Liver-derived T cell clones in autoimmune chronic active hepatitis: Accessory cell function of hepatocytes expressing class II major histocompatibility complex molecules

CLINICAL

IMMUNOLOGY

AND

IMMUNOPATHOLOGY

54, 382-394

(1990)

Liver-Derived T Cell Clones in Autoimmune Chronic Active Hepatitis: Accessory Cell Function of Hepatocytes Expressing Class II Major Histocompatibility Complex Molecules ALESSANDRA FRANCO, VINCENZO BARNABA, GIOVINA RUBERTI, ROSALBA BENVENUTO, CLARA BALSANO, AND ANTONINO MUSCA Fondazione Andrea Cesalpino, Istituto I Clinica Medica, Universitd “La Sapienza,”

Roma, Italy

Thirty T cell clones were generated from T cell blasts, infiltrating the liver of autoimmune chronic active hepatitis (CAH) patients, stimulated with autologous hepatocytes expressing class II major histocompatibility complex (MHC) molecules and interleukin 2 (IL2). Sixteen clones were CD4 + and 14 were CD8 + ; all were CD25 + and WI31 + , revealing that all cell lines expressed the Q/8 chains of T cell receptor. Five CD4+ and 4 CD8+ T clones proliferated in response to hepatocytes expressing both class I and class II antigens. The hepatocyte recognition was MHC restricted because only class II MHC-matched hepatocytes were able to stimulate the CD4 + T clones, while only class I-matched hepatocytes stimulated CD8 + T clones, and because MoAbs to monomorphic determinants of class II antigens or to class I antigens appeared to block the response of the CD4+ and CD8+ T clones, respectively. These findings, together with the observation that autologous irradiated peripheral blood mononuclear cells (iPBMC) were unable to stimulate the clones, indicate that the response of these clones was directed to a liver membrane antigen in association with class II or class I MHC molecules on the surface of the hepatocytes. All the CD8 + T clones and 5 CD4 + T clones expressed high cytotoxic activity in a lectin-dependent cell-mediated cytotoxicity assay; 10 CD8 + and 3 CD4 + T clones also showed natural killer (NK)-like function. The cytolytic machinery was also present in those clones (both CD8 and CD4) recognizing the HLA-matched hepatocytes. All liver-derived T clones were able to produce high amounts of interferon (IFN)-7, as well as being capable of secreting IL2, following PHA stimulation. 8 1990 Academic

Press, Inc.

INTRODUCTION

Chronic active hepatitis (CAH)’ may be caused by a variety of distinct liver disease states, including hepatitis B and non A-non B virus infections, drug reactions, alcoholic liver disease, and autoimmune disease involving the liver (1); the hallmark of this pathology is the presence of infiltrating T lymphocytes in portal tracts and lobular area in liver tissue (2). Autoimmune CAH is a disease of unknown etiology, in which the histological hallmark of infiltrating T lymphocytes is accompanied by high titers of serum nonorgan-specific and liver membrane autoantibodies (3) and by the frequent association with the HLA antigens B8, DR3 ’ Abbreviations used: CAH, chronic active hepatitis; MHC, major histocompatibility complex; MoAb; Monoclonal antibody; rIL2, recombinant interl&kin 2; PHA, phytohemagglutinin; PBMC, peripheral blood mononuclear cells; CTL, cytotoxic T lymphocytes; NK, natural killer; LDCC, lectindependent cell-mediated cytotoxicity; IFN, interferon; LAK, lymphokine-activated killer; TCR, T cell receptor; APC, antigen-presenting cells. 382 0090-1229190 Copyri&t All rights

$1.50

0 1990 by Aczhmic Press, Inc. of reproduction in any form reserved.

LIVER-DERIVED

T CELL

CLONES

IN

AUTOIMMUNE

CAH

383

(4). Although peripheral blood lymphocytes from CAH patients have been shown to be cytotoxic to autologous hepatocytes (5-7), the cytotoxicity assay used in these studies is obsolete because of low sensitivity, and we cannot be sure that peripheral lymphocytes reflect the functions of intrahepatic immunocytes. Recent studies reported the isolation of liver-derived T cell clones from patients with hepatitis B virus-induced CAH and the characterization of these clones in terms of hepatitis B core antigen (HBcAg) specificity, antigen-specific helper function (8), and cytolytic activity (9, 10). Moreover, the possibility of growing long-term T cell lines and clones (11) has established in autoimmune pathology that epithelial cells expressing class II major histocompatibility complex (MHC) antigens can perform accessory cell functions with respect to autoreactive T cell clones (12, 13). We recently demonstrated that interferon (IFN)-y induces class II antigens on normal human hepatocytes and that the same molecules are also spontaneously expressed on hepatocytes during chronic hepatitis (14). In this study, we isolated liver-derived T cell clones from autoimmune CAH patients with hepatocytes expressing class II MHC molecules to investigate the possibility that class II antigen-positive hepatocytes could serve to perpetuate the immune response by the presentation of liver membrane antigens to T cells. MATERIALS

