Specific liver autoreactivity in schistosomiasis mansoni.

To investigate the possible involvement of autoimmune mechanisms in the development of hepatosplenic schistosomiasis (HSS), 234 patients with chronic Schistosoma mansoni infections were screened for a wide range of non-organ-specific autoantibodies as well as for antibodies reacting with the GOR peptide and with a liver-specific autoantigen, the hepatic asialoglycoprotein receptor (ASGP-R). Thirty-five (15.0%) were seropositive for antinuclear, smooth muscle or gastric parietal cell antibodies at low titres (< or = 1:80), and 15/176 (8.5%) had anti-GOR, all of whom had concomitant hepatitis C viral (HCV) infections. Anti-ASGP-R was found in 64 (27.4%) of the 234 patients at titres similar to those found in 18 untreated auto-immune hepatitis patients studied concurrently. Anti-ASGP-R seropositivity occurred significantly (P < 0.005) more frequently in patients with HSS (62/190, 32.6%) than in those with hepatointestinal schistosomiasis (2/44, 4.5%), but did not correlate with severity of liver disease or with the presence of the non-organ-specific autoantibodies. Anti-ASGP-R was found significantly (P < < 0.0005) less frequently in HSS patients who had had a splenectomy for portal hypertension (5/86, 5.8%) than in those who had not had a splenectomy (57/104, 54.8%). The findings suggest that liver-specific autoreactivity may play a role in the development of HSS.

Alcohol dehydrogenase: a target of humoral autoimmune response in liver disease.

BACKGROUND & AIMS: Liver-specific membrane lipoprotein (LSP) is a heterogeneous liver preparation that has been widely used to study autoreactivity in liver disease. The aim of this study was to identify autoantigens in LSP. METHODS: Guinea pig anti-LSP serum was used to screen a human liver complementary DNA (cDNA) library. Humoral immune responses to isolated potential autoantigens were investigated by immunoblotting in 91 pediatric patients with various liver diseases, 20 adult patients with alcoholic liver disease and 20 with autoimmune thyroid disease, 37 healthy children, and 20 healthy adults. RESULTS: A 1.6-kilobase cDNA insert isolated from the cDNA library was found to encode amino acids 61-374 of the human alcohol dehydrogenase (ADH)-gamma 1 subunit. Antibodies to this or other ADH subunits were found significantly more frequently in autoimmune liver diseases (19 of 39 patients; 49%), Wilson's disease (5 of 13 patients; 38%), and alcoholic liver disease (10 of 20 patients; 50%) than in normal controls (P < 0.0001, P < 0.005, and P < 0.05, respectively) and correlated with disease activity in autoimmune liver disease. CONCLUSIONS: ADH has been identified as a new antigenic component of the LSP using a xenogeneic antiserum to immunoprobe a human cDNA liver library and seems to be a target autoantigen in liver disease. This approach may be useful in identifying other potential autoantigens.

Autoimmune mechanisms in chronic hepatitis B and delta virus infections.

OBJECTIVE: To investigate the possibility that hepatitis delta virus infection may be responsible for, or enhance, the autoreactivity seen in patients chronically infected with hepatitis B virus. DESIGN: Sera from 68 patients with chronic hepatitis B infection, 27 of whom had concomitant delta virus (HDV) infection and 19 of whom were infected with hepatitis C virus (HCV), were screened for (1) the non-organ-specific autoantibodies usually associated with autoimmune hepatitis, (2) antibodies against the putative autoantigen, GOR, which have been described in patients with HCV infection and (3) antibodies against a hepatocyte-specific autoantigen, the asialoglycoprotein receptor (ASGP-R). METHODS: Anti-GOR antibodies were detected using an enzyme-linked immunosorbent assay against a synthetic GOR peptide, non-organ-specific autoantibodies using standard indirect immunofluorescence and anti-ASGP-R antibodies using a radioimmunoassay. RESULTS: Liver-specific autoreactivity, manifested by circulating anti-ASGP-R antibodies, was found in 41 (60.3%) patients and correlated independently with the histological severity of liver damage but not with HDV infection. In contrast, non-organ-specific autoantibodies (antinuclear, anti-smooth muscle, anti-gastric parietal cell and anti-liver-kidney microsomal type 1) were found (mostly at low titres) in only 15 (22.1%) patients and did not correlate with either HDV infection or with histological severity. Basal cell layer antibodies and type 3 liver-kidney microsomal antibodies were not seen in any patient. Antibodies against GOR were found in only five (7.4%) patients, all of whom showed evidence of exposure to HCV. CONCLUSION: The findings suggest that HBV-induced liver-specific autoreactions might contribute to periportal liver damage in patients with chronic hepatitis B infection, but do not support the notion that the delta virus can induce an autoimmune response or that HCV coinfection suppresses autoreactivity in this situation.

