Tumour necrosis factor á impairs function of liverderived T lymphocytes and natural killer cells inpatients with primary sclerosing cholangitis
X Bo, U Broome, M Remberger, S Sumitran-Holgersson
AbstractBackground—Primary sclerosing cholan-gitis (PSC) is considered to be a chronicautoimmune disease where infiltrating Tlymphocytes have been implicated in thedestruction of bile ducts. Altered functionof these T cells may reflect abnormalitiesin the immune response leading to tissuedamage.Aim—We investigated the proliferativeand functional capacity of freshly isolatedliver derived T lymphocytes (LDLs) andnatural killer (NK) cells from PSC pa-tients.Methods—The proliferative responses tocommon mitogens such as phytohaemag-glutinin (PHA), concanavalin A (Con A),and lipopolysaccharide (LPS) were stud-ied, and the cytotoxic function of Tlymphocytes was measured using alloge-neic target cells. NK (CD56+/16+) cytotoxicfunction was measured using the two celllines K562 (NK sensitive) and Rajilymphoma cells (NK resistant).Results—Compared with patients withprimary biliary cirrhosis (PBC), auto-immune hepatitis (AIH), and normal con-trols (without liver disease), in PSC: (1)LDLs contained a low percentage of Tcells; (2) there was significantly decreasedexpression of interleukin (IL)-2 receptor(p<0.001) on activated T cells (HLA-DR+);(3) LDLs but not peripheral blood lym-phocytes had significantly impaired pro-liferative responses to mitogens such asPHA, Con A, and LPS (p< 0.001); (4) nocytotoxic activity of PSC liver T and NKcells was recorded; (5) significantly higherlevels of tumour necrosis factor á (TNF-á)and IL-1â but lower levels of IL-2, IL-10,and interferon ã were found in the super-natants of mitogen stimulated LDL cul-tures (p<0.001); (6) higher percentages offreshly isolated PSC LDLs containedintracytoplasmic TNF-á and IL-1â; and(7) pretreatment of PSC LDLs in vitrowith neutralising TNF antibodies signifi-cantly enhanced proliferative responsesand allowed IL-2 receptor expressionfollowing stimulation. In addition, theimpaired cytolytic activity of both NK andT cells was partially restored. Impairedproliferative or functional capacity of liverderived T cells was not observed in eitherPBC or AIH patients.Conclusions—We suggest that reduced Tcell reactivity in liver infiltrating cellsobtained from patients with PSC is due to
high local production of TNF-á. Our find-ings indicate that the use of anti-TNFantibodies as an alternative treatment forPSC patients should be evaluated.(Gut 2001;49:131–141)
Keywords: autoimmune liver diseases; biliary epithelialcells; cytokines
Primary sclerosing cholangitis (PSC) is charac-terised by inflammatory fibrosis of the intra-hepatic and extrahepatic biliary tracts.1 2
Patients with PSC frequently develop progres-sive hepatobiliary fibrosis with eventual devel-opment of cirrhosis, portal hypertension, liverfailure, and premature death. The aetiology ofPSC is unknown although lymphocytic infiltra-tion in areas of portal damage suggests anautoimmune mediated destructive process.3
Infiltration by T lymphocytes is a featureshared by several organ specific autoimmunediseases. Periportal mononuclear cell infiltra-tion with at least 80% T lymphocytes is foundin most liver biopsy specimens of PSCpatients.4 Some of these T cells show evidenceof activation, as indicated by expression ofHLA-DR.5 In 70% of cases PSC is associatedwith inflammatory bowel disease (IBD), mostcommonly ulcerative colitis. This close associ-ation further suggests that PSC may be anautoimmune disease.6
The pathophysiological mechanisms under-lying the development of PSC are unknown. InPSC the final target of the inflammatoryprocess appears to be the biliary epithelium.7
The cause of epithelial cell destruction is notknown. There is increasing evidence thatlocally produced cytokines may play a criticalrole, and may be directly responsible forepithelial cell injury by altering epithelial cellfunction or integrity.8 The eVects of cytokinesin other chronic inflammatory disorders suchas rheumatoid arthritis,9 Crohn’s disease,10
psoriasis,11 and glomerulonephritis12 havealready been reported. In IBD patients, pro-inflammatory cytokines such as tumournecrosis factor á (TNF-á), interleukin (IL)-1â,and interferon ã (IFN-ã) have been reported to
Abbreviations used in this paper: BECs, biliaryepithelial cells; IBD, inflammatory bowel disease;FITC, fluorescein isothiocynate; IL-1â, interleukin 1â;IFN-ã, interferon ã; LDLs, liver derived lymphocytes;PHA, phytohaemagglutinin; Con A, concanavalin A;LPS, lipopolysaccharide; PBC, primary biliarycirrhosis; PSC, primary sclerosing cholangitis; TNF-á,tumour necrosis factor á; AIH, autoimmune hepatitis;PBS, phosphate buVered saline; NK, natural killer;PBLs, peripheral blood lymphocytes; cpm, counts perminute.
Gut 2001;49:131–141 131
Division of ClinicalImmunology F-79,Karolinska Institute,Huddinge UniversityHospital, S-141 86Huddinge, SwedenX BoM RembergerS Sumitran-Holgersson
Division ofGastroenterologyK-63, KarolinskaInstitute, HuddingeUniversity Hospital,S-141 86 Huddinge,SwedenU Broome
Correspondence to:Dr S [email protected]
Accepted for publication12 December 2000
www.gutjnl.com
be increased and it is suggested that IBD isattributable to an imbalance in Th1 and Th2CD4 T cell activity with a predominant Th1cell response.13–15
Neither the detection or presence of cy-tokines in the livers of PSC patients nor the roleplayed by cytokines in the destruction of biliaryepithelium has been elucidated previously. Inaddition, so far the eVects of proinflammatorycytokines on freshly isolated cells from the sitesof inflammation in patients with autoimmuneliver diseases have not been studied. In mostreports to date, the functional capacity of liverderived lymphocytes (LDLs) have been stud-ied using cells propagated in vitro, indicatingthe possibility of strong selection bias, andtherefore such cells may represent only a frac-tion of the LDLs present in situ.
Autoimmune injury has been suggested tobe responsible for bile duct destruction inPSC16 and enhanced autoreactivity of periph-eral T lymphocytes in these patients has beenreported.17 The aim of our study was tofunctionally characterise lymphocytes freshlyisolated from the livers of PSC patients toevaluate alterations in the function of LDLsthat may reflect abnormalities that could beassociated with tissue damage. In all experi-ments, cell samples were compared with thosefrom normal controls and from patients withother autoimmune liver diseases.
Materials and methodsPATIENTS AND CONTROLS
Seven PSC patients (four females), median age42 (36–63) years, all with end stage cirrhoticdisease, were studied. Five patients also hadulcerative colitis but none had cholangiocarci-noma. The diagnosis of PSC was based onaccepted clinical, histological, and cholangi-ographic criteria.7 Four primary biliary cirrho-sis (PBC), three autoimmune hepatitis (AIH)patients, and eight healthy donor livers werealso included. Diagnoses of PBC and AIH werebased on accepted criteria.18 19 All patients withPBC were treated with ursodeoxycholic acid.Patients with AIH were treated with steroidsand azathioprine. All patients with PBC andAIH had end stage liver disease with biopsyproved cirrhosis. None had hepatitis B or Cinfections.
