Posted 11/12/2002

Immunopathogenesis of Primary Biliary Cirrhosis

from Seminars in Liver Disease

By:

M. Eric Gershwin, M.D.

Division of Rheumatology, Allergy and Clinical Immunology, School of Medicine, University of California at Davis, Davis, California

Emmet B. Keeffe, M.D.

Liver Transplant Program, Stanford University Medical Center, Palo Alto, California

Akiyoshi Nishio, M.D.

Department of Gastroenterology, Tenri Hospital, Nara, Japan

Abstract and Introduction

AbstractAlthough the autoantigens of antimitochondrial antibodies (AMA) have been defined and epitope mapped for both autoreactive B and T cells, the pathogenesis of primary biliary cirrhosis (PBC) still remains a mystery. The data gathered so far address several important aspects of this intriguing puzzle. First, biliary epithelial cells (BECs) seem to be immunologically active because they express molecules such as major histocompatibility complex (MHC) antigens, and adhesion and costimulatory molecules. Second, although pyruvate dehydrogenase complex (PDC)-E2, the major autoantigen in PBC, is upregulated in BECs when examined immunohistochemically, this abnormal staining seems to be secondary to immune complexes of AMA bound to PDC-E2 present in the BECs. Third, in addition to CD4+ T cells, CD8+ T cells also recognize the inner lipoyl domain of PDC-E2. Fourth, modification of mitochondrial antigens by xenobiotics may lead to the induction of the disease. These findings help to clarify the pathogenic mechanism of PBC and suggest that (l) induction may be secondary to a primary response to a xenobiotic that is normally metabolized in an estrogen-dependent pathway and (2) pathology is mediated by and orchestrated by a highly directed and specific CD4, CD8 and autoantibody response to the lipoyl domain of the mitochondrial autoantigens, with tissue destruction based on the immunoglobulin A (IgA) receptor, apoptosis, and the mucosal organization of biliary and salivary duct cells.

Objectives: Upon completion of this article the reader should be able to (1) understand the immune response in primary biliary cirrhosis, (2) describe the significance of a positive antimitochondrial antibody, and (3) understand current theories of the pathogenesis of disease.

Accreditation: Tufts University School of Medicine is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. TUSM takes full responsibility for the content, quality, and scientific integrity of this continuing education activity.

Credit: Tufts University School of Medicine designates this education activity for a maximum of 1.0 hour credit toward the AMA Physicians Recognition Award in category one. Each physician should claim only those hours that he/she actually spent in the educational activity.

IntroductionNormally, the immune system distinguishes self from nonself and reacts against all foreign antigens. If there is immune reactivity against self-components, autoimmune diseases, either organ specific or systemic, may occur. BECs are major targets in several liver diseases induced by breakdown of self-tolerance, resulting in the development of "vanishing bile duct syndromes" that include PBC. PBC is a chronic cholestatic liver disease characterized by progressive destruction of small and mid-sized intrahepatic bile ducts, leading to cirrhosis and liver failure. It has been considered an autoimmune disease based on the lymphoid infiltration noted in portal tracts and the high prevalence of disease-specific autoantibodies; these include not only AMA but also antibodies to components of the nuclear pore complex.

Although the autoantigens for AMA have been molecularly identified and epitope mapped for autoreactive B and T cells, the mechanism by which self-tolerance is broken and autoimmunity causes tissue damage to only BECs remains enigmatic. One attractive hypothesis is molecular mimicry, which has been implicated in other autoimmune diseases.[1-3] In fact, molecular mimicry of autoepitopes by either peptides of microorganisms or xenobiotics in the setting of cryptic T cell epitopes and degeneracy of T cell receptors (TCRs) is a likely mechanism for disruption of self-tolerance.[4]

