Alcohol 27 (2002) 37–41

0741-8329/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.PII:

S0741-8329(02)00213-6

Lymphocyte-mediated liver injury in alcohol-related hepatitis

Robert G. Batey*, Qi. Cao, Belinda Gould

Department of Gastroenterology, John Hunter Hospital, New Lambton Heights, Newcastle 2305, NSW, Australia

Received 23 November 2001; received in revised form 11 February 2002; accepted 11 February 2002

Abstract

The pathogenesis of alcohol-related liver disease (ALD) remains inadequately explained. Increasing alcohol intake is associated withan increased risk of ALD, but many heavy drinkers develop no liver damage. An explanation for ALD susceptibility requires theoriesthat extend beyond a biochemical understanding of alcohol metabolism. Several hepatic cell populations are involved in the pathogenesisof liver injury. The liver-associated lymphocyte (LAL) response to alcohol intake plus immune stimulation may determine susceptibilityto liver damage. We have isolated rat LALs and demonstrated the following: (1) Liver-associated lymphocytes differ from the peripheralblood lymphocyte pool; the CD8:CD4 ratio is higher in the LAL population than in peripheral blood. (2) Tumor necrosis factor-alpha(TNF-

�

) and interleukin-6 production by these cells is suppressed by regular alcohol intake. (3) Tumor necrosis factor-

�

and interleukin-6 production by LALs is increased after parenteral administration of concanavalin A (Con A) and by Con A in in vitro LAL cultures ob-tained from healthy (control) and ethanol-consuming rats. (4) In vivo stimuli that lead to increased cytokine production by LALs lead,within 12–24 h, to increased hepatocyte necrosis [elevated alanine aminotransferase (ALT) levels] and apoptosis. (5) Liver-associatedlymphocytes isolated from ethanol-consuming rats, transferred to non-ethanol-consuming rats, confer on the latter animals an ethanol-consuming response to Con A. (6) Cytokine release by LALs is quantitatively as significant as that from Kupffer cells after exposure to li-popolysaccharide. (7) In co-culture studies inhibition of TNF-

�

activity reduces hepatocyte apoptosis induced in the presence of activatedLALs. (8) Finally, nuclear factor-kappa B inhibition decreases production of nitric oxide and TNF-

�

, with an associated reduction inhepatocyte apoptosis. In summary, our study findings support the suggestion that a role for LALs exists in the pathogenesis of alcohol andCon A–mediated liver disease. © 2002 Elsevier Science Inc. All rights reserved.

Keywords:

Alcohol-related liver disease; Cytokines; Liver-associated lymphocytes

1. Introduction

This article is a review of research studies that have beenundertaken in our laboratory. The experiments have beendesigned to increase understanding of the role of liver-asso-ciated lymphocytes (LALs) and liver-associated mononu-clear cells (MNCs) in the pathogenesis of alcohol-relatedliver disease (ALD). An evolving understanding of alcoholmetabolism has not provided an explanation for the fact thatliver injury develops in only a minority of people who drink

�

100 g of alcohol per day. Factors other than direct toxicityof alcohol or its metabolites need to be considered as possi-ble pathogenetic factors for alcohol-related liver injury.Once established, ALD can progress in the absence of alco-

hol consumption, supporting the suggestion that immuno-logical mechanisms may be responsible, at least in part, forthe pathologic changes in the liver in heavy drinkers. Ourresearch has evolved through the isolation of mononuclearcells from the livers of rats exposed to ethanol to a more de-tailed study of the LAL population. Our findings seem to in-dicate that these cells are capable of producing inflamma-tory and proinflammatory cytokines, which could be criticalto the evolution of ALD. Brief details of the experimentalconditions used in the studies are presented, results are sum-marized, and the article concludes with a brief review of thecurrent literature in this field.