AND METHODS

Patients The study population consisted of two female patients (patient 1 and patient 2) with hepatitis B surface antigen (HBsAg) seronegative CAH, with high titers of antinuclear (> 1:80) and anti-smooth muscle (> 1: 160) autoantibodies and associated Sjogren’s syndrome and thyroiditis, respectively (“autoimmune” CAH). Aspartate aminotransferase and alanine aminotransferase ranged from 54 to 171 (normal, ~37) Ill/liter and from 32 to 253 (normal, ~43) III/liter; IgG range was 1471 to 2870 (normal, ~1760) mg/dl. Patient I was a 44-year-old female with an inferred HLA-Al;B8;DR3 haplotype; while patient 2 was a 38-year-old female with an inferred HLA-Al;B8;Cwl;DR3 haplotype. The histological diagnosis was assessed according to the suggestions of an international committee (15). HBsAg and HB core Ag were negative in hepatocyte cytoplasms or nuclei by immunofluorescence. Prior to the study, the patients had not been on immunosuppressive therapy. Detection of expression of class II MHC antigens on 4-pm-thick cryostat sections with various concentrations of monoclonal antibody (MoAb) to framework determinants of class II molecules and fluoresceine-labeled F(ab), goat antimouse IgG revealed hepatocytes positive for class II antigens in both patients. This study was approved by the Medical Ethics Committees of the Rome University; however, the liver biopsy specimens were used in part for histological diagnosis and in part for our study. Monoclonal

Antibodies

The murine MoAb W6/32 to monomorphic determinants of class 1 antigens was purchased from Cappel Laboratories (Cochranville, PA). The MoAb Q5/13 (16)

384

FRANC0

ET

AL.

to framework determinants of class II MHC antigen was kindly provided by Dr. P. Giacomini (Regina Elena Cancer Institute, Rome, Italy). OKT3 (antXD3), OKT4 (antiCD4), OKT8 (antiCD8), OKIa (anti-DR), OKT26a (anti-CD25 IL-2 receptor-bearing lymphocytes), and OKMl (anti-CD1 1) monoclonal antisera were purchased from Ortho (Raritan, NY) and MoAb anti-Leu 11 (anti-CD16) was from Beckton-Dickinson (Mountain View. CA). The WT31 MoAb, directed at a common epitope of the IX/B heterodimer of the antigen T cell receptor, was purchased from Sanbio (Uden, Holland). Generation of Liver-Derived Cloned T Cell Lines The liver biopsy specimens were washed extensively to eliminate contaminating blood lymphocytes and were then cut into small fragments in RPM1 1640 (Flow Laboratory, UK) supplemented with 10% heat-inactivated fetal calf serum (Flow Lab.), 1% glutamine (Flow Lab.), 25 mM Hepes (GIBCO), 1% sodium pyruvate (Flow Lab.), 100 U/ml penicillin (Flow Lab.), 100 p&ml streptomycin (Flow Lab.), and 2.5 p&ml fungizone (Flow Lab.) (complete medium). Lymphocytes were recovered from liver cell population by Ficoll-Hypaque density gradient centrifugation (17) and both populations were washed three times. The cell viability, determined by the trypan blue dye exclusion test, was more than 95% for lymphocytes and more than 50% for hepatocytes. The hepatocytes were plated (1 x 105) onto 24-well plates (Falcon) in which autologous lymphocytes (5 x 105) derived from liver biopsy were added with autologous irradiated (3000 R) peripheral blood mononuclear cells (iPBMC) (3 x 10’). As control, peripheral lymphocytes, depleted of adherent cells, were incubated with autologous hepatocytes with the same procedure described above. The wells were fed twice weekly with 50 U/ml human recombinant interleukin 2 (rIL2, Biogen N.V., Geneva, Switzerland). Two to 3 weeks later, sufficient liver-derived T cells were generated, harvested, washed three times, and cloned by limiting dilution. For cloning, different procedures were used for the two patients. T cell blasts, obtained from patient 1, were seeded at 0.3 cell/well in 96-well flat-bottomed plates (Falcon plastics) in the presence of rIL2 (25 U/ml) and autologous iPBMC; while T blasts, obtained from patient 2, were seeded at 0.3 cell/well in the presence of 1 pg/ml phytoemoagglutinin (PHA-P; Wellcome, Beckenham, G.B.), 25 U/ml rIL2, and allogeneic iPBMC. After 2-3 weeks, cell growth was detected using an inverted microscope. Growing cells of patient 1 were expanded in 24-well plates in rIL2-containing medium and HLH-matched hepatocytes plus autologous iPBMC every 1-2 weeks; while cells of patient 2 were expanded in rIL2-containing medium and allogeneic iPBMC plus PHA every 2-3 weeks. Isolation of Hepatocyte Stimulator Cells The hepatocytes, used as stimulators in the proliferation assay of T clones, were isolated from liver biopsies of CAH patients, who may or may not have shared class I or class II antigens with the T cell clone donors. Tissue specimens were cut into small fragments in complete medium containing 0.01% collagenase type IV (Sigma, St. Louis, MO) for 3 hr at 37°C and 5% CO,. After three washes with complete medium, the cell viability determined by trypan blue dye exclusion

