Glycogen synthase kinase 3 involvement in the excessive proinflammatory response to LPS in patients with decompensated cirrhosis

Research Article

Glycogen synthase kinase 3 involvement in theexcessive proinflammatory response to LPS in patients

with decompensated cirrhosisq

Nicolas Coant1,2,�, Marika Simon-Rudler1,�, Thierry Gustot1,3,⇑,�, Magali Fasseu1, Sonia Gandoura1,Kévin Ragot4, Waël Abdel-Razek5, Dominique Thabut1, Philippe Lettéron1,2, Eric Ogier-Denis1,2,

Romy Ouziel3, Jacques Devière3, Gérard Lizard4, Zéra Tellier6, Didier Lebrec1,2,5, Richard Moreau1,2,5

1INSERM, U773, Centre de Recherche Biomédicale Bichat-Beaujon CRB3, Paris and Clichy, France; 2Université Denis Diderot-Paris 7, Site Bichat75018, France; 3Dept. of Gastroenterology and Hepato-Pancreatology, Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium;4INSERM, U866, Faculté des Sciences Gabriel, Dijon, France; 5Service d’Hépatologie, Hôpital Beaujon, Assistance Publique Hôpitaux de

Paris,Clichy, France; 6Laboratoire français du Fractionnement et des Biotechnologies, Courtaboeuf, France

Background & Aims: In decompensated cirrhosis, the early innate IL-10 was increased. LPS-induced, AKT-mediated phosphoryla-
immune response to the Toll-like receptor 4 (TLR4) agonist, lipo-polysaccharides (LPS), is characterized by a hyper-production ofpro-inflammatory cytokines and hypo-production of the anti-inflammatory cytokine IL-10. In LPS-stimulated non-cirrhoticimmune cells, the constitutively active glycogen synthase kinase(GSK) 3 favors pro- vs. anti-inflammatory cytokines, by acting ongene induction. However, in these cells, TLR4 dampens its ownpro-inflammatory response by inducing early (within minutes)AKT-mediated phosphorylation of GSK3b (one of two GSK3 iso-forms) on Ser9. Phosphorylation of GSK3b (Ser9) inhibits its activity,decreases pro-inflammatory cytokines, and increases IL-10. Thus,we investigated the role of GSK3 in LPS-induced cytokine produc-tion by peripheral blood mononuclear cells (PBMCs) or monocytesfrom patients with advanced cirrhosis and normal subjects.Methods: Cells were pre-incubated with or without GSK3 inhib-itor (SB216763 or lithium chloride) for 1 h and then stimulatedwith LPS. Cytokine production was assessed at mRNA andsecreted proteins levels, by real-time RT-PCR at 1 h and ELISAat 20 h, respectively. GSK3b phosphorylation was assessed usingWestern blotting.Results: In cirrhotic and normal PBMCs pretreated with GSK3inhibitors, LPS-induced production of pro-inflammatory proteinsTNF-a and IL-12p40 was significantly decreased while that of
Journal of Hepatology 20

Keywords: Cirrhosis; Alcohol; Innate immunity; Lipopolysaccharides; AKT; Gly-cogen synthase kinase 3.Received 15 September 2010; received in revised form 9 December 2010; accepted 15December 2010; available online 18 February 2011q This study has been presented in part at the 45[th] annual EASL Meeting inVienna, Austria, 2010.⇑ Corresponding author. Address: Dept. of Gastroenterology and Hepato-Pancreatology, Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium.Tel.: +32 2 555 61 98; fax: +32 2 555 61 97.E-mail address: [email protected] (T. Gustot).

� These authors contributed equally to this work.Abbreviations: TLR4, Toll-like receptor 4; LPS, lipopolysaccharides; TNF-a, tumornecrosis factor-a; IL, interleukin; GSK3, glycogen synthase kinase 3; PI3K, phos-phatidylinositol 3-kinase; PBMCs, peripheral blood mononuclear cells; LiCl,lithium chloride; CBA, cytometric bead array; RT-PCR, reverse transcriptase-polymerase chain reaction.

tion of GSK3b on Ser9 found in normal monocytes, was abolishedin cirrhotic cells.Conclusions: GSK3 is involved in the early TLR4-mediated pro-inflammatory response in patients with decompensated cirrhosis.This was associated with a defect in AKT-mediated GSK3b phos-phorylation resulting in unrestricted ‘pro-inflammatory’ activityof the enzyme.� 2011 European Association for the Study of the Liver. Publishedby Elsevier B.V. All rights reserved.

Introduction

Patients with cirrhosis have a high risk of developing sepsis (i.e.,infection plus systemic inflammation) and sepsis-induced organfailures resulting in death [1]. Septic shock mortality reaches80% in patients with cirrhosis compared to 50% in non-cirrhoticpatients [2,3]. In cirrhosis, the early stage of sepsis is accompa-nied by a markedly imbalanced cytokine response, which con-verts responses that are normally beneficial for fightinginfections into excessive, damaging inflammation [1,4]. In vivoand ex vivo studies have shown that during the first hours ofexposure to bacterial components (e.g., lipopolysaccharide, LPS,a Gram-negative bacteria byproduct), cirrhotic immune cellsoverproduce pro-inflammatory cytokines (such as tumor necrosisfactor (TNF)-a) and under-produce the anti-inflammatory cyto-kine interleukin (IL)-10 [5–9]. However, little is known aboutmechanisms involved in the early deregulation of the inflamma-tory response to bacterial components in cirrhotic immune cells.

Non-cirrhotic immune cells express two glycogen synthasekinase 3 (GSK3) isoforms, GSK3a and GSK3b, which are constitu-tively active serine–threonine kinases encoded by two separategenes, GSK3A and GSK3B, respectively [10]. In non-cirrhoticimmune cells, LPS via its ‘cognate’ receptor, Toll-like receptor 4(TLR4) stimulates the production of pro- and anti-inflammatorycytokines which is modulated by GSK3 that tilts the balance in

11 vol. 55 j 784–793

mailto:[email protected]

Fig. 1. Role of GSK3b in Toll-like receptor 4 (TLR4) signaling in immune cellsfrom normal subjects. After lipopolysaccharide (LPS) recognition, TLR4 dimer-izes and undergoes conformational changes inducing activation of intracellularcascades. TLR4 activation induces phosphorylation of inhibitor of nuclear factor(NF)-jB (IjB) by IjB kinase (IKK), its polyubiquitination and degradation,releases NF-jB to the nucleus and induces pro-inflammatory genes, such astumor necrosis factor (TNF), interleukin (IL) 12B, encoding TNF-a, and IL-12p40,respectively. The TLR4-induced activation of mitogen-activated protein kinase(MAPK) signaling pathway via transcription factor, activator protein-1 (AP-1)contributes to cytokine gene induction. In addition, the constitutively activeGSK3b favors NF-jB-elicited induction of pro-inflammatory genes. Activation ofTLR4 also induces negative regulator pathways to control the inflammatoryprocess. For example, TLR4 activation results in phosphoinositide 3-kinase (PI3K)-mediated AKT (protein kinase B) phosphorylation, inducing phosphorylation ofGSK3b (Ser9) that causes its inhibition and the resulting decrease in pro-inflammatory cytokine gene transcription. Moreover, TLR4-induced, AKT-mediated GSK3b increases transcription of the anti-inflammatory cytokine IL10reducing the effects of pro-inflammatory cytokines. The TLR4/PI3K/AKT pathwayalso phosphorylates GSK3a (Ser21) but this phosphorylation does not effectcytokine production, in fact its effects on cell functions are unknown.

