Suppressive and pro-inflammatory roles for IL-4 in the pathogenesis of experimental drug-induced liver injury.

The pathogenesis of immune-mediated drug-induced liver injury (DILI) following halogenated anesthetics, carbamazepine or alcohol has not been fully elucidated. Detecting cytochrome P450 2E1 (CYP2E1) IgG4 auto-antibodies in anesthetic DILI patients suggests a role for IL-4 in this hapten-mediated process. We investigated IL-4-mediated mechanisms using our model of experimental DILI induced by immunizing BALB/c (WT) and IL-4(-/-) (KO) mice with S100 liver proteins covalently modified by a trifluoroacetyl chloride (TFA) hapten formed following halogenated anesthetic metabolism by CYP2E1. WT mice developed more hepatitis, TFA and S100 antibodies (p<0.01), as well as T-cell proliferation to CYP2E1 and TFA (p<0.01) than KO mice. Additionally, WT CD4(+) T cells adoptively transferred hepatitis to naïve Rag(-/-) mice (p<0.01). Pro-inflammatory cytokines were expectedly decreased in TFA hapten-stimulated KO splenocyte supernatants (p<0.001); however, IL-2 and IFN-gamma (p<0.05), as well as IL-6 and IL-10 (p<0.001) levels were elevated in CYP2E1-stimulated KO splenocyte supernatants, suggesting dual IL-4-mediated pro-inflammatory and regulatory responses. Anti-IL-10 administered to KO mice increased hepatitis, TFA and CYP2E1 antibodies in KO mice confirming a critical role for IL-4. This is the first demonstration of dual roles for IL-4 in the pathogenesis of immune-mediated DILI by suppressing auto-antigen-induced regulatory responses while promoting hapten-induced pro-inflammatory responses.

IP-10 protects while MIP-2 promotes experimental anesthetic hapten - induced hepatitis.

MIP-2 and IFN-gamma inducible protein-10 (IP-10) and their respective receptors, CXCR2 and CXCR3, modulate tissue inflammation by recruiting neutrophils or T cells from the spleen or bone marrow. Yet, how these chemokines modulate diseases such as immune-mediated drug-induced liver injury (DILI) is essentially unknown. To investigate how chemokines modulate experimental DILI in our model we used susceptible BALB/c (WT) and IL-4-/- (KO) mice that develop significantly reduced hepatitis and splenic T cell priming to anesthetic haptens and self proteins following TFA-S100 immunizations. We detected CXCR2+ splenic granulocytes in all mice two weeks following immunizations; by three weeks, MIP-2 levels (p<0.001) and GR1+ cells were elevated in WT livers, suggesting MIP-2-recruited granulocytes. Elevated splenic CXCR3+CD4+T cells were identified after two weeks in KO mice indicating elevated IP-10 levels which were confirmed during T cell priming. This result suggested that IP-10 reduced T cell priming to critical DILI antigens. Increased T cell proliferation following co-culture of TFA-S100-primed WT splenocytes with anti-IP-10 (p<0.05) confirmed that IP-10 reduced T cell priming to CYP2E1 and TFA. We propose that MIP-2 promotes and IP-10 protects against the development of hepatitis and T cell priming in this murine model.

A novel model of drug hapten-induced hepatitis with increased mast cells in the BALB/c mouse.

Clinical evidence suggests that idiosyncratic hepatitis following administration of halogenated volatile anesthetics is mediated by autoimmune responses. No murine model to study mechanisms of anesthetic-induced or any other form of drug-induced idiosyncratic hepatitis exists. Anesthetics are believed to trigger hepatitis by covalently linking a trifluoroacetyl (TFA) chloride hapten to hepatic proteins, forming haptenated self-proteins. To test this hypothesis, we developed a hapten-induced model of hepatitis by immunization with syngeneic S100 liver proteins covalently coupled to TFA (TFA-S100). We found that TFA-S100 induced hepatitis was more severe than disease induced by S100 plus adjuvants or by the adjuvant alone and was characterized by neutrophil, mast cell, and eosinophil infiltration. TFA-specific IgG1 antibodies directly correlated with hepatitis, whereas S100 autoantibodies did not. TNF-alpha, IL-1beta, and IL-6 released from splenocytes collected 2 weeks after TFA-S100 inoculation were increased resembling the elevated serum cytokines reported in patients with autoimmune hepatitis (AIH). Three weeks after inoculation, the peak of hepatitis, we noted decreased numbers of Kupffer cells and lower levels of IL-6 and IL-10 in the liver, cytokines produced by Kupffer cells. This is the first report, to our knowledge, of a hapten-induced model of hepatitis with immune and autoimmune features.