Antibody-directed complement-mediated cytotoxicity to hepatocytes from patients with chronic hepatitis B.

The susceptibility of hepatocytes from patients with chronic hepatitis B to complement-dependent cytotoxicity mediated by heterologous antibodies to hepatitis B virus core (anti-HBc) and surface (anti-HBs) antigens and to hepatic asialoglycoprotein receptor was examined using a microcytotoxicity assay. The anti-HBc-induced cytotoxicity was found to be markedly enhanced against hepatocytes isolated from patients with chronic active hepatitis (72.6 +/- 9.5% (mean +/- s.e.m.); n = 6) over that against hepatocytes from individuals with chronic persistent hepatitis or inactive liver cirrhosis (40.6 +/- 18.6%; n = 4) (P = 0.019). Overall, values of the anti-HBc-directed cytotoxicity were higher in patients positive for HBcAg in hepatocytes and seropositive for hepatitis B virus e antigen (HBeAg). Hepatocytotoxicity was also exerted by anti-HBs and anti-asialoglycoprotein receptor antibodies in the presence of complement, but it was not seemingly related to disease activity. These results indicate that hepatitis B virus core and surface antigens and asialoglycoprotein receptor at the hepatocyte surface can be recognized by antibodies, and raise the possibility that complement-dependent cytolysis may contribute to the injury of hepatitis B virus-infected hepatocytes. The data also suggest that liver cells of patients with severe chronic hepatitis might be more susceptible to anti-HBc antibody-directed complement-mediated cytotoxicity than those with inactive liver histology.

Cellular and humoral immune reactions against autoantigens and hepatitis C viral antigens in chronic hepatitis C.

BACKGROUND/AIMS: Previous reports have suggested that the hepatitis C virus (HCV) may induce autoimmune hepatitis. The aim of this study was to examine this hypothesis by investigating humoral and cellular immune responses to HCV-related antigens and various autoantigens in patients with chronic HCV infections. METHODS: Lymphoproliferative responses in vitro and/or circulating antibodies to an HCV core peptide, the putative autoantigen GOR, the liver-specific hepatic asialoglycoprotein receptor (ASGP-R), and other autoantigens were investigated in 27 adults with chronic hepatitis C. RESULTS: Five patients with HCV (18.5%) showed cellular immune responses to ASGP-R and two others had antibodies to ASGP-R, whereas 6 of 14 patients (42.8%) showed cellular responses to GOR and 7 of 14 patients (50%) showed responses to HCV core. Other autoantibodies were detected in three patients (11%). Nine patients with autoimmune hepatitis studied concurrently for comparison showed cellular and/or humoral responses to ASGP-R but not to GOR. Only 2 of 11 patients with other chronic liver disorders showed immune responses to any antigen tested. CONCLUSIONS: Specific immunocompetence against HCV-related antigens can often be shown in patients with chronic hepatitis C but is infrequently accompanied by autoreactions against liver-specific or nonspecific antigens. A reported association between T-cell responses to HCV core and lack of liver damage could not be confirmed.

Demographics of anti-asialoglycoprotein receptor autoantibodies in autoimmune hepatitis.

BACKGROUND/AIMS: The asialoglycoprotein receptor (ASGPR) is an established, liver-specific autoantigen. This multicenter study investigated the specificity of anti-ASGPR autoantibodies for autoimmune hepatitis (AIH) in different ethnic groups. METHODS: Nine hundred fourteen sera from European, Japanese, and North American (U.S.) patients with chronic inflammatory liver disorders were tested. An enzyme-immunoassay using human ASGPR and a radioimmunoassay against rabbit ASGPR, performed independently on coded sera, were compared. RESULTS: The highest frequency (76%) of anti-human ASGPR was found in AIH patients (11/24 U.S.; 21/25 European; 28/30 Japanese), particularly in those with active disease before treatment (53/62, 85%), and decreased in titer with response to immunosuppressive therapy. These antibodies were found at low titers in 43 (11%) of 385 patients with viral hepatitis and in 25 (7.6%) of 328 patients with other chronic inflammatory liver disorders (P < 0.0005 compared with all AIH patients). Twenty of 37 sera tested by enzyme-immunoassay and radioimmunoassay were positive, and nine were negative for anti-ASGPR by both assays (78% concordance); six sera were exclusively positive on human substrate. CONCLUSIONS: Circulating anti-ASGPR autoantibodies are closely associated with autoimmune hepatitis independent of geographic or ethnic criteria. Two anti-ASGPR assays currently in use show high reliability.

Virus-induced autoimmunity in hepatitis C virus infections: a rare event.

Serial serum samples from 16 Italian patients presenting with acute hepatitis C virus (HCV) infections (which progressed to chronic hepatitis in six) were screened for the non-organ-specific autoantibodies most frequently associated with autoimmune hepatitis (AIH), as well as for antibodies against the hepatic asialoglycoprotein receptor (ASGP-R) and against the GOR peptide. One patient had low titres (1:10-1:80) of liver-kidney microsomal (LKM-1) antibodies during the recovery phase and three others had transient low titres of anti-smooth muscle (IgM class, 1:10) or anti-ASGP-R (1:150-1:300). Anti-GOR was detected in 43 (65%) of 66 sera from 13 of these patients. There was no correlation between any of these findings and progression to chronicity. By comparison, 18 patients with AIH studied concurrently before institution of immunosuppressive therapy all had antinuclear and/or smooth muscle antibodies, or LKM-1, at 1:40-1:640 and anti-ASGP-R at 1:300-1:2,100. None of these 18 had evidence of HCV infection and all were seronegative for anti-GOR. The findings indicate that the autoantibodies usually associated with AIH are rare in HCV infections but the virus can very occasionally induce a transient autoimmune response. Anti-GOR appears to be an antibody specifically related to HCV infection and is probably not a marker of induced autoimmunity, and it does not predict progression to chronic hepatitis.

Antibodies to hepatitis C virus in autoimmune liver disease: evidence for geographical heterogeneity.

To resolve conflicting reports about the occurrence of antibodies against hepatitis C virus (HCV) in patients with autoimmune chronic active hepatitis (AI-CAH), sera from UK and Italian patients were tested with the original anti-HCV assay (Ortho) and a novel anti-HCV assay (UBI) based entirely on synthetic HCV peptides. 28 (60%) of 47 Italian patients with type-1 AI-CAH were anti-HCV-positive by Ortho ELISA, 25 of whom were also strongly positive by the UBI assay. 15 (60%) of 25 UK patients with type-1 AI-CAH were HCV-positive by Ortho ELISA but only 2 were positive by the UBI assay. Similarly, 29 (88%) of 33 Italian patients with type-2 AI-CAH, but 0 of 10 UK patients, were very strongly anti-HCV-positive with the UBI assay. Italian patients with AI-CAH appear to have a high frequency of genuine exposure to HCV, whereas seropositivity by the Ortho HCV ELISA in UK patients is likely to represent a false-positive result. These findings indicate important geographical and/or genetic influences in autoimmune liver disease among different populations.

T-cell-directed hepatocyte damage in autoimmune chronic active hepatitis.

To investigate the function of activated T lymphocytes in autoimmune chronic active hepatitis, 7 of 15 T-cell clones from the peripheral blood of 8 patients were studied. These clones showed specificity for liver-membrane antigen with proliferation when stimulated by rabbit liver cell membranes. 6 of these clones reacted with liver-specific lipoprotein complex, and 1 clone (and 3 subclones) responded to the asialoglycoprotein receptor (ASGPR), both known targets of immune attack in autoimmune chronic active hepatitis. 2 of these clones stimulated autologous B lymphocytes to produce liver-membrane-specific autoantibodies and antibody to the ASGPR. These results suggest that liver-membrane-specific activated T lymphocytes in peripheral blood may be important in the autoimmune attack of chronic active hepatitis.

Auto-immune features in patients with idiopathic chronic active hepatitis who are seronegative for conventional auto-antibodies.

In patients with chronic active hepatitis (CAH), the absence of the conventional serum auto-antibodies (antinuclear, smooth muscle and liver-kidney microsomal) is often taken as evidence against an auto-immune aetiology and as indicative that the disease is unlikely to respond to immunosuppressive therapy. We report 12 British patients (11 female) who presented with histologically florid CAH (11 with cirrhosis or fibrosis and seven with ascites) but without significant titres of these antibodies or any other demonstrable aetiological feature (cryptogenic CAH), who have been followed up for a median of 5.25 years (range: 0.75-16 years). Ten had hypergammaglobulinaemia and/or specific elevations of serum IgG concentrations at presentation and five of 10 patients tested were found to have the HLA allotypes B8 and DR3. Remission was initially induced with prednisolone with or without azathioprine in all patients. Six patients subsequently relapsed on one or more occasions, either spontaneously while on maintenance therapy or during attempts to withdraw corticosteroids, and required increases or reintroduction of immunosuppressive therapy to regain disease control. Retrospective analysis of pretreatment samples from 11 of the patients revealed that all had been seropositive at presentation for auto-antibodies against the liver membrane lipoprotein preparation known as liver-specific membrane lipoprotein (LSP) and/or against the hepatic asialoglycoprotein receptor (ASGP-R), titres of which subsequently fluctuated in direct relation to response to treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

Antibodies against the hepatic asialoglycoprotein receptor perfused in situ preferentially attach to periportal liver cells in the rat.

Autoantibodies reacting with the galactose-specific hepatic asialoglycoprotein receptor--a liver-specific component expressed on the surfaces of hepatocytes--are often found in patients with chronic active hepatitis of presumed autoimmune origin. As part of an investigation into whether these anti-asialoglycoprotein receptor antibodies might be involved in the development of periportal liver damage in chronic active hepatitis, livers of ether-anesthetized rats were perfused in situ with polyclonal guinea pig anti-rabbit asialoglycoprotein receptor or murine monoclonal anti-human galactose-specific hepatic asialoglycoprotein receptor antibodies in excess at less than 8 degrees C or, as a control, with guinea pig anti-human plasma protein antibodies or normal guinea pig serum. Rapid (1 min) antegrade (by way of portal vein) or retrograde (through hepatic veins by way of vena cava) perfusions were performed in a nonrecirculating (once-through) mode in Ca+(+)-free medium. Blocks of liver tissue were immediately snap-frozen and the distribution of the antibody examined in cryostat sections by using an avidin-biotin immunohistochemical technique. In all of the perfusions with anti-asialoglycoprotein receptor (six antegrade, seven retrograde), the antibodies were found to be prominently and almost exclusively deposited on liver cells in the periportal areas. No deposition of immunoglobulins was detected in livers perfused with the control guinea pig sera. The findings suggest that the asialoglycoprotein receptor is expressed at high density mainly on cells in zone 1 of the hepatic lobule, and this may have implications for the development of periportal liver damage in chronic active hepatitis.

Autoreactivity to hepatocellular antigens in primary biliary cirrhosis and primary sclerosing cholangitis.

To investigate the possible involvement of autoimmune reactions in the periportal hepatocellular damage that is often seen in primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC), sera from 35 patients with PBC and 31 with PSC were tested for autoantibodies against the liver-specific lipoprotein preparation, LSP, and against one of its liver-specific constituents, the asialoglycoprotein receptor (ASGP-R), and compared with results in 24 untreated patients with autoimmune chronic active hepatitis (AI-CAH). Anti-LSP antibodies were found in 48.5% of the PBC and 10% of the PSC patients, vs. 100% of those with AI-CAH, while anti-ASGP-R was found in 23% of PBC, 10% of PSC and 96% of AI-CAH patients. In PBC (but not in PSC) these antibodies correlated with severity of periportal inflammation and piecemeal necrosis but tended to be associated with the later stages of the diseases and both seropositivity for, and titres of, anti-LSP and anti-ASGP-R were significantly influenced by the presence of HLA DR3 (positively) and DR2 (negatively) in these patients. DR3 was also associated with significantly higher, and DR2 with lower, serum IgG concentrations in PBC. The findings suggest that, in PBC, DR2 and DR3 may be associated, respectively, with one or more genes that code for down-regulation or for elevation of overall immunoresponsiveness and that autoreactivity to hepatocellular antigens in PBC is more likely to be a consequence than a cause of hepatocellular injury. Periportal liver damage in PSC seems to involve different mechanisms.

Immunohistochemical demonstration of the asialoglycoprotein receptor in rat liver by a sensitive avidin-biotin technique.

An immunohistochemical technique for the detection of the hepatic asialoglycoprotein receptor (ASGP-R) in cryostat sections of liver by polyclonal and monoclonal anti-ASGP-R antibodies is described. The procedure is based on the alkaline-phosphatase-avidin-biotin complex (ABC-AP) system and important features include fixation of the sections with periodate-lysine-paraformaldehyde (with or without dichromate) and an absolute requirement for blocking of endogenous biotin activity. The sensitivity of the technique is such that binding to ASGP-R can be detected with femtomolar concentrations of monoclonal anti-ASGP-R antibodies and, with polyclonal antisera, approaches that of a radioimmunoassay.

Different immune mechanisms leading to autoimmunity in primary sclerosing cholangitis and autoimmune chronic active hepatitis of childhood.

Children with primary sclerosing cholangitis or autoimmune chronic active hepatitis have similar high levels of immunoglobulin G and non-organ-specific autoantibodies and may have similar histological features. To investigate a possible immunopathogenesis of primary sclerosing cholangitis, we have studied a series of regulatory and/or effector immune mechanisms in eight children with primary sclerosing cholangitis, comparing them to 14 children with autoimmune chronic active hepatitis and 24 healthy children as controls. Antibodies to a liver membrane protein preparation were found in all children with autoimmune chronic active hepatitis tested and in seven of eight with primary sclerosing cholangitis, whereas antibodies against the hepatic asialoglycoprotein receptor were present in three of six patients with autoimmune chronic active hepatitis and in two of the eight with primary sclerosing cholangitis. Lymphocyte cytotoxicity values to autologous hepatocytes were similarly elevated in primary sclerosing cholangitis (median: 50%; range: 38 to 83%) and in autoimmune chronic active hepatitis (median: 52%; range 37 to 87%) compared to controls (median: 8%; range: 0 to 27%) (p less than 0.01 for both). In contrast, T suppressor cell number and function were normal in patients with primary sclerosing cholangitis (median: 23%; range: 19 to 28%; and median: 54%; range: 44 to 61%), but significantly decreased in patients with autoimmune chronic active hepatitis (median: 15%; range: 9 to 21%; and median: 9%; range: -40 to +21%) when compared to controls (median: 24%; range: 16 to 29%; and median: 53%; range: 8 to 77%) (p less than 0.01 for both).(ABSTRACT TRUNCATED AT 250 WORDS)

Detection of anti-liver cell membrane antibody using a human hepatocellular carcinoma cell line.

A radioimmunometric technique for the detection of autoantibodies to liver membrane antigens has been developed using Alexander cells, a human hepatocellular carcinoma cell line. After incubation of Alexander cells with serum, antimembrane antibodies were detected by addition of 125I-labeled Protein A. Binding ratios in 15 children with uncontrolled autoimmune chronic active hepatitis and in seven children with primary sclerosing cholangitis were significantly higher than in 18 age-matched normal controls. Nine patients with inactive autoimmune chronic active hepatitis, 13 with alpha 1-antitrypsin deficiency and five with fulminant hepatic failure had ratios similar to controls. In nine patients with Wilson's disease, there was a modest but significant increase in binding ratio. In four children with autoimmune chronic active hepatitis, binding ratios fell during effective immunosuppressive therapy. Sera from patients with systemic lupus erythematosus or rheumatoid arthritis gave normal results, excluding that binding derives from Fc-mediated immune complex capture. A positive correlation was found between Alexander cell binding values and anti-liver-specific protein antibody titers, suggesting that the two assays detect antibodies against shared antigenic determinants. The Alexander cell assay is a simple, rapid and sensitive technique to detect antibody to liver cell membrane antigens.

Sensitive avidin biotin based technique for identifying liver membrane antigens in tissue sections.

An enzyme immunohistochemical technique for the localisation of liver membrane antigens in tissue sections by antisera raised in guinea pigs against the liver preparation known as "liver specific membrane lipoprotein (LSP)" was developed, based on the alkaline phosphatase avidin biotin complex (ABC AP) system. Of a wide range of fixatives and fixation conditions investigated, a short (five minute) exposure of cryostat sections to Bouin's fluid provided the most satisfactory results and--together with procedures to block endogenous biotin and alkaline phosphatase--yielded clear sections with no background staining or other artefacts to interfere with specific staining patterns. The sensitivity of the technique approaches that of a radioimmunoassay, as shown by the staining of the sinusoidal domains of hepatocellular plasma membranes by the guinea pig anti-LSP antisera at dilutions up to 1/50,000. Apart from its reliability and sensitivity the procedure offers additional advantages over techniques such as indirect immunofluorescence in that it provides a permanent preparation with well defined morphological details which can be seen by ordinary light microscopy.

