One third of HLA DQ2 homozygous patients with type 1 diabetes express celiac disease-associated transglutaminase autoantibodies.

Type 1 diabetes and celiac disease are both immunologic disorders where specific HLA alleles are associated with disease risk. We have developed a radioassay for autoantibodies to tissue transglutaminase (tTG) following the report that this enzyme is 'the' endomysial autoantigen (EMA) of celiac disease. The radioassay for transglutaminase autoantibodies is similar to that utilized for detecting anti-islet autoantibodies. The 'cut-off' for the IgA autoantibody assay was established as 3 x 100th percentile of 184 healthy control subjects at an index of 0.05. Ninety-eight of 847 patients with type 1 diabetes (11.6%) had tissue transglutaminase autoantibodies (tTG). All EMA-positive patients were positive (49/49) for transglutaminase autoantibodies, as were 49/540 EMA-negative patients. Twenty transglutaminase-positive patients consented to intestinal biopsy and 15 biopsies were positive for celiac disease. All patients with a transglutaminase level greater than 0.70 (13/13) had a positive biopsy, while none (0/3) with a level <0.3 had a positive biopsy. The prevalence of transglutaminase autoantibodies was higher in diabetic patients with HLA DQ2 or DQ8. One third of DQ2 homozygous patients (22/68) expressed transglutaminase autoantibodies vs. less than 2% of patients lacking DQ2 or DQ8. A simple radioassay for IgA transglutaminase autoantibodies detects all endomysial antibody positive patients and detects transglutaminase autoantibodies in 5% of endomysial autoantibody negative patients. The prevalence of transglutaminase autoantibodies is associated with DQ2 and DQ8 and in particular DQ2 homozygosity. Autoimmunity to transglutaminase is remarkably prevalent amongst patients with type 1 diabetes expressing certain class II HLA alleles.     

Contribution of celiac disease autoantibodies to the disease process

Evaluation of: Teesalu K, Panarina M, Uibo O, Uibo R, Utt M. Autoantibodies from patients with celiac disease inhibit transglutaminase 2 binding to heparin/heparan sulfate and interfere with intestinal epithelial adhesion. Amino Acids doi:10.1007/s00726-011-1020-1 (2011) (Epub ahead of print).

Transglutaminase 2 is a multifunctional protein involved in cellular adhesion. Moreover, transglutaminase 2 has been identified as the autoantigen in celiac disease, and in untreated celiac disease, in addition to being present in the serum, the transglutaminase 2-targeted autoantibodies are bound to their antigen in the basement membrane underlining the small-bowel mucosal epithelium. Furthermore, the disease-specific transglutaminase 2-targeted autoantibodies have been experimentally shown to exert various biological effects on different cell types. Using Caco-2 intestinal epithelial cells, it has now also been demonstrated that serum transglutaminase 2-targeted autoantibodies from untreated celiac patients inhibit the adhesion of these cells. These findings provide an important direction for future research to improve our general understanding of celiac disease pathogenesis and especially the role of the disease-specific autoantibodies during the progression of the disorder.     

Natural Hidden Autoantibodies to Tissue Transglutaminase Cross-React with Fibrinogen

Introduction  

Patients with celiac disease display autoantibodies against tissue transglutaminase (TG2), and the high sensitivity and specificity of these autoantibodies render them a reliable tool for diagnosis. However, we found that denatured sera from healthy persons also showed reactivity against TG2.     