Antimitochondrial and Other Autoantibodies

Antimitochondrial antibodies are found in approximately 95% of patients with PBC and are considered as the serological hallmark of the disease. The targets of AMA in PBC sera are members of an enzyme family, the 2-oxoacid dehydrogenase complexes (2-OADC), which are located on the inner membrane of the mitochondria and catalyze the oxidative decarboxylation of various -keto acid substrates. Components of 2-OADC include the E2 subunit of PDC (PDC-E2), the E2 subunit of the 2-oxoglutarate dehydrogenase complex (OGDC-E2), the E2 subunit of the branched chain 2-oxoacid dehydrogenase complex (BCOADC-E2), and the dihydrolipoamide dehydrogenase binding protein (E3BP).^[5-13] The most predominant reactivity of AMA in sera from PBC patients is directed against PDC-E2. Reactivity against OGDC-E2 and BCOADC-E2 is lower, around 50 to 70%. Antibodies to PDC-E1 are present in lower titers. Approximately 10% of patients react only to OGDC-E2 or BCOADC-E2, or both (Table 1). Antimitochondrial reactivity is usually observed against some, or even all, of the 2-OADC, but serological cross- reactivity is only found between PDC-E2 and E3BP. Several studies using oligopeptides or recombinant proteins have shown that the predominant epitopes of PDC-E2, E3BP, OGDC-E2, and BCOADC-E2 are all conformational lipoate-binding sites.^[14-17]

The highly conserved structure in the E2 subunits of 2-OADC and their lipoyl domains suggest that lipoic acid may be a part of the immunodominant epitope. The contribution of lipoic acid to the reactivity of the autoantibodies has been studied, and the results are somewhat contradictory.^[18-20] However, the data clearly show that AMA are capable of binding to both lipoylated and unlipoylated PDC-E2.

In addition to AMA, PBC sera have been reported to contain other disease-specific autoantibodies, including antinuclear antibodies (ANA). ANA have been identified in more than 50% of patients with PBC. These include autoantibodies against the nuclear pore protein, gp210, which is a 210-kD transmembrane glycoprotein believed to be involved in the attachment of pore complex constituents within the nuclear membrane. Antibodies directed against gp210 are found in about 25% (range, 10 to 40%) of patients with AMA-positive PBC and up to 50% of those with AMA-negative PBC. The disease specificity for the detection of such autoantibodies by immunoblotting is more than 90%.^[21] In contrast to AMA, antibodies against gp210 been proposed to have prognostic significance, but the data are unclear. Other autoantigens include the nuclear pore protein, p62, which is another nuclear pore glycoprotein specifically recognized by PBC sera and present in 32% of patients with PBC.^[22] The presence of anti-gp210 and p62 antibodies in PBC sera appears to be mutually exclusive. In about 20 to 30% of PBC patients, autoantibodies are directed against Sp100, a nucleoprotein of 100 kD molecular weight and appear to exhibit a high specificity for PBC.^[23] Finally, few PBC sera show reactivity against the inner nuclear membrane protein, lamin B receptor (LBR), which binds both nuclear lamins and double-stranded DNA. Although less than 1% of PBC sera contain antibodies to LBR, there does appear to be disease specificity for such reactivity.^[24] In patients with negative AMA, the presence of PBC-associated ANA may be the only seroimmunological clue for establishing the diagnosis of PBC.

T Cell Response in PBC

The particular specificity of bile duct destruction, associated with the lymphoid infiltration in the portal tracts and the aberrant expression of MHC class II antigen on biliary epithelium have suggested that BECs are the target of an intense autoimmune response. The profile of periductal inflammatory cells may change according to the phase of cholangitis and also to the histological stage of PBC. However, it has been speculated that autoreactive liver-infiltrating T cells mediate bile duct destruction through cytotoxicity or lymphokine production, or both. There is a controversy regarding whether CD4⁺ or CD8⁺ cells are the dominant cell type in portal tract lesions. Immunohistochemical studies of liver tissues in PBC reported a predominance of T cell infiltration, mainly CD3, CD4-positive T cells expressing the TCFR, particularly in and around the portal areas.^[25] Early studies focused on the phenotypic characterization of isolated liver-infiltrating T cells and demonstrated a marked enrichment for CD8⁺ cytotoxic T lymphocytes (CTLs) in patients with PBC.^[26] Later, several studies were reported on analysis of T-cell lines that proliferate in the presence of putative mitochondrial antigens.