Production of murine monoclonal antibodies against a liver-specific, species cross-reactive antigen in the liver-specific lipoprotein (LSP) preparation.

Previous attempts in several laboratories have failed to produce murine monoclonal antibodies (MAbs) against liver-specific, species-cross-reactive, cell surface-expressed antigens in the normal liver preparation known as liver-specific membrane lipoprotein (LSP). In the present study, BALB/c mice were pretreated with a single dose of cyclophosphamide (20 mg/kg), hyperimmunized with human LSP and hybridomas produced by fusion of spleen cells from these mice with murine myeloma (P3-NS1-Ag4-1) cells. To bias selection in favour of MAbs reacting with species cross-reactive epitopes, hybridoma supernatants were screened by ELISA against rabbit LSP. From 70 stable hybridomas, four MAbs were obtained that react with rabbit LSP. One is an IgM antibody and the other three are of IgG2a class. All four react with the hepatic asialo-glycoprotein receptor (HL), a liver-specific, species-cross-reactive component that is normally expressed on the surfaces of hepatocytes. Using a rapid screening technique (an 'additive' ELISA), preliminary evidence was obtained indicating that these four MAbs between them recognise three different epitopes on the HL molecule. The results suggest that this is a viable approach for the production of MAbs against autoantigens to which autoreactivity may normally be suppressed.

Serial study of liver-directed autoantibodies and autoreactive T-lymphocytes in acute viral hepatitis B.

To explore the mechanisms underlying liver-directed autoimmune reactions in acute Hepatitis B Virus (HBV) infection, we followed five subjects who were identified in the early incubation phase (30-70 days before the first elevation of transaminases). We assessed serially cellular (using a T-lymphocyte migration inhibitory factor assay) and humoral (RIA) immunity to LSP (a macromolecular, liver-derived lipoprotein complex) and hepatic lectin (HL), the liver-specific receptor for desialylated glycoproteins, which appears to be a major target antigen for autoreactions in autoimmune chronic active hepatitis. Anti-LSP and anti-HL autoantibodies were found, at some stage during acute HBV infection, in 4/5 subjects, whereas cellular immunity to the same antigens was detected in only two patients. Sustained production of anti-HL antibodies was noted only in patients showing cellular immunity to this antigen and was apparently secondary to liver damage, whereas anti-LSP antibodies were first detected at the onset of liver injury when there was no evidence of T-cell immunity to the same antigenic complex. One explanation for this apparent dichotomy between cellular and humoral responses to LSP is that a helper T-cell response to the major envelope component of HBV, HBsAg, which precedes by 10-20 days the development of anti-LSP antibodies, promotes a humoral reaction to autoantigens contained in the LSP preparation, coexpressed with HBsAg, on the surface of infected hepatocytes.

Liver autoreactivity in acute virus A, B and non-A, non-B hepatitis.

As part of an investigation into the question of whether virus-induced autoreactivity might contribute to liver damage in viral hepatitis, serial studies (from onset through recovery) of circulating liver autoantibodies have been performed in patients with uncomplicated acute virus A (AVH-A), B (AVH-B) and non-A, non-B (AVH-NANB) hepatitis in whom the time of onset of symptoms could be precisely documented. One hundred and forty-four sera from 35 patients were tested by radioimmunoassay for autoantibodies against the liver-derived lipoprotein complex, LSP, and also against one of its constituents--the asialoglycoprotein receptor, known as hepatic lectin (HL). Anti-LSP antibodies were found in all 10 patients with AVH-A, in 17/18 with AVH-B and in 3/7 with AVH-NANB at titres that declined during recovery. Anti-HL antibodies were detected concurrently in 6 of the AVH-A patients and in 5 with AVH-B but on only 1 occasion in 1 patient with AVH-NANB. Transient cellular immunity to LSP, assayed by a T-lymphocyte migration inhibitory factor test, was detected in 4 of the 6 AVH-B patients tested, 2 of whom also showed concurrent reactivity to HL, but these cellular immune responses did not correlate with production of anti-LSP and/or anti-HL. The findings indicate that humoral immune responses to liver cell surface antigens are frequently triggered by hepatitis A and B viruses, possibly via induction of autoreactive, T-cell independent, liver antigen-specific B lymphocytes. These liver-specific autoreactions have the potential to contribute to hepatocellular damage in virus A and B hepatitis but it seems unlikely that autoimmunity plays a significant pathogenetic role in NANB viral infections.