JOURNAL OF HEPATOLOGY

favor of pro- vs. anti-inflammatory cytokines (Fig. 1) [11,12]. Thisis why selective cell-permeable inhibitors of GSK3 decreasepro-inflammatory cytokines and stimulate anti-inflammatorymolecules in LPS-stimulated immune cells [11,12]. On the otherhand, within the first minutes of LPS exposure, TLR4 activatesphosphatidylinositol 3-kinase (PI3K) and its mediator AKT whichinduces phosphorylation of GSK3b on Ser9 (Fig. 1) [11,13]. Phos-phorylation of GSK3b (Ser9) results in its inhibition leading to adecrease in pro-inflammatory cytokines and an increase in IL-10 [11,14,15]. Therefore, a defect in GSK3b phosphorylation incirrhotic immune cells may deregulate the inflammatoryresponse to LPS. Interestingly, LPS, via the TLR4/PI3K/AKT path-way, also induces phosphorylation of GSK3a on Ser21 resultingin its inhibition. However, in immune cells, for unknown reasons,LPS-induced phosphorylation of GSK3a (Ser21) does not affectcytokine production [11]. There are examples of non redundancy


between the two GSK3 isoforms in non immune cells [16,17].Since there is no information on GSK3 in cirrhotic immune cells,the aim of the present ex vivo study was to investigate the role ofGSK3 in the alteration of the innate immune response to LPS inimmune cells from patients with decompensated cirrhosis. Wefirst assessed the effects of cell-permeable inhibitors of GSK3on cytokine production in LPS-stimulated immune cells frompatients and normal subjects. Then, we studied GSK3 expressionand LPS-induced phosphorylation of GSK3 in immune cells fromboth groups.

Patients and methods

Patients

Enrolled patients were hospitalized in our Liver Unit. Inclusion criteria were ageover 18 years, diagnosis of alcohol-induced decompensated cirrhosis (B or C inthe Child-Pugh classification) based on liver biopsy or classical clinical, biochem-ical, and radiological criteria. They had a history of excessive alcohol ingestion(50 g per day) but no other causes of liver disease. The viral serologies for HBVand HCV were negative. Alcohol consumption was stopped for at least 3 days.The clinical characteristics of patients are shown in Table 1. Exclusion criteriawere: evidence of recent gastrointestinal bleeding, current bacterial infections,and treatment with corticosteroids, pentoxifylline, and other immunosuppressivedrugs in the past 30 days, and presence of hepatocellular carcinoma, other canceror human immunodeficiency virus infection. Healthy subjects (controls) are hos-pital employee volunteers. The study was approved by our Institutional ReviewBoard, and written informed consent was obtained from all patients.

Isolation and culture of peripheral blood mononuclear cells (PBMCs) and monocytes

Peripheral venous blood was obtained from volunteers and patients with cirrho-sis by venopuncture and collected into EDTA-treated, sterile, pyrogen-free tubes(Gibco, Grenoble, France). PBMCs and monocytes isolation are described in Sup-plementary material.

Pharmacological inhibitors

GSK3 inhibitors (SB216763 and lithium chloride (LiCl)) and the PI3K inhibitor,LY294002, were purchased from Sigma. There are described in Supplementarymaterial.

Measurement of cytokine productions using enzyme-linked immunosorbent assay(ELISA)

Human TNF-a, IL-12p40, and IL-10 were measured in culture supernatant byELISA Bender MedSystems™ (Vienna, Austria) or BD Biosciences (San Diego,CA) (Supplementary material).

Measurement of IL-10 and IL-12 production using the multiplexed cytometric beadarray (CBA)

The production of inflammatory cytokines was investigated in the culture mediaof cirrhotic monocytes with a flow cytometric bead-based assay: the multiplexedcytometric bead array (CBA) [18–22] (Supplementary material).

Reverse transcriptase-polymerase chain reaction (RT-PCR)

Total RNA was extracted from cell using Trizol� reagent (Invitrogen, Cergy-Pontoise, France) (Supplementary material). The differential expression ofcytokine genes was assessed by Real time quantitative polymerase chain reaction(PCR) by using the SYBR� Green based detection system on LightCycler �480(Roche Diagnostics) [23].

The expressions of GSK3A, GSK3B, AKT1, AKT2, and AKT3 mRNAs were assessedby qualitative PCR monocyte lysates.

1 vol. 55 j 784–793 785

Table 1. Characteristics of patients with decompensated alcoholic cirrhosis.*

Number of patients 63 )sraey(egA 58 (32-79)

Gender (female/male) 21/47Ascites/refractory ascites 51/20 Child-Pugh score 11 (7-15) MELD score 22 (8-38) Serum albumin (g/L) 24 (13-42) Total bilirubin ( mol/L) 28 (13-632) Prothrombin time (%)/INR 40 (12-91)/1.7 (1.1-3.7) Serum creatinine ( mol/L) 72.5 (40-248)Leukocytes count (103/mm3) 7 (1.9-19.9)

* Values are number of patients or median [min–max].

**

#**

TNF-

α (p

g/m

l)

0

2000

4000

6000

8000 UnstimulatedLPS

**

**

IL-1

2p40

(pg/

ml)

0

1000

2000

4000

3000


**

##**

IL-1

0 (p

g/m

l)

500

1500

1000


Research Article

Western blotting

Cell lysates were obtained for Western blotting analysis (Supplementarymaterial).

Statistical analysis

Data are expressed as median [min–max] or mean ± SE. Mann–Whitney U testand Wilcoxon signed rank test or Student’s t test were used where appropriate.Analyses were performed using the SPSS 17.0. p values <0.05 were consideredsignificant.

0

Cirrhosis Normal

Fig. 2. Peripheral blood mononuclear cells (PBMCs) from patients withcirrhosis overproduce TNF-a and underproduce IL-10. Freshly isolated PBMCsfrom patients with cirrhosis (n = 10) and normal subjects (n = 10) were left un-stimulated or stimulated with lipopolysaccharide (LPS, 1 lg/ml) for 24 h. Cell-freesupernatants were analyzed by ELISA for production of TNF-a, IL-12p40, and IL-10. CTR + LPS ⁄⁄p <0.01, PBMCs from cirrhotic patients vs. those from controls#p <0.05, ##p <0.01. Results are shown as means ± standard errors.