Galactosylation of Serum IgA1 <Emphasis Type="Italic">O</Emphasis>-Glycans in Celiac Disease

In celiac disease, gluten ingestion provokes small-bowel mucosal injury and production of IgA autoantibodies against transglutaminase 2 (TG2). It has been suggested that in celiac patients IgA could mediate the transepithelial passage of gluten peptides in a mechanism involving the transferrin receptor. As IgA1 with galactose-deficient O-linked glycans has elevated affinity for the transferrin receptor, we assessed whether total serum IgA1 and IgA1 anti-TG2 autoantibodies in celiac patients are aberrantly glycosylated. We report that males with celiac disease have higher total serum levels of galactose-deficient IgA1 than non-celiac males. Furthermore, O-glycans of the disease-specific TG2 IgA1 autoantibodies in celiac patients exhibited elevated galactose deficiency. A gluten-free diet had no effect on the total serum levels of galactose-deficient IgA1, whereas the amount of galactose-deficient anti-TG2 IgA1 decreased. Thus, the undergalactosylated IgA1 molecules are not pathognomonic for celiac disease, but galactose deficiency in IgA1 could be an aggravating factor.     

IgA cross-reactivity between a nuclear autoantigen and wheat proteins suggests molecular mimicry as a possible pathomechanism in celiac disease

Celiac disease patients display IgA antibody reactivity to wheat as well as to human proteins. We used serum IgA from celiac patients and, for control purposes, from patients with Crohn's disease, ulcerative colitis and from healthy individuals to identify celiac disease-specific IgA autoantigens in nitrocellulose-blotted extracts from various human cell types (epithelial, endothelial, intestinal cells, fibroblasts). The pattern, recognition intensity and time course of IgA autoreactivity was monitored using serial serum samples obtained from celiac children before and under gluten-free diet. By immunoblot inhibition and subcellular (cytosolic, nuclear) cell fractionation we identified a 55 kDa nuclear autoantigen expressed in intestinal, endothelial cells and in fibroblasts which was recognized by IgA antibodies of approximately half of the celiac disease patients and cross-reacted with wheat proteins. IgA reactivity to the 55 kDa autoantigen disappeared during gluten-free diet and was inhibited after pre-absorption of sera with wheat proteins but not with tissue transglutaminase, previously reported as the unique celiac disease-specific autoantigen. In conclusion, we defined a novel 55 kDa celiac disease-specific nuclear IgA autoantigen which shares epitopes with wheat proteins and which is different from tissue transglutaminase and calreticulin. Although the newly defined autoantigen was recognized much less frequently than tissue transglutaminase, our data suggest molecular mimicry between wheat and human proteins as a possible pathomechanism for the induction and/or maintenance of mucosal tissue damage in celiac disease.     

Diagnostic immunology in celiac disease

Serum autoantibodies to transglutaminase and endomysium are found in the majority of patients with celiac disease, an autoimmune multisystem disorder affecting approximately 1% of Western and Middle-Eastern populations. Detection of these antibodies plays a crucial role in the diagnosis of celiac disease. The aim of this review is to summarize recent publications in this field, with particular focus on the applications and limitations of celiac autoantibody testing in routine clinical practice.     

Tissue transglutaminase antibody determination in celiac disease. Analysis of diagnostic specificity of anti-human IgA-type assays

Biopsy is onerous and, for this reason, immunodiagnostics in sera of celiac disease patients are an "additional diagnostic standard." The objective of the study was to investigate the variability in diagnostic specificity of ELISAs for the detection of IgA anti-tissue transglutaminase antibodies in serum of celiac disease patients who underwent biopsy. All patients were included in the study on the basis that they had a small intestinal biopsy. We studied 18 patients with histological proven celiac disease (7 male, 11 female, mean age +/- SD: 35+/-19 years) from Graz, Austria. Healthy control subjects were also entered into the study. The determinations of the anti-tissue transglutaminase antibodies were simultaneously performed together with the endomysium and gliadin antibody markers. We analysed the 216 serum values according to Cochran's non-parametric Q-test. The complexity to the analysis reflects the complexity of the diagnostic situation with the patients. No real differences were found in the reactions of the anti-human IgA-type anti-tissue transglutaminase ELISAs. Based on these results, an association was established between the outcomes of anti-human IgA-type ELISAs for the specific antigen and patients with histologically proven celiac disease, treated for celiac disease after histology was carried out and the diagnosis was made, and healthy controls. The detection of IgA anti-tissue transglutaminase antibodies in serum is a promising alternative to the indirect immunofluorescence determination of IgA-type endomysium antibodies. One ELISA for the specific antigen showed some advantage with respect to its extended scale of detection. Immunopathology of celiac disease can be based on the results of the appropriate IgA anti-tissue transglutaminase ELISAs under uncomplicated gastrointestinal conditions.     