ISOLATION OF MONONUCLEAR CELLS FROM
PERIPHERAL BLOOD
Peripheral blood lymphocytes (PBLs) wereisolated from patients using density gradientcell separation. PBLs were obtained before anyblood products were given. Cells were sus-pended in RPMI 1640 cell culture medium(Gibco, Paisley, UK) supplemented with 10%heat inactivated fetal calf serum, 2 mML-glutamine, 25 mM HEPES, and antibiotics(complete medium), and adjusted to a concen-tration of 1×106 cells/ml.
ISOLATION OF LIVER LYMPHOCYTES
Liver tissues were obtained from organsremoved from recipients and organ donors atthe time of liver transplantation. Permissionwas granted from the local ethics committee.
Initially, the livers were washed extensively withphosphate buVered saline (PBS) to remove allblood and contaminating peripheral lym-phocytes. Tissues were then chopped withsterile scalpels and enzymatically digested withsterile medium containing 0.05% (w/v) colla-genase and 0.002% (w/v) DNase (both fromSigma, Steinhem, Germany) in a 37°C waterbath for 30 minutes. The digested cell suspen-sions were passed through a nylon mesh filter(25 µm pore size) to remove undissociated tis-sue and cell clumps. The isolated cells werewashed once with PBS to remove mucous-likematerial and centrifuged at 350 g for 10minutes. Thereafter lymphocytes were sepa-rated on Lymphoprep using density gradientcentrifugation.
Except for phenotyping and intracellularcytokine labelling, all experiments were per-formed after overnight incubation of LDLs onplastic to remove adherent cells. Hence themononuclear cell populations consisted mainlyof T lymphocytes, CD56+/16+cells, and some Bcells.
ISOLATION OF CD56+/16+ CELLS FROM LDLsCells were added to culture flasks and incu-bated overnight to remove adherent cells. Allnon-adherent cells were collected from theflasks, washed once with PBS, and centrifugedat 200 g for five minutes. Anti-CD3 antibodies(clone SK7; cat. No 347340; Becton-Dickinson, San Jose, California, USA)20 µl/106 cells were added to the pellet andincubated for 30 minutes at room temperature.Cells were washed twice with PBS, andanti-CD3 reactive lymphocytes were bound torat antimouse IgG1 coated magnetic beads (cat.No. 110.12; Dynal Inc., Oslo, Norway) 30 µlbeads/106 cells, following incubation on a rock“n” roller for 30 minutes at room temperature.Anti-CD3 reactive lymphocytes were removedby placing the mixture in a magnetic field. Thesupernatant was centrifuged to obtain cells thatwere depleted of CD3+ lymphocytes. The useof rat antimouse IgG1 coated magnetic beadswas repeated a second time to remove anyremaining CD3+ cells. The cells were nextincubated with anti-CD19 antibody coatedmagnetic beads 20 µl/106 cells (cat. No 111.04;Dynal Inc.), and incubated at room tempera-ture on a rock “n” roller for 20 minutes. Anti-CD19 reactive lymphocytes were removed byplacing the mixture in a magnetic field. Theaverage phenotype of the negatively selectednatural killer (NK) cell enriched fraction was95% CD3−, CD16+, and CD56+ cells, and lessthan 1% CD3+ cells, as indicated by flow cyto-metric analysis.
ISOLATION OF BILIARY EPITHELIAL CELLS
Human intrahepatic biliary epithelial cells(BECs) were isolated from the livers of twonormal healthy liver donors using a methodpreviously described.20 In brief, liver tissue wasmechanically disrupted and enzymatically di-gested as described above and the non-hepatocyte fraction was collected by densitygradient centrifugation using Percoll. BECs
132 Bo, Broome, Remberger, et al
www.gutjnl.com
were further purified by immunomagnetic iso-lation using Dynabeads conjugated to the epi-thelial cell specific antibody BerEp4 (Dynal).Approximately 2×107 immunomagnetic beadswere washed and incubated with the cellsextracted from the density gradient interface.The mixture was incubated at 4°C for 30–40minutes with gentle agitation. Cells coupled tothe beads were isolated and washed by applyinga magnetic particle concentrator (Dynal).Immunoisolated BECs were resuspended in a1:1 mixture of Ham’s F12 and DMEM media(both from Gibco), supplemented with 5%fetal bovine serum, 5 ng/ml epidermal growthfactor, 30 ng/ml cholera toxin, 0.4 µg/mlhydrocortisone hemisuccinate, 2 nM tri-iodo-thyronine and 5 µg/ml insulin (all from Sigma),and 10 ng/ml human recombinant hepatocytegrowth factor (R&D Systems, Abingdon, UK),and seeded in gelatin coated tissue cultureflasks. Cells were grown at 37°C in ahumidified atmosphere of 5% CO2 in air. Cellmonolayers were passaged at confluence usingtrypsin-EDTA.
LABELLING OF PBLs AND LDLs FOR T AND NK
CELL MARKERS WITH CONJUGATED ANTIBODIES
Isolated PBLs/LDLs (1×104) were incubatedwith 5 µl of fluorescein isothiocynate (FITC)or phycoerythrin conjugated monoclonal anti-bodies in 50 µl of PBS-0.1% sodium azidesolution for 30 minutes at 4°C. Cells werewashed once with PBS-sodium azide and fixedin 1% (wt/vol) paraformaldehyde solution inPBS.
The monoclonal antibodies (Becton-Dickinson, Mountain View, California, USA)used in the study included the following:anti-CD3, anti-CD4, anti-CD8, anti-CD56,anti-CD16, anti-CD25, and anti-HLA-DR.Mouse isotypes IgG1 and IgG2a were used ascontrols in all experiments. Two colour analysiswas performed on a FACSorter (Becton Dick-inson). For intracellular cytokine staining ofLDLs, FITC conjugated antihuman TNF-á(IC210F) and anti-human IL-1â (IC201F)(both antibodies from R&D systems) wereused. The procedure for intracellular stainingwas performed according to the manufactur-er’s instructions.
PHENOTYPING OF BECs
Single colour fluorescence was used to pheno-typically characterise the BECs. Primary anti-bodies for intracellular staining were directedto cytokeratin 7 and 19 (Dako, A/S, Denmark).For labelling of intracellular antigens,21 cellswere pelleted and washed once in 300 µl ofPBS containing 0.5% (w/v) saponin (Sigma).Primary antibodies at predetermined concen-trations were added to the cells and incubatedfor 30 minutes at 4°C. Cells were washed twicewith cold PBS with saponin. An appropriateFITC conjugated goat antimouse secondaryantibody (Immunotech, Marseille, France) wasadded and incubated for 20 minutes at 4°C.Cells were washed once prior to analysis on theflow cytometer. Antibodies against surfacereceptors included antihuman HEA-125 (epi-thelial specific antigen) (Serotec, Oxford, UK)
used for phenotyping of epithelial cells, FITClabelled antihuman TNF receptor I (TNF RI,clone 16803.1), TNF receptor II (TNF RII,clone 22235.311; both from R&D Systems),and anti-CD95 (Fas) (Becton-Dickinson).