Comparative immunoreactivity of anti-trifluoroacetyl (TFA) antibody and anti-lipoic acid antibody in primary biliary cirrhosis: searching for a mimic.

Previous studies documenting the existence of cross-reactivity between the lipoated (but not unlipoated) forms of the inner lipoyl domain (E2L2) of PDC-E2 [the major autoantigen in Primary biliary cirrhosis (PBC)] and trifluoroacetylated (TFA) proteins, led us to hypothesize that PBC may be due to an initial insult with an environmental agent that cross-reacts with TFA. Therefore, we performed a comparative study of the reactivity of rabbit anti-TFA antibody and anti-lipoic acid (LA) antibody against the mitochondrial autoantigens of human PBC and various TFA and LA conjugated proteins. Whereas both anti-TFA and anti-LA reacted with PDC-E2, the wild-type lipoated form of E2L2, OGDC-E2, E3-BP and LA-KLH, neither reacted with BCOADC-E2 or the non-lipoated form of E2L2. Of interest was that while anti-TFA reacted with PDC-E2, TFA-RSA and LA-KLH, it failed to inhibit PDC-E2 enzyme function. In contrast, anti-LA demonstrated cytoplasmic and mitochondrial staining, and inhibited PDC enzyme activity. Hence, although considerable cross reactivity exists between anti-TFA and anti-LA, the molecular nature of the interaction is clearly different. One of 14 PBC sera reacted weakly with TFA-albumin, whereas four of 14 PBC sera reacted with LA-KLH. Immunohistochemically, both anti-TFA and anti-LA antibodies reacted focally with periportal hepatocytes and bile ducts in both PBC and controls. However, anti-LA produced much stronger focalized staining of the bile ducts of diseased liver. This study suggests that while anti-TFA antibody recognizes lipoic acid-linked enzymes and proteins, the epitope recognized differs from that of anti-LA antibody and PBC autoantibodies. It is unlikely that a response to TFA is the triggering event in PBC. Anti-LA antibodies share a higher degree of similarity to PBC sera providing suggestive evidence that anti-LA antibodies or anti-LA like antibodies (mimotopes) may help define the initiator of the autoimmune response.

Kupffer cells from halothane-exposed guinea pigs carry trifluoroacetylated protein adducts.

The anesthetic, halothane, is bioactivated by the liver cytochrome P450 system to trifluoroacetyl-chloride, which can readily acylate liver protein. Covalent binding of the trifluoroacetyl moiety may result in hapten formation leading to the induction of an immune response and ultimately halothane hepatitis. In this study the presence of trifluoroacetylated-protein adducts in Kupffer cells was investigated to learn how the immune system might come in contact with the proteins. Guinea pigs were exposed to 1.0% halothane, 40% oxygen for 4 h. Kupffer cells were isolated on days 1 through 9 post-exposure, by liver perfusion and purification by elutriation. Using gel electrophoresis and Western blotting techniques, it has been demonstrated that Kupffer cells obtained from halothane-treated guinea pigs, do carry trifluoroacetyl-protein adducts as recognized by an anti-trifluoroacetyl-rabbit serum albumin antibody. Apparent molecular weights of polypeptides bound by trifluoroacetyl were of a wide range, 25-152 kDa. Bands were most prominent in the larger Kupffer cells with more appearing at lower molecular weights. Trifluoroacetyl-protein adducts were not detected in lung, spleen, lymph node or peripheral blood macrophages. This work suggests a role for Kupffer cells in the presentation of altered proteins in the liver to cells of the immune system.

Demonstration of a cellular immune response in halothane-exposed guinea pigs.