Van de Water et al^[27] obtained T-cell lines from liver biopsies of patients with PBC and showed that cloned T-cell lines specifically responded when stimulated with PDC-E2 or BCOADC-E2 but not with control proteins. Shimoda et al^[28] derived six T cell clones specific for PDC-E2 from the peripheral blood mononuclear cells (PBMC) of four patients with PBC using a panel of overlapping peptides spanning the full-length PDC-E2. The surface phenotypes of these T cell clones were CD4⁺, CD45RO⁺, and TCFR. The minimal epitopes of these T cell clones all mapped to the amino acid sequence 163-176 (GDLLAEIETDKATI) within the inner lipoyl domain of PDC-E2. Moreover, all T cell clones responded to the sequence 36-49 (GDLIAEVETDKATV), which corresponds to the outer lipoyl domain of PDC-E2. These two reactive peptides share a common amino acid motif E, D, and K at 170, 172, and 173, respectively (ExDK sequence motif), whereas other peptides that share amino acid sequence homology with PDC-E2 but do not have the ExDK motif did not induce proliferation (Table 2). One particular cloned T-cell line also cross-reacted with an exogenous antigenic peptide (EQSLITVEGDKASM), homologous to part of the sequence of PDC-E2 from Escherichia coli, which has an ExDK motif. However, none of these PDC-E2-specific T cell clones cross-reacted with the HLA-DR chain peptide 82-95 (QGALANIAVDKANL), or the human glycogen phosphorylase peptide 354-367 (LVDLERMDWDKAWD), despite both having some amino acid homology with PDC-E2 but lacking the requisite ExDK motif. Furthermore, Shimoda et al^[29] also demonstrated, specifically in PBC, a high precursor frequency of PDC-E2 163-176-reactive T cells in the hilar lymph nodes and livers of patients, 100- to 150-fold greater than that for PBMC from the same patients. In addition, there was cross-reactivity of PDC-E2 163- 176-specific T cell clones with the OGDC-E2 peptide 100-113 (DEVVCEIETDKTSV), thereby identifying the epitope motif "ExETDK" common to the two E2 subunits.

TCR analysis of these PDC-E2-specific T cells revealed significant diversity.^[30] In contrast, in the third complementarity-determining region (CDR) 3, G was frequently found and the GxG or GxS sequence motif was found in all T cell clones, suggesting the preferential usage of limited motif in CDR3 region in PDC-E2 163-176-specific T cells.^[31] However, the results of T cell repertoire studies have largely been obtained from small number of clones and limit any conclusion.

There is very limited information on autoantigen-specific CTL responses available, compared with what is available concerning autoreactive CD4⁺ T cell responses. Kita et al^[32] have recently identified an HLA-A2- restricted CTL epitope of the E2 component of PDC-E2 using PBMC from patients with PBC. This peptide of the amino acid sequence 159-167 (KLSEGDLLA) of PDC-E2, a region very close to the epitope recognized by MHC class II-restricted CD4⁺ cells and by antibody, induces specific MHC class I- restricted CD8⁺ CTL lines from 10/12 HLA-A2+ PBC patients, but not controls, after in vitro stimulation with antigen-pulsed dendritic cells (DCs) (Table 2). PDC-E2-specific CTLs could also be generated by pulsing DCs with full-length recombinant PDC-E2 protein. Furthermore, using soluble PDC-E2 complexed with either PDC-E2-specific human monoclonal antibodies (mAbs) or affinity-purified autoantibodies against PDC-E2, the generation of PDC-E2-specific CTLs occurred at 100-fold and 10-fold less antigen concentration, respectively, compared with soluble antigen alone. They also found a 10-fold increase in the frequency of PDC-E2 159-167-specific CTLs in the liver as compared with the blood in PBC. In addition, the precursor frequency of the CTLs in blood was significantly higher in early stage PBC. Although the effector functions of the PDC-E2 159-167-specific CTLs are not well-understood, of interest is the fact that, after stimulation with the peptide, the response of PDC-E2 159-167-specific T cells is heterogeneous with respect to interferon- production. These data document the enrichment of autoantigen-specific CD8⁺ T cells in the PBC liver, suggesting that CD8⁺ T cells play a significant role in the immunopathogenesis of PBC.