LIVER-DERIVED

T CELL

CLONES

IN

AUTOIMMUNE

CAH

385

test was always higher than 90%. The interval between obtaining a liver biopsy and having hepatocytes ready for cell proliferation assays ranged from 3.5 to 4 hr. Liver cells were counted and used in part for the proliferation assay and in part for class II antigens analysis by indirect immunofluorescence. Analysis of class II molecules on hepatocytes was performed as previously described (14). Proliferation

Assay

Cloned T cell lines were washed extensively and were incubated (1 x 104) in 200 ~1 of complete medium in quadruplicate with or without 1 x 104/well HLAmatched or non HLA-matched hepatocytes or lo5 autologous iPBMC, in 96-well flat-bottomed plates (Falcon plastics) at 37°C in 5% CO, for 72 hr. Eighteen hours before harvesting cultures onto glass fiber discs, using an automatic cell harvester (Skatron, Lierbyen, Norway), 1 $Yi of [3H]thymidine (So&, Saluggia, Italy) was added to each well. Radioactivity in the filters was measured in a liquid scintillation counter (LS 1800 counter Beckman Instruments, Irvine, CA). Results are expressed as mean + SEM. Blocking Experiments

Inhibition of proliferative response to HLA-matched hepatocytes by MoAbs was performed. In some experiments, 10 &ml MoAbs to monomorphic determinants of either class II or class I antigens were added in quadruplicate to the wells, containing HLA-matched hepatocytes and liver-derived T cell clones. The cultures were incubated for 72 hr, then exposed to 1 &i of [3H]thymidine, and incorporation was measured by liquid scintillation counter, as described above. Results are given as mean +- SEM. Surface Phenotype Analysis

Cloned T cells were analyzed by indirect immunofluorescence on a fluorescence microscopy, as described in detail in a previous paper (18), with the use of the murine MoAbs described above. Cytotoxicity

Assay

Cytotoxic T lymphocyte (CTL) activity was tested in a 4-hr 51Cr release assay. Cloned T cells were used as effector cells at effecter/target cell ratios 10/l and 3/l. Effector cells were incubated into V-bottom microtiter wells (Falcon 3047), to which 5 x lo3 “Cr-labeled target cells were added. The cell lines used as targets in the cytotoxicity assay were the NK-sensitive K562 cell line (NK activity) and the NK-resistant Raji cell line in the presence of 1 l&ml PHA (lectin-dependent cell-mediated cytotoxicity, LDCC) (19). Microplates were centrifuged at 1OOgfor 5 min and incubated at 37°C for 4 hr. The plates were again centrifuged and the supernatants (SN) were assayed for ‘lCr release. Specific lysis was calculated as percentage cytotoxicity: 100 x (cpm experimental release - cpm spontaneous release)/(total cpm-cpm spontaneous release). Lymphokines Production

Cloned T cells (2

x

105) were washed extensively

and cultured (0.2 ml complete

386

FRANC0

ET

AL.

medium in flat-bottomed wells) in the presence or in the absence of PHA (1 pg/ml) at 37°C. SN were removed after 24 hr for IL2 and after 48 hr for IFN-y determinations. IL2 activity was assessed using a commercial assay kit (InterTest 2, Genzyme, Boston, MA); for the quantitative measurements of IFN-y, the Centocor interferon-y RIA (Centocor Inc., Malvem, PA) was used. RESULTS