Halothane hepatitis is considered to be a result of an idiosyncratic autoimmune reaction brought about by the formation of neoantigens that have been generated by covalent binding of halothane biotransformation intermediates. The guinea pig is being examined as an animal model to investigate an immune-mediated mechanism for halothane hepatotoxicity. Male Hartley guinea pigs were exposed to 1% halothane/40% oxygen for 4 hr, three times with 40-day intervals. Kupffer cells and splenocytes were isolated from animals on various days after each halothane exposure. Splenocytes were cocultured in a lymphocyte transformation test with various concentrations of TFA(trifluoroacetylated)-antigens for 7 days and proliferation was measured by 3H-thymidine incorporation. In a second experiment, Kupffer cells were cocultured with autologous as well as allogeneic splenocytes with or without concanavalin A to determine whole cell sensitization and accessory function by Kupffer cells from treated animals. A 4-fold increase in splenocyte proliferation occurred in response to TFA-guinea pig albumin. No significant increase in proliferation could be detected with TFA-lysine or guinea pig albumin. A 14-fold increase in splenocyte proliferation also occurred in response to Kupffer cells from halothane-exposed animals. Autologous splenocytes demonstrated more of a response from treated versus control animals, indicating possible involvement of major histocompatibility complex II antigens. These results indicate recognition of TFA-antigens and Kupffer cells as antigen-presenting cells in halothane-exposed guinea pigs. This study provides good evidence that a cellular immune response is involved in the guinea pig after halothane exposure.

The lipoic acid containing components of the 2-oxoacid dehydrogenase complexes mimic trifluoroacetylated proteins and are autoantigens associated with halothane hepatitis.

Anti-CF3CO antibodies, monospecific toward trifluoroacetylated proteins (CF3CO-proteins), which are elicited in experimental animals and humans exposed to the anesthetic agent halothane, cross-react with an unknown protein of approximately 52 kDa, constitutively expressed in tissues of experimental animals and humans not previously exposed to the agent. Using anti-CF3CO antibody, the protein(s) of 52 kDa could be immunoprecipitated from solubilized rat heart homogenate. Two-dimensional gel electrophoretic analysis revealed the presence of distinct major (P1, P2) and minor (P3, P4, P5) protein components with apparent molecular masses of 52 kDa. From each of the components P1 and P2, the amino acid sequences of three peptides were determined and found to exhibit 100% identity with the corresponding amino acid sequences of the E2 subunit of the rat 2-oxoglutarate dehydrogenase complex (OGDC). Additionally to the E2 subunit of OGDC, anti-CF3CO antibody also recognized on immunoblots the purified E2 subunit of the branched chain 2-oxoacid dehydrogenase complex (BCOADC) and protein X, a constituent of the pyruvate dehydrogenase complex (PDC), in a manner sensitive to competition by N6-(trifluoroacetyl)-L-lysine (CF3CO-Lys), 6(RS)-lipoic acid, and N6-(6(RS)-lipoyl)-L-lysine (lipoyl-Lys). Furthermore, a discrete population of autoantibodies was identified in sera of patients with halothane hepatitis which could not discriminate between the lipoylated target epitope present on the E2 subunit of OGDC and epitopes on CF3CO-RSA, used as model for CF3CO-proteins. These data suggest that the autoantigenicity of these proteins in halothane hepatitis is based on the molecular mimicry of CF3CO-Lys by lipoic acid, the prosthetic group common to protein X and the E2 subunits of OGDC and BCOADC.

Molecular mimicry in halothane hepatitis: biochemical and structural characterization of lipoylated autoantigens.

Exposure of human individuals to halothane causes, in about 20% of all cases, a mild transient form of hepatotoxicity. A small subset of exposed individuals, however, develops a potentially severe and life-threatening form of hepatic damage, coined halothane hepatitis. Halothane hepatitis is thought to have an immunological basis. Sera of afflicted individuals contain a wide variety of autoantibodies against hepatic proteins, in both trifluoroacetylated form (CF3CO-proteins) and, at least in part, in native form. CF3CO-proteins are elicited in the course of oxidative biotransformation of halothane, and include the trifluoroacetylated forms of protein disulfide isomerase, microsomal carboxylesterase, calreticulin, ERp72, GRP 78, and ERp99. Current evidence suggests that CF3CO-proteins arise in all halothane-exposed individuals; however, the vast majority of individuals appear to immunochemically tolerate CF3CO-proteins. The lack of immunological responsiveness of these individuals towards CF3CO-proteins might be due to tolerance, induced through the occurrence of structures in the repertoire of self-determinants, which immunochemically and structurally mimic CF3CO-proteins very closely. In fact, lipoic acid, the prosthetic group of the constitutively expressed E2 subunits of the family of mammalian 2-oxoacid dehydrogenase complexes and of protein X, was shown by immunochemical and molecular modelling analysis to be a perfect structural mimic of N6-trifluoroacetyl-L-lysine (CF3 CO-Lys), the major haptenic group of CF3CO-proteins. As a consequence of molecular mimicry, autoantibodies in patients' sera not only recognize CF3CO-proteins, but also the E2 subunit proteins of the 2-oxoacid dehydrogenase complexes and protein X, as autoantigens associated with halothane hepatitis. Furthermore, a fraction of patients with halothane hepatitis exhibit irregularities in the hepatic expression levels of these native, not trifluoroacetylated autoantigens. Collectively, these data suggest that molecular mimicry of CF3CO-Lys by lipoic acid, or the impairment thereof, might play a role in the susceptibility of individuals for the development of halothane hepatitis.