PROLIFERATION ASSAYS
Triplicate cultures of 100 µl of LDLs or PBLs(1×105) suspended in complete medium weredistributed into a 96 well tissue culture plate. Asample of 100 µl of either phytohaemagglutinin(PHA 5 µg/ml), concanavalin A (Con A 5µg/ml), lipopolysaccharide (LPS 10 ng/ml), ormedium alone were added to the wells andincubated for 48 hours at 37°C. 3H thymidinewas then added, harvested after another 24hours of incubation with a semiautomated cellharvester (Titertek 550; Flow laboratories,Irvine, Ayrshire, UK), and counted in a liquidscintillation counter (LKB Instruments Inc.,Bromma, Sweden). Results are expressed asarithmetic mean counts per minute (cpm) oftriplicate cultures (SEM). The experimentswere also performed with only CD3+ T cellfractions.
CELL MEDIATED LYMPHOLYSIS ASSAY
Two non-adherent cell lines were used astargets for the cytotoxicity experiments: K562(NK cell sensitive) and Raji (NK resistant).These cells were obtained from the AmericanType Culture Collection (ATCC, Rockville,Maryland, USA) and maintained in RPMI1640 complete medium. Immediately beforeeach cytotoxicity assay, both target cell typeswere harvested and pelleted by centrifugation(400 g, 10 minutes). Target cells (1×106) in100 µl complete medium were labelled with100 µCi of 51CrO4Na2 (Amersham, Bromma,Sweden) for one hour at 37°C in a humidified5% CO2 atmosphere and washed three timesbefore being resuspended in complete mediumat a concentration of 1×105 cells/ml. Freshlyisolated eVector cells (CD56+/16+/CD3− cellsfrom livers of PSC/PBC/AIH/normals) wereresuspended at an appropriate concentrationand aliquoted (100 µl) in triplicate in 96 Ubottom microculture plates (Costar, Cam-bridge, Massachusetts, USA). The labelledtarget cells were incubated with the CD56+/16+
cells at eVector to target ratios of 50:1, 25:1,12:1, and 6:1 in a total incubation volume of200 µl. The plates were centrifuged at 200 g forfive minutes, incubated at 37°C for four hours,and subsequently aliquots of supernatant (100µl) were counted for radioactivity. Spontaneous51Cr release was determined by counting thesupernatant radioactivity present in the cul-tures of target cells in medium alone. Maximalrelease was determined by incubating thetarget cells with 5% Triton X-100. Cytotoxicitywas expressed as percentage specific 51Crrelease using the formula:
% specific release = ((experimentalcpm−spontaneous cpm)/(maximum cpm–spontaneous cpm))×100.
For detection of the functional capacity ofliver derived cytotoxic T cells, LDLs from each
EVect of TNF on liver T cells in PSC 133
www.gutjnl.com
PSC/PBC/AIH patient were mixed with alloge-neic irradiated (2000 rads) LDLs (fromnormals) at a cell concentration of 1.5×106/mland a responder:stimulator ratio of 1:1 (MLC),and incubated in flasks for five days at 37°C.Three days prior to the cell mediated lym-pholysis assay, PHA (10 µg/ml) was added tofresh allogeneic LDL target cells and incubatedat 37°C. On day 6, PHA stimulated LDLs werewashed once in PBS and labelled with 100 µCiof 51Cr for one hour at 37°C, washed threetimes, and resuspended at 1×105 cells/ml incomplete medium. EVector cells from the fiveday MLC cultures were washed, counted, anddistributed in U bottomed microtitre plates at
eVector to target ratios of 50:1, 25:1, 12:1, and6:1. After incubation for four hours at 37°C,the amounts of radioactivity released into thesupernatants were measured and the percent-age specific cytotoxicity was calculated asabove. Control experiments were performedusing LDLs from the four normal livers stimu-lated against each other.
CYTOKINE ASSAYS
PSC, PBC, AIH, and normal LDL culturesupernatants from unstimulated, mitogenstimulated, and cultures neutralised with anti-TNF-á or anti-IL-1â were collected after 48hours, sterile filtered, and kept frozen at −70°Cuntil assayed. The cytokines TNF-á, IFN-ã,IL-1â, IL-2, IL-10, and IL-12 were measuredby standard sandwich ELISA techniques usingthe Quantikine sandwich enzyme immu-noassay from R&D systems. Assays wereperformed according to the manufacturer’sinstructions. In brief, samples and standardswere incubated in anticytokine antibody coatedwells. After washing, a polyclonal antibodyconjugated to horseradish peroxidase wasadded. After another incubation and washing,a chromogen-peroxide substrate solution wasadded. Colour development was stopped withsulphuric acid and the intensity of the colourwas measured.
NEUTRALISING ANTIBODY STUDIES
Anti-hTNF-á (MAB210) and anti-hIL-1â(MAB201) neutralising antibodies were pur-chased from R&D systems. The anti-hTNF-ámonoclonal antibodies can neutralise thebiological activity of both soluble and mem-brane bound TNF-á. LDLs were incubatedwith neutralising antibodies (10 µg/ml) for twohours and the proliferation assay carried out asdescribed previously. An additional controlincluded cells with the neutralising antibodiesalone. For the cytolytic assays, LDLs wereincubated with the neutralising antibodiesovernight.
STATISTICAL ANALYSIS
When comparing diVerences in proliferativeresponse and cytotoxic functions betweendiVerent groups, the Kruskall Wallis test wasused. The Mann-Whitney U test was usedwhen diVerences between PBL and LDLswithin the same patient were compared.Calculations were made with Statistica/Wc
(Statsoft, Inc, Tulsa, Oklahoma, USA).
ResultsSIGNIFICANTLY DECREASED EXPRESSION OF IL-2
RECEPTOR ON ACTIVATED T CELLS DERIVED FROM
PSC LIVERS
We phenotypically characterised T and NKcells from PBLs and LDLs, including expres-sion of activation markers, for example IL-2receptor (CD25) and HLA-DR (table 1).Among the cells obtained from control livertissues, T lymphocytes and NK cells (CD56+/CD16+/CD3−) represented the two majorpopulations. Lower proportions of CD3+ andNK cells were found in PSC livers compared
Table 1 Phenotypic analysis of peripheral blood and liver derived lymphocytes fromnormal healthy liver donors, and from primary sclerosing cholangitis (PSC), primarybiliary cirrhosis (PBC), and autoimmune hepatitis (AIH) patients
Normals (n=8) PSC (n=7) PBC (n=4) AIH (n=3)
Peripheral blood lymphocytesCD3+ 63 (6) 63 (8) 65 (7) 63 (6)CD4+/CD3+ 33.8 (3) 38.14 (2) 38.8 (3) 33.8 (3)CD8+/CD3+ 20.5 (1) 22.93 (5) 30.5(1) 20.5(1)CD45RO+/CD3+ 19.8 (1) 44.38 (1) 49.8 (1) 19.8 (1)CD45RA+/CD3+ 30.2 (2) 35.89 (1) 40.2 (2) 30.2 (2)CD25+/CD3+ 1.3 (2) 1.2 (2) 2.3 (2) 1.3 (2)HLA-DR+/CD3+ 2.32 (1) 2.36 (2) 2.12 (1) 2.32 (1)CD56+16+/CD3- 9.8 (2) 15.57 (2) 10.8 (2) 9.8 (2)
Liver derived lymphocytesCD3+ 64 (7) 37.26 (5) 63 (6) 65 (6)CD4+/CD3+ 16.01 (3) 11.89 (3) 37.8 (3) 13.5 (3)CD8+/CD3+ 47.83 (7) 25.57 (1) 25.5(1) 52.5(1)CD45RO+/CD3+ 30.09 (2) 23.46 (1) 49.8 (1) 50.8 (1)CD45RA+/CD3+ 28.01 (1) 18.19 (2) 12.2 (2) 16.2 (2)CD25+/CD3+ 3.36 (1) 0.78 (1) 4.3 (2) 10.3 (2)HLA-DR+/CD3+ 6.28 (5) 22.25 (3) 20.32 (1) 22.32 (1)CD56+16+/CD3- 47.29 (5) 36.36 (4) 24.8 (2) 19.8 (2)
Values are mean (SEM).