Results

Enhanced pro-inflammatory profile in LPS-stimulated cirrhoticPBMCs

Un-stimulated cirrhotic and normal PBMCs did not producedetectable amounts of TNF-a, IL-12p40, and IL-10. In both groups,exposure of LPS for 24 h stimulated the production of all mea-sured cytokines (Fig. 2). As expected [24], LPS-induced TNF-aproduction was significantly higher in cirrhotic than normalPBMCs (p <0.05). In addition, LPS-induced IL-12p40 productiontended to be higher in cirrhotic PBMCs compared to controls(Fig. 2). We also found that LPS-induced IL-10 production wassignificantly lower in cirrhotic than normal PBMCs (p <0.01).Thus, TNF-a over-production and IL-10 under-production arehallmarks of LPS-stimulated cirrhotic immune cells.

In this study, three patients had biopsy proven-alcoholic hep-atitis. The LPS-induced production of TNF-a, IL-12p40, and IL-10did not significantly differ between patients with alcoholic hepa-titis and those without (data not shown).

Next, we investigated in cirrhotic PBMCs the kinetics of LPSinduction of mRNAs encoding TNF-a, IL-12p40, and IL-10 (i.e.,TNF, IL12B, and IL10 mRNAs, respectively). For this, we isolatedPBMCs in 33 patients with decompensated cirrhosis. Then,PBMCs were stimulated with LPS or left unstimulated. The rela-tive mRNA expressions were measured in cell lysates at differenttime points after the onset of exposure to LPS (30 min, 1, 4, and20 h). The maximum induction was found at 1 h for TNF mRNA,4 h for IL12B mRNA, and 4 h for IL10 mRNA. After 20 h of LPS stim-ulation, the relative levels of each mRNA were not statisticallydifferent from those of unstimulated cells (Fig. 3).

786 Journal of Hepatology 201

Cell-permeable inhibitors of GSK3 differentially modulateproduction of pro-inflammatory cytokines and IL-10 in LPS-stimulated cirrhotic immune cells.

In normal immune cells, cell-permeable inhibitors of GSK3differentially regulate LPS-induced cytokine production: theyinhibit production of pro-inflammatory cytokines and stim-ulate the anti-inflammatory IL-10 [11,12,14,15]. The effectsof GSK3 inhibitors on LPS-stimulated cirrhotic immune cellsare unknown. We, therefore, studied the effects of 10 lMSB216763 (an arylindolemaleimide which is a selective cellpermeable inhibitor of the two GSK3 isoforms in an ATP com-petitive manner) [25] on LPS-induced cytokine productionassessed by protein levels of TNF-a, IL-12p40, and IL-10secreted in the culture medium. For this, freshly isolated PBMCsfrom patients with cirrhosis and normal subjects were pre-incubated for 1 h with medium only, 10 mM SB216763, andthen left un-stimulated or stimulated with 1 lg/ml of LPS for20 h. In cirrhotic PBMCs pre-incubated with SB216763, therewas a significant reduction in LPS-induced production ofTNF-a (p <0.05), and IL-12p40 (p <0.05) (Fig. 4A). In normal

1 vol. 55 j 784–793

**

**

TNF-α

mR

NA

(fold

cha

nge)

0

100

200

300

400

30 min 1 hr 4 hr 20 hrTime

30 min 1 hr 4 hr 20 hrTime

30 min 1 hr 4 hr 20 hr

Time

*

*

IL12

mR

NA

(fold

cha

nge)

0

200

400

600

800

1000

**

IL10

mR

NA

(fold

cha

nge)

0

10

5

15

20

25

Fig. 3. Kinetics of mRNA induction in LPS-stimulated peripheral bloodmononuclear cells (PBMCs) from patients with cirrhosis (n = 33). Freshlyisolated PBMCs from patients with cirrhosis were left un-stimulated or stimu-lated with lipopolysaccharide (LPS, 1 lg/ml) for 30 min, 1, 4, or 20 h. LPS vs.unstimulated ⁄p <0.05, ⁄⁄p <0.01. Whole-cell lysates were analyzed for cytokinemRNA levels using real time RT-PCR. Results are shown as means ± standarderrors.


PBMCs, we observed a significant decrease in LPS-induced IL-12p40 production but only a trend for reduction for LPS-inducedTNF-a with SB216763 pre-incubation. On the other hand, inboth groups, SB216763 significantly enhanced LPS-inducedIL-10 production (Fig. 4A).


Interestingly, using LiCl (20 mM, an ATP noncompetitiveinhibitor of the two GSK3 isoforms) [26], we observed a similarpattern of effects to that of SB216763. In cirrhotic and normalPBMCs pre-incubated with LiCl, there were decreases in LPS-induced productions of TNF-a (p <0.05) and IL-12p40 (p <0.05)(Fig. 4B). In PBMCs pre-incubated with LiCl, there was a signifi-cant increase in LPS-induced IL-10 production in cirrhotic butnot normal cells.

Because monocytes are the main producers of IL-12 in theblood [27], we then investigated the effects of LiCl on freshly iso-lated LPS-stimulated monocytes from patients with decompen-sated cirrhosis. Freshly isolated monocytes were pre-incubatedfor 1 h with medium only, 20 mM LiCl and then left un-stimulatedor stimulated with 1 lg/ml of LPS for 20 h. Cell-free supernatantswere analyzed by CBA for IL-10 and IL-12p70 production andELISA for IL-12p40 production. As expected with ‘pure’ monocytes[28], cirrhotic monocytes did not produce IL-12p70 whicheverthe condition. On the other hand, LPS-stimulated monocytesproduced IL-10 (142 [26–279] pg/ml) and IL-12p40 (389 [267–610] pg/ml). In LiCl-pre-incubated cells, LPS-induced productionof IL-10 was increased (by 720% [600–1240%]; p <0.05) and thatof IL-12p40 decreased (by 98% (35–99%); p <0.05). These resultsindicate that GSK3 inhibition alters the cytokine balance of cir-rhotic monocytes toward an anti-inflammatory phenotype uponLPS stimulation.

