Cytotoxic T Lymphocyte Antigen-4 (CTLA-4) Gene Polymorphisms and Susceptibility to Type 1 Autoimmune Hepatitis

Transcription

1 Cytotoxic T Lymphocyte Antigen-4 (CTLA-4) Gene Polymorphisms and Susceptibility to Type 1 Autoimmune Hepatitis KOSH AGARWAL, 1 ALBERT J. CZAJA, 2 DAVID E. J. JONES, 1 AND PETER T. DONALDSON 1 Genetic susceptibility to type 1 autoimmune hepatitis is indicated by a preponderance of female subjects and strong associations with human leukocyte antigens (HLA) DRB1*0301 and DRB1*0401. The gene encoding cytotoxic T-lymphocyte antigen-4 (CTLA-4) on chromosome 2q33 may also influence autoimmunity. To determine the frequency and significance of the exon 1 adenine (A)-guanine (G) base-exchange polymorphism for CTLA-4 in patients with type 1 autoimmune hepatitis, 155 northern European Caucasoid patients and 102 ethnically-matched control subjects were tested by polymerase chain reaction. The genotype distribution was significantly different in patients compared to controls (AA 50/155 patients vs. 51/102 controls; AG 84/155 patients vs. 38/102 controls; GG 21/155 patients vs. 13/102 controls, , P.011). This difference was caused by a significant over-representation of the G allele in patients compared to controls (105/155 patients vs. 51/102 controls, , P.004, odds ratio 2.12). The GG genotype was associated with a significantly higher mean serum aspartate transaminase level (P.03), greater frequency of antibodies to thyroid microsomal antigens (P.004) and was found more commonly in patients with HLA DRB1*0301 (P.02). Treatment outcomes, however, were not affected by the genotype. The CTLA-4 G allele is more common in patients with type 1 autoimmune hepatitis and may represent a second susceptibility allele. Furthermore, there may be synergy between the HLA-DRB1*0301 and the GG genotype in terms of disease risk. (HEPATOLOGY 2000;31:49-53.) Autoimmune hepatitis (AIH) is an unresolving hepatocellular inflammation of unknown cause. The diagnosis requires the presence of interface hepatitis on histological examination, hypergammaglobulinemia, abnormally elevated serum Abbreviations: AIH, autoimmune hepatitis; ANA, antinuclear antibodies; SMA, smooth muscle antibodies; LKM-1, liver-kidney microsomal antibodies type 1; HLA, human leukocyte antigen; CTLA-4, cytotoxic T-lymphocyte antigen-4; A, adenine; G, guanine; Thr, threonine; ALA, alanine; bp, base pair. From the 1 Centre for Liver Research, Faculty of Clinical Medical Sciences, The Medical School, University of Newcastle, Framlington Place, Newcastle upon Tyne, UK; and the 2 Department of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN. Received August 17, 1999; accepted October 14, Address reprint requests to: Peter T. Donaldson, Ph.D., Centre for Liver Research, Faculty of Clinical Medical Sciences, The Medical School, University of Newcastle, Framlington Place, Newcastle upon Tyne, UK NE2 4HH. P.T.Donaldson@ ncl.ac.uk; fax: Copyright 2000 by the American Association for the Study of Liver Diseases /00/ $3.00/0 49 transaminase levels, and liver-associated autoantibodies in serum. 1 Type 1 AIH is characterized by the presence of circulating antinuclear antibodies (ANA) and/or smooth muscle antibodies (SMA) and it is the major form of AIH in Caucasoid adults. Type 2 AIH is characterized by antibodies to liver-kidney microsome type 1 (anti-lkm1) and occurs mainly in children. 2,3 Genetic factors contribute to susceptibility for most autoimmune diseases, including Graves disease, myasthenia gravis, multiple sclerosis, primary Sjögren s syndrome, systemic lupus erythematosus, insulin-dependent diabetes mellitus, and rheumatoid arthritis. 4 In type 1 AIH, genetic involvement is indicated by a marked female preponderance and strong, well-established, human leukocyte antigen (HLA) associations. 5,6 Recent studies suggest that approximately 85% of northern European and North American Caucasoid patients with type 1 AIH have the DRB1*0301 and/or DRB1*0401 alleles. 7,8 The genetic basis for HLA-encoded susceptibility to type 1 AIH is not fully understood, but a hypothesis based on sharing of a common 6 amino acid motif, LLEQKR, at positions of the DR polypeptide of the expressed DR molecule has been developed. 