Trifluoroacetylation potentiates the humoral immune response to halothane in the guinea pig.

Halothane hepatitis appears to result from an inappropriate immune response to the products of halothane metabolism. Attempts to produce an animal model for halothane hepatitis have been largely unsuccessful. Although guinea pigs produce neoantigens following treatment with halothane, the subsequent antibody response is weak, possibly accounting for the failure to produce halothane hepatitis in these animals. In order to increase the antibody response to halothane neoantigens, three methods for trifluoroacetylating proteins were used. Guinea pigs were either treated with S-ethylthiotrifluoroacetate, autologous lymphocytes trifluoroacetylated ex vivo, or immunized with trifluoroacetylated mycobacterial protein, followed by exposure to halothane, and examined for anti-halothane metabolite antibodies (anti-TFA antibodies). Animals treated with S-ethylthiotrifluoroacetate developed anti-TFA antibodies, and following exposure to halothane exhibited an enhanced antibody response. Treatment with trifluoroacetylated lymphocytes also resulted in an enhanced anti-TFA antibody response following halothane exposure. Immunization with trifluoroacetylated mycobacterial proteins resulted in very high anti-TFA antibody titers. However, subsequent exposure to halothane had no observable effect on specific antibody titers. Exposure to halothane, regardless of treatment, resulted in the production of anti-microsomal protein antibodies. Signs of halothane hepatitis were not observed, indicating that enhancement of the humoral immune response does not appear to be sufficient for production of halothane hepatitis.

Human hepatocytes express trifluoroacetylated neoantigens after in vitro exposure to halothane.

Biotransformation of anaesthetic halothane by cytochrome P450-dependent monooxygenases resulted in the production of reactive intermediate trifluoroacetyl (TFA) halide, capable of covalently binding to hepatocyte proteins. TFA-modified liver proteins can act as antigens and are implicated in the pathogenesis of halothane hepatitis in humans. The aim of this study was to investigate the formation of TFA-neoantigens in halothane-treated primary cultures of adult human hepatocytes and to evaluate the usefulness of this in vitro model for studying immune-mediated halothane hepatotoxicity. Cultured human hepatocytes were incubated with halothane under constant temperature, atmosphere and anaesthetic concentration conditions. The results obtained show that halothane-treated hepatocytes isolated from seven different donors produced TFA-antigens as detected by immunocytochemical and western immunoblot analysis using rabbit anti-TFA antiserum. TFA-adducts were localized mainly in the endoplasmic reticulum and in small amounts on the plasma membrane of parenchymal cells. By immunoblotting, several neoantigens, with molecular masses from 42 to 100 kDa, were detected in halothane-exposed hepatocytes. These observations are consistent with the formation of TFA-adducts through metabolism of the anaesthetic and suggest that primary cultures of human hepatocytes represent a suitable in vitro model to study the pathogenesis of immune-mediated halothane hepatotoxicity.

Absence of anti-trifluoroacetate antibody after halothane anaesthesia in patients exhibiting no or mild liver damage.

