats fed control, ethanol, control + betaine, or ethanol +betaine diets as detailed in Section 2. (a) For these gels, 250 μg of mitochondrial protein were subjected to 1D BN-PAGE. (c) The 1D gel stripswere overlaid across a Tris-Tricine-SDS-PAGE gel to resolve the individual polypeptides that comprise each oxidative phosphorylation systemcomplex. The proteins that comprise each complex (I, V, III, IV, and II) appear as vertically aligned spots on the 2D gel. (b) Comparisonof the relative quantities of complexes I, V, III, and IV in liver mitochondria from the control; control + betaine; ethanol, and ethanol +betaine. Statistical analyses for data presented in panel B: 2-factor ANOVA on raw densitometry values—complex I: ethanol P = 0.0002,betaine P = 0.92, interaction P = 0.02; complex V: ethanol P = 0.016, betaine P = 0.72, interaction P = 0.23; complex III: ethanolP = 0.005, betaine P = 0.92, interaction P = 0.31; complex IV: ethanol P = 0.15, betaine P = 0.62, interaction P = 0.045; complex II:ethanol P = 0.54, betaine P = 0.98, interaction P = 0.80. (d) Complex IV, subunit 1 protein levels were measured by immunoblotting. Thebar graph results represent the mean volume integration units (V.I.U) ± S.E.M. for n = 8 animals per group. Values not sharing a commonletter are statistically different, P < 0.05. The top figure shows representative immunoblots of subunit 1 protein for control, ethanol, andbetaine supplemented control and ethanol-fed rats.


together, future experiments will be directed at determiningwhether alcohol-dependent disruption in the activity of spe-cific mitochondrial SAM-dependent methyltransferase(s) asa consequence of low SAM : SAH ratio is directly responsiblefor alterations in mitochondrion proteome and function.

In conclusion, this study shows for the first time thatbetaine prevents or blunts chronic ethanol-mediated alter-ations to the oxidative phosphorylation system proteome.We propose that the mitochondrial protection affordedby betaine is associated with the maintenance of hepaticSAM : SAH ratios and by blocking the ethanol-induced in-duction of NOS2 and NO, a key source of protein mod-ification and damage. Moreover, this study demonstratesthat proteomic-based methods like BN-PAGE are powerfultools to aid in the identification in the molecular targets ofdisease. Thus, maintenance of mitochondria function may beanother key molecular target underlying the hepatoprotec-tive effect of betaine against ethanol hepatotoxicity and fattyliver disease.

Abbreviations

BN-PAGE: Blue native gel electrophoresisCYP2E1: Cytochrome P450 2E1GSH: GlutathioneNOS2: Inducible nitric oxide synthaseSAM: S-adenosylmethionineSAH: S-adenosylhomocysteine.

Conflict of Interests

The authors of this study confirm that they have no com-mercial associations that pose a conflict of interests in con-nection with this paper.

Acknowledgments

The authors gratefully acknowledge the financial supportof the following: Department of Veterans Affairs NationalMerit Review grant (K. K. Kharbanda) and National Instituteof Health grants, R21AA017296 (K. K. Kharbanda) andAA15172 and AA18841 (S. M. Bailey). Ms. King is supportedby a NIH Research Supplement to Promote Diversity inHealth-Related Research linked to parent grant AA15172.

References

[1] T. F. Taraschi, W. S. Thayer, J. S. Ellingson, and E. Rubin,“Effects of ethanol on the structure and function of rat livermitochondrial and microsomal membranes,” Annals of theNew York Academy of Sciences, vol. 488, pp. 127–139, 1986.

[2] A. Cahill, X. Wang, and J. B. Hoek, “Increased oxidative dam-age to mitochondrial DNA following chronic ethanol con-sumption,” Biochemical and Biophysical Research Communica-tions, vol. 235, no. 2, pp. 286–290, 1997.

[3] S. M. Bailey, G. Robinson, A. Pinner et al., “S-adenosyl-methionine prevents chronic alcohol-induced mitochondrialdysfunction in the rat liver,” American Journal of Physiology,vol. 291, no. 5, pp. G857–G867, 2006.

[4] P. Sykora, K. K. Kharbanda, S. E. Crumm, and A. Cahill,“S-adenosyl-L-methionine co-administration prevents the

ethanol-elicited dissociation of hepatic mitochondrial ribo-somes in male rats,” Alcoholism: Clinical and ExperimentalResearch, vol. 33, no. 1, pp. 1–9, 2009.

[5] M. Yan, P. Zhu, H. M. Liu, H. T. Zhang, and L. Liu, “Ethanolinduced mitochondria injury and permeability transition poreopening: role of mitochondria in alcoholic liver disease,”World Journal of Gastroenterology, vol. 13, no. 16, pp. 2352–2356, 2007.

[6] A. L. King, T. M. Swain, D. A. Dickinson, M. J. Lesort, and S.M. Bailey, “Chronic ethanol consumption enhances sensitivityto Ca2+-mediated opening of the mitochondrial permeabilitytransition pore and increases cyclophilin D in liver,” AmericanJournal of Physiology, vol. 299, no. 4, pp. G954–G966, 2010.

[7] C. C. Cunningham and S. M. Bailey, “Ethanol consumptionand liver mitochondria function,” Biological Signals and Recep-tors, vol. 10, no. 3-4, pp. 271–282, 2001.

[8] S. M. Bailey and C. C. Cunningham, “Contribution of mito-chondria to oxidative stress associated with alcoholic liver dis-ease,” Free Radical Biology and Medicine, vol. 32, no. 1, pp. 11–16, 2002.

[9] J. B. Hoek, A. Cahill, and J. G. Pastorino, “Alcohol and mito-chondria: a dysfunctional relationship,” Gastroenterology, vol.122, no. 7, pp. 2049–2063, 2002.

[10] S. M. Bailey, “A review of the role of reactive oxygen and nitro-gen species in alcohol-induced mitochondrial dysfunction,”Free Radical Research, vol. 37, no. 6, pp. 585–596, 2003.

[11] S. M. Bailey, A. Landar, and V. Darley-Usmar, “Mitochondrialproteomics in free radical research,” Free Radical Biology andMedicine, vol. 38, no. 2, pp. 175–188, 2005.

[12] M. L. Martınez-Chantar, F. J. Corrales, L. A. Martınez-Cruzet al., “Spontaneous oxidative stress and liver tumors inmice lacking methionine adenosyltransferase 1A,” The FASEBJournal, vol. 16, no. 10, pp. 1292–1294, 2002.

[13] A. J. Barak, H. C. Beckenhauer, M. E. Mailliard, K. K.Kharbanda, and D. J. Tuma, “Betaine lowers elevated S-ade-nosylhomocysteine levels in hepatocytes from ethanol-fedrats,” Journal of Nutrition, vol. 133, no. 9, pp. 2845–2848, 2003.

[14] K. K. Kharbanda, M. E. Mailliard, C. R. Baldwin, H. C.Beckenhauer, M. F. Sorrell, and D. J. Tuma, “Betaine attenuatesalcoholic steatosis by restoring phosphatidylcholine gener-ation via the phosphatidylethanolamine methyltransferasepathway,” Journal of Hepatology, vol. 46, no. 2, pp. 314–321,2007.

[15] K. K. Kharbanda, “Alcoholic liver disease and methionine me-tabolism,” Seminars in Liver Disease, vol. 29, no. 2, pp. 155–165, 2009.

[16] K. K. Kharbanda, “Role of transmethylation reactions in alco-holic liver disease,” World Journal of Gastroenterology, vol. 13,no. 37, pp. 4947–4954, 2007.

[17] K. K. Kharbanda, D. D. Rogers, M. E. Mailliard et al., “A com-parison of the effects of betaine and S-adenosylmethionineon ethanol-induced changes in methionine metabolism andsteatosis in rat hepatocytes,” Journal of Nutrition, vol. 135, no.3, pp. 519–524, 2005.

[18] K. K. Andringa, A. L. King, H. B. Eccleston et al., “Analysis ofthe liver mitochondrial proteome in response to ethanol andS-adenosylmethionine treatments: novel molecular targets ofdisease and hepatoprotection,” American Journal of Physiology,vol. 298, no. 5, pp. G732–G745, 2010.



[20] K. K. Kharbanda, D. D. Rogers II, H. C. Beckenhauer et al.,“Adenosine-induced apoptosis mediated via increased hepat-ocellular levels of S-adenosylhomocysteine is attenuated bybetaine administration,” Hepatology, vol. 40, p. 572A, 2004.

[21] K. K. Kharbanda and A. J. Barak, “Defects in methioninemetabolism: its role in ethanol-induced liver injury,” inComprehensive Handbook of Alcohol-Related Pathology, V. R.Preedy and R. R. Watson, Eds., vol. 2, pp. 735–747, AcademicPress, Elsevier Sciences, 2005.

[22] K. K. Kharbanda, D. D. Rogers, M. E. Mailliard et al., “Role ofelevated S-adenosylhomocysteine in rat hepatocyte apoptosis:protection by betaine,” Biochemical Pharmacology, vol. 70, no.12, pp. 1883–1890, 2005.

[23] K. K. Kharbanda, M. E. Mailliard, C. R. Baldwin, M. F. Sorrell,and D. J. Tuma, “Accumulation of proteins bearing atypicalisoaspartyl residues in livers of alcohol-fed rats is preventedby betaine administration: effects on protein-l-isoaspartylmethyltransferase activity,” Journal of Hepatology, vol. 46, no.6, pp. 1119–1125, 2007.

[24] C. S. Lieber and L. M. Decarli, “Liquid diet technique ofethanol administration: 1989 update,” Alcohol and Alcoholism,vol. 24, no. 3, pp. 197–211, 1989.

[25] N. A. Osna, D. L. Clemens, and T. M. Donohue, “Interferongamma enhances proteasome activity in recombinant Hep G2cells that express cytochrome P4502E1: modulation by eth-anol,” Biochemical Pharmacology, vol. 66, no. 5, pp. 697–710,2003.

[26] H. Moshage, B. Kok, J. R. Huizenga, and P. L. M. Jansen,“Nitrite and nitrate determinations in plasma: a critical eval-uation,” Clinical Chemistry, vol. 41, no. 6, pp. 892–896, 1995.

[27] A. Venkatraman, S. Shiva, A. Wigley et al., “The role ofiNOS in alcohol-dependent hepatotoxicity and mitochondrialdysfunction in mice,” Hepatology, vol. 40, no. 3, pp. 565–573,2004.

[28] S. E. McKim, E. Gabele, F. Isayama et al., “Inducible nitricoxide synthase is required in alcohol-induced liver injury:studies with knockout mice,” Gastroenterology, vol. 125, no. 6,pp. 1834–1844, 2003.

[29] Z. Yu and B. C. Kone, “Hypermethylation of the induciblenitric-oxide synthase gene promoter inhibits its transcription,”The Journal of Biological Chemistry, vol. 279, no. 45, pp.46954–46961, 2004.

[30] Y. Jiang, J. Zhang, J. Xiong, J. Cao, G. Li, and S. Wang, “Li-gands of peroxisome proliferator-activated receptor inhibithomocysteine induced DNA methylation of inducible nitricoxide synthase gene,” Acta Biochimica et Biophysica Sinica, vol.39, no. 5, pp. 366–376, 2007.

[31] D. Wu and A. I. Cederbaum, “Oxidative stress and alcoholicliver disease,” Seminars in Liver Disease, vol. 29, no. 2, pp. 141–154, 2009.

[32] B. J. Roberts, B. J. Song, Y. Soh, S. S. Park, and S. E. Shoaf,“Ethanol induces CYP2E1 by protein stabilization: role ofubiquitin conjugation in the rapid degradation of CYP2E1,”The Journal of Biological Chemistry, vol. 270, no. 50, pp.29632–29635, 1995.

[33] N. A. Osna, R. L. White, T. M. Donohue, M. R. Beard, D. J.Tuma, and K. K. Kharbanda, “Impaired methylation as a novelmechanism for proteasome suppression in liver cells,” Bio-chemical and Biophysical Research Communications, vol. 391,no. 2, pp. 1291–1296, 2010.

[34] F. Esfandiari, J. A. Villanueva, D. H. Wong, S. W. French, andC. H. Halsted, “Chronic ethanol feeding and folate deficiencyactivate hepatic endoplasmic reticulum stress pathway in mi-

cropigs,” American Journal of Physiology, vol. 289, no. 1, pp.G54–G63, 2005.

[35] A. A. Caro and A. I. Cederbaum, “Inhibition of CYP2E1 cat-alytic activity in vitro by S-adenosyl-L-methionine,” Biochem-ical Pharmacology, vol. 69, no. 7, pp. 1081–1093, 2005.

[36] T. Hirano, N. Kaplowitz, H. Tsukamoto, S. Kamimura, andJ. C. Fernandez-Checa, “Hepatic mitochondrial glutathionedepletion and progression of experimental alcoholic liver dis-ease in rats,” Hepatology, vol. 16, no. 6, pp. 1423–1427, 1992.

[37] J. C. Fernandez-Checa, M. Ookhtens, and N. Kaplowitz,“Effect of chronic ethanol feeding on rat hepatocytic glutathi-one. Compartmentation, efflux, and response to incubationwith ethanol,” Journal of Clinical Investigation, vol. 80, no. 1,pp. 57–62, 1987.

[38] C. Garcia-Ruiz, A. Morales, A. Colell et al., “Feeding S-ade-nosyl-L-methionine attenuates both ethanol-induced deple-tion of mitochondrial glutathione and mitochondrial dysfunc-tion in periportal and perivenous rat hepatocytes,” Hepatology,vol. 21, no. 1, pp. 207–214, 1995.

[39] S. M. Bailey, V. B. Patel, T. A. Young, K. Asayama, and C. C.Cunningham, “Chronic ethanol consumption alters the glu-tathione/glutathione peroxidase-1 system and protein oxida-tion status in rat liver,” Alcoholism: Clinical and ExperimentalResearch, vol. 25, no. 5, pp. 726–733, 2001.

[40] G. Kanbak, M. Inal, and C. Baycu, “Ethanol-induced hepato-toxicity and protective effect of betaine,” Cell Biochemistry andFunction, vol. 19, no. 4, pp. 281–285, 2001.

[41] S. H. Kenyon, C. J. Waterfield, D. S. Asker et al., “Effect ofhydrazine upon vitamin B12-dependent methionine synthaseactivity and the sulphur amino acid pathway in isolated rathepatocytes,” Biochemical Pharmacology, vol. 57, no. 11, pp.1311–1319, 1999.

[42] J. Balkan, S. Oztezcan, M. Kucuk, U. Cevikbas, N. Kocak-Toker, and M. Uysal, “The effect of betaine treatment ontriglyceride levels and oxidative stress in the liver of ethanol-treated guinea pigs,” Experimental and Toxicologic Pathology,vol. 55, no. 6, pp. 505–509, 2004.

[43] C. L. Powell, B. U. Bradford, C. P. Craig et al., “Mechanism forprevention of alcohol-induced liver injury by dietary methyldonors,” Toxicological Sciences, vol. 115, no. 1, pp. 131–139,2010.

[44] G. S. Shadel, “Coupling the mitochondrial transcription ma-chinery to human disease,” Trends in Genetics, vol. 20, no. 10,pp. 513–519, 2004.


Research Article

Dose-Dependent Change in Elimination Kinetics ofEthanol due to Shift of Dominant Metabolizing Enzyme fromADH 1 (Class I) to ADH 3 (Class III) in Mouse

Takeshi Haseba,1 Kouji Kameyama,2 Keiko Mashimo,1 and Youkichi Ohno1

1 Department of Legal Medicine, Nippon Medical School, 1-1-5 Sendagi, Bunkyo-ku, Tokyo 113-8602, Japan2 Department of Pathology, Nippon Medical School, 1-1-5 Sendagi, Bunkyo-ku, Tokyo 113-8602, Japan

Correspondence should be addressed to Takeshi Haseba, [email protected]

Received 27 May 2011; Accepted 23 August 2011


Copyright © 2012 Takeshi Haseba et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

ADH 1 and ADH 3 are major two ADH isozymes in the liver, which participate in systemic alcohol metabolism, mainly distributingin parenchymal and in sinusoidal endothelial cells of the liver, respectively. We investigated how these two ADHs contribute to theelimination kinetics of blood ethanol by administering ethanol to mice at various doses, and by measuring liver ADH activityand liver contents of both ADHs. The normalized AUC (AUC/dose) showed a concave increase with an increase in ethanol dose,inversely correlating with β. CLT (dose/AUC) linearly correlated with liver ADH activity and also with both the ADH-1 and -3contents (mg/kg B.W.). When ADH-1 activity was calculated by multiplying ADH-1 content by its Vmax/mg (4.0) and normalizedby the ratio of liver ADH activity of each ethanol dose to that of the control, the theoretical ADH-1 activity decreased dose-dependently, correlating with β. On the other hand, the theoretical ADH-3 activity, which was calculated by subtracting ADH-1activity from liver ADH activity and normalized, increased dose-dependently, correlating with the normalized AUC. These resultssuggested that the elimination kinetics of blood ethanol in mice was dose-dependently changed, accompanied by a shift of thedominant metabolizing enzyme from ADH 1 to ADH 3.

1. Introduction

Alcohol dehydrogenase (ADH; EC 1.1.1.1) in the liver isgenerally accepted to be the primary enzyme responsible forethanol metabolism. This is supported by evidence that thelevel of liver ADH activity is closely correlated with the rateof ethanol metabolism [1–3] and that the metabolism invivo is markedly depressed in animals treated with pyrazolesof ADH inhibitors [4, 5] and in ones genetically lackingADH [6]. The process by which blood ethanol is eliminatedwas traditionally assumed to follow zero-order [7] or singleMichaelis-Menten (M-M) kinetics [8, 9], even though mam-malian livers actually contain three kinds of ADH isozymes(Class I, II, III) with different Kms for ethanol [10, 11].Thus, it was commonly thought that the elimination processwas regulated by Class I ADH (ADH 1), which distributesmainly in parenchymal liver cells [12], because this classicallyknown ADH has the lowest Km among the three liver ADH

isozymes and because its activity saturates at millimolarlevels of ethanol. Indeed, mice genetically lacking ADH 1have been used to demonstrate that ADH 1 is a key enzymein systemic ethanol metabolism [13, 14]. However, studieson these ADH-1-deficient animals have also shown thatethanol metabolism in vivo cannot be explained solely byADH 1 [13, 14]. Although the microsomal ethanol oxidizingsystem (MEOS) including CYP2E1 as a main component,and catalase have been discussed for many years as candidatesfor non-ADH 1 pathways [15, 16], these studies have failed toclarify their roles in ethanol metabolism in mice geneticallylacking these enzymes [17–19]. Moreover, the process of theelimination of blood ethanol has been shown to involve first-order kinetics [20–23], suggesting that alcohol-metabolizingenzymes with a very high Km participate in systemic ethanolmetabolism. ADH 3 (Class III), another major ADH, whichdistributes mainly in sinusoidal endothelial cells of the liver[12], has very high Km for ethanol. Therefore, it shows very


little activity when assayed by the conventional method withmillimolar levels of ethanol as a substrate; but its activityincreases up to the molar level of ethanol [10, 24]. Addi-tionally, this ADH has been demonstrated to be markedlyactivated under hydrophobic conditions, which lower itsKm [14, 25]. Previously, liver ADH activity was assumed tobe attributable solely to ADH 1 because it was responsiblefor most of the activity due to its low Km [10, 24]. However,we have used ethanol-treated mice to show that liver ADHactivity assayed by the conventional method depends notonly on ADH 1 but also on ADH 3 and governs the elimi-nation rate of blood ethanol [3]. Moreover, we have recentlydemonstrated using Adh3-null mice that ADH 3 participatesin systemic ethanol metabolism dose-dependently [14].

These data suggest that systemic ethanol metabolism inmice involves both liver ADH 1 and ADH 3, possibly throughthe regulation of their contents and/or enzymatic kinetics.However, how these two ADH isozymes contribute to theelimination kinetics of ethanol is largely unknown.

In the present study, we investigated how these two liverADHs contribute to the elimination kinetics of ethanol inmice by statistically analyzing the pharmacokinetic parame-ters of blood ethanol and the enzymatic parameters of ADH,based on a two-ADH model that ascribes liver ADH activityto both ADH 1 and ADH 3.

2. Methods

2.1. Measurement of Pharmacokinetic Parameters of BloodEthanol. As previously described [3], male ddY mice (9weeks old) were injected with ethanol (i.p.) at a dose of 1,2, 3, 4.5, or 5 g/kg body weight, while the control mice wereinjected with saline (0 g/kg). For each dose, blood sampleswere taken from the tails of mice (n = 3) at scheduled times(0.5, 1, 2, 4, 8, and 12 h) after ethanol administration.

Blood ethanol concentration was measured with a head-space gas chromatograph [3]. The rate of ethanol elimina-tion from blood was expressed as a β-value (mmol/L/h),which was calculated from a regression line fitted to theblood ethanol concentrations at various times by the lin-ear least-squares method [26]. The area under the bloodconcentration-time curve (AUC) was calculated by trape-zoidal integration using the extrapolation of time coursecurves to obtain the normalized AUC (AUC/dose) and bodyclearance of ethanol (CLT : the reciprocal of the normalizedAUC) [23].

All animals received humane care in compliance with ourinstitutional guidelines “The Regulations on Animal Exper-imentation of Nippon Medical School,” which was based on“The Guidelines of the International Committee on Labora-tory Animals 1974”.

2.2. Measurement of Liver ADH Parameters. In order toobtain liver samples, mice were sacrificed by cervical dislo-cation at scheduled times during ethanol metabolism at eachdose (0.5, 1, and 2 h for 1 and 2 g/kg; 0.5, 1, 2, 4, and 8 hfor 3 g/kg; 0.5, 1, 2, 4, 8, and 12 h for 0, 4.5, and 5 g/kg)(n = 3 at each time in each dose). Each liver was homog-enized in 6 vol (w/v) of extraction buffer (0.5 mM NAD,

0.65 mM DTT/5 mM Tris-HCL, pH 8.5) and centrifuged at105, 000 ×g for 1 h to obtain a liver extract.

ADH activity was measured at pH 10.7 by the conven-tional assay with 15 mM ethanol as a substrate, using liverextract during the times of ethanol metabolism at each dose.The ADH 1 and ADH 3 contents of liver were measured byEIA using isozyme-specific antibodies on the same samplesas those used for ADH activity [3], excluding the samplesat doses of 2 and 4.5 g/kg. The ADH activity and content ofliver were expressed in terms of liver weight/kg body weightbecause these units are not influenced by hepatomegaly orvariations in the total liver weight with respect to bodyweight. These liver ADH parameters were averaged overthe ethanol-metabolizing time for each dose of ethanol andtermed the liver ADH activity, the liver ADH 1 content, andthe liver ADH 3 content.

The apparent Km and Vmax of ADH activity weredetermined from a Lineweaver-Burk plot with ethanol (0.1–100 mM) as a substrate, using liver extracts obtained at 1 and4 h after ethanol administration for all doses (n = 3 at eachtime in each dose). Vmax is expressed in units/mg of the sumof the ADH 1 and ADH 3 contents.

2.3. Two-ADH-Complex Model of Liver ADH Activity. Thetwo-ADH-complex model, which ascribes liver ADH activityto both ADH 1 and ADH 3, is described by the function[y (ADH activity) = f (ADH 1 activity, ADH 1 content,ADH 3 activity, ADH 3 content)] for each liver extract. TheVmax of ADH 1 in liver extract is assumed to be a constant4.0 units/mg, regardless of ethanol dose, because purifiedmouse ADH 1 usually exhibits a relatively constant Vmax ofaround 4.0 units/mg, a value that was obtained with around15 mM ethanol as a substrate at pH 10.7 [3]. In the complexmodel, therefore, ADH 1 activity was calculated from [ADH1 content × 4.0], while ADH 3 activity was assumed tobe [ADH activity − ADH 1 activity] in each liver. Theseassumptions are based on two facts: (1) ADH 2 (the thirdADH isozyme in liver) is only responsible for a very smallportion of total ADH activity in mice liver (<3%) [3], and (2)ADH 3 is activated depending on the conditions of medium[14, 25]. The calculated ADH 1 and ADH 3 activities werethen averaged over the ethanol-metabolizing time for eachdose of ethanol and normalized by the ratio of the averageliver ADH activity of each ethanol group to that of thecontrol. These normalized ADH activities were termed thetheoretical ADH 1 and ADH 3 activities. These parameterswere used for statistical analyses and correlation studies.

3. Results

3.1. Effect of Dose on Pharmacokinetics of Blood Ethanol.Figure 1 shows the time course of blood ethanol concentra-tion in mice after the administration of ethanol at variousdoses. Blood ethanol elimination roughly followed zero-order or M-M kinetics, reaching a constant Vmax at everydose of ethanol, as shown by the regression lines fitted tothe blood ethanol concentrations at various times (r2 =0.996, 0.996, 0.999, 1.000, and 0.945 for doses of 1, 2, 3,


20

40

60

80

100

120

Blo

odet

han

olco

nce

ntr

atio

n(m

M)

00

2 4 6 8 10 12 14 16

Time after ethanol administration (h)

Figure 1: Time course of blood ethanol concentration in mice afterethanol administration (i.p.) at various doses. Each plot representsthe mean ± SD of 3 mice. © 1 g/kg; � 2 g/kg; � 3 g/kg; � 4.5 g/kg;� 5 g/kg.

4.5, and 5 g/kg, resp.). The β values were 16.9, 16.5, 14.5,8.7, and 6.9 mmol/L/h and the blood ethanol concentrationsextrapolated to a time of zero (C0) were 25.2, 54.1, 74.8,94.9, and 104.2 mM for doses of 1, 2, 3, 4.5, and 5 g/kg,respectively. The β values were almost constant at lowdoses (1 and 2 g/kg) but decreased when the dose exceeded2 g/kg (r2 = 0.997) (Figure 2(a)). On the other hand, thenormalized AUC (AUC/dose), which negatively correlatedwith β (r2 = 0.974) (Figure 2(b)), showed a concave increasewith dose (r2 = 0.991) (Figure 2(a)) and, therefore, exhibiteda linear correlation with the square of the dose (r2 = 0.993)(data not shown). The CLT of ethanol, that is, the reciprocalof the normalized AUC, decreased dose-dependently alonga concave curve (data not shown). This differed from thebehavior of β, which exhibited a convex decrease.

3.2. Effect of Ethanol Dose on Liver ADH Parameters. LiverADH activity (the average over the ethanol-metabolizingtime for each ethanol dose) was higher for the 1 g/kg dose(P < 0.001), but lower for doses above 2 g/kg (P < 0.005 for4.5 and 5 g/kg) than that of the control (Figure 3(a)). LiverADH 1 content (the average over the ethanol-metabolizingtime) increased for the 1 g/kg dose (P < 0.0001) butdecreased at higher doses (P < 0.05 for 3 g/kg, P < 0.0001 for5 g/kg). Liver ADH 3 content (the average over the ethanol-metabolizing time) also increased for the 1 g/kg dose (P <0.0001) and showed no significant decrease at higher doses(Figure 3(b)). Within ethanol groups, liver ADH activity andliver ADH 1 content decreased dose-dependently (Figures3(a) and 3(b)), while the ratio of ADH 3 content to ADH 1content increased dose-dependently (Figure 3(c)). Both theADH 1 and ADH 3 contents correlated linearly with liverADH activity (r2 = 1.000 for each) (Figure 4). The Vmax/Km

of ADH activity of liver extract increased dose-dependently,when measured at 1 or 4 h after administration of ethanol(Figure 5).

3.3. Correlation Between Liver ADH Parameters and Phar-macokinetic Parameters. Although β showed a convex cor-relation with liver ADH activity, the CLT showed a linearcorrelation with that activity (r2 = 0.972) (Figure 6), andwith both liver ADH 1 and ADH 3 contents (r2 = 0.988 and0.987, resp.) (Figure 7).

3.4. Two-ADH-Complex Model of Liver ADH Activity. Analy-sis of the data based on the two-ADH-complex model of liverADH activity revealed that the theoretical ADH 1 activity inthe liver decreased dose-dependently, whereas the theoreticalADH 3 activity increased dose-dependently (r2 = 1.000 foreach) (Figure 8). As shown in Figure 9, the increase in theratio of theoretical activities of ADH 3 to ADH 1 correlatedpositively with the normalized AUC (r2 = 1.000), butnegatively with β (r2 = 0.984).

4. Discussion

The elimination rate of alcohol from the blood (β) is usuallyassumed to be constant regardless of the blood ethanollevel and to correspond to the rate constant of zero-orderor the Vmax of single Michaelis-Menten (M-M) eliminationkinetics [7–9]. However, the present study in mice showedthat β decreased dose-dependently at higher doses (3–5 g/kg)(Figure 2(a)), which was accompanied by a decrease in liverADH activity (Figure 3(a)). β was found to be constant onlywhen liver ADH activity was sufficiently high at low doses ofethanol (1 and 2 g/kg), in which case the liver ADH activitywas greater than that of the control. These results mean that,as the ethanol dose increases, the elimination kinetics ofethanol in mice changes from M-M to other kinetics, whichinvolves the decrease of liver ADH activity. Similar resultshave been reported for rats; β or the clearance rate decreaseddose-dependently at doses above 2 g/kg, accompanied bydose-dependent decreases of liver ADH activity [27, 28].

AUC, which represents the total amount of ethanolinvolved in systemic exposure, is an important pharmacoki-netic parameter on the bioavailability or toxicity of ethanol.In the present study, the normalized AUC (AUC/dose)showed a concave increase against ethanol dose (Figure 2(a)),probably due to the decrease of liver ADH activity at higherdoses of ethanol (Figure 3(a)). Therefore, it showed a linearcorrelation with the square of the dose, but not with doseitself (see Section 3). These data also indicate that over a widerange of doses the ethanol pharmacokinetics in mice does notsimply follow zero-order [7] or M-M kinetics [9], in whichthe relation between the normalized AUC and ethanol doseshows a proportional correlation.

Several studies have suggested that the elimination ofblood ethanol involves first-order kinetics. In humans [29]and rabbits [23], β gradually increased, even at dosesof 2 or 3 g/kg, even though the concentration of bloodethanol exceeded that at which the activity of ADH 1, the key


6

8

10

12

14

16

18

0 1 2 3 4 5 6

valu

e(m

mol

e/L

/h)

()

Dose of ethanol (g/kg)

0

50

100

150

200

(a)

Nor

mal

ized

AU

C(A

UC

/Dos

e)

(h×

kg/L

)(□

)

6 8 10 12 14 16 18

(mmol/l/h)

0

50

100

150

200

(b)

Figure 2: (a) Effect of ethanol dose on elimination rate (β) and normalized AUC (AUC/dose) of blood ethanol. (b) Correlation of normalizedAUC with β in mice for various doses of ethanol. β (©) and normalized AUC (�) were calculated from the regression line fitted to the bloodethanol concentrations at each dose in Figure 1.

100

150

200

250

300

Liv

erA

DH

acti

vity

(un

its/

kgB

.W.)

( )

0 1 2 3 4 5 6

#

a

#


(a)

20

30

40

50

60

Live

rA

DH

c

c

onte

nt

(mg/

kgB

.W.)

0 1 2 3 4 5 6

#

##

0.7

0.8

0.9

1

1.1

1.2

0 1 2 3 4 5 6

(g/ kg)

ADH3/ADH1

b( ) ( )


(b)

Figure 3: (a) Effect of ethanol dose on liver ADH activity. Three mice were sacrificed at scheduled times during ethanol metabolism aftervarious doses of ethanol: 0.5, 1, and 2 h for 1 and 2 g/kg (9 mice in each dose); 0.5, 1, 2, 4, and 8 h for 3 g/kg (15 mice in the dose); 0.5,1. 2, 4, 8, and 12 h for 0, 4.5, and 5.0 g/kg (18 mice in each dose), and livers were then removed to prepare liver extracts. The liver ADHactivity was measured by the conventional assay with 15 mM ethanol as a substrate at pH 10.7 using liver extracts and is expressed in termsof liver weight/kg body weight. The activities were averaged in each group of ethanol dose to obtain the mean ± SD. (b) Effect of ethanoldose on ADH 1 (©) and ADH 3 (�) content of liver. In addition to liver ADH activity, the liver extracts were used to measure ADH isozymecontents by EIA using isozyme-specific antibodies. Liver ADH isozyme contents were also averaged in each group of ethanol dose to obtainthe mean ± SD. (c) Effect of ethanol dose on ratio of ADH 3 content to ADH 1 content.


125

150

175

200

225

250

Liv

erA

DH

acti

vity

(un

its/

kgB

.W.)

25 30 35 40 45 50

Liver ADH isozyme content (mg/ kg B.W.)

Figure 4: Correlation of liver ADH activity with ADH 1 (©) andADH 3 (�) contents of liver. Each plot represents the value obtainedfrom Figures 3(a) and 3(b).

16

18

20

22

24

0 1 2 3 4 5 6


(un

it/m

gof

AD

H1

and

3)/m

M

App

aren

tV

max

/Km

Figure 5: Effect of ethanol dose on catalytic efficiency (Vmax/Km) ofliver ADH activity. The apparent Vmax and Km of liver ADH activitywere measured using liver extracts from mice 1 h (©) and 4 h (�)after the administration of each dose of ethanol. Vmax is expressedper mg of the sum of the ADH 1 and ADH 3 contents. Each plotrepresents the average value of 3 mice.

metabolic enzyme, is saturated [10, 24]. This type of elimi-nation of blood ethanol is probably due to the participationin ethanol metabolism of higher Km enzyme(s) without adecrease of liver ADH activity. Fujimiya et al. [23] haveproposed a parallel first-order and M-M kinetics for thistype of ethanol elimination, in which the relation betweenthe normalized AUC and ethanol dose is also linearlyproportional. However, our present results for mice suggestthat, just as in humans and rabbits, β decreases at higher

6

8

10

12

14

16

18

120 160 200 2400

0.02

0.04

0.06

0.08

0.1

(L/k

g/h

)(

)

Liver ADH activity (units/kg B.W.)

valu

e(m

mol

/L/h

)(

)

□B

ody

clea

ran

ceof

eth

anol

(CLT

)

Figure 6: Correlation of β and body clearance (CLT) with liverADH activity. β value (©) was from Figure 2. CLT value (�) was thereciprocal of the normalized AUC in Figure 2. Liver ADH activitywas from Figure 3(a).

0

0.02

0.04

0.06

0.08

0.1

25 30 35 40 45 50

(Dos

e/A

UC

)(L

/kg/

h)

Liver ADH isozyme content (mg/kg B.W.)

Bod

ycl

eara

nce

ofet

han

ol(C

L T)

Figure 7: Correlation of body clearance (CLT) with liver ADH 1and ADH 3 contents. CLT value was from Figure 6. Liver ADH 1(©) and ADH 3 (�) contents were from Figure 3(b).

doses of ethanol than 3 g/kg due to a decrease in liver ADHactivity.

The first-order kinetics in alcohol elimination fromthe blood has been clearly observed in highly intoxicatedmen with several hundred mM of blood ethanol [20, 21].ADH− deer mice, which have a low liver ADH activity dueto genetically lacking ADH 1 [6], also eliminated bloodethanol following kinetics similar to first-order one up to anethanol dose of 6 g/kg, at which the maximum blood ethanolconcentration reached around 130 mM [30]. These cases ofethanol elimination are probably carried out by a very high-Km enzyme rather than the key enzyme of ADH 1.


15

20

25

30

35

40

0 1 2 3 4 5 6

140

145

150

135

155


Th

eore

tica

lAD

H

3

acti

vity

()

Th

eore

tica

lAD

H

1 acti

vity

()□

Figure 8: Effect of ethanol dose on theoretical liver ADH 1 andADH 3 activities in two-ADH-complex model. Liver ADH 1 activitywas estimated by multiplying the ADH 1 content by the Vmax/mgof ADH 1 (4.0 units/mg). The ADH 3 activity was calculated bysubtracting the ADH 1 activity from the total liver ADH activity.The total liver ADH activity was from Figure 3(a) and liver ADH 1content from Figure 3(b). The theoretical ADH 1 (©) and ADH 3(�) activities were obtained by normalizing by the ratio of the totalADH activity to that for the control.

As non-ADH 1 pathways, MEOS and catalase have beenassumed to participate in ethanol metabolism when theblood ethanol level is high because their Kms for ethanol ishigher than that of ADH 1 [16, 31–33]. However, neither ofthese enzymes can explain the first-order kinetics observedat such high levels of blood ethanol in humans and ADH−

deer mice because their activities saturate around 50 mMof ethanol [34, 35]. Moreover, any contributions of thesetwo enzymes to systemic alcohol metabolism have not beendemonstrated even by using CYP2E1-null or acatalasemicmouse, which genetically lacks MEOS or catalase activity,respectively [17–19]. On the other hand, ADH 4, whichmainly localizes in the stomach and also has a higher Km

for ethanol than ADH 1 [36], may play an important rolein first-pass metabolism (FPM) to lower BAC and AUC[37]. However, the effect of FPM on BAC is distinct onlyat low doses of ethanol, which becomes unclear at 2 g/kgand more [37, 38]. In addition, ethanol was injected to miceintraperitoneally in our study. Therefore, the contribution ofADH 4 to BAC and β value may be negligible in this study.

We have recently proposed the participation of ADH 3,which has a very high Km for ethanol, as a non-ADH 1pathway of ethanol metabolism. Experiments on ADH 3−/−

mice showed that ADH 3 dose-dependently contributed tothe elimination of blood ethanol, probably through first-order kinetics [14]. We focused on liver ADH activity andtwo ADH isozymes, ADH 1 and ADH 3, to analyze elimina-tion kinetics of blood alcohol because the total ADH activityof the liver is closely correlated with the elimination rateof blood alcohol [1–3] and both ADH isozymes have been

6

8

10

12

14

16

18

(m

mol

/l/h

)(

)

0

50

100

150

200

Nor

mal

ized

AU

C(A

UC

/Dos

e)(□

)

0.2 0.22 0.24 0.26

Ratio of theoretical activities of two ADHs(ADH3/ADH1)

Figure 9: Correlation of β and normalized AUC (AUC/dose) ofblood ethanol with theoretical ratio of activities of the two ADHs(ADH 3/ADH 1). The values of β (©) and normalized AUC (�)were from Figure 2. Theoretical activities of ADH 1 and ADH 3 werefrom Figure 8.

demonstrated in vivo to contribute to alcohol metabolism[13, 14].

Although β does not always correlate with total liverADH activity when the activity is excessive [39, Figure 6],body clearance (CLT) exhibited a linear correlation with liverADH activity (Figure 6). CLT , which is the reciprocal of thenormalized AUC, is an important parameter indicating theethanol elimination capacity of the whole body. Many studieshave demonstrated that the rate of ethanol elimination in thewhole body (CLT or μmoles/h/animal) correlates with thetotal liver ADH activity [1, 2, 28, 40]. However, the ethanolelimination in the body cannot be explained solely by ADH1 [6, 13, 14]. The present study showed that CLT , whichcorrelated with liver ADH activity (Figure 6), also correlatedwith both contents of ADH 1 and ADH 3 (Figures 4 and 7).Therefore, it is considered that the capacity to eliminateethanol from the whole body involves not only ADH 1 butalso ADH 3, depending primarily on the level of total liverADH activity [3].