7,8 It is proposed that the critical amino acid within this motif is the lysine residue at position 71 and this residue, which is on the lip of the antigen-binding groove influences both the steric configuration of the antigenpresenting complex and its subsequent recognition by immunocytes. 8 This motif is common to both DRB1*0301 and DRB1*0401, and the presence of a lysine residue at position 71 may facilitate the binding and presentation of an as yet unidentified autoantigenic peptide. A number of genes outside the major histocompatibility complex may also play a role in susceptibility to type 1 AIH. Investigations so far have identified susceptibility loci within the T-cell receptor constant region genes, 9 the T-cell receptor variable region genes, 10 and the immunoglobulin heavy chain constant region genes. 11 None of these genes, however, including the HLA alleles, DRB1*0301 and DRB1*0401, account for all instances of disease and other genetic and/or environmental factors must be involved. Potential candidates include most genes encoding immunoregulatory proteins, especially the cytokine genes, genes encoding adhesion molecules, and genes encoding proteins involved in antigen processing, antigen presentation, and cell signalling. Recent studies of the interleukin-1, interleukin-10, and tumor necrosis factor polymorphisms in type 1 AIH have been disappointing. 12,13 Perhaps this is not surprising in light of the prominent role played by T cells in the pathogenesis of

2 50 AGARWAL ET AL. HEPATOLOGY January 2000 AIH. 14,15 Better candidates are genes that control the magnitude, phenotype, and/or termination of T-cell responses. Cytotoxic T lymphocyte antigen-4 (CTLA-4) is a T-cell surface molecule that interacts, in competition with the costimulatory molecule CD28, with the ligands B7-1 and B7-2 on antigen presenting cells. 16,17 It is increasingly being appreciated that the balance between the binding of CTLA-4 and of CD28 to their common ligands plays a significant role in the peripheral control of the cellular immune response. Factors modulating the expression or function of either CTLA-4 of CD28 could, therefore, play a role in altering this balance resulting in loss of peripheral control and the expression of autoreactive responses. The CTLA-4 gene located on chromosome 2q33 includes a single base exchange polymorphism in exon 1, where there is an adenine (A) for guanine (G) substitution at position 49 resulting in a threonine (Thr) for alanine (Ala) substitution in the expressed protein. 18 The aims of this study were to determine the frequency of the A to G base-exchange polymorphism in exon 1 of the CTLA-4 gene in a large, well-characterized series of patients with type 1 AIH and to assess the relationship between this polymorphism and the clinical expressions and behavior of the disease. PATIENTS AND METHODS Study Population. The study population consisted of 155 adult patients seen by one investigator (A.J.C.) at the Mayo Clinic, Rochester, MN. All patients fulfilled the diagnostic criteria recommended by the International Autoimmune Hepatitis Group 19 and they were classified as type 1 AIH based on antibody profiles. 3 All patients were Caucasoid and of northern European background. Their mean age at presentation to the Mayo Clinic was 46 1 years (range, years) and 127 were women (82%). Each lacked serological evidence of infection with hepatitis B virus and hepatitis C virus by second generation enzyme-linked immunosorbent assay. Control DNA was obtained from 102 healthy, white adults in the Mayo Clinic volunteer blood donor program. The mean age of the control group was 40 years (range, years) and 59 were women (58%). Each normal subject had indicated the absence of major illness on a standard blood donation questionnaire. All were residents of Olmsted County, MN, and their ethnic backgrounds were similar to those of the patients reflecting varying mixtures of German, French, English, Norwegian, Irish, Polish, Bohemian, Dutch, Swiss, Danish, Scots, and Swedish, that make up the North American white population. We were unable to identify any examples of mixed, non-caucasoid parentage in either the patients or the controls. All patients and normal subjects were genotyped for HLA DRB1, DQA1, and DQB1 as part of previously published studies 6,8 and HLA data are presented only for comparison with the CLTA-4 genotypes. Class I HLA (A and B loci) were determined in 153 patients (99%) and 98 normal subjects (96%) by a standard microlymphocytoxicity technique. 