Collectively, these data demonstrate that B cell, helper cell, and cytotoxic T cell epitopes all contain a shared peptide sequence within the inner lipoyl domain of PDC-E2. This finding is similar to that of multiple sclerosis, in which myelin basic protein (MBP)-specific autoantibodies react with an immunodominant MBP peptide that is also the autoepitope of the MBP-specific T cell clones.^[33] This would be consistent with more efficient uptake and processing of the autoantigen via the Ig receptor on autoimmune B cells and protection of the epitopes from degradation during antigen processing. It would be of interest to determine whether B and T cell epitopes of OGDC-E2 and BCOADC-E2 occupy similar sites on the molecule. If so, the immune response could be triggered by a sequence that bears a common structural motif among all three molecules; such findings would be encouraging for the hypothesis of disease initiation by molecular mimicry. Furthermore, the finding not only that autoantigen-antibody immune complexes can stimulate autoimmune T cells but also that presentation of the autoantigen is of a higher relative efficiency defines a unique role for AMA in the pathogenesis of PBC.

Biliary Epithelium

A puzzling feature of PBC, similar to certain other autoimmune diseases, is that the autoimmune attack is predominantly organ specific but the mitochondrial autoantigens are not tissue specific. Several mechanisms have been proposed regarding the effector mechanisms that mediate selective bile duct damage in PBC. Recent studies suggested that BECs are immunologically active because they express and secrete molecules associated with immune recognition of target cells such as MHC antigens, adhesion and costimulatory molecules, cytokines, and their receptors (Table 3).

The mechanism for the participation of HLA in the pathogenesis of PBC is not fully understood. However, recent understanding of the relationship between structure and function of MHC molecules favors the hypothesis that the upregulation of MHC antigens in presenting peptide antigens to the autoreactive T cells is important in disease susceptibility. Membrane expression of MHC class I expression increases on hepatocytes during cholestasis. It is also reported that the increase of chenodeoxycholic acid in cholestasis activates protein kinase C and protein kinase A, resulting in enhanced MHC class I molecules on the human hepatocytes in primary culture.^[34]

MHC class II (HLA-DR) and intercellular adhesion molecule (ICAM)-1, which are critical for the interaction and adhesion between lymphoid cells and target cells, have both been shown to be upregulated on the BECs in PBC.^[35,36] Work with isolated human BECs in vitro has demonstrated that MHC molecules and ICAM-1 are induced in these cells in response to proinflammatory cytokines.^[37] BECs also express high levels of several other adhesion molecules, such as vascular cell adhesion molecule (VCAM)-1 and leukocyte function associated antigen (LFA)-3, that are also important for mediating adhesion to lymphocytes.^[38,39] Although upregulation of MHC and adhesion molecules may be secondary to the inflammatory response, not a cause of the disease, these could serve to augment the local immune response and accelerate bide duct destruction.

Functional activation of lymphocytes requires a second costimulatory signal via interaction between CD28 and B7-1 (CD80) or B7-2 (CD86) in addition to the interaction between peptide-bound MHC molecules and their cognate TCR.^[40,41] The expression of CD80 and CD86 is mainly restricted to professional antigen presenting cells (APCs), such as dendritic cells. The ability of BECs to express either of the CD28 ligands, CD80, or CD86, or a combination of these, would allow them to act as APCs and thereby provoke and maintain a T cell-mediated response. However, there is some controversy concerning the expression of CD80/ 86 by BECs. Immunohistochemical studies showed expression of B7-1 and B7-2 on BECs in PBC,^[38,42] but comprehensive studies have failed to demonstrate CD80 or CD86 protein and mRNA level in cultured BECs.^[43] The fact that memory CD4⁺ T cells are less dependent on the interaction between CD28 and CD80/86 supports the view that although initial priming at early PBC stage may require such costimulation, the latter events may not.

BECs from PBC patients overexpress interleukin (IL) -6 and tumor necrosis factor (TNF)- compared with other hepatobiliary diseases.^[44] TNF receptor and IL-6 receptor were also detected on these damaged bile ducts, suggesting an autocrine effect.^[44] The increased expression of IL-6 and TNF- could affect the proliferation, maturation, and regulation of B-cell and T-cell lineage infiltrates around bile ducts.^[45] IL-6 promotes terminal differentiation of B cells leading to immunoglobulin secretion. TNF- has been shown to induce the expression of MHC class II, adhesion molecules, and a variety of other antigens on the bile ducts. It may also increase the cytotoxic activities of T cells. IL-6 may be responsible for BEC proliferation via an autocrine pathway^[46] and TNF- may be involved in the BEC damage. The autocrine role of TNF- on cell damage, including apoptosis, has already been shown in renal epithelial cells and hepatitis B and C virus-infected hepatocytes.^[47] TNF- is also known to disturb the barrier function of the bile ducts, which may lead to leakage of toxic substances into the bile, leading to a local inflammatory process and cholangitis.^[48]