In the two-ADH-complex model, which ascribes liverADH activity to both ADH 1 and ADH 3, the theoreti-cal ADH 1 activity decreased dose-dependently (Figure 8),which is experimentally supported by the dose-dependentdecrease in liver ADH 1 content (Figure 3(b)). On theother hand, the theoretical ADH 3 activity increased dose-dependently (Figure 8). This is supported by the dose-dependent increase in the apparent Vmax/Km of ADH activityof liver extract, which is expressed in units/mg of the sumof the ADH 1 and ADH 3 contents (Figure 5). The kineticactivation of liver ADH 3 at large doses of ethanol (3–5 g/kg) was also suggested by our previous study [3]. Inaddition, the theoretical ADH 3 activity also correlated withthe ratio of the ADH 3 to the ADH 1 content, which increased


dose-dependently (Figure 3(c)). All these experimental datasupport the idea that the activity of ADH 3 increases dose-dependently due to changes in its content and/or enzymekinetics in the liver.

The changes in β and the normalized AUC againstethanol dose, which showed an inverse linear correlation(Figure 2(b)), may be ascribed to the changes in ADH 1and ADH 3 activities in the liver (Figure 9). TheoreticalADH 3 activity and normalized AUC show similar dose-dependent increases, whereas theoretical ADH 1 activityand β show similar dose-dependent decreases (Figures 2(a)and 8). The hypothesis that the increase in ADH 3 activityaccompanying the decrease in ADH 1 activity in the liverincreases the normalized AUC and decreases β (Figure 9) issupported by the fact that the ethanol-oxidizing efficiencyof ADH 3 is much less than that of ADH 1 due to its lowaffinity for ethanol. Thus, the two-ADH-complex model ofliver ADH activity explains well the dose-dependent changesin the pharmacokinetic parameters in mice. The greaterparticipation of ADH 3 and the smaller participation of ADH1 into ethanol metabolism increase AUC, which in turn raisesthe ratio of ADH 3 activity to ADH 1 activity (Figure 9).This interdependent increase in the activity ratio and AUCmay elevate the bioavailability or toxicity of ethanol. Thisdynamic theory of the elimination kinetics of ethanol basedon the two-ADH-complex model seems to be applicable toalcoholism; regarding patients with alcoholic liver disease,we already reported that the ADH 3 activity increased butthe ADH 1 activity decreased with an increase in alcoholintake. Furthermore, the ratio of ADH 3 to ADH 1 activityis significantly related to the incidence of alcoholic cirrhosisof the liver [41].

5. Conclusion

The present study suggests that the elimination kineticsof ethanol in mice changes dose-dependently from M-Mkinetics to first-order kinetics due to a shift of the dominantmetabolizing enzyme from low-Km ADH 1 to very high-Km ADH 3. Such a change in the enzymatic pathway ofethanol metabolism may elevate the toxicity of ethanol bynonlinearly increasing AUC due to a decrease in liver ADHactivity and sustaining the metabolism through an increasein ADH 3 activity. Thus, ADH 1 and ADH 3, which distributemainly in parenchymal cells and in sinusoidal endothelialcells of the liver, respectively, seem to regulate pathologicaleffects of alcohol by sharing alcohol metabolism, dependingon their catalytic efficiencies, intralobular locations, andresponsive potentials to ethanol dose.

Acknowledgment

This work was financially supported in part by the JapanSociety for Promotion of Science (no. 11470120).

References

[1] L. Lumeng, W. F. Bosron, and T. K. Li, “Quantitative corre-lation of ethanol elimination rates in vivo with liver alcohol

dehydrogenase activities in fed, fasted and food-restrictedrats,” Biochemical Pharmacology, vol. 28, no. 9, pp. 1547–1551,1979.

[2] K. E. Crow and M. J. Hardman, “Regulation of rates of ethanolmetabolism,” in Regulation of Rates of Ethanol Metabolism. inHuman Metabolism of Alcohol, Regulation, Enzymology, andMetabolites of Ethanol, K. E. Crow and R. D. Batt, Eds., vol. 11,pp. 3–16, CRC Press, Boca Raton, Fla, USA, 1989.

[3] T. Haseba, M. Kurosu, and Y. Ohno, “Dose- and time-dependent changes of mouse liver alcohol dehydrogenase(ADH) activity involving both Class I and Class III ADHs afteradministration of ethanol,” Legal Medicine, pp. 202–211, 2003.

[4] B. V. Plapp, K. G. Leidal, R. K. Smith, and B. P. Murch, “Kinet-ics of inhibition of ethanol metabolism in rats and the rate-limiting role of alcohol dehydrogenase,” Archives of Biochem-istry and Biophysics, vol. 230, no. 1, pp. 30–38, 1984.

[5] R. A. Page, K. E. Kitson, and M. J. Hardman, “The importanceof alcohol dehydrogenase in regulation of ethanol metabolismin rat liver cells,” Biochemical Journal, vol. 278, no. 3, pp. 659–665, 1991.

[6] K. G. Burnett and M. R. Felder, “Ethanol metabolism inPeromyscus genetically deficient in alcohol dehydrogenase,”Biochemical Pharmacology, vol. 29, no. 2, pp. 125–130, 1980.

[7] M. Gibaldi and D. Perrier, “Nonlinear pharmacokinetics,” inPharmacokinetics, Marcel Dekker, New York, NY, USA, 2ndedition, 1982.

[8] F. Lundquist and H. Wolthers, “The kinetics of alcohol elim-ination in man,” Acta pharmacologica et Toxicologica, vol. 14,no. 3, pp. 265–289, 1958.

[9] R. E. Rangno, J. H. Kreeft, and D. S. Sitar, “Ethanol ‘dose-dependent’ elimination: Michaelis-Menten V classical kineticanalysis,” British Journal of Clinical Pharmacology, vol. 12, no.5, pp. 667–673, 1981.

[10] B. L. Vallee and T. J. Bazzone, “Isozymes of human liver alco-hol dehydrogenase,” in Isozyme, Current Topics in Biologicaland Medical Research, J. G. Scandolios, M. C. Rattazzi, and G.S.Whitt, Eds., vol. 8, pp. 219–2447, Alan R. Liss, New York, NY,USA, 1983.

[11] T. Haseba, “Acidic pI-alcohol dehydrogenase of mouse liver:purification and characterization,” Japanese Journal of AlcoholStudies and Drug Dependence, vol. 20, no. 4, pp. 333–349, 1985.

[12] T. Haseba, S. Sato, M. Ishizaki, I. Yamamoto, M. Kurosu, andT. Watanabe, “Intralobular and intracellular localization ofalcohol dehydrogenase (ADH) isozymes in mouse liver: basicADH (class I) and acidic ADH (class III),” Biomedical Research,vol. 12, no. 3, pp. 199–209, 1991.

[13] L. Deltour, M. H. Foglio, and G. Duester, “Metabolic defi-ciencies in alcohol dehydrogenase Adh1, Adh3, and Adh4 nullmutant mice. Overlapping roles of Adh1 and Adh4 in ethanolclearance and metabolism of retinol to retinoic acid,” Journalof Biological Chemistry, vol. 274, no. 24, pp. 16796–16801,1999.

[14] T. Haseba, G. Duester, A. Shimizu, I. Yamamoto, K.Kameyama, and Y. Ohno, “in vivo contribution of Class IIIalcohol dehydrogenase (ADH3) to alcohol metabolismthrough activation by cytoplasmic solution hydrophobicity,”Biochimica et Biophysica Acta, vol. 1762, no. 3, pp. 276–283,2006.

[15] R. Teschke and J. Gellert, “Hepatic microsomal ethanol-oxi-dizing system (MEOS): metabolic aspects and clinical impli-cations,” Alcoholism: Clinical and Experimental Research, vol.10, no. 6, pp. 20–32, 1986.

[16] B. U. Bradford, C. B. Seed, J. A. Handler, D. T. Forman, and R.G. Thurman, “Evidence that catalase is a major pathway


of ethanol oxidation in vivo: dose-response studies in deermice using methanol as a selective substrate,” Archives ofBiochemistry and Biophysics, vol. 303, no. 1, pp. 172–176, 1993.

[17] H. Kono, B. U. Bradford, M. Yin et al., “Cyp2e1 is not involvedin early alcohol-induced liver injury,” American Journal ofPhysiology, vol. 277, no. 6, pp. G1259–G1267, 1999.

[18] M. L. Weiner, C. Freeman, H. Trochimowicz et al., “13-Weekdrinking water toxicity study of hydrogen peroxide with 6-week recovery period in catalase-deficient mice,” Food andChemical Toxicology, vol. 38, no. 7, pp. 607–615, 2000.

[19] V. Vasiliou, T. L. Ziegler, P. Bludeau, D. R. Petersen, F. J.Gonzalez, and R. A. Deitrich, “CYP2E1 and catalase influenceethanol sensitivity in the central nervous system,” Pharmaco-genetics and Genomics, vol. 16, no. 1, pp. 51–58, 2006.

[20] K. B. Hammond, B. H. Rumack, and D. O. Rodgerson, “Bloodethanol: a report of unusually high levels in a living patient,”Journal of the American Medical Association, vol. 226, no. 1, pp.63–64, 1973.

[21] S. O’Neill, K. F. Tipton, J. S. Prichard, and A. Quinlan, “Sur-vival after high blood alcohol levels. Association with first-order elimination kinetics,” Archives of Internal Medicine, vol.144, no. 3, pp. 641–642, 1984.

[22] N. H. G. Holford, “Clinical pharmacokinetics of ethanol,”Clinical Pharmacokinetics, vol. 13, no. 5, pp. 273–292, 1987.

[23] T. Fujimiya, K. Yamaoka, and Y. Fukui, “Parallel first-order and Michaelis-Menten elimination kinetics of ethanol.Respective role of alcohol dehydrogenase (ADH), non-ADHand first-order pathways,” Journal of Pharmacology and Exper-imental Therapeutics, vol. 249, no. 1, pp. 311–317, 1989.

[24] E. M. Algar, T. L. Seeley, and R. S. Holmes, “Purification andmolecular properties of mouse alcohol dehydrogenase isoz-ymes,” European Journal of Biochemistry, vol. 137, no. 1-2, pp.139–147, 1983.

[25] J. M. Moulis, B. Holmquist, and B. L. Vallee, “Hydrophobicanion activation of human liver xx alcohol dehydrogenase,”Biochemistry, vol. 30, no. 23, pp. 5743–5749, 1991.

[26] E. M. P. Widmark, “Verteilung und Unwandlung des Aethy-lalkohols im Organismus des Hundes,” Biochem Z, vol. 267,pp. 128–134, 1933.

[27] T. J. Braggins and K. E. Crow, “The effects of high ethanoldoses on rates of ethanol oxidation in rats. A reassessmentof factors controlling rates of ethanol oxidation in vivo,”European Journal of Biochemistry, vol. 119, no. 3, pp. 633–640,1981.

[28] M. Sharkawi, “In vivo inhibition of liver alcohol dehydroge-nase by ethanol administration,” Life Sciences, vol. 35, no. 23,pp. 2353–2357, 1984.

[29] Y. Mizoi, S. Hishida, I. Ijiri, M. Kinoshita, T. Okada, andJ. Adachi, “A study of individual differences in alcoholmetabolism, especially on blood alcohol concentration andWidmark’s β value,” Japanese Journal of Studies on Alcohol, vol.8, pp. 179–192, 1973 (Japanese).

[30] E. B. Glassman, G. A. McLaughlin, and D. T. Forman, “Roleof alcohol dehydrogenase in the swift increase in alcoholmetabolism (SIAM). Studies with deermice deficient in alco-hol dehydrogenase,” Biochemical Pharmacology, vol. 34, no. 19,pp. 3523–3526, 1985.

[31] S. Kubota, J. M. Lasker, and C. S. Lieber, “Molecular regulationof ethanol-inducible cytochrome P450-IIEI in hamsters,”Biochemical and Biophysical Research Communications, vol.150, no. 1, pp. 304–310, 1988.

[32] J. A. Handler and R. G. Thurman, “Catalase-dependentethanol oxidation in perfused rat liver. Requirement for fatty-

acid-stimulated H2O2 production by peroxisomes,” EuropeanJournal of Biochemistry, vol. 176, no. 2, pp. 477–484, 1988.

[33] J. Alderman, S. Kato, and C. S. Lieber, “The microsomalethanol oxidizing system mediates metabolic tolerance toethanol in deermice lacking alcohol dehydrogenase,” Archivesof Biochemistry and Biophysics, vol. 271, no. 1, pp. 33–39, 1989.

[34] C. S. Lieber and L. M. DeCarli, “Hepatic microsomal ethanol-oxidizing system. In vitro characteristics and adaptive proper-ties in vivo,” Journal of Biological Chemistry, vol. 245, no. 10,pp. 2505–2512, 1970.

[35] R. G. Thurman, H. G. Ley, and R. Scholz, “Hepatic microso-mal ethanol oxidation: Hydrogen peroxide formation and therole of catalase,” European Journal of Biochemistry, vol. 25, no.3, pp. 420–430, 1972.

[36] M. Zgombic-Knight, Hwee Luan Ang, M. H. Foglio, andG. Duester, “Cloning of the mouse class IV alcohol dehy-drogenase (retinol dehydrogenase) cDNA and tissue-specificexpression patterns of the murine ADH gene family,” Journalof Biological Chemistry, vol. 270, no. 18, pp. 10868–10877,1995.

[37] R. J. K. Julkunen, C. Di Padova, and C. S. Lieber, “First passmetabolism of ethanol—a gastrointestinal barrier against thesystemic toxicity of ethanol,” Life Sciences, vol. 37, no. 6, pp.567–573, 1985.

[38] K. Ukita, T. Fujimiya, and Y. Fukui, “Pharmacokinetic studyof ethanol elimination: evaluation of first-pass effect,” inBiomedical and Social Aspects of Alcohol and Alcoholism, K.Kuriyama, A. Takada, and H. Ishii, Eds., pp. 111–114, ExcerptaMedica, Amsterdam, The Netherlands, 1988.

[39] A. Zorzano, L. Ruiz del Arbol, and E. Herrera, “Effect of liverdisorders on ethanol elimination and alcohol and aldehydedehydrogenase activities in liver and erythrocytes,” ClinicalScience, vol. 76, no. 1, pp. 51–57, 1989.

[40] S. Cheema-Dhadli, F. A. Halperin, K. Sonnenberg, V. MacMil-lan, and M. L. Halperin, “Regulation of ethanol metabolismin the rat,” Biochemistry and Cell Biology, vol. 65, no. 5, pp.458–466, 1987.

[41] T. Haseba and Y. Ohno, “A new view of alcohol metabolismand alcoholism-Role of the high-Km class III alcohol dehy-drogenase (ADH3),” International Journal of EnvironmentalResearch and Public Health, vol. 7, no. 3, pp. 1076–1092, 2010.


Review Article

Autologous Bone Marrow Stem Cells in the Treatment ofChronic Liver Disease

Madhava Pai,1 Duncan Spalding,1 Feng Xi,1 and Nagy Habib1, 2

1 Department of HPB Surgery, Hammersmith Hospital, Imperial College, Hammersmith Campus, London, UK2 Faculty of Medicine, Imperial College London, Hammersmith Campus, Du Cane Road, London W12 0NN, UK

Correspondence should be addressed to Nagy Habib, [email protected]

Received 29 July 2011; Accepted 16 September 2011


Copyright © 2012 Madhava Pai et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Chronic liver disease (CLD) is increasing worldwide yet there has been no major advance in effective therapies for almost fivedecades. There is mounting evidence that adult haematopoietic stem cells (HSC) are capable of differentiating into many types oftissue, including skeletal and cardiac muscle, neuronal cells, pneumocytes and hepatocytes. These recent advances in regenerativemedicine have brought hope for patients with liver cirrhosis awaiting transplantation. New findings in adult stem cell biology aretransforming our understanding of tissue repair raising hopes of successful regenerative hepatology. Although all clinical trials todate have shown some improvement in liver function and CD34+ cells have been used safely for BM transplantation for over 20years, only randomised controlled clinical trials will be able to fully assess the potential clinical benefit of adult stem cell therapy forpatients with CLD. This article focuses on the potential of bone marrow stem cells (BMSCs) in the management of CLD and theunresolved issues regarding their role. We also outline the different mechanisms by which stem cells may impact on liver disease.

1. Introduction

The liver has a remarkable capacity to regenerate in responseto injury; however, in severe cases its regenerative capacitymay prove to be insufficient, and the liver injury mayprogress to end-stage liver disease (ESLD) and subsequentliver failure. Up to two million people suffer from chronicliver disease in the UK, many of whom remain unaware oftheir illness [1, 2]. Chronic liver disease is the fifth leadingcause of death in the UK after cancer, cardiovascular disease,stroke, and respiratory disease [2]. Alcoholic liver disease(ALD), one of the major medical complications of alcoholabuse, is the commonest cause of ESLD in Europe and NorthAmerica and is one of the most controversial indicationsfor transplantation. Alcohol abuse accounts for 80% of allliver cirrhosis cases seen in district general hospitals in theUnited Kingdom [3] and for a substantial and increasingproportion of all liver transplants performed. Chronic liverdisease (CLD) due to alcohol abuse continues to rise [4]. In2005, 4,160 people died in England and Wales from alcoholicliver disease, an increase of 37% since 1999 [1]. Alcoholicliver cirrhosis (ALC) has an unfavourable prognosis, with

a mortality of 49% and 90% after 1 and 15 yr of followup,respectively [5].

At present orthotopic liver transplantation is the onlytherapeutic option for patients with acute and chronicESLDs. Liver transplantation, however, has the disadvantageof requiring lifelong immunosuppression and followup,with 10–15% of patients dying whilst on the waiting listdue to the shortage of donated organs [6]. In 2005, onlyone-third of patients waiting for a liver transplant weretransplanted [6]. With the number of donor organs likelyto decrease over the coming decades, research into thealternative methods of treatment of whole-organ transplantis essential. Hepatocyte transplantation has been suggestedas an alternative to liver transplantation, especially forhepatic disorders caused by inherited protein deficiency [7].The widespread application of hepatocyte transplantation,however, is also limited by organ availability, by problemswith viability of isolated hepatocytes after cryopreservation,and by the potential formation of hepatocyte aggregatesduring injection subsequently obstructing liver sinusoids andresulting in portal hypertension or fatal emboli.


Recent advances in the understanding of stem cellbiology and plasticity have raised expectations for usingstem cells as a new type of cellular therapy in regener-ative medicine. In particular, adult hematopoietic stem-cell (HSC-) based treatment is evolving as a viable clinicalalternative. These cells are capable of differentiating intomany types of tissues, including skeletal and cardiac muscles,neuronal cells, pneumocytes, and hepatocytes [8]. Althoughstem cell therapy is not classically considered within therealm of clinical medicine, this technology will becomeincreasingly important for clinicians in the future.

The BM compartment is largely made up of HSCs,committed progenitor cells, and noncirculating stromalcells called mesenchymal stem cells (MSC) which havethe ability to develop into mesenchymal lineages [9, 10].Haematopoietic stem cells are adult stem cells that can beidentified by their ability to differentiate into all blood celltypes and reconstitute the haematopoietic system in a hostthat has been lethally myeloablated [11]. It was previouslythought that adult stem cells were lineage restricted, butrecent studies have shown that BM-derived progenitorsparticipate in the regeneration of ischaemic myocardium[12], damaged skeletal muscle [13], and neurogenesis [14],in addition to haematopoiesis. This paper focuses on thepotential of HSCs in the management of CLD, concentratingon experimental models in animal and human tissue alongwith the current status of clinical trials.

2. The Role of Bone Marrow Stem Cells (BMSCs)in Liver Repair

One of the first demonstrations of the ability of BMSCs toreconstitute liver was reported by Petersen and colleagues in1999. Lethally irradiated female rats with induced hepaticinjury, treated with 2-aminoacetylfluorine to prevent hepaticproliferation, and were rescued using bone marrow trans-plants from syngeneic males. The Y-chromosome markersdipeptidyl peptidase IV enzyme (DPPIV) and L21-6 antigenwere used to identify liver cells of BM origin. This cross-sex model allowed the identification of male liver cells in thefemale rats’ livers indicating that BM-derived HSCs have thecapacity to transdifferentiate into hepatocytes [15].

Although evidence of transdifferentiation to hepatocytesis compelling from animal studies, few have examined thispossibility in humans. Alison and associates detected Y-chromosome-positive cells in a retrospective analysis of thelivers of 9 female recipients of bone marrow transplants frommale donors. Cells were confirmed as being hepatocytes dueto their expression of cytokeratin-8 [16]. The authors alsolooked for the presence of Y-chromosome-positive cells in11 female livers transplanted to male recipients that werelater removed due to recurrent disease, finding a number thatexpressed cytokeratin-8 (0.5%–2%). This confirmed thatcirculating extrahepatic stem cells colonise the liver [16].

3. Mechanism of Hepatocyte Regeneration

There is much controversy concerning the mechanism bywhich BMSCs contribute to hepatocyte regeneration or to

liver repair. Transdifferentiation into hepatocytes representsgenomic plasticity in response to the microenvironment andhas been shown in several experiments in vivo [17–19].However, some authors have proposed that conversion tohepatocytes may occur via cell fusion [20, 21]. The so-called“bystander effect” is postulated to be due to factors secretedby BMSCs that are chemoattracted to the site of injury,leading to the stimulation of mitosis of endogenous livercells. This mechanism is thought to recruit endogenous BMfor cardiac repair following myocardial infarction followingadministration of granulocyte colony-stimulating factor (G-CSF) [22].

Other possible explanations for target organ regenerationand improvement in function include facilitating the releaseof vascular endothelial growth factor (VEGF) by stem cells,thus, increasing the blood supply to cells and helping torepair damaged tissue [23]. Stem cells may also act by up-regulating the Bcl-2 gene and suppressing apoptosis [24] orby suppressing inflammation in the diseased organ via theinterleukin-6 (IL-6) pathway [25]. Both of these processesare thought to contribute to the regeneration of normal cellsin the damaged organ. Finally, HSCs may stimulate tissue-specific stem cells, such as oval cells in the liver, facilitatingregeneration of the target organ [26].

4. Animal Studies

Jang and colleagues transplanted enriched CD45+ HSCs intolethally irradiated mice treated with a single dose of carbontetrachloride (CCl4) [19]. In this model, 7.6% of liver cellswere of donor origin within 7 days of transplantation. Therewas early amelioration of liver disease with some improve-ment in liver function in transplanted mice compared tocontrols. The most promising study to date demonstratesliver disease reversal following transplantation of enrichedHSCs into fumarylacetoacetate hydrolase- (FAH-) deficientmice, an animal model of tyrosinemia type I [27]. Bonemarrow from metabolically competent donor mice wastransplanted into a lethally irradiated FAH-deficient mousestrain, resulting in the proliferation of large numbers ofdonor LacZ+ hepatocytes and restoration of liver biochem-ical function. However, an animal study to investigatewhether transplantation of HSCs CD34+ could improvehepatic fibrosis by their differentiation into hepatocytesfound differing results [28]. HSCs from human umbilicalcord blood were purified, transduced with a lentiviral vectorcontaining the green fluorescent protein (GFP) gene, andinjected via the portal vein into rats with liver cirrhosisinduced by the four-month administration of thioacetamide.Rats were killed at 15 and 60 days following transplantation.Up to 37% and 22% fluorescent cells were observed inthe blood of control and cirrhotic rats respectively, at 15days after transplantation. At 60 days after transplantation;however, fluorescent cells were completely absent from theblood. Fluorescence was not detected in liver sections ateither 15 or 60 days after transplantation. A polymerase chainreaction study to detect the GFP gene ruled out silencingof the transgene. These results suggest that the transplanted


cells did not engraft in the liver and were eliminated from therats.

Oyagi and associates have transplanted MSCs, inducedto adopt hepatocyte phenotype in vitro, intravenously intononirradiated CCl4-damaged recipients and observed botha rise in serum albumin and a histological decrease inhepatic fibrosis [29]. Similarly, Jiang et al. transplantedROSA26 mouse multipotent adult progenitor cells (MAPC)into nonobese diabetic/severe combined immunodeficiencydisease (NOD/SCID) mice sacrificing them 4–24 weeks later.Recipient livers contained 5–10% donor cells colocalisedwith the hepatocyte markers CK18 and albumin. Since therewas no noxious liver injury creating a donor cell survivaladvantage, there was no increase in the number of donor cellsin the 6-month after transplant period [30].

Persistent injury has been found to induce efficient trans-differentiation of BMCs into functional hepatocytes [18].Green fluorescent protein- (GFP-) transfected BMCs fromnontreated mice injected into those with liver cirrhosisinduced by CCl4 efficiently migrated into the periportalarea of liver lobules after one day, repopulating 25% ofthe recipient liver by 4 weeks. In contrast, no GFP-positiveBMCs were detected following transplantation into controlmice with undamaged livers. Serum albumin levels weresignificantly elevated to compensate for chronic liver failurein BMC transplantation suggesting that recipient conditionsand microenvironments are key factors for successful celltherapy using BMCs.

5. Clinical Studies

Several studies have demonstrated the presence of cellsof bone marrow origin in the human liver. Alison andcolleagues [16] elegantly demonstrated that adult humanliver cells can be derived from stem cells originating inbone marrow. Analysing livers from female patients who hadreceived a bone marrow transplantation from a male donor,they found Y-chromosome- and CK8- positive hepatocytes,thus, suggesting that extrahepatic stem cells can engraft inthe liver. Theise et al. [31] also studied autopsy and liverbiopsy tissue from recipients of sex-mismatched therapeuticbone marrow and orthotopic liver transplantations. Theyidentified hepatocytes and cholangiocytes of bone marroworigin by immunocytochemistry staining for CK8, CK18,and CK19 and FISH analysis for the Y-chromosome. Theyfound up to 43% of hepatocytes and 38% of cholangiocyteswere engrafted, showing that these cells can be derivedand differentiated from bone marrow to replenish the liver.Other studies, however, have found lower numbers. Korblingand associates [32], for example, confirmed bone marrow-derived hepatocytes in liver biopsies of sex-mismatchedbone marrow transplantation, but these represented only4%–7% of hepatocytes. Similarly, Ng et al. [33] showedthat only a small proportion of hepatocytes (1.6%) wererecipient derived in the liver allografts. The inconsistencyof these studies may relate to the use of varying techniquesand markers to identify recipient-derived hepatocytes in thetransplanted patients.

Although studies have shown that bone marrow stemcells can give rise to hepatocytes, the use of bone marrowstem cells as therapeutic agents is still in its infancy. Thesestudies generally involve the mobilisation of bone marrowstem cells using granulocyte colony-stimulating factor (G-CSF) or infusion of collected bone marrow stem cells, eitherperipherally or directly into the hepatic vasculature (Table 1).Our group conducted a phase I clinical trial of the infusionof CD34+ cells into the portal vein or the hepatic arteryof five patients with chronic liver disease with no adverseeffects [34]. Although these patients received relatively lownumbers of cells (2 × 106), a moderate improvement inserum bilirubin was seen in 3 of the 5 patients whichlasted for more than 18 months [35]. Our experience isin keeping with the observations made by Am Esch IIand associates, who in their first publication demonstratedincreased liver regeneration in 3 patients following intra-portal administration of autologous CD133+ BM cells intothe left lateral portal vein branches during right portal veinembolisation (PVE). By CT criteria, left lateral segmenthypertrophy was 2.5-fold higher compared to 3 patients thathad right PVE only [36]. In their second publication, whichincluded patients from the first study, they recruited a totalof 13 patients [37]. There was a significant increase in thedaily liver growth in patients who had stem cell infusions inaddition to PVE (n = 6) when compared to patients withPVE alone (n = 7).

Terai and colleagues have also shown improvement inliver function following peripheral infusion of autologousBM cells in patients with liver cirrhosis. Nine patientswho received a peripheral vein infusion of an average of5.2 × 109 autologous mononuclear cells (CD34+, CD45+,and ckit+) demonstrated significant improvement in theChild-Pugh Scores and serum levels of albumin. Liverbiopsies were taken in 3 patients revealing an increasein proliferating cell nuclear antigen staining, an indirectmarker of hepatocyte turnover [38]. Yannaki and associateshave reported 2 patients with alcohol-induced liver cir-rhosis treated with autologous mobilised HSCs [39]. Eachpatient underwent three rounds of G-CSF mobilisation andperipheral vein infusion of CD34+ cells. The procedure waswell tolerated, and both patients improved their baselineChild-Turcotte-Pugh (CTP) and model for end-stage liverdisease (MELD) scores during 30 months of followup. Afurther 2 patients with hepatitis-B-related decompensatedliver cirrhosis treated with mobilised autologous peripheralblood monocytes (PBMC) also showed an improvement inserum albumin, bilirubin, alanine aminotransferase (ALT),aspartate aminotransferase (AST), and CTP scores forgreater than one year following transfusion [40].

Lyra and associates performed a study on 10 patientswith chronic end-stage liver disease, receiving committedprogenitor cells and no BM cells via the hepatic artery. Thisstudy showed improvement in serum bilirubin, albumin,and international normalised ratio (INR) [41]. They wenton to perform the first randomised controlled study ofautologous BMC transplantation in liver disease [42]. Thirtypatients were randomised to receive either a placebo orBMC, in the form of an autologous mononuclear cell

4 International Journal of HepatologyT

abl

e1:

Clin

ical

tria

lsof

bon

em

arro

wst

emce

llsin

liver

dise

ase.

Ref

eren

cePa

tien

tgr

oup

Nu

mbe

r(n

)Si

tein

ject

ion

BM

cells

Ou

tcom

e

Am

Esc

hII

etal

.[36

]Pa

tien

tsw

ith

liver

can

cer,

no

cirr

hos

isSt

udy

:3C

ontr

ol:3

Port

alve

inA

uto

logo

us,

bon

em

arro

w,

CD

133+

2.5-

fold

incr

ease

inle

ftlo

bein

stu

dygr

oup

onC

Tvo

lum

etry

Gor

don

etal

.[34

]

Hep

atit

isB

orC

Alc

ohol

icci

rrh

osis

and

prim

ary

scle

rosi

ng

chol

angi

tis

5Po

rtal

vein

3H

epat

icar

tery

2

Au

tolo

gou

spe

riph

eral

bloo

d,po

st-G

-CSF

CD

34+

Impr

ovem

ent

inal

bum

inor

bilir

ubi

nin

3/5

pati

ents

Tera

iet

al.[

38]

Hep

atit

isB

orC

Cir

rhos

is9

Peri

pher

alve

inA

uto

logo

us,

iliac

cres

t,u

nso

rted

BM

Impr

ovem

ent

inth

eC

hild

-Pu

ghSc

ore

and

albu

min

Yan

nak

iet

al.[

39]

Alc

ohol

icci

rrh

osis

2Pe

riph

eral

vein

Au

tolo

gou

sPe

riph

eral

bloo

d,po

st-G

-CSF

CD

34+

Impr

ovem

ent

inth

eC

hild

-Pu

ghan

dM

ELD

Scor

e

Furs

tet

al.[

37]∗

Pati

ents

wit

hliv

erca

nce

r,n

oci

rrh

osis

Stu

dy:6

Con

trol

:7Po

rtal

vein

Au

tolo

gou

s,bo

ne

mar

row

,C

D13

3+In

crea

sein

daily

liver

grow

th

Lyra

etal

.[41

]A

lcoh

olic

,hep

atit

isC

,ch

oles

tati

can

dcr

ypto

gen

icci

rrh

osis

10H

epat

icar

tery

Au

tolo

gou

s,Il

iac

cres

t,u

nso

rted

BM

Impr

ovem

ent

inC

hild

’sP

ugh

scor

e,bi

liru

bin

,an

dal

bum

in

Gas

barr

inie

tal

.[43

]D

rug-

indu

ced

acu

teliv

erfa

ilure

1Po

rtal

vein

Au

tolo

gou

sPe

riph

eral

bloo

d,po

st-G

-CSF

CD

34+

Impr

ovem

ent

incl

otti

ng

(PT

),A

ST,a

nd

ALT

up

to30

days

Pati

ent

died

ofse

psis

atda

y60

Yan

etal

.[40

]H

epat

itis

Bci

rrh

osis

2H

epat

icar

tery

Au

tolo

gou

sp

erip

her

albl

ood

mon

ocyt

espo

st-G

-CSF

Impr

ovem

ent

inal

bum

in,

Bili

rubi

n,A

ST,A

LT,a

nd

the

Ch

ild-P

ugh

Scor

e

Moh

amad

nej

adet

al.[

45]

Dec

ompe

nsa

ted

cirr

hos

is4

Hep

atic

arte

ryA

uto

logo

us

bon

em

arro

w-h

aem

atop

oiet

icC

D34

+

cells

Two

pati

ents

show

edm

inor

impr

ovem

ent

inal

bum

inon

eh

adn

eph

ropa

thy

and

hep

ator

enal

syn

drom

e—tr

ial

term

inat

ed

Moh

amad

nej

adet

al.[

46]

Dec

ompe

nsa

ted

cirr

hos

is4

Peri

pher

alve

inA

uto

logo

us

bon

em

arro

wm

esen

chym

alst

emce

llsM

ELD

Scor

eim

prov

emen

tin

2pa

tien

ts

Paie

tal

.[49

]A

lcoh

olic

liver

cirr

hos

is9

Hep

atic

arte

ryA

uto

logo

us,

post

-G-C

SF,C

D34

+

cells

expa

nde

din

vitr

oSi

gnifi

can

tim

prov

emen

tin

bilir

ubi

nan

dliv

eren

zym

es

Kh

anet

al.[

44]

Hep

atit

isB

orC

4H

epat

icar

tery

Au

tolo

gou

s,po

st-G

-CSF

,CD

34+

cells

Impr

ovem

ent

inliv

erfu

nct

ion

Kh

araz

iha

etal

.[47

]H

epat

itis

Bor

CA

lcoh

olic

cirr

hos

isan

dC

rypt

ogen

ic8

Peri

pher

alor

port

alve

inA

uto

logo

us

bon

em

arro

wm

esen

chym

alst

emce

llsSi

gnifi

can

tim

prov

emen

tin

ME

LDSc

ore

Lyra

etal

.[42

]H

epat

itis

Bor

CA

lcoh

olic

and

cryp

toge

nic

cirr

hos

is

Stu

dy:1

5C

ontr

ol:1

5H

epat

icar

tery

Au

tolo

gou

sm

onon

ucl

ear-

enri

ched

bon

em

arro

wce

llsIm

prov

emen

tin

liver

fun

ctio

n

Kim

etal

.[48

]H

epat

itis

B10

Peri

pher

alve

inA

uto

logo

us

bon

em

arro

wce

llsIm

prov

emen

tin

albu

min

,qu

alit

yof

life,

and

the

Ch

ild-P

ugh

Scor

e∗

Th

isin

clu

des

Am

Esc

hII

etal

.[36

].


preparation infused into the hepatic artery. After 3 monthsfollowup, the treated patients had a significant improvementin albumin compared to controls (16% versus 2%) and asignificant reduction in their Child-Pugh status (−8% versus+4%). There was no change in the INR between the twogroups. Gasbarrini et al. have also reported the successfuluse of autologous CD34+ BMSCs via the portal vein as arescue treatment in a patient with drug-induced acute liverfailure [43]. A liver biopsy performed at 20 days followinginfusion showed increased hepatocyte replication aroundnecrotic foci; there was also improvement in synthetic liverfunction within the first 30 days. The patient, however,died secondary to multiorgan failure related to bacterialinfection. In a study by Khan and colleagues, four patientswith liver insufficiency were given G-CSF to mobilise stemcells. CD34+ cells (0.1 × 108) were injected into the hepaticartery [44]. All patients showed improvements in serumalbumin, bilirubin, and ALT after one month from the cellinfusion.

In contrast to the previous studies, a trial of 4 patientswith decompensated cirrhosis treated with CD34+ stemcells via the hepatic artery was stopped prematurely dueto one patient developing nephropathy and hepatorenalsyndrome secondary to radiocontrast [45]. The same groupshowed that MSC transplantation in a further 4 patientswith decompensated liver cirrhosis, this time via a peripheralvein, was well tolerated and resulted in a MELD scoreimprovement in two [46]. Another study injecting MSC intoeither peripheral veins or the portal vein of patients withESLD having a MELD score ≥10 (n = 8) showed significantimprovement in their MELD score [47]. Kim et al. reportedsignificant improvement in serum albumin, quality of life,and the Child-Pugh Score in ten patients with advanced livercirrhosis due to hepatitis B infection following autologousbone marrow infusion [48]. Finally, our group has publishedthe results of administering autologous expanded mobilisedadult BM CD34+ cells via the hepatic artery in 9 patientswith alcoholic liver cirrhosis. Significant decreases in serumbilirubin, ALT, and AST levels were observed, whilst theChild-Pugh Scores and radiological ascites improved in 7 and5 patients, respectively [49].

6. Limitations of Studies and Future Issues

Although all clinical trials to date have shown some improve-ment in liver function, it must be remembered that thenatural history of cirrhosis tends to be variable. Thus,one would expect some patients to improve with time,particularly in compliant patients who can be followedup and remain abstinent from alcohol. The liver containsapproximately 2.8 × 1011 hepatocytes, and the requiredmass of cells to correct a single enzyme biochemical defectis likely to be significantly less than that required fortreatment of chronic or acute liver failure. There is evidenceto suggest that transplantation of only 1–5% of the totalliver mass may be sufficient to restore adequate functionalactivity [7, 50]. Cells can be delivered to patients viaa peripheral vein, the portal vein, hepatic artery, or an

intrasplenic injection. As both fulminant and chronic, liverfailure requires the replacement of greater than 10% offunctional liver; the cell mass required for transplantationwill be significantly higher. The liver cell mass is restoredprimarily through division of mature hepatocytes. Stimula-tion of regeneration, such as a partial hepatectomy, promotesincreases in carbamoyl phosphate synthetase I activity withsubsequent liver hypertrophy [51]. This early experiencesuggests that this therapeutic approach has the potential ofboth enhancing and accelerating hepatic regeneration in aclinical setting.

Unlike hepatocytes, the use of BMSCs for liver regen-eration does not depend on the procurement of cadavericlivers whose donors are often immunologically disparateand also in short supply. The use of adult stem cells isattractive as it overcomes the moral and ethical barriers of EScell manipulation. Further advantages of the use of BMSCsare that they are multipotent, there is already considerableexperience in their use, they are easily accessible, and thereis unlimited supply. Conversely, concerns have been raisedabout the adverse long-term effects of stem cell therapy.There is evidence to suggest that treatment with BMSCsmay provide liver fibrogenic cells (hepatic stellate cells andmyofibroblasts) which contribute to fibrosis and could havea deleterious effect on already decompensated cirrhotic livers[52, 53]. Similarly, there are concerns that hepatocellularcarcinoma (HCC) originates from hepatic oval cells andBMSCs [54]. Much of the data concerning the malignantpotential of BMSCs, however, originates from geneticallymodified rodent models and may not be present in humans[55].

New findings in adult stem cell biology are transformingour understanding of tissue repair raising hopes of successfulregenerative hepatology. Although all clinical trials to datehave shown some improvement in liver function and CD34+

cells have been used safely for BM transplantation for over20 years, only randomised controlled clinical trials will beable to fully assess the potential clinical benefit of adult stemcell therapy for patients with liver insufficiency secondary toALD.

References

[1] British Liver Trust, “Facts about liver disease,” 2007, http://www.britishlivertrust.org.uk/home/about-us/media-centre/facts-about-liver-disease.aspx.

[2] “UK National Statistics,” 2009, http://www.statistics.gov.uk/hub/index.html.

[3] K. Walsh and G. Alexander, “Alcoholic liver disease,” Postgrad-uate Medical Journal, vol. 76, no. 895, pp. 280–286, 2000.

[4] R. Williams, “The pervading influence of alcoholic liverdisease in hepatology,” Alcohol and Alcoholism, vol. 43, no. 4,pp. 393–397, 2008.

[5] H. Bell, J. Jahnsen, E. Kittang, N. Raknerud, and L. Sand-vik, “Long-term prognosis of patients with alcoholic livercirrhosis: a 15-year follow-up study of 100 Norwegian patientsadmitted to one unit,” Scandinavian Journal of Gastroenterol-ogy, vol. 39, no. 9, pp. 858–863, 2004.

[6] United Network of Organ Sharing Online (UNOS), “Data,”2006, http://www.unos.org/donation/index.php?topic=data.


[7] I. J. Fox, J. R. Chowdhury, S. S. Kaufman et al., “Treatment ofthe Crigler-Najjar syndrome type I with hepatocyte transplan-tation,” The New England Journal of Medicine, vol. 338, no. 20,pp. 1422–1426, 1998.

[8] S. L. Preston, M. R. Alison, S. J. Forbes, N. C. Direkze, R.Poulsom, and N. A. Wright, “The new stem cell biology:something for everyone,” Journal of Clinical Pathology—Molecular Pathology, vol. 56, no. 2, pp. 86–96, 2003.

[9] M. F. Pittenger, A. M. Mackay, S. C. Beck et al., “Multilineagepotential of adult human mesenchymal stem cells,” Science,vol. 284, no. 5411, pp. 143–147, 1999.

[10] P. Bianco, M. Riminucci, S. Gronthos, and P. G. Robey, “Bonemarrow stromal stem cells: nature, biology, and potentialapplications,” Stem Cells, vol. 19, no. 3, pp. 180–192, 2001.

[11] E. L. Herzog, L. Chai, and D. S. Krause, “Plasticity of marrow-derived stem cells,” Blood, vol. 102, no. 10, pp. 3483–3493,2003.

[12] D. Orlic, J. Kajstura, S. Chimenti et al., “Mobilized bonemarrow cells repair the infarcted heart, improving functionand survival,” Proceedings of the National Academy of Sciencesof the United States of America, vol. 98, no. 18, pp. 10344–10349, 2001.

[13] E. Gussoni, Y. Soneoka, C. D. Strickland et al., “Dystrophinexpression in the mdx mouse restored by stem cell transplan-tation,” Nature, vol. 401, no. 6751, pp. 390–394, 1999.

[14] E. Mezey, K. J. Chandross, G. Harta, R. A. Maki, and S. R.McKercher, “Turning blood into brain: cells bearing neuronalantigens generated in vivo from bone marrow,” Science, vol.290, no. 5497, pp. 1779–1782, 2000.

[15] B. E. Petersen, W. C. Bowen, K. D. Patrene et al., “Bone marrowas a potential source of hepatic oval cells,” Science, vol. 284, no.5417, pp. 1168–1170, 1999.

[16] M. R. Alison, R. Poulsom, R. Jeffery et al., “Hepatocytes fromnon-hepatic adult stem cells,” Nature, vol. 406, no. 6793, p.257, 2000.

[17] D. S. Krause, N. D. Theise, M. I. Collector et al., “Multi-organ,multi-lineage engraftment by a single bone marrow-derivedstem cell,” Cell, vol. 105, no. 3, pp. 369–377, 2001.

[18] S. Terai, I. Sakaida, N. Yamamoto et al., “An in vivo modelfor monitoring trans-differentiation of bone marrow cells intofunctional hepatocytes,” Journal of Biochemistry, vol. 134, no.4, pp. 551–558, 2003.

[19] Y. Y. Jang, M. I. Collector, S. B. Baylin, A. M. Diehl, and S.J. Sharkis, “Hematopoietic stem cells convert into liver cellswithin days without fusion,” Nature Cell Biology, vol. 6, no. 6,pp. 532–539, 2004.

[20] G. Vassilopoulos, P. R. Wang, and D. W. Russell, “Transplantedbone marrow regenerates liver by cell fusion,” Nature, vol. 422,no. 6934, pp. 901–904, 2003.

[21] X. Wang, H. Willenbring, Y. Akkari et al., “Cell fusion isthe principal source of bone-marrow-derived hepatocytes,”Nature, vol. 422, no. 6934, pp. 897–901, 2003.

[22] H. Takano, M. Ohtsuka, H. Akazawa et al., “Pleiotropiceffects of cytokines on acute myocardial infarction: G-CSFas a novel therapy for acute myocardial infarction,” CurrentPharmaceutical Design, vol. 9, no. 14, pp. 1121–1127, 2003.

[23] J. Tang, Q. Xie, G. Pan, J. Wang, and M. Wang, “Mesenchymalstem cells participate in angiogenesis and improve heart func-tion in rat model of myocardial ischemia with reperfusion,”European Journal of Cardio-thoracic Surgery, vol. 30, no. 2, pp.353–361, 2006.

[24] Z. Chen, C. C. Chua, Y. S. Ho, R. C. Hamdy, and B. H.L. Chua, “Overexpression of Bcl-2 attenuates apoptosis andprotects against myocardial I/R injury in transgenic mice,”

The American Journal of Physiology—Heart and CirculatoryPhysiology, vol. 280, no. 5, pp. H2313–H2320, 2001.

[25] M. Wang, B. M. Tsai, P. R. Crisostomo, and D. R. Meldrum,“Pretreatment with adult progenitor cells improves recoveryand decreases native myocardial proinflammatory signalingafter ischemia,” Shock, vol. 25, no. 5, pp. 454–459, 2006.

[26] T. W. Austin and E. Lagasse, “Hepatic regeneration fromhematopoietic stem cells,” Mechanisms of Development, vol.120, no. 1, pp. 131–135, 2003.

[27] E. Lagasse, H. Connors, M. Al-Dhalimy et al., “Purifiedhematopoietic stem cells can differentiate into hepatocytes invivo,” Nature Medicine, vol. 6, no. 11, pp. 1229–1234, 2000.

[28] M. J. Saez-Lara, C. Frecha, F. Martin et al., “ransplantation ofhuman CD34+ stem cells from umbilical cord blood to ratswith thioacetamide-induced liver cirrhosis,” Xenotransplanta-tion, vol. 13, no. 6, pp. 529–535, 2006.

[29] S. Oyagi, M. Hirose, M. Kojima et al., “Therapeutic effect oftransplanting HGF-treated bone marrow mesenchymal cellsinto CCl4-injured rats,” Journal of Hepatology, vol. 44, no. 4,pp. 742–748, 2006.

[30] Y. Jiang, B. N. Jahagirdar, R. L. Reinhardt et al., “Pluripotencyof mesenchymal stem cells derived from adult marrow,”Nature, vol. 418, no. 6893, pp. 41–49, 2002.

[31] N. D. Theise, S. Badve, R. Saxena et al., “Derivation ofhepatocytes from bone marrow cells in mice after radiation-induced myeloablation,” Hepatology, vol. 31, no. 1, pp. 235–240, 2000.

[32] M. Korbling, R. L. Katz, A. Khanna et al., “Hepatocytes andepithelial cells of donor origin in recipients of peripheral-blood stem cells,” The New England Journal of Medicine, vol.346, no. 10, pp. 738–746, 2002.

[33] I. O. Ng, K. L. Chan, W. H. Shek et al., “High frequency ofchimerism in transplanted livers,” Hepatology, vol. 38, no. 4,pp. 989–998, 2003.

[34] M. Y. Gordon, N. Levicar, M. Pai et al., “Characterizationand clinical application of human CD34+ stem/progenitor cellpopulations mobilized into the blood by granulocyte colony-stimulating factor,” Stem Cells, vol. 24, no. 7, pp. 1822–1830,2006.

[35] N. Levicar, M. Pai, N. A. Habib et al., “Long-term clinicalresults of autologous infusion of mobilized adult bone marrowderived CD34+ cells in patients with chronic liver disease,” CellProliferation, vol. 41, supplement 1, pp. 115–125, 2008.

[36] J. S. Am Esch II, W. T. Knoefel, M. Klein et al., “Portalapplication of autologous CD133+ bone marrow cells to theliver: a novel concept to support hepatic regeneration,” StemCells, vol. 23, no. 4, pp. 463–470, 2005.

[37] G. Furst, J. S. Am Esch, L. W. Poll et al., “Portal veinembolization and autologous CD133+ bone marrow stem cellsfor liver regeneration: initial experienceSafety and efficacy ofautologous bone marrow stem cell transplantation throughhepatic artery for the treatment of chronic liver failure:a preliminary studyImprovement of liver function in livercirrhosis patients after autologous mesenchymal stem cellinjection: a phase I-II clinical trial,” Radiology, vol. 243, no.1, pp. 171–179, 2007.

[38] S. Terai, T. Ishikawa, K. Omori et al., “Improved liver functionin patients with liver cirrhosis after autologous bone marrowcell infusion therapy,” Stem Cells, vol. 24, no. 10, pp. 2292–2298, 2006.

[39] E. Yannaki, A. Anagnostopoulos, D. Kapetanos et al., “Lastingamelioration in the clinical course of decompensated alcoholiccirrhosis with boost infusions of mobilized peripheral blood


stem cells,” Experimental Hematology, vol. 34, no. 11, pp.1583–1587, 2006.

[40] L. Yan, Y. Han, J. Wang, J. Liu, L. Hong, and D. Fan,“Peripheral blood monocytes from patients with HBV relateddecompensated liver cirrhosis can differentiate into functionalhepatocytes,” The American Journal of Hematology, vol. 82, no.11, pp. 949–954, 2007.

[41] A. C. Lyra, M. B. Soares, L. F. da Silva et al., “Feasiblity andsafety of autologous bone marrow mononuclear cell trans-plantation in patients with advanced chronic liver disease,”World Journal of Gastroenterology, vol. 13, no. 7, pp. 1067–1073, 2007.

[42] A. C. Lyra, M. B. Soares, L. F. M. Da Silva et al., “Infusion ofautologous bone marrow mononuclear cells through hepaticartery results in a short-term improvement of liver functionin patients with chronic liver disease: a pilot randomizedcontrolled study,” European Journal of Gastroenterology andHepatology, vol. 22, no. 1, pp. 33–42, 2010.

[43] A. Gasbarrini, G. L. Rapaccini, S. Rutella et al., “Rescuetherapy by portal infusion of autologous stem cells in a caseof drug-induced hepatitis,” Digestive and Liver Disease, vol. 39,no. 9, pp. 878–882, 2007.

[44] A. A. Khan, N. Parveen, V. S. Mahaboob et al., “Safety andefficacy of autologous bone marrow stem cell transplantationthrough hepatic artery for the treatment of chronic liverfailure: a preliminary study,” Transplantation Proceedings, vol.40, no. 4, pp. 1140–1144, 2008.

[45] M. Mohamadnejad, M. Namiri, M. Bagheri et al., “Phase 1human trial of autologous bone marrow-hematopoietic stemcell transplantation in patients with decompensated cirrhosis,”World Journal of Gastroenterology, vol. 13, no. 24, pp. 3359–3363, 2007.

[46] M. Mohamadnejad, K. Alimoghaddam, M. Mohyeddin-Bonab et al., “Phase 1 trial of autologous bone marrowmesenchymal stem cell transplantation in patients withdecompensated liver cirrhosis,” Archives of Iranian Medicine,vol. 10, no. 4, pp. 459–466, 2007.

[47] P. Kharaziha, P. M. Hellstrom, B. Noorinayer et al., “Improve-ment of liver function in liver cirrhosis patients after autol-ogous mesenchymal stem cell injection: a phase I-II clinicaltrial,” European Journal of Gastroenterology & Hepatology, vol.21, no. 10, pp. 1199–1205, 2009.

[48] J. K. Kim, Y. N. Park, J. S. Kim et al., “Autologous bone marrowinfusion activates the progenitor cell compartment in patientswith advanced liver cirrhosis,” Cell Transplantation, vol. 19, no.10, pp. 1237–1246, 2010.

[49] M. Pai, D. Zacharoulis, M. N. Milicevic et al., “Autologousinfusion of expanded mobilized adult bone marrow-derivedCD34+ cells into patients with alcoholic liver cirrhosis,” TheAmerican Journal of Gastroenterology, vol. 103, no. 8, pp. 1952–1958, 2008.

[50] M. Muraca, G. Gerunda, D. Neri et al., “Hepatocyte transplan-tation as a treatment for glycogen storage disease type 1a,” TheLancet, vol. 359, no. 9303, pp. 317–318, 2002.

[51] A. C. Selden and H. J. F. Hodgson, “Growth factors and theliver,” Gut, vol. 32, no. 6, pp. 601–603, 1991.

[52] F. P. Russo, M. R. Alison, B. W. Bigger et al., “The bone marrowfunctionally contributes to liver fibrosis,” Gastroenterology,vol. 130, no. 6, pp. 1807–1821, 2006.

[53] D. Cassiman, A. Barlow, S. Vander Borght, L. Libbrecht, and V.Pachnis, “Hepatic stellate cells do not derive from the neuralcrest,” Journal of Hepatology, vol. 44, no. 6, pp. 1098–1104,2006.

[54] X. Z. Wu and D. Chen, “Origin of hepatocellular carcinoma:role of stem cells,” Journal of Gastroenterology and Hepatology,vol. 21, no. 7, pp. 1093–1098, 2006.

[55] M. L. Dumble, E. J. Croager, G. C. T. Yeoh, and E. A. Quail,“Generation and characterization of p53 null transformedhepatic progenitor cells: oval cells give rise to hepatocellularcarcinoma,” Carcinogenesis, vol. 23, no. 3, pp. 435–445, 2002.


Research Article

Alcohol Activates TGF-Beta but Inhibits BMPReceptor-Mediated Smad Signaling and Smad4 Binding toHepcidin Promoter in the Liver

Lisa Nicole Gerjevic, Na Liu, Sizhao Lu, and Duygu Dee Harrison-Findik

Division of Gastroenterology and Hepatology, Department of Internal Medicine,University of Nebraska Medical Center, 95820 UNMC, Omaha, NE 68198-5820, USA

Correspondence should be addressed to Duygu Dee Harrison-Findik, [email protected]

Received 28 May 2011; Accepted 7 August 2011


Copyright © 2012 Lisa Nicole Gerjevic et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Hepcidin, a key regulator of iron metabolism, is activated by bone morphogenetic proteins (BMPs). Mice pair-fed with regular andethanol-containing L. De Carli diets were employed to study the effect of alcohol on BMP signaling and hepcidin transcription inthe liver. Alcohol induced steatosis and TGF-beta expression. Liver BMP2, but not BMP4 or BMP6, expression was significantlyelevated. Despite increased BMP expression, the BMP receptor, and transcription factors, Smad1 and Smad5, were not activated.In contrast, alcohol stimulated Smad2 phosphorylation. However, Smad4 DNA-binding activity and the binding of Smad4 tohepcidin promoter were attenuated. In summary, alcohol stimulates TGF-beta and BMP2 expression, and Smad2 phosphorylationbut inhibits BMP receptor, and Smad1 and Smad5 activation. Smad signaling pathway in the liver may therefore be involvedin the regulation of hepcidin transcription and iron metabolism by alcohol. These findings may help to further understand themechanisms of alcohol and iron-induced liver injury.

1. Introduction

Alcoholic liver-disease patients frequently display evidenceof iron overload [1–5]. Alcohol-induced iron overloadenhances the production of free radicals and proinflamma-tory cytokines [6, 7]. However, the underlying mechanismsof iron accumulation observed in alcoholic liver diseaseare unclear. We and others have recently shown a rolefor hepcidin in alcohol-induced increases in iron transport[8–13]. Hepcidin is a circulatory antimicrobial peptidesynthesized by the liver [14, 15]. It plays a pivotal role iniron homeostasis by inhibiting iron uptake in the duodenumand iron export in reticuloendothelial macrophages [16,17]. Alcohol downregulates hepcidin expression in the liver,which leads to an increase in duodenal iron transporterexpression [9]. However, how alcohol suppresses hepcidintranscription in the liver is still unclear.

Bone morphogenetic proteins (BMPs) belong to thetransforming growth factor beta (TGF-β), superfamily of

growth factors [18]. BMP2, BMP4, BMP6 and BMP9have all been reported to regulate hepcidin transcription[19–22]. However, transgenic mouse studies have recentlysuggested that BMP6, is involved in the regulation ofhepcidin expression in vivo [23, 24]. Moreover, iron hasbeen shown to induce BMP6 mRNA expression and Smad5phosphorylation [25–27]. Similar to TGF-β receptor, thebinding of BMP ligands to type I and type II BMP receptorserine/threonine kinases leads to the phosphorylation andactivation of type I BMP receptor (BMPR-I) [28]. ActivatedBMPR-I in turn phosphorylates the receptor-regulated Smad(R-Smad) family of transcription factors: Smad1, Smad5,and Smad8 [29]. On the other hand, activated TGF-βreceptor induces the phosphorylation of Smad2 and Smad3.Upon phosphorylation, these R-Smads form a complex withthe common mediator of Smad signaling, Smad4. The Smadcomplexes subsequently translocate into the nucleus wherethey participate in the regulation of gene transcription [30,31]. Of note, liver-specific disruption of Smad4 leads to


(a) (b)

Figure 1: Liver Histology. The fixed liver sections of mice fed with regular (a) or ethanol-containing (b) Lieber-De Carli liquid diets werestained with hematoxylin and eosin (a, b), as described in Materials and Methods (original magnification 20X). The arrows indicate steatosis(b).

a decrease in hepcidin expression and accumulation of ironin liver, kidney, and pancreas [32].

The involvement of the profibrogenic cytokine, TGF-β in alcohol-induced liver injury has been well-established[33, 34]. However, the role of BMPs and BMP receptor-mediated signaling in alcoholic liver disease is largelyunknown. In this study, we examine the effect of alcohol onBMP expression and BMP receptor-mediated regulation ofhepcidin transcription in the liver in vivo. Alcohol and ironplay a synergistic role in the pathogenesis of alcoholic liverdisease. These studies will help us to further understand themechanisms of liver injury induced by iron and alcohol.


2.1. Animal Experiments. Animal experiments were ap-proved by the Animal Ethics Committee at the University ofNebraska Medical Center. C57BL/6 NCR male mice (NIH)were housed individually and pair-fed with either regularor ethanol-containing Lieber De Carli liquid diets (Dyets,Inc., cat no: 710027, 710260, resp.), as described previously[12]. The ethanol content of the diet was gradually increasedover a 9-day period to 5% (no ethanol for 3 days, 1% for 2days, 2% for 2 days, and 3% for 2 days). Mice were exposedto 5% ethanol for 4 weeks. For iron experiments, micewere fed initially with a custom prepared egg-white-basedsolid rodent diet [35] containing 0.02% carbonyl iron (F614,Bio-Serv, Inc.) for one week to achieve a basal hepcidinexpression level. Subsequently, they were fed with 0.2% or2% carbonyl iron diets for 3 weeks to achieve normal andiron overload states, respectively, as published previously[12].

2.2. RNA Isolation, cDNA Synthesis, and Real-Time Quan-titative PCR Analysis. RNA isolation, cDNA synthesis, andquantitative PCR were performed, as published previously[9]. The sequences of Taqman fluorescent probe (5

′6-

[FAM]; 3′

[TAMRA-Q]) and primers are shown in Table 1.

2.3. Western Blotting, Immunoprecipitation, and Immuno-histochemistry. Total liver cell lysates were prepared by

homogenizing mouse livers in lysis buffer [10 mM Tris/HCl(pH 7.4), 100 mM NaCl, 5 mM EDTA, 10% glycerol, 1 mMPMSF, complete protease inhibitor cocktail (Roche Diag-nostics Corp.), phosphatase inhibitor cocktail A (SantaCruz, sc-45044), and 1% Triton-X-100]. The lysates weresubsequently incubated on ice for 20 min. and centrifuged(3000x g) for 5 min. at 4◦C. Supernatants were employed forwestern blot or immunoprecipitation experiments. Westernblots were performed, as described previously [12, 36].Anti-phospho-Smad2, anti-phosho-Smad1/5, anti-Smad2,and anti-Smad5 antibodies were obtained commercially (cellsignaling). For immunoprecipitations, 500 μg of liver lysateprotein was incubated with BMPR-I antibody or normalrabbit IgG (Santa Cruz) and protein A/G PLUS-Agarose pre-blocked with BSA (Santa Cruz). Immunocomplexes elutedby nonreducing SDS buffer were resolved on 10% polyacry-lamide gels and immunoblotted with anti-phosphoserine(Millipore) or BMPR-I antibodies (Santa Cruz). Alkalinephosphatase-conjugated anti-mouse (Millipore) or anti-rabbit (SouthernBiotech) light chain-specific immunoglob-ulins were used as secondary antibodies. Immunostainingof paraffin embedded liver sections with TGF-β (Abcam) orBMP2 (Santo Cruz) antibodies were performed by VectastainABC kit (Vector Labs), according to manufacturer’s instruc-tions.

2.4. Electrophoretic Mobility Gelshift Assay (EMSA). Mouseliver nuclear lysate isolation and EMSA were performed,as described [9]. Briefly, the consensus and mutant Smad4oligonucleotides (Santa Cruz) were labeled by T4 polynu-cleotide kinase and 32P-γ-ATP (Perkin Elmer, 3.000 Ci/moL,10 mCi/mL). 7 μg of nuclear extract protein and 100.000 cpmof 32P-labeled Smad probes were used for each bindingreaction. Protein and DNA complexes were resolved on7% nondenaturing polyacrylamide gels and radiolabeledbands were visualized by autoradiography. For competitionassays, unlabeled consensus Smad oligonucleotide in 30-fold excess was incubated with nuclear lysates on iceprior to the addition of the 32P-labeled consensus Smadprobe.


(a) (b)

(c) (d)

Control Alcohol Alcohol

TGF-

(e)

Control Alcohol Alcohol

Gapdh

(f)

0

0.5

1

1.5

2

2.5

Fold

TG

F-ex

pres

sion

Control Alcohol

(g)

Figure 2: TGF-β Expression. The fixed liver sections of mice fed with regular (a) or ethanol-containing (b) Lieber-De Carli liquid diets wereimmunostained with an anti-TGF-β antibody, as described in Materials and Methods. Liver sections from mice injected (i.p.) with sunfloweroil, as control (c) or carbon tetrachloride (d) for 8 weeks were also immunostained with the anti-TGF-β antibody to serve as positive controlsfor TGF-β staining. The arrows indicate TGF-β expression (b, d) (original magnification 20X). Whole cell lysates isolated from the livers ofmice fed with regular (control) or ethanol-containing (alcohol) Lieber-De Carli liquid diets were employed to determine TGF-β proteinexpression by western blotting, as described in Materials and Methods (e). An anti-gapdh antibody was employed to confirm equal proteinloading (f). TGF-β protein expression, normalized to gapdh, in alcohol-treated mice was expressed as fold expression of that in control miceand was quantified by scanning autoradiographs by a densitometer (g).


Control Alcohol

Phospho-Smad2

(a)

Control Alcohol

Smad2

(b)

0

1

2

3

Control Alcohol

Fold

p-Sm

ad2

expr

essi

on

(c)

Figure 3: Smad2 Activation. The phosphorylation of Smad2 protein in the liver lysates of mice fed with regular (control) or ethanol-containing (alcohol) Lieber-De Carli liquid diets was determined by western blotting employing an anti-phospho-Smad2 antibody, asdescribed in Materials and Methods. (b) An antibody recognizing the total levels of Smad2 protein was employed as a control to confirmequal protein loading. (c) Autoradiographs from different experiments (n = 3) were scanned by a densitometer, and phospho-Smad2 (p-Smad2) expression in each sample was quantified by normalizing to total Smad2 protein expression. Normalized phospho-Smad2 expressionin alcohol-treated mice was expressed as fold expression of that in control mice.

Table 1: Mouse-specific sequences of real-time quantitative PCR probe and primers.

Gene Forward primer (5′-3

′) Reverse primer (5

′-3

′) Taqman probe (5

′-3

′)

BMP2 GCATCCAGCCGACCCTT GCCTCAACTCAAATTCGCTGA TCCCGGCCTTCGGAAGACGTC

BMP4 GGACTTCGAGGCGACACTTC TTGCTAGGCTGCGGACG ACAGATGTTTGGGCTGCGCCG

BMP6 CCTCTTCTTCGGGCTTCCTC CCTTTTGCATCTCCCGCTT ATCGGCGGCTCAAGACCCACG

Hepcidin ACTCGGACCCAGGCTGC AGATAGGTGGTGCTGCTCAGG TGTCTCCTGCTTCTCCTCCTTGCCA

2.5. Chromatin Immunoprecipitation (CHIP). CHIP wasperformed, as described [37]. Chromatin isolated fromformalin-fixed mouse liver was sheared by sonication andimmunoprecipitated by using control IgG (cell signaling) oranti-Smad4 antibody (cell signaling) and protein A/G beads(Santa Cruz). An aliquot of precleared chromatin was savedas total input DNA prior to the immunoprecipitation. Coim-munoprecipitated DNA and total input DNA were analyzedby PCR using primers (forward 5

′-gccatactgaaggcactga

′3;

reverse 5′-gtgtggtggctgtctagg-3

′) specific for mouse hepcidin

promoter.

2.6. Statistical Analysis. Statistical analysis of differences intreatment groups was performed by using the nonparametricMann-Whitney test and Student’s t-test.

3. Results

In order to study the effect of chronic alcohol consumptionon the expression of different bone morphogenetic proteins

(BMPs) and signaling in the liver, we employed wild-typemice pair-fed with regular (control) or ethanol-containingLieber De Carli diets, as described in Materials and Methods.Mice fed with alcohol for 4 weeks displayed significant lipidaccumulation in the liver, compared to control mice fedwith regular L. De Carli diet, as shown by hematoxylin andeosin staining (Figures 1(a) and 1(b)). Similarly, chronicalcohol consumption resulted in increased transforminggrowth factor beta (TGF-β) expression in the liver, as shownby immunostaining and western blotting (Figure 2). TGF-βis known to induce the phosphorylation and activation ofthe transcription factor, Smad2 [29]. Accordingly, westernblot analysis indicated a significant twofold increase in thelevel of phospho-Smad2 protein expression in the livers ofalcohol-fed mice compared to control mice (Figures 3(a)and 3(c)). The level of total Smad2 protein expression inthe liver was not altered by alcohol (Figure 3(b)). BMPsalso belong to the TGF-β superfamily of growth factors andactivate the Smad signaling pathway [18]. However, the effect


0

1

2

3

4

∗

Control Alcohol Control Alcohol Control AlcoholBMP2 BMP4 BMP6

Fold

BM

Pm

RN

Aex

pres

sion

(nor

mal

ized

toga

pdh

)

(a)

(b) (c)

Figure 4: Alcohol and BMP2, BMP4, BMP6 Expression. (a) cDNA was synthesized from liver RNA of mice fed with regular (control) orethanol-containing (alcohol) Lieber-De Carli liquid diets, as described in Materials and Methods. It was employed in real-time PCR todetect bone morphogenetic protein (BMP) expression. The mRNA expression in alcohol-fed mice was expressed as fold of that in pair-fedcontrol mice fed with regular diet (mean ± S.E.M.; n = 3, 4 mice per group). Asterisks indicate statistical significance (P < 0.05). The fixedliver sections of mice fed with regular (b) or ethanol-containing (c) Lieber-De Carli liquid diets were immunostained with an anti-BMP2antibody, as described in Materials and Methods. The arrows indicate BMP2 expression (original magnification 20X).

0

0.5

1

1.5

Control Alcohol

(nor

mal

ized

toga

pdh

)Fo

ldh

epci

din

mR

NA

expr

essi

on

Figure 5: Alcohol and Hepcidin Expression. cDNA was synthesized from liver RNA of mice fed with regular (control) or ethanol-containing(alcohol) Lieber-De Carli liquid diets, as described in Materials and Methods. It was employed in real-time PCR to detect hepcidin expression.The mRNA expression in alcohol-fed mice was expressed as-fold of that in pair-fed control mice fed with regular diet (mean± S.E.M.; n = 3,4 mice per group). Asterisks indicate statistical significance (P < 0.05).


Con

trol

Alc

ohol

Con

trol

Alc

ohol

Nor

mal

Ove

rloa

d

L. De Carli Diet Carbonyl iron diet

Phospho-Smad 1/5

(a)

Con

trol

Alc

ohol

Con

trol

Alc

ohol

Nor

mal

Ove

rloa

d

L. De Carli Diet Carbonyl iron diet

Smad 5

(b)

0

1

2

3

Control Alcohol Normal OverloadL. De Carli Diet Carbonyl iron diet

Fold

p-Sm

ad5

Exp

ress

ion

(c)

Figure 6: Activation of Smad1 and Smad5. (a) The phosphorylation of Smad1 and Smad5 proteins in the liver lysates of mice fed withregular (control) or ethanol-containing (alcohol) Lieber-De Carli liquid diets, and of mice fed with diets containing 0.2% (normal) or 2%(overload) carbonyl iron was determined by western blotting employing an anti-phospho-Smad1/5 antibody, as described in Materials andMethods. (b) An antibody recognizing the total levels of Smad5 protein was employed as a control to confirm equal protein loading. (c)Autoradiographs from different experiments (n = 3) were scanned by a densitometer, and phospho-Smad1/5 (p-Smad5) expression in eachsample was quantified by normalizing to total Smad5 protein expression. Normalized phospho-Smad expression in alcohol or iron-fed micewas expressed as fold expression of that in the control mice.

of alcohol on BMP expression is unknown. Compared tocontrol mice, mice with chronic alcohol exposure displayedan increase in BMP2, BMP4, and BMP6 mRNA expressionin the liver (Figure 4(a)). However, the median responsedifferences in BMP4 and BMP6 expression between alcohol-fed and control mice were not statistically significant (P >0.05) (Figure 4(a)). In contrast, the alcohol-induced increasein BMP2 mRNA expression in the liver was statisticallysignificant (P < 0.05) (Figure 4(a)). The livers of alcohol-treated mice also exhibited an increase in BMP2 proteinexpression compared to control mice (Figures 4(b) and 4(c)).Mice with chronic alcohol exposure displayed a significant(P < 0.05) decrease in hepcidin mRNA expression in the liver(Figure 5).

Bone morphogenetic proteins induce intracellular sig-naling via the phosphorylation of the transcription factors,Smad1, Smad5, and Smad8. We performed western blotsby using an antibody which recognizes both Smad1 andSmad5 phosphorylated on serine residues, as described inMaterials and Methods. Unlike Smad 2 (see above), ourwestern blot analysis did not detect a significant change in thephosphorylation of Smad1 and Smad5 proteins in the livers

of alcohol-treated mice, compared to the controls (Figures6(a) and 6(c)). Since iron has been reported to induce BMPsignaling and Smad5 phosphorylation [25, 27], the livers ofmice fed with iron diets (see Materials and Methods) wereemployed as internal controls. Accordingly, our western blotanalysis detected a significant increase in Smad1 and Smad5phosphorylation in the livers of mice with iron overload,compared to control mice with a normal iron state (Figures6(a) and 6(c)). The level of total Smad5 protein expressionin the liver was not altered by alcohol or iron treatments(Figure 6(b)).

Bone morphogenetic proteins bind to and signal throughtype I and type II serine/threonine kinase receptors, BMPR-I and BMPR-II. Upon ligand binding, BMPR-I is phos-phorylated. To determine the activation of BMPR-I, weperformed immunoprecipitation experiments followed bywestern blotting, as described in Materials and Methods.BMPR-I immunocomplexes from mouse livers were blottedwith an anti-phosphoserine antibody. The level of BMPR-I phosphorylation on serine residues in alcohol-fed micewas not significantly different than that in control mice(Figures 7(a) and 7(c)). However, BMPR-I phosphorylation


Con

trol

Alc

ohol

Nor

mal

Ove

rloa

d

Ove

rloa

d

Carbonyl iron diet

BMPR-I IgG Rabbit IgG

BMPR-I(phosphoserine)

(a)

Con

trol

Alc

ohol

Nor

mal

Ove

rloa

d

Ove

rloa

d

Carbonyl iron diet

BMPR-I IgG Rabbit IgG

BMPR-I

(b)

0

1

2

3

Fold

p-B

MP

R-I

expr

essi

on

Control Alcohol Normal OverloadL. De Carli Diet Carbonyl iron diet

(c)

Figure 7: Activation of BMPR-I. Bone morphogenetic protein receptor type I (BMPR-I) was immunoprecipitated from the liver lysates ofmice fed with regular (control) or ethanol-containing (alcohol) Lieber-De Carli liquid diets, and of mice fed with diets containing 0.2%(normal) or 2% (overload) carbonyl iron, as described in Materials and Methods. The phosphorylation of BMPR-I on serine residuesin the immune complexes was determined by western blotting employing an anti-phosphoserine antibody (a), and the total levels ofBMPR-I protein were detected by western blotting with an anti-BMPR-I antibody (b). (c) Autoradiographs from different experiments(n = 2) were scanned by a densitometer, and phospho-BMPR-I (p-BMPR-I) expression in each sample was quantified by normalizingto immunoprecipitated BMPR-I protein level. Normalized p-BMPR-I expression in alcohol or high iron-fed mice was expressed as foldexpression of that in the control mice or in mice fed with normal iron diet.

was induced in the livers of mice fed with high irondiets, which were used as internal controls (Figures 7(a)and 7(c)). We also confirmed by western blotting thatequal levels of BMPR-I protein were immunoprecipitatedfrom the livers of alcohol or iron-treated and control mice(Figure 7(b)). Furthermore, control samples, which wereimmunoprecipitated with normal rabbit IgG also showed nosignificant BMPR-I phosphorylation (Figure 7(a)).

Smad4 forms a complex with phosphorylated R-Smadsand regulates the transcription of target genes. In orderto determine the effect of alcohol on Smad4 DNA-bindingactivity in the liver, we performed electromobility shiftassays, as described in Materials and Methods. The DNA-binding activity of Smad4 in liver nuclear lysates from micewith chronic alcohol exposure was not significantly differentthan that of control mice (Figure 8). However, iron, usedas internal control, induced Smad DNA-binding activity(Figure 8). The specificity of DNA-binding activity was alsoconfirmed with both competition tests, using unlabeled(cold) Smad consensus oligonucleotides, and by employing32P-labeled mutant Smad oligonucleotide as a probe in

gelshift assays, as described in Materials and Methods(Figure 8).

In order to determine the effect of chronic alcoholexposure on Smad4-mediated transcription of hepcidin, weperformed chromatin immunoprecipitation experiments, asdescribed in Materials and Methods. The binding of Smad4to hepcidin promoter was significantly attenuated in thelivers of mice treated with alcohol compared to controlmice (Figures 9(a) and 9(d)). We have also confirmed thatthe level of total input DNA (see Materials and Methods)was similar in all samples (Figure 9(b)). Furthermore, nosignificant amplification of hepcidin promoter was observedin chromatin samples, which were immunoprecipitated withthe control IgG (Figure 9(c)).

4. Discussion

Hepcidin, mainly synthesized in the liver, is the key regulatorof iron homeostasis and its expression is also regulated byiron. Alcohol has been shown to suppress hepcidin tran-scription in the liver leading to elevated iron absorption in


− − − − − − −+

Free

Free

FreeControl Alcohol Normal Overload Overload Overload

Smad

Smad cons.oligonucleotide

Carbonyl iron

mutant mutant

Figure 8: Smad4 DNA-Binding Activity. Smad DNA-binding activity was determined by electrophoretic mobility gelshift assays (EMSA),as described in Materials and Methods. 7 μg of nuclear lysate protein isolated from the livers of mice fed with regular (control) or ethanol-containing (alcohol) Lieber-De Carli liquid diets, and of mice fed with diets containing 0.2% (normal) or 2% (overload) carbonyl iron wasemployed in EMSA. Cold competition with unlabeled Smad consensus (cons.) oligonucleotides and 32P-labeled mutant Smad probe wereemployed to test the specificity of DNA binding, as described in Materials and Methods. The arrows indicate specific Smad DNA complexesand unbound (free) probes. The results are representative of multiple EMSA experiments (n = 3).

the duodenum [9–11, 13]. However, how alcohol attenuatesliver hepcidin transcription and function is not completelyunderstood.

Various signaling mechanisms including BMP-mediatedSmad signaling are involved in the regulation of hepcidintranscription in the liver [23–25, 27]. The deletion of Smad4in the liver also attenuates hepcidin expression and causespronounced hepatic iron accumulation in mice [32]. BMPsbelong to the TGF-β superfamily of growth factors. TGF-β isone of the main profibrogenic cytokines, which is involvedin the progression of alcoholic liver disease [33, 34, 38].However, the role of BMPs in alcoholic liver disease isunclear. Here, we show that alcohol significantly induces theexpression of BMP2 in the liver in vivo. Although alcoholis known to alter iron homeostasis [3, 5, 39], unlike iron[25, 27], chronic alcohol exposure did not significantlyupregulate the expression of BMP6 in the liver.

Ligand binding induces the phosphorylation of type IBMP receptor (BMPR-I). Upon phosphorylation, BMPR-I stimulates BMP signaling by phosphorylating Smad1,Smad5, and Smad8. Interestingly, despite an alcohol-inducedincrease in BMP expression, the phosphorylation of Smad1and Smad5 was not elevated in the livers of mice withchronic alcohol exposure. In contrast, TGF-β-mediatedphosphorylation of Smad2 was induced by alcohol. BMPR-I receptor is expressed on the plasma membrane. It is feasible

that alcohol-mediated inhibition of BMP-mediated Smadsignaling may occur in proximity to the cell surface. Ourimmunoprecipitation studies clearly demonstrate a lack ofBMPR-I phosphorylation in the livers of mice with chronicalcohol exposure. The specificity of immune complexes wasconfirmed by employing control antibodies and IgG light-chain-specific secondary antibodies for western blotting (seeFigure 7). Furthermore, we have also confirmed that iron, asan internal control, upregulates the phosphorylation of bothBMPR-I, and Smad1 and Smad5 proteins. Of note, we havepreviously reported that alcohol renders hepcidin insensitiveto body iron levels and abolishes its protective role in ironoverload [12]. However, whether or not alcohol interfereswith iron-mediated activation of hepcidin transcription viaBMP/Smad signaling in the liver warrants further investiga-tion.

The inhibition of BMP receptor activation and signal-ing by alcohol may involve various mechanisms. Alcoholmetabolism is well known to increase the NADH:NAD+

ratio and induce hypoxia in the liver. Accordingly, hypoxiahas recently been suggested to inhibit hepcidin expressionby attenuating Smad signaling in human Huh7 hepatomacells [40]. Furthermore, hypoxia-induced changes in theNADH : NAD+ ratio have been reported to attenuate BMPreceptor activation in lung cells [41]. However, it shouldbe noted that alcohol-induced hypoxia is limited to the


Control Alcohol

321 bp

(a)

Control Alcohol

321 bp

(b)

Control Alcohol

321 bp

(c)

1.5

1

0.5

0

Fold

smad

4bi

ndi

ng

Control Alcohol

(d)

Figure 9: Smad4 Binding to Mouse Hepcidin Promoter. Chro-matin was isolated from the livers of mice fed with regular(control) or ethanol-containing (alcohol) Lieber-De Carli liquiddiets. Chromatin immunoprecipitation was performed with anti-Smad4 antibody (a) or normal rabbit IgG (control) (c). Thecoimmunoprecipitated DNA and total input DNA (control) (b)were subjected to PCR to amplify a 321 base pair mouse hepcidinpromoter region, as described in Materials and Methods. (d)Ethidium bromide-stained agarose gels from different experiments(n = 3) were scanned to quantify the amount of Smad4 co-immunoprecipitated DNA by normalizing to input DNA.

centrilobular region of the liver [42]. It is therefore possiblethat other mechanisms besides hypoxia may be involved inalcohol-induced inhibition of BMP-mediated Smad signal-ing in the liver. For example, changes in inhibitory Smadsor a competition between R-Smads activated by TGF-β(Smad2, Smad3) and BMPs (Smad1, Smad5, and Smad8).Of note, TGF-β receptor has been reported to interact withBMPR-I and inhibit BMP-mediated Smad signaling [43].

Accordingly, we have observed the activation of Smad2,but not of Smad1 or Smad5, in the livers of mice withchronic alcohol exposure. Conversely, recombinant BMP6has been shown to inhibit TGF-β-mediated Smad signaling[44]. Nevertheless, our findings showing alcohol-inducedinhibition of Smad DNA-binding activity and the binding ofSmad4 to hepcidin promoter strongly suggest that alcoholcan directly interfere with nuclear Smad DNA complexes.The regulation of Smad signaling complexes by alcoholmay therefore be one of the mechanisms by which alcoholsuppresses hepcidin transcription in the liver in vivo.

Alcohol metabolism in the liver produces toxic metabo-lites, such as acetaldehyde and lipid peroxidation products[45, 46]. They, in turn, activate TGF-β production andlead to the secretion of extracellular matrix proteins [33].BMPs have been reported to interact with and antagonizeTGF-β blocking its profibrogenic activity [47, 48]. Byblocking TGF-β, BMPs can also modulate cell adhesion andmigration [49, 50]. It is therefore possible that the inductionof BMP expression in the liver in response to chronicalcohol exposure is associated with antifibrogenic responsemechanisms. Furthermore, the inverse effect of alcohol onTGF-β and BMP-mediated Smad signaling may be one of themechanisms involved in the progression of liver fibrosis inalcoholic liver disease.

The alcohol-induced inhibition of Smad4 binding tohepcidin promoter and suppression of hepcidin transcrip-tion in the liver is expected to gradually elevate intestinaliron uptake and iron storage in Kupffer cells. Accordingly,we have previously reported the elevation of hepatic ironlevels in rats with chronic alcohol exposure [12]. Iron andalcohol are known to act synergistically to induce liverinjury [51–53]. Interestingly, the inhibition of BMP signalinghas also been reported in Hfe knockout mice, an animalmodel for the commonest iron overload disorder, genetichemochromatosis [54, 55]. This study therefore indicates arole for Smad signaling in the regulation of iron metabolismby alcohol, which may have implications for alcoholic liverdisease and also genetic hemochromatosis in conjunctionwith alcohol.

5. Conclusions

Bone morphogenetic protein signaling has recently beenshown to induce hepcidin transcription in the liver. BMPand TGF-β both belong to the same family of growthfactors and stimulate the Smad signaling pathway. Alcoholis known to induce TGF-β expression, which plays a role inliver fibrinogenesis, whereas the effect of alcohol on BMPsignaling is unknown. Here, we show that similar to TGF-β, BMP protein expression was also upregulated in theliver. However, alcohol exerted different effects on TGF-β-mediated Smad2 activation and BMP-mediated Smad1and Smad5 activation. The inhibitory effect of alcohol onBMP-mediated Smad signaling may occur in proximity tothe cell surface by interfering with the activation of BMPreceptor type I. This subsequently resulted in the inhibitionof Smad4 binding to hepcidin promoter in the livers of


mice with chronic alcohol exposure. Collectively, thesefindings strongly suggest that the simultaneous inhibition ofBMP-mediated Smad activation and stimulation of TGF-β-mediated Smad activation by alcohol may be involved in thesuppression of liver hepcidin transcription and deregulationof iron metabolism by alcohol in vivo. Iron and alcohol actsynergistically to induce liver injury. Further understandingof the role of alcohol in Smad signaling and hepcidintranscription will help to elucidate the mechanisms of liverinjury observed in patients with alcoholic liver disease orwith genetic hemochromatosis and alcohol abuse.

Acknowledgments

The authors thank Dr. Robert G. Bennett (Omaha VAMedical Center) for providing the livers from mice treatedwith carbon tetrachloride. These studies were supported byfunds from the University of Nebraska Medical Center andR01 Grant (AA017738) to D. Harrison-Findik.

References

[1] J. B. Whitfield, G. Zhu, A. C. Heath, L. W. Powell, and N.G. Martin, “Effects of alcohol consumption on indices ofiron stores and of iron stores on alcohol intake markers,”Alcoholism: Clinical and Experimental Research, vol. 25, no. 7,pp. 1037–1045, 2001.

[2] M. G. Irving, J. W. Halliday, and L. W. Powell, “Associationbetween alcoholism and increased hepatic iron stores,” Alco-holism: Clinical and Experimental Research, vol. 12, no. 1, pp.7–13, 1988.

[3] G. N. Ioannou, J. A. Dominitz, N. S. Weiss, P. J. Heagerty,and K. V. Kowdley, “The effect of alcohol consumption onthe prevalence of iron overload, iron deficiency, and irondeficiency anemia,” Gastroenterology, vol. 126, no. 5, pp. 1293–1301, 2004.

[4] A. S. Tavill and A. M. Qadri, “Alcohol and iron,” Seminars inLiver Disease, vol. 24, no. 3, pp. 317–325, 2004.

[5] P. Duane, K. B. Raja, R. J. Simpson, and T. J. Peters, “Intestinaliron absorption in chronic alcoholics,” Alcohol and Alcoholism,vol. 27, no. 5, pp. 539–544, 1992.

[6] L. G. Valerio, T. Parks, and D. R. Petersen, “Alcohol mediatesincreases in hepatic and serum nonheme iron stores in a ratmodel for alcohol-induced liver injury,” Alcoholism: Clinicaland Experimental Research, vol. 20, no. 8, pp. 1352–1361, 1996.

[7] M. Lin, R. A. Rippe, O. Niemela, G. Brittenham, and H.Tsukamoto, “Role of iron in NF-kappa B activation andcytokine gene expression by rat hepatic macrophages,” TheAmerican Journal of Physiology, vol. 272, pp. G1355–G1364,1997.

[8] D. D. Harrison-Findik, “Is the iron regulatory hormonehepcidin a risk factor for alcoholic liver disease?” WorldJournal of Gastroenterology, vol. 15, no. 10, pp. 1186–1193,2009.

[9] D. D. Harrison-Findik, D. Schafer, E. Klein et al., “Alcoholmetabolism-mediated oxidative stress down-regulates hep-cidin transcription and leads to increased duodenal irontransporter expression,” Journal of Biological Chemistry, vol.281, no. 32, pp. 22974–22982, 2006.

[10] K. R. Bridle, T. K. Cheung, T. L. Murphy et al., “Hepcidinis down-regulated in alcoholic liver injury: implications for

the pathogenesis of alcoholic liver disease,” Alcoholism: Clinicaland Experimental Research, vol. 30, no. 1, pp. 106–112, 2006.

[11] T. Ohtake, H. Saito, Y. Hosoki et al., “Hepcidin is down-regulated in alcohol loading,” Alcoholism: Clinical and Experi-mental Research, vol. 31, no. 1, pp. S2–S8, 2007.

[12] D. D. Harrison-Findik, E. Klein, C. Crist, J. Evans, N.Timchenko, and J. Gollan, “Iron-mediated regulation of liverhepcidin expression in rats and mice is abolished by alcohol,”Hepatology, vol. 46, no. 6, pp. 1979–1985, 2007.

[13] J. M. Flanagan, H. Peng, and E. Beutler, “Effects of alcoholconsumption on iron metabolism in mice with hemochro-matosis mutations,” Alcoholism: Clinical and ExperimentalResearch, vol. 31, no. 1, pp. 138–143, 2007.

[14] C. Pigeon, G. Ilyin, B. Courselaud et al., “A new mouseliver-specific gene, encoding a protein homologous to humanantimicrobial peptide hepcidin, is overexpressed during ironoverload,” Journal of Biological Chemistry, vol. 276, no. 11, pp.7811–7819, 2001.

[15] C. H. Park, E. V. Valore, A. J. Waring, and T. Ganz, “Hepcidin,a urinary antimicrobial peptide synthesized in the liver,”Journal of Biological Chemistry, vol. 276, no. 11, pp. 7806–7810, 2001.

[16] G. Nicolas, L. Viatte, M. Bennoun, C. Beaumont, A. Kahn, andS. Vaulont, “Hepcidin, a new iron regulatory peptide,” BloodCells, Molecules & Diseases, vol. 29, no. 3, pp. 327–335, 2002.

[17] E. Nemeth, M. S. Tuttle, J. Powelson et al., “Hepcidinregulates iron efflux by binding to ferroportin and inducingits internalization,” Science, vol. 306, no. 5704, pp. 2090–2093,2004.

[18] D. M. Kingsley, “The TGF-β superfamily: new members,new receptors, and new genetic tests of function in differentorganisms,” Genes and Development, vol. 8, no. 2, pp. 133–146,1994.

[19] L. Lin, E. V. Valore, E. Nemeth, J. B. Goodnough, V. Gabayan,and T. Ganz, “Iron transferrin regulates hepcidin synthesisin primary hepatocyte culture through hemojuvelin andBMP2/4,” Blood, vol. 110, no. 6, pp. 2182–2189, 2007.

[20] J. Truksa, H. Peng, P. Lee, and E. Beutler, “Bone morpho-genetic proteins 2, 4, and 9 stimulate murine hepcidin 1expression independently of Hfe, transferrin receptor 2 (Tfr2),and IL-6,” Proceedings of the National Academy of Sciences ofthe United States of America, vol. 103, no. 27, pp. 10289–10293,2006.

[21] J. L. Babitt, F. W. Huang, Y. Xia, Y. Sidis, N. C. Andrews, and H.Y. Lin, “Modulation of bone morphogenetic protein signalingin vivo regulates systemic iron balance,” Journal of ClinicalInvestigation, vol. 117, no. 7, pp. 1933–1939, 2007.

[22] J. L. Babitt, F. W. Huang, D. M. Wrighting et al., “Bonemorphogenetic protein signaling by hemojuvelin regulateshepcidin expression,” Nature Genetics, vol. 38, no. 5, pp. 531–539, 2006.

[23] B. Andriopoulos, E. Corradini Jr., Y. Xia et al., “BMP6 is akey endogenous regulator of hepcidin expression and ironmetabolism,” Nature Genetics, vol. 41, no. 4, pp. 482–487,2009.

[24] D. Meynard, L. Kautz, V. Darnaud, F. Canonne-Hergaux, H.Coppin, and M. P. Roth, “Lack of the bone morphogeneticprotein BMP6 induces massive iron overload,” Nature Genet-ics, vol. 41, no. 4, pp. 478–481, 2009.

[25] L. Kautz, D. Meynard, A. Monnier et al., “Iron regulatesphosphorylation of Smad1/5/8 and gene expression of Bmp6,Smad7, Id1, and Atoh8 in the mouse liver,” Blood, vol. 112, no.4, pp. 1503–1509, 2008.


[26] S. Arndt, U. Maegdefrau, C. Dorn, K. Schardt, C. Hellerbrand,and A. K. Bosserhoff, “Iron-induced expression of bonemorphogenic protein 6 in intestinal cells is the main regulatorof hepatic hepcidin expression in vivo,” Gastroenterology, vol.138, no. 1, pp. 372–382, 2010.

[27] L. Kautz, C. Besson-Fournier, D. Meynard, C. Latour, and M.P. Roth, “Iron overload induces BMP6 expression in the liverbut not in the duodenum,” Haematologica, vol. 96, pp. 199–203, 2011.

[28] D. Chen, M. Zhao, and G. R. Mundy, “Bone morphogeneticproteins,” Growth Factors, vol. 22, no. 4, pp. 233–241, 2004.

[29] J. Massague, J. Seoane, and D. Wotton, “Smad transcriptionfactors,” Genes and Development, vol. 19, no. 23, pp. 2783–2810, 2005.

[30] E. Piek, C. H. Heldin, and P. T. Dijke, “Specificity, diversity,and regulation in TGF-β superfamily signaling,” FASEB Jour-nal, vol. 13, no. 15, pp. 2105–2124, 1999.

[31] C. H. Heldin, K. Miyazono, and P. Ten Dijke, “TGF-βsignalling from cell membrane to nucleus through SMADproteins,” Nature, vol. 390, no. 6659, pp. 465–471, 1997.

[32] R. H. Wang, C. Li, X. Xu et al., “A role of SMAD4 iniron metabolism through the positive regulation of hepcidinexpression,” Cell Metabolism, vol. 2, no. 6, pp. 399–409, 2005.

[33] S. V. Siegmund, S. Dooley, and D. A. Brenner, “Molecularmechanisms of alcohol-induced hepatic fibrosis,” DigestiveDiseases, vol. 23, no. 3-4, pp. 264–274, 2006.

[34] C. Meyer, N. M. Meindl-Beinker, and S. Dooley, “TGF-beta signaling in alcohol induced hepatic injury,” Frontiers inBioscience, vol. 15, pp. 740–749, 2010.

[35] P. Coyle, J. C. Philcox, and A. M. Rofe, “Metallothionein-nullmice absorb less Zn from an egg-white diet, but a similaramount from solutions, although with altered intertissue Zndistribution,” Journal of Nutrition, vol. 129, no. 2, pp. 372–379,1999.

[36] D. D. Harrison-Findik, E. Klein, J. Evans, and J. Gollan, “Reg-ulation of liver hepcidin expression by alcohol in vivo doesnot involve Kupffer cell activation or TNF-alpha signaling,”The American Journal of Physiology—Gastrointestinal and LiverPhysiology, vol. 296, pp. G112–G118, 2009.

[37] A. S. Weinmann, S. M. Bartley, T. Zhang, M. Q. Zhang, and P.J. Farnham, “Use of chromatin immunoprecipitation to clonenovel E2F target promoters,” Molecular and Cellular Biology,vol. 21, no. 20, pp. 6820–6832, 2001.

[38] K. Breitkopf, S. Haas, E. Wiercinska, M. V. Singer, and S.Dooley, “Anti-TGF-β strategies for the treatment of chronicliver disease,” Alcoholism: Clinical and Experimental Research,vol. 29, no. 11, pp. 121S–131S, 2005.

[39] D. D. Harrison-Findik, “Role of alcohol in the regulation ofiron metabolism,” World Journal of Gastroenterology, vol. 13,no. 37, pp. 4925–4930, 2007.

[40] T. B. Chaston, P. Matak, K. Pourvali, S. K. Srai, A. T.McKie, and P. A. Sharp, “Hypoxia inhibits hepcidin expressionin HuH7 hepatoma cells via decreased SMAD4 signaling,”American Journal of Physiology—Cell Physiology, vol. 300, no.4, pp. C888–C895, 2011.

[41] X. Wu, M. S. Chang, S. A. Mitsialis, and S. Kourembanas,“Hypoxia regulates bone morphogenetic protein signalingthrough C-terminal-binding protein 1,” Circulation Research,vol. 99, no. 3, pp. 240–247, 2006.

[42] S. W. French, N. C. Benson, and P. S. Sun, “Centrilobularliver necrosis induced by hypoxia in chronic ethanol-fed rats,”Hepatology, vol. 4, no. 5, pp. 912–917, 1984.

[43] N. Dumont and C. L. Arteaga, “A kinase-inactive type II TGF-beta receptor impairs BMP signaling in human breast cancercells,” Biochemical and Biophysical Research Communications,vol. 301, pp. 108–112, 2003.

[44] J. D. Yan, S. Yang, J. Zhang, and T. H. Zhu, “BMP6 reversesTGF-β1-induced changes in HK-2 cells: implications for thetreatment of renal fibrosis,” Acta Pharmacologica Sinica, vol.30, no. 7, pp. 994–1000, 2009.

[45] A. I. Cederbaum, “Role of lipid peroxidation and oxidativestress in alcohol toxicity,” Free Radical Biology and Medicine,vol. 7, no. 5, pp. 537–539, 1989.

[46] H. Tsukamoto, Y. Takei, C. J. McClain et al., “How is the liverprimed or sensitized for alcoholic liver disease?” Alcoholism:Clinical and Experimental Research, vol. 25, no. 5, pp. 171S–181S, 2001.

[47] S. Wang and R. Hirschberg, “BMP7 antagonizes TGF-β-dependent fibrogenesis in mesangial cells,” The AmericanJournal of Physiology, vol. 284, no. 5, pp. F1006–F1013, 2003.

[48] G. M. Mitu, S. Wang, and R. Hirschberg, “BMP7 is a podocytesurvival factor and rescues podocytes from diabetic injury,”The American Journal of Physiology—Renal Physiology, vol.293, no. 5, pp. F1641–F1648, 2007.

[49] E. Fransvea, U. Angelotti, S. Antonaci, and G. Giannelli,“Blocking transforming growth factor-beta up-regulates E-cadherin and reduces migration and invasion of hepatocellularcarcinoma cells,” Hepatology, vol. 47, no. 5, pp. 1557–1566,2008.

[50] S. Yang, J. Du, Z. Wang et al., “BMP-6 promotes E-cadherinexpression through repressing δEF1 in breast cancer cells,”BMC Cancer, vol. 7, article 211, 2007.

[51] H. Tsukamoto, W. Horne, S. Kamimura et al., “Experimentalliver cirrhosis induced by alcohol and iron,” Journal of ClinicalInvestigation, vol. 96, no. 1, pp. 620–630, 1995.

[52] A. I. Cederbaum, “Iron and CYP2E1-dependent oxidativestress and toxicity,” Alcohol, vol. 30, no. 2, pp. 115–120, 2003.

[53] L. M. Fletcher and L. W. Powell, “Hemochromatosis andalcoholic liver disease,” Alcohol, vol. 30, no. 2, pp. 131–136,2003.

[54] E. Corradini, C. Garuti, G. Montosi et al., “Bone morpho-genetic protein signaling is impaired in an HFE knockoutmouse model of hemochromatosis,” Gastroenterology, vol.137, no. 4, pp. 1489–1497, 2009.

[55] L. Kautz, D. Meynard, C. Besson-Fournier et al., “BMP/Smadsignaling is not enhanced in Hfe-deficient mice despiteincreased Bmp6 expression,” Blood, vol. 114, no. 12, pp. 2515–2520, 2009.


Review Article

CYP2E1 Sensitizes the Liver to LPS- and TNF α-Induced Toxicityvia Elevated Oxidative and Nitrosative Stress and Activation ofASK-1 and JNK Mitogen-Activated Kinases

Arthur I. Cederbaum, Lili Yang, Xiaodong Wang, and Defeng Wu

Department of Pharmacology and Systems Therapeutics, Mount Sinai School of Medicine, P.O. Box 1603, One Gustave L. Levy Place,New York, NY 10029, USA

Correspondence should be addressed to Arthur I. Cederbaum, [email protected]

Received 17 May 2011; Revised 10 August 2011; Accepted 10 August 2011


Copyright © 2012 Arthur I. Cederbaum et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

The mechanisms by which alcohol causes cell injury are not clear. A major mechanism is the role of lipid peroxidation andoxidative stress in alcohol toxicity. Many pathways have been suggested to play a role in how alcohol induces oxidative stress.Considerable attention has been given to alcohol elevated production of lipopolysaccharide (LPS) and TNFα and to alcoholinduction of CYP2E1. These two pathways are not exclusive of each other; however, interactions between them, have not beenextensively evaluated. Increased oxidative stress from induction of CYP2E1 sensitizes hepatocytes to LPS and TNFα toxicityand oxidants, activation of inducible nitric oxide synthase and p38 and JNK MAP kinases, and mitochondrial dysfunctionare downstream mediators of this CYP2E1-LPS/TNFα-potentiated hepatotoxicity. This paper will summarize studies showingpotentiated interactions between these two risk factors in promoting liver injury and the mechanisms involved including activationof the mitogen-activated kinase kinase kinase ASK-1. Decreasing either cytosolic or mitochondrial thioredoxin in HepG2 cellsexpressing CYP2E1 causes loss of cell viability and elevated oxidative stress via an ASK-1/JNK-dependent mechanism. Wehypothesize that similar interactions occur as a result of ethanol induction of CYP2E1 and TNFα.

1. Introduction

The ability of acute and chronic ethanol treatment to increaseproduction of reactive oxygen species and enhance peroxi-dation of lipids, protein, and DNA has been demonstratedin a variety of systems, cells, and species, including humans[1]. Despite a tremendous growth in understanding alcoholmetabolism and actions, the mechanism(s) by which alcoholcauses cell injury are still not clear. A variety of leadingmechanisms have been briefly summarized [2–4], and it islikely that many of them ultimately converge as they reflecta spectrum of the organism’s response to the myriad ofdirect and indirect actions of alcohol. A major mechanismthat is a focus of considerable research is the role of lipidperoxidation and oxidative stress in alcohol toxicity. Undercertain conditions, such as acute or chronic alcohol exposure,production of reactive oxygen species (ROS) is enhancedand/or the level or activity of antioxidants is reduced. The

resulting state, which is characterized by a disturbance in thebalance between ROS production, on one hand, and ROSremoval and repair of damaged complex molecules, on theother is called oxidative stress.

ROS have been implicated in many of the major diseasesthat plague mankind, including the toxicity of O2 itself;hyperbaric O2; ischemia-reperfusion injury; cardiovascu-lar diseases; atherosclerosis; carcinogenesis; diabetes; neu-rodegenerative diseases, including Parkinson’s disease andAlzheimer’s disease; toxicity of heavy metals, for example,iron; asbestos injury; radiation injury; vitamin deficiency;drug (e.g., redox cycling agents) toxicity; aging; inflamma-tion; smoke toxicity; emphysema; toxicity of acute andchronic ethanol treatment [2–6]. ROS can be producedfrom many systems in cells including the mitochondrialrespiratory chain [7], the cytochrome P450s [8, 9], oxida-tive enzymes such as xanthine oxidase, aldehyde oxidase,


cyclooxygenase, monoamine oxidase, and the NADPH oxi-dase complex [10], autooxidation of heme proteins suchas ferrohemoglobin or myoglobin, or biochemicals such ascatecholamines, quinones, or tetrahydrobiopterins. In ad-dition to these cellular sources of ROS, environmental sour-ces of ROS include radiation, UV light, smoke, and certaindrugs which are metabolized to radical intermediates orwhich can redox cycle. ROS are toxic to cells because theycan react with most cellular macromolecules inactiving en-zymes or denaturing proteins, causing DNA damage suchas strand breaks, base removal, or base modifications whichcan result in mutation, peroxidation of lipids which can re-sult in destruction of biological membranes and produce re-active aldehydic products such as malondialdehyde or 4-hy-droxynonenal. A variety of enzymatic and non-enzymaticmechanisms have evolved to protect cells against ROS, in-cluding the superoxide dismutases, which remove O2

−; cata-lase and the glutathione (GSH) peroxidase system which re-move H2O2; glutathione transferases which can remove re-active intermediates and lipid aldehydes, metallothioneins,heme oxygenase, thioredoxin which remove various ROS;ceruloplasmin and ferritin which help remove metals such asiron which promote oxidative reactions; nonenzymatic, low-molecular-weight antioxidants such as GSH itself, vitaminE, ascorbate (vitamin C), vitamin A, ubiquinone, uric acid,and bilirubin [11, 12]. Oxidative stress or toxicity by ROSreflects a balance between the rates of production of ROScompared to the rates of removal of ROS plus repair of dam-aged cellular macromolecules. While excess ROS can causetoxicity, macrophages and neutrophils contain an NADPHoxidase which produces ROS to destroy foreign organisms[13], and the enzyme myeloperoxidase catalyzes a reactionbetween H2O2 and chloride to produce the powerful oxidanthypochlorite (bleach) to help destroy foreign invaders. Inaddition, ROS at low concentrations, especially H2O2, maybe important in signal transduction mechanisms in cells andthus be involved in cellular physiology and metabolism [14].

Many pathways have been suggested to play a key role inhow ethanol induces “oxidative stress” [1–4]. Some of theseinclude redox state changes (decrease in the NAD+/NADHredox ratio) produced as a result of ethanol oxidation byalcohol and aldehyde dehydrogenases; production of thereactive product acetaldehyde as a consequence of ethanoloxidation by all major oxidative pathways; damage to mito-chondria which results in decreased ATP production; director membrane effects caused by hydrophobic ethanol inter-action with either phospholipids or protein components orenzymes; ethanol-induced hypoxia, especially in the peri-central zone of the liver acinus as oxygen is consumed inorder for the liver to detoxify ethanol via oxidation; etha-nol effects on the immune system and altered cytokine pro-duction; ethanol-induced increase in bacterial-derived endo-toxin with subsequent activation of Kupffer cells; ethanolinduction of CYP2E1; ethanol mobilization of iron whichresults in enhanced levels of low-molecular-weight nonhemeiron; effects on antioxidant enzymes and chemicals, particu-larly mitochondrial and cytosolic glutathione; one electronoxidation of ethanol to the 1-hydroxy ethyl radical; con-version of xanthine dehydrogenase to the xanthine oxidase

form. Again, many of these pathways are not exclusive of oneanother, and it is likely that several, indeed many, systemscontribute to the ability of ethanol to induce a state of oxi-dative stress.

2. Kupffer Cells and Alcoholic Liver Disease

Kupffer cells are stimulated by chronic ethanol treatmentto produce free radicals and cytokines, including TNFα,which plays a role in ALD [15, 16]. This stimulationis mediated by bacterial-derived endotoxin, and ALD isdecreased when gram-negative bacteria are depleted fromthe gut by treatment with lactobacillus or antibiotics [17].The TNFα receptor superfamily consists of several memberssharing a sequence homology, the death domain, locatedin the intracellular portion of the receptor. These “death”receptors, including Fas, TNF-R1, and TRAIL-R1/TRAIL-R2, are expressed in hepatocytes and when stimulated bytheir respective ligands, FasL, TNFα, or TRAIL, hepato-cyte injury can occur [18]. Lipopolysaccharide (LPS) isa component of the outer wall of gram-negative bacteriathat normally inhabit the gut. LPS penetrates the gutepithelium only in trace amounts; however, LPS absorptioncan be elevated under pathophysiological conditions such asalcoholic liver disease [19]. When LPS is released from gram-negative bacteria and enters the blood stream, the liver tightlyregulates the entry and processing of LPS by virtue of itsability to clear LPS and respond to LPS [20]. In additionto its ability to clear LPS, the liver also responds to LPSand produces cytokines. LPS directly causes liver injury bymechanisms involving inflammatory cells such as Kupffercells, and chemical mediators such as superoxide, nitricoxide, and tumor necrosis factor (TNFα) and other cytokines[21–23]. In addition, LPS potentiates liver damage inducedby hepatotoxins including ethanol [24–29]. In experimentalalcoholic liver disease, the combination of LPS and chronicethanol produce hepatic necrosis and inflammation [27–29]. Ethanol alters gut microflora, the source of LPS,and ethanol increases the permeability of the gut, thusincreasing the distribution of LPS from the gut into the portalcirculation (endotoxemia). This causes activation of Kupffercells, the resident macrophages in liver, resulting in release ofchemical mediators including cytokines and reactive oxygenspecies (ROS), and subsequently, alcoholic liver disease[30]. Destruction of Kupffer cells with gadolinium chlorideattenuated ALD [15]. A major advance was the finding thatanti-TNFα antibodies protect against ALD [16]. NADPHoxidase was identified as a key enzyme for generating ROSin Kupffer cells after ethanol treatment [31]. Moreover, inmice deficient in a subunit of NADPH oxidase, p47phox, theethanol-induced increase in ROS and TNFα and liver injurywas decreased [32]. The role of TNFα in ALD was furthervalidated by the findings that the ethanol-induced pathologywas nearly blocked in TNFα receptor1 knockout mice [33].

The transcription factor nuclear factor-kappaB (NF-κB)in Kupffer cells regulates activation of many inflammatorygenes, including TNFα. Endotoxin activates NF-κB, leadingto the hypothesis that inhibition of NF-κB in Kupffer cellswould prevent ALD [34]. Administration of an adenovirus


encoding for the IkB superrepressor to rats chronicallyinfused with ethanol blunted the ethanol-induced activationof NF-κB, TNFα production, and pathological changes. Ageneral scheme to explain these results is that chronic ethanoltreatment elevates endotoxin levels, endotoxin activatesKupffer cells to produce free radicals via NADPH oxidase, thefree radicals activate NF-κB, leading to an increase in produc-tion of TNFα, followed eventually by tissue damage [29].

3. CYP2E1

CYP2E1 metabolizes a variety of small, hydrophobic sub-strates including solvents such as chloroform and carbontetrachloride, aromatic hydrocarbons such as benzene andtoluene, alcohols such as ethanol and pentanol, aldehy-des such as acetaldehyde, halogenated anesthetics suchas enflurane and halothane, nitrosamines such as N,N-dimethylnitrosamine, and drugs such as chlorzoxazone andacetaminophen [35–41]. From a toxicological point ofview, interest in CYP2E1 revolves around the ability ofthis P450 to metabolize and activate many toxicologicallyimportant compounds such as ethanol, carbon tetrachloride,acetaminophen, benzene, halothane, and many other halo-genated substrates. Procarcinogens including nitrosaminesand azo compounds are effective substrates for CYP2E1.Toxicity by the above compounds is enhanced after inductionof CYP2E1, for example, by ethanol treatment, and toxicityis reduced by inhibitors of CYP2E1 or in CYP2E1 knockoutmice [42].

Molecular oxygen itself is likely to be the most importantsubstrate for CYP2E1 [8, 9]. CYP2E1, relative to severalother P450 enzymes, displays high NADPH oxidase activityas it appears to be poorly coupled with NADPH-cytochromeP450 reductase [43, 44]. CYP2E1 was the most efficientP450 enzyme in the initiation of NADPH-dependent lipidperoxidation in reconstituted membranes among five dif-ferent P450 forms investigated. Furthermore, anti-CYP2E1IgG inhibited microsomal NADPH oxidase activity andmicrosomal lipid peroxidation dependent on P450, butnot lipid peroxidation initiated by the action of NADPH-cytochrome P450 reductase [43]. In our laboratory, we foundthat microsomes isolated from rats fed ethanol chronicallywere about twofold to threefold more reactive in generatingsuperoxide radical and H2O2, and in the presence of ferriccomplexes, in generating hydroxyl radical and undergoinglipid peroxidation [45–48]. CYP2E1 levels were elevatedabout threefold to fivefold in liver microsomes after feedingrats the Lieber-DeCarli diet for four weeks. In all the abovesystems, the enhanced effectiveness of microsomes isolatedfrom the ethanol-fed rats was prevented by addition ofchemical inhibitors of CYP2E1 and by polyclonal antibodyraised against CYP2E1 purified from pyrazole-treated rats,confirming that the increased activity in these microsomeswas due to CYP2E1.

Since CYP2E1 can generate ROS during its catalyticcycle, and its levels are elevated by chronic treatment withethanol, CYP2E1 has been suggested as a major contributorto ethanol-induced oxidative stress, and to ethanol-inducedliver injury [49–53]. Experimentally, a decrease in CYP2E1

induction was found to be associated with a reduction inalcohol-induced liver injury [54–58]. A CYP2E1 transgenicmouse model was developed that overexpressed CYP2E1.When treated with ethanol, the CYP2E1 overexpressingmice displayed higher transaminase levels and histologicalfeatures of liver injury compared with the control mice [59].We developed an adenoviral vector which expresses hu-man CYP2E1 and showed that infection of HepG2 cells withthis adenovirus potentiated acetaminophen toxicity as com-pared to HepG2 cells infected with a LacZ expressing ade-novirus [60]. Administration of the CYP2E1 adenovirus invivo to mice elevated CYP2E1 levels and activity and pro-duced significant liver injury compared to the LacZ-infectedmice as reflected by histopathology and elevated transam-inase levels [61]. However, other studies suggested thatCYP2E1 may not play a role in alcohol liver injury basedupon studies with gadolinium chloride or CYP2E1 knockoutmice [62, 63]. Bradford et al. [64] using CYP2E1 andNADPH oxidase knockout mice concluded that CYP2E1was required for ethanol induction of oxidative stress toDNA, whereas NADPH oxidase was required for ethanol-in-duced liver injury. As mentioned earlier, it is likely thatseveral mechanisms contribute to alcohol-induced liverinjury and that ethanol-induced oxidative stress is likely toarise from several sources, including CYP2E1, mitochondria,and activated Kupffer cells.

4. LPS/TNFα-CYP2E1 Interactions

As discussed above, abnormal cytokine metabolism is amajor feature of alcoholic liver disease. Rats chronically fedethanol were more sensitive to the hepatotoxic effects ofadministration of LPS and had higher plasma levels of TNFαthan control rats [65, 66]. In the intragastric model of chron-ic ethanol administration, the development of liver injurycoincided with an increase in TNFα, associated with anincrease in serum LPS [29]. Anti-TNFα antibody preventedalcohol liver injury in rats [16], and mice lacking the TNFR1receptor did not develop alcohol liver injury [33]. Taken asa whole, these and other studies clearly implicate TNFα asa major risk factor for the development of alcoholic liverinjury. One complication in this central role for TNFα is thathepatocytes are normally resistant to TNFα-induced toxicity.This led to the hypothesis that besides elevating TNFα, alco-hol somehow sensitizes or primes the liver to becomesusceptible to TNFα [67, 68]. Known factors which sensitizethe liver to TNFα are inhibitors of mRNA or proteinsynthesis, which likely prevent the synthesis of protectivefactors, inhibition of NF-κB activation in hepatocytes tolower synthesis of such protective factors, depletion ofGSH, especially mitochondrial GSH, lowering of S-adenosylmethionine (SAM) coupled to elevation of S-adenosyl hom-ocysteine (SAH), that is, a decline in the SAM/SAH ratio,or inhibition of the proteasome. Combined treatment withethanol plus TNFα is more toxic to hepatocytes and HepG2E47 cells which express high levels of CYP2E1 than controlhepatocytes with lower levels of CYP2E1 or HepG2 C34cells which do not express CYP2E1 [69]. RALA hepatocyteswith increased expression of CYP2E1 were sensitized to


0100200300400500

Control Pyrazole LPS Pyrazole + LPS

ALT

(un

its/

L)

0

200

400

600

800

1000

AST

(un

its/

L)

500

1000

1500

2000

Saline P + L C + P + L

ALTAST

# #

(a) (b) (c)

0

Control Pyrazole LPS Pyrazole + LPS

(a)

(a)

(b)

(b)

(c)

(c)

(a)

(b)

(c) (e)

(d)

(A) (B)

(C) (D)

ALT

/AST

(un

its/

L)

Figure 1: Effect of pyrazole or LPS or LPS plus pyrazole on serum ALT (a) or AST (b) or liver histopathology (c). In (c), panels refer to (A)saline; (B) pyrazole-treated; (C) LPS-treated; (D) LPS plus pyrazole treated. Arrows show necrotic foci with inflammatory cell infiltration.Note: combined treatment with pyrazole plus TNFα produces liver injury. The CYP2E1 inhibitor, chlormethiazole (CMZ), protects againstLPS plus pyrazole toxicity in mice. (d) ALT/AST levels. Sal: Saline-treated; P + L: pyrazole plus LPS treated: C + P + L, CMZ plus pyrazole plusLPS treated. (e) histopathology (top panels); 3-nitrotyrosine protein adducts (middle panels); 4-hydroxynonenal adducts (bottom panels).In all panels, panel (a) is the saline treated; panel (b) is the LPS plus pyrazole treated; panel (c) refers to the CMZ plus LPS plus pyrazoletreated.

TNFα-mediated cell death [70]. These results suggest thatincreased oxidative stress from CYP2E1 may sensitize isolat-ed hepatocytes to TNFα-induced toxicity.

Either LPS or CYP2E1 is considered independent riskfactors involved in alcoholic liver disease, but mutual rela-tionships or interactions between them are unknown. Weinitiated studies to evaluate whether CYP2E1 contributesor potentiates LPS- or TNFα-mediated liver injury in vivo.These studies may provide an experimental model to betterunderstand mechanisms of ethanol-induced liver damage.

5. Pyrazole Potentiates LPS Toxicity [71, 72]

Male, Sprague-Dawley rats (160–180 g) were injected intra-peritoneally with pyrazole (PY), 200 mg per kg body wt,once a day for 2 days to induce CYP2E1. After an overnightfast, either saline or LPS (Sigma, serotype 055: BS, 10 mg/kg

body wt) was injected via the tail vein. Rats were killed 8–10 hr after the LPS or saline injection and blood and livertissue collected. Neither pyrazole alone or LPS alone causedliver injury as reflected by transaminase (ALT, AST) levelsor liver histopathology (Figures 1(a) and 1(b)). However,the combination of LPS plus pyrazole increased AST andALT levels about fourfold over the levels in the pyrazolealone or LPS alone groups (Figures 1(a) and 1(b)). LPS-plus-pyrazole-treatment induced extensive necrosis of hepato-cytes, mainly located both in periportal and pericentral zonesof the liver, accompanied by strong infiltration of inflam-matory cells (Figure 1(c)). LPS alone treatment caused someapoptosis and activation of caspases 3 and 9, whereas pyra-zole treatment alone had no effect. LPS plus pyrazoletreatment was not any more effective than LPS alone in in-creasing apoptosis, unlike the increases in necrosis and in-flammation.


To assess whether oxidative stress occurs after the varioustreatments, malondialdehyde (MDA) levels as a reflectionof lipid peroxidation were assayed. Whereas pyrazole aloneor LPS alone did not elevate MDA levels over those foundwith saline controls, the combination of LPS plus pyrazoleincreased MDA levels about 65% (P < 0.05 compared tothe other 3 groups). Protein carbonyl formation as a markerfor oxidized protein formation was determined. Low levelsof protein carbonyls were found in saline control livers.Treatment with either LPS alone or pyrazole alone elevatedprotein carbonyl levels; however, striking increases in proteincarbonyls were found in the combined LPS plus pyrazolegroup. In situ detection of superoxide was measured usingthe oxidation-dependent fluorescent dye dihydroethidium.Red fluorescence was weak in saline control livers, wasslightly increased in either the LPS or pyrazole livers, and washighest in the LPS plus pyrazole livers. 3-Nitrotyrosine (3-NT) protein adducts were detected by a slot blot technique.3-NT adducts were highest in livers from the LPS-plus-pyrazole-treated mice. Thus, several parameters of oxida-tive/nitrosative stress were elevated in livers from the LPSplus pyrazole-treated mice.

CYP2E1 catalytic activity (oxidation of P-nitrophenol top-nitrocatechol) was increased about 2-fold by either thepyrazole alone or the pyrazole plus LPS treatments. LPSalone slightly but not significantly decreased CYP2E1 activi-ty. Levels of CYP2E1 protein, measured by immunoblot anal-ysis, showed similar trends, being increased about 2-fold bypyrazole or pyrazole plus LPS treatments. These results showthat pyrazole treatment enhanced LPS-induced necrosis,not apoptosis. This enhanced liver injury is associated withelevated levels of CYP2E1 and increased oxidative/nitrosativestress generated by the combination of LPS plus elevatedCYP2E1.

To validate the role of CYP2E1 in the potentiation ofLPS toxicity by pyrazole, experiments with chlormethiazole(CMZ) an inhibitor of CYP2E1 and with CYP2E1 knockoutmice were carried out [71]. C57BL/6 mice were injectedintraperitoneally with pyrazole, 150 mg/kg body wt once aday for 2 days or 0.9% saline. After an overnight fast, LPS,4 mg/kg body wt, or saline was injected IP. CMZ was inject-ed in some mice at a concentration of 50 mg/kg body wt15 hours before and 30 minutes after the LPS treatment.Mice were killed 3, 8, or 24 h after LPS or saline injection. Inother experiments, CYP2E1 knockout mice, kindly providedby Dr. Frank Gonzalez, NCI, NIH, and their genetic back-ground SV129 controls were treated with pyrazole and LPSas above. Initial experiments showed that neither pyra-zole alone nor LPS alone produced liver injury under thoseconditions. However, the LPS-plus-pyrazole-treatment pro-duced significant liver injury in mice, as was previouslyshown in rats. Little injury occurred at 3 or 8 hr afterthe LPS administration, but did occur at 24 h. The injuryin the LPS-plus-pyrazole-treated mice was associated withan elevation in oxidative/nitrosative stress as reflected byincreases in 3-NT and 4-hydroxynonenal (HNE) protein ad-ducts. Administration of CMZ to the LPS-plus-pyrazole-treated mice decreased the elevated ALT and AST levels byabout 55 and 65%, respectively, (Figure 1(d)). Pathological

evaluation showed large necrotic areas in the livers fromthe LPS-plus-pyrazole-treated-mice, but only small necroticfoci were observed after treatment with CMZ (Figure 1(e)).The treatment with CMZ also lowered the elevated oxida-tive/nitrosative stress produced by the LPS plus pyrazoletreatment as only weak signals for formation of 4-HNEadducts and 3-NT adducts were found after the CMZ treat-ment (Figure 1(e)). The pyrazole plus LPS treatment pro-duced a 2-fold increase in CYP2E1 catalytic activity, whichwas prevented after the administration of CMZ. Thus, CMZblocked the elevation of CYP2E1 in the LPS-plus-pyrazole-treated mice, and this was associated with a decline inoxidative/nitrosative stress and blunting of liver injury.

To further evaluate a role for CYP2E1 in the LPS pluspyrazole toxicity, CYP2E1 knockout or wild-type con-trol SV129 mice were treated with LPS plus pyrazole. Aswith C57Bl/6 mice, liver injury was observed in the wild-type SV129 mice treated with LPS plus pyrazole, but notmice treated with LPS alone or pyrazole alone. Serum ALTand AST levels were about 50% lower in LPS-plus-pyra-zole-treated CYP2E1 knockout mice as compared to wild-type mice. Pathological evaluation showed large necroticareas and widespread necrotic foci in wild-type mice, where-as almost normal histology was found in the LPS-plus-pyrazole-treated CYP2E1 knockout mice. Positive TUNELstaining was also significantly lower in the CYP2E1 nullmice compared to wild-type mice. Immunoblots confirmedthe absence of CYP2E1 protein in the knockout mice, whilestrong signals from CYP2E1 were detected in immunoblotsof the wild type mice. Thus, in both rats and mice,the CYP2E1 inducer pyrazole potentiates LPS-inducedliver injury. This potentiation is associated with elevatedoxidative/nitrosative stress and is blocked by the CYP2E1inhibitor CMZ and blunted in CYP2E1 knockout mice. Wehypothesize that CYP2E1-mediated oxidative stress maysynergize with LPS-generated oxidative stress in this modelto produce liver injury.

6. Pyrazole Potentiates TNFα Toxicity [73, 74].

Since TNFα levels are elevated after LPS administration andTNFα plays an important role in the effects of LPS, wedetermined if pyrazole treatment to induce CYP2E1 po-tentiates TNFα toxicity as it did with LPS toxicity. Basically,the same approaches described above were used, with injec-tion of TNFα (50 ug/kg body wt.) replacing the LPS treat-ment.

Figure 2(a) shows that ALT and AST levels were low inthe saline control mice and in the pyrazole-treated micechallenged with saline. Treatment of control mice with TNFαelevated transaminase levels by about 2-3-fold. Treatment ofthe pyrazole mice with TNFα elevated transaminase levelsmore than 3-fold over the TNFα-saline control treated mice.Liver sections were stained with H&E for morphologicalevaluation. The saline and TNFα treated mice showednormal liver morphology. Liver from pyrazole treatedmice showed some vacuolar degeneration. Liver from theTNFα-plus-pyrazole-treated mice showed several necroticloci (arrows), and typical pathology morphology changes


0

100

200

300

400

500

600

700

800

900

Control PY TNF TNF + PY

ALTAST

ALT

/AST

(un

its/

L)

(a)

Saline TNF

PY TNF + PY

(b)

0

2

4

6

8

10

12

14

TB

AR

S(n

mol

es/m

g)

LyateMito

Saline PY TNF TNF+ PY

(c)

0

50

100

150

200

250

300

350

400

ALT AST

KOWT

ALT

/AST

(un

its/

L)

(d)

(A) (B)

(e)

Figure 2: Pyrazole potentiates TNFα hepatotoxicity and oxidative stress in mice. Mice were treated with either saline or pyrazole alone orTNFα alone or pyrazole plus TNFα followed by assays for (a) serum ALT/AST, (b) histopathology (arrows show necrotic zones), (c) lipidperoxidation as reflected by levels of TBARs in liver cell lysates and in isolated mitochondrial fractions. Note: combined treatment with TNFαplus pyrazole produces liver injury. (d) Serum ALT and AST levels in pyrazole plus TNFα-treated wild type (WT) and CYP2E1 knockout(KO) mice. (e) Histopathology in pyrazole plus TNFα-treated KO (panel a) and WT (panel b). Note: liver injury is decreased in CYP2E1knockout mice compared to WT mice.

including nuclear pyknosis, karyorrhexis, and karyolysiswere observed (Figure 2(b)). The treatment with pyrazoledid not significantly alter the levels of thiobarbituric acid-reactive substrates (TBARS) in the total liver extract or themitochondria (Figure 2(c)). TNFα treatment of control miceelevated levels of TBARS about 2-3 fold. TBARS in thehomogenates and the mitochondria were further elevatedwhen TNFα was administered to the pyrazole-treated mice.Highest liver and mitochondrial TBARs levels were observedin the pyrazole-plus-TNFα-treated mice (Figure 2(c)). LiverGSH levels were similar in the saline, pyrazole-treated, andTNFα-treated mice but were decreased about 40% in the liverextracts from the pyrazole-plus-TNFα-treated mice. GSHlevels were lowered 40% in the liver mitochondria from thepyrazole plus TNFα-treated mice compared to the TNFαalone treated mice. These results suggest that the combinedpyrazole plus TNFα treatment produces elevated oxidative

stress in the liver compared to TNFα alone or pyrazole alone,and that mitochondrial oxidative stress may occur in livers ofthe pyrazole-plus-TNFα-treated mice.

As expected, CYP2E1 activity as reflected by the NADPH-dependent microsomal oxidation of p-nitrophenol and thecontent of CYP2E1 (Western blot analysis) were elevated 2-to 3-fold by pyrazole or by pyrazole plus TNFα treatment,over the saline or TNFα alone treated mice. Thus, TNFαalone or in combination with pyrazole did not alter CYP2E1activity or content. Also, induction of CYP2E1 alone bypyrazole is not sufficient to induce liver injury; rather, a sec-ond “hit,” for example, TNFα is required. What is the evi-dence that induction of CYP2E1 by pyrazole is important forthe elevated injury found in the pyrazole-plus-TNFα-treatedmice? We used CYP2E1 knockout mice to address this ques-tion. Large increases in ALT and AST levels were found afterTNFα administration to pyrazole-treated SV129 wild type


SalineTNF

TNF + PY + CsA

PY

TNF + PY

OD

540

nm

CaCl2

0.16

0.14

0.12

0.1

0.08

0.06

0.04

0.02

0Time (s )

(a)

Rel

ativ

efl

uor

esce

nce

Succinate

ADP

TNF + PY

PY

Saline

TNF

0 30 60 90 120 150 180 210 240 270 300

200

190

180

170

160

150

140

130

120

Time (s )

(b)

Figure 3: Treatment with TNFα plus pyrazole causes mitochondrial injury. TNFα plus pyrazole causes mitochondrial injury as reflectedby (a) increased mitochondrial swelling (decreased absorbance at 540 nm) or (b) decreased mitochondrial membrane potential as assayedby succinate-dependent decline in rhodamine 123 fluorescence. The decline in fluorescence after addition of succinate is reflective of themitochondrial membrane potential. Note: this decline is very small in the mitochondria isolated from TNFα plus pyrazole-treated mice.In (a) the increased swelling produced by TNFα plus pyrazole is prevented by cyclosporine A (CsA), an inhibitor of the mitochondrialpermeability transition.

mice. TNFα treatment of pyrazole-treated CYP2E1 knockoutmice did not elevate transaminase levels (Figure 2(d)).Similarly, TBARs levels in liver homogenates and isolatedmitochondria were not elevated in the TNFα plus pyrazole-treated CYP2E1 knockout mice but were increased in thewild-type mice. Normal liver pathology was observed afterpyrazole plus TNFα treatment of CYP2E1 knockout mice.(Figure 2(e)). The failure of TNFα to induce liver injury inpyrazole-treated CYP2E1 knockout mice supports a criticalrole for CYP2E1 in the potentiated injury observed in thewild-type mice.

7. Mitochondrial Dysfunction

Alcohol can cause mitochondrial dysfunction [75, 76]. Wehypothesized that mitochondria are an eventual target fordeveloping liver injury induced by TNFα when CYP2E1 iselevated by pyrazole. Initiation of a mitochondrial perme-ability transition was determined by assessing mitochondrialswelling in the absence and presence of 100 μM calcium. Suc-cinate (10 mM) was the respiratory substrate. As shown inFigure 3(a), in the absence of calcium, swelling (decrease inabsorbance at 540 nm) was low with all mitochondrial prepa-rations although there was some basal swelling with themitochondria from the pyrazole plus TNFα-treated mice.The addition of 100 μM calcium caused a low rate of swellingin the saline or TNFα alone mitochondria; swelling was

somewhat elevated in the pyrazole alone mitochondria.Swelling was very rapid without any lag phase with themitochondria from the pyrazole-plus-TNFα-treated mice(Figure 3(a)). Importantly, this rapid swelling was blockedby cyclosporine A (2 μM), a classic inhibitor of the mito-chondrial permeability transition. Calcium elevates mito-chondrial swelling in the saline-, TNFα alone-, and pyrazolealone groups, which was most pronounced in the TNFα pluspyrazole group. The calcium-induced swelling was sensitiveto cyclosporine A in all groups. The basal swelling, in theabsence of added calcium, was also higher in the TNFαplus pyrazole group, further suggestive of mitochondrial dys-function.

The electrochemical potential of the proton gradient gen-erated across the mitochondrial membrane (ΔΨ) was as-sessed by monitoring fluorescence quenching of rhodamine123. Addition of 10 mM succinate at one minute causeda decrease in fluorescence reflective of a high ΔΨ corre-sponding to state 4 of respiration (Figure 3(b)). The declinein fluorescence averaged about 40 arbitrary units per minutewith mitochondria from the saline or TNFα alone treatedmice and 30 arbitrary units per minute with mitochondriafrom the pyrazole-treated mice. However, the decline in fluo-rescence was only about 14 arbitrary units with mitochon-dria from the TNFα-plus-pyrazole-treated mice. Addition ofADP at 3 minutes caused an enhancement of fluorescencewhich corresponds to state 3 respiration as part of the proton


motive force is utilized to synthesize ATP. This enhancementof fluorescence averaged 15, 14, 12, and 4 arbitrary unitsper minute for mitochondria from the saline, TNFα alone,pyrazole alone, and TNFα plus pyrazole treated mice, respec-tively. Taken as a whole, these initial data suggest a smalldecline in ΔΨ in mitochondria from the pyrazole-treatedmice and a more pronounced decline in mitochondria fromthe pyrazole-plus-TNFα-treated mice.

8. Cyclosporine A (CsA) Prevents PyrazolePlus LPS-Induced Liver Injury [77]

We evaluated whether cyclosporine A (CsA), an inhibitorof the mitochondrial permeability transition, could protectagainst the TNFα-plus-pyrazole-induced liver injury. Suchan experiment could validate that mitochondrial dysfunctionis a key downstream target in this injury. Male C57BL/6 micewere treated with saline, pyrazole, LPS, or pyrazole plusLPS plus corn oil or pyrazole plus LPS plus 1 dose of CsA(100 mg/kg body wt, dissolved in corn oil). Serum ALT andAST levels were elevated in the PY + LPS + corn oil groupcompared to the other 3 groups. CsA treatment attenuatedthis increase in transaminases. H&E staining of liver tissueshowed that the PY + LPS + corn oil treatment inducedextensive liver zonal necrosis and that the CsA treatment pre-vented this. Mitochondrial swelling was increased in mito-chondria isolated from the PY + LPS + corn oil treated micecompared to mitochondria from the saline + corn oil mice.The in vivo treatment with CsA prevented this increase inmitochondrial swelling, which likely explains the protectionagainst LPS-plus-pyrazole-induced liver injury. The LPSplus pyrazole elevation of 4-HNE and 3-NT protein adductswere also decreased by CsA, suggesting that mitochondrialdysfunction plays an important role in the increase inoxidative/nitrosative stress.

9. Activation of MAP Kinases

Mitogen-activated protein kinases (MAPKs) are serine-thre-onine kinases that mediate intracellular signaling associatedwith a variety of cellular activities including cell proliferation,differentiation, survival, death, and transformation. Themammalian MAPK family consists of extracellular signal-regulated kinase (ERK), p38 MAPK, and c-Jun NH2-termi-nal kinase (JNK; also known as stress-activated protein kin-ase or SAPK) [78]. The MAPK signaling cascade consistsof three distinct members of the protein kinase family, in-cluding MAP kinase (MAPK), MAPK kinase (MAPKK), andMAPKK kinase (MAPKKK). MAPKKK phosphorylates andthereby activates MAPKK, and the activated form of MAPKKin turn phosphorylates and activates MAPK. ActivatedMAPK may translocate to the cell nucleus and regulate theactivities of transcription factors and thereby control geneexpression [79, 80]. In either in vivo or in vitro modelsof alcoholic liver disease, an increase of gene expression ofthe MAPK pathway was found [81, 82]. Compatible data inprotein expression levels were seen in many studies. Intra-peritoneal injection of alcohol to rats induced rapid phos-phorylation of p38 MAPK, and JNK after only 1 hr of ethanol

injection, and this was accompanied with apoptosis of theliver [83]. In human stellate cells, increased phosphorylationof p38 MAPK and JNK was found to be associated withethanol-induced stellate cell activation, toxicity, and apop-tosis [84]. JNK and p38 MAPK may become activated simul-taneously, while some studies have shown that JNK and p38MAPK may even react in the opposite way according to thespecific treatments. In one study, after chronic alcohol feed-ing, LPS stimulation of Kupffer cells increased p38 MAPKactivity, whereas it decreased JNK activity [85]. In humanmonocytes, acute alcohol exposure increased JNK phosphor-ylation, while chronic alcohol exposure decreased JNK ac-tivity [86]. Apparently, further studies are needed to clarifywhy MAPK can react differently depending on the stimuli orin different cell lines.

MAP kinases such as JNK or p38 MAPK have beenshown to play important roles in several models of liverinjury, including CYP2E1-dependent toxicity [69, 70, 87–92]. We evaluated possible activation of MAP kinases inour pyrazole/LPS or pyrazole/TNFα hepatotoxicity modelsby assaying for the phosphorylated MAPK. As shown inFigure 4(a), LPS treatment alone did not cause significantJNK activation or p38 MAPK activation as reflected by thelow p-JNK and pp38 MAPK levels relative to total JNKand p38 MAPK levels. Similar low ratios were found forthe saline or the pyrazole alone treated mice (Figure 4(a)).However, both JNK and p38 MAPK were activated in liversof the pyrazole plus LPS-treated mice. A similar activation ofJNK and p38 MAPK was observed after pyrazole plus TNFαbut not in mice treated with TNFα or pyrazole alone [73].ERK was not altered by TNFα alone or pyrazole plus TNFαtreatment. To evaluate the significance of these changes inMAPK activation, the effect of SP600125, an inhibitor ofJNK, and SB203580, an inhibitor of p38MAPK, on the hepat-otoxicity was determined. The TNFα plus pyrazole elevationof transaminases was blunted by administration of SP600125(15 mg/kg) or SB203580 (15 mg/kg) (Figure 4(b)). TheMAPK inhibitors also lowered the necrosis (Figure 4(c))and partially blocked the increased oxidative stress producedby the pyrazole plus TNFα treatment, but had no effect onCYP2E1 activity or protein levels. These results suggest theCYP2E1 elevation of TNFα liver injury and oxidative stressis MAPK dependent. The activation of JNK in the pyra-zole plus TNFα group was blocked by SP600125 but notSB203580 whereas the activation of p38 MAPK was blockedby SB203580 but not SP600125.

10. Activation of ASK-1 and DownstreamMap Kinase Kinases

The upstream mediators of JNK and p38 MAPK activationwere not identified in these previous studies. For mechanisticand therapeutic implications, it would be important toevaluate the MAP kinase kinase kinase and MAP kinasekinase which activate JNK and p38 MAPK in this PY plusTNFαmodel. Apoptosis signal-regulating kinase 1 (ASK-1) isa member of the MAP3K family which is responsive to stress-induced cell damage. Activation of ASK-1 can determinecell fate by regulation of both the MKK4/MKK7-JNK and


1 2 3 4 5 6 7 8

Saline PY LPS LPS + PY

pJNK

JNK

pJNK1/JNK1: 0.9 1.2 1.4 3.4

pp38

pp38/ : 1 0.8 0.9 2.8

1 2 3 4 5 6 7 8

p38

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

p38

(a)

0200400600800

1000120014001600

ALT

,AST

(un

its/

L)

ALTAST

#

##

Salin

e

PY

TN

F

TN

F+

PY

TN

F+

PY

+SB

TN

F+

PY

+SP

(b)

Saline

TNF

TNFPYSB

PY

TNFPY

TNFPYSP

(c)

Figure 4: MAP kinase activation. (a) LPS plus pyrazole treatment activates JNK and p38 MAPK. The pJNK/JNK and the pp38MAPK/p38MAPK ratios are shown below the blots. (b) Either the JNK inhibitor SP600125 (SP) or the p38 MAPK inhibitor SB203508(SB) prevents TNFα plus pyrazole-induced elevation of ALT and AST or (c) liver pathology.

the MKK3/MKK6-p38 MAPK signaling cascades [93]. ASK-1 is activated by oxidative stress, ER stress, and inflammatorycytokines such as TNFα [94]. In resting cells, ASK-1 formsan inactive complex with reduced thioredoxin (Trx). Underconditions of stress by TNFα or ROS, ASK-1 dissociates fromTrx and becomes activated [95] (Figure 5). Oxidation of Trxby ROS causes dissociation of ASK-1 from the oxidized Trxwhich switches the inactive form of ASK-1 to the activekinase. The Trx-ASK complex is thought to be a redox sensor,which functions as a molecular switch turning the cellularredox state into a MAP kinase signaling pathway [96]. Ac-tivated ASK-1 then promotes activation (phosphorylation)of the downstream MAPKK, MKK4/MKK7 which can ac-tivate JNK, and MKK3/MKK6 which can activate p38 MAPK[93–96] (Figure 5). We evaluated whether CYP2E1 plus-TNFα-induced ROS promote release of ASK-1 from the Trx-ASK1 complex and activate ASK-1 followed by the phos-phorylation of MKK4/MKK7 and/or MKK3/MKK6 whichsubsequently regulate the phosphorylation of JNK and p38MAPK and contribute to the liver injury.

Wild-type mice treated with PY plus TNFα developedliver injury between 8 and 12 h after TNFα administration asreflected by the high levels of ALT and AST at 12 h. Oxidativestress is a likely key factor to trigger signaling and liver injuryin CYP2E1-mediated hepatotoxicity [97]. A time course foroxidative stress after PY plus TNFα treatment was studied.GSH was decreased in wild-type mice after 4 h and remainedat lower levels for at least 12 h as compared to the TNFαalone group. Lipid peroxidation increased significantly at 4 hin the PY-plus-TNFα-treated mice and remained elevated upto 12 h. These results show that oxidative stress occurs atan earlier time after administration of TNFα than does liverinjury in the TNFα-plus-PY-treated mice. Treatment with PYincreased the levels of CYP2E1 prior to the administration ofTNFα, and CYP2E1 levels remained about 2-fold elevated atleast until 8 h after administration of TNFα in the PY-treatedmice.

Since previous results showed a key role of JNK andp38 MAPK in the TNFα-plus-PY-induced liver injury, weevaluated whether upstream MAPKK and MAPKKK were


ASK-1-Trx ASK-1 MKK 7/4 + MKK 3/6

JNKp38 MAPK

pp38 MAPKpJNK

ROS/RNS inhibit MKPS

(Inactive)

(Active)

(Active)

ROS/RNS

Figure 5: Scheme for activation of ASK-1 by ROS/RNS anddownstream MAP kinase kinases (MKK4/7 and MKK 3/6) andMAP kinase JNK and p38 MAPK by ROS/RNS. Dissociation ofthe inhibitory thioredoxin (TRX) from the TRX-ASK-1 complex byROS/RNS activates ASK-1. MKPS and MAP kinase phosphataseswhich deactivate activated JNK and p38 MAPK by dephosphoryla-tion are inhibited by ROS/RNS thereby sustaining the activation ofJNK and p38 MAPK.

activated, the time course for their activation in relation tothe hepatic injury, and the role of CYP2E1. We focused onASK-1 since this MAPKKK has been shown to be importantas a target for TNFα signaling [93, 94, 96]. TNFα orpretreatment with PY alone did not activate ASK-1. TNFαplus PY treatment activated ASK-1 3-fold compared withthe 0 hour control at 4 h after TNFα treatment. Activationof ASK-1 decreased at 8 and 12 h. Immunoprecipitationexperiments showed that ASK-1 was bound to Trx-1 at 0 hbut was released from the Trx-ASK1 complex at 4 h andremained free from binding to Trx1 at 8 and 12 h. NoASK-1 release from the Trx-ASK1 complex was found inTNFα alone treated mice. ASK-1 was not activated in PY-plus-TNFα treated CYP2E1−/− mice, and no ASK-1 wasreleased from the Trx-ASK1 complex in CYP2E1−/− mice.Thus, activation of ASK-1 by treatment with TNFα plus PYis associated with its release from the Trx-ASK1 complex,occurs prior to the liver injury, and requires CYP2E1.

MKK4/7 and MKK3/6 are the MAPKK which activatedownstream JNK or p38 MAPK, respectively, [98]. They arealso targets for activation by ASK-1 [99, 100]. Treatment ofwild-type mice with PY plus TNFα activated MKK4 at 4,8, and 12 h compared with the TNFα alone groups [101].No activation of MKK4 was found in TNFα or TNFα + PYtreated CYP2E1−/− mice [92]. MKK7 was activated onlyat 12 h. MKK3 was activated as early as 4 h in the TNFα-plus-PY-treated mice, while MKK6 was activated at 8 h. JNKwas activated in the TNFα + PY mice at 8 and 12 h, andp38 MAPK was activated at 12 h when compared with TNFαalone. In CYP2E1−/− mice, neither MKK4/7, MKK3/6,JNK, nor p38 MAPK was activated. Thus, the time courseexperiments suggest MKK4 may be the MAPK responsiblefor activation of JNK, while either MKK3 or MKK6 may bethe MAPKK responsible for the activation of p38 MAPK.

In summary, a time course of in vivo liver injury inducedby PY plus TNFα was carried out to determine the sequence

of events and relationships between induction of CYP2E1,oxidative stress, the activation of ASK-1, MKK3/MKK6,MKK4/MKK7, p38 MAPK and JNK with the developmentof liver injury [101]. The liver injury occurs at 8 to 12 h afterthe addition of TNFα. Since ROS is postulated to be a criticalfactor in the mechanism by which TNFα plus PY induce liverinjury, development of ROS should precede the liver injury.Indeed, hepatic GSH levels were decreased and TBARS levelswere elevated 4h after administration of TNFα to PY-treatedmice. Thus oxidative stress precedes the liver injury. A like-ly contributor to the increase in oxidative stress is the induc-tion of CYP2E1 by the pyrazole treatment as no injury oroxidative stress was observed in CYP2E1 knockout mice.CYP2E1 levels were already elevated at the time of TNFαadministration (0 h) since the mice were treated for two daysprior to this injection of TNFα on day 3. ASK-1, a member ofthe MAPKKK family, activates both MKK4/MKK7-JNK andMKK3/MKK6-p38 MAPK signaling cascades. ASK-1 wasactivated in PY-treated mice at 4 h after the administrationof TNFα. Immunoprecipitation analysis showed that ASK-1 was dissociated from the inactive Trx-ASK complex at4 h, consistent with the activation of ASK-1 at 4 h. InCYP2E1−/− mice, pyrazole plus TNFα treatment failed toactivate ASK-1 and ASK-1 was not dissociated from the Trx-ASK1 complex. If CYP2E1-generated ROS is important forthe release and activation of ASK-1, elevation of CYP2E1and in oxidative stress should occur as early events. Increasesin CYP2E1 and ROS occur at 4 h, at least consistent withthe activation of ASK-1 at 4 h, although future experimentswith shorter time intervals will be necessary to evaluate theserelationships in more detail. Our results implicate a role forASK-1 in CYP2E1 potentiation of TNFα-induced liver injury.Future experiments with ASK-1 knockout mice [102] wouldbe interesting to further validate the role of ASK-1 in thePY/TNFα model. JNK or p38 MAPK activities are increasedupon phosphorylation by MAPK kinase (MKK4/MKK7or MKK3/MKK6) [98]. The activity of ASK-1 modulatesand regulates the phosphorylation of MKK4/MKK7 andMKK3/MKK6. PY plus TNFα treatment increased MKK4phosphorylation at 4, 8, and 12 h, while activation of MKK7was delayed until 12 h. MKK3 and MKK6 phosphorylationswere also increased at 4 to 8 h. In CYP2E1−/− mice, noMAPKK was activated at any observation time point. TNFαalone did not significantly activate the MAPKK in wild-typeor CYP2E1−/− mice. The activation of MKK4 and MKK3/6(4–8 h) occur prior to the onset of liver injury (8–12 h).

The role of CYP2E1 in the activation of ASK-1, MKK4/MKK7 or MKK3/MKK6 is apparent, since TNFα treatmentonly induced such activations in wild type mice treated withPY to induce CYP2E1 but not in CYP2E1−/− mice. We hy-pothesize that TNFα alone- or CYP2E1 alone-generated ROSstress is not sufficient to trigger the dissociation of ASK-1 from the Trx-ASK complex. The CYP2E1 sensitization ofTNFα-induced liver injury may occur through a synergisticeffect with TNFα to produce an enhanced ROS stress con-sistent with the so-called “Two Hit” hypothesis. We speculatethat similar interactions between CYP2E1 and TNFα may beimportant for alcohol-induced liver injury.


11. Thioredoxin-CYP2E1-ASK-1-JNK1Interactions

The thioredoxin system plays a key role in modulating redoxsignaling pathways which regulate physiological as well aspathophysiological processes [103, 104]. The thioredoxinsystem includes thioredoxin, thioredoxin reductase, andthioredoxin peroxidases. Thioredoxin has a conserved cat-alytic site (-Cys-Gly-Pro-Cys-Lys-) that undergoes reversibleoxidation to the cystine disulfide. Oxidized thioredoxin isa major substrate for thioredoxin reductase, and reducedthioredoxin serves as an electron carrier to reduce peroxire-doxins. The oxidized thioredoxin is reduced back to the re-duced form by thioredoxin reductase [105, 106]. There aretwo main thioredoxins: thioredoxin-1 (TRX-1), a cytosolicform; thioredoxin-2 (TRX-2), a mitochondrial form [105].Modification of thiols in thioredoxin interrupts signalingmechanisms involved in cell growth, proliferation, and apop-tosis. The role of thioredoxin in the regulation of the ac-tivation of apoptosis signal-regulating kinase-1 (ASK-1) anddownstream apoptosis pathways has been reported in multi-ple studies [95, 96, 106, 107]. Thioredoxin can associate withthe N-terminal portion of ASK-1 in vitro and in vivo. Ex-pression of thioredoxin inhibited ASK-1 kinase activity andthe subsequent ASK-1-dependent apoptosis [107]. In restingcells, endogenous ASK-1 constitutively forms a complexwhich includes thioredoxin. Upon ROS stimulation, theASK-1 unbinds from thioredoxin and forms a fully activatedhigher-molecular-mass complex. As discussed above, TNFαincreases oxidative stress in mice with elevated CYP2E1, withsubsequent activation of ASK-1 via a mechanism involvingthioredoxin-ASK-1 dissociation, followed by activation ofdownstream MKK and MAPK [101].

Both TRX-1 and TRX-2 are involved in the protectionfrom oxidative stress. TRX-2 plays an important role in pro-tecting the mitochondria against oxidative stress and in pro-tecting cells from ROS-induced apoptosis. Supplementationof human recombinant TRX-1 to mice fed a Lieber-DeCarliethanol diet decreased several markers of oxidative stress,inflammatory cytokine expression, and apoptosis in liver[108]. Since thioredoxin is a reducing molecule which candecrease oxidative stress, we evaluated [109] whether thiore-doxin can inhibit the oxidative stress induced by CYP2E1,and whether there is any difference in the function of TRX-1 versus TRX-2 in blunting CYP2E1 oxidant stress. SiRNAfor either TRX-1 or TRX-2 was added to HepG2 cells withCYP2E1 expression (E47 cells) or without CYP2E1 expres-sion (C34 cells) to test (1) whether thioredoxin decreasesoxidative stress and injury induced by CYP2E1; (2) consider-ing the compartmentation of thioredoxin, whether TRX-1 orTRX-2 has a stronger protective effect in preventing againstthis injury and oxidative stress; (3) what the mechanism ofthe protection by thioredoxin from cell death in CYP2E1-expressing cells is [109].

Both E47 and C34 cells were treated with either controlsiRNA, TRX-1 siRNA, or TRX-2 siRNA, or both TRX-1 andTRX-2 siRNA for 72 hrs. TRX-1 expression was decreasedby 90% by either TRX-1 siRNA alone or TRX-1 and TRX-2 siRNA together in both cell lines. TRX-2 expression was

decreased by 80–90% by TRX-2 siRNA alone or TRX-1 andTRX-2 siRNA together in both cell lines. TRX-1 siRNA isspecific for cytosolic thioredoxin and had no effect on levelsof mitochondrial thioredoxin, and TRX-2 siRNA is specificfor decreasing mitochondrial thioredoxin and had no effecton levels of cytosolic thioredoxin.

Knockdown of TRX-1 or TRX-2 or both decreased cellviability of E47 cells by 40–60%, but cell viability of C34cells was not affected with the knockdown of either TRX-1or TRX-2 or both (Figure 6(a)). These results indicate thatcell death induced by thioredoxin knockdown under theseconditions is CYP2E1 dependent and that decreasing eitherTRX-1 or TRX-2 promotes this toxicity. To assess the modeof cell death, experiments studying uptake of propidiumiodide or annexin V staining were carried out. Uptake of pro-pidium iodide into E47 cells was elevated upon knockdownof either TRX-1 or TRX-2 or both. Annexin V staining, takenas a reflection of apoptosis, was also elevated in the E47 cellsupon knockdown of TRX-1 or TRX-2 or both. Thus, the celldeath appears to be a mix of necrosis plus apoptosis, thatis, necroptosis. We next evaluated whether knocking downof thioredoxin intracellularly by siRNA induces ROS pro-duction and lipid peroxidation. Total ROS was detectedboth by fluorescence microscopy, flow cytometry assay, andspectrofluorimetry assay. An increase of ROS productionwas detected in E47 cells but not in C34 cells after 72 hrstreatment with either TRX-1 or TRX-2 siRNA or both.Quantification of ROS production by spectrofluorimetryindicated that total ROS production was elevated 50–100%by thioredoxin knockdown in E47 cells (Figure 6(b)). Therewere no increases in ROS production in C34 cells uponthioredoxin knockdown. There were significant increases ofROS production when either TRX-1 or TRX-2 was lowered.This suggests that TRX-1 alone or TRX-2 alone is not suf-ficient to protect the E47 cells from oxidative stress. It wouldappear that both TRX-1 or TRX-2 are essential for theprotection of E47 cells from oxidative stress. The productionof superoxide was assayed using dihydroethidium (DHE) asthe probe. Knockdown of TRX-1 or TRX-2 or both increasedDHE fluorescence in E47 cells, but not C34 cells. 4-HNEadduct formation was analyzed by immunocytochemistrywith fixed E47 and C34 cells. At baseline, 4-HNE adductexpression is higher in E47 cells than C34 cells when controlsiRNA was applied, similar to the increase in fluorescenceof E47 compared to C34 cells. There was no increase of4-HNE adducts in C34 cells, but a significant increase of4-HNE adducts was observed in E47 cells when comparingeither TRX-1 or TRX-2 or both siRNA treatment to controlsiRNA treatment. Treatment with either TRX-1 or TRX-2siRNA or both did not cause a significant change of totalglutathione level in C34 cells, while a 50% decrease wasfound in E47 cells (Figure 6(c)). This suggests that withknockdown of thioredoxin, glutathione was consumed as amajor reducing molecule and antioxidant. Addition to theculture medium of glutathione ethyl ester prevented E47cell death caused by either TRX-1 or TRX-2 siRNA or bothtogether (Figure 6(d)). The lowering of, as well as the pro-tection by, glutathione suggests that the knockdown of thi-oredoxin-induced cell death is related to oxidative stress.


0

20

40

60

80

100

120

ControlsiRNA

TRX-1siRNA

TRX-2siRNA

TRX-1 + TRX-2siRNA

Cel

lvia

bilit

y(%

)

C34 cellsE47 cells

(a)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

(a.u

.)/u

gpr

otei

n

E47 cellsC34 cells

Control siRNATRX-1 siRNA

TRX-2 siRNATRX-1 + TRX-2 siRNA

(b)

E47 cells C34 cells

Control siRNA

TRX-1 siRNA

TRX-2 siRNA

TRX-1 + TRX-2 siRNA

0

0.2

0.4

0.6

0.8

1

1.2

Rel

ativ

eG

SHle

vel

(c)

0

20

40

60

80

100

120

ControlsiRNA

TRX-1siRNA

TRX-2siRNA

TRX-1 + TRX-2siRNA

Cel

lvia

bilit

y(%

)

&&

&

No GSSEPlus GSSE

(d)

Figure 6: Effect of thioredoxin (TRX) knockdown on E47 (express CYP2E1) and C34 (do not express CYP2E1) HepG2 cell viability. E47and C34 cells were treated with control siRNA or cytosolic TRX-1 siRNA or mitochondrial TRX-2 siRNA or both TRX-1 and TRX-2 siRNAsfor 72 hours. (a) Cell viability was determined by a MTT assay. (b) ROS production was determined by a fluorescence assay. Arbitraryunits of fluorescence by the E47 and C34 cells. (c) Cellular levels of glutathione (GSH). The GSH level in each group was expressed as thevalue relative to that of the control siRNA treatment group in E47 cells. (d) Supplementation with GSH restores E47 cell viability after TRXknockdown. At 24 hours, 5 mM glutathione ethyl ester (GSSE) was added to the cell culture medium, and the cells were incubated with theindicated siRNA for 48 hours followed by MTT assay. Note: both cytosolic and mitochondrial TRX are important in protection of HepG2cells from CYP2E1-generated oxidant stress.

Since thioredoxin is bound to ASK-1 and inhibits theactivation of ASK-1, experiments were carried out to evaluatewhether thioredoxin knockdown activates ASK-1 and down-stream MAPK signaling pathways in the E47 cells. IncreasedASK-1 phosphorylation was seen by immunohistochemistryin E47 cells upon treatment with TRX-1 or TRX-2 siRNAor both at 5, 24, and 48 hrs, but not after 72 hrs ofsiRNA treatment. Western blot analysis revealed a 2–4-foldincrease in the pASK-1/ASK-1 ratio 24 hrs and 48 hrs butnot after 72 hrs of thioredoxin knockdown. ASK-1 activatesdownstream MAPK such as JNK and p38 MAPK, ultimatelyby promoting their phosphorylation to pJNK or pp38 MAPK

[95, 96]. Increased JNK1 but not JNK2 phosphorylation wasseen in E47 cells treated with either TRX-1 or TRX-2 siRNAfor 48 hrs (Figure 7(b)). No such activation persisted at 72hrs after treatment. Thus, activation of JNK1 occurs afterthe earlier activation of ASK-1 (5–48 hrs) and declines whenactivation of ASK-1 terminates (72 hrs). p38 MAPK was notactivated under these conditions as there was no increase inpp38 MAPK levels. One downstream target of JNK1 is cJUNphosphorylation. There was an increase in the pc-JUN/c-JUN ratio 72 hrs after treatment with siRNA for TRX-1,TRX-2, or both, a time point after the activation of JNK1(48 hrs).


0

20

40

60

80

100

120

ControlsiRNA

TRX-1siRNA

TRX-2siRNA

TRX-1 + TRX-2siRNA

Cel

lvia

bilit

y(%

)No L-JNKI1Plus L-JNKI1

# &# &# &

(a)

− JNKi

+ JNKi

pJNK-2

pJNK-1

JNK-2

JNK-1

48 hrs 72 hrs

ControlsiRNA

TRX-1siRNA

TRX-2siRNA

TRX-1 +TRX-2siRNA

1 1.8 2.4 4

ControlsiRNA

TRX-1siRNA

TRX-2siRNA

TRX-1 +TRX-2siRNA

pJNK-2

pJNK-1

JNK-2

JNK-1

1 1.03 1.02 1.07

(b)

Figure 7: A JNK inhibitor protects the E47 cells from loss of viability produced by TRX knockdown. The E 47 cells were incubated withand without 5 uM of the JNK inhibitor L-JNKI1for 3 hours followed by treatment with the indicated siRNA for 48 or 72 hours. (a) MTTassay to determine cell viability. (b) The effect of TRX knockdown on the activation of JNK in the absence and presence of the JNK inhibitor.Numbers under the blots refer to the pJNK/JNK ratio.

Could the CYP2E1 plus thioredoxin knockdown inducedcell death be mediated through ASK-1 and JNK1 signalingpathways? The JNK inhibitor, L-JNKI1, which specificallyinhibits the phosphorylation of JNK, lowered the declinein E47 cell viability from 45–50% in the absence of L-JNKI1 to about 20–30% in the presence of L-JNKI1 plusTRX-1, or TRX-2 siRNA, or both TRX-1 and TRX-2 siRNAtreatment (Figure 7(a)). Under these conditions, L-JNKI1strongly blunted the activation of JNK which occurs 48 hrsafter thioredoxin knockdown; the pJNK1/JNK1 ratio waselevated 2- to 4-fold by siRNA for TRX-1 or TRX-2 or bothin the absence of JNKI1, whereas no increase in pJNK1/JNK1was observed in the presence of the inhibitor (Figure 7(b)).The partial protection by L-JNKI1 suggests that the celldeath induced by thioredoxin knockdown was partly via JNKsignaling pathways, although non-JNK-dependent pathwaysare also likely involved.

In conclusion, both cytosolic and mitochondrial thiore-doxin are important in protecting HepG2 cells from celldeath by oxidative stress induced by CYP2E1. Thioredox-

in knockdown increased cellular production of ROS andincreased lipid peroxidation in HepG2 cells expressingCYP2E1. The signaling pathway which induced cell deathby thioredoxin knockdown may involve, at least in part, theactivation of ASK-1 and JNK1. This protection by both TRX-1 and TRX-2 against CYP2E1-dependent toxicity may play arole in the ability of thioredoxin to protect against ethanol-induced hepatotoxicity [108] and suggests that antioxidativeprotection in both the cytosol and mitochondria is necessaryfor effective protection against liver injury potentiated byCYP2E1.

12. Effect of N-Acetylcysteine (NAC)

We evaluated [74] the effect of NAC, a general antioxidantand a precursor of GSH, on the potentiation of TNFα toxicityby pyrazole as a proof of principle that oxidative stress playsan important role in the overall liver injury. C57BL/6 micewere treated with pyrazole for two days and then challengedwith either saline or TNFα. Some mice in each group were


0

200

400

600

800

1000

1200

1400

1600

Saline PY TNF TNF + PY

ALT

(un

its/

L)

(a)

Saline PY TNF TNF + PY0

100

200

300

400

500

600

700

800

900

AST

(un

its/

L)

(b)


TB

AR

S(n

mol

es/m

g)

3.5

3

2.5

2

1.5

1

0.5

0

B6-129PF2/JNOS2−/−

(c)


GSH

(nm

oles

/mg)

0

10

20

30

40

50

60

B6-129PF2/JNOS2−/−

(d)

Figure 8: TNFα-plus-pyrazole-induced hepatotoxicity and oxidative stress are decreased in iNOS knockout mice. B6-129 WT mice andB6-129 iNOS knockout mice (NOS2−/−) were treated with either saline or pyrazole alone or TNFα alone or pyrazole plus TNFα for 3 daysfollowed by assays of (a) ALT/AST, (b) TBARS, and (c) GSH. Note: liver injury and oxidant stress were much lower in the NOS2−/− micethan the WT mice indicating a role for NO and NO metabolites in the TNFα-plus-pyrazole-induced liver injury and oxidative stress.

Pyrazole CYP2E1 O− •2

TNF iNOS NO

ROSONOO•

NF- BMitochondrial

dysfunction

Protective factors

JNK, p38

Figure 9: Model for the potentiation of TNFα-induced hepatotoxicity, oxidative stress mitochondrial dysfunction, and activation of MAPKby pyrazole induction of CYP2E1. Pyrazole induction of CYP2E1 coupled to TNFα induction of iNOS results in elevated oxidative/nitrosativestress in hepatocytes. This results in activation of JNK and p38 MAPK which, along with the elevated ROS/RNS, damage mitochondrialfunction ultimately leading to liver injury.

also treated with 150 mg/kg NAC on the second day oftreatment with pyrazole and on day 3 prior to the challengewith TNFα. The elevation in ALT and AST and the necrosiscaused by the pyrazole plus TNFα treatment were loweredby NAC. The increase in TBARs produced by pyrazole plus

TNFα and the decline in liver GSH were both preventedby NAC. Treatment with NAC had no effect on CYP2E1protein levels or CYP2E1 catalytic activity. The activation ofJNK or p38 MAPK by the pyrazole plus TNFα treatment,compared to pyrazole alone, was blocked by NAC. The


pyrazole plus TNFα treatment elevated levels of iNOS 2.6-fold, and this increase in iNOS was blunted by NAC to a 1.4-fold increase. These results with NAC suggest that elevatedoxidative stress is central to the activation of JNK and p38MAPK, to peroxynitrite formation, and to the liver injuryproduced by treatment with pyrazole plus TNFα.

13. Pyrazole/TNFα Hepatotoxicity iniNOS Knockout Mice

The inducible nitric oxide synthase (iNOS) has been shownto play an important role in alcohol-induced liver injury[110]. We hypothesized that induction of CYP2E1 by pyra-zole and induction of iNOS by LPS/TNFα result in the for-mation of the powerful oxidant peroxynitrite, ONOO, de-rived from the reaction between O2

−• and NO. 3-Nitroty-rosine protein adducts (3-NT) were elevated in the liver afterpyrazole plus LPS treatment [71, 72]. We believe that ONOOplays a key role in the oxidative/nitrosative stress and hepato-toxicity produced by the pyrazole plus LPS/TNFα treatment.If correct, oxidative/nitrosative stress and hepatotoxicity pro-duced by pyrazole plus LPS/TNFα treatment should beblunted in iNOS null mice. NOS2 (iNOS) knockout mice(B6-129P2) and genetic background control B6-129PF2/Jmice were purchased from Jackson Laboratory and treatedwith saline or pyrazole alone or TNFα alone or pyrazoleplus TNFα. The pyrazole plus TNFα treatment elevated ALTor AST levels about 2-fold (P < 0.05) in iNOS null miceas compared to treatment with saline or pyrazole alone orTNFα alone (Figure 8(a)). Pyrazole plus TNFα elevated ALTand AST about four- to fivefold in the genetic backgroundmice (Figure 8(a)) (P < .01 compared to the increase inALT and AST in iNOS null mice). In NOS2−/−mice, TNFαplus PY induced some hepatocyte degeneration change inthe pericentral area but no loci of necrosis were found. Inthe control wild-type B6-129PF2/J mice, TNFα plus PY in-duced more severe liver injury and necrotic loci were foundin several pericentral areas. TNFα plus PY slightly in-creased lipid peroxidation in NOS2−/−mice compared withsaline-, PY-, or TNFα-treated mice. Lipid peroxidation wasmore significantly elevated by TNFα-plus-PY treatment inB6-129PF2/J mice (4-fold increase) compared to the othergroups and to the TNFα plus PY treated NOS2−/− mice(2-fold increase, Figure 8(b)). TNFα plus PY lowered GSHlevels by 25% in NOS2−/− mice, while a more pronounceddecline in GSH occurred in the control mice (67% decrease,Figure 8(c)). Levels of CYP2E1 were elevated to comparableextents by pyrazole in the wild-type and the iNOS knockoutmice (about 2.5–3-fold); thus, the lower liver injury in theiNOS knockout mice is not due to lower levels of CYP2E1.These results suggest that while TNFα plus PY does inducesome liver injury and oxidant stress in the NOS2−/− mice,a more severe liver injury and oxidant stress is inducedby TNFα plus PY in the control mice. We hypothesizethat NO derived from iNOS reacts with superoxide radicalproduced from CYP2E1 to generate the powerful oxidantperoxynitrite which plays a critical role in the liver injuryproduced by TNF plus pyrazole. The absence of iNOS in theknockout mice with the accompanying decline in NO would

prevent formation of significant amounts of peroxynitriteeven though superoxide continues to be produced from theelevated CYP2E1 and therefore liver injury is lowered.

14. Conclusions

This paper has focused on two major contributors to mech-anisms by which ethanol causes liver injury, induction ofCYP2E1, and elevated endotoxin (LPS) levels followed by in-creased production of TNFα. Each of these has been ex-tensively studied, but there are few studies in which bothfactors have been evaluated simultaneously. We have shownthat induction of CYP2E1 by pyrazole potentiates LPS- orTNF-induced hepatotoxicity. Evidence for a role for CYP2E1comes from studies in which the CYP2E1 inhibitor CMZblocks the liver injury, and from studies with CYP2E1 knock-out mice where pyrazole plus LPS toxicity is blunted. Thepotentiated toxicity is associated with an increase in oxidativeand nitrosative stress. Prevention of such increases, for ex-ample, treatment with the antioxidant NAC or administra-tion of TNFα plus pyrazole to iNOS knockout mice, bluntsthe liver injury thus validating that the elevated oxidative/nitrosative stress plays a key role in producing the liver in-jury rather than occurs because of liver injury. JNK andP38 MAP kinases are activated by the combined pyrazoleplus LPS/TNFα treatment. Preventing activation of JNK withSP600125 or activation of P38 MAPK with SB203580 de-creases the liver injury. Inhibition of CYP2E1 or use ofCYP2E1 knockout or iNOS knockout mice or preventingthe oxidative/nitrosative stress decreases the activation ofJNK and P38 MAPK. We hypothesize that the increase inoxidative/nitrosative stress and the activation of MAP kinasesultimately impact on mitochondrial integrity and function asshown by the increase in mitochondrial swelling and declinein mitochondrial membrane potential. Protection of mito-chondrial integrity with cyclosporine A prevents the TNFα-plus-pyrazole-induced hepatotoxicity and oxidative stress. InHepG2 cells expressing CYP2E1, both cytosolic and mito-chondrial TRX are necessary for protection against CYP2E1-generated oxidative stress, and cell toxicity. We hypothesizethat similar interactions involving activation of MAP kinases,oxidative stress and mitochondrial dysfunction occur as aresult of ethanol induction of CYP2E1 and elevation ofLPS/TNFα, and our working scheme is shown in Figure 9.Induction of CYP2E1 by pyrazole or ethanol increases super-oxide radical production, while elevation of LPS/TNFα byethanol activates iNOS and NO production. The interactionbetween superoxide and NO produces the powerful oxidantperoxynitrite. Downstream targets for ROS and RNS includeactivation of ASK-1 and subsequently JNK and mitochon-drial dysfunction which contribute to loss of viability.

References

[1] E. Albano, “Alcohol, oxidative stress and free radical damage,”Proceedings of the Nutrition Society, vol. 65, no. 3, pp. 278–290, 2006.

[2] S. C. Bondy, “Ethanol toxicity and oxidative stress,” Toxicol-ogy Letters, vol. 63, no. 3, pp. 231–241, 1992.


[3] A. I. Cederbaum, “Introduction—serial review: alcohol,oxidative stress and cell injury,” Free Radical Biology andMedicine, vol. 31, no. 12, pp. 1524–1526, 2001.

[4] R. Nordmann, C. Ribiere, and H. Rouach, “Implication offree radical mechanisms in ethanol-induced cellular injury,”Free Radical Biology and Medicine, vol. 12, no. 3, pp. 219–240,1992.

[5] J. A. Knight, “Free radicals: their history and current status inaging and disease,” Annals of Clinical and Laboratory Science,vol. 28, no. 6, pp. 331–346, 1998.

[6] J. P. Kehrer, “Free radicals as mediators of tissue injury anddisease,” Critical Reviews in Toxicology, vol. 23, no. 1, pp. 21–48, 1993.

[7] B. Chance, H. Sies, and A. Boveris, “Hydroperoxidemetabolism in mammalian organs,” Physiological Reviews,vol. 59, no. 3, pp. 527–604, 1979.

[8] R. E. White, “The involvement of free radicals in the mecha-nisms of monooxygenases,” Pharmacology and Therapeutics,vol. 49, no. 1-2, pp. 21–42, 1991.

[9] J. Blanck, O. Ristau, A. A. Zhukov, A. L. Archakov, H. Rein,and K. Ruckpaul, “Cytochrome P-450 spin state and leakinessof the monooxygenase pathway,” Xenobiotica, vol. 21, no. 1,pp. 121–135, 1991.

[10] H. De Groot, “Reactive oxygen species in tissue injury,”Hepato-Gastroenterology, vol. 41, no. 4, pp. 328–332, 1994.

[11] B. Halliwell, “Antioxidant defence mechanisms: From thebeginning to the end (of the beginning),” Free RadicalResearch, vol. 31, no. 4, pp. 261–272, 1999.

[12] B. P. Yu, “Cellular defenses against damage from reactiveoxygen species,” Physiological Reviews, vol. 74, no. 1, pp. 139–162, 1994.

[13] G. M. Rosen, S. Pou, C. L. Ramos, M. S. Cohen, and B. E.Britigan, “Free radicals and phagocytic cells,” FASEB Journal,vol. 9, no. 2, pp. 200–209, 1995.

[14] H. M. Lander, “An essential role for free radicals and derivedspecies in signal transduction,” FASEB Journal, vol. 11, no. 2,pp. 118–124, 1997.

[15] Y. Adachi, B. U. Bradford, W. Gao, H. K. Bojes, and R.G. Thurman, “Inactivation of Kupffer cells prevents earlyalcohol-induced liver injury,” Hepatology, vol. 20, no. 2, pp.453–460, 1994.

[16] Y. Iimuro, R. M. Gallucci, M. Luster, H. Kono, and R.G. Thurman, “Antibodies to tumor necrosis factor alfaattenuate hepatic necrosis and inflammation caused bychronic exposure to ethanol in the rat,” Hepatology, vol. 26,no. 6, pp. 1530–1537, 1997.

[17] A. A. Nanji, U. Khettry, and S. M. H. Sadrzadeh, “Lactobacil-lus feeding reduces endotoxemia and severity of experimentalalcoholic liver disease,” Proceedings of the Society for Exper-imental Biology and Medicine, vol. 205, no. 3, pp. 243–247,1994.

[18] X. M. Yin and W. X. Ding, “Death receptor activation-induced hepatocyte apoptosis and liver injury,” CurrentMolecular Medicine, vol. 3, no. 6, pp. 491–508, 2003.

[19] R. K. Rao, A. Seth, and P. Sheth, “Recent advances inalcoholic liver disease. I. Role of intestinal permeability andendotoxemia in alcoholic liver disease,” American Journal ofPhysiology—Gastrointestinal and Liver Physiology, vol. 286,no. 6, pp. G881–G884, 2004.

[20] G. L. Su, “Lipopolysaccharides in liver injury: molecularmechanisms of Kupffer cell activation,” American Journal ofPhysiology—Gastrointestinal and Liver Physiology, vol. 283,no. 2, pp. G256–G265, 2002.

[21] S. Suzuki, S. Nakamura, A. Serizawa et al., “Role of Kupffercells and the spleen in modulation of endotoxin-induced liver

injury after partial hepatectomy,” Hepatology, vol. 24, no. 1,pp. 219–225, 1996.

[22] J. H. Wang, H. P. Redmond, R. W. G. Watson, and D.Bouchier-Hayes, “Role of lipopolysaccharide and tumornecrosis factor-alpha in induction of hepatocyte necrosis,”American Journal of Physiology, vol. 269, no. 2, pp. G297–G304, 1995.

[23] J. A. Hewett, P. A. Jean, S. L. Kunkel, and R. A. Roth,“Relationship between tumor necrosis factor-alpha andneutrophils in endotoxin-induced liver injury,” AmericanJournal of Physiology, vol. 265, no. 6, pp. G1011–G1015, 1993.

[24] I. V. Deaciuc, M. Nikolova-Karakashian, F. Fortunato, E. Y.Lee, D. B. Hill, and C. J. McClain, “Apoptosis and dysreg-ulated ceramide metabolism in a murine model of alcohol-enhanced lipopolysaccharide hepatotoxicity,” Alcoholism, vol.24, no. 10, pp. 1557–1565, 2000.

[25] J. Hansen, D. L. Cherwitz, and J. I. Allen, “The role oftumor necrosis factor-alpha in acute endotoxin-inducedhepatotoxicity in ethanol-fed rats,” Hepatology, vol. 20, no.2, pp. 461–474, 1994.

[26] B. S. Bhagwandeen, M. Apte, L. Manwarring, and J. Dick-eson, “Endotoxin induced hepatic necrosis in rats on analcohol diet,” Journal of Pathology, vol. 152, no. 1, pp. 47–53,1987.

[27] A. Koteish, S. Yang, H. Lin, X. Huang, and A. M. Diehl,“Chronic ethanol exposure potentiates lipopolysaccharideliver injury despite inhibiting Jun N-terminal kinase andcaspase 3 activation,” Journal of Biological Chemistry, vol. 277,no. 15, pp. 13037–13044, 2002.

[28] P. Mathurin, Q. G. Deng, A. Keshavarzian, S. Choudhary, E.W. Holmes, and H. Tsukamoto, “Exacerbation of alcoholicliver injury by enteral endotoxin in rats,” Hepatology, vol. 32,no. 5, pp. 1008–1017, 2000.

[29] R. G. Thurman, “Mechanisms of hepatic toxicity. II. Alco-holic liver injury involves activation of Kupffer cells byendotoxin,” American Journal of Physiology, vol. 275, no. 4,pp. G605–G611, 1998.

[30] N. Enomoto, K. Ikejima, S. Yamashina et al., “Kupffer cellsensitization by alcohol involves increased permeability togut-derived endotoxin,” Alcoholism, vol. 25, supplement 6,pp. 51S–54S, 2001.

[31] H. Kono, I. Rusyn, T. Uesugi et al., “Diphenyleneiodo-nium sulfate, an NADPH oxidase inhibitor, prevents earlyalcohol-induced liver injury in the rat,” American Journal ofPhysiology—Gastrointestinal and Liver Physiology, vol. 280,no. 5, pp. G1005–G1012, 2001.

[32] H. Kono, I. Rusyn, M. Yin et al., “NADPH oxidase-derived free radicals are key oxidants in alcohol-induced liverdisease,” Journal of Clinical Investigation, vol. 106, no. 7, pp.867–872, 2000.

[33] M. Yin, M. D. Wheeler, H. Kono et al., “Essential role oftumor necrosis factorα in alcohol-induced liver injury inmice,” Gastroenterology, vol. 117, no. 4, pp. 942–952, 1999.

[34] T. Uesugi, M. Froh, G. E. Arteel et al., “Delivery of IkappaBsuperrepressor gene with adenovirus reduces early alcohol-induced liver injury in rats,” Hepatology, vol. 34, no. 6, pp.1149–1157, 2001.

[35] H. M. Bolt, P. H. Roos, and R. Thier, “The cytochromeP-450 isoenzyme CYP2E1 in the biological processing ofindustrial chemicals,” International Archives of Occupationaland Environmental Health, vol. 76, no. 3, pp. 174–185, 2003.

[36] C. S. Lieber, “Cytochrome P-4502E1: its physiological andpathological role,” Physiological Reviews, vol. 77, no. 2, pp.517–544, 1997.


[37] D. R. Koop, “Oxidative and reductive metabolism by cyto-chrome P450 2E1,” FASEB Journal, vol. 6, no. 2, pp. 724–730,1992.

[38] J. L. Raucy, J. C. Kraner, and J. M. Lasker, “Bioactivation ofhalogenated hydrocarbons by cytochrome P4502E1,” CriticalReviews in Toxicology, vol. 23, no. 1, pp. 1–20, 1993.

[39] B. J. Song, A. I. Cederbaum, D. R. Koop, M. Ingelman-Sund-berg, and A. Nanji, “Ethanol-inducible cytochrome P450(CYP2E1): biochemistry, molecular biology and clinical rele-vance,” Alcoholism, vol. 20, supplement 8, pp. 138A–146A,1996.

[40] E. Tanaka, M. Terada, and S. Misawa, “Cytochrome P450 2E1:its clinical and toxicological role,” Journal of Clinical Phar-macy and Therapeutics, vol. 25, no. 3, pp. 165–175, 2000.

[41] C. S. Yang, J. S. H. Yoo, H. Ishizaki, and J. Hong, “CytochromeP450IIE1. Roles in nitrosamine metabolism and mechanismsof regulation,” Drug Metabolism Reviews, vol. 22, no. 2-3, pp.147–159, 1990.

[42] S. S. T. Lee, J. T. M. Buters, T. Pineau, P. Fernandez-Salguero,and F. J. Gonzalez, “Role of CYP2E1 in the hepatotoxicity ofacetaminophen,” Journal of Biological Chemistry, vol. 271, no.20, pp. 12063–12067, 1996.

[43] G. Ekstrom and M. Ingelman-Sundberg, “Rat liver micro-somal NADPH-supported oxidase activity and lipid perox-idation dependent on ethanol-inducible cytochrome P-450(P-450IIE1),” Biochemical Pharmacology, vol. 38, no. 8, pp.1313–1319, 1989.

[44] L. D. Gorsky, D. R. Koop, and M. J. Coon, “On the stoichi-ometry of the oxidase and monooxygenase reactions cata-lyzed by liver microsomal cytochrome P-450,” Journal of Bio-logical Chemistry, vol. 259, no. 11, pp. 6812–6817, 1984.

[45] E. Dicker and A. I. Cederbaum, “Hydroxyl radical generationby microsomes after chronic ethanol consumption,” Alco-holism, vol. 11, no. 3, pp. 309–314, 1987.

[46] S. M. Klein, G. Cohen, C. S. Lieber, and A. I. Cederbaum,“Increased microsomal oxidation of hydroxyl radical scav-enging agents and ethanol after chronic consumption ofethanol,” Archives of Biochemistry and Biophysics, vol. 223, no.2, pp. 425–432, 1983.

[47] S. Puntarulo and A. I. Cederbaum, “Increased NADPH-de-pendent chemiluminescence by microsomes after chronicethanol consumption,” Archives of Biochemistry and Biophys-ics, vol. 266, no. 2, pp. 435–445, 1988.

[48] J. Rashba-Step, N. J. Turro, and A. I. Cederbaum, “IncreasedNADPH- and NADH-dependent production of superoxideand hydroxyl radical by microsomes after chronic ethanoltreatment,” Archives of Biochemistry and Biophysics, vol. 300,no. 1, pp. 401–408, 1993.

[49] A. A. Caro and A. I. Cederbaum, “Oxidative stress, toxicol-ogy, and pharmacology of CYP2E1,” Annual Review of Phar-macology and Toxicology, vol. 44, pp. 27–42, 2004.

[50] T. Castillo, D. R. Koop, S. Kamimura, G. Triadafilopoulos,and H. Tsukamoto, “Role of cytochrome P-450 2E1 inethanol-, carbon tetrachloride— and iron-dependent micro-somal lipid peroxidation,” Hepatology, vol. 16, no. 4, pp. 992–996, 1992.

[51] M. Morimoto, M. A. Zern, A. L. Hagbjork, M. Ingelman-Sundberg, and S. W. French, “Fish oil, alcohol, and liverpathology: role of cytochrome P450 2E1,” Proceedings of theSociety for Experimental Biology and Medicine, vol. 207, no. 2,pp. 197–205, 1994.

[52] A. A. Nanji, S. Zhao, S. M. H. Sadrzadeh, A. J. Dannenberg,S. R. Tahan, and D. J. Waxman, “Markedly enhancedcytochrome P450 2E1 induction and lipid peroxidation is

associated with severe liver injury in fish oil-ethanol-fed rats,”Alcoholism, vol. 18, no. 5, pp. 1280–1285, 1994.

[53] S. W. French, K. Wong, L. Jui, E. Albano, A. L. Hagbjork, andM. Ingelman-Sundberg, “Effect of ethanol on cytochromeP450 2E1 (CYP2E1), lipid peroxidation, and serum proteinadduct formation in relation to liver pathology pathogene-sis,” Experimental and Molecular Pathology, vol. 58, no. 1, pp.61–75, 1993.

[54] Y. Lu, J. Zhuge, X. Wang, J. Bai, and A. I. Cederbaum,“Cytochrome P450 2E1 contributes to ethanol-induced fattyliver in mice,” Hepatology, vol. 47, no. 5, pp. 1483–1494, 2008.

[55] E. Albano, P. Clot, M. Morimoto, A. Tomasi, M. Ingelman-Sundberg, and S. W. French, “Role of cytochrome P4502E1-dependent formation of hydroxyethyl free radical in thedevelopment of liver damage in rats intragastrically fed withethanol,” Hepatology, vol. 23, no. 1, pp. 155–163, 1996.

[56] M. Morimoto, A. L. Hagbjork, A. A. Nanji et al., “Role ofcytochrome P4502E1 in alcoholic liver disease pathogenesis,”Alcohol, vol. 10, no. 6, pp. 459–464, 1993.

[57] M. Morimoto, A. L. Hagbjork, Y. J. Y. Wan et al., “Modulationof experimental alcohol-induced liver disease by cytochromeP450 2E1 inhibitors,” Hepatology, vol. 21, no. 6, pp. 1610–1617, 1995.

[58] Z. Q. Gouillon, D. Lucas, J. Li et al., “Inhibition of ethanol-induced liver disease in the intragastric feeding rat modelby chlormethiazole,” Experimental Biology and Medicine, vol.224, no. 4, pp. 302–308, 2000.

[59] K. Morgan, S. W. French, and T. R. Morgan, “Productionof a cytochrome P450 2E1 transgenic mouse and initialevaluation of alcoholic liver damage,” Hepatology, vol. 36, no.1, pp. 122–134, 2002.

[60] J. X. Bai and A. I. Cederbaum, “Adenovirus mediatedoverexpression of CYP2E1 increases sensitivity of HepG2cells to acetaminophen induced cytotoxicity,” Molecular andCellular Biochemistry, vol. 262, no. 1-2, pp. 165–176, 2004.

[61] J. X. Bai and A. I. Cederbaum, “Adenovirus-mediated expres-sion of CYP2E1 produces liver toxicity in mice,” ToxicologicalSciences, vol. 91, no. 2, pp. 365–371, 2006.

[62] H. Kono, B. U. Bradford, M. Yin et al., “Cyp2e1 is notinvolved in early alcohol-induced liver injury,” AmericanJournal of Physiology, vol. 277, no. 6, pp. G1259–G1267, 1999.

[63] D. R. Koop, B. Klopfenstein, Y. Iimuro, and R. G. Thurman,“Gadolinium chloride blocks alcohol-dependent liver toxic-ity in rats treated chronically with intragastric alcohol despitethe induction of CYP2E1,” Molecular Pharmacology, vol. 51,no. 6, pp. 944–950, 1997.

[64] B. U. Bradford, H. Kono, F. Isayama et al., “CytochromeP450 CYP2E1, but not nicotinamide adenine dinucleotidephosphate oxidase, is required for ethanol-induced oxidativeDNA damage in rodent liver,” Hepatology, vol. 41, no. 2, pp.336–344, 2005.

[65] S. Kamimura and H. Tsukamoto, “Cytokine gene expressionby Kupffer cells in experimental alcoholic liver disease,”Hepatology, vol. 22, no. 4, pp. 1304–1309, 1995.

[66] R. Honchel, M. B. Ray, L. Marsano et al., “Tumor necrosisfactor in alcohol enhanced endotoxin liver injury,” Alco-holism, vol. 16, no. 4, pp. 665–669, 1992.

[67] V. Purohit and D. A. Brenner, “Mechanisms of alcohol-in-duced hepatic fibrosis: a summary of the Ron Thurman sym-posium,” Hepatology, vol. 43, no. 4, pp. 872–878, 2006.

[68] H. Tsukamoto, Y. Takei, C. J. McClain et al., “How is the liverprimed or sensitized for alcoholic liver disease?” Alcoholism,vol. 25, supplement 5, pp. 171S–181S, 2001.


[69] J. G. Pastorino and J. B. Hoek, “Ethanol potentiates tumornecrosis factor-alpha cytotoxicity in hepatoma cells and pri-mary rat hepatocytes by promoting induction of the mito-chondrial permeability transition,” Hepatology, vol. 31, no. 5,pp. 1141–1152, 2000.

[70] H. Liu, B. E. Jones, C. Bradham, and M. J. Czaja, “Increasedcytochrome P-450 2E1 expression sensitizes hepatocytes to c-Jun-mediated cell death from TNF-alpha,” American JournalPhysiology-Gastrointestinal and Liver Physiology, vol. 282, no.2, pp. G257–G266, 2002.

[71] Y. Lu and A. I. Cederbaum, “Enhancement by pyrazole oflipopolysaccharide-induced liver injury in mice: role of cyto-chrome P450 2E1 and 2A5,” Hepatology, vol. 44, no. 1, pp.263–274, 2006.

[72] Y. Lu, X. Wang, and A. I. Cederbaum, “Lipopolysaccharide-induced liver injury in rats treated with the CYP2E1 inducerpyrazole,” American Journal of Physiology—Gastrointestinaland Liver Physiology, vol. 289, no. 2, pp. G308–G319, 2005.

[73] D. Wu and A. Cederbaum, “Cytochrome P4502E1 sensitizesto tumor necrosis factor alpha-induced liver injury throughactivation of mitogen-activated protein kinases in mice,”Hepatology, vol. 47, no. 3, pp. 1005–1017, 2008.

[74] D. Wu, C. Xu, and A. Cederbaum, “Role of nitric oxide andnuclear factor-kappaB in the CYP2E1 potentiation of tumornecrosis factor alpha hepatotoxicity in mice,” Free RadicalBiology and Medicine, vol. 46, no. 4, pp. 480–491, 2009.

[75] S. M. Bailey, “A review of the role of reactive oxygen andnitrogen species in alcohol-induced mitochondrial dysfunc-tion,” Free Radical Research, vol. 37, no. 6, pp. 585–596, 2003.

[76] J. B. Hoek, A. Cahill, and J. G. Pastorino, “Alcohol andmitochondria: a dysfunctional relationship,” Gastroenterol-ogy, vol. 122, no. 7, pp. 2049–2063, 2002.

[77] J. Zhuge and A. I. Cederbaum, “Inhibition of the mito-chondrial permeability transition by cyclosporin A preventspyrazole plus lipopolysaccharide-induced liver injury inmice,” Free Radical Biology and Medicine, vol. 46, no. 3, pp.406–413, 2009.

[78] J. A. McCubrey and R. A. Franklin, “Reactive oxygen inter-mediates and signaling through kinase pathways,” Antioxi-dants and Redox Signaling, vol. 8, no. 9-10, pp. 1745–1748,2006.

[79] T. W. Sturgill and J. Wu, “Recent progress in characterizationof protein kinase cascades for phosphorylation of ribosomalprotein S6,” Biochimica et Biophysica Acta, vol. 1092, no. 3,pp. 350–357, 1991.

[80] H. Ichijo, E. Nishida, K. Irie et al., “Induction of apoptosisby ASK1, a mammalian MAPKKK that activates SAPK/JNKand p38 signaling pathways,” Science, vol. 275, no. 5296, pp.90–94, 1997.

[81] F. Bardag-Gorce, B. A. French, J. Dedes, J. Li, and S. W.French, “Gene expression patterns of the liver in response toalcohol: in vivo and in vitro models compared,” Experimentaland Molecular Pathology, vol. 80, no. 3, pp. 241–251, 2006.

[82] J. Li, F. Bardag-Gorce, J. Oliva, J. Dedes, B. A. French, andS. W. French, “Gene expression modifications in the livercaused by binge drinking and S-adenosylmethionine feeding.The role of epigenetic changes,” Genes and Nutrition, vol. 5,no. 2, pp. 169–179, 2010.

[83] A. R. Aroor, T. T. James, D. E. Jackson, and S. D. Shukla,“Differential changes in MAP kinases, histone modifications,and liver injury in rats acutely treated with ethanol,”Alcoholism, vol. 34, no. 9, pp. 1543–1551, 2010.

[84] A. Szuster-Ciesielska, K. Plewka, J. Daniluk, and M.Kandefer-Szerszen, “Zinc supplementation attenuates etha-nol- and acetaldehyde-induced liver stellate cell activation by

inhibiting reactive oxygen species (ROS) production and byinfluencing intracellular signaling,” Biochemical Pharmacol-ogy, vol. 78, no. 3, pp. 301–314, 2009.

[85] R. Kishore, M. R. McMullen, and L. E. Nagy, “Stabilization oftumor necrosis factor alpha mRNA by chronic ethanol: roleof A + U-rich elements and p38 mitogen-activated proteinkinase signaling pathway,” Journal of Biological Chemistry,vol. 276, no. 45, pp. 41930–41937, 2001.

[86] S. Oak, P. Mandrekar, D. Catalano, K. Kodys, and G. Szabo,“TLR2- and TLR4-mediated signals determine attenuationor augmentation of inflammation by acute alcohol inmonocytes,” Journal of Immunology, vol. 176, no. 12, pp.7628–7635, 2006.

[87] J. G. Pastorino, N. Shulga, and J. B. Hoek, “TNF-alpha-induced cell death in ethanol-exposed cells depends on p38MAPK signaling but is independent of Bid and caspase-8,”American Journal of Physiology—Gastrointestinal and LiverPhysiology, vol. 285, no. 3, pp. G503–G516, 2003.

[88] M. J. Czaja, “The future of GI and liver research: editorialperspectives III. JNK/AP-1 regulation of hepatocyte death,”American Journal Physiology-Gastrointestinal and Liver Phys-iology, vol. 284, no. 6, pp. G875–G879, 2003.

[89] B. K. Gunawan, Z. X. Liu, D. Han, N. Hanawa, W. A.Gaarde, and N. Kaplowitz, “c-Jun N-terminal kinase playsa major role in murine acetaminophen hepatotoxicity,”Gastroenterology, vol. 131, no. 1, pp. 165–178, 2006.

[90] H. Liu, C. R. Lo, and M. J. Czaja, “NF-kappaB inhibitionsensitizes hepatocytes to TNF-induced apoptosis through asustained activation of JNK and c-Jun,” Hepatology, vol. 35,no. 4, pp. 772–778, 2002.

[91] J. M. Schattenberg, R. Singh, Y. Wang et al., “JNK1 but notJNK2 promotes the development of steatohepatitis in mice,”Hepatology, vol. 43, no. 1, pp. 163–172, 2006.

[92] T. Uehara, B. Bennett, S. T. Sakata et al., “JNK mediateshepatic ischemia reperfusion injury,” Journal of Hepatology,vol. 42, no. 6, pp. 850–859, 2005.

[93] A. Matsuzawa and H. Ichijo, “Redox control of cell fateby MAP kinase: physiological roles of ASK1-MAP kinasepathway in stress signaling,” Biochimica et Biophysica Acta,vol. 1780, no. 11, pp. 1325–1336, 2008.

[94] J. M. Kyriakis and J. Avruch, “Mammalian mitogen-activatedprotein kinase signal transduction pathways activated bystress and inflammation,” Physiological Reviews, vol. 81, no.2, pp. 807–869, 2001.

[95] G. Fujino, T. Noguchi, A. Matsuzawa et al., “Thioredoxinand TRAF family proteins regulate reactive oxygen species-dependent activation of ASK1 through reciprocal modula-tion of the N-terminal homophilic interaction of ASK1,”Molecular and Cellular Biology, vol. 27, no. 23, pp. 8152–8163,2007.

[96] H. Liu, H. Nishitoh, H. Ichijo, and J. M. Kyriakis, “Activationof apoptosis signal-regulating kinase 1 (ASK1) by tumornecrosis factor receptor-associated factor 2 requires priordissociation of the ASK1 inhibitor thioredoxin,” Molecularand Cellular Biology, vol. 20, no. 6, pp. 2198–2208, 2000.

[97] D. Wu and A. I. Cederbaum, “Oxidative stress and alcoholicliver disease,” Seminars in Liver Disease, vol. 29, no. 2, pp.141–154, 2009.

[98] X. Wang, A. Destrument, and C. Tournier, “Physiologicalroles of MKK4 and MKK7: insights from animal models,”Biochimica et Biophysica Acta, vol. 1773, no. 8, pp. 1349–1357,2007.

[99] A. Wullaert, K. Heyninck, and R. Beyaert, “Mechanisms ofcrosstalk between TNF-induced NF-κB and JNK activation


in hepatocytes,” Biochemical Pharmacology, vol. 72, no. 9, pp.1090–1101, 2006.

[100] J. Matsukawa, A. Matsuzawa, K. Takeda, and H. Ichijo, “TheASK1-MAP kinase cascades in mammalian stress response,”Journal of Biochemistry, vol. 136, no. 3, pp. 261–265, 2004.

[101] D. Wu and A. Cederbaum, “Activation of ASK-1 and down-stream MAP kinases in cytochrome P4502E1 potentiatedtumor necrosis factor alpha liver injury,” Free Radical Biologyand Medicine, vol. 49, no. 3, pp. 348–360, 2010.

[102] H. Nakagawa, S. Maeda, Y. Hikiba et al., “Deletion of apop-tosis signal-regulating kinase 1 attenuates acetaminophen-induced liver injury by inhibiting c-Jun N—Terminal kinaseactivation,” Gastroenterology, vol. 135, no. 4, pp. 1311–1321,2008.

[103] A. Holmgren, “Thioredoxin,” Trends in Biochemical Sciences,vol. 54, pp. 237–271, 1985.

[104] T. Ebrahimian and R. M. Touyz, “Thioredoxin in vascularbiology: role in hypertension,” Antioxidants and Redox Sig-naling, vol. 10, no. 6, pp. 1127–1136, 2008.

[105] G. Powis and W. R. Montfort, “Properties and biologicalactivities of thioredoxins,” Annual Review of Pharmacologyand Toxicology, vol. 41, pp. 261–295, 2001.

[106] G. Powis and W. R. Montfort, “Properties and biologicalactivities of thioredoxins,” Annual Review of Biophysics andBiomolecular Structure, vol. 30, pp. 421–455, 2001.

[107] M. Saitoh, H. Nishitoh, M. Fujii et al., “Mammalian thiore-doxin is a direct inhibitor of apoptosis signal-regulatingkinase (ASK) 1,” EMBO Journal, vol. 17, no. 9, pp. 2596–2606, 1998.

[108] J. I. Cohen, S. Roychowdhury, P. M. DiBello, D. W. Jacobsen,and L. E. Nagy, “Exogenous thioredoxin prevents ethanol-induced oxidative damage and apoptosis in mouse liver,”Hepatology, vol. 49, no. 5, pp. 1709–1717, 2009.

[109] L. Yang, D. Wu, X. Wang, and A. I. Cederbaum, “Depletionof cytosolic or mitochondrial thioredoxin increases CYP2E1-induced oxidative stress via an ASK-1-JNK1 pathway inHepG2 cells,” Free Radical Biology and Medicine, vol. 51, no.1, pp. 185–196, 2011.

[110] S. E. McKim, E. Gabele, F. Isayama et al., “Inducible nitricoxide synthase is required in alcohol-induced liver injury:studies with knockout mice,” Gastroenterology, vol. 125, no.6, pp. 1834–1844, 2003.


Review Article

Role of Adaptive Immunity in Alcoholic Liver Disease

Emanuele Albano

Department of Medical Sciences and Interdisciplinary Research Centre for Autoimmune Diseases (IRCAD),University “Amedeo Avogadro” of East Piedmont, Via Solaroli 17, 28100 Novara, Italy

Correspondence should be addressed to Emanuele Albano, [email protected]

Received 31 May 2011; Accepted 8 July 2011


Copyright © 2012 Emanuele Albano. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Stimulation of innate immunity is increasingly recognized to play an important role in the pathogenesis of alcoholic liverdisease (ALD), while the contribution of adaptive immunity has received less attention. Clinical and experimental data showthe involvement of Th-1 and Th-17 T-lymphocytes in alcoholic hepatitis. Nonetheless, the mechanisms by which alcohol triggersadaptive immunity are still incompletely characterized. Patients with advanced ALD have circulating IgG and T-lymphocytesrecognizing epitopes derived from protein modification by hydroxyethyl free radicals and end products of lipid-peroxidation. Hightiters of IgG against lipid peroxidation-derived antigens are associated with an increased hepatic production of proinflammatorycytokines/chemokines. Moreover, the same antigens favor the breaking of self-tolerance towards liver constituents. In particular,autoantibodies against cytochrome P4502E1 (CYP2E1) are evident in a subset of ALD patients. Altogether these results suggestthat allo- and autoimmune reactions triggered by oxidative stress might contribute to hepatic inflammation during the progressionof ALD.

1. Introduction

According to the World Health Organization, alcohol-relateddiseases are the third cause of death and disability in mostwell-developed countries and a leading cause of disease in thedeveloping countries in Eastern Europe, Central and SouthAmerica, and East Asia [1]. Although several organs are in-jured by ethanol, alcoholic liver disease (ALD) is the mostcommon medical consequence of excessive alcohol intakeaccounting for about 70% of the recorded mortality [2].Thus, understanding the mechanisms responsible for alcoholliver injury has a relevant clinical and social impact. Currentview indicates that multiple factors including oxidative stress,endoplasmic reticulum stress, metabolic alterations, and in-terferences with the transduction of intracellular signals areinvolved in ALD pathogenesis [3]. Furthermore, emergingevidence suggests that chronic inflammation represents thedriving force in the evolution of alcohol liver injury. In thiscontext, a large number of studies have investigate the roleof innate immunity in ALD ([4–8] for recent review), whilethe contribution of adaptive immunity to alcohol-inducedhepatic inflammation has received much less attention. Thispaper will give an overview of the possible implications of

adaptive immunity in the inflammatory processes associatedwith ALD.

Early studies have shown that liver inflammatory infil-trates in alcoholic hepatitis and active alcoholic cirrhosiscontain both CD8+ and CD4+ T-lymphocytes [9]. Ineither chronic alcohol-treated mice or alcohol abusers liver-infiltrating T cells express an activation/memory phenotypeand respond to T-cell receptor stimulation by producing Th-1 cytokines such as interferon-γ (IFN-γ) and TNF-α [10, 11].A Th-1 cytokine pattern is also evident in peripheral bloodT cells from active drinkers with or without ALD [12]. Inagreement with these observations, Paronetto [13] reportedthe presence in ALD patients of circulating antibodiestargeting alcohol-altered autologous hepatocytes. Hyperpro-duction of polyclonal gamma globulins is also frequent in al-cohol abusers in association with tissues deposition of IgA[14]. Moreover, ALD patients do not rarely have signs ofauto-immunity consisting in increased titers of circulatingantibodies directed against non-organ-specific and liver-specific autoantigens [15]. In particular, antiphospholipidantibodies can be observed in up to 80% of patients withalcoholic hepatitis or cirrhosis, but are not infrequent inheavy drinkers with milder liver damage [16, 17]. A further


evidence supporting the implication of adaptive immunityin ALD comes from the recent demonstration that IL-17-producing T helper (Th-17) lymphocytes are evident inhepatic inflammatory infiltrates of patients with alcoholichepatitis/cirrhosis in concomitance with an increase in IL-17plasma levels [18]. The implication of Th-17 T-cells in ALDis particularly important considering the increasing impor-tance ascribed to these cells in the pathogenesis of severalchronic inflammatory diseases including viral hepatitis B andC and primary biliary cirrhosis [19].

2. Role of Oxidative Stress inAlcohol-Induced Immune Reactions

The mechanisms by which alcohol triggers adaptive immu-nity are still incompletely characterized. Pioneering studyby Israel and colleagues [20] has shown that the adductsoriginating from acetaldehyde binding to hepatic proteinscause the production of specific antibodies when injectedinto experimental animals. The presence of antiacetaldehydeantibodies has been subsequently confirmed in rats chroni-cally exposed to alcohol as well as in alcoholic patients [21,22]. Guinea pig immunization with acetaldehyde-modifiedhemoglobin followed by alcohol feeding reproduces severalfeatures of alcoholic hepatitis [23]. However, the interest forace taldehyde-induced immune responses is hampered by theuncertainty regarding identity of the antigens involved andby the low specificity for ALD of anti-acetaldehyde antibod-ies [24]. Subsequent studies demonstrate that another meta-bolite of ethanol, that is, hydroxyethyl free radical (HER),produced during cytochrome P4502E1- (CYP2E1-) depen-dent ethanol oxidation, can interact with proteins generatingantigens distinct from those derived from acetaldehyde [25].Anti-HER IgG are detectable in chronically ethanol-fed ratsas well as in alcohol abusers [26, 27]. Human anti-HER IgGare recognized as main antigen HER-modified CYP2E1 [28],and their presence strictly correlates with CYP2E1 activity[29].

Oxidative stress is one of the feature of alcohol hepato-toxicity and significantly contributes to liver injury [30]. Oneof the consequence of oxidative stress is the stimulation oflipid peroxidation with the generation of wide array of re-active lipid breakdown products including aldehydes such asmalondialdehyde (MDA), 4-hydroxynonenal (4-HNE), andlipid hydroperoxides that are readily detectable in the serumand the liver of ALD patients and alcohol-fed rodents [30].Many lipid peroxidation products are highly reactive andby interacting with cellular constituents generate antigenicproducts that have been implicated in the stimulationof immune responses associated with atherosclerosis andseveral autoimmune diseases [31, 32]. We have observedthat a large proportion (55–70%) of the patients with ad-vanced ALD (alcoholic hepatitis and/or cirrhosis), but notheavy drinkers with fatty liver only, have elevated titersof circulating IgG against proteins adducted by MDA, 4-HNE, and oxidized arachidonic acid [33]. In about 35%of these patients, the presence of anti-MDA antibodies isassociated with the detection of peripheral blood CD4+ T

cells responsive to MDA adducts, indicating the capabilityof lipid oxidation antigens to trigger both the humoral andcellular branches of adaptive immunity [34]. At present,the chemical identity of the different antigens involved isstill incompletely characterized. Studies by Tuma and Thielehave shown that the reaction between MDA, acetaldehyde,and lysine ε-amino groups generates condensation productsnamed malondialdehyde-acetaldehyde adducts (MAA) thatnot only are highly immunogenic, but also can stimulate in-flammation [35]. MAA adducts have been detected in theliver of ethanol-fed rats, and their formation is responsiblefor the increased titers of IgG recognizing MAA-modifiedproteins in patients with advanced ALD [36]. Consideringthat appreciable amounts of acetaldehyde and MDA aregenerated in the liver during alcohol intoxication, it is wellpossible that MAA adducts might significantly contributeto immune response in ALD. It is noteworthy that rodentsand humans have circulating natural antibodies, mainly ofthe IgM class, targeting oxidation-derived epitopes includingMDA and MAA adducts [37, 38]. These natural antibodiesdisplay protective action against atherosclerosis by scaveng-ing oxidized LDL and preventing inflammation [37]. In ourhands, the antibody responses observed in ALD patientsspecifically involve only IgG, while IgM against MDA andMAA adducts do not differ from those in healthy controls[33, 36]. This suggests that the extensive lipid peroxidationcaused by alcohol abuse might overcome the scavengingcapacity of natural IgM antibodies and, in combination withinflammatory stimuli, favors the activation of B- and T-cellclones recognizing a variety of oxidation-derived antigens.Oxidative stress also likely accounts for the developmentof ALD-associated antiphospholipid antibodies, since theyrecognize as antigens oxidized phospholipids, namely, oxi-dized cardiolipin and phosphatidylserine [39, 40]. Inter-estingly, the presence of IgG targeting lipid-peroxidation-derived antigens is evident in patients with chronic hepatitisC (CHC) consuming moderate amounts of alcohol andincreases in a dose-dependent manner with alcohol intake[41]. This is consistent with recent observations about thesynergic action of ethanol and hepatitis C virus in promotingoxidative stress within the hepatocytes [42].

3. Mechanisms Possibly Involved in theRecruitment of Adaptive Immunity inAlcoholic Liver Disease

It is known since long time that excessive alcohol consump-tion affects the innate and adaptive immunity increasingthe susceptibility to infections and compromising tissue res-ponse to injury [5]. In particular, both acute and chronicalcohol intakes depress antigen presentation by monocytesand dendritic cells, affect the expression costimulatory mole-cules, and reduce T-cell proliferation [5]. It is possible that,during the evolution of ALD, the production of proinflam-matory cytokines/chemokines by Kupffer cells and NKTlymphocytes might overcome alcohol-dependent immunedepression promoting the response of intraportal lymphoidfollicles to antigens derived from oxidatively damaged


hepatocytes. Alcohol-induced oxidative stress can specif-ically facilitate this process through several mechanism.In either rodents and humans, chronic alcohol exposureincreases the levels of circulating bioactive oxidized phos-pholipids able to interact with plated activating factor (PAF)receptors and to stimulate inflammation [43]. Further-more, oxidized lipid and protein adducted by lipid peroxi-dation end products act as danger-associated molecularpatterns (DAMPs) and are capable of activating inflamma-tory and immune cells through the interaction with patternrecognition receptors (PRRs) such as scavenger receptors(SRA-1,2, CD36, SR-B1, and LOX-1) and toll-like receptors(TLR-4) [44]. In particular, CD36 efficiently recognizesfree and protein-bound oxidized lipids favoring their inter-nalization by macrophages and antigen-presenting cellswith their subsequent presentation to immune cells [44].The interaction between CD36 and TLR-4 also promotesproinflammatory activation of macrophages and vascularendothelial cells in response to oxidation-specific epitopes[44]. In the liver, the presentation of oxidative-stress-derivedantigens by hepatic stellate cells (HSCs) might represent anadditional pathway for CD4+ and CD8+ T-cell activation,as human HSCs are efficient professional antigen presentingcells [45] and have the capacity to internalize oxidativelymodified proteins through CD36 [46]. According to this sce-nario, time course experiments performed in enteral alcohol-fed rats show that the hepatic mRNAs expression of Th-1 cytokines (TNF-α, IL-12) has a biphasic pattern peakingafter 14 days of alcohol feeding and then rising again after 35days [47]. Oxidative stress is evident already after few days ofalcohol exposure, but lipid-peroxidation-derived antibodiesare evident in concomitance with the late increase cytokineproduction [47] and their formation is prevented by theantioxidant N-acetylcysteine [48]. Moreover, stimulation ofB cells through Toll-like receptor 9 (TLR-9) has been impli-cated in causing hyperimmunoglobulinemia in patients withalcoholic cirrhosis [49]. The immune responses in ALDmight be influenced by alterations of the immunoregulatorymechanisms. Hepatic steatosis and oxidative stress lower re-gulatory T-cells (Tregs) population in the liver [50]. Thismight impact on B- and T-cell activation by oxidation-speci-fic epitopes, as, during the evolution of atherosclerosis, Tregsimpairment favors lymphocyte responses against oxidizedLDL [51]. Finally, further stimuli to adaptive immunity inALD might involve osteopontin and the adipokines leptinand adiponectin. Osteopontin is a cytokine produced bymany cell types, including macrophages and T-lymphocytes,that promotes macrophage and T-cell activation, stimulateslymphocyte Th-1 and Th-17 differentiation, and induces B-cell proliferation and antibody production [52, 53]. Increas-ed osteopontin is a feature of several inflammatory andautoimmune diseases, where it directs the recruitment ofautoreactive T cells and lymphocyte Th-1 and Th-17 differ-entiation [53]. An increased liver production of osteopontincorrelates with the extent of inflammation in alcohol-fedrodents [54], while hepatic osteopontin mRNA expressionis higher in patients with alcoholic hepatitis than in heavydrinkers with fatty liver only [55]. Leptin and adiponectinoriginating from the adipose tissue have immunoregulatory

functions: leptin stimulates lymphocyte survival and prolif-eration favoring Th-1 reactions, while adiponectin downreg-ulates macrophage activity and B- and T-cells proliferation[56]. Alcohol influences adipocyte production of adipokinesincreasing serum leptin in patients with alcoholic liverdisease (ALD) [57] and lowering adiponectin secretion [58].

Little is known about the origin of antiphospholipidantibodies often associated with ALD. In the recent years,increasing evidence has linked defects in the disposal ofapoptotic cells with the development of antiphospholipidantibodies [59, 60]. In particular, phagocytosis of apoptoticbodies by immature dendritic cells in an inflammatorycontext can lead to the presentation of apoptotic cell-derivedself-antigens to T-lymphocytes [59, 60]. We have observ-ed that antiphospholipid antibodies from the sera of ALDpatients bind to apoptotic, but not to living cells, by specifi-cally targeting oxidized phosphatidylserine expressed on thecell surface [40]. This is consistent with the detection ofoxidized phosphatidylserine on surface of apoptotic bodies[61] and suggests that ALD-associated antiphospholipidantibodies may originate from alterations in the disposalof apoptotic hepatocytes. Indeed, growing evidence pointsto the importance of apoptosis-derived antigens as targetsof antiphospholipid antibodies [62]. Hepatocyte apoptosisis one of the consequence of chronic alcohol intake [30],and an increase in apoptotic bodies is well evident in liverbiopsies of ALD patients [63]. At the same time, alcoholimpairs the capacity of neighboring hepatocytes to disposeapoptotic cells through asialoglycoprotein receptor-mediatedphagocytosis [64] and primes Kupffer cells to produce TNF-α and IL-6 when exposed to apoptotic bodies [65].

The presence of both non-organ-specific and liver-specific autoantibodies is a common feature in ALD. Amongthese latter, we have reported that alcohol-fed rats as wellas about 40% of the patients with advanced ALD developcirculating IgG directed against CYP2E1 [66, 67]. Anti-CYP2E1 autoantibodies from ALD patients are similar tothose associated with halothane hepatitis and recognize atleast two distinct conformational epitopes on the moleculesurface, in a position compatible with the targeting ofCYP2E1 present on the outer layer of the hepatocyte plasmamembranes [68]. The development of anti-cytochrome P450(CYP) isoenzymes is rather common in liver diseases suchas type-2 autoimmune hepatitis, drug-induced hepatitis, andhepatitis C [69]. In drug-induced hepatitis, it has beenpostulated that the binding of reactive metabolites to CYPspromotes both humoral immune responses against the drug-derived epitope(s) and, at the same time, favors the activationof normally quiescent autoreactive lymphocytes recognizingthe native CYP molecules [69]. ALD patients with anti-HERantibodies have a 4-fold increased risk of developing anti-CYP2E1 autoreactivity as compared to patients without anti-HER IgG [67]. This indicates that CYP2E1 alkylation byHER is involved in the development of ALD anti-CYP2E1autoantibodies. Nonetheless, additional factors might favorthe breaking of self-tolerance in ALD. We have observed thata polymorphism (Thr17→Ala substitution) in the cytotoxic-T-lymphocyte-associated antigen-4 (CTLA-4) gene increasesby 3.8-fold the risk of developing anti-CYP2E1 IgG without


influencing the formation of anti-HER antibodies [67]. ALDpatients having both anti-HER IgG and mutated CTLA-4 show a prevalence of anti-CYP2E1 autoreactivity 23-foldhigher than those negative for both these factors [67].CTLA-4 is a membrane receptor expressed by activated T-lymphocytes and by CD25+ CD4+ Tregs that downmodulatesT-cell-mediated responses to antigens [70]. Accordingly, inhumans, CTLA-4 genetic polymorphisms are risk factors forseveral autoimmune diseases, including primary biliary cir-rhosis and type-1 autoimmune hepatitis [71]. More recently,Thiele and coworkers have shown that mice immunizationwith MAA-modified liver cytosolic proteins in the absenceof adjuvants is able to promote autoimmune injury in theliver [72]. This indicates that, in some ALD patients, thecombination of CYP2E1 modification by HERs, impairedCTLA-4 control of T-cell proliferation, and the stimulationof immune system by lipid peroxidation products can leadto the breaking of self-tolerance. Preliminary data shows thathigh titers of anti-CYP2E1 autoantibodies correlate with theextension of lymphocyte infiltration and the frequency ofapoptotic hepatocytes, suggesting that in a subset of ALDautoimmune mechanisms might contribute to tissue injury.

4. Possible Role of Adaptive Immunityin the Progression of Alcohol Liver Damage

So far little is known about the mechanisms by which adap-tive immunity might contribute to hepatic inflammation inALD. However, data emerging from experimental modelsof atherosclerosis indicate that IFN-γ, TNF-α, and CD40ligand (CD154) produced by Th-1 CD4+ T-cell responsiveto antigens in oxidized LDL drive plaque macrophages toproduce reactive oxygen species, NO, and proinflamma-tory cyto/chemokines [73]. Studies in alcohol-fed rodentsshow that the development of antibodies against lipid-peroxidation-derived epitopes associate with the hepaticexpression of proinflammatory cytokines and histologicalevidence of steatohepatitis [47, 48]. In humans, a prospectivesurvey has evidenced an association between the presenceof antibodies toward alcohol-modified hepatocytes and anincreased risk of developing alcoholic liver cirrhosis [74].On the same line, heavy drinkers with lipid-peroxidation-derived antibodies have a 5-fold higher prevalence of elevatedplasma TNF-α levels than alcohol abusers with these anti-bodies within the control range [75]. Moreover, in thesesubjects, the combination of high TNF-α and lipid-per-oxidation-induced antibodies increases by 11-fold the risk ofdeveloping advanced ALD [75]. Interestingly, the combina-tion of steatosis and high titers of antibodies against lipid-peroxidation-derived adducts is an independent predictorof advanced fibrosis/cirrhosis in alcohol-consuming patientswith chronic hepatitis C [76]. Further evidence in favor of apossible contribution of oxidative-stress-mediated immunityin sustaining hepatic inflammation comes from studies innonalcoholic steatohepatitis (NASH). We have observed thatantibodies against lipid-peroxidation-derived antigens simi-lar to those detected in ALD, are present in rodent models ofNASH [77] as well as in about 40% of adult NAFLD/NASH

patients [78] and in 60% of children with NASH [79].Experiments comparing the development of NASH inducedby mice feeding with a methionine/choline deficient (MCD)diet demonstrate that C57BL/6, but not BALB/C, micehave an increased prevalence of liver infiltrating T and Bcells and generate antibodies against MDA adducts. Hepaticnecroinflammation and the expression of TNF-α and IL-12are significantly higher in C57BL/6 mice than in BALB/Cmice. Among MCD-treated mice, a significant positivecorrelation is also evident between the titers of anti-MDAantibodies and the number of hepatic necroinflammatoryfoci [80]. In line with these findings, in children withNAFLD, high titers of anti-MDA adduct IgG, but not ofother oxidative stress markers, are associated with moresevere lobular inflammation and with 13-fold increasedrisk of overt steatohepatitis [79]. We have observed that inmice with NASH, anti-MDA IgG target specific antigens inhepatic necroinflammatory foci leading to the formationof immunocomplexes. Furthermore, Rensen and coworkershave detected extensive deposition of different complementfractions in liver biopsies form NASH patients that associateswith increased hepatocyte apoptosis, granulocyte infiltrationand higher liver expression of IL-1β, IL-6, and IL-8 mRNAs[81]. This suggests the possibility that complement activationmight bridge adaptive and innate immune responses duringALD evolution. Supporting this view, recent observations byNagy’s group demonstrate complement activation within theliver of alcohol-fed mice that leads to increased productionof C3a anaphylatoxin [82]. Moreover, mice deficient of C3and C5 are protected against hepatic injury, while the lack ofthe complement-regulating protein CD55/DAF exacerbatesalcohol hepatotoxicity [82].

5. Conclusion and Future Perspectives

In conclusion, data so far obtained indicates that alcohol-induced oxidative modifications of hepatic constituentstrigger specific immune responses and, in some conditions,may favor the breaking of the self-tolerance toward liver con-stituents. The development of adaptive immune responsesis likely favored by the capacity of alcohol to stimulate bydifferent ways innate immunity [4–8]. Although still indirect,available evidence also suggests that the development ofhumoral and cellular immunity may contribute to hepaticinflammation during the evolution of ALD. Nonetheless,during the progression of ALD to cirrhosis, additionalmechanisms may be at work. Recent data indicate that theinteractions between myofibroblast-like hepatic stellate cells(HSC/MFs), Kupffer cells, and CD4+ T-lymphocytes are cri-tical for the regulation of fibrogenic responses [83, 84]. Inparticular, by expressing adhesion molecules (VCAM-1 andICAM-1) and chemokines (CCL2, CXCL9/10, CXCL16, andfractalin), HSC/MFs are essential for the liver recruitmentof lymphocytes and provide the milieu for their survivaland activation [84]. On the other hand, Th-2 cytokines(IL-4 and IL-13) from CD4+ T cells directly sustain matrixdeposition by HSC/MFs and favor “alternative” M2 activa-tion of macrophages with the production of chemokines


(CCL6 and CCL17), IL-10, TGF-β, and VEGFs that drivehealing responses and angiogenesis [83, 84]. At present, itis not known whether T-cell stimulation by oxidative-stress-derived antigens might promote Th-2 lymphocyte develop-ment and M2 macrophage polarization, contributing tofibrogenesis.

From the clinical point of view, prospective studies arerequired to dissect out the precise role of immune responsesin the progression of human ALD. If supported by furtherdata, the concept that adaptive immunity has a role in alcoholhepatotoxicity might lead to the development of simple im-munometric assays to discriminate ALD patients at risk ofprogressing to hepatitis and/or fibrosis and moreover, theidentification of alcohol abusers with a prominent immuneor autoimmune component in their hepatic disease thatmight lead to a targeted use of immune-suppressive therapyin ALD.

Conflict of Interests

The author has no conflict of interests on the matter concern-ing the present paper.

Acknowledgment

The author’s researches have been supported by grants fromthe Regional Government of Piedmont (Ricerca Sanitaria Fi-nalizzata 2002–04 and Ricerca Scientifica Applicata 2004).

References

[1] WHO, “Global status report on alcohol 2004-country profile,”2004, http://www.who.int/substance abuse/publications/.

[2] N. Sheron, N. Olsen, and I. Gilmore, “An evidence-basedalcohol policy,” Gut, vol. 57, no. 10, pp. 1341–1344, 2008.

[3] E. Albano, “Alcohol-induced hepatic injury,” in EndogenousToxins, P. J. O’Brien and W. R. Bruce, Eds., vol. 2, pp. 451–483,Wiley VCH, Weinheim, Germany, 2009.

[4] I. N. Hines and M. D. Wheeler, “Recent advances in alcoholicliver disease III. Role of the innate immune response inalcoholic hepatitis,” American Journal of Physiology, vol. 287,no. 2, pp. G310–G314, 2004.

[5] G. Szabo and P. Mandrekar, “A recent perspective on alcohol,immunity, and host defense,” Alcoholism: Clinical and Experi-mental Research, vol. 33, no. 2, pp. 220–232, 2009.

[6] R. Rao, “Endotoxemia and gut barrier dysfunction in alcoholicliver disease,” Hepatology, vol. 50, no. 2, pp. 638–644, 2009.

[7] B. Gao, E. Seki, D. A. Brenner et al., “Innate immunity inalcoholic liver disease,” American Journal of Physiology, vol.300, no. 4, pp. G516–G525, 2011.

[8] A. M. Miller, N. Horiguchi, W. I. Jeong, S. Radaeva, and B.Gao, “Molecular mechanisms of alcoholic liver disease: innateimmunity and cytokines,” Alcoholism: Clinical and Experimen-tal Research, vol. 35, no. 5, pp. 787–793, 2011.

[9] A. Chedid, C. L. Mendenhall, T. E. Moritz et al., “Cell-mediated hepatic injury in alcoholic liver disease,” Gastroen-terol, vol. 105, no. 1, pp. 254–266, 1993.

[10] K. Song, R. A. Coleman, C. Alber et al., “TH1 cytokine res-ponse of CD57+ T-cell subsets in healthy controls and patientswith alcoholic liver disease,” Alcohol, vol. 24, no. 3, pp. 155–167, 2001.

[11] K. Song, R. A. Coleman, X. Zhu et al., “Chronic ethanolconsumption by mice results in activated splenic T cells,”Journal of Leukocyte Biology, vol. 72, no. 6, pp. 1109–1116,2002.

[12] J. F. Laso, I. J. Madruga, and A. Orfao, “Cytokines and alcoholliver disease,” in Ethanol and the Liver, C. D. I. N. Sherman,V. R. Preedy, and R. R. Watson, Eds., pp. 206–219, Taylor andFrancis, London, UK, 2002.

[13] F. Paronetto, “Immunologic reactions in alcoholic liver dis-ease,” Seminars in Liver Disease, vol. 13, no. 2, pp. 183–195,1993.

[14] C. A. Laskin, E. Vidins, L. M. Blendis, and C. A. Soloninka,“Autoantibodies in alcoholic liver disease,” American Journalof Medicine, vol. 89, no. 2, pp. 129–133, 1990.

[15] I. G. McFarlane, “Autoantibodies in alcoholic liver disease,”Addiction Biology, vol. 5, no. 2, pp. 141–151, 2000.

[16] A. Chedid, K. R. Chadalawada, T. R. Morgan et al., “Phospho-lipid antibodies in alcoholic liver disease,” Hepatology, vol. 20,no. 6, pp. 1465–1471, 1994.

[17] C. Biron, N. Lalloyer, J. L. Tonnelot, D. Larget, J. C. Gris, andJ. F. Schved, “Anticardiolipin antibodies and acute alcoholicintoxication,” Lupus, vol. 4, no. 6, pp. 486–490, 1995.

[18] A. Lemmers, C. Moreno, T. Gustot et al., “The interleukin-17 pathway is involved in human alcoholic liver disease,”Hepatology, vol. 49, no. 2, pp. 646–657, 2009.

[19] L. Hammerich, F. Heymann, and F. Tacke, “Role of IL-17and Th17 cells in liver disease,” Clinical and DevelopmentalImmunology, vol. 2011, Article ID 345803, 12 pages, 2011.

[20] Y. Israel, E. Hurwitz, O. Niemela, and R. Arnon, “Monoclonaland polyclonal antibodies against acetaldehyde-containingepitopes in acetaldehyde-protein adducts,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 83, no. 20, pp. 7923–7927, 1986.

[21] O. Niemela, F. Klajner, H. Orrego, E. Vidinis, L. Blendis, andY. Israel, “Antibodies against acetaldehyde-modified proteinepitopes in human alcoholics,” Hepatology, vol. 7, no. 6, pp.1210–1214, 1987.

[22] J. Koskinas, J. G. Kenna, G. L. Bird, G. J. M. Alexander, and R.Williams, “Immunoglobulin A antibody to a 200-kilodaltoncytosolic acetaldehyde adduct in alcoholic hepatitis,” Gas-troenterology, vol. 103, no. 6, pp. 1860–1867, 1992.

[23] H. Yokoyama, H. Ishii, S. Nagata, S. Kato, K. Kamegaya, andM. Tsuchiya, “Experimental hepatitis induced by ethanol afterimmunization with acetaldehyde adducts,” Hepatology, vol.17, no. 1, pp. 14–19, 1993.

[24] L. W. Klassen, D. Tuma, and M. F. Sorrell, “Immune mech-anisms of alcohol-induced liver disease,” Hepatology, vol. 22,no. 1, pp. 355–357, 1995.

[25] C. Moncada, V. Torres, E. Varghese, E. Albano, and Y. Israel,“Ethanol-derived immunoreactive species formed by freeradical mechanisms,” Molecular Pharmacology, vol. 46, no. 4,pp. 786–791, 1994.

[26] P. Clot, G. Bellomo, M. Tabone, S. Arico, and E. Albano,“Detection of antibodies against proteins modified by hydrox-yethyl free radicals in patients with alcoholic cirrhosis,”Gastroenterology, vol. 108, no. 1, pp. 201–207, 1995.

[27] E. Albano, P. Clot, M. Morimoto, T. Tomasi, M. Ingelman-Sundberg, and S. W. French, “Role of cytochrome P4502E1-dependent formation of hydroxyethyl free radical in thedevelopment of liver damage in rats intragastrically fed withethanol,” Hepatology, vol. 23, no. 1, pp. 155–163, 1996.

[28] E. Albano, S. W. French, and M. Ingelman-Sundberg,“Hydroxyethyl radicals in ethanol hepatotoxicity,” Frontiers inBioscience, vol. 4, pp. D533–D540, 1999.


[29] P. Clot, E. Albano, E. Elliasson et al., “Cytochrome P4502E1hydroxyethyl radical adducts as the major antigenic deter-minant for autoantibody formation among alcoholics,” Gas-troenterology, vol. 111, pp. 206–216, 1996.

[30] E. Albano, “Alcohol, oxidative stress and free radical damage,”Proceedings of the Nutrition Society, vol. 65, no. 3, pp. 278–290,2006.

[31] W. Palinski and J. L. Witztum, “Immune responses to oxidativeneoepitopes on LDL and phospholipids modulate the devel-opment of atherosclerosis,” Journal of Internal Medicine, vol.247, no. 3, pp. 371–380, 2000.

[32] B. T. Kurien and R. H. Scofield, “Autoimmunity and oxida-tively modified autoantigens,” Autoimmunity Reviews, vol. 7,no. 7, pp. 567–573, 2008.

[33] E. Mottaran, S. F. Stewart, R. Rolla et al., “Lipid peroxidationcontributes to immune reactions associated with alcoholicliver disease,” Free Radical Biology and Medicine, vol. 32, no.1, pp. 38–45, 2002.

[34] S. F. Stewart, M. Vidali, C. P. Day, E. Albano, and D. E. J. Jones,“Oxidative stress as a trigger for cellular immune responses inpatients with alcoholic liver disease,” Hepatology, vol. 39, no.1, pp. 197–203, 2004.

[35] G. M. Thiele, T. K. Freeman, and L. W. Klassen, “Immunologicmechanisms of alcoholic liver injury,” Seminars in LiverDisease, vol. 24, no. 3, pp. 273–287, 2004.

[36] R. Rolla, D. Vay, E. Mottaran et al., “Detection of circulatingantibodies against malondialdehyde-acetaldehyde adducts inpatients with alcohol-induced liver disease,” Hepatology, vol.31, no. 4, pp. 878–884, 2000.

[37] C. J. Binder, P. X. Shaw, M. K. Chang et al., “The role of naturalantibodies in atherogenesis,” Journal of Lipid Research, vol. 46,no. 7, pp. 1353–1363, 2005.

[38] M. Y. Chou, L. Fogelstrand, K. Hartvigsen et al., “Oxidation-specific epitopes are dominant targets of innate natural anti-bodies in mice and humans,” Journal of Clinical Investigation,vol. 119, no. 5, pp. 1335–1349, 2009.

[39] R. Rolla, D. Vay, and E. Mottaran, “Anti-phospholipidantibodies associated with alcoholic liver disease specificallyrecognize oxidized phospholipids,” Gut, vol. 49, no. 6, pp. 852–859, 2001.

[40] D. Vay, C. Rigamonti, M. Vidali et al., “Anti-phospholipid anti-bodies associated with alcoholic liver disease target oxidizedphosphatidylserine on apoptotic cell plasma membranes,”Journal of Hepatology, vol. 44, no. 1, pp. 183–189, 2006.

[41] C. Rigamonti, E. Mottaran, E. Reale et al., “Moderate alcoholconsumption increases oxidative stress in patients with chro-nic hepatitis C,” Hepatology, vol. 38, no. 1, pp. 42–49, 2003.

[42] J. Choi and J. H. J. Ou, “Mechanisms of liver injury: III.Oxidative stress in the pathogenesis of hepatitis C virus,”American Journal of Physiology, vol. 290, no. 5, pp. G847–G851, 2006.

[43] L. Yang, C. Latchoumycandane, M. R. McMullen et al.,“Chronic alcohol exposure increases circulating bioactiveoxidized phospholipids,” Journal of Biological Chemistry, vol.285, no. 29, pp. 22211–22220, 2010.

[44] Y. I. Miller, S. H. Choi, P. Wiesner et al., “Oxidation-specificepitopes are danger-associated molecular patterns recognizedby pattern recognition receptors of innate immunity,” Circula-tion Research, vol. 108, no. 2, pp. 235–248, 2011.

[45] F. Winau, C. Quack, A. Darmoise, and S. H. Kaufmann,“Starring stellate cells in liver immunology,” Current Opinionin Immunology, vol. 20, no. 1, pp. 68–74, 2008.

[46] W. Schneiderhan, A. Schmid-Kotsas, J. Zhao et al., “Oxidizedlow-density lipoproteins bind to the scavenger receptor, CD36,

of hepatic stellate cells and stimulate extracellular matrixsynthesis,” Hepatology, vol. 34, no. 4, pp. 729–737, 2001.

[47] M. J. J. Ronis, A. Butura, S. Korourian et al., “Cytokineand chemokine expression associated with steatohepatitis andhepatocyte proliferation in rats fed ethanol via total enteralnutrition,” Experimental Biology and Medicine, vol. 233, no. 3,pp. 344–355, 2008.

[48] M. J. J. Ronis, A. Butura, B. P. Sampey et al., “Effects of N-acetylcysteine on ethanol-induced hepatotoxicity in rats fedvia total enteral nutrition,” Free Radical Biology and Medicine,vol. 39, no. 5, pp. 619–630, 2005.

[49] B. Massonnet, A. Delwail, J. M. Ayrault, C. Chagneau-Derrode, J. C. Lecron, and C. Silvain, “Increased immunoglob-ulin A in alcoholic liver cirrhosis: exploring the responseof B cells to Toll-like receptor 9 activation,” Clinical andExperimental Immunology, vol. 158, no. 1, pp. 115–124, 2009.

[50] X. Ma, J. Hua, A. R. Mohamood, A. R. A. Hamad, R. Ravi,and Z. Li, “A high-fat diet and regulatory T cells influencesusceptibility to endotoxin-induced liver injury,” Hepatology,vol. 46, no. 5, pp. 1519–1529, 2007.

[51] J. George, “Mechanisms of Disease: the evolving role ofregulatory T cells in atherosclerosis,” Nature Clinical PracticeCardiovascular Medicine, vol. 5, no. 9, pp. 531–540, 2008.

[52] E. M. Gravallese, “Osteopontin: a bridge between bone and theimmune system,” Journal of Clinical Investigation, vol. 112, no.2, pp. 147–149, 2003.

[53] J. Morimoto, S. Kon, Y. Matsui, and T. Uede, “Osteopontin; asa target molecule for the treatment of inflammatory diseases,”Current Drug Targets, vol. 11, no. 4, pp. 494–505, 2010.

[54] S. H. Ramaiah and S. Ritting, “Role of osteopontin inregulating hepatic inflammatory responses and toxic liverinjury,” Expert Opinion on Drug Metabolism and Toxicology,vol. 3, no. 4, pp. 519–526, 2007.

[55] D. Seth, M. D. Gorrell, S. Cordoba, G. W. McCaughan, andP. S. Haber, “Intrahepatic gene expression in human alcoholichepatitis,” Journal of Hepatology, vol. 45, no. 2, pp. 306–320,2006.

[56] H. Tilg and A. R. Moschen, “Adipocytokines: mediatorslinking adipose tissue, inflammation and immunity,” NatureReviews Immunology, vol. 6, no. 10, pp. 772–783, 2006.

[57] S. Naveau, G. Perlemuter, M. Chaillet et al., “Serum leptin inpatients with alcoholic liver disease,” Alcoholism: Clinical andExperimental Research, vol. 30, no. 8, pp. 1422–1428, 2006.

[58] S. A. Polyzos, K. A. Toulis, D. G. Goulis, C. Zavos, andJ. Kountouras, “Serum total adiponectin in nonalcoholicfatty liver disease: a systematic review and meta-analysis,”Metabolism, 2010.

[59] M. R. Elliott and K. S. Ravichandran, “Clearance of apoptoticcells: implications in health and disease,” Journal of CellBiology, vol. 189, no. 7, pp. 1059–1070, 2010.

[60] D. V. Krysko, K. D’Herde, and P. Vandenabeele, “Clearanceof apoptotic and necrotic cells and its immunological conse-quences,” Apoptosis, vol. 11, no. 10, pp. 1709–1726, 2006.

[61] V. E. Kagan, G. G. Borisenko, B. F. Serinkan et al., “Appetizingrancidity of apoptotic cells for macrophages: oxidation, exter-nalization and recognition of phosphatidylserine,” AmericanJournal of Physiology, vol. 285, pp. L1–L17, 2003.

[62] C. Alessandri, F. Conti, M. Pendolino, R. Mancini, and G.Valesini, “New autoantigens in the antiphospholipid syn-drome,” Autoimmunity Reviews, vol. 10, no. 10, pp. 609–616,2011.

[63] S. Natori, C. Rust, L. M. Stadheim, A. Srinivasan, L. J. Burgart,and G. J. Gores, “Hepatocyte apoptosis is a pathologic featureof human alcoholic hepatitis,” Journal of Hepatology, vol. 24,no. 2, pp. 248–253, 2001.


[64] B. L. McVicker, D. J. Tuma, J. A. Kubik, A. M. Hindemith, C. R.Baldwin, and C. A. Casey, “The effect of ethanol on asialogly-coprotein receptor-mediated phagocytosis of apoptotic cellsby rat hepatocytes,” Hepatology, vol. 36, no. 6, pp. 1478–1487,2002.

[65] B. L. McVicker, D. J. Tuma, K. K. Kharbanda, J. L. Kubik,and C. A. Casey, “Effect of chronic ethanol administration onthe in vitro production of proinflammatory cytokines by ratKupffer cells in the presence of apoptotic cells,” Alcoholism:Clinical and Experimental Research, vol. 31, no. 1, pp. 122–129,2007.

[66] S. D. Lytton, A. Helander, Z. Q. Zhang-Gouillon et al.,“Autoantibodies against cytochromes P-4502E1 and P-4503Ain alcoholics,” Molecular Pharmacology, vol. 55, no. 2, pp. 223–233, 1999.

[67] M. Vidali, S. F. Stewart, R. Rolla et al., “Genetic and epigeneticfactors in autoimmune reactions toward cytochrome P4502E1in alcoholic liver disease,” Hepatology, vol. 37, no. 2, pp. 410–419, 2003.

[68] M. Vidali, M. Hidestrand, E. Eliasson et al., “Use of molecularsimulation for mapping conformational CYP2E1 epitopes,”Journal of Biological Chemistry, vol. 279, no. 49, pp. 50949–50955, 2004.

[69] D. P. Bogdanos and G. N. Dalekos, “Enzymes as target antigensof liver-specific autoimmunity: the case of cytochromesP450s,” Current Medicinal Chemistry, vol. 15, no. 22, pp. 2285–2292, 2008.

[70] K. J. Scalapino and D. I. Daikh, “CTLA-4: a key regulatorypoint in the control of autoimmune disease,” ImmunologicalReviews, vol. 223, no. 1, pp. 143–155, 2008.

[71] S. C. L. Gough, L. S. K. Walker, and D. M. Sansom,“CTLA4 gene polymorphism and autoimmunity,” Immuno-logical Reviews, vol. 204, pp. 102–115, 2005.

[72] G. M. Thiele, M. J. Duryee, M. S. Willis et al., “Autoimmunehepatitis induced by syngeneic liver cytosolic proteins bio-transformed by alcohol metabolites,” Alcoholism: Clinical andExperimental Research, vol. 34, no. 12, pp. 2126–2136, 2010.

[73] G. K. Hansson and P. Libby, “The immune responsein atherosclerosis: a double-edged sword,” Nature ReviewsImmunology, vol. 6, no. 7, pp. 508–519, 2006.

[74] S. Takase, M. Tsutsumi, H. Kawahara, N. Takada, and A.Takada, “The alcohol-altered liver membrane antibody andhepatitis C virus infection in the progression of alcoholic liverdisease,” Hepatology, vol. 17, no. 1, pp. 9–13, 1993.

[75] M. Vidali, J. Hietala, G. Occhino et al., “Immune responsesagainst oxidative stress-derived antigens are associated withincreased circulating tumor necrosis factor-α and acceleratedliver damage in heavy drinkers,” Free Radical Biology andMedicine, vol. 45, no. 3, pp. 306–311, 2008.

[76] M. Vidali, G. Occhino, A. Ivaldi, C. Rigamonti, M. Sartori, andE. Albano, “Combination of oxidative stress and steatosis is arisk factor for fibrosis in alcohol-drinking patients with chro-nic hepatitis C,” American Journal of Gastroenterology, vol. 103,no. 1, pp. 147–153, 2008.

[77] J. N. Baumgardner, K. Shankar, L. Hennings, E. Albano, T.M. Badger, and M. J. J. Ronis, “N-acetylcysteine attenuatesprogression of liver pathology in a rat model of nonalcoholicsteatohepatitis,” Journal of Nutrition, vol. 138, no. 10, pp.1872–1879, 2008.

[78] E. Albano, E. Mottaran, M. Vidali et al., “Immune responsetowards lipid peroxidation products as a predictor of progres-sion of non-alcoholic fatty liver disease to advanced fibrosis,”Gut, vol. 54, no. 7, pp. 987–993, 2005.

[79] V. Nobili, M. Parola, A. Alisi et al., “Oxidative stressparameters in paediatric non-alcoholic fatty liver disease,”International Journal of Molecular Medicine, vol. 26, no. 4, pp.471–476, 2010.

[80] V. Maina, S. Sutti, M. Vidali et al., “Strain differences inmice susceptibility to steatohepatitis induced by methion-ine/choline deficient diet underscore a role of immune mech-anisms in NASH,” Journal of Hepatology, vol. 52, supplement1, p. S308, 2010.

[81] S. S. Rensen, Y. Slaats, A. Driessen et al., “Activation ofthe complement system in human nonalcoholic fatty liverdisease,” Hepatology, vol. 50, no. 6, pp. 1809–1817, 2009.

[82] J. I. Cohen, X. Chen, and L. E. Nagy, “Redox signaling and theinnate immune system in alcoholic liver disease,” Antioxidants& Redox Signaling, vol. 15, no. 2, pp. 523–534, 2001.

[83] A. P. Holt, M. Salmon, C. D. Buckley, and D. H. Adams,“Immune interactions in hepatic fibrosis,” Clinics in LiverDisease, vol. 12, no. 4, pp. 861–882, 2008.

[84] T. A. Wynn and L. Barron, “Macrophages: master regulators ofinflammation and fibrosis,” Seminars in Liver Disease, vol. 30,no. 3, pp. 245–257, 2010.


Review Article

Aberrant Hepatic Methionine Metabolism andGene Methylation in the Pathogenesis and Treatment ofAlcoholic Steatohepatitis

Charles H. Halsted and Valentina Medici

Department of Internal Medicine, University of California, Davis, 451 E. Health Sciences Drive, Room 6323, Davis,CA 95616, USA

Correspondence should be addressed to Charles H. Halsted, [email protected]

Received 26 May 2011; Accepted 8 July 2011


Copyright © 2012 C. H. Halsted and V. Medici. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

The pathogenesis of alcoholic steatohepatitis (ASH) involves ethanol-induced aberrations in hepatic methionine metabolism thatdecrease levels of S-adenosylmethionine (SAM), a compound which regulates the synthesis of the antioxidant glutathione and isthe principal methyl donor in the epigenetic regulation of genes relevant to liver injury. The present paper describes the effectsof ethanol on the hepatic methionine cycle, followed by evidence for the central role of reduced SAM in the pathogenesis of ASHaccording to clinical data and experiments in ethanol-fed animals and in cell models. The efficacy of supplemental SAM in theprevention of ASH in animal models and in the clinical treatment of ASH will be discussed.

1. Introduction

Alcoholic steatohepatitis (ASH) represents the intermedi-ate stage of liver injury in the progression of steatosis,inflammation, and hepatocellular apoptosis and necrosis thatoccurs in the development of alcoholic cirrhosis. This paperwill address the role of ethanol-induced aberrant hepaticmethionine metabolism in the pathogenesis of ASH, inparticular through its epigenetic effects on the expressionsof genes relevant to alcoholic liver injury. Initially, we willdiscuss the normal hepatic methionine metabolic cycle,including its interactions with folate, vitamin B12, andvitamin B6, and its relevance to the epigenetic regulationof gene expression. This will be followed by experimentalevidence for multiple effects of ethanol on the methioninemetabolic cycle that lead to liver injury through alteredmethylation and expressions of genes relevant to steatosis,apoptosis, and oxidative stress. We will summarize datafrom experimental and clinical studies that support theseconcepts.

2. Normal Hepatic Methionine Metabolism

Hepatic methionine metabolism can be visualized in twoparts, the transmethylation cycle for production of S-adenosylmethionine (SAM) and its metabolism to homo-cysteine and the transsulfuration pathway for the reduc-tion of homocysteine for the synthesis of glutathione(GSH) (Figure 1). Endogenous 5-methyltetrahydrofolate (5-MTHF) that is derived from dietary folate is substrate forthe initial vitamin-B12-dependent reaction of methioninesynthase (MS) that converts homocysteine to methionine.The additional pathway of betaine homocysteine methyl-transferase (BHMT) uses the choline product betaine assubstrate with homocysteine for methionine synthesis andis considered a salvage pathway when MS is compromisedby toxins including ethanol. Methionine is metabolized toSAM by methionine adenosyltransferase (MAT), which is aproduct of the MAT1A gene in adult liver. SAM is utilized ata rate of 6–8 grams per day [1], primarily by different methyl-transferase reactions that include phosphatidylethanolamine


SAM

SAH

DNA

SAM

GSH

Homocysteine

Betaine

Methionine

5-MTHF

Methyl-DNA,methyl-histones

SAHH

MTs

Cystathionase

ABU

(B6)

(B6)

(B12)

Cysteine

Cystathionine

PEMT

CβS

MAT

THF

MS

DMG

+

PC

PE

BHMT

Figure 1: Normal hepatic methionine metabolism. 5-MTHF:5-methyltetrahydrofolate; THF: tetrahydrofolate; MS: methion-ine synthase; BHMT: betaine hydroxy methyltransferase; DMG:dimethylglycine; MAT: methionine adenosyl transferase; SAM: S-adenosylmethionine; MT: methyltransferase; PE: phosphatidyleth-anolamine; PEMT: phosphatidylethanolamine methyltransferase;PC: phosphatidylcholine; SAH: S-adenosylhomocysteine; SAHH:SAH hydrolase; CβS: cystathionine beta synthase; ABU: α-amino-butyrate; GSH: glutathione.

transferase (PEMT) to yield phosphatidylcholine (PC) andmany others that methylate gene-specific DNA and varioushistone residues. In addition, SAM levels are regulated bythe activity of glycine-n-methyltransferase (GNMT) withsarcosine as its end product (not shown in Figure 1). WhileSAM is the sole substrate for methyltransferase reactions,its product S-adenosylhomocysteine (SAH) is the principalinhibitor of the same reactions. Consequently, the SAM-to-SAH ratio is used conveniently as an index of methylationcapacity since the Km and Ki for most of these reactionsare in similar range [2]. SAH is hydrolyzed to homocysteineby SAH hydrolase (SAHH), which is a reversible enzymedepending upon the abundance of each product. The initialreaction in the transsulfuration pathway is facilitated by SAMand involves the reduction of homocysteine through thecystathionine beta synthase (CβS) reaction to cystathionine[3]. Both CβS and the subsequent cystathionase reactionrequire vitamin B6 as cofactor. Therefore, it may be predictedthat conditions that limit the production of SAM and theavailability of vitamin B6 reduce the production of GSHand its antioxidant properties. Summarizing the key points,homocysteine is regulated by three enzymes, MS, SAHH,and CβS, each of which can be compromised to result inhomocysteine elevation which is involved in the pathogenesisof ASH. The metabolite SAM plays a key role in regulating

gene methylation and expression and a second one byregulating the production of the principal antioxidant GSH.

3. Effects of Ethanol Exposure on HepaticMethionine Metabolism

There is abundant and accumulating evidence that experi-mental ethanol exposure and chronic alcoholism influencethe expressions and activities of several enzymes in thehepatic methionine cycle. For example, transcript expres-sions of MS, MAT1A, and CβS were decreased in liverbiopsies from ASH patients [4], while decreased expressionsand activities of MS, MAT1A, and SAHH, but increasedGNMT, were found in chronic ethanol-fed micropigs withhistologically proven ASH [5]. Reduced activities of MS withcompensatory increase in BHMT expressions and activitieswere also found in ethanol-fed rats [6–8]. Since MAT1Ais susceptible to oxidation and nitrosylation of its cysteine121 residue, its reduced activity in ASH may be attributedto the generation of reactive oxygen species (ROS) that areproduced by the ethanol metabolizing enzyme cytochromep4502E1 (CYP2E1) [9, 10]. Thus, since SAM regulates thegeneration of GSH through its facilitating effect on CβS[3], the reduced antioxidant potential caused by reducedGSH would predictably decrease the effectiveness of MAT1Ato generate SAM. These relationships were supported byin vitro studies in chemically treated primary hepatocytesin which elevation of prooxidant CYP2E1 was associatedwith reduced SAM levels [11]. Also, studies in ethanol-fedbaboons and micropigs showed close correlations betweenliver levels of GSH and SAM [12, 13].

4. Effects of Aberrant Methionine Metabolismon the Pathogenesis of ASH

The conventional understanding of the pathogenesis of ASHincludes a series of events that are triggered by ethanol-induced transport of enterotoxic lipopolysaccharide (LPS)from the intestine to hepatic Kupffer cells that are inturn induced to release cytokines such as tumor necrosisfactor alpha (TNFα) which initiates pathways of hepatocyteapoptosis, inflammation, and necrosis. These processes arefacilitated by the effect of the ethanol metabolizing hepato-cyte enzyme CYP2E1 which is a powerful producer of ROS[14]. In addition, the metabolism of ethanol by alcohol toacetaldehyde alters the NADH/NAD redox potential whichpromotes steatosis through its effects on fatty acid oxidationin the liver [15]. Steatosis, which is an early sign in devel-opment of ASH, represents the combination of increasedhepatic lipogenesis and decreased fatty acid oxidation andreduced triglyceride export. Many of these avenues of injuryhave been linked to ethanol-induced aberrant methioninemetabolism in experimental animal models, as describedbelow.

Studies that used the model of the intragastric ethanol-fed C57BL6 mouse linked enhanced lipogenesis and apop-tosis to hyperhomocysteinemia, which is known to triggerendoplasmic reticulum (ER) stress pathways that include


upregulation of the chaperone glucose-regulated protein(GRP78), the transcription factor sterol response elementbinding protein (SREBP1-c) and its downstream lipogenesisgenes, and the growth arrest and DNA damage (GADD153)pathway for apoptosis. Since the histopathology of steatosisand apoptosis and these gene activations were prevented byconcomitant provision of betaine which lowered homocys-teine through the BHMT salvage pathway (Figure 1), theauthors concluded that there is a specific role for ethanol-induced hyperhomocysteinemia in the pathogenesis of ASHby activation of ER stress pathways [16]. More recent studiesin the ethanol-fed rat model of ASH associated apoptosiswith increased liver SAH and steatosis with reduced activityof PEMT which regulates triglyceride export from the liver[17–19]. Others associated apoptosis in ethanol-fed micewith increased levels of SAH which sensitized isolatedprimary hepatocytes to the effects of TNFα [20], while alater study in the same model associated this effect withreduced mitochondrial SAM (mSAM) caused by competitiveinhibition of its mitochondrial transport by elevated SAH[21]. The original finding of a mitochondrial transporterfor cytoplasmic SAM that is inhibited by SAH [22] wascomplemented by discovery of a mitochondrial GSH trans-porter that is inactivated by ethanol but is sustained bycoadministration of SAM [23]. A role for mSAM in preven-tion of the prooxidant effects of ethanol was supported byintervention studies that demonstrated a protective effect ofSAM against decreased mitochondrial respiration [24] andincreased levels of mitochondrial superoxide and induciblenitric oxide synthase (iNOS) [25]. SAM may also play a rolein prevention of hepatocellular carcinoma (HCC) in ALD byinhibition of angiogenesis genes [26].

5. Role of Altered Gene Methylation Capacity inthe Pathogenesis of ASH

The pathogenesis of ASH includes activations of many genesinvolved in liver injury, each of which are potentially andepigenetically regulated by alterations in hepatic levels ofSAM and SAH and their methylation capacity ratio. Ourlaboratory explored these relationships in two ethanol-fedanimal models, the micropig and the genetically alteredmouse. An initial study demonstrated the suitability of themicropig for study of the pathogenesis of ASH, in which itscharacteristic histopathology was induced by feeding ethanolat 40% of kcal over 12 months [27] together with progressiveincrease in serum homocysteine levels and reduced genemethylation capacity as shown by decrease in the liver SAMand the SAM-to-SAH ratio [28]. Subsequently, we includedfolate deficiency in the dietary regimen in order to accentuatechanges in methionine metabolism through reduction of theinitial methyl donor 5-MTHF (Figure 1), with the additionalrationale of the well-known association of folate deficiencywith chronic alcoholism [29, 30]. In contrast to the findingsin the previous study [27] only 3 months were requiredfor the development of the histopathology of ASH, whileincreases in timed serum homocysteine levels and loweringof the hepatic SAM-to-SAH ratio were the greatest in the

group fed the combined folate-deficient and ethanol dietcompared to groups fed either diet alone [12]. Using liversamples from the same groups of micropigs, we foundprogressive increase in transcript and protein levels ofgenes relevant to steatosis and liver injury from micropigsfed control, folate-deficient, ethanol, and ethanol withfolate-deficient diets, together with increased expressionsof CYP2E1 and ER stress pathway genes for the markerGRP 78, lipogenesis that included SREBP 1-c and fattyacid synthase (FAS), and apoptosis including GADD 153[31]. The finding that these gene expressions correlated withchanges in the SAM-to-SAH ratio supported the relationshipof the pathogenesis of ASH to ethanol-induced reduction ofgene methylation capacity. This concept was supported bysubsequent studies in which the coadministration of SAMsustained normal histology while preventing changes in theSAM-to-SAH ratio, and activations of the same ER stressgenes for lipogenesis and apoptosis [32], as well as genesrelevant to oxidative liver injury including CYP2E1, nicoti-namide adenine dinucleotide phosphate oxidase (NOX1),and inducible nitric oxide synthase (iNOS) [33].

6. ASH Is Mediated through theEpigenetic Effects of Ethanol-InducedAltered Methionine Metabolism

Epigenetics refers to the effects of external factors on theexpressions of genes that are unrelated to changes in theirDNA sequences. Gene expression is regulated at the levelsof both DNA and its histone infrastructures that existwithin nuclear chromatin. In the context of the presentpaper, such changes can be attributed to the regulatoryeffects of increased or decreased methylation on DNA andhistones, since folate contains the parent methyl group andSAM is the principal methyl donor in all methyltransferasereactions (Figure 1). As described in recent reviews, theeffect of chronic ethanol exposure on the activation orsuppression of selected ASH-related genes can be ascribed todifferent methylation effects at histone 3 lysine (K) residues,whereby gene expression is enhanced by methylation atH3meK4 but is silenced by methylation at H3meK9 [34, 35].Furthermore, ethanol exposure enhances gene expressionby acetylation at H3AcK9[35], whereas sirtuin1, a knownhistone de-acetylase, is reduced in livers of ethanol-fed mice[36]. Therefore, the effects of ethanol exposure on specifichistone methyltransferases, acetylases, and deacetylases maybe crucial to the epigenetic control of the expressions of genesrelevant to liver injury. Further epigenetic regulation occursat the level of DNA, where increased methylation of pro-moter region cytosine residues suppresses gene expressionand vice versa. For example, global DNA hypomethylationoccurred in the ethanol-fed SAM-deficient rat [37] andMAT1A knockout mouse with enhanced carcinogenesis [38],as well as in our folate-deficient chronic alcoholic micropigmodel in association with increased DNA oxidation andstrand breaks [12]. Others found reduced expression of theDNA methyltransferase DNMT3b that correlated negatively


with increased blood alcohol levels in chronic alcoholicstogether with increased global DNA methylation [39].

We explored epigenetic regulation of ASH in a novelmodel of the C57BL6 CβS-deficient mouse fed intragastricethanol at 37% of kCal for 4 weeks. This model is particularlyrelevant since, as noted above, the same method was usedto induce ASH in C57BL6 wild-type mice with findings ofaltered methionine metabolism, all of which were correctedby the methyl donor betaine [16]. Since CβS heterozygosityis known to increase serum homocysteine levels [40] andethanol exposure inhibits MS expression and activity [5,6] we reasoned that both ethanol exposure and genotypewould accentuate aberrant methionine metabolism with thegreatest effect in the combined group. Our study showed allthe histopathological features of severe ASH in the combinedethanol-fed heterozygous group, which also showed thegreatest reduction in the SAM-to-SAH ratio and the greatestincrease in expressions of the ER stress genes SREBP 1-c, GRP78, and GADD153 [41]. While immunohistologicalstaining of liver slices showed that methylated histone residueH3K9 was specifically reduced in centrilobular regions whichwere the sites of maximal steatosis in the ethanol-fedheterozygous group, the chromatin immunoprecipitation(ChIP) assay using antibody to H3K9 showed its reducedpresence in the promoter regions of the same ER stressgenes. Since H3K9 associates with gene suppression, itsreduced presence in liver from the combined ethanol-fedheterozygous group is consistent with activation of thesegenes. Furthermore, the transcript expression of G9a, ahistone methyltransferase that is required for methylationof K9 residues, was specifically reduced in both ethanol-fed groups in contrast to other K9 methyltransferases [41].The mechanism for reduced expression of the histonemethyltransferase G9a is unclear, but others have showneffects of ethanol-induced ROS on chromatin remodelingas a mechanism for changes in histone acetylation reactions[42]. Summarizing, these studies provided an epigeneticmechanism for the relationship of ethanol-induced aberrantmethionine metabolism on the epigenetic regulation ofexpression of selected ER stress genes that are relevant tolipogenesis and apoptosis in ASH [41].

Using the ChIP assay in a mouse macrophage RAWcell line and in vivo in LPS-treated mice, others showedthat SAM inhibited the LPS-induced activations of TNFαand iNOS by blocking the binding of H3meK4 to thepromoter regions of these genes [43]. Data from two groupssuggest epigenetic regulation of proteosome inhibition as amechanism for enhanced liver injury. In one study in anethanol-fed rat model, proteosome inhibition was associatedwith alterations in histone acetylation and methylation [44],whereas methylation of a specific subunit by provision ofSAM prevented proteosome degradation in ethanol-exposedhepatocytes [45].

7. Clinical Studies

A study of patients with cirrhosis of diverse etiologiesincluding alcoholism found significant elevations in serumhomocysteine and cystathionine, which was suggestive of

a block in the transsulfuration pathway for homocysteineelimination and production of GSH [46]. Compared tohealthy subjects, we found increased serum homocysteinelevels in alcoholics with or without ASH in associationwith elevated serum SAH levels [47]. While others havedescribed elevated homocysteine levels in chronic alcoholicsin association with deficiencies of folate and vitamin B6[48], we found that ASH patients also had marked increasein serum levels of cystathionine, the substrate for vitamin-B6-dependent cystathionase (Figure 1), with comparablereduction in its product α-aminobutyrate (ABU). Further-more, cystathionine levels and the ABU-to-cystathionineratio correlated with vitamin B6 levels and the ABU-to-cystathionine ratio was a positive predictor of the presenceof ASH among alcohol drinkers [47].

The growing experimental evidence linking impairedmethionine metabolism with the onset and development ofalcoholic liver disease prompted several clinical trials thatattempted to demonstrate a role for SAM in its treatment.One study showed that 15 days of intravenous SAM at2 g/day increased red blood cell GSH in 20 patients with ASH[49]. Hepatic GSH levels were low in liver biopsies from 17patients with ASH or other chronic liver diseases and werenormalized by SAM treatment at 1.2 g/day in a 6-monthcontrolled study [50]. A later 2-year multicenter Europeantrial of 123 ASH patients found reduction in mortalityor incidence of liver transplant from 30% in the placebogroup to 16% in the SAM group, which was significantafter exclusion of Childs class C patients from the analysis[51]. A 2006 meta-analysis of 9 studies found insufficientevidence for the effective use of SAM in the treatmentof alcoholic liver disease [52]. Whereas the PC derivativepolyenylphosphatidylcholine (PPC) was known to restoreliver SAM levels in ethanol fed rats [53], there was no benefitof this compound according to liver histopathology in 412ASH patients who underwent liver biopsies at baseline andafter 24 months of treatment [54].

We conducted a randomized treatment trial in 26 ASHpatients who participated in our baseline study [47] andthen received SAM at 1.2 g/d or placebo daily for 24 weeks[55]. All biochemical parameters of liver function improvedover time with no differences between the groups. Althoughserum SAM levels rose over time with treatment, there wereno changes in other methionine metabolites in either group.Furthermore, there were no differences between groups inhistopathology scores for steatosis, inflammation, or fibrosisin 14 patients who underwent liver biopsies before and aftertreatment [55].

Summarizing, the results from several clinical trials ofSAM in treatment of ASH are inconclusive as to its effect.Previous positive studies included patients with other causesof liver disease and showed no comprehensive histologicaldata on the efficacy of SAM. While our study is thefirst to provide data on potential changes in methioninemetabolite levels and documents the lack of differencesin histopathology, its limitations include relatively shortexposure to SAM and relatively small number of subjects.Unlike studies in animal models that showed the efficacy ofSAM in prevention of ASH [13, 32, 33], the use of SAM in the


treatment of established liver disease may be compromisedby lack of retention of SAM by injured hepatocytes as wellas decreased numbers of normally functioning hepatocytes.Also, based on our methionine metabolite findings at base-line [47] and the role of vitamin B6 in the transsulfurationpathway (Figure 1), there may be a requirement for vitaminB6 supplementation in addition to SAM in longer-termtreatment of ASH.

8. Conclusions

Abundant evidence now exists for the central role of aberranthepatic methionine metabolism in the pathogenesis of ASH.Ethanol and its ROS products inhibit key enzymes inthe methionine cycle, in particular those involved in thesynthesis of SAM. Since SAM is the principal methyl donorfor histones and DNA and facilitates the production ofGSH, its reduction influences the methylation and henceepigenetic regulation of many genes involved in alcoholicliver injury as well as the capacity for antioxidant defensein the pathogenesis of ASH. Whereas SAM supplementationhas been proven to prevent ASH and its mechanisms inexperimental animal models, clinical trials have failed todemonstrate a definitive effect of SAM in the treatment ofestablished ASH.

Acknowledgments

The authors’ original work that is referenced in this paperwas supported by Grants P30DK035747, R01AA014145, andR01AA014562 to C. H. Halsted from the National Institutesof Health, USA Neither author has a conflict of financialinterests in this work or in any of the original studiespresented herein.

References

[1] J. M. A. L. Mato, F. J. Corrales, and M. A. Pajares, “S-adenosylmethionine and the liver,” in The Liver: Biology andPathobiology, I. M. B. J. Arias, N. Fausto, W. B. Jakoby, D. A.Schachter, and D. A. Shafritz, Eds., pp. 461–470, Raven Press,New York, NY, USA, 1994.

[2] S. Clarke and K. Banfield, “S-adenosylmethionine-dependentmethyltransferases,” in Homocysteine in Health and Disease,R. Carmel and D. Jacobsen, Eds., pp. 63–78, CambridgeUniversity Press, Cambridge, UK, 2001.

[3] J. D. Finkelstein, W. E. Kyle, J. J. Martin, and A. M. Pick,“Activation of cystathionine synthase by adenosylmethionineand adenosylethionine,” Biochemical and Biophysical ResearchCommunications, vol. 66, no. 1, pp. 81–87, 1975.

[4] T. D. Lee, M. R. Sadda, M. H. Mendler et al., “Abnormalhepatic methionine and glutathione metabolism in patientswith alcoholic hepatitis,” Alcoholism: Clinical and Experimen-tal Research, vol. 28, no. 1, pp. 173–181, 2004.

[5] J. A. Villanueva and C. H. Halsted, “Hepatic transmethylationreactions in micropigs with alcoholic liver disease,” Hepatol-ogy, vol. 39, no. 5, pp. 1303–1310, 2004.

[6] A. Barak, H. Beckenhauer, and D. Tuma, “Ethanol feedinginhibits the activity of hepatic N5-methyltetrahydrofolate-homocysteine methyltransferase in the rat,” IRCS MedicalScience, vol. 13, no. 8, pp. 760–761, 1985.

[7] A. J. Barak, H. C. Beckenhauer, M. Junnila, and D. J.Tuma, “Dietary betaine promotes generation of hepatic S-adenosylmethionine and protects the liver from ethanol-induced fatty infiltration,” Alcoholism: Clinical and Experimen-tal Research, vol. 17, no. 3, pp. 552–555, 1993.

[8] K. C. Trimble, A. M. Molloy, J. M. Scott, and D. G. Weir,“The effect of ethanol on one-carbon metabolism: increasedmethionine catabolism and lipotrope methyl-group wastage,”Hepatology, vol. 18, no. 4, pp. 984–989, 1993.

[9] F. Ruiz, F. J. Corrales, C. Miqueo, and J. M. Mato, “Nitric oxideinactivates rat hepatic methionine adenosyltransferase In vivoby S-nitrosylation,” Hepatology, vol. 28, no. 4 I, pp. 1051–1057,1998.

[10] E. Sanchez-Gongora, F. Ruiz, J. Mingorance, W. An, F. J.Corrales, and J. M. Mato, “Interaction of liver methionineadenosyltransferase with hydroxyl radical,” FASEB Journal,vol. 11, no. 12, pp. 1013–1019, 1997.

[11] J. Zhuge and A. I. Cederbaum, “Depletion of S-adenosyl-l-methionine with cycloleucine potentiates cytochrome P4502E1 toxicity in primary rat hepatocytes,” Archives of Biochem-istry and Biophysics, vol. 466, no. 2, pp. 177–185, 2007.

[12] C. H. Halsted, J. A. Villanueva, A. M. Devlin et al., “Folate defi-ciency disturbs hepatic methionine metabolism and promotesliver injury in the ethanol-fed micropig,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 99, no. 15, pp. 10072–10077, 2002.

[13] C. S. Lieber, A. Casini, L. M. DeCarli et al., “S-adenosyl-L-methionine attenuates alcohol-induced liver injury in thebaboon,” Hepatology, vol. 11, no. 2, pp. 165–172, 1990.

[14] H. Tilg and A. M. Diehl, “Mechanisms of disease: cytokinesin alcoholic and nonalcoholic steatohepatitis,” New EnglandJournal of Medicine, vol. 343, no. 20, pp. 1467–1476, 2000.

[15] M. You and D. W. Crabb, “Recent advances in alcoholic liverdisease II. Minireview: molecular mechanisms of alcoholicfatty liver,” American Journal of Physiology—Gastrointestinaland Liver Physiology, vol. 287, no. 1, pp. G1–G6, 2004.


[17] K. K. Kharbanda, M. E. Mailliard, C. R. Baldwin, H. C.Beckenhauer, M. F. Sorrell, and D. J. Tuma, “Betaine attenuatesalcoholic steatosis by restoring phosphatidylcholine gener-ation via the phosphatidylethanolamine methyltransferasepathway,” Journal of Hepatology, vol. 46, no. 2, pp. 314–321,2007.

[18] K. K. Kharbanda, D. D. Rogers, M. E. Mailliard et al., “A com-parison of the effects of betaine and S-adenosylmethionineon ethanol-induced changes in methionine metabolism andsteatosis in rat hepatocytes,” Journal of Nutrition, vol. 135, no.3, pp. 519–524, 2005.

[19] K. K. Kharbanda, S. L. Todero, B. W. Ward, J. J. Cannella,and D. J. Tuma, “Betaine administration corrects ethanol-induced defective VLDL secretion,” Molecular and CellularBiochemistry, vol. 327, no. 1-2, pp. 75–78, 2009.

[20] Z. Song, Z. Zhou, S. Uriarte et al., “S-adenosylhomocysteinesensitizes to TNF-α hepatotoxicity in mice and liver cells: apossible etiological factor in alcoholic liver disease,” Hepatol-ogy, vol. 40, no. 4, pp. 989–997, 2004.

[21] Z. Song, Z. Zhou, M. Song et al., “Alcohol-induced S-adenosylhomocysteine accumulation in the liver sensitizes toTNF hepatotoxicity: possible involvement of mitochondrialS-adenosylmethionine transport,” Biochemical Pharmacology,vol. 74, no. 3, pp. 521–531, 2007.


[22] D. W. Horne, R. S. Holloway, and C. Wagner, “Transportof S-adenosylmethionine in isolated rat liver mitochondria,”Archives of Biochemistry and Biophysics, vol. 343, no. 2, pp.201–206, 1997.

[23] A. Colell et al., “Transport of reduced glutathione in hep-atic mitochondria and mitoplasts from ethanol-treated rats:effect of membrane physical properties and S-adenosyl-L-methionine,” Hepatology, vol. 26, no. 3, pp. 699–708, 1997.

[24] P. Sykora, K. K. Kharbanda, S. E. Crumm, and A. Cahill,“S-adenosyl-L-methionine co-administration prevents theethanol-elicited dissociation of hepatic mitochondrial ribo-somes in male rats,” Alcoholism: Clinical and ExperimentalResearch, vol. 33, no. 1, pp. 1–9, 2009.

[25] S. M. Bailey, G. Robinson, A. Pinner et al., “S-adenosyl-methionine prevents chronic alcohol-induced mitochondrialdysfunction in the rat liver,” American Journal of Physiology—Gastrointestinal and Liver Physiology, vol. 291, no. 5, pp. G857–G867, 2006.

[26] S. C. Lu, K. Ramani, X. Ou et al., “S-adenosylmethioninein the chemoprevention and treatment of hepatocellularcarcinoma in a rat model,” Hepatology, vol. 50, no. 2, pp. 462–471, 2009.

[27] C. H. Halsted, J. Villanueva, C. J. Chandler et al., “Centrilob-ular distribution of acetaldehyde and collagen in the ethanol-fed micropig,” Hepatology, vol. 18, no. 4, pp. 954–960, 1993.

[28] C. H. Halsted, J. Villanueva, C. J. Chandler et al., “Ethanolfeeding of micropigs alters methionine metabolism andincreases hepatocellular apoptosis and proliferation,” Hepatol-ogy, vol. 23, no. 3, pp. 497–505, 1996.

[29] V. Herbert, R. Zalusky, and C. S. Davidson, “Correlation offolate deficiency with alcoholism and associated macrocytosis,anemia, and liver disease,” Annals of Internal Medicine, vol. 58,pp. 977–988, 1963.

[30] D. Savage and J. Lindenbaum, “Anemia in alcoholics,”Medicine, vol. 65, no. 5, pp. 322–338, 1986.

[31] F. Esfandiari, J. A. Villanueva, D. H. Wong, S. W. French, andC. H. Halsted, “Chronic ethanol feeding and folate deficiencyactivate hepatic endoplasmic reticulum stress pathway inmicropigs,” American Journal of Physiology—Gastrointestinaland Liver Physiology, vol. 289, no. 1, pp. G54–G63, 2005.

[32] F. Esfandiari, M. You, J. A. Villanueva, D. H. Wong, S. W.French, and C. H. Halsted, “S-adenosylmethionine attenuateshepatic lipid synthesis in micropigs fed ethanol with a folate-deficient diet,” Alcoholism: Clinical and Experimental Research,vol. 31, no. 7, pp. 1231–1239, 2007.

[33] J. A. Villanueva, F. Esfandiari, M. E. White, S. Devaraj, S. W.French, and C. H. Halsted, “S-adenosylmethionine attenuatesoxidative liver injury in micropigs fed ethanol with a folate-deficient diet,” Alcoholism: Clinical and Experimental Research,vol. 31, no. 11, pp. 1934–1943, 2007.

[34] P. Mandrekar and G. Szabo, “Signalling pathways in alcohol-induced liver inflammation,” Journal of Hepatology, vol. 50, no.6, pp. 1258–1266, 2009.

[35] S. D. Shukla, J. Velazquez, S. W. French, S. C. Lu, M. K. Ticku,and S. Zakhari, “Emerging role of epigenetics in the actions ofalcohol,” Alcoholism: Clinical and Experimental Research, vol.32, no. 9, pp. 1525–1534, 2008.

[36] M. You, X. Liang, J. M. Ajmo, and G. C. Ness, “Involvementof mammalian sirtuin 1 in the action of ethanol in the liver,”American Journal of Physiology—Gastrointestinal and LiverPhysiology, vol. 294, no. 4, pp. G892–G898, 2008.

[37] S. C. Lu, Z. Z. Huang, H. Yang, J. M. Mato, M. A. Avila, andH. Tsukamoto, “Changes in methionine adenosyltransferaseand S-adenosylmethionine homeostasis in alcoholic rat liver,”

American Journal of Physiology—Gastrointestinal and LiverPhysiology, vol. 279, no. 1, pp. G178–G185, 2000.

[38] S. C. Lu and J. M. Mato, “Role of methionine adenosyltrans-ferase and S-adenosylmethionine in alcohol-associated livercancer,” Alcohol, vol. 35, no. 3, pp. 227–234, 2005.

[39] D. Bonsch, B. Lenz, R. Fiszer, H. Frieling, J. Kornhuber, and S.Bleich, “Lowered DNA methyltransferase (DNMT-3b) mRNAexpression is associated with genomic DNA hypermethylationin patients with chronic alcoholism,” Journal of Neural Trans-mission, vol. 113, no. 9, pp. 1299–1304, 2006.

[40] M. Watanabe, J. Osada, Y. Aratani et al., “Mice deficient incystathionine β-synthase: animal models for mild and severehomocyst(e)inemia,” Proceedings of the National Academy ofSciences of the United States of America, vol. 92, no. 5, pp. 1585–1589, 1995.

[41] F. Esfandiari, V. Medici, D. H. Wong et al., “Epigenetic reg-ulation of hepatic endoplasmic reticulum stress pathways inthe ethanol-fed cystathionine beta synthase-deficient mouse,”Hepatology, vol. 51, no. 3, pp. 932–941, 2010.

[42] I. Rahman, J. Marwick, and P. Kirkham, “Redox modulationof chromatin remodeling: impact on histone acetylation anddeacetylation, NF-κB and pro-inflammatory gene expression,”Biochemical Pharmacology, vol. 68, no. 6, pp. 1255–1267, 2004.

[43] A. I. Ara, M. Xia, K. Ramani, J. M. Mato, and S. C. Lu,“S-adenosylmethionine inhibits lipopolysaccharide-inducedgene expression via modulation of histone methylation,”Hepatology, vol. 47, no. 5, pp. 1655–1666, 2008.

[44] J. Oliva, J. Dedes, J. Li, S. W. French, and F. Bardag-Gorce,“Epigenetics of proteasome inhibition in the liver of rats fedethanol chronically,” World Journal of Gastroenterology, vol. 15,no. 6, pp. 705–712, 2009.

[45] N. A. Osna, R. L. White, T. M. Donohue, M. R. Beard, D.J. Tuma, and K. K. Kharbanda, “Impaired methylation as anovel mechanism for proteasome suppression in liver cells,”Biochemical and Biophysical Research Communications, vol.391, no. 2, pp. 1291–1296, 2010.

[46] M. P. Look, R. Riezler, C. Reichel et al., “Is the increasein serum cystathionine levels in patients with liver cirrhosisa consequence of impaired homocysteine transsulfurationat the level of γ-cystathionase?” Scandinavian Journal ofGastroenterology, vol. 35, no. 8, pp. 866–872, 2000.

[47] V. Medici, J. M. Peerson, S. P. Stabler et al., “Impairedhomocysteine transsulfuration is an indicator of alcoholic liverdisease,” Journal of Hepatology, vol. 53, no. 3, pp. 551–557,2010.

[48] M. L. Cravo, L. M. Gloria, J. Selhub et al., “Hyperhomo-cysteinemia in chronic alcoholism: correlation with folate,vitamin B-12, and vitamin B-6 status,” American Journal ofClinical Nutrition, vol. 63, no. 2, pp. 220–224, 1996.

[49] C. Loguercio, G. Nardi, F. Argenzio et al., “Effect of S-adenosyl-L-methionine administration on red blood cellcysteine and glutathione levels in alcoholic patients with andwithout liver disease,” Alcohol and Alcoholism, vol. 29, no. 5,pp. 597–604, 1994.

[50] G. Vendemiale, E. Altomare, T. Trizio et al., “Effects of oralS-adenosyl-L-methionine on hepatic glutathione in patientswith liver disease,” Scandinavian Journal of Gastroenterology,vol. 24, no. 4, pp. 407–415, 1989.

[51] J. M. Mato et al., “S-Adenosylmethionine in alcoholic livercirrhosis: a randomized, placebo-controlled, double-blind,multicenter clinical trial,” Journal of Hepatology, vol. 30, no.6, pp. 1081–1089, 1999.


[52] A. Rambaldi and C. Gluud, “S-adenosyl-L-methionine foralcoholic liver diseases,” Cochrane Database of SystematicReviews, no. 2, p. CD002235, 2006.

[53] S. I. Aleynik and C. S. Lieber, “Polyenylphosphatidylcholinecorrects the alcohol-induced hepatic oxidative stress by restor-ing s-adenosylmethionine,” Alcohol and Alcoholism, vol. 38,no. 3, pp. 208–212, 2003.

[54] C. S. Lieber, D. G. Weiss, R. Groszmann, F. Paronetto,and S. Schenker, “II. Veterans affairs cooperative study ofpolyenylphosphatidylcholine in alcoholic liver disease,” Alco-holism: Clinical and Experimental Research, vol. 27, no. 11, pp.1765–1772, 2003.

[55] V. Medici et al., “S-adenosylmethionine treatment of alcoholicliver disease: a double-blind, randomized, placebo-controlledtrial,” Alcoholism Clinical and Experimental Research. In press.

Advances in Alcoholic Liver Diseasedownloads.hindawi.com/journals/specialissues/682927.pdf ·...

Documents

Transcript of Advances in Alcoholic Liver Diseasedownloads.hindawi.com/journals/specialissues/682927.pdf ·...