Generation of Liver-Derived T Cell Clones T cell clones, derived from liver expressing class II MHC molecules on hepatocytes of autoimmune CAH patients, were analyzed. When hepatic lymphomononuclear cells from CAH patients were cultured in the presence of autologous hepatocytes expressing class II antigens and autologous iPBMC, a strong growth was observed on an inverted microscopy after 5-6 days. Since no blast clusters were observed when the peripheral blood lymphocytes from the same patients were incubated with autologous hepatocytes, the possibility that the T clones considered in this study were derived from circulating lymphocytes contaminating liver biopsies is very unlikely. Owing to their extensive growth, T blasts from two hepatocyte-stimulated cultures (patient 1 and patient 2) were expanded with IL2 and then cloned. After 2 to 3 weeks, proliferating cells were apparent in 8 out of 480 wells containing T blasts cloned in the presence of rIL2 and autologous iPBMC (patient 1); while 22 cloned lines were generated from 280 wells containing T blasts cloned in the presence of PHA, rIL2, and allogeneic iPBMC (patient 2). The comparison of the clonal efficiency in the two patients suggests that the procedure used in patient 2 is more successful in order to obtain a major number of hepatocyte-specific T clones generated from liver-infiltrating lymphocytes. Thereafter, the T cell clones were tested for their capacity to proliferate in response to hepatocytes expressing class II MHC antigens isolated from CAH patients, who may or may not have shared MHC antigens with T cell lines donors, and for their cytotoxic potential. In parallel to the cytolytic ability of the clones, phenotypic analysis revealed that all cell lines expressed the CD3/Ti antigen recognition structure because they reacted with anti-CD3 and WT31 MoAbs, although they showed distinct phenotypes: 2 T clones were CD3 + , CD4 - , CD8 + and 6 were CD3 + , CD4 + , CD8 - in patient 1; while 12 were CD3 + , CD4 - , CD8 + and 10 were CD3 + , CD4 + , CD8 - in patient 2. All clones resulted TAC + (CD25), while none expressed large granular lymphocyte (LGL) molecules (CD16 and CD1 I). Proliferative Response to Hepatocytes and MHC Restrictions of Liver-Derived T Cell Clones The capacity of liver T cell clones to respond to hepatocytes was evaluated. Two CD4 + T clones from patient 1 and four CD8 + and three CD4 + T clones from patient 2 proliferated in response to HLA-matched hepatocytes. The CD4 + T clones from both patients proliferated in the presence of hepatocytes expressing class II antigens isolated from CAH patients who shared the same HLA-DR molecules as the T cells donors (Experiments 1 and 2 in Table 1; Experiments 1

LIVER-DERIVED

T CELL

CLONES

IN

AUTOIMMUNE

387

CAH

and 2 in Table 2) and not in the presence of hepatocytes of CAH patients with different HLA-DR molecules (Experiment 3 in Table 1; Experiments 5 and 6 in Table 2); while the CD8 + T clones proliferated in response to hepatocytes from CAH patients sharing class I HLA antigens (Experiments 1 and 2 in Table 2) and not in response to hepatocytes from CAH patients sharing no HLA-A,B,C determinant with T cells donor (Experiment 3 in Table 2). Moreover, the T clones of patient 2 did not proliferate in response to autologous iPBMC, excluding the possibility of an autologous mixed lymphocyte response (Experiment 7 in Table 2). In addition, Table 3 confirms that the hepatocyteinduced proliferation was MHC restricted since at least for clones PM5 and PM1 1 with phenotype CD4 and for clone PM7 with phenotype CDS, MoAbs to monomorphic determinants of class II antigens or to class I antigens inhibited the proliferation of CD4 + clones and CD8 + clones, respectively. These MoAbs had no toxic effect on IL2-induced proliferation of the T cell lines (Table 3). Analysis of Cytolytic Activity

of Liver-Derived

T Cell Clones

Several experiments were performed to detect the antigen-specific cytotoxic capacity of the T clones, using autologous hepatocytes as target cells, in a 5*Cr release assay. The hepatocytes were not capable of functioning as target cells because of their high spontaneous 51Cr release, ruling out the possibility of setting up an adequate hepatocyte-specific cytotoxicity test and demonstrating whether liver-derived T clones could actually destroy normal or compromised liver cells. Thus, the cytolytic potential of the T clones from patient 2 was assessed by an LDCC assay, detecting cytolytic cells of any specificity, and by an NK assay. The results in Fig. 1 show that all CD8 + and 5 out of 10 CD4 + T clones were able to lyse unrelated target cells when cultured in the presence of PHA; while 10 out of 12 CD8 + T clones and 3 out of 10 CD4 + T clones also showed NK activity. It is TABLE PROLIFERATIVE

RESPONSE

1

OF LIVER-DERIVED T CELL CLONES WITH HEPATOCYTES’

Exp

Stimulator

HLA antigen@

Class II molecule expression’

1.

Hepatocytes -

(Al, B8, DR3)

2.

Hepatocytes -

3.

Hepatocytes -

FROM

PATIENT

1 STIMULATED

CD4+ T clones (HLA-Al,

B8, DR3)

sv3

sv7

+

13,411 t 425 721 t 84

ND

(ND, ND, DR3)

+

ND

16,425 + 515 631 ? 112

(A3, B35, DR5)

+

1,110 t 88 825 2 25

466-t 188 745 + 226

a T cells were cultured with or without hepatocytes. Cell proliferation was estimed by incorporation of [3H]thymidine. Results are expressed as the mean cpm 2 SEM. ND, not determined. * HLA antigens of CAH patient donors of stimulator hepatocytes. ’ Class II MHC antigen expression on stimulator hepatocytes was revealed with MoAb to framework determinants of class II molecules, as described under Materials and Methods.

Hepatocytes -

Hepatocytes

Hepatocytes

Hepatocytes -

Hepatocytes

Autologous irradiated

1.

2.

3.

4.

5.

6.

7.

DR3)

(ND, ND, ND, DR41DRS)

(ND, ND, ND, DR4DRwlO)

(AlO, B35, Cw3, DR3)

tA30. 812, Cw4, DR3)

(Al, B8, Cwl,

(Al, BE. Cw4, DR3)

HLA antigen@

RESPONSE

+

+

+

-

+

+

+

Class II molecule expression’

OF LIVER-DERIVED

250 rt 35 320 k 93

443 + 109 410 _c 115

322 -c 138 343 -e 24

380 c 97 266 + 24

ND

12,231 -’ 420 6.53 C 155

FROM

720 f 37 443 + 65

ND

415 A 88 366 k 37

222 + 61 315 + 65

ND

ND

17,154 2 510 209k 15

PM11

CD4

CLONES

TABLE2

9,827 It 364 307 5 65

PM5

T CELL

PM18

PM7

ND

625 2 129 699 + 99

255 -t 24 157 + 8

ND

715 + 28 610 k 127

581 -c 207 486 -c 37

427 t 65 271 * 35

125 % 15 213 + 22

ND

ND

ND

ND

ND

611 + 65 321 + 32

PM9 1320 k 460 289 2 35

PMlS 6915 2 389 114 + 32

535 + 200 453 -1- 198

ND

ND

ND

666 t 217 237 + 33

11,210 k 382 1078 -+ 72

CD8

HEPATOCYTES”

6355 k 592 356 + 24

5981 ‘_ 389 304 _c 68

1,454 + 162 895 t_ 155

ND

8,720 k 433 1,114 + 381

I 1,532 t_ 454 1,247 + 207

WITH

B8, Cw2, DRl/3)

2 STIMULATED

T clones (HLA-AI.

PATIENT

PM21

864 it_ 215 432 L 77

ND

ND

1,131 2 166 985 2 221

ND

10,261 f 342 1,044 f 281

u T cells were cultured with or without hepatocytes or autologous IPBMC. Cell proliferation was estimed by incorporation of [3H]thymidine. Results are expressed as the mean cpm + SEM. ND, not determined. h HLA antigens of CAH patient donors of stimulator hepatocytes: inferred haplotypes (Expt 1-t) or phenotypes (Expt 5-6) c Class II MHC antigen expression on stimulator hepatocytes was revealed with MoAb to framework determinants of class II molecules, as described under Materials and Methods.

PBMC

Stimulator

Hepatocytes

Bxp

PROLIFERATIVE -

7

0

LIVER-DERIVED EFFECT

OF ANTI-MHC LIVER-DERIVED

Hepatocytes* +

rIL2 -

+

-

T CELL CLONES IN AUTOIMMUNE

CAH

389

TABLE 3 ANTIGEN MONOCLONAL ANTIBODIES ON THE PROLIFERATIVE RESPONSE OF T CELL CLONES STIMULATED WITH HLA-MATCHED HEPATOCYTES” __..__T cell clones MoAb

PM5 (CD4)

PM11 (CD4)

PM? (CD8)

89925460 14,010 + 320 6122 2 315 Anti-class I (W6/32) 9888 k 325 13.732 + 344 1225 -t 115 + Anti-class II (Q5/13) 1121 2 198 2,615 rl- 290 6255 +- 255 + Anti-class I (W6/32) ND 12,111 t 420 7032 -f 360 + Anti-class II (Q5/13) ND 13,425 r 333 5411 c 252 --. --’ T cells were incubated with HLA-matched hepatocytes in the presence or absence of anti-MHC antigen MoAbs. Some wells received rIL2 plus MoAb in absence of hepatocytes. Cell proliferation was estimated by incorporation of [3H]thymidine. Results are expressed as the mean cpm _t SEM. b The hepatocytes were obtained from CAH patients sharing the same HLA antigens (Al, B8, DR3) of T clone donor.

interesting to note that all of the clones (both CD8 + or CD4 + ) specifically recognizing the hepatocytes had a cytotoxic potential. In contrast, no clone had cytolytic activity against NK-resistant Raji cells in the absence of PHA, indicating that these cytotoxic T clones were not lymphokine-activated killer (LAK) cells. Lymphokine Production

All liver-derived T clones from patient 2 were assayed for their ability to produce IFN-)I and IL2 following PHA stimulation. As shown in Fig. 2, both CD4 + and CDS + T clones released high amount of IFN-r, as well as being capable of 100

. .

80 (Y 9

:. : * I

.3

.

100

1 . 80 x

y" u)

60

t

60

.

P ;

40

40

CD4

: w m ; L -

CD8

FIG. 1. NK and LDCC cytolytic activities of liver-derived T cell clones. The clones were tested against an NK-sensitive human K562 cell line (NK assay) and against an NK-resistant human Raji cell line with or without 1 pg/ml PHA (LDCC assay). Results are given as percentage specific lysis at a 10/l effectorkarget ratio. *T cell clones recognizing HLA-matched hepatocytes.

FRANC0

390 150

l

ET AL.

0

-6

.