Several studies have shown that autoantigens relevant to PBC may be present on the apical region of BECs of livers from PBC patients when examined immunohistochemically.^[49-51] Furthermore, this upregulation is present early in the natural history of PBC^[52] and is also present in BECs in allografts of patients with recurrent PBC after liver transplantation,^[53] suggesting a role for these antigens in the pathogenesis or progression, or both, of PBC. mAbs to OGDC-E2 and BCOADC-E2 also showed a disease-specific pattern of reactivity.^[54] It is not clear at present, however, whether the molecules detected at this special location are the mitochondrial antigens or molecules that are cross-reactive with the mitochondrial proteins. Enhanced synthesis, impaired degradation, and abnormal targeting of the mitochondrial proteins to the surface and apical regions of the BEC are each or all a possible explanation. Increased synthesis of these autoantigens in bile duct cells seems unlikely in patients with patients with PBC.^[55] An alternative possibility is a trafficking defect, in which PDC-E2 is aberrantly transported to the cytoplasmic membrane. Although such alteration might occur as a result of a point mutation in the mitochondrial presequence, in a way similar to the mistargeting of alanine,^[56] it is yet to be proved. Nonetheless, the abundance of such disease-specific determinants raises the possibility that novel molecules exist in the BECs of PBC, which cross-react with the mitochondrial proteins. Interestingly, AMA, especially of the IgA isotype, and the autoantigens PDC-E2, OGDC-E2, and BCOADC-E2 have all been readily detected in the bile of patients with PBC.^[57] Recent data suggest that this specific staining is due to a large fragment of PDC-E2, if not the entire molecules, because the disease-specific mAbs recognized the same region of PDC-E2 but not the same epitope.^[58] Taken together, it is likely that the abnormal staining on the BECs is secondary to immune complexes of IgA and IgG bound to PDC-E2 via the polyimmunoglobulin receptor on bile duct cells. It should be also noted that in apoptotic BECs, unlike control cells, PDC-E2 may not be degraded, resulting in a potential source of immunogenic PDC-E2 in patients with PBC.^[59]

Recent studies have shown that patients with PBC have reduced levels of the anion transporter AE2, both in lymphocytes and in bile duct cells.^[60] This transporter is implicated in the chloride bicarbonate exchanger, and it has been suggested that reduced activity of AE2 could contribute to cholestasis and extrahepatic manifestation such as the polyglandular features frequently seen in PBC.

There is evidence that BECs, in the presence of activated CTLs, undergo apoptosis in PBC using nick-end labeling methods for the detection of DNA fragmentation,^[61,62] although the mechanisms responsible for inducing such apoptosis are presently unclear. Related studies have shown not only increased expression of perforin and granzymes in PBC but also Fas (CD95) upregulated on the BEC membrane, associated with infiltration of CD95 ligand-expressing mononuclear cells.^[63] It is possible that both of these pathways are involved. Recently, toxic bile salts have been shown to induce apoptosis in cultured hepatocytes, suggesting that intracellular retention of the bile salts may contribute to apoptosis of hepatocytes, and possibly BECs, during cholestasis in PBC.^[64]

Genetic and Familial Analysis

It is well-established that genetic factors play an important role in modulating autoimmune diseases by conferring susceptibility to or providing protection from disease onset, progression, and severity. Although the prevalence of PBC in the general population ranges from 10 to 200 per million, depending on the population studied, the occurrence of PBC in first-degree relatives is nearly 100-fold higher than it is in the general population.^[65] When different generations are affected, the second generation usually presents earlier and tends to have more rapid progression.^[66] Furthermore, family members of PBC patients have positive AMA as well as other autoantibodies more frequently than controls do. These data provide clues regarding the etiology of the disease and support a genetic component.

