Introduction

Autoimmune hepatitis is a consequence of dysregulated immune responses that overcome tolerance for self-antigens [1,2,3]. Cytokine-producing inflammatory cells, antigen-activated promiscuous CD4+ T cells, and liver-infiltrating cytotoxic T cells (CTLs) create a microenvironment that promotes liver damage [4,5,6,7]. The common clinical outcomes of this complex immune response are progressive hepatic injury and fibrosis [8, 9]. Environmental factors could trigger autoimmune hepatitis by overcoming homeostatic mechanisms that modulate immune tolerance and immune ignorance [10,11,12].

The linear or conformational mimicry of foreign antigens with self-antigens could promote the loss of self-tolerance and immune ignorance by increasing the array of immune targets and the avidity of promiscuous T cell effectors [4, 13,14,15,16,17,18,19,20,21,22]. Diverse pathogens have been evaluated as the principal sources of environmentally-derived molecular mimics of self-antigens in autoimmune hepatitis [23,24,25,26], and vaccination [27,28,29,30,31,32,33,34,35,36,37,38,39,40,41] and the intestinal microbiome [25, 26, 42,43,44,45,46,47] have emerged as sources of antigens that could also promote autoimmunity by molecular mimicry. Molecular mimicry could be a mechanism that complements epigenetic transformations in shaping the autoreactive response to environmental stimuli [48,49,50,51,52,53,54,55].

The molecular mimicry of environmental antigens has been implicated in the development of primary biliary cholangitis (PBC). Structural mimicries and immune cross-reactions have been established between the immune dominant pyruvate dehydrogenase complex-E2 (PDC-E2) of PBC and certain bacteria (Escherichia coli [56,57,58] and Novoshingobium aromaticivorans) [59,60,61]). Furthermore, limited population studies have supported the bacterial associations with the occurrence of clinical disease [58, 61,62,63]. Xenobiotics, especially 2-octynoic acid [64], have also been shown to mimic and modify the lipoyl domain of PDC-E2 by chemical conjugation [65]. Transformation of the altered native PDC-E2 into a homologue of native PDC-E2 could promote the production of antimitochondrial antibodies [66] and the development of a PBC-like disease in an experimental model [67,68,69]. The challenge has been to establish these environmentally-induced molecular mimicries as pathological bases for PBC in humans [70].

Improvement in the management of autoimmune hepatitis is recognized as an unmet clinical need [71, 72], and satisfaction of this need requires clarification of the factors that induce and sustain the autoreactive process. Identification of critical environmental elements and the mechanisms by which they can promote autoimmune hepatitis could also suggest strategies that prevent or limit the disease [73].

The goals of this review are to describe the experimental and clinical evidence that supports molecular mimicry as a causative mechanism of autoimmune hepatitis, indicate limitations and uncertainties of this premise, and encourage investigations that assess diverse environmental antigens as immunogenic and pathogenic molecular mimics of disease-relevant self-antigens. The validation of molecular mimicry as a pathogenic mechanism may direct future management strategies.

Methods

English abstracts were identified in PubMed using the search phrases, “environmental factors in autoimmune hepatitis”, “molecular mimicry”, “vaccination and autoimmune hepatitis”, and “intestinal microbiome and autoimmune hepatitis”. Key aspects of pertinent abstracts were recorded, and full-length articles that expanded pivotal concepts constituted the primary bibliography. Secondary and tertiary bibliographies were developed from the references cited in the selected full-length articles of the preceding bibliography. Several hundred abstracts and 184 full length articles were examined.

Molecular Mimicry and Environmental Pathogens

Molecular mimicry is a largely theoretical mechanism by which environmental pathogens could trigger autoimmune hepatitis [18, 21, 74]. The process requires the selection, processing, and presentation of an immunogenic peptide intrinsic to the pathogen [3, 75]. The peptide selected for presentation by antigen presenting cells (APCs) has to resemble, not duplicate, the structure of a self-antigen [75], and it has to activate an immune response by overcoming self-tolerance and immune ignorance [6, 21] (Fig. 1). Up-regulation of MHC molecules by the invading pathogen and an inflammatory milieu (cytokines, chemokines, and inflammatory cells) associated with the infection (bystander response) could promote the presentation of pathogen-derived immunogenic peptides [21, 76]. Activation and differentiation of CD4+ and CD8+ T cells sensitized to these peptides could then induce an immune-mediated, inflammatory liver disease depending on the number and avidity of the T cell effectors for the homologous tissue antigen [18, 21, 75].

Fig. 1
figure 1

Molecular mimicry and environmental antigens. Environmental antigens from exogenous and endogenous sources may have structural or conformational homology with a disease-related antigen. Peptides of at least 5 amino acids (aa) that are derived from these sources could constitute an immunogenic unit that is homologous to an immune dominant epitope of a self-antigen. Intestinal dysbiosis and gut-derived metabolic processes may affect levels of lipopolysaccharide (LSP) and short chain fatty acids (SCFA) which in turn may increase permeability of the intestinal mucosal barrier. Gut-derived immunogenic peptides could then translocate to systemic sites and encounter antigen presenting cells. In dendritic cells, the immunogenic homologues can undergo proteasomal processing and presentation by class I or class II molecules of the major histocompatibility complex (MHC). A molecular chaperone may facilitate the proteasomal processing and substrate presentation. Heat shock proteins (HSP) are theoretical candidates for this role (dotted dual-headed arrow). Antigen cross-presentation can allow a processed environmental antigen to be presented by MHC class I molecules to CD8+ T cells, and the activated CD8+ T cells may then differentiate to liver-infiltrating cytotoxic T cells (CTL). The processed homologue may also be presented by MHC class II molecules to CD4+ T cells. The activated CD4+ T cells may then induce a type 1 helper T cell (Th1) response with increased production of interleukin 2 (IL-2) and interferon-gamma (IFN-γ). Macrophages and natural killer (NK) cells may also be activated. The IFN-γ may in turn activate the class II transactivator (CIITA) gene which increases the expression of class II MHC molecules and presentation of the processed homologue. The activated CD4+ T cells may also produce a Th2 response characterized by certain interleukins (IL-4, IL-5, IL-13, and IL-10) and the activation of B cells and plasma cells which in turn may generate autoantibodies