It has been shown that the circulating antibodies, which bind to rat hepatic microsomal proteins obtained after in vivo exposure to halothane, are detectable by immunoblotting in patients with "halothane hepatitis (HH)," and that rabbit immunized anti-sera against trifluoroacetylated rabbit serum albumin (TFA-RSA) recognizes rat microsomal distorted polypeptides in almost the same way as do sera from patients with HH. In this paper, we report first the development of a novel method of synthesizing TFA-RSA using p-nitrophenyl TFA, and second the results of tests for circulating anti-TFA antibodies in the serum of 86 patients who had received halothane anaesthesia and developed no (67 patients) or mild (19 patients, the maximum activity of serum alanine aminotransaminase 519 IU.L-1) liver damage. Serum was selected from stored sera of post-transfusion patients. The new method of synthesizing TFA-RSA was convenient and was able to be done at neutral pH. Rabbit sera obtained after immunization with the newly synthesized TFA-RSA recognized the same polypeptides (109 kDa, 92 kDa, 80 kDa, 76 kDa, 64 kDa and 59 kDa) as the established anti-sera against TFA-RSA, and these reactions were inhibited in the presence of TFA-lysine. Circulating antibodies were not detected in our patients who had developed no or mild liver damage. The present finding supports the hypothesis that the appearance of circulating antibodies against microsomal distorted proteins are specific to patients with HH. Furthermore, we have shown here that the halothane-induced mild increase in ALT activity is not associated with the appearance of those circulating antibodies, supporting the pathophysiological difference between HH and halothane-induced mild hepatic damage.

Halothane hepatitis patients have serum antibodies that react with protein disulfide isomerase.

Clinical and laboratory evidence suggests that the fulminant liver failure sometimes associated with the inhalation anesthetic halothane may be an immune-mediated toxicity. Most importantly, the vast majority of patients with a clinical diagnosis of halothane hepatitis have serum antibodies, which react with one or more specific liver microsomal proteins that have been covalently altered by the trifluoroacetyl chloride metabolite of halothane. The serum antibodies are specific to halothane hepatitis patients and are not seen in sera of patients with other types of liver pathology. In this study, a 57-kD trifluoroacetylated liver microsomal neoantigen associated with halothane hepatitis and native 57-kD protein were purified from liver microsomes of halothane-treated and -untreated rats, respectively. When the purified trifluoroacetylated 57-kD and native 57-kD proteins were used as test antigens in an enzyme-linked immunosorbent assay, serum antibodies from halothane hepatitis patients (n = 40) reacted with both of these proteins to a significantly greater extent than did serum antibodies from control patients (n = 32). On the basis of its apparent monomeric molecular mass, isoelectric point and NH2-terminal amino acid and tryptic peptide sequences, the 57-kD protein has been identified as rat liver protein disulfide isomerase. Antibodies raised against rat liver protein disulfide isomerase also reacted with a protein of approximately 58-kD in human liver microsomes. The results of this investigation suggest that trifluoroacetylated protein disulfide isomerase is one of the immunogens associated with halothane hepatitis. In certain patients it might lead either to specific antibodies or, possibly, to specific T cells, which could be responsible for halothane hepatitis.

Screening for antibodies associated with halothane hepatitis.

The diagnosis of halothane hepatitis (HH) may be assisted by detection of antibodies reacting to trifluoroacetylated proteins (anti-TFA antibodies). An enzyme-linked immunosorbent assay (ELISA) utilizing trifluoroacetylated rabbit serum albumin (TFA-RSA) as antigen detected anti-TFA antibodies in 67% of sera from patients for whom a clinical diagnosis of HH was made. Anti-TFA antibodies were detected in 33% of sera when using an ELISA with liver microsomal protein from halothane-treated rabbits as antigen. Absorption of the sera with untreated rabbit liver microsomal protein before using the microsomal protein ELISA resulted in detection of anti-TFA antibodies in 42% of sera. Using the presumptive hapten N-epsilon-trifluoroacetyl-1-lysine to block antibody binding in an ELISA resulted in positive detection in 50% of sera: the results did not always agree with the other ELISA methods. The TFA-RSA ELISA was the most sensitive method and, combined with the TFA-lysine blocking ELISA, resulted in 92% of sera from HH patients testing positive for HH-associated antibodies.

Detection of cysteine conjugate metabolite adduct formation with specific mitochondrial proteins using antibodies raised against halothane metabolite adducts.

Antibodies raised against halothane metabolite adducts cross-react with S-(1,1,2,2-tetrafluoroethyl)-L-cysteine (TFEC) and S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine metabolite adducts. Using these antibodies in immunohistochemical experiments, metabolite binding was localized to the damaged areas of the proximal tubule after treatment of male rats with TFEC. Immunoblot analysis of subcellular fractions of rat kidney tissue after in vivo treatment with TFEC revealed a high specificity for binding of metabolites to proteins of the mitochondrial fraction. These proteins may represent target molecules which play a role in cysteine conjugate induced nephrotoxicity.

Halothane metabolism: immunochemical evidence for molecular mimicry of trifluoroacetylated liver protein adducts by constitutive polypeptides.