The MHC molecules are highly polygenic and polymorphic. They are capable of presenting numerous peptide antigens to T cell receptors, thereby eliciting immune responses against infectious agents and altered self-molecules such as those expressed by certain tumors. Therefore, studies of genetic background have largely focused on the MHC in efforts to clarify the association, if any, with disease susceptibility. Although several earlier studies investigated a potential association, it is now widely accepted that MHC class I molecules are not associated with susceptibility to PBC.^[67] By contrast, the association with MHC class II molecules has been demonstrated not only in Caucasian but also in Asian populations. Based on serological HLA typing, some studies have shown that PBC is associated with HLA-DR3,^[68] -DR8,^[69-71] and possibly -DR4^[72] in Caucasians; an association with HLA-DR2 has been reported in Japanese.^[73]

Recently, molecular biology techniques have been introduced for HLA typing. Morling et al^[74] showed that susceptibility of PBC in Danish population was associated with the extended haplotype HLA-B8, DRB3*01/02/03, DQA1*0501, DQB1*0201. Begovich et al^[75] typed and compared the distribution of the polymorphic class II antigens. The study demonstrated the increased frequency in patients of DRB1*0801-DQA1*0401/0601-DQB1*4 haplotype and a decreased frequency of DRB1*1501-DQA1*0102-DQB1*0602 and DRB1*1302-DQA1*0102-DQB1*0604. In a study of German Caucasians, HLA-DPB1*0301 was clearly associated with PBC.^[76] Underhill et al^[77] reported weak association of PBC with DR8-DQB1*0402, which includes DPB1*0301 in the United Kingdom. In Japanese patients, an increased association with DQ3 was shown to be due to linkage disequilibrium with DPB1*0501, postulating that a single amino acid residue (leucine) at position 35 of the DPB polypeptide is crucial for disease susceptibility.^[78] These varied results among studies suggest that the overall association between MHC class II and PBC is weaker than what has been reported for other autoimmune diseases.^[79,80] Otherwise, it may reflect the difference in the sensitivity and methods of analysis, the difference in ethnic groups of the patients studied, and possible errors in diagnosis.

It has become increasingly evident that non-MHC genes are involved in modulating autoimmune diseases. The proinflammatory cytokine genes and genes for their receptors and inhibitors are responsible for the natural history of several autoimmune and infectious diseases. Although a number of such genes, including TNF-,^[81,82] IL-10,^[83] transporters associated with antigen processing protein (TAP)1 and TAP2,^[84] and national resistance-associated macrophage protein 1,^[85] have been investigated in patients with PBC, most studies are negative or controversial as to the role of polymorphism of cytokine genes in the pathogenesis of PBC.

These results might indicate that PBC patients are genetically highly heterogeneous or that susceptibility to PBC is determined by the interaction of numerous genes, or both, the contribution of each of which may only be determined by examining large number of PBC patients.

Infection and PBC

The characteristic histological feature of PBC is nonsuppurative destructive cholangitis that progresses to cirrhosis. Because granulomas, occurring either as aggregates of histiocytes or as noncaseating lesions adjacent to damaged bile ducts, are commonly found in PBC, it has been suggested that mycobacteria have a role in the pathogenesis of PBC.^[86] However, other studies have cast doubt on this hypothesis.^[87] These data are still inconclusive because the detection sensitivity of the polymerase chain reaction may not detect mycobacteria in archival liver tissue samples used, and if mycobacteria do have a role in the pathogenesis of PBC, they may only be necessary for initiating autoimmunity by molecular mimicry. Of note, treatment with the antimicrobial agent rifampicin has no effect on the course of the disease.^[88]