0

CD4 CD4 CDS CLONES C fODI’fE 5 CLONES CLONES I101 (12) (IO) (101 FIG. 2. IFN-y and IL2 productions by liver-derived T cell clones. Cloned T cells (2 X ld) were cultured in the presence or in the absence of PHA (1 p&ml). SN were removed after 24 hr for IL2 and after 48 hr for IFN-y determinations. Numbers in parentheses indicate the number of clones assayed for IFN-?I and IL2 production.

secreting IL2. None of the clones showed any detectable production kines without PHA stimulation.

of lympho-

DISCUSSION

This report investigated T cell clones derived from lymphocytes infiltrating the liver from autoimmune CAH patients in terms of their phenotype and functional capacity. The study of infiltrating T lymphocytes was hampered by the small number of liver-derived lymphocytes from liver biopsy specimens. Using techniques for growing long-term T clones (1 l), it has been possible to establish cloned T cell lines from liver-derived T blasts stimulated with autologous hepatocytes expressing class II MHC antigens. Several studies have demonstrated that during the course of different inflammatory processes, class II MHC molecules (20) are also expressed on nonlymphoid cells (21-24), and that some cytokines enhance class II molecule expression on cells normally negative for these antigens (25-29). Since recent findings have established that hepatocytes may display these effects during chronic hepatitis (14, 30), we examined whether or not such activated liver cells can perform accessory cell functions with respect to liver-derived T cell clones in autoimmune CAH, as demonstrated in other autoimmune diseases (13, 31, 32). In this study, five CD4 + and four CD8 + liver-derived T clones from patients with autoimmune CAH proliferated in response to hepatocytes expressing both class I and class II MHC molecules, obtained from patients with CAH, who may or may not have shared class I or class II molecules with donors of T clones. These data suggest that the class II expression on hepatocytes reflects acquisition

LIVER-DERIVED

T CELL

CLONES

IN AUTOIMMUNE

CAH

391

of a capacity for antigen presentation by these cells for the class II-restricted CD4+ T clones. The use of specific cloned T cells derived from a target organ such as the liver may explain the very high values observed in some of the proliferation assays, in contrast to other studies using autoantigens to stimulate noncloned, albeit tissuederived, lymphocytes (33). The hepatocyte recognition was MHC restricted because only hepatocytes derived from class II MHC-matched patients were able to stimulate the CD4 + T clones, while only hepatocytes derived from class I-matched patients stimulated the CD8 + T clones, and because MoAbs to monomorphic determinants of class II antigens or to class I antigens appeared to block the response of the CD4 + and CD8 + T clones, respectively. In our study, the HLA-DR3 allele seems to be one of the restricting molecules for hepatocyte recognition by CD4+ T cell clones, but this restriction may be fortuitous. Autoimmune CAH is associated with the serologic marker HLA-DR3 phenotype, but this association is incomplete; some of the autoimmune CAH patients lacked the disease-associated HLA-DR3 specificity. It is possible that more than one MHC class II molecule can present liver membrane determinants to the specific T cells. It has been reported that single peptides can stimulate helper T cells in the context of several MHC class II molecules (34, 35). Alternatively, the incomplete association between HLA molecules and autoimmune CAH could be explained by the hypothesis that gene conversion-like events could be involved in shuflling oligonucleotides, encoding epitopes not identified by conventional serologic typing, into different HLA-D frameworks (36, 37). The observation that the autologous irradiated PBMC, containing cells normally expressing HLA-DR molecules, were unable to stimulate the clones indicates that the responses of these CD4 + and CD8 + T clones were not directed to class II or class I molecules alone, but presumably to a liver membrane antigen (autoantigen?) in association with class II or class I MHC molecules on the surface of the hepatocytes. Antigen-specific T cell activation normally results from the formation of a ternary complex involving the T cell receptor (TCR), the antigen, and class I or class II MHC molecules (38, 39). We analyzed the liver-derived cloned T cell lines for cytotoxic activity. Unfortunately, several attempts, using human hepatocytes as target cells in a 51Cr release assay, have been revealed unsuccessful because of the inability of hepatocytes to retain “Cr. Recently, c y to1 y tic activity of peripheral blood lymphocytes from hepatitis A patients has been detectable by a microcytotoxicity assay using %--labeled hepatitis A virus-infected skin fibroblasts (40). Lacking HLA-matched hepatocyte lines necessary for the study of specific T cell cytotoxicity, the cytolytic potential of the clonal progeny of hepatic T cells was assessed by an LDCC assay or by an NK assay. All of the CD8 + T clones tested, but also a half of CD4 + T clones, lysed Raji target cells in the presence of PHA. Since the LDCC assay detects the cytolytic activity of T lymphocytes irrespective of their specificity (19), our data indicate that in most of the clones the cytotoxic machinery was present, as demonstrated also in chronic hepatitis B or in primary biliary cirrhosis (10, 11). Moreover, the functional analysis of clones