2. Methods

In all studies, female Wistar rats, ranging in age from 8to 10 weeks at the commencement of the studies, have beenused. The strain used will consume 40% ethanol in water,developing fatty liver with mild to moderate inflammationand fibrosis if exposure continues for 26 weeks (Keegan et

* Corresponding author. Department of Gastroenterology, JohnHunter Hospital, Locked Bag 1, Hunter Regional Mail Centre, Newcastle,2310, NSW, Australia. Tel.:

�

61-2-4921-3486; fax:

�

61-2-4921-3998.

E-mail address

: rbatey@hunter.health.nsw.gov.au

(R.G. Batey).Editor: T.R. Jerrells

38

R.G. Batey et al. / Alcohol 27 (2002) 37–41

al., 1995). In all studies, animals were provided a drinkingsolution of ethanol [increasing in concentration from 5%(vol./vol.) to 40% (vol./vol.) in 2% sucrose (wt./vol.)], iso-caloric sucrose, or a constant solution of 2% sucrose (Kee-gan et al., 1995). In the most recent studies, ethanol has alsobeen incorporated into 2% (wt./vol.) agar gels to increaseintake. Animals are maintained on the 40% ethanol for 6 or8 weeks (this varied in different protocols) before studies ofthe role of MNCs or LALs in hepatic injury. Details of theanimal housing and daily monitoring are included in ourpublications (Batey et al., 1998, 1999; Cao et al., 1998a,1998b, 1999; Keegan et al., 1995).

Ethanol-consuming animals were subjected to additionaltreatments designed to increase our understanding of factorsthat may modify susceptibility to alcohol-mediated injury.These were

1. intravenous concanavalin A (Con A) administrationat a dose of 20 mg/kg in 300

�

l of phosphate-bufferedsaline (PBS). Control rats received 300

�

l of PBS in-travenously.

2. intravenous lipopolysaccharide (LPS) administrationat a dose of 1 mg per kilogram of body weight in 300

�

l of PBS. Lipopolysaccharide was derived from

Es-cherichia coli

(serotype 026: B6, Sigma-Aldrich FineChemicals, Sydney, Australia). Control rats received300

�

l of PBS intravenously.

2.1. Liver perfusion for the isolation of liver-associatedmononuclear cells

After animals were anesthetized, the livers were preper-fused in situ through the portal vein with sterile PBS contain-ing 0.1% ethylenediaminetetraacetic acid (EDTA) at a pres-sure of 10 cm water, giving a flow rate of 10 ml/min. Next,livers were perfused with 300 ml of PBS/EDTA solution at apressure of 40 cm water (flow rate of 30 ml/min). The perfu-sate was centrifuged at 1,500 rpm for 15 min, and cells wereprepared as previously described (Batey et al., 1998).

2.2. Isolation of CD4

�

, CD5

�

, and CD8

�

T cells

Isolated MNCs, obtained from liver perfusate, were cul-tured with mouse anti-rat CD5 (MRC OX-19), CD4 (W3/25),or CD8 at 4

°

C for 45 min. Incubated cells were collected bycentrifugation at 800

g

for 10 min, resuspended, and washedtwice with Hanks’ balanced salt solution (HBSS). Dynabeadsm-450 sheep anti-mouse immunoglobulin G (IgG) was addedto the cells (2

�

10

7

Dynabeads per milliliter). After incuba-tion for 10 min, the cells of interest were obtained by a mag-netic particle concentrator and washed three times withRPMI-1640. In some experiments, the above-described pro-cedure was followed, but cells were separated by using aFACscan cell sorter. The isolated T cells were regulated into20

�

10

6

cells per milliliter for T cell transfer, 1

�

10

6

cellsper milliliter for culture, or 5

�

10

6

cells per milliliter for co-culture with hepatocytes. Viability of liver-associated T cellswas

�

98%, as confirmed with the use of trypan-blue exclu-

sion. The purity of CD4

�

, CD8

�

, and CD5

�

T cells, as deter-mined by flow cytometry, was

�

95%.Hepatocytes for co-culture studies were isolated by stan-

dard techniques (Seglen, 1976), and apoptosis was mea-sured as described previously.