Figure 1 Biliary epithelial cells (BECs, second passage) isolated from two healthy andthree primary sclerosing cholangitis (PSC) livers were tested for expression of CD95 andthe two tumour necrosis factor (TNF) receptors to study the role of BECs as targets forlymphocyte mediated or direct cytokine mediated damage. TNF-á and IL-1 stimulatedPSC BECs demonstrated increased expression of the TNF RI and RII receptors but lowexpression of CD95. Normal BECs expressed CD95 and TNF RII but had lowerexpression of TNF RI.
200
160
40
120
0
80Co
un
ts
100 101 102
Fluorescence103 104
200
160
40
120
0
80Co
un
ts
100 101 102
Fluorescence
Control BECsstimulated
PSC BECsstimulated
103 104
200
160
40
120
0
80Co
un
ts
100 101 102 103 104
200
160
40
120
0
80Co
un
ts
100 101 102 103 104
200
160
40
120
0
80Co
un
ts
100 101 102 103 104
200
160
40
120
0
80Co
un
ts
100 101 102 103 104
Anti-CD95
Anti-TNF RI
Anti-TNF RII
134 Bo, Broome, Remberger, et al
www.gutjnl.com
with PBC, AIH, and normal livers. A signifi-cantly higher fraction of liver derived CD3+ cellsfrom all three disease groups expressed HLADR compared with normals. However, the mostinteresting diVerence was very low expression ofCD25 on activated T cells from PSC patientscompared with PBC, AIH, and normals(p<0.001). In contrast, as shown in table 1,comparing PBLs and LDLs in PSC patients, wefound that the proportions of CD3+ and CD25+
T cells in PBLs were normal.
TNF-á AND IL-1â EXPOSED BILIARY EPITHELIAL
CELLS EXPRESS HIGHER LEVELS OF TNF-RI AND RII
BECs were routinely characterised duringgrowth by immunofluorescence labelling andflow cytometry. Results indicated that the cul-tures were more than 95% pure for BECs onthe basis of expression of cytokeratins 19 and 7.Other contaminating cells (endothelial cellsand fibroblasts) and smooth muscle cells wereless than 2%.
BECs isolated from the livers of three PSCpatients and two normal controls (second pas-sage) were tested for expression of CD95 and
the two TNF receptors RI and RII. A markedmorphological diVerence was observed in nor-mal BECs compared with PSC BECs. NormalBECs showed typical epitheloid morphologywhile PSC BECs were more elongated andvacuolated (manuscript in preparation).Freshly isolated BECs from normals or PSClivers did not express either of the TNF recep-tors. Both types of BECs expressed cytokeratin7. However, lower expression of cytokeratin 19(data not shown) and CD95 was observed onPSC BECs compared with control BECs (fig1). Cytokine stimulation with TNF-á andIL-1â showed that PSC BECs expressed higherlevels of the two TNF receptors (RI and RII)with no change in CD95 compared withcontrols (fig 1).
DIMINISHED PROLIFERATIVE RESPONSES OF
MITOGEN STIMULATED T CELLS ISOLATED FROM
PSC PATIENTS
As seen in fig 2, a low proliferative responseinduced by PHA, Con A, and LPS in LDLsisolated from PSC patients was seen. This wassignificantly diVerent from LDLs obtained
Figure 2 In general, liver derived lymphocytes (LDLs) isolated from controls (n=8) on stimulation with mitogens such asconcanavalin A (Con A) and phytohaemagglutinin (PHA) for 72 hours showed lower proliferation responses than thoseobserved in peripheral blood lymphocytes from the same on stimulation. No diVerence in response was observed afterpretreatment with anti-tumour necrosis factor (TNF) monoclonal antibodies. Significantly diminished proliferation ofLDLs from the seven primary sclerosing cholangitis (PSC) patients was observed on stimulation which was restored toalmost normal by pretreatment with anti-TNF monoclonal antibodies (p<0.001). No significant diVerences in proliferativeresponses in mitogen stimulated LDLs from primary biliary cirrhosis (PBC) or autoimmune hepatitis (AIH) patients(either before or after anti-TNF monoclonal antibody pretreatment) was observed compared with controls (NS). Datarepresent mean counts per minute (cpm) of triplicate cultures. LPS, lipopolysaccharide.
Co
n A
PH
A
LPS
Co
ntr
ol
100 000Control LDLs
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
PSC LDLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
AIH LDLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
PBC LDLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
No anti-TNF-α antibodyAnti-TNF-α antibody
EVect of TNF on liver T cells in PSC 135
www.gutjnl.com
from PBC, AIH, and normal livers (p<0.001)(fig 2). Similar results were obtained when purepopulations of CD3+ T cells were used (datanot shown).
Comparison between mitogen stimulatedproliferative responses of PBLs and LDLs fromPSC patients showed a significant diVerence.They were significantly higher in PBLs than inLDLs (fig 3) (p<0.001) but lower comparedwith PBLs from normal controls. In cells fromPBC and AIH patients, no significant diVer-ences were observed in proliferative responsesof either PBLs or LDLs compared withnormals (NS).
As no proliferative responses were observedin LDLs from PSC patients, we performed thesame assays using shorter periods of time (4, 8,12, 16, 24, 36, and 48 hours). Diminished pro-liferative responses were observed at all timepoints (data not shown).
IMPAIRED CYTOLYTIC ACTIVITY OF T
LYMPHOCYTES AND NK CELLS ISOLATED FROM
THE LIVERS OF PSC PATIENTS
We found that the cytotoxic functions of T andNK cells obtained from the livers of PSCpatients were significantly impaired comparedwith PBC, AIH, and normal controls
(p<0.001) (fig 4). PSC liver CD56+/16+ cellsdid not kill K562 or Raji. However, NK cellsfrom PBC, AIH, and normal livers spontane-ously killed K562 but not the Raji cell line, andcytotoxic T cells clearly showed high cytotoxicfunction against allogeneic target cells (fig 4).