6,8 Our study had been approved by the Institutional Review Board of the Mayo Clinic as part of a program project grant. Clinical Assessments. Patients were screened at presentation for the presence of SMA, ANA, and anti-lkm1 by indirect immunofluorescence on murine kidney and stomach sections as described previously A serum titer of 1:40 or higher was considered positive for SMA and ANA. A serum titer of 1:10 or higher was considered positive for anti-lkm1. All patients (100%) were tested for SMA; 154 (99%) were tested for ANA; and 140 patients (90%) were tested for anti-lkm1. All patients had SMA, ANA, or both and 137 patients were seronegative for anti-lkm1. Two patients were indeterminant for anti-lkm1 and 1 had anti-lkm1. The latter individual was retained in the analysis because she also had ANA. 23 Thyroglobulin and microsomal thyroid antibodies were sought in 152 patients (98%) by standard techniques and a titer of at least 1:100 was considered positive. Concurrent non-hepatic disorders of an immunologic nature were sought in a uniform fashion in each patient in accordance with previously published guidelines. 6 The presence of thyromegaly, synovitis, colitis, and skin lesions were recorded systematically. If clinical history, physical examination, and/or laboratory findings showed abnormalities, additional studies were performed as indicated. Patients had clinical and laboratory features of active disease at the time of presentation to the Mayo Clinic and in 152 individuals who underwent liver biopsy examination, histological disease was also present. Previous treatments or ongoing therapy with corticosteroids did not preclude inclusion in our analysis. The manifestations of active disease at accession were presumed to be unaffected by previous treatments or ongoing therapy that had not induced disease remission. The term at presentation as used in this manuscript means when first seen at the Mayo Clinic and does not connote treatment status or refer to the initial onset of clinically significant disease. Treatment Outcomes. One hundred forty-one patients (91%) were treated with prednisone alone (51 patients) or a higher dose of prednisone in combination with azathioprine (90 patients). Each of these therapies had been shown previously to be equally effective in the management of type 1 autoimmune hepatitis. 1 Thirteen patients (8%) had disease of insufficient severity to warrant corticosteroid treatment and 1 patient (1%) received an investigational drug. Only the 141 patients receiving conventional corticosteroid regimens were assessed for associations between the CTLA-4 polymorphisms and treatment outcome (Table 3). Remission, treatment failure, and incomplete response constituted the treatment outcomes. Remission connoted disappearance of symptoms, normal conventional laboratory tests of liver inflammation (except for a 2-fold or less elevation of the serum aspartate transaminase level), and improvement of the liver biopsy tissue to normal, portal hepatitis, or inactive cirrhosis. Treatment failure connoted deterioration despite compliance with therapy and an incomplete response indicated improvement insufficient to satisfy remission criteria. Patients with an incomplete response required continuous corticosteroid therapy. Recrudescence of disease after remission and withdrawal of medication connoted relapse whereas continued satisfaction of remission criteria after termination of therapy connoted a sustained remission. Fibrosis with a complete regenerative nodule in sequential liver tissue samples indicated the development of cirrhosis. Histological Assessments. Liver tissue specimens were obtained in 152 patients at the time of accession (98%) and the histological findings were interpreted under code by experienced hepatopathologists at the Mayo Clinic. Each examiner reported the features in a standard fashion and each adhered to conventional histological criteria for a tissue diagnosis. 