A monoform antibody [anti-TFA antibody] against TFA-protein adducts (TFA-adducts) was obtained by affinity purification of a polyclonal antiserum, raised in rabbits against TFA-rabbit serum albumin, on a N-epsilon-TFA-L-lysine matrix coupled to Affi-Gel 102. The anti-TFA antibody did recognize TFA-adducts of distinct molecular mass on Western blots of hepatocyte homogenates or microsomal membranes obtained from rats pretreated with halothane. The anti-TFA antibody also recognized cross-reactive polypeptides with apparent molecular masses of 52 kDa and 64 kDa on Western blots of hepatocyte homogenates obtained from rats not treated with halothane or metabolites thereof. The 52-kDa and 64-kDa cross-reactive polypeptides were localized in the 3,000 x g particulate fraction of liver homogenates. Recognition, on Western blots, of TFA-adducts and both the 52-kDa and 64-kDa cross-reactive polypeptides by anti-TFA antibody was sensitive to competition by N-epsilon-TFA-L-lysine (IC50 less than 100 microM) and N-epsilon-acetyl-L-lysine (IC50 approximately 10 mM). Treatment with piperidine (1 M) did abolish the recognition of TFA-adducts but not that of the 52-kDa and the 64-kDa cross-reactive polypeptides by anti-TFA antibody on Western blots. In antibody-exchange experiments, anti-TFA antibody was affinity-adsorbed on Western blots to the 52-kDa or the 64-kDa cross-reactive polypeptides of the rat heart, followed by spontaneous transfer to target TFA-adducts present on Western blots of rat liver microsomal membranes. The majority of these target TFA-adducts were recognized by anti-TFA antibody transferring from the source 52-kDa or 64-kDa cross-reactive polypeptides. When examined up to 10 days after exposure of rats to a single dose of halothane, no influence on the constitutive level of expression, in the liver, of either cross-reactive polypeptide was observed. In contrast, TFA-adducts were persistent for greater than 90 hr but less than 10 days. In addition to the liver, the 52-kDa and the 64-kDa cross-reactive polypeptides were prominently expressed in the heart and the kidney and, to a much lesser degree, in the spleen, the thymus, the lung, and skeletal muscle of the rat. Considerable variation in the level of expression of the 52-kDa and the 64-kDa cross-reactive polypeptides was recognized in livers of the six human individuals tested so far.(ABSTRACT TRUNCATED AT 250 WORDS)

Antibodies raised against trifluoroacetyl-protein adducts bind to N-trifluoroacetyl-phosphatidylethanolamine in hexagonal phase phospholipid micelles.

The delayed fulminant form of halothane hepatotoxicity is thought to be triggered by an immune response to haptenic adducts formed by a metabolite, trifluoroacetyl chloride. In this study we demonstrate that antibodies purified from the sera of rabbits sensitized to a trifluoroacetyl-protein adduct will cross-react with a trifluoroacetyl-phosphatidylethanolamine adduct. Trifluoroacetyl adducts of both rabbit serum albumin (TFA-RSA) and dioleoylphosphatidylethanolamine (TFA-DOPE) were prepared. The TFA-RSA was coupled to an Affigel-10 affinity column to purify hapten-selective immunoglobulin (Ig) G antibodies (anti-TFA-RSA IgG) from the sera of rabbits given i.m. injections of TFA-RSA. The TFA-DOPE was purified by high-performance liquid chromatography and the structure confirmed with direct chemical ionization mass spectrometry. Lamellar liposomes containing a mixture of 5% TFA-DOPE, 71% DOPE and 24% dioleoyl-phosphatidylcholine, as well as hexagonal phase micelles containing 5% TFA-DOPE and 95% DOPE, were prepared by sonication. Anti-TFA-RSA IgG antibodies were added to each of these lipid mixtures for 30 min, fluorescein-conjugated goat-antirabbit IgG antibodies were added next for an additional 30 min and then binding of anti-TFA-RSA IgG antibodies to TFA-DOPE was quantified by flow cytometry. Anti-TFA-RSA IgG antibodies bound to TFA-DOPE only when it was incorporated into hexagonal phase micelles. These findings suggest that TFA-phosphatidylethanolamine adducts that reside in nonlamellar domains on the hepatocyte surface could be recognition sites for anti-TFA-adduct antibodies and potentially participate in immune-mediated hepatotoxicity.