All animal and human subject studies were reviewed andapproved by the Hunter Area Health Service Research Eth-ics Committee.

3. Results and discussion

3.1. Animal studies

Results of early experiments demonstrated that chronicdietary ethanol consumption in rats resulted in the suppres-sion of interleukin (IL)-1

�

and tumor necrosis factor-alpha(TNF-

�

) production by intrahepatic perfusate mononuclearcells (Batey et al., 1998). Production of IL-6 remained con-stant throughout the study. Injection of LPS led to an in-creased production of IL-6 and TNF-

�

by the intrahepaticmononuclear cells (ethanol-fed rats

�

control rats), and thisincrease was mirrored in elevated serum IL-1

�

and TNF-

�

levels, which preceded an elevation in the hepatic enzymealanine aminotransferase (ALT) by 6–12 h. The proportionof mononuclear cells, lymphocytes (~60%), monocytes andlarge granular lymphocytes (~20%), and Kupffer cells(~18%) in the liver perfusate of control and ethanol-fed ani-mals was similar. Ethanol consumption alone does not leadto activation of MNCs or recruitment of T cells to the liver.However, stimuli such as LPS can up-regulate TNF-

�

pro-duction by the mononuclear cell population. The initialstudies did not allow the specific cells responsible for pro-ducing TNF-

�

to be defined. Results of these studies werethe first to support the suggestion that intrahepatic T cellsmight be closely related to the evolving lesion in ethanol-exposed animals.

In 1997, hepatitis had not been shown to develop in ratsafter Con A injection, all studies having been reported inmice. We exposed rats to an intravenous injection of Con Aand isolated LALs from the liver perfusate 24 h later. Wedemonstrated that this model also results in an acute hepati-tis and characterized the LAL response to the Con A injec-tion (Cao et al., 1998b). Con A led to an increase in the per-centages of CD4

�

, CD5

�

, and CD25

�

T cells in blood andthe liver perfusate. The CD25

�

cells have been shown to beinvolved in the pathogenesis of Con A hepatitis in themouse. Increased levels of IL-6 and TNF-

�

were derivedfrom liver perfusate CD4

�

cells. In a subsequent study, weexamined the effect of Con A injection in ethanol-fed rats(Cao et al., 1998a). In this study, a single injection of Con Aafter 8 weeks of exposure to 40% ethanol in drinking waterled to a severe hepatic injury, preceded by increases in IL-6,TNF-

�

, and ALT levels in peripheral blood. Increased num-bers of activated CD25

�

T cells and increased IL-6 andTNF-

�

production by cultured intrahepatic CD4

�

T cellswere found in ethanol-fed, Con A–exposed rats. In an at-

R.G. Batey et al. / Alcohol 27 (2002) 37–41

39

tempt to document the important role of LALs in this injury,studies were designed in which 20

�

10

6

T cells obtainedfrom ethanol-consuming rats were adoptively transferred tonondrinking rats, which were then injected with Con A.

Results of these studies showed that the LAL cells confera drinking phenotype on nondrinking animals. Cells ob-tained from control animals did not alter the response of therecipients to Con A. The findings seem to indicate that etha-nol consumption induced a disregulated T cell population,which, when activated by a polyclonal T cell activator (ConA), produced excessive amounts of injurious cytokines(Batey et al., 1999; Cao et al., 1999).

In the two models of liver injury used in our studies (ConA or LPS injection in ethanol-consuming rats), we have dem-onstrated different patterns of T cell and Kupffer cell cyto-kine (IL-6 and TNF-

�

) production. In the LPS model,Kupffer cell cytokine release exceeds that of LALs,whereas in the Con A model, LAL cytokine production permillion cells is greater. The LALs are capable of produc-ing similar concentrations of the measured cytokines, asdo Kupffer cells.