27 Thirty-six patients had cirrhosis at accession and 116 patients were at risk to develop cirrhosis. Of these 116 patients, 107 were treated with conventional corticosteroid regimens and underwent follow-up liver biopsy examination. These latter patients were assessed for the development of cirrhosis. Determination of CTLA-4 Gene Polymorphisms. A 328-base pair (bp) fragment of the first exon of the CTLA-4 gene, containing the A to G base-exchange polymorphism at position 49 (exon 1), was amplified using the primers 5 -CCACGGCTTCCTTTCTCGTA-3 (sense) and 5 -AGTCTCACTCACCTTTGCAG-3 (antisense) in a 50-µL reaction mix containing: a KCL reaction buffer (Bioline, UK), 100 µmol/l dntp, 0.6 µmol/l of each primer, 0.5 to 0.8 µg of genomic DNA, and2uoftaq polymerase (Bioline, UK) for 35 cycles at 93 C for 1 minute, 56.5 C for 1 minute, followed up by 72 C for 1 minute. After amplification, 10 µl of the 328-bp amplicon was digested with2uoftherestriction endonuclease Bst71I (isochizomer of Bbv 1) for 5 hours at 50 C. This restriction enzyme cleaves the

3 HEPATOLOGY Vol. 31, No. 1, 2000 AGARWAL ET AL. 51 polymerase chain reaction amplicon if the base G is present at position 49 (exon 1). Cleavage results in 2 digestion products of 244 bp and 84 bp. If A is present at position 49, cleavage does not occur and a single 328-bp product results. Three different CTLA-4 genotypes (GG, 84-bp 244-bp products; AG, 84-bp 244-bp 328-bp products; and AA, 328-bp product) were identified by electrophoresis in a 1.5% (wt/vol) agarose gel with ethidium bromide staining and ultraviolet transillumination. Control samples of predetermined genotype were included in each batch as well as negative controls (no DNA added). Statistical Analyses. Allele and genotype frequencies were compared using 2 and Fisher exact tests, as appropriate. All analyses were performed on a personal computer using the Epistat statistical software (Epi-info; CDC, Atlanta, GA). Clinical differences were sought using the unpaired t test to evaluate the significance of differences in means for continuous variables and the Fisher exact test for comparison of dichotomous variables. The Mann-Whitney test was used to compare nonparametric variables in independent samples. Because the variables for comparison had been formulated a priori and then assessed systematically in each study group, an unadjusted P value of.05 was used to determine statistical significance. Data are presented as the mean SE in tables and text. RESULTS Frequency of CTLA-4 Genotypes. The observed genotype frequencies for patients and controls were in Hardy-Weinberg equilibrium. Comparing patients and controls, the AA genotype occurred less frequently in patients with type 1 AIH than in normal subjects and the AG genotype occurred more commonly ( , P.013) (Table 1). This difference was caused by an over-representation of the G allele in the patient group. Indeed, 68% of the patients had the AG or GG genotype compared with 50% of the normal subjects ( , P.004; odds ratio 2.10; 95% CI, ) (Table 1). Comparing the gene frequencies, the CTLA-4 codon 17 Ala allele (G at position 49) was found on 126 of the 310 chromosomes in the patient population (41%) and only 64 of the 204 chromosomes from the control group (31%) ( , P.033). Associations of CTLA-4 Genotypes With Clinical Features. Patients with the GG genotype had significantly higher serum aspartate transaminase levels at presentation to the Mayo Clinic than patients with other genotypes (GG: vs. AA/AG: , P.03) and they also had a higher frequency of thyroid microsomal antibodies (GG: 57% vs. AA/AG: 34%, P.05) (Table 2). In the latter instance, the difference in occurrence was most marked in patients with the GG genotype compared to those of the AA genotype (57% vs. 20%, P.004) (Table 2). The frequency of thyroid disease, however, was similar among all genotypes (Table 2). TABLE 1. Frequency of CTLA-4 Alleles in Type 1 Autoimmune Hepatitis and Normal Subjects CTLA-4 Polymorphism Type 1 Autoimmune Hepatitis (N 155) Normal Subjects (N 102) AA 50 (32)* 51 (50)* AG 84 (54)* 38 (37)* GG 21 (14)* 13 (13)* AG or GG 105 (68) 51 (50) NOTE. Numbers in parentheses are percentages. Significantly different from each other at level of: * 2 (2 3, AA vs. AG vs. GG) P (2 2, AA vs. AG or GG) P.004. TABLE 2. Clinical