PDC-E2 is an evolutionally well-conserved molecule among species, especially at the lipoic acid binding sites. Sera from patients with PBC have been shown to react with both human and E. coli PDC-E2,^[18] and such reactivity of AMA to both human and bacterial molecules has stimulated speculation that PBC may be induced by exposure to enterobacterial antigens. In older studies, rabbits immunized with enterobacterial rough-colony (R) mutants generate AMA; enterobacterial wild forms, which are not R mutants, do not generate AMA.^[89] Although some studies have indicated that there is an increased prevalence of R forms of E. coli in the stools of patients with PBC,^[90] this has not been well-confirmed by others. Other evidence implicating E. coli includes an increased incidence of asymptomatic E. coli urinary tract infections (UTIs) in patients with PBC^[91] and the observation that patients with recurrent UTIs also have an increased incidence of positive AMA.^[92] R mutants also differ from wild-type enterobacteria mainly in the defective synthesis of the saccharide protein of lipopolysaccharide. This results in a different molecular organization of the cell membrane and the possibility that R mutants might possess different characteristics, especially those related to immunogenic properties. Phylogenetic conservation and the wide prevalence of bacterial and mitochondrial polypeptides suggest the presence of immune tolerance to such antigens in humans. Immune tolerance may be terminated when polypeptides are presented in a strong immunogenic form or when a defect in the immune system occurs. However, the hypothesis that infection of R forms of enterobacteria may induce AMA in PBC patients is not a proven one; the association between recurrent UTI and PBC has not been confirmed by all studies, and AMA found in patients with UTI are of low titer and of different specificity.^[18]

A preliminary study on experimentally infected laboratory animals with various Helicobacter strains suggested that bacterial translocation might occur from the stomach and the intestine to the liver. This may involve uptake and intracellular survival in macrophages and other professional phagocytes activated in the stomach during Helicobacter infections.^[93] The pathogenesis may then be similar to infection of the bile tree with bile-tolerant and bile-adapted strains of Salmonella typhi and other enteric organisms known to invade the bile tree in chronic infection and in carriers and to increase the risk of later development of primary biliary carcinoma.

Interestingly, injection of porcine bile into the murine biliary tree causes immune reactions with tissue damage, similar to PBC.^[93] It is tempting to speculate that various Helicobacter species and possibly other new pathogens in mice, rodents, and dogs may also invade the human bile tree and liver.

In PBC, there has been only one microbiological study of bacteria in bile, and it failed to culture any bacteria.^[94] However, using a PCR technique,^[95] Nilsson et al^[96] reported that bile and liver biopsy samples were positive for Helicobacter DNA in nearly half of patients with PBC and primary sclerosing cholangitis (PSC). Another PCR analysis of the bile demonstrated the presence of the 16S bacterial ribosomal RNA gene in gallbladder bile from PBC and control patients by PCR. Sequencing revealed that Staphylococcus aureus or other gram-positive bacteria were predominant (75%).^[97] In contrast, Tanaka et al^[98] were unable to find evidence of mycobacterial or bacterial genomic products in explanted livers from patients with PBC. The conflicting results may be partly due to the difference in the disease stages at which samples were obtained.

Although the possibility of bacterial infection, particularly with intracellular organisms, is an attractive hypothesis in the pathogenesis of PBC, there is at present no concrete evidence to suggest an ongoing infectious process in PBC. However, this does not preclude infection at the time of disease induction because PBC could be a "hit-and-run" disease; the etiologic agent would therefore be absent in the later stages of the disease.

Although viruses are often cited as possible causative factors in autoimmune diseases, including rheumatoid arthritis, Sjögren's syndrome, systemic lupus erythematosus (SLE), and multiple sclerosis, no specific virus has been implicated in the pathogenesis of PBC. After demonstration that human intracisternal A-type peptide (HIAP) could be isolated from the salivary glands of patients with Sjögren's syndrome,^[99] immunoblotting studies have indicated that many patients with PBC have seroreactivity with retroviral proteins (35% HIV-1 p24; 51% HIAP).^[100] However, this pattern of seroreactivity is also seen in patients with PSC, and it is unlikely that mitochondrial antigens share epitopes with the HIV-1 p24 and HIAP retroviral proteins. Nevertheless, the involvement of viral infection, or activation of human endogenous retroviruses, as a possible mechanism triggering the onset of PBC should not be overlooked. The activation of viral elements itself or expression of specific proteins that may share a molecular mimic to the mitochondrial antigens in PBC remains an unproven but attractive hypothesis. It should be noted that several lines of evidence suggest that viruses, including Epstein-Barr virus, HIV, human T cell leukemia virus-1, and hepatitis virus C, may be implicated in pathogenesis of Sjögren's syndrome, which is also an autoimmune disease occurring in epithelial cells of exocrine glands and is often present in patients with PBC.^[101]