Antibodies in celiac disease: implications beyond diagnostics

Celiac disease is a multisystemic dietary, gluten-induced autoimmune disorder characterized by the presence of transglutaminase (TG) 2 serum autoantibodies. Distinct autoantibodies targeting members of the TG family (TG2, TG3 and TG6) are found deposited in small-bowel mucosa and in extraintestinal tissues affected by the disease. Serum autoantibodies against other self-antigens also emerge in untreated celiac disease patients. Although villous atrophy and crypt hyperplasia in small-bowel biopsy samples are still the gold standards in diagnostics, celiac disease-specific antibodies are widely used as diagnostic aids. Gluten-induced small-bowel mucosal T-cell response is the cornerstone in the pathogenesis of the disorder, but humoral immunity may also play a central role. This review article is focused on the autoantibodies that occur in the context of celiac disease. The article summarizes the diagnostic utility of different celiac-related antibodies and discusses their roles in the pathogenesis of the disease.     

Contribution of celiac disease autoantibodies to the disease process

Transglutaminase 2 is a multifunctional protein involved in cellular adhesion. Moreover, transglutaminase 2 has been identified as the autoantigen in celiac disease, and in untreated celiac disease, in addition to being present in the serum, the transglutaminase 2-targeted autoantibodies are bound to their antigen in the basement membrane underlining the small-bowel mucosal epithelium. Furthermore, the disease-specific transglutaminase 2-targeted autoantibodies have been experimentally shown to exert various biological effects on different cell types. Using Caco-2 intestinal epithelial cells, it has now also been demonstrated that serum transglutaminase 2-targeted autoantibodies from untreated celiac patients inhibit the adhesion of these cells. These findings provide an important direction for future research to improve our general understanding of celiac disease pathogenesis and especially the role of the disease-specific autoantibodies during the progression of the disorder.     

The role of tissue transglutaminase (transglutaminase type II) for the intestinal manifestations of murine semi-allogenic graft-versus-host disease

The intestinal manifestation of acute murine semi-allogenic graft-versus-host (GvH) disease is characterized by the occurrence of lymphocytic infiltrates in the lamina propria, by crypt hyperplasia and villous atrophy. In a histological respect, this animal model resembles human celiac disease. Tissue transglutaminase (tTG) (transglutaminase type II) has been identified to be the major B cell autoantigen in celiac disease. Furthermore, tissue transglutaminase has been implicated to be involved in its pathogenesis. Therefore, we aimed to investigate whether tissue transglutaminase is expressed in the intestines of GvH animals and whether its inhibition has any effect on the intestinal histology. Sera of patients with celiac disease and anti-tTG antibodies were purified. These antibodies were used for immuno-histochemistry of jejunal cryosection from GvH and syngenic control animals at day 6 after lymphocyte transfer. Furthermore, GvH mice were treated with antitTG antibodies and with the inhibitor of tissue transglutaminase monodansyl-cadaverine. The effect of this treatment on the development of crypt hyperplasia and villous atrophy were examined by light microscopy of hematoxylin-eosin (H&E) stained jejunal paraffin sections. We found a strong subepithelial expression of tissue transglutaminase in GvH animals but not in syngenic control mice. The localization of tTG seemed to be associated with the extracellular matrix (ECM). However, neither the treatment of GvH animals with anti-tTG antibodies nor the application of mono-dansyl-cadaverine prevented the development of crypt hyperplasia and villous atrophy. Similar to the situation in human celiac disease tissue, transglutaminase is highly expressed in the intestine of animals undergoing a semi-allogenic graft-versus-host reaction. However, this enzyme is probably not involved in the development of crypt hyperplasia and villous atrophy in this animal model.     