As LDLs derived from the livers of PSCpatients had impaired proliferative and func-tional capacity, we decided to screen superna-tants of mitogen stimulated cultures for thepresence of cytokines, as certain proinflamma-tory cytokines may inhibit lymphoproliferativeresponses in vitro.
HIGH LEVELS OF TNF-á AND IL-1â IN THE
SUPERNATANTS OF PHA AND CON A STIMULATED
LDLs FROM PSC PATIENTS
Supernatants were collected after 48 hoursfrom cell cultures stimulated with either PHA,Con A, or LPS and analysed for the presence ofTNF-á, IFN-ã, IL-1â, IL-2, IL-10, and IL-12.As shown in table 2, significantly higher levelsof TNF-á and IL-1â, and lower concentrationsof IL-2, IL-10, and IFN-ã were detected in thesupernatants of LDL cell cultures from PSCpatients stimulated with PHA and Con A com-pared with PBC, AIH, and normal controls(p<0.001). Supernatants collected after 4, 8,
Figure 3 Peripheral blood lymphocytes (PBLs) isolated from controls (n=8), primary sclerosing cholangitis (PSC)(n=7), primary biliary cirrhosis (PBC) (n=4), and autoimmune hepatitis (AIH) (n=3) patients showed normalproliferative responses on stimulation with mitogens such as concanavalin A (Con A) and phytohaemagglutinin (PHA) for72 hours. No diVerence in responses was observed after pretreatment with anti-tumour necrosis factor (TNF) monoclonalantibodies. Data represent mean counts per minute (cpm) of triplicate cultures. LPS, lipopolysaccharide.
Co
n A
PH
A
LPS
Co
ntr
ol
100 000Control PBLs
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
100 000
75 000
50 000
25 000
0
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
PSC PBLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
AIH PBLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
Co
n A
PH
A
LPS
Co
ntr
ol
PBC PBLs
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
)
No anti-TNF-α antibodyAnti-TNF-α antibody
136 Bo, Broome, Remberger, et al
www.gutjnl.com
12, 16, 24, and 36 hours gave the same results(data not shown).
Flow cytometric analysis of three daymitogen stimulated LDLs showed that T lym-phocytes and CD56+/16+ cells were the majorpopulations present in cultures.
INCREASED INTRACELLULAR EXPRESSION OF
TNF-á AND IL-1â IN LDLs FROM PSC LIVERS
LDLs from all PSC, PBC, AIH, and controllivers were stained for intracellular expressionof TNF-á and IL-1â. We found varying butincreased percentages of PSC LDLs possessingTNF-á and IL-1â. LDLs from one PBCpatient had a high proportion of IL-1â but notof TNF-á secreting cells. Only one control livershowed binding slightly higher than back-ground (table 3). A representative example ofthe increased intracellular binding of anti-TNF-á and anti-IL-1â in LDLs from one PSCpatient compared with one control is shown infig 5.
TREATMENT WITH ANTI-TNF MONOCLONAL
ANTIBODY RESTORES THE PROLIFERATIVE
RESPONSES OF PSC LDLs
The final concentration of the anti-TNFmonoclonal antibody was determined in initialtitration experiments of time and concentra-tions to detect the optimal time and amountrequired for neutralising the TNF-á eVect(data not shown). Treatment with 10 µg/mlovernight was found to be optimal for our pur-poses. An isotype matched irrelevant mono-clonal antibody (Sigma) was used as a control.No non-specific stimulation using the controlantibodies was observed (see fig 6). Theanti-TNF-á antibody at a final concentrationof 10 µg/ml neutralised the eVect of 100 ng/mlTNF-á. Anti-TNF treated cell culture super-natants showed normal levels of IL-2, IL-10,and IFN-ã but lower TNF-á and IL-1â levels(table 2).
Anti-TNF treated LDLs from PSC patientsdemonstrated significantly enhanced prolifera-tion (p<0.001) comparable with controls (fig2). However, no diVerence in the proliferativeresponses of PBC, AIH, and normal LDLsbefore or after anti-TNF treatment wasobserved (figs 2 and 3).
As high levels of IL-1â were also detected insupernatants from mitogen stimulated LDLcultures, we decided to study the eVect of anti-IL-1â monoclonal antibody on the proliferativeresponses of these cells. As in the case of anti-TNF antibody, titration studies indicated that aconcentration of 10 µg/ml with overnight treat-ment was optimal for our purposes (the PBCpatient with the high levels of IL-1â was usedas a positive control). However, no eVect ofanti-IL-1â antibody was observed on levels ofother cytokines produced and detected in thesupernatants or on the proliferative responsesof LDLs from PSC, PBC, or AIH.
Although a combination of anti-TNF andanti-IL-1â enhanced the proliferative re-sponses of LDLs to some extent, the best eVectwas obtained using anti-TNF alone. Thusanti-IL-1â alone or in combination withanti-TNF was not as eVective in restoringimpaired proliferative responses as anti-TNFalone (fig 6).
TREATMENT WITH ANTI-TNF MONOCLONAL
ANTIBODY RESTORES PARTIALLY THE CYTOLYTIC
CAPACITY OF PSC T AND NK CELLS
Anti-TNF treated cytotoxic T and CD56+16+
cells from PSC patients demonstrated amoderate increase in cytotoxic activity to allo-geneic target cells and K562, respectively,compared with activity prior to anti-TNF anti-body treatment (fig 4).
DiscussionIn the present study we functionally character-ised freshly isolated cells from blood and thesite of inflammation (that is, the liver) in PSCpatients. Our results show that cell proliferationand the function of liver derived T lymphocytesis impaired in PSC patients compared withnormal controls and other patients withautoimmune liver diseases. We believe this is aconsequence of exposure (maybe even chronic)
Figure 4 Cytotoxic liver derived lymphocytes (LDLs) (T cells) generated by stimulationagainst normal allogeneic LDLs from control, primary biliary cirrhosis (PBC), andautoimmune hepatitis (AIH) livers showed high cytotoxic function against allogeneic targetcells. In the case of controls, only the average per cent lysis (from eight experiments innormals, four in PBC, and three in AIH) is shown. Cytotoxic LDLs (T cells ) isolated fromprimary sclerosing cholangitis (PSC) livers (n=7) on the other hand did not show anycytotoxic activity against allogeneic target cells. The cytotoxic activity of T cells was thuscompletely abolished in the livers of PSC patients (p<0.001). Natural killer (NK) cells(CD56+16+) isolated from control, PBC, and AIH livers showed normal cytotoxic activityagainst K562 (NK cell sensitive cell line) but not against Raji’s (NK resistant cell line)(only mean per cent lysis is shown). NK cells isolated from PSC livers (n=7) on the otherhand did not show any cytotoxic activity against K562. NK cell cytotoxic activity was thuscompletely abolished in the livers of PSC patients (p<0.001). However, overnight treatmentof cytotoxic LDLs (T cells) isolated from PSC livers with tumour necrosis factor (TNF)antibodies (abs) (10 mg/ml) showed a moderate increase in cytotoxic activity againstallogeneic target cells. Similarly, NK cells (CD56+/16+) showed enhanced cytotoxic activityagainst K562 (NK cell sensitive cell line).