Next, we used real-time RT-PCR to gain further insights intothe mechanism by which GSK3 inhibitors modulated cytokineproduction in cirrhotic PBMCs. After LPS stimulation (1 lg/ml)for 1 h, PBMCs exhibited significant induction of TNF mRNA,IL12B mRNA, and IL10 mRNA (Fig. 5). In the presence ofSB216763, there was an inhibition of LPS-induced IL12B andTNF mRNAs and a significant induction of IL10 mRNA (Fig. 5).Therefore, the effects of SB216763 on levels of LPS-inducedsecreted cytokines (i.e., IL-10, IL-12p40, and TNF-a; Fig. 3A)resulted from effects of the GSK3 inhibitor on correspondingLPS-induced mRNA levels. These findings are consistent with amodulation by GSK3 inhibitors of LPS-induced cytokine genetranscription [11–13,16]. Moreover, SB216763 induced IL10mRNA expression in LPS-untreated PBMCs, suggesting thatGSK3 inhibits this expression in cultured cirrhotic PBMCs withoutstimuli.

Defect in TLR4-mediated GSK3b phosphorylation in cirrhoticmonocytes

In normal immune cells, within the first minutes of LPS exposure,TLR4 engagement via the PI3K activates AKT, which in turn phos-phorylates a key serine residue near the N terminus of GSK3(Ser21 of GSK3a and Ser9 of GSK3b), resulting in the inactivationof GSK3 [11–13,26]. In normal immune cells, GSK3a phosphory-lation does not affect cytokine production. In contrast, in thesecells, inhibition of GSK3b was found to be a negative feedbackmechanism by which TLR4 dampens its own pro-inflammatoryresponse and favors IL-10 production (Fig. 1) [11,12,14,15].GSK3 expression and modulation of its phosphorylation by theTLR4/PI3K/AKT pathway is unknown in cirrhotic immune cells.We, therefore, addressed this question in freshly isolated cir-rhotic and normal monocytes. First, we measured GSK3A andGSK3B mRNA levels using semi-quantitative RT-PCR analysis inwhole-cell lysates from patients (n = 3) and normal subjects(n = 3). Then, we studied GSK3 protein expression and early

1 vol. 55 j 784–793 787

*

TNF-

α (p

g/m

l)

0

2000

4000

6000

8000

*

*

IL-1

2p40

(pg/

ml)

0

2000

1000

4000

3000

5000

6000

*

*

IL-1

0 (p

g/m

l)

0

500

1500

1000

2000

2500

*IL

-10

(pg/

ml)

0

500

1500

1000

2000

Unstimulated

LPSSB216763

LPS + SB216763

A B

Cirrhosis NormalCirrhosis Normal

*

*TNF-

α (p

g/m

l)

0

2000

4000

6000

Unstimulated

LPSLiCl

LPS + LiCl

**IL

-12p

40 (p

g/m

l)

0

2000

1000

4000

3000

5000

Fig. 4. Inhibitors of GSK3 differentially modulate pro- and anti-inflammatory cytokine secretion by LPS-stimulated peripheral blood mononuclear cells (PBMCs)from patients with cirrhosis and normal subjects. Freshly isolated PBMCs from patients with cirrhosis (n = 8) and normal subjects (n = 9) were pre-incubated for 1 h withmedium only, (A) 10 mM SB216763 or (B) 20 mM LiCl and then left un-stimulated or stimulated with 1 lg/ml of LPS for 20 h. Cell-free supernatants were analyzed by ELISAfor production of TNF-a, IL-12p40, and IL-10. LPS vs. LPS + SB216763 ⁄p <0.05, LPS vs. LPS + LiCl ⁄p <0.05. Results are shown as means ± standard errors.

Research Article

(30 min) phosphorylation by Western blotting using specific anti-bodies. Due to the fact that Western blotting analysis of mono-cyte proteins performed on an individual basis (i.e., oneanalysis per patient or normal subject) provided very weak sig-nals (T.G., R.M. unpublished data) [7], we pooled protein extractsfrom five individuals per group.

In un-stimulated monocytes, the levels of GSK3A mRNA andtotal GSK3a protein were significantly lower in cirrhotic thannormal cells (Fig. 6A and B). In both cirrhotic and normal un-stimulated cells, there was slight Ser21 phosphorylation ofGSK3a (Fig. 6A), suggesting some constitutive GSK3a phosphor-ylation. LPS stimulation (1 lg/ml) for 30 min resulted in slightGSK3a phosphorylation on Ser21 in both cirrhotic and normalmonocytes (Fig. 6A).

In un-stimulated monocytes, the levels of GSK3B mRNA andtotal GSK3b protein were only slightly lower in cirrhotic com-pared to normal cells (Fig. 6A and B). In both cirrhotic and normalun-stimulated cells, there was no significant phosphorylation ofGSK3b on Ser9 (Fig. 6A). In LPS-stimulated monocytes, at30 min, there was a significant phosphorylation of GSK3b onSer9 in normal but not cirrhotic cells (Fig. 6A). In addition, wefound that LPS stimulation of cirrhotic and normal monocytesmediated the phosphorylation of AKT (Ser473) (Fig. 6A) and that


the PI3K inhibitor LY294002 abolished the ability of LPS to inducephosphorylation of AKT (data not shown). Thus, cirrhotic mono-cytes have a defect in TLR4/PI3K-mediated phosphorylation ofGSK3b (Ser9) but not of GSK3a (Ser21) and this defect seems tobe distal to PI3K. Since TLR4-mediated phosphorylation of GSK3b(Ser9) is an ‘anti-inflammatory’ mechanism [11,12,14,15], ourresults suggest that a defect in TLR4-induced phosphorylation ofGSK3b on Ser9 may contribute to the excessive pro-inflammatoryresponse in LPS-stimulated cirrhotic immune cells (Fig. 2).

Then, we asked if the defect in Ser9 phosphorylation of GSK3bwas either specific for ‘cirrhotic’ TLR4 pathway or was alsoobserved with other ‘non TLR4-mediated’ stimuli. We, therefore,used LiCl (20 mM) because this drug is known to stimulate AKT tophosphorylate the two GSK3 isoforms (Ser21 of GSK3a and Ser9of GSK3b) [29]. In the absence or presence of LPS, cirrhotic andnormal monocytes exhibited LiCl-induced phosphorylation ofAKT (Fig. 6A) and GSK3a (Ser21) (Fig. 6A). Interestingly, in theabsence or presence of LPS, LiCl induced Ser9 phosphorylationof GSK3b in normal but not in cirrhotic cells (Fig. 6A). Therefore,cirrhotic monocytes have a defect in both TLR4-dependent and -independent AKT-mediated phosphorylation of GSK3b on Ser9.