Features at Presentation and CTLA-4 Polymorphisms Clinical Features AA (N 50) CTLA-4 Polymorphisms AG (N 84) GG (N 21) AA or GA (N 134) Age (yr) Gender (female: male) 37:13 73:11 17:4 110:24 Immune diseases 12 (24) 33 (39) 9 (43) 45 (34) AST (nl, 31 U/L) Bilirubin (nl, 1.1 mg/dl) Globulin (nl, g/dl) Immunoglobulin G (nl, 700-1,500 mg/dl) 2, , , , Thyroid disease 5 (10) 18 (21) 5 (24) 23 (17) Microsomal antibody 10/49 (20) 34/82 (41) 12 (57) # 44 (34)# Thyroglobulin antibody 3/49 (6) 13/82 (16) 3 (14) 16/131 (12) HLA DRB1* /49 (51) 37 (44) 16 (76) 62/133 (47) HLA DRB1*0301 / * (12) 9 (11) 3 (14) 15 (11) HLA DRB1* (40) 40 (48)# 5 (24)# 60 (45) A1-B8-DRB* (40) 27 (32) 12 (57) 47 (35) DRB1*0301 / DRB1* (42) 29 (34)* 15 (71)* 50 (37) DRB1*0401 / DRB1* (30) 32 (38) 4 (19) 47 (35) Cirrhosis at entry 10 (20) 21/81 (26) 5 (24) 31/131 (24) Abbreviation: AST, aspartate transaminase. Significantly different from each other at: *P.003. P.004. P.01. P.02. P.03. P.04. #P.05. When patients were stratified according to the presence or absence of HLA DRB1*0301, there was a striking correlation between possession of the GG genotype and DRB1*0301. Sixteen of 21 (76%) patients with the GG genotype had DRB1*0301 compared to 62 of 133 (47%) patients with the AG or AA genotypes (P.02) (Table 2). An inverse relationship between possession of the GG genotype and the HLA DRB1*0401 allele was also noted. Five of 21 patients with the GG genotype (24%) had DRB1*0401 compared to 60 of 134 patients (45%) with the AG or AA genotypes (Table 2). The difference in occurrence of DRB1*0401, however, was significant only between patients with the GG and AG genotypes (Table 2). Associations of CTLA-4 Genotypes With Treatment Outcome. CTLA-4 genotypes did not affect treatment outcome despite the association of the GG genotype with HLA DRB1*0301 and the AA genotype with HLA DRB1*0401 (Table 3). Immediate treatment outcomes, including the frequencies of remission, treatment failure and incomplete response, and long-term treatment outcomes, including the frequencies of progression to cirrhosis, death from liver failure, and requirement for liver transplantation, were similar in all genotypes (Table 3). DISCUSSION This population-based, case-control study investigated CTLA-4 gene polymorphisms in type 1 AIH. Our findings indicate that there is a significantly higher frequency of the G allele in these patients than in geographically and ethnically

4 52 AGARWAL ET AL. HEPATOLOGY January 2000 TABLE 3. CTLA-4 Polymorphisms and Treatment Outcome Treatment Outcome AA (N 47) CTLA-4 Polymorphisms AG (N 75) GG (N 19) AA or GA (N 122) Remission 32 (68) 53 (71) 10 (53) 85 (70) Sustained remission 9/32 (28) 15/53 (28) 2/10 (20) 24/85 (28) Relapse 22/32 (69) 39/53 (74) 8/10 (80) 61/85 (72) Treatment failure 7 (15) 7 (9) 4 (21) 14 (11) Incomplete response 8 (17) 15 (20) 5 (26) 23 (19) Developed cirrhosis 8/30 (27) 12/40 (30) 5/11 (45) 20/70 (29) Liver transplantation 5 (11) 2 (3) 1 (5) 7 (6) Death from hepatic failure 3 (6) 2 (3) 0 (0) 5 (4) Transplantation or hepatic death 8 (17) 4 (5) 1 (5) 12 (10) Death (total) 7 (15) 5 (7) 1 (5) 12 (10) Duration follow-up (mo) NOTE. Numbers in parentheses are percentages. matched control subjects. In addition, the GG genotype is associated with the HLA DRB1*0301 allele more commonly than other genotypes. Furthermore, patients with homozygosity for the G allele have higher serum levels of aspartate transaminase and are more likely to have antibodies to thyroid microsomal antigens. However, these latter findings, in particular those with respect to aspartate transaminase, were statistically borderline and their significance in this study is questionable. There are 2 possible explanations for these findings. First, the CTLA-4 exon 1 gene polymorphism may be functionally linked with susceptibility to type 1 AIH. Second, the CTLA-4 exon 1 gene polymorphism may be a linkage marker for an unidentified susceptibility gene for type 1 AIH on chromosome 2q. A pivotal role for CTLA-4 may be the maintenance of peripheral self-tolerance by modulation of T-cell immune responses. 