ControlsPBCAIH
50:1 25:1 12:1 6:1
90
80
70
60
50
40
30
20
% L
ysis
Control/PBC/AIH T cells
50:1 25:1 12:1 6:1
100
0
% L
ysis
PSC T cells
80
40
60
20
50:1 25:1 12:1 6:1
100
0
% L
ysis
PSC NK cells
Ratio Ratio
80
40
60
20
50:1 25:1 12:1 6:1
100
0
% L
ysis
Control/PBC/AIH NK cells
Allogeneic target cells
Allogeneic target cells
With TNF abs
No TNF abs No TNF abs
With TNF abs
K562 or Raji as target cells
K562 as target cells
Raji as target cells
80
40
60
20
EVect of TNF on liver T cells in PSC 137
www.gutjnl.com
to high levels of TNF in vivo, as (a) freshly iso-lated LDLs from PSC patients showed in-creased expression of intracellular TNF-á, (b)in vitro, supernatants from mitogen stimulated,freshly isolated LDLs showed the presence ofsignificantly high levels of TNF-á compared
with controls, and (c) addition of anti-TNFantibodies restored the diminished proliferativeresponses in vitro. Alternatively, one mayspeculate that T cells had undergone TNF-áinduced apoptosis and that addition of anti-TNF antibodies inhibited apoptosis. However,our investigations showed that even though asmall number of cells underwent apoptosis(approximately 15% using annexin V), the restof the cells survived (data not shown) at theend of the assay but did not proliferate.
Our results are supported by the findings ofSpengler and colleagues22 who reported thepresence of relatively high levels of TNF-á inculture supernatants of PHA stimulated T celllines from liver biopsies in patients with diVer-ent stages of PSC while decreased levels wereobserved in PBC patients. These resultsindicate that increased levels of TNF-á arealready present in PSC patients even in theearly stages of the disease and are not restrictedto only the late stages. In addition, Berg andcolleagues,23 based on the cytokine profileobtained using supernatants from blood lym-phocytes, have classified PSC as TH1 (T helper1 subclass of T cells) mediated as the TH2/TH1
ratio was found to be 1:4 while in PBC the ratiowas 1:1. TH1 cells produce proinflammatorycytokines, including IFN-ã, IL-2, and poten-tially TNF. Therefore, based on reports fromother studies, including the current one, it maybe hypothesised that LDLs in PSC are chroni-cally exposed to TNF-á in vivo and thereforethis accounts for the impaired in vitro prolifera-tive and functional capacities. Thus even
Table 2 Cytokine production in supernatants of mitogen stimulated or anti-tumour necrosis factor (TNF) monoclonalantibody treated lymphocytes from primary sclerosing cholangitis (PSC) and control livers
IL-2 (pg/ml) IL-10 (pg/ml) IL-12 (pg/ml) IFN-ã (pg/ml) IL-1â (pg/ml) TNF-á (pg/ml)
NormalsNo anti-TNF ab
PHA 5800 (178) 120 (10) <50 10500 (994) 63 (2) <44Con A 5000 (140) 53 (2) <50 10000 (1005) 32 (2) <44LPS <70 100 (8) <50 89 (5) <10 <44
With anti-TNF abPHA 5445 (156) 128 (15) <50 10500 (1009) 139 (5) <44Con A 5000 (163) 50 (2) <50 8000 (886) <10 <44LPS <70 98 (10) <50 <27 <10 <44
PSCNo anti-TNF-ab
PHA <70 <50 <50 2044 (695) >3000 >10000Con A <70 <50 <50 1846 (476) >3000 >10000LPS <70 <50 <50 <27 2796 (256) 6650 (5550)
With anti-TNF abPHA 5880 (525) 148 (20) <50 6835 (578) <222 545 (35)Con A 6480 (256) 64 (5) <50 6552 (590) <222 144 (25)LPS <70 <50 <50 63 (12) <10 <44
PBCNo anti-TNF ab
PHA 3100 (184) 420 (30) <50 6500 (994) 2003 (225) 544 (25)Con A 4000 (165) 553 (25) <50 7000 (1005) 3200 (206) 323 (47)LPS <70 100 (8) <50 89 (5) <10 <44
With anti-TNF abPHA 3145 (186) 428 (25) <50 5500 (900) 1139 (210) 104 (10)Con A 4030 (153) 550 (20) <50 6800 (886) 1110 (256) 124 (17)LPS <70 98 (10) <50 <27 <10 <44
AIHNo anti-TNF ab
PHA 5108 (180) 110 (10) <50 9500 (794) 68 (2) 2839 (60)Con A 5200 (120) 43 (2) <50 10000 (8565) 42 (2) 1738 (20)LPS <70 106 (8) <50 89 (5) <10 <44
With anti-TNF abPHA 5430 (128) 138 (15) <50 10400 (989) 140 (5) 534 (2)Con A 5120 (174) 50 (2) <50 9000 (846) <10 487 (10)LPS <70 100 (10 <50 <27 <10 <44
IL, interleukin; IFN-ã, interferon ã; TNF-á, tumour necrosis factor á; PHA, phytohaemagglutinin; Con A, concanavalin A; LPS,lipopolysaccharide; PSC, primary sclerosing cholangitis; PBC, primary biliary cirrhosis; AIH, autoimmune hepatitis; ab, antibody.Values are mean (SEM).
Table 3 Percentage of liver derived lymphocytes (LDLs)positive for intracytoplasmic staining of anti-tumournecrosis factor á (TNF-á) and interleukin 1 â (IL-1â)cytokines from primary sclerosing cholangitis (PSC) andcontrol patients
% of LDLs stainedwith anti-TNF-á ab
% of LDLs stainedwith anti-IL-1â ab
Normals (n=8)1 10 132 12 153 13 134 26 185 15 166 13 187 17 128 14 15
PSC (n=7)1 55 502 50 383 58 294 40 335 56 456 60 397. 45 24
PBC (n=4)1 10 202 13 183 15 434 12 23
AIH (n=3)1 25 152 20 183 28 22
PBC, primary biliary cirrhosis; AIH, autoimmune hepatitis; ab,antibody.
138 Bo, Broome, Remberger, et al
www.gutjnl.com
though the current observations were made inpatients with end stage liver disease, they mayreflect immunological changes already occur-ring during the initial stages of the disease. Ourfindings are specific to PSC patients asimpaired proliferation was not observed ineither PBC or AIH patients with end stage liverdisease and is therefore not only explained bythe end stage nature of the disease.
In addition, we found significantly higherlevels of TNF-á and IL-1â but significantlylower levels of IL-2, IL-10, and IFN-ã in thesupernatants of mitogen stimulated LDLsobtained from PSC livers compared with con-trols. This eVect was reversible as exposure ofcells (overnight) to the neutralising eVect ofanti-TNF-á antibodies restored the prolifera-tive responses. Addition of neutralising anti-IL-1â antibodies alone did not reverse im-paired proliferative responses to mitogenstimulation. It should be noted that theanti-TNF-á monoclonal antibodies used neu-tralise both soluble and membrane boundTNF-á while the anti-IL-1â antibody neutral-ises only soluble IL-1â and thus is probably lesseVective in restoring impaired responses.