In recent studies, we have used co-culture techniques toexamine the mechanisms by which freshly isolated LALs,obtained from ethanol-consuming rats, influence hepatocyteapoptosis. The co-culture of CD4

�

LALs (5

�

10

6

cells permilliliter) with hepatocytes (1

�

10

6

cells per milliliter) wasdone with the use of culture inserts. The T cell culture me-dium contained Con A (5

�

g/ml). We used a series of ex-periments to assess the effect of monoclonal antibodiesagainst TNF-

�

(40

�

g/ml) or interferon (IFN)-gamma (40

�

g/ml) and pyrrolidine dithiocarbamate, a nuclear factor-kappa B (NF-

�

B) inhibitor (100

�

mol/l) in the culture me-dium on the T cell contribution to hepatocyte apoptosis. Theindividual agents provided insight into the mechanisms bywhich apoptosis might be initiated.

3.2. Human subject studies

We have also studied peripheral blood mononuclear cell(PBMC) cultures stimulated with LPS (1

�

g/ml culture con-centration) and Con A (5

�

g/ml culture concentration) andwhole blood cultures stimulated with LPS (1

�

g/ml cultureconcentration) in the following populations:

1. healthy controls;2. heavy drinkers (defined as consumption of alcohol at

�

80 g/day for

�

5 years for men and

�

40 g/day for

�

5 years for women) with results of liver functiontests within normal limits;

3. drinkers with abnormal results for liver function tests;4. patients with alcohol-related hepatitis; and5. patients with alcohol-related cirrhosis who were ei-

ther currently drinking or abstinent.

We studied these groups to measure IL-6, TNF-

�

, andIFN-

�

production. There were no significant differences inPBMC production of TNF-

�

, IFN-

�

, or IL-6 among the

study groups. With the use of whole blood cultures, wefound that individuals with alcoholic hepatitis had signifi-cantly lower production of TNF-

�

(

P

.008), IFN-

�

(

P

.008), and IL-6 (

P

.017) than controls. Patients with cir-rhosis had lower production levels of TNF-

�

(

P

.003) andIFN-

�

(

P

.017) than all other groups. The significance ofthese changes remains to be fully elucidated.

4. Review of the literature

4.1. Cytokine measurements in alcohol-related liver disease

Many investigators have sought to define the nature,extent, and significance of blood and hepatic cytokinechanges associated with ALD. Study results are not con-sistent, reflecting variable study design and the wide varia-tion in cytokine levels in healthy populations (Cook, 1998;van der Linden et al., 1998; Wilson et al., 1991). Geneticvariability in the response of individuals to stimuli, lead-ing to cytokine release, may provide an explanation for thedifferences in susceptibility to diseases such as ALD(Grove et al., 1997, 2000). In human beings, serum levelsand peripheral blood mononuclear cell production of TNF-

�

and IL-6 tend to be increased in alcohol-related hepatitisand cirrhosis, whereas IFN-

�

levels tend to be reduced(Bird et al., 1990; Daniluk & Kandefer-Szerszen, 1994b,1997; Daniluk et al., 1996; Deviere et al., 1989, 1990,1991, 1992; Felver et al., 1990). Levels of TNF-

�

havebeen correlated with survival in patients with alcohol-relatedhepatitis (Khoruts et al., 1991). Levels of IL-8 have also beenshown to be elevated in alcohol-related hepatitis. It may bedirectly involved in increasing the neutrophil infiltrate inthe liver in this condition (Hill et al., 1993). The mono-cytes of patients with alcohol-related hepatitis have beenshown to produce increased amounts of TNF-

�

. No pro-spective studies of individuals progressing from a healthystate to ALD have been published to identify a pattern ofcytokine release that predicts future severe liver disease.Animal models of ALD provide variable results, but over-all the study findings support the suggestion that risingserum levels of TNF-

�

and IL-6 precede overt ALD(Enomoto et al., 1998; Honchel et al., 1992; Kamimura &Tsukamoto, 1995; Pennington et al., 1997). Some investi-gators have examined liver mononuclear cell and Kupffercell production of cytokines (Batey et al., 1998; Daniluk& Kandefer-Szerszen, 1994a; Enomoto et al., 1998; Ka-mimura & Tsukamoto, 1995).

4.2. Lymphocyte changes in alcohol-related liver disease

Alcohol-related liver disease is associated with changesin peripheral blood lymphocyte numbers. Changes dependon the severity of the liver injury, with lymphopenia charac-terizing the presence of cirrhosis. T cells from human be-ings who regularly consume

�

80 g of alcohol per day ap-pear to be “activated,” and the activated state persists afteralcohol has been withdrawn (Cook, 1998; Laso et al., 1996).

40

R.G. Batey et al. / Alcohol 27 (2002) 37–41

4.3. Liver-associated lymphocytes and alcohol-relatedliver disease

The human liver has an intrinsic lymphocyte populationnormally resident in the portal tract. CD3

�

T lymphocytespredominate, and B lymphocytes make up only 5% of thispopulation. Up to 15% of hepatic CD3

�

cells express the

�

T cell receptor (TCR) compared with 2.7% of CD3

�

cells inblood. Almost 50% of these

�

T cells co-express CD8. Re-sults of studies have increasingly clarified our understand-ing of the intrahepatic lymphocyte population. Cells ex-pressing both the T cell marker CD3 and the natural killer(NK) cell marker CD56 (natural killer T cell) constitutemore than 30% of all hepatic CD3

�

lymphocytes. Thesecells are not commonly found in peripheral blood, indicat-ing the specific nature of the hepatic or liver-associatedlymphocyte population. Study results have revealed thepresence of CD4

–

CD8

–

T cells in the liver, and findings ob-tained from studies that were largely carried out on miceindicate that these develop extrathymically, responding toIL-3. These cells express reduced levels of the CD3 TCRcomplex and preferentially express certain TCR V

�

genes(Kita et al., 2001). Although the LAL population can be de-scribed in great detail, still to be defined is how LALs con-tribute to the pathogenesis of liver disease. Kupffer cells,stellate cells, hepatic endothelial cells, and LALs are influ-enced by the state of activation of the other cells, and cyto-kine profiles taken from any situation other than specificcell cultures may reflect increased production by more thanone cell line.

Quantitative immunohistochemical analysis of LAL sub-sets in liver biopsy samples obtained from patients withALD reveals that increased numbers of CD4

�

and CD8

�

Tcells correlate with regenerating nodules, lobular inflamma-tion, and central sclerosis, whereas B cells and NK cells arerare or absent in these situations (Chedid et al., 1993; Sakaiet al., 1993). Long-term ethanol consumption seems to de-press NK cell function in animals (Abdallah et al., 1988;Meadows et al., 1989). In human beings, an increase in NKcells (CD3

– CD56� positive) and NK activity, as well as Tcells co-expressing CD3 and CD56, is detected in peripheralblood in subjects with alcohol-related hepatitis. Conversely,hepatic NK cell numbers are not increased in biopsy sam-ples obtained from patients with ALD, unlike changes seenin chronic active and chronic viral hepatitis (Chedid et al.,1993).

4.4. Whole blood culture techniques

The whole blood culture technique was first described 26years ago, but only more recently has it been used to studyvarious disease states (Hall & Gordon, 1976). T cell func-tion and cytokine production in rheumatoid arthritis(Zangerle et al., 1992) and in HIV infection (Filella et al.,1998) and colorectal carcinoma (Elsasser-Beile et al., 1992)have been described. Optimal conditions for evaluatingwhole blood cultures are yet to be fully defined, and results

with this technique do need to be reproduced in more thanone laboratory. The claimed benefit of the use of wholeblood is that it allows the different cell lines to interact invitro, regulating cytokine production more physiologically.Our use of whole blood culture technology was driven by adesire to determine whether this approach gave more repro-ducible results than did the use of isolated PBMCs. Theshortcomings of the whole blood technique, including theinability of the researcher to control lymphocyte numbers inthe peripheral blood, are acknowledged.