16,17,28 Indeed, the functional importance of this molecule is emphasized by the development of lethal selfreactive lymphoproliferative disease in CTLA-4 deficient mice. 29 These studies indicate that CTLA-4 is a negative regulator of T-cell activation and that it is vital for the control of lymphocyte homeostasis. Blockade of the CTLA-4/B7 interaction promotes disease onset and accelerates disease progression in murine models of autoimmunity such as the T-cell receptor transgenic model of diabetes 30 and experimental allergic encephalomyelitis. 31 In the nonobese diabetic (NOD) transgenic model of diabetes, autoreactive T cells escape clonal deletion, contribute to the peripheral T-cell repertoire, and promote the development of diabetes several months later. 30 Treatment with anti-ctla-4 monoclonal antibodies induces diabetes rapidly, but only if treatment is given before the onset of insulitis. 30 CTLA-4 engagement per se, rather than a lack of costimulation, is required for the induction of anergy in vivo and the existence of autoreactive peripheral T-cell populations. 28,32 Antigen-specific autoreactive T-cell clones exist in type 1 AIH 15,33 and it is possible that functional abnormalities of the CTLA-4 T-cell interaction and costimulatory signal prevent termination of peripheral T-cell responses, which in turn, contribute to its immunopathogenesis. Despite strong circumstantial evidence linking the CTLA-4 gene with control of the immune response, the true functionality of the CTLA-4 position 49 polymorphism is uncertain. The Thr/Ala substitution in the leader peptide may have a subtle effect on posttranslational processing or localization of the CTLA-4 protein. The Thr/Ala polymorphism is also in strong linkage with a downstream (AT) n repeat, which may affect RNA stability as established previously in cytokine genes. 34 Full understanding of the basis for our observations must await functional definition of this polymorphism. Another explanation for our findings is that the exon 1 polymorphism of the CTLA-4 gene may simply be acting as a linkage marker for an unidentified susceptibility gene for type 1 AIH on chromosome 2q. The CTLA-4 gene is located in the 2q33 band in close proximity to the genes encoding the natural resistance macrophage protein 1 and genes for the interleukin 8 receptor. 16,17,35 The natural resistance macrophage protein 1 gene has been implicated in susceptibility to infectious disease, but no associations or polymorphisms with the interleukin 8 gene family have yet been described. The HLA DRB1*0301 allele is known to be associated with more severe disease in patients with type 1 AIH 6,8,36 though the apparent association between the CTLA-4 GG genotype and DRB1*301 may compound this effect. The CTLA-4 GG genotype, however, was not associated with a poorer treatment outcome than other genotypes and this finding suggests that the effect of the CTLA-4 GG genotype on immunopathogenesis is not pronounced and/or is suppressable with corticosteroid therapy. Importantly, there were only 21 GGpositive patients in our study and larger numbers are necessary to enhance statistical power. Furthermore, both DRB1*0301 and the GG genotype are associated with disease severity and the apparent synergy between the 2 genes may be false. Additional studies of this relationship are necessary. Genetic associations with polymorphisms of exon 1 of the CTLA-4 gene have been widely reported in other autoimmune diseases. 18,33-35,37-39 The 2q33 region of the human genome has been identified as IDDM12 in studies of insulindependent diabetes and the A to G substitution at position 49 has been associated with increased susceptibility to several autoimmune diseases including primary biliary cirrhosis, 37 primary sclerosing cholangitis, 38 as well as insulin-dependent diabetes mellitus, 18,35 and autoimmune thyroid disease. 39 Therefore, it is unlikely that CTLA-4 gene polymorphisms contribute to disease in an antigen-specific manner. Indeed this is no surprise because regulation of the immune response by CTLA-4 is not antigen specific. The clustering of autoimmune diseases around particular genes suggests that these genes predispose individuals to autoimmunity in general and explains, in part, the coincidence of disparate autoimmune diseases in some families and individuals. 