The impaired cytolytic activity of both NKand T cells was partially restored aftertreatment with TNF antibodies. It may be pos-sible that repeated treatment with higher dosesor longer incubation time with TNF antibodiesmay be required to see complete reversal of thecytotoxic capacity of these cells. Due to lack ofsuYcient cells, such studies could not beperformed.
Currently, we do not know what stimulatesLDLs in PSC to produce high levels of TNF-á.Elevated levels of TNF-á in PSC may have agenetic background. Genetic variation in theTNF-á gene may influence inflammatory andimmune mediated diseases. The −308 poly-morphism in the TNF-á promoter region (theTNF2 allele) is associated with a six to seven-fold increase in baseline and inducible levels ofTNF-á production.24 Recently, Bernal and col-leagues25 studied the association of TNF poly-morphism with PSC and concluded thatgenetic susceptibility to PSC may be deter-mined by polymorphism within the TNF genesparticularly with the TNF2 allele. Our resultssupport this finding. However, it is importantto mention that as studies were performed withmaterial obtained from patients with end stagecirrhotic disease, our findings may not reflectthe earlier state of the disease.
The elegant in vitro and in vivo (animal)studies of T lymphocytes chronically exposedto TNF performed by Cope and colleagues26 27
showed that chronic exposure to TNF sup-presses a broad range of T cell responses andmay target T cell receptor signalling28 throughpartial rather than complete uncoupling ofearly signalling events within T cells. The pro-posed immunosuppressive eVects of prolongedexposure to TNF are, among others, antiprolif-erative eVects, which influence the precursorfrequency of peripheral autoreactive lym-phocytes, defective eVector T and/or B cellresponses, and attenuation of antigen-receptorsignalling.29 Our findings of decreased prolif-erative responses, low levels of IL-2 receptorexpression, as well as decreased levels of IL-2,IL-10, and IFN-ã in the supernatants ofmitogen stimulated LDLs obtained from thesite of inflammation in PSC patients supportthe findings of Cope et al.
The results of this study also emphasise theimportance of using the right source of cells forphenotypic and functional studies as PBLs and
Figure 5 Increased immunofluorescent staining of intracellular cytokines tumour necrosisfactor á (TNF-á) and interleukin 1â (IL-1â) was seen in freshly isolated liver derivedlymphocytes from all seven primary sclerosing cholangitis (PSC) patients but not fromcontrols. A representative picture of this finding from one PSC patient and one normalcontrol is shown. Cells were initially fixed, permeabilised, and stained with fluoresceinisothiocynate conjugated anti-TNF/anti-IL-1â antibodies. Unlabelled anti-cytokinemonoclonal antibody was used to decrease high background staining.
100
100
Co
un
ts
Anti-TNF-α
101
102
103
104
101 102 103 104100
100
Co
un
ts
Anti-IL-1β
101
102
103
104
101 102 103 104
100
100
Co
un
ts
101
102
103
104
101 102 103 104100
100
Co
un
ts101
102
103
104
101 102
Fluorescence Fluorescence
Control PSC
103
10% 55%
13% 50%
104
Figure 6 EVect of anti-cytokine antibodies on restorationof diminished proliferative responses to mitogen stimulationby concanavalin A (Con A) and phytohaemagglutinin(PHA). Liver derived lymphocytes (LDLs) from primarysclerosing cholangitis (PSC) livers remained untreated orpretreated with anti-tumour necrosis factor (TNF)monoclonal antibodies (10 µg/ml) and/or interleukin 1(IL-1) monoclonal antibodies (10 µg/ml), or controlmonoclonal antibodies (10 µg/ml) overnight. Con A orPHA was added and 3H thymidine incorporation wasdetected after 72 hours. Anti-TNF monoclonal antibodiesalone restored the diminished proliferative responses in theLDLs from PSC patients.
No
an
tib
od
y
An
ti-T
NF
An
ti-I
L-1
An
ti-T
NF
+ an
ti-I
L-1
Iso
typ
e co
ntr
ol
50000
40000
30000
20000
10000
0
3 H t
hym
idin
e in
corp
ora
tio
n (
cpm
) Con APHA
EVect of TNF on liver T cells in PSC 139
www.gutjnl.com
LDLs from PSC patients showed significantlyvarying results. The most striking finding of thephenotypic analysis of LDLs was the very lowexpression of the IL-2 receptor on activated Tlymphocytes from PSC livers compared withPBC, AIH, and normal controls, even thoughLDLs from all three disease groups had signifi-cantly high levels of activated T cells (CD3+/HLA-DR+).
Damage of intrahepatic bile ducts byactivated T cells is assumed to be a key eventin the pathogenesis of PSC.17 Based on ourfindings we put forward two hypotheses.Firstly, assessment of the normal functionalcapacity of freshly isolated cytotoxic PSCLDLs demonstrated decreased or no cytolyticactivity. The possibility that T cells mediatebiliary epithelial cell injury in PSC by a directcytotoxic eVect seems less likely in view of ourinability to detect any normal cytotoxic activ-ity among LDLs from these patients. NK cellsmay not be involved in the destruction as theirnormal killing capacity (based on resultsobtained with K562 cells) was also abolished.These eVector cells may play an important rolein the early stages of the disease but their rolemay diminish with time because of chronicexposure to TNF-á and possibly IL-1â. It isconceivable that TNF-á alone or in combina-tion with IL-1â may directly mediate BECinjury. In addition, expression of receptorTNF-R1 on BECs further supports the possi-bility of induction of apoptosis in BECs byhigh levels of TNF-á in PSC patients (fig 1).Induction of apoptosis in BECs by TNF-ácontaining supernatants from PSC LDLcultures is currently under investigation(manuscript in preparation). Even if TNF-ádoes not directly mediate BEC injury, it mayadversely aVect BEC in many ways.30 Anotherimportant clinical implication is the absence ofIL-2 and its role in PSC. IL-2 deficient micespontaneously develop a chronic, non-granulomatous inflammation of the submu-cosa and mucosa of the colon, and caecum,which is reported to resemble ulcerative colitisin humans.31 Ciacci et al have shown thatintestinal epithelial cells have IL-2 receptorsand proliferate in the presence of IL-2.32 Thusthe absence of IL-2 in concert with chronicTNF and IL-1â production may further causeBEC injury. The second hypothesis is that asalready activated cells respond poorly tomitogen stimulation, allostimulation, and pro-duce less of certain cytokines, it is likely thatwe are detecting the presence of an antigenspecific T cell clone that is expanded in PSCpatients. Broome et al have reported the pref-erential usage of the Vâ3 region of the T cellreceptor in liver T lymphocytes in PSCpatients.3 This study needs to be confirmedand further analysis using all of the currentlyavailable monoclonal antibodies against the Vregions of the T cell receptor needs to be per-formed.
Our findings may have important clinicalimplications in that it may be possible to treatPSC patients with anti-TNF-á antibodies.Significant clinical benefits have been ob-served in patients with rheumatoid arthritis,
multiple sclerosis, and Crohn’s disease usinganti-TNF therapy. However, whether longterm chronic TNF blockade aggravates theautoimmune disease process has still to beelucidated. As yet there are no convincingclinical data suggesting that long term TNFblockade might exacerbate autoimmunity.Nevertheless, based on our results we suggestthat treatment of PSC patients using anti-TNF antibodies needs to be evaluated in thefuture.