392

FRANC0

ET AL.

derived from hepatic T cells revealed that a proportion of both CD8 + and CD4 + cytolytic precursors had NK-like function against a NK-sensitive target (K562 line). These cytolytic clones did not express LGL molecules and did not display LAK phenomenon (41). It should be noted that the cytotoxic machinery was also present in T clones (both CD8 and CD4) recognizing the HLA-matched hepatocytes, suggesting that these cells could recognize some antigens on the cell membrane of hepatocytes in a MHC-restricted manner and subsequently proceed to kill the presenting cells. Since hepatocytes express class II MHC antigens during chronic inflammatory liver diseases (14, 30), and also in our patients, a specific cytotoxic interaction between hepatic class II-restricted CD4+ T cells and hepatocytes could be involved in the pathogenesis of liver cell injury in autoimmune CAH. The class II MHC molecule expression on hepatocytes may be modulated by IFN-y released by infiltrating T lymphocytes, since the liver-derived T clones were able to produce high amounts of this lymphokine. We recently demonstrated that IFN-y induces expression of class II molecules on normal human hepatocytes in vitro (14). The nature of the liver cell membrane antigen or the association with MHC molecules still needs to be clarified. Recently, the study of the structure of class I molecules seems to have defined that the TCR binds the antigen and MHC molecule as a complex epitope and not as two independent entities (42). Moreover, at least for class II-restricted T cells responding to soluble antigens, the antigen is usually not seen in its native form but only after some type of metabolic “processing” by the antigen-presenting cells (APC) that denatured and perhaps enzymatically cleaved the protein antigens (43,44). More recent evidence suggests that processing may also be involved in antigen recognition by at least some cytotoxic class I-restricted T cells (45, 46). The use of the liver-derived cloned T lines should elucidate the nature and the processing pathways of the liver membrane epitopes recognized by cytotoxic class I- or class II-restricted T cells in autoimmune CAH. In this regard, experiments using histocompatible APC from blood together with liver membrane preparations from various sources to stimulate the liverderived T clones need to be performed; these studies may further define the APC function of hepatocytes to present liver membrane antigens to T cells. ACKNOWLEDGMENTS We thank Dr. Manlio Ferrarini for helpful discussions and Dr. Marianna Suppa for her valuable advice in the manuscript preparation. This work was supported by “Centro Nazionale delle Ricerche” Grant 86.00271.04.

REFERENCES 1. Sherlock, S., “Diseases of Liver and Biliary System,” 6th ed., Blackwell, Oxford, 1981. 2. Colucci, G., Colombo, M., Del Ninno, E., and Paronetto, F., Gastroenterology 85, 1138, 1983. 3. Mackay, I. R., and Frazer, I. H., “Advances in Hepatitis Research” (F. V. Chisari, Ed.), p. 179, Masson Pub., 1984. 4. Mackay, I. R., and Tait, B. D., Gastroenterology 79, 95, 1980.