Environmental Factors and PBC

There is large variation in the reported prevalence of PBC and, although part of the variation in prevalence may reflect methodological factors, there is likely to be a true variation in incidence. In addition to the weak genetic association with the disease susceptibility, studies on differing prevalence of PBC among ethnically (and genetically) equivalent populations located in geographically different regions as a result of immigration have led to the search for environmental provocation of PBC.^[65] Some cases of PBC tend to "cluster" within areas. This phenomenon was first described in Sheffield, where cases of PBC clustered in one area supplied by the Rivelin reservoir; the point prevalence here was 115 per million compared with 13 per million elsewhere in Sheffield.^[102] A subsequent study by Chetwynd et al^[103] showed significant evidence of spatial clustering; this clustering was unrelated to urban or rural location or to diagnostic activity, suggesting a true clustering effect. However, results of epidemiological studies in particular regions have not so far revealed any causative environmental exposures.

Xenobiotics are foreign compounds that may either alter or complex to defined self-proteins, inducing a change in the molecular structure of the native protein sufficient to induce to an immune response. Such immune responses may then result in the recognition of not only the modified or altered protein but also the unmodified native protein.^[104] The chronic presence of the self-protein serves to perpetuate the immune response initiated by the xenobiotic-induced adduct and leads to autoimmunity.^[105,106] There have been a number of chemical xenobiotics associated with human autoimmune diseases;^[106] these include SLE, in which factors such as ultraviolet light and drugs, including procainamide and hydralazine, may trigger the disease. Many xenobiotics are metabolized in the liver, thereby increasing the potential for liver-specific alteration of proteins. In fact, a liver-specific autoimmune disease can be observed in some patients exposed to chlorofluorohydrocarbon anesthetics.^[107,108] Previous work has reported that immunization with halothane, whose trifluoroacetyl (TFA) metabolite covalently links to lysine on cytochrome p450 2E1,^[109] induces the formation of antibodies that cross-react not only with the haptenated (TFA) immunogen but also with the lipoated PDC-E2.^[110,111] This finding has important implications in the pathogenic mechanisms associated with PBC, an autoimmune disease marked by the presence of AMAs.

The targets of AMAs are the E2 components of the 2-OADC pathway, particularly PDC-E2, and the primary B cell epitope of PDC-E2 recognized by AMAs includes a lipoated lysine residue. Therefore, there is a possibility that the lipoic acid residue of PDC-E2, modified by xenobiotics, becomes immunogenic and initiates or perpetuates AMA response. To address this hypothesis, Long et al^[112] synthesized the inner lipoyl domain of PDC-E2, replacing the lipoic acid moiety with synthetic structures designed to mimic a xenobiotically induced lipoyl hapten, and they quantified the reactivity of these structure with sera from PBC patients. Interestingly, AMAs from all seropositive patients with PBC but no controls reacted against 3 of the 18 organic modified autoepitopes significantly better than they reacted to the native domain of PDC-E2.^[113] By structural analysis, the features that correlated with autoantibody binding included synthetic domain peptides with a halide or methyl halide in the meta- or para-position containing no strong hydrogen bond accepting groups on the phenyl ring of the lysine substitutes and synthetic domain peptides with a relatively low rotation barrier about the linkage bond. Many chemicals, including pharmaceuticals and household detergents, have the potential to form such halogenated derivatives as metabolites. These data reflect that an organic chemical can serve as a mimetope for an autoantigen and provide evidence for potential mechanism by which environmental organic compounds may cause PBC rather than microbiologic agents.

Concluding Remarks

Although extensively studied, the mechanism by which immune tolerance is disrupted and autoimmunity induces selective tissue damage to BECs remains unknown. One intriguing hypothesis is molecular mimicry of autoepitopes by either peptides of infectious agents or xenobiotics. Because of the highly conserved nature of the mitochondrial proteins, work concerning the T cell epitopes on the autoantigens -- for both CD4⁺ and CD8⁺ T cells -- will help to clarify the possibility of molecular mimicry. Although there is at present no concrete evidence to suggest an ongoing infection at the time of disease induction, future epidemiological studies including xenobiotic exposure may provide more insights into pathogenesis. Further, the future use of immunotherapy is likely to significantly improve our treatment of patients with autoimmune disease, including primary biliary cirrhosis.