The reasons why cirrhotic monocytes have differences inphosphorylation between GSK3b and GSK3a are unclear. In fact,

1 vol. 55 j 784–793

**

#

TNF-α

mR

NA

(2e-Δ

ΔCt )

0

40

20

60

80

100

Unstimulated

LPSSB216763

LPS + SB216763

**

##

IL12

mR

NA

(2e-Δ

ΔCt )

0

40

20

60

**

§§

##

IL10

mR

NA

(2e-Δ

ΔCt )

0

20

10

30

Fig. 5. The GSK3 inhibitor SB216763 differentially modulates pro- and anti-inflammatory cytokine mRNA expression in LPS-stimulated peripheral bloodmononuclear cells (PBMCs) from patients with cirrhosis. PBMCs from patients(n = 6) were pre-incubated for 1 h with medium only or SB216763 and then leftun-stimulated or stimulated with LPS (1 lg/ml) for 1 h. CTR vs. LPS ⁄⁄p <0.01; LPSvs. LPS + SB216763 #p <0.05, ##p <0.01; CTR vs. CTR + SB216763 §§p <0.01. Whole-cell lysates were analyzed for cytokine mRNA levels using real time RT-PCR.Results are shown as means ± standard errors.


there are three AKT isoforms: AKT1, AKT2, and AKT3 [30,31]. Ithas been shown that AKT2 preferentially phosphorylates GSK3awhile AKT1 and AKT3 phosphorylate GSK-3b [32]. Thus, we won-dered if AKT isoforms were differentially expressed in monocytesfrom patients with cirrhosis and measured the expressions of thethree AKT isoforms using qualitative PCR in cirrhotic and normalmonocytes. The three AKT isoforms were found in both groups;the expression levels of AKT2 and AKT3 mRNAs were similar incirrhotic and normal monocytes (Fig. 6B). In contrast, the expres-sion levels of AKT1 mRNAs were significantly lower in cirrhoticthan normal monocytes (Fig. 6B). These findings suggest that anunder-expression of AKT1 may play a role in the low LPS-inducedphosphorylation of GSK3b in cirrhotic immune cells (Fig. 6C).

Discussion

This study investigated the early ex vivo response to LPS in PBMCsfrom patients with decompensated alcoholic cirrhosis and fromnormal subjects. In both groups, LPS induced significant increases


in the production of both pro-inflammatory cytokines (TNF-a andIL-12p40) and the anti-inflammatory IL-10. However, as expected[5,22], the LPS-induced production of TNF-a was significantlyhigher and that of IL-10 lower in cirrhotic than normal PBMCs.It should be noted that LPS-induced productions of IL-12p40 (thisstudy) and IL-1b (T.G., R.M., unpublished data) were similar inboth cirrhotic and normal PBMCs. Together, these findings indi-cate that TNF-a hyper-production and defective IL-10 productioncharacterize an early ‘pro-inflammatory phenotype’ of LPS-stim-ulated cirrhotic PBMCs. Interestingly, a defect in IL-10 productionhas previously been shown to be involved in TNF-a hyper-pro-duction in LPS-stimulated cirrhotic PBMCs [6].

It is important to note that our patients had stable decompen-sated cirrhosis and not acutely decompensated cirrhosis. Indeed,they did not have acute variceal hemorrhage, recent enhance-ment of jaundice, ongoing sepsis, or development of extra-hepatic organ failure(s). When we studied kinetics of inductionof TNF, IL12B, and IL10 mRNAs in cirrhotic PBMCs, we found amaximum induction at 1 or 4 h, followed by a downregulationat 20 h, in particular of IL10 mRNA. Thus, we did not find evidencein these patients of ‘anti-inflammatory/immunosuppressiveprogram’ during prolonged exposure to LPS. Our patients weredifferent from those with acute-on-chronic liver failure who havean acute deterioration of liver function associated with the devel-opment of extra-hepatic organ failures and biological markers foran immunosuppressive state both in vivo and ex vivo [33].

Since the constitutively active GSK3 enhances LPS-inducedpro-inflammatory cytokines and inhibits IL-10 in normal immunecells [11,12,14,15], we investigated the role of GSK3 in responseto LPS in cirrhotic immune cells. For this purpose, we studiedthe effects of cell-permeable inhibitors of GSK3, SB216763, andLiCl, on cytokine production by LPS-stimulated cirrhotic PBMCs.In these cells, we found that GSK3 inhibitors had significanteffects on LPS-induced cytokine secretion in the extracellularmilieu: they significantly decreased production of TNF-a andIL-12p40, and enhanced IL-10 production. As expected [11,13],the effects of GSK3 inhibitors on secreted cytokines were a resultof an impact on LPS-elicited induction of genes encoding thesecytokines. Indeed, we found that GSK3 inhibition by SB216763significantly reduced LPS-induced TNF and IL12B mRNA levelsand increased LPS-induced IL10 mRNA. Together, these findingsshow that constitutive GSK3 activity tilts the balance in favor ofpro- vs. anti-inflammatory cytokines in LPS-stimulated cirrhoticimmune cells.

Within the first minutes of exposure to LPS, TLR4 engagementactivates PI3K leading to the recruitment of AKT and its activation(i.e., phosphorylation) [11]. Activated AKT phosphorylates Ser21of GSK3b (one of two GSK3 isoforms) and this inhibits the consti-tutive activity of GSK3b. In the present study, we used two differ-ent ways to activate the AKT/GSK3b pathway: TLR4 engagementby LPS and exposure to LiCl. Unlike normal monocytes, cirrhoticmonocytes stimulated with LPS or incubated with LiCl had adefect in AKT-mediated phosphorylation of GSK3b on Ser9, sug-gesting an impaired inhibition of GSK3b activity. It should beemphasized that selective suppression of GSK3b in non cirrhoticcells is associated with an altered response to LPS, i.e., a decreasein pro-inflammatory cytokines and an increase in IL-10[11,12,14,15]. Together these findings suggest that, an unre-stricted GSK3b activity may play a role in TNF-a overproduction,and defective IL-10 production in LPS-stimulated cirrhoticimmune cells.

1 vol. 55 j 784–793 789

α Ser21β Ser9p-GSK3

p-AKT Ser473

Total GSK3α

Total AKT

Total GSK3β

β Actin

Control LiCl LPS LPS+

LiClControl LiCl LPS LPS

+LiCl

CirrhosisNormal

*

GSK3A mRNA

Arbi

trary

uni

ts

n.sGSK3B mRNA

Arbi

trary

uni

ts

*

AKT1 mRNA

Arbi

trary

uni

ts

n.s.AKT2 mRNA

Arbi

trary

uni

ts

n.s.