40,41 Increasing evidence from genomic scanning and candidate gene approaches in autoimmune diseases suggests that identification of overlapping (common) susceptibility loci may provide valuable insight into the complex genetic interactions, which lead to autoimmune disease. The end stage phenotype of a specific autoimmune disease may be clinically distinct and/or organ specific with fundamental shared immunoregulatory processes affecting cytokine production, pro/antiinflammatory cytokine ratios, apoptosis, effector T-cell regulation, and antibody production. 40,41 Other factors such as environmental exposure or additional genetic loci may modulate disease or target tissue phenotype. Case-control candidate gene studies are frequently criticized and they have a number of recognized problems. Large family groupings with type 1 AIH, however, are not available for genetic studies and our investigation is based on a

5 HEPATOLOGY Vol. 31, No. 1, 2000 AGARWAL ET AL. 53 uniquely characterized and systematically evaluated homogeneous group of patients and ethnically similar control subjects. The power of such studies is directly dependent on the number of patients assessed and their relative value is based on the accuracy of diagnosis and the appropriateness of controls. Our study fully satisfies these requirements. In summary, our findings suggest that the CTLA-4 gene may be a non major histocompatibility complex susceptibility determinant in type 1 AIH. Association and linkage studies in several different populations 18-22,34 indicate that the CTLA-4 gene contributes to the autoimmune process in susceptible individuals and its effect is probably not disease specific. REFERENCES 1. Krawitt EL. Autoimmune hepatitis. N Engl J Med 1996;334: Homberg JC, Abauf N, Bernard O, Islam S, Alvarez F, Khalil SH, Poupon R, et al. Chronic active hepatitis associated with anti liver/kidney microsome antibody type 1: a second type of autoimmune hepatitis. HEPATOLOGY 1987;7: Czaja AJ, Manns MP. The validity and importance of subtypes of autoimmune hepatitis: a point of view. Am J Gastroenterol 1995;90: Heward J, Gough CL. Genetic susceptibility to the development of autoimmune disease. Clin Sci 1997;93: Donaldson PT, Doherty DG, Hayllar KM, McFarlane IG, Johnson PJ, Williams R. Susceptibility to autoimmune chronic active hepatitis: human leukocyte antigens DR4 and A1-B8-DR3 are independent risk factors. HEPATOLOGY 1991;13: Czaja AJ, Strettell MDJ, Thompson LJ, Santrach PJ, Moore SB, Donaldson PT, Williams R. Associations between alleles of the major histocompatibility complex and type 1 autoimmune hepatitis. HEPATOLOGY 1997;25: Doherty DG, Donaldson PT, Underhill JA, Farrant JM, Duthie A, Mieli-Vergani G, McFarlane IG, et al. Allelic sequence variation in the HLA class II genes and proteins in patients with autoimmune hepatitis. HEPATOLOGY 1994;19: Strettell MDJ, Donaldson PT, Thompson LJ, Santrach PJ, Moore SB, Czaja AJ, Williams R. Allelic basis for HLA-encoded susceptibility to type 1 autoimmune hepatitis. Gastroenterology 1997;112: Manabe K, Hibberd ML, Donaldson PT, Underhill JA, Doherty DG, Demaine AG, Mieli-Vergani G, et al. T-cell receptor constant germline polymorphisms and susceptibility to autoimmune hepatitis. Gastroenterology 1994;106: Arenz M, Meyer zum Buschenfelde K-H, Lohr HF. Limited T cell receptor V -chain repertoire of liver-infiltrating T cells in autoimmune hepatitis. J Hepatol 1998;28: Whittingham S, Mathews JD, Schanfield MS, Tait BD, Mackay IR. Interaction of HLA and Gm in autoimmune chronic active hepatitis. Clin Exp Immunol 1981;43: Cookson S, Constantini PK, Clare A, Underhill JA, Bernal W, Czaja AJ, Donaldson PT. Frequency and nature of cytokine gene polymorphisms in type 1 autoimmune hepatitis. HEPATOLOGY 1999;30: Czaja AJ, Cookson S, Constantini PK, Clare M, Underhill JA, Donaldson PT. Cytokine polymorphisms associated with clinical features and treatment outcome in type 1 autoimmune hepatitis. Gastroenterology 1999;117: Wen L, Peakman M, Lobo-Yeo A, McFarlane BM, Mowat AP, Mieli- Vergani G, Vergani D. T-cell directed hepatocyte damage in autoimmune chronic active hepatitis. Lancet 1990;336: Lohr H, Manns M, Kyriatsoulis A, Lohse AW, Trautwein C, Meyer zum Buschenfelde K-H, Fleischer B. Clonal analysis of liver-infiltrating T cells in patients with LKM-1 