Diagnostic value of anti-Saccharomyces cerevisiae and antineutrophil cytoplasmic antibodies for inflammatory bowel disease: high prevalence in patients with celiac disease

Both celiac disease and inflammatory bowel disease (IBD) are characterized by chronic diarrhea and the presence of distinct (auto)antibodies. In the present study we wanted to determine the prevalence of serological markers for inflammatory bowel disease, i.e., perinuclear antineutrophil cytoplasmic antibodies (pANCA) and/or anti-Saccharomyces cerevisiae antibodies (ASCA), in 37 patients with biopsy-confirmed celiac disease (Marsh IIIb/c). The majority of the patients was positive for IgA (auto)antibodies typically associated with celiac disease, i.e., antiendomysium antibodies (EMA) (86.5%), antigliadin antibodies (AGA) (73%), and antirecombinant human tissue transglutaminase antibodies (rh-tTGA) (86.5%). Four patients with selective IgA deficiency could be identified by analyzing EMA, AGA, and rh-tTGA for the IgG isotype. The prevalence of pANCA and ASCA, markers that are used for IBD, was unexpectedly high in our cohort of patients with celiac disease: 8 patients were positive for pANCA (IgG) and 16 patients were positive for ASCA (IgG and/or IgA). These results indicate that the presence of pANCA or ASCA in the serum of patients with chronic diarrhea does not exclude celiac disease. A prospective study is required to determine whether pANCA and/or ASCA identify patients at risk for developing secondary autoimmune disease.     

Anti-tissue transglutaminase, anti-endomysium and anti-R1-reticulin autoantibodies-the antibody trinity of coeliac disease.

Anti-tissue transglutaminase has been recently described as the predominant autoantigen in coeliac disease. We purified serum anti-tissue transglutaminase antibodies from three patients with coeliac disease by column chromatography and eluted tissue section-bound R1-anti-reticulin antibodies from sections of rat tissue for two of these. Lastly, we generated seven mouse MoAbs to guinea pig tissue transglutaminase. Each preparation was examined for anti-tissue transglutaminase, anti-endomysium, anti-R1 reticulin and anti-gliadin antibodies. Column-purified patient antibodies and 2/7 mouse MoAbs gave characteristic anti-endomysium/anti-R1 reticulin reactivity on rat, monkey and human tissue. All positive sera gave indistinguishable patterns of immunofluorescence on rat liver, kidney and stomach, monkey oesophagus, and human umbilical cord. Anti-R1-reticulin eluted from sections showed anti-tissue transglutaminase reactivity in 2/2 cases, but 0/2 showed anti-gliadin reactivity. In both, tissue section-eluted anti-R1 reticulin gave endomysial staining on monkey oesophagus. None of the mouse monoclonals, or any of the purified patient's anti-tissue transglutaminase or anti-R1-reticulin antibody showed any reactivity with gliadin. These data confirm tissue transglutaminase as the predominant autoantigen in coeliac disease and suggest that both anti-endomysium and anti-R1 reticulin reactivities seen in coeliac disease arise due to an immune response to tissue transglutaminase. Rigorous immunoabsorption was sufficient to abrogate reactivity in the tissue transglutaminase ELISA, but failed to completely absorb anti-endomysium and anti-reticulin activity. The possibility remains that some of the anti-endomysium and anti-reticulin activity was directed against antigens other than tissue transglutaminase.     