5. Future studies

It is suggested that further studies to define the role ofliver-associated T cells in alcoholic liver disease should in-clude co-culture studies to determine the influence of liver-associated T cells on Kupffer cell activation in healthy andalcohol-exposed livers. Studies to characterize the differ-ence in the responses of peripheral blood and LALs in ConA–induced hepatitis in the rat and in ALD in the rat and hu-man subjects need to be undertaken to determine whetherthere are peripheral blood studies that could indicate inwhich patients there is an increased risk of liver injury de-veloping. Whole blood culture techniques should be evalu-ated further in this context as the technique may prove eas-ier to use than the isolated PBMC culture.

The use of gene-knockout mice to allow a study of theimportance of specific cytokines—i.e., IL-6, TNF-�, IFN-�,IL-10, and IL-12—could advance understanding of the roleof T cell, Kupffer cell, and stellate cell activation in thepathogenesis of ALD.

It is hoped that future studies will be designed and con-ducted to define the importance of the LAL population inALD, as this condition continues to be responsible for sig-nificant morbidity and mortality worldwide.

References

Abdallah, R. M., Starkey, J. R., & Meadows, G. G. (1988). Toxicity ofchronic high alcohol intake on mouse natural killer cell activity. ResCommun Chem Pathol Pharmacol 59, 245–258.

Batey, R., Cao, Q., Madsen, G., Pang, G., Russell, A., & Clancy, R. (1998).Decreased tumor necrosis factor-alpha and interleukin-1alpha produc-tion from intrahepatic mononuclear cells in chronic ethanol consump-tion and upregulation by endotoxin. Alcohol Clin Exp Res 22, 150–156.

Batey, R. G., Clancy, R. L., Pang, G. T., & Cao, Q. (1999). Alcoholic hep-atitis as a T-cell mediated disorder: an hypothesis. Alcohol Clin ExpRes 23, 1207–1209.

Bird, G. L., Sheron, N., Goka, A. K., Alexander, G. J., & Williams, R. S.(1990). Increased plasma tumor necrosis factor in severe alcoholic hep-atitis. Ann Intern Med 112, 917–920.

Cao, Q., Batey, R., Pang, G., & Clancy, R. (1998a). Altered T-lymphocyteresponsiveness to polyclonal cell activators is responsible for liver cellnecrosis in alcohol-fed rats. Alcohol Clin Exp Res 22, 723–729.

Cao, Q., Batey, R., Pang, G., & Clancy, R. (1999). Ethanol-altered liver-as-sociated T cells mediate liver injury in rats administered ConcanavalinA (Con A) or lipopolysaccharide (LPS). Alcohol Clin Exp Res 23,1660–1667.

Cao, Q., Batey, R., Pang, G., Russell, A., & Clancy, R. (1998b). IL-6, IFN-

R.G. Batey et al. / Alcohol 27 (2002) 37–41 41

gamma and TNF-alpha production by liver-associated T cells and acuteliver injury in rats administered concanavalin A. Immunol Cell Biol 76,542–549.

Chedid, A., Mendenhall, C. L., Moritz, T. E., French, S. W., Chen, T. S.,Morgan, T. R., Roselle, G. A., Nemchausky, B. A., Tamburro, C. H.,Schiff, E. R., McClain C. J., Marsano, L. S., Allen, J. I., Samanta, A.,Weesner, R. E., & Henderson, W. G. (1993). Cell-mediated hepatic in-jury in alcoholic liver disease. Veterans Affairs Cooperative StudyGroup 275. Gastroenterology 105, 254–266.

Cook, R. T. (1998). Alcohol abuse, alcoholism, and damage to the immunesystem—a review. Alcohol Clin Exp Res 22, 1927–1942.