antibody-positive autoimmune chronic active hepatitis. Clin Exp Immunol 1991;84: Bluestone JA. Is CTLA-4 a master switch for peripheral T cell tolerance? J Immunol 1997;158: Thompson CB, Allison JP. The emerging role of CTLA-4 as an immune attenuator. Immunity 1997;7: Marron MP, Raffel LJ, Garchon HJ, Jacob CO, Serrano-Rios M, Larrad MTM, Teng WP, et al. Insulin-dependent diabetes mellitus (IDDM) is associated with CTLA-4 polymorphisms in multiple ethnic groups. Hum Mol Genet 1997;6: Johnson PJ, McFarlane IG, Alvarez F, Bianchi FB, Bianchi L, Burroughs A, Chapman R, et al. Meeting Report. International Autoimmune Hepatitis Group. HEPATOLOGY 1993;18: Czaja AJ, Cassani F, Cataleta M, Valentini P, Bianchi FB. Frequency and significance of antibodies to actin in type 1 autoimmune hepatitis. HEPATOLOGY 1996;24: Czaja AJ, Nishioka M, Morshed SA, Hachiya T. Patterns of nuclear immunofluorescence and reactivities to recombinant nuclear antigens in autoimmune hepatitis. Gastroenterology 1994;107: Czaja AJ, Cassani F, Cataleta M, Valentini P, Bianchi FB. Antinuclear antibodies and patterns of nuclear immunofluorescence in type 1 autoimmune hepatitis. Dig Dis Sci 1997;42: Czaja AJ, Manns MP, Homburger HA. Frequency and significance of antibodies to liver/kidney microsome type 1 in adults with chronic active hepatitis. Gastroenterology 1992;103: Fulthorpe AJ, Roitt IM, Doniach D, Couchman K. A stable sheep cell preparation for detecting thyroglobulin antibodies and its clinical application. J Clin Pathol 1961;14: Doniach D. Thyroid autoimmune disease. J Clin Pathol 1967;20: Bird T, Stephenson J. Evaluation of a tanned red cell technique for thyroid microsomal antibodies. J Clin Pathol 1973;26: Czaja AJ, Carpenter HA. Sensitivity, specificity and predictability of biopsy interpretations in chronic hepatitis. Gastroenterology 1993;105: McCoy KD, Le Gros G. The role of CTLA-4 in the regulation of T cell immune responses. Immunol Cell Biol 1999;77: Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, Lee KP, Thompson CB, et al. Lymphoproliferative disorders with early lethality in mice deficient in CTLA-4. Science 1995;270: Luhder F, Hoglund P, Allison JP, Benoist C, Mathis D. Cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) regulates the unfolding of autoimmune diabetes. J Exp Med 1998;187: Perrin PJ, Maldonado JH, Davis TA, June CH, Racke MK. CTLA-4 blockade enhances clinical disease and cytokine production during experimental allergic encephalomyelitis. J Immunol 1996;157: Perez VL, Parijs LV, Biuckians A, Zheng XX, Strom TB, Abbas AK. Induction of peripheral T cell tolerance in vivo requires CTLA-4 engagement. Immunity 1997;6: Lohr HF, Schaak JF, Gergen G, Fleischer B, Dienes H-P, Meyer zum Buschenfelde K-H. Phenotypical analysis and cytokine release of liverinfiltrating and peripheral blood T lymphocytes from patients with chronic active hepatitis of different etiology. Liver 1994;14: Shaw G, Kamen R. A conserved AU sequence from the 3 untranslated of GM-CSF mrna mediates selective mrna degradation. Cell 1986;46: Nistico L, Buzzeti R, Pritchard LE, Van der Auwera B, Giovannini C, Bosi E, Larrad MTM. The CTLA-4 region of chromosome 2q33 is linked to, and associated with, type 1 diabetes. Hum Mol Genet 1996;5: Czaja AJ, Carpenter HA, Santrach PJ, Moore SB. Significance of HLA DR4 in type 1 autoimmune hepatitis. Gastroenterology 1993;105: Agarwal K, Grove J, Daly AK, James OFW, SHS Pearce DEJ Jones, Bassendine MF. CTLA-4 gene polymorphism confers susceptibility to primary biliary cirrhosis [Abstract]. HEPATOLOGY 1998;28:404A. 38. Saarinen A, Broome U, Liger A, Olsson R, Hulcrantz R, Olerup O. CTLA-4 gene polymorphism associated with primary sclerosing cholangitis (PSC). J Hepatol 1998;28(Suppl): Donner H, Rau H, Walfish PG, Braun J, Siegmund T, Finke R, Herwig J, et al. CTLA-4 alanine-17 confers genetic susceptiblity to Graves Disease and to type 1 diabetes mellitus. J Clin Endocrinol Metab 1997;82: Becker KG, Simon RM, Bailey-Wilson JE, Freidlin B, Biddison WE, McFarland HF, Trent JM. Clustering of non-major histocompatibility complex susceptibility loci in human autoimmune diseases. Proc Natl Acad Sci U S A1998;95: Becker KG. Comparative genetics of type 1 diabetes and autoimmune disease. Diabetes 1999;48: