Autoimmunity Reviews 2 (2003) 322–331

Autoimmunity and hepatitis C Christian P. Strassburg*,1, Arndt Vogel, Michael P. Manns Department of Gastroenterology, Hepatology and Endocrinology, Hannover Medical School, Carl-Neuberg-Str. 1, D-30625 Hannover, Germany Accepted 24 February 2003

Abstract Hepatitis C is a widespread chronic liver disease leading to cirrhosis and to the complications of portal hypertension. Based on biochemical and clinical features, it is almost indistinguishable from autoimmune hepatitis, which is characterized by the absence of viral infection, and other causes of chronic liver diseases, and represents a classical autoimmune disease with loss of immunological tolerance of liver tissue. Although the differentiation between both diseases is not difficult due the availability of diagnostic viral markers, it is well recognized that not only are autoantibodies present in autoimmune hepatitis frequently detected in hepatitis C, but also that an array of immune-mediated symptoms and diseases occur in patients with chronic hepatitis C. This has prompted research aimed at identifying a link between hepatitis C and autoimmunity, and autoimmune hepatitis in particular. This review focuses on the general immunological mechanisms linking viral infections with autoimmunity and includes the specific features of hepatitis C- and D-associated autoimmunity. Virus infection remains at the center of molecular and cellular research aimed at identifying the forces driving human autoimmunity and autoimmune diseases. 䊚 2003 Elsevier Science B.V. All rights reserved. Keywords: Mimicry; Autoimmunity; Liver kidney microsomal antibodies; Animal models; Bystander activation; Hepatitis C; Immune-mediated disease

1. Introduction Autoimmunity is a pathophysiological mechanism that leads to chronic inflammatory diseases of a broad array of organs, which is manifest in diseases such as diabetes mellitus, adrenal insuf*Corresponding author. Tel.: q49-511-532-2853; fax: q49511-532-2093. E-mail address: [email protected] (C.P. Strassburg). 1 C.P.S. is a Heisenberg Scholar of the Deutsche Forschungsgemeinschaft (DFG).

ficiency, thyroiditis, multiple sclerosis and autoimmune hepatitis. By definition, autoimmunity is brought about by features of the immune system that are globally characterized as a loss of tolerance towards self-antigens. It is not a viral infection and the diagnosis of autoimmune diseases therefore centers around serological markers of autoimmunity, such as elevated gamma globulins and autoantibodies, and an exclusion of signs of viral or other infectious etiologies. In hepatology, two chronic inflammatory conditions share a number of features: autoimmune hepatitis and chronic viral hepatitis C. In both diseases, portal inflammation

1568-9972/03/$ - see front matter 䊚 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S1568-9972(03)00036-3

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

is present that is frequently indistinguishable by histological examination, aminotransferases are elevated and the typical sequelae of chronic hepatic disease, including cirrhosis and portal hypertension, develop. In early reports, liverykidney microsomal autoantibodies (LKM) directed against cytochrome P450 proteins were detected in chronic hepatitis C, which initiated speculation regarding an etiological parallel of both diseases. It has been well described in numerous studies that LKM autoantibodies are present in up to 5% of patients with hepatitis C, and that hepatitis C is associated with other immune-mediated diseases, which include autoimmune thyroid disease, mixed cryoglobulinemia and rheumatoid arthritis. Although this association is striking and exceeds those found with hepatitis B virus infection, no serious proof has been provided that hepatitis C is a cause of autoimmune hepatitis. The exact molecular, cellular and physiological mechanisms leading to autoimmunity are still incompletely elucidated. Human autoimmunity has a slow onset, which hinders the identification of the offending agent or mechanisms in most cases. It is generally assumed that the causes are multifactorial, including both genetic and permissive environmental triggers. Infectious agents have often been named as triggers of autoimmune diseases. One of the problems is the discrimination as to whether abnormal immunity is in fact genuinely directed against self rather than representing a response to infectious or infection-modified nonself. Hepatitis C (and hepatitis B as well as D) virus infection requires precise differentiation from autoimmune hepatitis in view of the therapeutic strategy employed: interferon and antivirals in viral hepatitis and immunosuppression in autoimmune hepatitis. If the wrong treatment strategy is employed, immunosuppression is likely to propagate replication, whereas interferon-a cannot only induce autoimmunity in general, but has also been shown to exacerbate autoimmune hepatitis. For the consideration of autoimmunity in hepatitis C virus infection, basic principles of autoimmunity have to be taken into account. Every organism is characterized by the potential of active autoreactive cells that can induce autoimmunity. The connection between autoimmunity and infec-

323

tion is two-fold: molecular mimicry can act as a direct mechanism for inducing autoimmunity, and the indirect activation of autoreactive B- and Tlymphocytes in the event of an infection can result in autoimmunity. Much has been learned regarding these mechanisms in animal models, which may or may not represent the situation present in humans. In this review the basic mechanisms regarding virus infection and autoimmunity are discussed and the particular association of autoimmunity in one of the most common chronic diseases of the liver, hepatitis C, is reviewed. One of the most attractive hypotheses on autoimmunity involves mimicry, which has led to extensive investigation of serologically defined epitopes and viral proteins. 2. Induction of autoimmunity by mimicry Protection against autoreactive T- and B-cells is achieved early in life by negative clonal selection and apoptosis in the thymus. In addition, the peripheral induction of tolerance by clonal deletion, induction of anergy, or ignorance serves to suppress autoimmunity w1x. Not all cells are deleted when central and peripheral tolerance is established, either because the antigen is inaccessible or because the antigen level is too low to lead to activation w1x. One mechanism to overcome tolerance is molecular mimicry (Fig. 1). It is regarded as a controversial hypothesis. During the last two decades a considerable body of publications has appeared focusing on epitope mimicry. This has provided the main arguments for this hypothesis, while failing to provide conclusive evidence thus far. According to the hypothesis a susceptible individual acquires an infection with an agent, which exhibits antigens that are immunologically similar to the host antigens, but nevertheless have the ability to induce an immune response when presented to T-cells. Either linear amino acid sequences of the molecules or their conformational epitopes can be shared. As a result, the immune response that is generated cross-reacts with host tissue antigens, which can potentially lead to tissue destruction. This autoimmune attack may progressively spread to other self-antigens. The foreign triggering antigen may or may not be present when

324

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

Fig. 1. Schematic and simplified presentation of humoral and cellular mechanisms postulated in virus-induced autoimmunity involving molecular mimicry. Viral antigens can be presented by professional antigen-presenting cells (APC) on class I and II major histocompatibility antigens when co-stimulated by CD4 and CD8 positive cells. As a consequence of molecular mimicry, CD8q cells can mount a direct attack on the hepatocyte, which is additionally stimulated by CD4q cells. These activate B-cells, which can produce antibody and, after expansion, can process autoantigens (gray arrow), which again can lead to the generation of autoreactive T-cells (curved gray arrow) when the appropriate co-stimulatory mediators are present. Note that there is no evidence of direct cytotoxicity by LKM-1 autoantibodies detected both in autoimmune hepatitis type 1 and hepatitis C virus (HCV) infection. However, autoantibodies have been shown to elicit direct cytotoxicity.

the ensuing autoimmune disease becomes clinically apparent. Cross-reactive immune responses can either develop on the humoral (antibody) or T-cellmediated level or on both. Cytotoxic CD8 (q) Tcells are the main effector cells in most autoimmune diseases. In addition, regulatory CD4 (q) T-cells are believed to play a crucial role in the control of pathological autoimmunity. Another level of recognition occurs at the peptide–B-cell interaction. B-cells display a variety of character-

istics that can contribute to autoimmunity. Autoantibodies are the hallmark of many autoimmune diseases w2x and may themselves initiate autoim¨ B-cells have been munity w3x. In addition, naıve shown to be able to differentiate into polarized Bcells with individual cytokine profiles and behave as classical antigen-presenting cells potentially capable of regulating the profile of an immune response, which then may initiate autoimmunity. In mouse models of diabetes mellitus and systemic lupus erythematosus, antibody production is not

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

an absolute requirement for the manifestation of autoimmune disease. This has led to the suggestion that B-cells may act as important antigen-presenting cells for autoreactive T-cells. In a first step, professional antigen-presenting cells present either a molecular mimic antigen or a cryptic self-antigen in the context of appropriate costimulatory molecules to T-cells, which in turn can activate the corresponding B-cells. After proliferation, they were able to process the autoantigen and present the novel-self peptides with costimulatory molecules, such as B7-2 to T-cells, resulting in a diverse, clonally expanded repertoire of auto-reactive T-cells w4x. This mechanism is limited by the fact that the supporting data involve primed Tcells and that B-cells most likely cannot activate ¨ antigen-specific T-cells. naıve 3. The role of autoepitopes in autoimmunity Molecular mimicry is present when immunorelevant cells can be activated by cross-reacting determinants. This can be the case in the absence of amino acid sequences identical to self-proteins that can activate auto-reactive T-cells. Some activating peptides do not share a single amino acid with the original peptide, which indicates that the detection of mimicry may not be easy and a prediction of critical epitopes may be complex. Along these lines, T-cells specific for myelin basic protein from patients with multiple sclerosis have been shown to recognize distinct but structurally related peptides from several bacterial and viral proteins. These would not have been identified by sequence alignment alone w5x. In addition, B-cell responses in autoimmunity can diversify to multiple sites of the original autoantigen as the disease progresses. A classical example of humoral autoimmunity in humans is systemic lupus erythematosus, in which an array of autoantibodies is found targeting intracellular antigens. Molecular mimicry with Epstein–Barr virus has been proposed for initiation of the disease. An early B-cell antibody is directed against a specific region of the snRNP autoantigen (amino acid sequence identity: PPPGMRPPP). Immunization of animals with this peptide or with a closely related sequence found in the Epstein–

325

Barr nuclear antigen-1 induces a lupus-like autoimmune syndrome. Further studies demonstrated that the highly diverse humoral autoimmune response in SLE can originate from a single protein, either from the molecular mimic discussed or other cryptic self-peptides w6,7x. This ability of the immune response to diversify to many sites on an antigen (epitope spreading) has probably evolved as an effective mechanism to clear foreign antigens, but may result in the amplification of the immune pathology in the context of an autoimmune reaction. With regard to chronic hepatitis C virus infection, it is noteworthy that LKM autoantibodies in HCV infection have been found to react with multiple targets of differing molecular masses, while in autoimmune hepatitis this response appears more focused. Whether this indicates a differing mechanism based on mimicry or the absence of mimicry remains speculative. 4. Molecular mimicry in animals It has been shown in transgenic animal models by molecular methodology that peripheral T-cells are capable of inducing autoimmunity. There are only a few non-transgenic animal models that provide conclusive proof of molecular mimicry as a mechanism capable of explaining autoimmunity. One of these involves the UL6 corneal antigen and herpetic stromal keratitis. Herpes stromal keratitis is a T-cell-mediated autoimmune disease of the eye triggered by herpes simplex virus type 1 w8x. A similar keratitis can be induced in some strains of mice when ocularly infected with HSV1. Mice that are resistant to HSV-1 infection express a specific Ig2ab isotype, which renders them immunologically tolerant towards the autoantigen w9x. The antigen of the Ig2a class antibody contains a 16-amino-acid peptide sequence with similarities to a protein expressed in corneal cells, which has not yet been identified but is the presumed autoantigen. A search of the GenBank database for HSV-1 proteins revealed a peptide of almost identical amino acid sequence in the UL6 protein, which has consequently been suggested to represent the molecular mimic capable of eliciting the HSK syndrome. In support of the corneal autoantigen hypothesis is the observation that viral

326

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

mutants that lack the mimicking peptide fail to induce HSK in susceptible mouse strains, and that keratitis can be adoptively transferred to virusinfected nude mice with UL6 and IgG2a immune T-cells. Although these findings may be the strongest yet developed in support of epitope mimicry, this is not undisputed. T-cell receptor transgenic mice can develop severe keratitis upon HSV-1 infection, despite the fact that they were unable to generate detectable immune responses to HSV antigens. The mechanism by which keratitis occurred in such mice was assumed to involve non-TCR-mediated bystander activation of CD4q T-cells in the cornea w10x. 5. Molecular mimicry in humans A classical example of epitope mimicry in man is rheumatic fewer following streptococcal pharyngitis. There are, however, only few, if any, examples in humans in which molecular mimicry give rise to classical autoimmune disease. Epidemiological data collected for many years have suggested an association between viruses and type 1 diabetes mellitus. Type 1 diabetes mellitus is characterized by a T-cell-mediated autoimmune destruction of pancreatic islet cells. Several reports in humans and animals indicated a link between Coxsackie virus infection and IDDM. The virus has been isolated from pancreatic tissues of patients with recent onset of IDDM, and immunization of mice with virus proteins or specific peptides induced cellular b-cell autoimmunity and diabetes. A sequence in GAD65 (amino acid sequence identity: PEVKEK) has a striking homology with an amino acid sequence of p2-C, which is highly conserved in Coxsackie virus B4, and molecular mimicry has been proposed between these two enzymes w11x. Two peptides p17 and p18 of GAD65 that share sequence similarity with Coxsackie virus (PEVKEK) appeared to be the major determinants of GAD65 preferentially recognized by T-cells from new-onset IDDM patients and their first-degree relatives w12x. In addition, Lonnrot et al. demonstrated that the epitopes in p2-C and GAD65 are antigenically highly similar, and that peptide immunization induces antibodies, which cross-react with these molecules w13x. How-

ever these various reports demonstrating a strong similarity between both epitopes do not prove that molecular mimicry is the cause of IDDM. In addition, another report has suggested that diabetes induced by Coxsackie virus is not due to molecular mimicry, but to bystander activation. Studies failed to demonstrate a significant accelation of IDDM in NOD mice after infection with Coxsackie virus B4. In contrast, mice with a transgenic diabetogenic T-cell receptor, which did not cross-react with the virus, rapidly developed diabetes following infection. This demonstrates that virally induced diabetes was due to bystander effects rather than molecular mimicry. Albeit representing an intriguing explanation, molecular mimicry is far from proven. 6. The role of bystander effects on autoimmunity In the process of an infection, formerly quiescent autoreactive T- and B-cells may become activated. The breakdown of tolerance may be the result of several mechanisms. Self-antigens are normally not accessible to the immune system, but become immunologically visible upon the initiation of inflammation. During infection, tissue damage and necrosis may uncover cryptic autoantigens, which are then presented at the site of infection or in draining lymph nodes w14x. In addition, in the course of an infection, antigen-presenting cells may upregulate MHC molecules and appropriate costimulatory molecules that are required for the activation of autoreactive cells. One theory is that this is the result of a local expression of proinflammatory cytokines, which in turn leads to the dysregulation of immunoreactivity in specific target tissues of genetically predisposed individuals. Thus, common viral infections with a tropism for a given tissue could produce organ-specific autoimmune diseases. 7. Autoimmune hepatitis and viruses Autoimmune hepatitis (AIH) is a chronic inflammatory disease of the liver characterized by

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

a loss of tolerance towards the hepatocellular epithelium, the cause of which still remains unknown. AIH is a chronic, mainly periportal hepatitis associated with hypergammaglobulinemia and circulating autoantibodies, which, in most cases, responds to immunosuppressive treatment. The disease is most prevalent among females, displays an immunogenetic association with the HLA A1-B8-DR3 or DR4 haplotype, and the presence of extrahepatic syndromes. The multifactorial definition of AIH illustrates that a single cause or a single test for the accurate diagnosis of AIH is not available. Therefore, the diagnosis is based upon a number of diagnostic criteria, as recently defined and modified by the International Autoimmune Hepatitis Group, and upon the exclusion of other causes of chronic hepatitis w15x. The combination of seroimmunological and molecular biological tests w2x helps in precise discrimination between autoimmune hepatitis and other causes of chronic hepatitis. The exclusion of replicating hepatitis virus infection, and the presence of female sex, hypergammaglobulinemia and response to immunosuppressive treatment are key components in an accurate diagnosis. Autoantibodies are a diagnostic hallmark of AIH detectable by serological in vitro assays. Autoantibody profiles have been utilized as a means of subclassification of AIH. According to this approach, AIH type 1 is characterized by the presence of antinuclear antibodies (ANA) andyor anti smooth muscle antibodies (SMA) directed predominantly against smooth muscle actin. AIH type 2 is characterized by anti liver–kidney microsomal autoantibodies (LKM-1) directed against cytochrome P450 (CYP) 2D6 w16x and with lower frequency against UDP-glucuronosyltransferases (UGT) w17x. AIH type 3 is characterized by autoantibodies against soluble liver antigens (SLAyLP) w18,19x. The proposed pathogenic concept of AIH is that a genetically predisposed host is exposed to an environmental agent, which triggers an autoimmune process against the liver epithelium. The presumed environmental factors are so far unknown. Various viruses have long been assumed to be causative agents of autoimmune hepatitis. However, for most viruses no conclusive concept

327

has been demonstrated. Some reports have suggested that measles or Epstein–Barr virus infection could trigger the autoimmune process; others have implicated hepatitis viruses as initiators of disease. In one instance, molecular mimicry was assumed. It was shown that the B-cell epitope of cytochrome P450, which is targeted by LKM-1 autoantibodies in AIH-2, displays homology with the immediate early antigen IE 175 of herpes simplex virus. One case has been reported, in which the only differences in HLA identical twins with discordant manifestation of AIH was exposure to HSV w20x. However, the above-discussed criteria for molecular mimicry are fulfilled for none of these viruses. 8. Autoimmunity and hepatitis C infection The discovery of HCV infection in patients with autoimmune diseases was a finding that prompted a search for a viral etiology of autoimmune hepatitis. However, refinement of testing systems has lead to the realization that HCV infection is not a likely cause of autoimmune hepatitis and that the initially observed coexistence was highly exaggerated. Nevertheless, HCV infection gives rise to a plethora of autoimmune or immune-mediated phenomena, which is why HCV infection is studied as a contributor to autoimmune diseases. In the initial reports analyzing LKM autoantibodies in HCV patients, the prevalence was found to reach 73–88% in Italian patients, 49% and 48% in French and German patients, respectively, and none in patients from the UK w21x. This finding demonstrated apparent geographic differences that can be explained by the geographic incidence of HCV infection, which is highest in Mediterranean countries. Therefore, the prevalence of LKM autoantibodies in patients with markers of chronic HCV infection was studied and showed only a prevalence of 3–5% w21x with apparently the lowest rate (or absence) of LKM autoantibodies in HCV patients of North American origin. These data indicate the limited usefulness of epidemiological evaluations to assess the role of HCV in hepatic autoimmunity. On the other hand, clinical data suggest that AIH and HCV-associated autoimmunity (LKM) differed in their presentation. Patients suffering from autoimmune hepatitis type

328

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

2 are usually young and of female sex (90%). They have a high biochemical ALT activity, a high titer of LKM-1 autoantibodies, usually a prompt response to immunosuppression and a dramatic exacerbation of disease activity upon interferon treatment. Extrahepatic manifestations of autoimmune disease are usually present. Conversely, HCV RNA-positive patients with LKM autoantibodies are older, with no predominance of the female sex; they have mild elevation of ALT levels, lower titers of LKM autoantibodies and the association with HLA DR3 or extrahepatic autoimmune syndromes is not as striking.. Other serological and clinical manifestations of HCV-associated autoimmunity and autoimmune diseases were subsequently reported. They include lymphocytic sialadenitis, similar but not identical ¨ to Sjogren’s syndrome, lichen planus, polyarteriitis nodosa, glomerulonephritis, thyroiditis, rheumatoid factor, mixed cryoglobulinemia, sporadic porphyria cutanea tarda and GOR antibodies (reviewed in w21x). In some of these associations, a causative role of the hepatitis C virus is suggested. Membranoproliferative glomerulonephritis and mixed cryoglobulinemia, for example, improve on administration of interferon alpha, the antiviralyimmunomodulatory effects of which are capable of eliminating the HCV virus w22x. Moreover, cryoprecipitates in cryoglobulinemia have been found to contain HCV core proteins and HCV RNA w23x. It is still controversial as to whether AIH associated with HCV infection exists. Although evidence that may suggest an involvement of the HCV virus has been reported w24x, it remains difficult to distinguish between the activation of latent autoimmune hepatitis (host factors) through the HCV virus, or even through interferon therapy for viral infection w25x, and the de novo induction of AIH by virtue of the virus (virus factor). Evidence against a genotypic determination of the induction of LKM autoantibodies by the HCV has been provided by Michitaka et al. w26x. On the other hand, a model of molecular mimicry appears possible in view of the great variability of the HCV virus. Epidemiological and clinical work-up have established the association of HCV and autoimmunity beyond any doubt, but have failed to

pinpoint valid differentiation criteria between virus-associated autoimmunity and genuine autoimmune disease. Studies have focused on differences in the immune response in both disease entities. LKM autoantibodies are being intensely studied to contribute to this controversial issue. They appear to be the result of differential immunological processes in AIH and HCV-associated autoimmunity. Initially, a mimicry effect was suggested by the identification of sequence similarities between the HCV genome and cytochrome P450 2D6, the target of LKM-1 autoantibodies w16x. Sequence similarities were also observed between the HSV IE175 protein and cytochrome P450 2D6. Molecular immunological evaluations have confirmed that LKM antibodies in HCV infection recognize cytochrome P450 2D6, but in almost 50% of cases are directed against other microsomal antigens of molecular mass of 50, 59 and 70 kDa w27x (Table 1). In 50% of sera, LKM autoantibodies can be detected by immunofluorescence and ELISA, but not by Western blot, which demonstrates the heterogeneity of LKM autoantibodies and the presence of autoantibodies directed against conformational epitopes in these patient sera. In AIH type 2, the majority of LKM-1 autoantibodies are directed against linear epitopes of cytochrome P450 2D6. The presence of conformational epitopes in some of these sera suggests that the immune response is antigen-driven. In view of this and the fact that to date specific virus factors (i.e. genotypes) have not been identified to play a significant role, host factors appear to be major players in the appearance of autoimmunity associated with viral infection. 9. Hepatitis D infection and autoimmunity The hepatitis D virus, an RNA viroid dependent on hepatitis B co-infection, was discovered in 1977 w28x. It was recognized soon after that in 13% of Italian patients suffering from HDV LKM autoantibodies were detectable that differed from LKM-1 autoantibodies in autoimmune hepatitis type 2 directed against cytochrome P450 2D6, and LKM-2 autoantibodies in dihydralazine-induced hepatitis directed against cytochrome P450 2C9 by their distinct immunofluorescence pattern on rat

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

329

Table 1 Heterogeneity of autoantibodies with autoantigen targets in the endoplasmic reticulum (microsomes) Antibody

MW (kDa)

Target antigen

Diseaseyassociation

LKM-1 LKM-2 LKM-3

50 50 55

Cytochrome P450 2D6 Cytochrome P450 2C9 UGT1A

LKM

50

Cytochrome P450 2A6

LM

52

Cytochrome P450 1A2

57 59 35 59 64 70

Disulfidisomerase Carboxylesterase ? ? ? ?

Autoimmune hepatitis type 2, hepatitis C Ticrynafen-induced hepatitis Hepatitis D-associated autoimmunity, autoimmune hepatitis type 2 Autoimmune polyendrocrine syndrome type 1 (APS-1yAPECED), hepatitis C Dihydralazine-induced hepatitis, hepatitis with APS-1yAPECED Halothane hepatitis Halothane hepatitis Autoimmune hepatitis Chronic hepatitis C Autoimmune hepatitis Chronic hepatitis C

LKM, liver–kidney microsomal autoantibodies; MW, molecular weight; UGT1A, family of 1-uridine-diphosphate-59glucuronosyltransferases.

liver–kidney cryostat sections. These autoantibodies were termed LKM-3. Similar to HCV infection, a multitude of serum autoantibodies was subsequently identified in sera of patients suffering from chronic HDV infection. These autoantibodies were identified by immunofluorescence and directed against basal cell layers (BCLA), thymic stellate epithelial cells (SECA), thymic reticular cells (TRA) and perithymocytic cells (PTA) w29,30x. Later, autoantibodies against nuclear laminin C were described w31x. In 1994, the molecular identity of the molecular target of LKM-3 autoantibodies, family 1 UDP glucuronosyltransferases (UGT-1), an enzyme family responsible for phasetwo drug metabolism, was described w32,33x. This discovery now allows for the characterization of another form of virus-associated autoimmunity. Interestingly, the original report described LKM-3 autoantibodies in one patient with autoimmune hepatitis type 2. This finding has since been confirmed and LKM-3 autoantibodies can be regarded as rare markers of AIH w17x. As in HCVassociated LKM autoantibodies, LKM-3 autoantibodies in HDV infection are present at lower titers than in AIH. In addition, autoepitope differences appear to exist between the two disease groups, with evidence of conformational epitopes w17x. LKM-3 autoantibodies appear to be specific for

HDV and do not occur in HCV infection. These preliminary data indicate that the immune response characterized by the molecular analysis of LKM-3 autoantibodies may differ in HDV-associated autoimmunity and AIH; this is subject to intense investigation. Further studies are needed to evaluate the clinical and pathophysiological significance of LKM-3 autoantibodies. Take-home messages ● Virus infections have been shown to serve as triggers of autoimmunity in animal models. ● Mimicry effects are the basis for an attractive hypothesis of breaking tolerance to induce autoimmunity. Conclusive evidence of this mechanism for human autoimmunity has not yet been provided. ● Hepatitis C and autoimmune hepatitis share LKM autoantibodies that differ regarding epitope specificity and diversity of recognized autoantigens. ● No conclusive evidence has been provided linking hepatitis C to the etiology of autoimmune hepatitis.

330

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331

● A broad array of immune-mediated diseases and symptoms are prevalent in hepatitis C infection. ● Distinction between hepatitis C and autoimmune hepatitis is important because therapeutic strategies are mutually exclusive: interferon and antivirals in hepatitis C, and immunosuppression in autoimmune hepatitis.

w14x

w15x

w16x

References w1x Goodnow CC. Balancing immunity and tolerance: deleting and tuning lymphocyte repertoires. Proc Natl Acad Sci USA 1996;93:2264 –71. w2x Strassburg CP, Obermayer-Straub P, Manns MP. Autoimmunity in liver diseases. Clin Rev Allergy Immunol 2000;18:127 –39. w3x Matsumoto I, Staub A, Benoist C, Mathis D. Arthritis provoked by linked T- and B-cell recognition of a glycolytic enzyme. Science 1999;286:1732 –5. w4x Mamula MJ. Epitope spreading: the role of self peptides and autoantigen processing by B-lymphocytes. Immunol Rev 1998;164:231 –9. w5x Wucherpfennig KW, Strominger JL. Molecular mimicry in T-cell-mediated autoimmunity: viral peptides activate human T-cell clones specific for myelin basic protein. Cell 1995;80:695 –705. w6x Topfer F, Gordon T, McCluskey J. Intra- and intermolecular spreading of autoimmunity involving the nuclear self-antigens La (SS-B) and Ro (SS-A). Proc Natl Acad Sci USA 1995;92:875 –9. w7x Bockenstedt LK, Gee RJ, Mamula MJ. Self-peptides in the initiation of lupus autoimmunity. J Immunol 1995;154:3516 –24. w8x Streilein JW, Dana MR, Ksander BR. Immunity causing blindness: five different paths to herpes stromal keratitis. Immunol Today 1997;18:443 –9. w9x Avery AC, Zhao ZS, Rodriguez A, et al. Resistance to herpes stromal keratitis conferred by an IgG2a-derived peptide. Nature 1995;376:431 –4. w10x Deshpande S, Zheng M, Lee S, et al. Bystander activation involving T-lymphocytes in herpetic stromal keratitis. J Immunol 2001;167:2902 –10. w11x Kaufman DL, Erlander MG, Clare-Salzler M, Atkinson MA, Maclaren NK, Tobin AJ. Autoimmunity to two forms of glutamate decarboxylase in insulin-dependent diabetes mellitus. J Clin Invest 1992;89:283 –92. w12x Atkinson MA, Bowman MA, Campbell L, Darrow BL, Kaufman DL, Maclaren NK. Cellular immunity to a determinant common to glutamate decarboxylase and Coxsackie virus in insulin-dependent diabetes. J Clin Invest 1994;94:2125 –9. w13x Lonnrot M, Hyoty H, Knip M, et al. Antibody crossreactivity induced by the homologous regions in glutamic acid decarboxylase (GAD65) and 2C protein of

w17x

w18x

w19x

w20x

w21x w22x

w23x

w24x

w25x

w26x

w27x

w28x

Coxsackie virus B4. Childhood Diabetes in Finland Study Group. Clin Exp Immunol 1996;104:398 –405. Behar SM, Porcelli SA. Mechanisms of autoimmune disease induction. The role of the immune response to microbial pathogens. Arthritis Rheum 1995;38:458 –76. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999;31:929 –38. Manns MP, Griffin KJ, Sullivan KF, Johnson EF. LKM1 autoantibodies recognize a short linear sequence in P450IID6, a cytochrome P-450 monooxygenase. J Clin Invest 1991;88:1370 –8. Strassburg CP, Obermayer-Straub P, Alex B, et al. Autoantibodies against glucuronosyltransferases differ between viral hepatitis and autoimmune hepatitis. Gastroenterology 1996;111:1576 –86. Manns M, Gerken G, Kyriatsoulis A, Staritz M, Meyer zum Buschenfelde KH. Characterisation of a new subgroup of autoimmune chronic active hepatitis by autoantibodies against a soluble liver antigen. Lancet 1987;1:292 –4. Wies I, Brunner S, Henninger J, et al. Identification of target antigen for SLAyLP autoantibodies in autoimmune hepatitis. Lancet 2000;355:1510 –5. Manns MP, Jentzsch M, Mergener K, et al. Discordant manifestation of LKM-1 antibody-positive autoimmune hepatitis in identical twins. Hepatology 1990;12:840. Strassburg CP, Manns MP. Autoimmune hepatitis versus viral hepatitis C. Liver 1995;15:225 –32. Johnson RJ, Gretch DR, Yamabe C, et al. Membranoproliferative glomerulonephritis associated with hepatitis C virus infection. N Engl J Med 1993;18:465 –70. Agnello V, Chung RT, Kaplan L. A role of hepatitis C virus infection in type II cryoglobulinemia. N Engl J Med 1992;19:1490. Mackie FD, Peakman M, Ma Y, et al. Primary and secondary liverykidney microsomal response following infection with hepatitis C virus. Gastroenterology 1994;106:1672 –5. Garcia-Buey L, Garcia-Monzon C, Rodriguez S, et al. Latent autoimmune hepatitis triggered during interferon therapy in patients with chronic hepatitis C. Gastroenterology 1995;108:1770 –7. Michitaka K, Durazzo M, Tillmann HL, Walker D, Philipp T, Manns MP. Analysis of hepatitis C virus genome in patients with autoimmune hepatitis type 2. Gastroenterology 1994;106:1603 –10. Durazzo M, Philipp T, Van Pelt FN, et al. Heterogeneity of liver–kidney microsomal autoantibodies in chronic hepatitis C and D virus infection. Gastroenterology 1995;108:455 –62. Rizzetto M, Swana G, Doniach D. Microsomal antibodies in active chronic hepatitis and other disorders. Clin Exp Immunol 1973;15:331 –44.

C.P. Strassburg et al. / Autoimmunity Reviews 2 (2003) 322–331 w29x Zauli D, Crespi C, Bianchi FB, Craxi A, Pisi E. Autoimmunity in chronic liver disease caused by hepatitis delta virus. J Clin Pathol 1986;39:897 –9. w30x Buti M, Amengual MJ, Esteban R, et al. Serological profile of tissue autoantibodies during acute and chronic delta hepatitis. J Hepatol 1989;9:345 –50. w31x Wesierska-Gadek J, Penner E, Hitchman E, Sauermann G. Antibodies to nuclear lamin C in chronic hepatitis delta virus infection. Hepatology 1990;12:1129 –33.

331

w32x Philipp T, Durazzo M, Trautwein C, et al. Recognition of uridine diphosphate glucuronosyl transferases by LKM-3 antibodies in chronic hepatitis D. Lancet 1994;344:578 –81. w33x Tukey RH, Strassburg CP. Human UDP-glucuronosyltransferases: metabolism, expression, and disease. Annu Rev Pharmacol Toxicol 2000;40:581 –616.

The World of Autoimmunity; Literature Synposis Enterovirus infection and beta cell autoimmunity Infectious agents are considered one of the major players in the pathogenesis of autoimmune diseases, including diabetes mellitus. Salminen et al. (J Med Virol 2003;69:91) investigated a Finnish birth cohort study for evidence of enterovirus infection from birth every 3–6 months. In a comparison between a group of children who became positive for diabetes-associated autoantibodies during the observation, and a group who remained autoantibody-negative, enterovirus infections were much more prevalent in the former, and enterovirus antibody levels were also higher in this group. More importantly, these infections were particularly frequent during the 6-month period preceding first detection of diabetes-associated autoantibodies. There were no differences, however, in the frequency of adenovirus infections between groups. These findings provide another evidence for the association of enterovirus infection and beta cells autoimmunity, and probably also of the pathogenesis of type 1 diabetes.

Schwann cells autoimmunity and inflammatory neuropathies Schwann cells play a role in the synthesis of peripheral nerve myelin. Kwa et al. (Brain 2003;126:361) report that 24% of Guillain-Barre syndrome patients and 26% of chronic inflammatory demyelinating polyneuropathy patients have sera reactivity towards Schwann cells. This prevalence of IgG antibodies against proliferating non-myelinating Schwann cells was almost absent in healthy subjects. The immunofluorescence was localized at the distal tips of the Schwann cells processes. They also provide additional data which support reactivity in these diseases not only towards myelin, but also towards other proteins and epitopes involved in Schwann cell–axon interaction.

Gluten-free diet in patients having high risk for type 1 diabetes Pastore et al. (J Clin Endocrinol Metab 2003;88:162) tested whether removal of gluten from human diet could reduce autoimmunity in human pre-clinical type 1 diabetes. The study included 17 first-degree relatives with at least two autoantibodies positive among islet cells antibodies, glutamic acid decarboxylase autoantibodies and insulin antibodies. All subjects took a 6-month period of gluten-free diet followed by a 6-month period of gluten-containing diet. Autoantibody titers did not change significantly during the study. However, acute insulin response to intravenous glucose tolerance test significantly increased in most subjects in the period of gluten-free diet, and decreased in most in the latter period. It is possible that gluten-free diet can preserve beta-cell function in subjects at risk for type 1 diabetes, even though humoral autoimmunity is not affected.