The study was financed by grants from the Magnus Bergvallfoundation and the Ruth and Richard Julins foundation, ClasGrotchinsky foundation, the Swedish Medical Society, theKarolinska Institute, and the Nanna Svartz Foundation. Wethank Professor Erna Möller and Dr Jan Holgersson forconstructive criticism of the manuscript.
1 Dickson ER, LaRusso NF, Wiesner RH. Primary sclerosingcholangitis. Hepatology 1984;4(suppl 1):33–5S.
2 LaRusso NF, Wiesner RH, Ludwig J, et al. Currentconcepts. Primary sclerosing cholangitis. N Engl J Med1984;310:899–903.
3 Broome U, Grunewald J, Scheynius A, et al. PreferentialVâ3 usage by liver T lymphocytes in patients with primarysclerosing cholangitis. J Hepatol 1997;26:527–34.
4 Ludwig J, Barham SS, LaRusso NF, et al. Morphologic fea-tures of chronic hepatitis associated with primary scleros-ing cholangitis and chronic ulcerative colitis. Hepatology1981;1:632–40.
5 Whiteside TL, Lasky S, Si L, et al. Immunologic analysis ofmononucelar cells in liver tissue and blood of patients withprimary sclerosing cholangitis. Hepatology 1985; 5:468–74.
6 Broome U, Olsson R, Loof L, et al. Natural history andprognostic factors in 305 Swedish patients with primarysclerosing cholangitis. Gut 1996;38:610–15.
7 Chapman RW, Arborgh BA, Rhodes JM, et al. Primary scle-rosing cholangitis: a review of its clinical features, cholangi-ography, and hepatic histology. Gut 1980;21:870–7.
8 Panja A, Goldberg S, Eckmann L, et al. The regulation andfunctional consequence of proinflammatory cytokine bind-ing on human intestinal epithelial cells. J Immunol1998;161:3675–84.
9 Shinmei M, Masuda K, Kikuchi T, et al. Interleukin 1,tumor necrosis factor, and interleukin 6 as mediators ofcartilage destruction. Semin Arthritis Rheum 1989;18:27–32.
10 Fais S, Capobianchi MR, Pallone F, et al. Spontaneousrelease of interferon gamma by intestinal lamina proprialymphocytes in Crohn’s disease. Kinetics of in vitroresponse to interferon gamma inducers. Gut 1991;32:403–7.
11 Grossman RM, Krueger J, Yourish D, et al. Interleukin 6 isexpressed in high levels in psoriatic skin and stimulatesproliferation of cultured human keratinocytes. Proc NatlAcad Sci USA 1989;86:6367–71.
12 Takemura T, Yoshioka K, Murakami K, et al. Cellularlocalization of inflammatory cytokines in human glomeru-lonephritis. Virchows Archiv 1994;424:459–64.
13 MacDonald TT, Hutchings P, Choy MY, et al. Tumournecrosis factor-alpha and interferon-gamma productionmeasured at the single cell level in normal and inflamedhuman intestine. Clin Exp Immunol 1990;81:301–5.
14 Niessner M, Volk BA. Altered Th1/Th2 cytokine profiles inthe intestinal mucosa of patients with inflammatory boweldisease as assessed by quantitative reversed transcribedpolymerase chain reaction (RT-PCR). Clin Exp Immunol1995;101:428–35.
15 Powrie F. T cells in inflammatory bowel disease: protectiveand pathogenic roles. Immunity 1995;3:171–4.
16 Boberg KM, Lundin KE, Schrumpf E. Etiology and patho-genesis in primary sclerosing cholangitis. Scand J Gastroen-terol Suppl 1994;204:47–58.
17 Lindor KD, Wiesner RH, LaRusso NF, et al. Enhancedautoreactivity of T-lymphocytes in primary sclerosingcholangitis. Hepatology 1987;7:884–8.
18 Neuberger J. Primary biliary cirrhosis. Lancet 1997;350:875–9.
19 Alvarez F, Berg PA, Bianchi FB, et al. InternationalAutoiimune Hepatitis Group Report: review of criteria fordiagnosis of autoimmune hepatitis. J Hepatol 1999;31:929–38.
20 Joplin R, Strain AJ, Neuberger JM. Immuno-isolation andculture of biliary epithelial cells from normal human liver.In Vitro Cell Dev Bio 1989;25:1189–92.
21 Leon MP, Bassendine MF, Gibbs P, et al. Hepatic allograftrejection: regulation of the immunogenicity of humanintrahepatic biliary epithelial cells. Liver Transpl Surg 1996;2:37–45.
22 Spengler U, Möller A, Jung MC, et al. T lymphocytes frompatients with primary biliary cirrhosis produce reducedamounts of lymphotoxin, tumor necrosis factor andinterferon-ã upon mitogen stimulation. J Hepatol 1992;15:129–35.
23 Berg PA, Klein R, Rocken M. Cytokines in primary biliarycirrhosis. Semin Liver Dis 1997;17:115–23.
140 Bo, Broome, Remberger, et al
www.gutjnl.com
24 Wilson AG, Symons JA, McDowell TL, et al. EVects of apolymorphism in the human tumor necrosis factor ápromoter on transcriptional activation. Proc Natl Acad SciUSA 1997;94:3195–9.
25 Bernal W, Moloney M, Underhill J, et al. Association oftumor necrosis factor polymorphism with primary scleros-ing cholangitis. J Hepatol 1999;30:237–41.
26 Cope AP, Londei M, Chu NR, et al. Chronic exposure totumor necrosis factor (TNF) in vitro impairs the activationof T cells through the T cell receptor/CD3 complex;reversal in vivo by anti-TNF antibodies in patients withrheumatoid arthritis. J Clin Invest 1994;94:749–60.
27 Cope AP, Aderka D, Wallach D, et al. Soluble TNF receptorproduction by activated T lymphocytes: diVerential eVectsof acute and chronic exposure to TNF. Immunology 1995;84:21–30.
28 Cope AP, Liblau RS, Yang XD, et al. Chronic tumornecrosis factor alters T cell responses by attenuatingT cell receptor signaling. J Exp Med 1997;185:1573–84.
29 Cope AP. Regulation of autoimmunity by proinflammatorycytokines. Curr Opin Immunol 1998;10:669–76.
30 Baumgart DC, Olivier WA, Reya T, et al. Mechanismsof intestinal epithelial cell injury and colitis in inter-leukin 2 (IL2)-deficient mice. Cell Immunol 1998;187:52–66.
31 Sadlack B, Merz H, Schorle H, et al. Ulcerative colitis-likedisease in mice with a disrupted interleukin-2 gene. Cell1993;75:253–61.
32 Ciacci C, Mahida YR, Dignass A, et al. Functionalinterleukin-2 receptors on intestinal epithelial cells. J ClinInvest 1993;92:527–32.
EVect of TNF on liver T cells in PSC 141
www.gutjnl.com