Daniluk, J., & Kandefer-Szerszen, M. (1994a). Influence of acute andchronic ethanol treatment on interferon production in mice. Arch Im-munol Ther Exp (Warsz) 42, 239–246.

Daniluk, J., & Kandefer-Szerszen, M. (1994b). Interferon production byperipheral blood cells of patients with alcoholic liver disease (ALD).Arch Immunol Ther Exp (Warsz) 42, 231–238.

Daniluk, J., & Kanderfer-Szerszen, M. (1997). Isoprinosine and levamisoleas stimulators of interferon production in blood leukocytes of patientswith alcoholic liver cirrhosis. Arch Immunol Ther Exp (Warsz) 45,183–187.

Daniluk, J., Kandefer-Szerszen, M., & Borowska, L. (1996). Tumor necro-sis factor and interferon production by peripheral blood leukocytes ofpatients with alcoholic cirrhosis. Arch Immunol Ther Exp (Warsz) 44,97–101.

Deviere, J., Content, J., Denys, C., Vandenbussche, P., Le Moine, O., Schan-dene, L., Vaerman, J. P., & Dupont E. (1992). Immunoglobulin A and in-terleukin 6 form a positive secretory feedback loop: a study of normalsubjects and alcoholic cirrhotics. Gastroenterology 103, 1296–1301.

Deviere, J., Content, J., Denys, C., Vandenbussche, P., Schandene, L., Wy-bran, J., & Dupont, E. (1989). High interleukin-6 serum levels and in-creased production by leucocytes in alcoholic liver cirrhosis. Correla-tion with IgA serum levels and lymphokines production. Clin ExpImmunol 77, 221–225.

Deviere, J., Content, J., Denys, C., Vandenbussche, P., Schandene, L., Wybran,J., & Dupont, E. (1990). Excessive in vitro bacterial lipopolysaccharide-in-duced production of monokines in cirrhosis. Hepatology 11, 628–634.

Deviere, J., Vaerman, J. P., Content, J., Denys, C., Schandene, L., Vanden-bussche, P., Sibille, Y., & Dupont, E. (1991). IgA triggers tumor necro-sis factor alpha secretion by monocytes: a study in normal subjects andpatients with alcoholic cirrhosis. Hepatology 13, 670–675.

Elsasser-Beile, U., von Kleist, S., Fischer, R., & Monting, J. S. (1992). Im-paired cytokine production in whole blood cell cultures from patientswith colorectal carcinomas as compared to benign colorectal tumors incontrols. J Clin Lab Anal 6, 311–314.

Enomoto, N., Ikejima, K., Bradford, B., Rivera, C., Kono, H., Brenner, D.A., & Thurman, R. G. (1998). Alcohol causes both tolerance and sensi-tization of rat Kupffer cells via mechanisms dependent on endotoxin.Gastroenterology 115, 443–451.

Felver, M. E., Mezey, E., McGuire, M., Mitchell, M. C., Herlong, H. F.,Veech, G. A., & Veech, R. L. (1990). Tumor necrosis factor alpha pre-dicts decreased long-term survival in severe alcoholic hepatitis. Alco-hol Clin Exp Res 14, 255–259.

Filella, X., Blade, J., Montoto, S., Molina, R., Coca, F., Montserrat, E., &Ballesta, A. M. (1998). Impaired production of interleukin 6 and tu-mour necrosis factor alpha in whole blood cell cultures of patients withmultiple myeloma. Cytokine 10, 993–996.

Grove, J., Daly, A. K., Bassendine, M. F., & Day, C. P. (1997). Associa-tion of a tumor necrosis factor promoter polymorphism with suscepti-bility to alcoholic steatohepatitis. Hepatology 26, 143–146.

Grove, J., Daly, A. K., Bassendine, M. F., Gilvarry, E., & Day, C. P.(2000). Interleukin 10 promoter region polymorphisms and susceptibil-ity to advanced alcoholic liver disease. Gut 46, 540–545.