AKT3 mRNA

Arbi

trary

uni

ts

NormalCirrhosisA B

C

Fig. 6. Defect in LPS (1 lg/ml)- or lithium chloride (LiCl, 20 mM)-induced, AKT-mediated Ser9 phosphorylation of GSK3b in peripheral blood monocytes frompatients with cirrhosis. (A) Freshly isolated monocytes from patients and normal subjects were incubated for 30 min with medium only, LiCl (20 mM), LPS (1 lg/ml) or LiClplus LPS. As detailed in the Results section, we pooled protein extracts obtained for each condition in 5 patients. Pooling was also performed for each condition in 5 normalsubjects. Protein extracts were analyzed by immunoblot with phosphorylation specific antibodies (left margins) followed by enhanced chemiluminescence detection toassess phosphorylation of AKT (Ser473), GSK3, and GSK3 (Ser21 and Ser9, respectively). (B) As described in the Patients and methods section, monocytes from patients andnormal subjects were plated after 1 h of adhesion. Whole-cell lysates were analyzed using semi-quantitative RT-PCR for GSK3A, GSK3B, AKT1, AKT2, and AKT3 mRNAs. Theexpression of each mRNA was normalized to b-actin expression. Values are representative means ± standard errors (from 3 patients and 3 normal subjects; ⁄p <0.005relative to normal subjects). (C) Proposed mechanisms for the decrease in GSK3 phosphorylation (Ser9) in cirrhotic monocytes. In normal monocytes, TLR4-inducedactivation of phosphoinositide 3-kinase (PI3K) results in AKT1 and AKT2 stimulation. AKT1 mediates phosphorylation of GSK3 on Ser9 while AKT2 mediatesphosphorylation of GSK3 on Ser21. In cirrhotic monocytes, AKT1 is underexpressed and this may play a role in the decreased phosphorylation of GSK3. A role of as-yet-unidentified phosphatases cannot be ruled out. In contrast, AKT2 is normally expressed in cirrhotic monocytes and phosphorylation of GSK3 on Ser21 is observed.

Research Article

In normal and cirrhotic monocytes, we found that AKT wasable to phosphorylate Ser21 of GSK3a (the second isoform ofGSK3), in response to the engagement of TLR4 or incubation with


LiCl. These findings are consistent with previous results obtainedin normal immune and non immune cells [26]. Thus, in cirrhoticmonocytes, unlike GSK3b, GSK3a is normally ‘phosphorylatable’

1 vol. 55 j 784–793

by TLR4 engagement or LiCl exposure. The reasons why GSK3 iso-forms are differentially regulated in cirrhotic monocytes areunclear. It is important to note that there are three AKT isoforms,AKT1, AKT2, and AKT3. AKT2 preferentially phosphorylatesGSK3a while AKT1 and AKT3 phosphorylate GSK3b. We found,using PCR, that the three AKT isoforms were expressed in cirrhoticand normal monocytes. Interestingly, whereas the expressions ofAKT2 and AKT3 mRNAs were similar in both groups, that of AKT1mRNA was significantly lower in cirrhotic than normal mono-cytes. Therefore, an under-expression of AKT1 may play a role inthe low level of Ser9 phosphorylation on GSK3b in cirrhoticimmune cells. However, a role of an as-yet-unidentified phospha-tase in defective GSK3b phosphorylation cannot be ruled out [13].
It has been shown that AKT1 (but not AKT2) plays a major rolein the induction of tolerance (i.e., desensitization) to LPS in nor-mal immune cells [34]. Our finding of decreased AKT1 expressionin cirrhotic monocytes suggests that these cells may be defectivein mounting tolerance to LPS. Studies are needed.

In normal immune cells, the function of the constitutive activityof GSK3a is unknown; it does not contribute to the regulation ofcytokine production. In cirrhotic monocytes, we found a markedunder-expression of GSK3a at mRNA and protein levels, suggest-ing that constitutive activity of GSK3a may be low in these cells.

Our study patients were in patients in whom blood was takenfor investigation after a short period of alcohol abstinence (due tohospitalization). This raises the question of the influence of under-lying chronic alcoholic consumption on the innate immuneresponse to LPS. In fact, the effect on the innate immune responseto LPS is different under acute vs. chronic exposure to alcohol.Acute ethanol exposure inhibits TNF-a production in LPS-stimulated monocytes/macrophages [35–38]. These acute effectsare not consistent with an alcohol-induced stimulation of GSK3bactivity (that would favor TNF-a production). The finding thatacute ethanol treatment decreases AKT phosphorylation andsubsequent GSK3b phosphorylation in murine cardiomyocytessuggests differences between cell types regarding the effect ofalcohol [39]. On the other hand, it has been shown that chronicexcessive alcohol consumption is associated with an enhancedproduction in LPS-stimulated monocytes/macrophages [38,40–42]. Different groups including ours found an enhancedLPS-induced TNF-a production in PBMCs or monocytes fromin-patients with alcoholic cirrhosis (who are subjected to a shortperiod of abstinence) [5,7–9,43]. Thus, patients with alcoholic cir-rhosis who are abstinent due to hospitalization have an innateimmune response to LPS which is very similar to that observedduring chronic alcohol consumption, suggesting prolonged effectsof alcohol on immune cell function. Another important point is therelationship of alcoholic hepatitis with the pro-inflammatory phe-notype of PBMCs. However, this study was not performed toaddress this question and only three of our patients had alcoholichepatitis. This number of patients is too low to draw firm conclu-sions based on our finding of similar LPS-induced cytokine pro-duction in patients with alcoholic hepatitis and in thosewithout. Future studies are needed to investigate the innateimmune response in patients with alcoholic hepatitis.

Another question is the nature of the innate immune responsein patients with nonalcoholic cirrhosis. Monocytes from patientswith primary biliary cirrhosis exhibit ex vivo hyper-production ofpro-inflammatory cytokines in response to engagement of sev-eral TLRs, including TLR4 [44]. On the other hand, little is knownon the innate immune response to LPS in PBMCs or monocytes


from patients with HCV- or HBV-related cirrhosis. Obviously,studies are needed in these patients because they may have alter-ations of their innate immune responses to TLR engagement. Invitro studies in non immune cells have shown that the HCVNS3/4A serine protease cleaves TRIF, an adaptator molecule ofTLR4 and TLR3 signaling, inhibiting the phosphorylation of IRF3and the transcription of pro-inflammatory cytokines [45,46]. Inmonocyte-derived dendritic cells from normal subjects, ex vivopretreatment with the HCV core protein reduces the LPS-inducedpro-inflammatory cytokine production [47]. The amount ofex vivo pro-inflammatory cytokine production by LPS-stimulatedperipheral blood monocytes is significantly lower in patientswith chronic HCV infection than in healthy subjects [47]. Inpatients with chronic hepatitis B, the presence of HBeAg reduceswhile its absence enhances the pro-inflammatory response ofperipheral blood monocytes to LPS [48].

LPS-induced TNF-a is an early mediator of sepsis which playsa crucial role in the development of organ failures and subse-quent death [49]. Since in vivo and ex vivo studies have shownthat LPS-stimulated cirrhotic immune cells have an early over-production of TNF-a, this over-production is probably involvedin the excessive severity of sepsis in patients with cirrhosis [1].Our finding that cell-permeable inhibitors of GSK3 significantlydecrease TNF-a hyper-production by LPS-stimulated cirrhoticPBMCs suggests that inhibition of GSK3 may be a novel approachin the treatment of sepsis in patients with cirrhosis.

In conclusion, during the first hours of exposure to LPS, cir-rhotic PBMCs are characterized by a pro-inflammatory phenotypewhich associates increased TNF-a and decreased IL-10. This wasassociated with a defect in AKT1 expression and in AKT-mediatedGSK3b phosphorylation, resulting in unrestricted ‘pro-inflamma-tory’ activity of the enzyme. In LPS-stimulated cirrhotic PBMCs,cell permeable inhibitors of GSK3 decrease the production ofpro-inflammatory cytokines and increase IL-10, showing a roleof constitutively active GSK3 in the ‘pro-inflammatory’ profile ofthese cells. GSK3 may be a target for novel therapeutic approachesaimed to decrease the excessive pro-inflammatory response tobacterial infections in patients with advanced cirrhosis.

Conflict of interest

The authors who have taken part in this study declared that theydo not have anything to disclose regarding funding or conflict ofinterest with respect to this manuscript.

Acknowledgments

This work was supported by the Institut National de la Santé et dela Recherche Médicale (INSERM), the Laboratoire Français duFractionnement et des Biotechnologies, and Ferring S.A.S. M.S.R.was supported by a grant from the Fondation pour la RechercheMédicale. T.G. is a post-doctoral researcher of the Belgian Fondsde la Recherche Scientifique (FNRS). R.M. is in receipt of an IN-SERM-APHP Fellowship.

Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jhep.2010.12.039.

1 vol. 55 j 784–793 791

http://dx.doi.org/10.1016/j.jhep.2010.12.039

Research Article
References
[1] Gustot T, Durand F, Lebrec D, Vincent JL, Moreau R. Severe sepsis in cirrhosis.Hepatology 2009;50:2022–2033.

[2] Moreau R, Hadengue A, Soupison T, Kirstetter P, Mamzer MF, Assous M,et al. Clinical, hemodynamic and metabolic characteristics and ICU outcomeof septic shock in patients with cirrhosis. Crit Care Med 1992;20:746–750.

[3] Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, et al.Sepsis Occurrence in Acutely Ill Patients Investigators. Sepsis in Europeanintensive care units: results of the SOAP study. Crit Care Med 2006;34:344–353.

[4] Tazi KA, Bièche I, Paradis V, Guichard C, Laurendeau I, Dargère D, et al. In vivoaltered unfolded protein response and apoptosis in livers from lipopolysac-charide-challenged cirrhotic rats. J Hepatol 2007;46:1075–1088.

[5] Le Moine O, Marchant A, De Groote D, Azar C, Goldman M, Deviere J. Role ofdefective monocyte interleukin-10 release in tumor necrosis factor-alphaoverproduction in alcoholics cirrhosis. Hepatology 1995;22:1436–1439.

[6] Byl B, Roucloux I, Crusiaux A, Dupont E, Devière J. Tumor necrosis factor aand interleukin 6 plasma levels in infected cirrhotic patients. Gastroenter-ology 1993;104:1492–1497.

[7] Tazi KA, Quioc JJ, Saada V, Bezeaud M, Lebrec D, Moreau R. Upregulation ofTNF-alpha production signaling pathways in monocytes from patients withadvanced cirrhosis: possible role of Akt and IRAK-M. J Hepatol 2006;45:280–289.

[8] Tazi KA, Quioc JJ, Abdel-Razek W, Tellier Z, Guichart C, Ogier-Denis E, et al.Protein array technology to investigate cytokine production by monocytesfrom patients with advanced alcoholic cirrhosis: an ex vivo pilot study.Hepatol Res 2009;39:706–715.

[9] Galbois A, Thabut D, Tazi KA, Rudler M, Mohammadi MS, Bonnefont-Rousselot D, et al. Ex vivo effects of high-density lipoprotein exposure on thelipopolysaccharide-induced inflammatory response in patients with severecirrhosis. Hepatology 2009;49:175–184.

[10] Woodgett JR. Molecular cloning and expression of glycogen synthase kinase-3/factor A. EMBO J 1990;9:2431–2438.

[11] Martin M, Rehani K, Jope RS, Michalek SM. Toll-like receptor-mediatedcytokine production is differentially regulated by glycogen synthase kinase3. Nat Immunol 2005;6:777–784.

[12] Rehani K, Wang H, Garcia CA, Kinane DF, Martin M. Toll-like receptor-mediated production of IL-1Ra is negatively regulated by GSK3 via the MAPKERK1/2. J Immunol 2009;182:547–553.

[13] Weintz G, Olsen JV, Fruhauf K, Niedzielska M, Amit I, Jantsch J, et al. Thephosphoproteome of toll-like receptor-activated macrophages. Mol Syst Biol2010;6:371.

[14] Hu X, Paik PK, Chen J, Yarilina A, Kockeritz L, Lu TT, et al. IFN-gammasuppresses IL-10 production and synergizes with TLR2 by regulating GSK3and CREB/AP-1 proteins. Immunity 2006;24:563–574.

[15] Ohtani M, Nagai S, Kondo S, Mizuno S, Nakamura K, Tanabe M, et al.Mammalian target of rapamycin and glycogen synthase kinase 3 differen-tially regulate lipopolysaccharide-induced interleukin-12 production indendritic cells. Blood 2008;112:635–643.

[16] Hoeflich KP, Luo J, Rubie EA, Tsao MS, Jin O, Woodgett JR. Requirement forglycogen kinase-3beta in cell survival and NF-kappaB activation. Nature2000;406:86–90.

[17] MacAulay K, Doble BW, Patel S, Hansotia T, Sinclair EM, Drucker DJ, et al.Glycogen synthase kinase 3alpha-specific regulation of murine hepaticglycogen metabolism. Cell Metab 2007;6:329–337.

[18] Vignali DA. Multiplexed particle-based flow cytometric assays. J Immunolmethods 2000;243:243–255.

[19] Kellar KL, Iannone MA. Multiplexed microsphere-based flow cytometricassays. Exp Hematol 2002;30:1227–1237.

[20] Varro R, Chen R, Sepulveda H, Apgar J. Bead-based multianalyte flowimmunoassays: the cytometric bead array system. Methods Mol Biol2007;378:125–152.

[21] Cook EB, Stahl JL, Lowe L, Chen R, Morgan E, Wilson J, et al. Simultaneousmeasurement of six cytokines in a single sample of human tears usingmicroparticle-based flow cytometry: allergics vs. non-allergics. J ImmunolMethods 2001;254:109–118.

[22] Prunet C, Montange T, Véjux A, Laubriet A, Rohmer JF, Riedinger JM, et al.Multiplexed flow cytometric analyses of pro- and anti-inflammatorycytokines in the culture media of oxysteral-treated human monocytic cellsand in the sera of atherosclerotic patients. Cytometry A 2006;69: 359–373.