Autoantibodies in celiac disease

Autoantibody production is an important feature of many autoimmune disorders, signifying a breakdown of immune tolerance to self-antigens. In celiac disease, an autoimmune enteropathy with multiple extra-intestinal manifestations, autoantibody reactivity to transglutaminase 2 (TG2) has been shown to closely correlate with the acute phase of the disease. It serves as a specific and sensitive marker of celiac disease, and is highly useful in aiding diagnosis and follow-up. Immune reactivity to other autoantigens, including transglutaminase 3, actin, ganglioside, collagen, calreticulin and zonulin, among others, has also been reported in celiac disease. The clinical significance of these antibodies is not known, although some may be associated with specific clinical presentations or extra-intestinal manifestations of celiac disease. This review examines the presence of anti-TG2 and other autoantibodies in celiac disease, discussing their diagnostic value, their potential role in disease pathogenesis and current hypotheses that explain how their release may be triggered.     

Celiac Disease and Gyneco‐obstetrics Complications: Can Serum Antibodies Modulate Tissue Transglutaminase Functions and Contribute to Clinical Pattern?

Citation Sóñora C, Muñoz F, Del Río N, Acosta G, Montenegro C, Trucco E, Hernández A. Celiac disease and gyneco‐obstetrics complications: can serum antibodies modulate tissue transglutaminase functions and contribute to clinical pattern? Am J Reprod Immunol 2011; 66: 476–487 Problem Untreated celiac disease (CD) is often associated with early miscarriages, infertility, and alterations in menstrual cycle. Tissue transglutaminase (tTG) antibodies could be involved by interfering with tTG transamidating activity and/or biological functions mediated by its interaction with fibronectin (FN). Method of study The correlation between the presence of extra‐digestive disorders and the reactivity of sera against tTG‐FN and its effects on tTG transamidating activity was analyzed in a group or 50 women with recently diagnosed CD. Results Heterogeneous behavior was observed among serum samples derived from patients with different complaints, suggesting that differences in fine specificity patterns could condition clinical outcome. Sera from women with gynecological and/or obstetric problems induced significant inhibition of in vitro enzymatic activity in comparison with those without these kinds of disorders. Conclusions The significant correlation observed between serum effects and clinical profile suggests a putative involvement of tTG‐specific antibodies in gynecological and/or obstetric disorders during active CD.     

Occurrence of IgA and IgG autoantibodies to calreticulin in coeliac disease and various autoimmune diseases

Calreticulin (CRT), a high-affintiy calcium binding protein and chaperone, was recently identified as one of the targets of autoantibodies in coeliac disease. We evaluated the level of IgA and IgG antibodies to CRT in sera from patients with coeliac disease and various autoimmune diseases. The level of antibodies to gliadin (shown previously to cross-react with CTR), isolated enterocytes and tissue transglutaminase were determined for comparison. The mean level of IgA antibodies to CRT was significantly higher (P< 0.001) in sera from coeliac patients with active disease (139.9+/-11.2 AU/+/-SE) than in healthy controls (20.9+/-1.7 AU). In sera of patients with systemic lupus erythematosus (SLE), insulin dependent diabetes mellitus (IDDM), multiple sclerosis (MS) and autoimmune thyroiditis (AT) or inflammatory bowel disease (IBD) the mean level (25.8+/-3.7 to 38.1+/-5.6 AU) did not exceed the cut-off value. A low level of these antibodies, however, was detected in some sera of patients with MS and IBD. The level of IgG anti-CRT antibodies was increased in coeliac patients (mean 125.4+/-8.0 AU, P< 0.001) when compared to that in healthy controls (33.9+/-2.3 AU). The IgG anti-CRT antibodies were also detected in about 30% of SLE patients sera (54.1+/-3.6 AU, P< 0.001), but the mean level reached only half that detected in coeliac patients.     