Hall, L. S., & Gordon, D. S. (1976). Reproducibility, efficacy and method-ology of mitogen-induced lymphocyte transformations by the wholeblood assay. J Immunol Methods 12, 31–38.

Hill, D. B., Marsano, L. S., & McClain, C. J. (1993). Increased plasma in-terleukin-8 concentrations in alcoholic hepatitis. Hepatology 18, 576–580.

Honchel, R., Ray, M. B., Marsano, L., Cohen, D., Lee, E., Shedlofsky, S.,& McClain, C. J. (1992). Tumor necrosis factor in alcohol enhancedendotoxin liver injury. Alcohol Clin Exp Res 16, 665–669.

Kamimura, S., & Tsukamoto, H. (1995). Cytokine gene expression byKupffer cells in experimental alcoholic liver disease. Hepatology 22,1304–1309.

Keegan, A., Martini, R., & Batey, R. (1995). Ethanol-related injury in therat: a model of steatosis, inflammation and pericentral fibrosis. J Hepa-tol 23, 591–600.

Khoruts, A., Stahnke, L., McClain, C. J., Logan, G., & Allen, J. I. (1991).Circulating tumor necrosis factor, interleukin-1 and interleukin-6 con-centrations in chronic alcoholic patients. Hepatology 13, 267–276.

Kita, H., Mackay, I. R., Van De Water, J., & Gershwin, M. E. (2001). Thelymphoid liver: considerations on pathways to autoimmune injury.Gastroenterology 120, 1485–1501.

Laso, F. J., Madruga, J. I., San Miguel, J. F., Cuidad, J., Lopez, A., AlvarezMon, M., & Orfao, A. (1996). Long lasting immunological effects ofethanol after withdrawal. Cytometry 26, 275–280.

Meadows, G. G., Blank, S., & Duncan, D. A. (1989). Modulation of naturalkiller immunomodulation and AIDS. In D. Seminara, R. R. Watson, &A. Pawlowski (Eds.), Alcohol, Immunomodulation, and AIDS: Pro-ceedings of the Alcohol-Immunology AIDS Conference, held in Tucson,Arizona, April 27–29, 1989 (p. 181). New York: Alan Liss.

Pennington, H. L., Hall, P. M., Wilce, P. A., & Worrall, S. (1997). Ethanolfeeding enhances inflammatory cytokine expression in lipopolysaccha-ride-induced hepatitis. J Gastroenterol Hepatol 12, 305–313.

Sakai, Y., Izumi, N., Marumo, F., & Sato, C. (1993). Quantitative immuno-histochemical analysis of lymphocyte subsets in alcoholic liver disease.J Gastroenterol Hepatol 8, 39–43.

Seglen, P. O. (1976). Preparation of isolated rat liver cells. Methods CellBiol 13, 29–83.

van der Linden, M. W., Huizinga, T. W., Stoeken, D. J., Sturk, A., & Wes-tendorp, R. G. (1998). Determination of tumour necrosis factor-alphaand interleukin-10 production in a whole blood stimulation system: as-sessment of laboratory error and individual variation. J Immunol Meth-ods 218, 63–71.

Wilson, B. M., Severn, A., Rapson, N. T., Chana, J., & Hopkins, P. (1991).A convenient human whole blood culture system for studying the regu-lation of tumour necrosis factor release by bacterial lipopolysaccharide.J Immunol Methods 139, 233–240.

Zangerle, P. F., De Groote, D., Lopez, M., Meuleman, R. J., Vrindts, Y.,Fauchet, F., Dehart, I., Jadoul, M., Radoux, D., & Franchimont, P.(1992). Direct stimulation of cytokines (IL-1 beta, TNF-alpha, IL-6,IL-2, IFN-gamma and GM-CSF) in whole blood: II. Application torheumatoid arthritis and osteoarthritis. Cytokine 4, 568–575.