LIVER-DERIVED

T CELL

CLONES

IN

AUTOIMMUNE

CAH

393

5. Mieli Vergani, G., Vergani, D., Jenkins, P. J., Portmann, B., Mowat. A. P., Eddleston, A.L. W. F. ,and Williams, R., Clin. Exp. Immunol. 38, 16, 1979. 6. Mondelli, M., Mieli Vergani, G., Alberti, A., Vergani, D., Portmann, B., Eddleston, A. L. W. F.. and Williams, R., J. Immunol. 129, 2773, 1982. 7. Bamaba, V., Levrero, M., France, A., Ruberti, G., Musca, A., Bonavita, M. S., and Balsano, F., Liver 6, 45, 1986. 8. Ferrari, C., Mondelli, M. U., Penna, A., Fiaccadori, F., and Chisari, F. V., J. Immunol. 139,539, 1987. 9. Hoffmann, R. M., Pape, G. R., Rieber, P.. Eisenburg, J., Dohrmann, J., Zachoval, R., Paumgartner, G., and Reithmuller, G., Eur. J. Zmmunol. 16, 1635, 1986. 10. Meuer, S. C., Moebius, U., Manns, M. M., Dienes, H. P., Ramadori, G., Hess, G., Hercend, T.. and Meyer zum Buschenfelde, K. H., Eur. J. Immunol. 18, 1447, 1988. 11. Fathman, C. G., and Futch, F. W., “Fundamental Immunology” (W. E. Paul, Ed.), p. 781, Raven Press, New York, 1984. 12. Londei, M., Bottazzo, G. F., and Feldmann, M., Science 228, 85, 1985. 13. Weetman, A. P., Volkman, D. J., Burman, K. D., Margolick, J. B., Petrick, P., Weintraub, B. D., and Fauci, A. S., Clin. Immunol. Immunopathol. 39, 139, 1986. 14. France, A., Bamaba, V., Natali, P. G., Balsano. C., Musca, A., and Balsano, F., Hepatology 8. 449, 1988. 15. Fogarty International Center Proceedings, “Nomenclature, Diagnostic Criteria and Diagnostic Methodology for Disease of the Liver and Biliary Tract,” Washington, DC., U.S. Govt. Printing Oflice, N. 22, 1976. 16. Quaranta, V. L., Walker, E., Pellegrino, M. A., and Ferrone, S.. J. fmmunol. 125, 1421, 1980. 17. Boyum, A., Stand. J. Clin. Lab. Invest. Suppl., 97, 9, 1983. 18. Bamaba, V., Zaccari, C., Levrero M., RUOCCO.G., and Balsano, F., Clin. Immunol. Immunopathol. 26, 83, 1983. 19. Moretta, A., Pantaleo, G., Moretta, L., Mingari, M. C., and Cerottini, J. C.. J. Exp. Med. 158, 571, 1983. 20. Bach. F. H., Zmmunol. Today 5, 89, 1985. 21. Hanafusa, T., Pujol-Barrel, R., Chiovato, L., Russel, R. C. G., Doniach, D., and Bottazzo, G. F., Lancet 2, 111, 1983. 22. Hall, B. M., Bishop, G. A., Duggin, G. G., Horvatt, J. S., Philips. J.. and Tiller, D. J., Lancer 2, 247, 1984. 23. Lampert, I. A., Switters, A. J.. and Chilsom, P. M., Nature (London) 293, 149. 1981. 24. Barclay, A. N., and Mason, D. W., J. Exp. Med. 156, 1665, 1982. 25. Wong, G. H. W., Bartlett, P. F., Clark-Lewis, I., Battaye, F., and Schrader. J. W.. Nature (London) 310,688, 1984. 26. Collins, T., Konnan, A. J., Wake, C. T., Boss, J. M., Kappes, D. J., Fiers, W., Ault, K. A., Gimbrone, M. J., Jr., Strominger, J. L., and Pober. J. S., Proc. Natl. Acad. Sci. USA 17, 4917, 1984. 27. Todd, I., Pujol-Borrell, R., Hammond, L. J., Bottazzo, G. F., and Feldmann, M., Clin. Exp. Immunol. 61, 265, 1985. 28. Momberg, F., Koch, W., Molter. P., Moldenhaner, G., Butcher, G. W., and Hammerling, G. J.. J. Immunol. 136, 940, 1986. 29. Pujol-Borrell, R., Todd, I., Doshi, M., Bottazzo, G. F., Sutton, R., Gray, D.. Adolf, G. R., and Feldmann, M., Nature (London) 326, 304, 1986. 30. van den Oord, J. J., de Vos, R., and Desmet, V. J., Hepatology 6, 981, 1986. 31. Londei, M., Lamb, J. R., Bottazzo, G. F., and Feldmann, M., Nature (London) 3112, 639, 1984. 32. Sun, D., and Wekerle, H., Nature (London) 320, 70, 1986. 33. Melms, A., Schalke. B. C. G., Kirchner, Th., Muller-Hermelink, H. K., Albert, E., and Wekerle. H., J. Clin. Invest. 81, 908, 1988. 34. Lamb, J. R.. MC Michael, A. J.. and Rothbard, J. B., Hum. Immunoi. 19, 79. 1987. 35. Buus, S., Sette, A., Colon, S. M., Miles, C., and Grey, H. M., Science 235, 1353, 1987. 36. Goronzy, J., Weyand, C. M., and Fathman, C. G., J. Clin. Invest. 77, 104, 1986.

394

FRANC0

ET AL.

37. Mengle-gaw, L., Conner, S., MC Devitt, H. O., and Fathman, C. G., J. Exp. Med. 160, 1184, 1984. 38. Schwartz, R. H., Annu. Rev. Immunol. 3, 237, 1985. 39. Livingstone, A. M., and Fathman, C. G., Annu. Rev. Immunol. 5, 477, 1987. 40. Vallbracht, A., Gabriel, P., Maier, K., Hartmann, F., Steinhardt, H. J., Muller, C., Wolf, A., Manncke, K. H., and Flehming, B., Hepatology 6, 1308, 1986. 41. Grimm, E. A., Mazumder, A., Zhang, H. Z., and Rosemberg, S. A., J. Exp. Med. 155, 1823, 1982. 42. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C., Nature (London) 329, 512, 1987. 43. Germain, R. N., Nature (London) 322, 687, 1986. 44. De Lisi, C., and Berzofsky, J. A., Proc. Natl. Acad. Sci. USA 82, 7048, 1985. 45. Townsend, A. R. M., Rothbard, J., Gotch, F. M., Bahadur, G., Wraith, D., and McMichael, A. J., Cell 44, 959, 1986. 46. Maryanski, J. L., Pala, P., Corradin, G., Jordan, B. R., and Cerottini, J. C., Nature (London) 324, 578, 1986. Received March 20, 1989; accepted with revision October 9, 1989