Tissue transglutaminase antibodies in celiac disease, comparison of an enzyme linked immunosorbent assay and a dot blot assay

AIMS: The purpose of our study is to determine the sensitivity, specificity and predictive values of an enzyme linked immunosorbent assay (ELISA) and a dot blot assay for the detection of IgA class anti-tissue transglutaminase antibodies (IgA-AtTGA) and to compare these results with those of IgA class anti-endomysium antibodies (IgA-AEA), IgA class anti-reticulin antibodies (IgA-ARA) and IgA class anti-gliadin antibodies (IgA-AGA). PATIENTS: Serum samples from 143 patients (97 children, 46 adults) with untreated celiac disease (CD) confirmed by intestinal biopsy and 74 disease controls (64 children, 10 adults) were studied. Methods. - The anti-tissue transglutaminase antibodies were detected by dot blot assay and an ELISA using guinea pig tissue transglutaminase (gp-tTG) as antigen. The anti-endomysium antibodies were detected by an indirect immunofluorescence technique on cryostat sections of human umbilical cord. The anti-reticulin antibodies were also investigated by indirect immunofluorescence on cryostat sections of kidney, liver and stomach of rat. The anti-gliadin antibodies were determined by an ELISA. RESULTS: The sensitivity of an ELISA for the detection of anti-tissue transglutaminase antibodies was 86% in children and 87% in adults and the sensitivity of dot blot assay was 57% in children and 54% in adults. The specificity of an ELISA and dot blot for the detection for anti-tissue transglutaminase antibodies was, respectively, 96% and 88% lower than that of anti-endomysium antibodies (100%). The sensitivity of anti-gliadin antibodies was 97% in children and 91% in adults and their specificity was 85%. The sensitivity of anti-reticulin antibodies was 94% in children and 87% in adults. Their specificity was 100%. CONCLUSIONS: The sensitivity and specificity of an ELISA for the detection of anti-tissue transglutaminase antibodies were better than that of dot blot assay. However, this dot blot assay could screen four celiac patients who have not had anti-tissue transglutaminase antibodies by an ELISA. The sensitivity of anti-endomysium antibodies was better than that of anti-tissue transglutaminase antibodies, anti-reticulin antibodies and anti-gliadin antibodies but in children aged less than 2 years, the sensitivity of anti-gliadin antibodies was better than that of anti-tissue transglutaminase antibodies.     

Interpretation of serological tests in the diagnosis of celiac disease: Anti-deamidated gliadin peptide antibodies revisited

Algorithms for celiac disease diagnosis provided by guidelines are based primarily on anti-tissue transglutaminase 2 (TG2) antibodies and/or anti-endomysium antibodies. The place of anti-deamidated gliadin peptide (DGP) antibodies is less well established. This study was designed to assess the clinical relevance of anti-DGP antibodies. Two thousand and twenty-six consecutive unselected patients systematically tested for anti-TG2, endomysium, gliadin, DGP antibodies and IgA dosage were investigated. The serological interpretation was assessed by analyzing the medical records of patients. From the 1984 newly investigated patients suspected of celiac disease, 10% had at least one celiac marker. Anti-TG2, anti-endomysium, anti-gliadin and anti-DGP antibodies were found in 1.1%, 0.6%, 6.8% and 4.1% of cases respectively, with different combinations. The diagnosis of celiac disease was retained in 0.45% of patients. When using the duodenal biopsies as a gold standard, analysis of the anti-DGP diagnosis performance showed that the specificity and the predictive positive value (PPV) were lower than that of the anti-TG2 assay. The combined detection of anti-TG2 and anti-DGP antibodies had a lower PPV than that of anti-TG2 and anti-endomysium antibodies (p?=?0.04). When analyzing the contribution of anti-DGP antibodies as an additional marker to both anti-TG2 and anti-endomysium antibodies, the PPV of the three associated antibodies was shown to be significantly lower than the PPV of the both anti-TG2 and anti-endomysium antibodies (p?=?0.04). As a conclusion, anti-DGP antibodies may not have the diagnosis value required as an additional screening test to anti-TG2 antibodies for identifying celiac disease patients in medical centers where anti-endomysium detection is available.     

Celiac disease prevalence and predisposing-HLA in a cohort of 93 Williams-Beuren syndrome patients

Williams-Beuren syndrome is considered to be at increased risk for celiac disease, as for recent literature data and celiac disease guidelines, despite pathogenic mechanisms are still unclear. Our study analyzed the prevalence of autoimmune disorders, HLA DQ2 and/or DQ8 haplotypes, of transglutaminase antibodies and of diagnosis of celiac disease in a cohort of 93 Williams-Beuren syndrome's patients (mean age 21.26 years). Our study showed an increased prevalence of celiac disease equal to 10.8% (10/93 patients). We did not find a significant different frequency of predisposing HLA in subjects with Williams-Beuren syndrome compared to literature data in the general population (49.5% vs. 42.9%, with p > .1), nor a susceptibility to autoimmunity. This suggests that the increased prevalence of celiac disease in Williams-Beuren syndrome cannot be ascribed to HLA haplotype and may be related to other factors that still need to be identified in these patients.     

Is celiac disease an autoimmune disorder?

Celiac disease, which results from an immune reaction to ingested cereal gluten proteins, has several autoimmune features. In particular, celiac disease patients produce highly disease specific IgA and IgG autoantibodies to tissue transglutaminase when they are on a gluten-containing diet, and they have small intestinal intraepithelial lymphocytes which can mediate direct cytotoxicity of enterocytes expressing MIC molecules in an antigen non-specific manner. Similar to typical autoimmune disorders, celiac disease has a multifactorial aetiology with complex genetics, and several autoimmune diseases are commonly presented by patients with celiac disease. Much has been learned about the immunology of celiac disease in recent years, and there is overwhelming evidence that the immune response to gluten is central to the pathogenesis. In light of this, the many autoimmune phenomena associated with celiac disease are thought-provoking, and they challenge us to rethink the boundaries between autoimmunity and immunopathology.     

Do Autoantibodies Predict Autoimmune Liver Disease in Primary Sjögren's Syndrome? Data of 180 Patients upon a 5 Year Follow‐Up

Objective: To evaluate the clinical value of autoantibodies as serological markers to predict autoimmune liver diseases in primary Sjögren's syndrome (SS). Materials and methods: 180 patients who met the European diagnostic criteria for SS but without a history of liver disease were studied upon a 5 year follow‐up. Sera taken at enrolment were evaluated by immunofluorescence analysis (IF‐AMA) on rat liver, stomach and kidney sections, enzyme‐linked immunosorbent assay using rat mitochondrial, microsomal and soluble liver antigens and Western blot (WB) analysis using rat mitochondrial antigens. Results: At presentation, 152 (84%) sera had autoantibodies. Antinuclear antibodies (ANA) were expressed in 58% of patients and displayed three distinct patterns (speckled, homogenous and anticentromere). Smooth muscle autoantibodies (SMAs) and parietal cell autoantibodies were found in 39 and 4.5% of patients, respectively. Three patients presented antimitochondrial antibodies by IF‐AMA, and two of them developed symptomatic primary biliary cirrhosis (PBC). Two patients without IF‐AMA and without evidence of cholestasis had PBC‐specific AMA (anti‐PDC–E2 and anti‐BCKADC–E2). However, these two patients and the third IF‐AMA‐positive woman remained free from symptoms and biochemical signs of PBC. Autoimmune hepatitis (AIH) (n = 2), ‘overlap syndrome’ of AIH and chronic hepatitis C (n = 1) and autoimmune cholangiopathy (AIC) (n = 1) were diagnosed in four patients. Conclusions: Patients with IF‐AMA usually develop symptomatic PBC upon a 5 year follow‐up. Our findings support the idea that patients without IF‐AMA, who express PBC‐specific AMA, are in early, asymptomatic stage of the disease. High‐titre SMA and IF‐AMA are the most specific indicators for AIH and PBC.