[23] Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods2001;25:402–408.


[24] Deviere J, Content J, Denys C, Vandenbussche P, Schandene L, Wybran J, et al.Excessive in vitro bacterial lipopolysaccharide-induced monokines incirrhosis. Hepatology 1990;11:628–634.

[25] Coghlan MP, Culbert AA, Cross DAE, Corcoran SL, Yates JW, Pearce NJ,et al. Selective small molecule inhibitors of glycogen synthase kinase-3modulate glycogen metabolism and gene transcription. Chem Biol 2000;7:793–803.

[26] Klein PS, Melton DA. A molecular mechanism for the effect of lithium ondevelopment. Proc Natl Acad Sci USA 1996;93:8455–8459.

[27] Trinchieri G. Interleukin-12 and the regulation of innate resistance andadaptive immunity. Nat Rev Immunol 2003;3:133–146.

[28] Rodionova E, Conzelmann M, Maraskovsky E, Hess M, Kirsch M, Giese T, et al.GSK-3 mediates differentiation and activation of proinflammatory dendriticcells. Blood 2007;109:1584–1592.

[29] Beaulieu JM, Marion S, Rodriguiz RM, Medvedev IO, Sotnikova TD, Ghisi V,et al. A beta-arrestin 2 signaling complex mediates lithium action onbehavior. Cell 2008;132:125–136.

[30] Kumar CC, Madison V. AKT crystal structure and AKT-specific inhibitors.Oncogene 2005;24:7493–7501.

[31] Nicholson KM, Anderson NG. The protein kinase B/Akt signaling pathway inhuman malignancy. Cell Signal 2002;14:381–395.

[32] Brognard J, Sierecki E, Gao T, Newton AC. PHLPP and a second isoform,PHLPP2, differentially attenuate the amplitude of Akt signalling by regulat-ing distinct Akt isoforms. Mol Cell 2007;25:917–931.

[33] Wasmuth HE, Kunz D, Yagmur E, Timmer-Stranghöner A, Vidacek D, SiewertE, et al. Patients with acute on chronic liver failure display ‘‘sepsis-like’’immune paralysis. J Hepatol 2005;42:195–201.

[34] Androulidaki A, Iliopoulos D, Arranz A, Doxaki C, Schworer S, ZacharioudakiV, et al. The kinase Akt1 controls macrophage response to lipopolysaccharideby regulating microRNAs. Immunity 2009;31:220–231.

[35] Verma BK, Fogarasi M, Szabo G. Down-regulation of tumor necrosis factoralpha activity by acute ethanol treatment in human peripheral bloodmonocytes. J Clin Immunol 1993;13:8–22.

[36] Szabo G, Chavan S, Mandrekar P, Catalano D. Acute alcohol consumptionattenuates interleukin-8 (IL-8) and monocytes cheoattractant peptide-1(MCP-1) induction in response to ex vivo stimulation. J Clin Immunol1999;19:67–76.

[37] Szabo G, Mandrekar P, Girouard L, Catalano D. Regulation of humanmonocyte functions by acute ethanol treatment: decreased tumor necrosisfactor-alpha, interleukin-1 beta and elevated interleukin-10, and trans-forming growth factor-beta production. Alcohol Clin Exp Res 1996;20:900–907.

[38] Mandrekar P, Bala S, Catalano D, Kodys K, Szabo G. The opposite effectsof acute and chronic alcohol on lipopolysaccharide-induced inflammationare linked to IRAK-M in human monocytes. J Immunol 2009;183:1320–1327.

[39] Ma H, Yu L, Byra EA, Hu N, Kitagawa K, Nakayama KI, et al. Aldehydedehydrogenase 2 knockout accentuates ethanol-induced cardiac depres-sion: role of protein phosphatase. J Mol Cell Cardiol 2010;49:322–329.

[40] Gobejishvili L, Barve S, Joshi-Barve S, Uriarte S, Song Z, McClain C. Chronicethanol-mediated decrease in cAMP primes macrophages to enhancedLPS-inducible NF-kappaB activity and TNF expression: relevance toalcoholic liver disease. Am J Physiol Gastrointest Liver Physiol 2006;291:G681–688.

[41] Gobejishvili L, Barve S, Joshi-Barve S, McClain C. Enhanced PDE4B expressionaugments LPS-inducible TNF expression in ethanol-primed monocytes:relevance to alcoholic liver disease. Am J Physiol Gastrointest Liver Physiol2008;295:G718–724.

[42] Bala S, Marcos M, Kodys K, Csak T, Catalano D, Mandrekar P, et al.Upregulation of microRNA-155 in macrophages contributes to increasedtumor necrosis factor alpha (TNF{alpha}) production via increased mRNAhalf-life in alcoholic liver disease. J Biol Chem 2010 [Epub ahead ofprint].

[43] Albillos A, Herza Ad Ade L, Reyes E, Monserrat J, Munoz L, Nieto M, et al.Tumor necrosis factor-alpha expression by activated monocytes and alteredT-cell homeostasis in ascetic alcoholic cirrhosis: amelioration with norflox-acin. J Hepatol 2004;40:624–631.

[44] Mao TK, Lian ZX, Selmi C, Ichiki Y, Ashwood P, Ansari AA, et al. Alteredmonocyte responses to defined TLR ligands in patients with primary biliarycirrhosis. Hepatology 2005;42:802–808.

[45] Foy E, Li K, Wang C, Sumpter Jr R, Ikeda M, Lemon SM, et al. Regulation ofinterferon regulatory factor-3 by the hepatitis C virus serine protease.Science 2003;300:1145–1148.

1 vol. 55 j 784–793

[46] Li K, Foy E, Ferreon JC, Nakamura M, Ferreon AC, Ikeda M, et al. Immune
evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptator protein TRIF. Proc Natl Acad Sci USA 2005;102:2992–2997.

[47] Chung H, Watanabe T, Kudo M, Chiba T. Hepatitis C virus core proteininduces homotolerance and cross-tolerance to Toll-like receptor ligands byactivation of Toll-like receptor 2. J Infect Dis 2010;202:853–861.


[48] Visvanathan K, Skinner NA, Thompson AJ, Riordan SM, Sozzi V, Edwards R,et al. Regulation of Toll-like receptor-2 expression in chronic hepatitis B bythe precore protein. Hepatology 2007;45:102–110.

[49] Grivennikov SI, Tumanov AV, Liepinsh DJ, Kruglov AA, Marakusha BI,Shakhov AN, et al. Distinct and nonredundant in vivo functions of TNFproduced by T cells and macrophages/neutrophils: protective and deleteri-ous effects. Immunity 2005;22:93–104.

1 vol. 55 j 784–793 793

Documents

Glycogen synthase kinase 3 involvement in the excessive proinflammatory response to LPS in patients with decompensated cirrhosis