Genetic Predisposition and Plasticity of the APCs

Antigenic peptides presented by the class II MHC molecules (HLA DR, DP, and DQ) are derived from extracellular proteins (foreign antigens) [77] that are processed mainly by proteases within endosomal compartments [78, 79] (Table 1). The antigen binding groove of class II MHC molecules consists of two chains of amino acids that are encoded by separate genes in the class II MHC region of chromosome 6 [80]. The antigen binding groove of the HLA-DR molecules associated with autoimmune hepatitis consists of a monomorphic α-chain and polymorphic β-chain [81,82,83,84]. The DRα-chain is encoded by the non-polymorphic DRA gene, and the DRβ-chain is encoded by the polymorphic DRB1 genes [85].

Table 1 Requisites for molecular mimicry of environmental antigens in autoimmune hepatitis

The selection, positioning, and presentation of the antigenic peptide to naïve CD4+ helper T cells in autoimmune hepatitis is influenced by the hypervariable regions encoded within the DRβ-chain [86] (Table 1). Sequence polymorphisms of the DRβ-chain determine the type of peptides that can bind in the pockets of the antigen binding groove [86], and this molecular alignment can affect susceptibility to autoimmune hepatitis [87, 88].

Structural alterations in the antigenic peptide and variations in the MHC alleles that encode the DRβ chain can change susceptibility by altering the conformation of the antigen-presenting complex [89,90,91] (Table 1). The conformational plasticity of the antigen binding groove of class II MHC molecules allows it to sample, select, and accommodate diverse antigenic peptides derived from the environment [91]. Furthermore, the antigen binding groove is open at both ends, and it can accommodate relatively large, environmentally derived peptides of 13–25 residues [91, 92].

The molecular mimicry of self-antigens by peptides of environmental origin requires presentation of the homologous peptides to CD4+ helper T cells, and the class II MHC molecules on dendritic cells, monocytes, macrophages, and B cells can met this requirement [80] (Table 1). Furthermore, the pro-inflammatory cytokine, interferon-gamma (IFN-γ), can induce the expression of class II MHC molecules on other cells by activating the class II transactivator (CIITA) gene [93,94,95] (Fig. 1). CIITA is the major regulator of MHC class II transcription [93], and it may expand the number and variety of APCs and the population of activated promiscuous CD4+ T cells.

Antigen cross-presentation by dendritic cells is another mechanism by which environmental peptides that mimic self-antigens can initiate an autoreactive response (Table 1). Antigen cross-presentation is a process by which antigens from the extracellular environment can be presented by MHC class I molecules and directly stimulate the development of antigen-specific CD8+ T cells [96,97,98,99] (Fig. 1). This process is distinguished from the canonical pathway in which extracellular (foreign) antigens are presented by MHC class II molecules and activate CD4+ helper T cells [100]. Dendritic cells can present processed antigens by MHC class I and MHC class II molecules [101,102,103], and they are the principal population engaged in antigen cross-presentation. Molecular chaperones that carry substrates generated by the proteasomal degradation of the foreign antigen can engage in the priming process [97]. Enhanced immune reactivity to the extrinsic antigen may be a consequence of this interaction depending on the antigen and the microenvironment [98, 104].

Causative Effect of Molecular Mimicry in Experimental Models

Molecular mimicry by a viral pathogen has been proposed as a basis for the immune-mediated liver damage in murine models of autoimmune hepatitis. Infection with an adenovirus expressing the human cytochrome P450 IID6 (hCYP2D6) [75, 105, 106] and immunization with hCYP2D6 and human formiminotransferase cyclodeaminase [107, 108] have induced laboratory, histological, and immunological changes of autoimmune hepatitis (Table 2). CYP2D6 [109,110,111] and formiminotransferase cyclodeaminase [112, 113] have been closely associated with type 2 autoimmune hepatitis, and molecular mimicry between the human antigens and mouse homologues in the experimental models has been proposed as the basis for the pathological [75, 105, 107, 108], serological [105, 107, 108], and fibrotic [114] changes resembling autoimmune hepatitis.

Table 2 Evidence for molecular mimicry of environmental antigens in autoimmune hepatitis

Molecular mimicry has also been demonstrated in a study exposing wild-type and transgenic mice to the same human antigen associated with autoimmune hepatitis [75]. The infection of wild-type mice with adenovirus expressing hCYP2D6 has generated T cell activity against epitopes that have had intermediate homology between hCYP2D6 and mouse cytochrome (mCYP) homologues [75]. In contrast, the same adenovirus infection of transgenic mice expressing hCYP2D6 has failed to generate an hCYP2D6-specific T cell response [75]. Molecular mimicry rather than identity between the adenoviral hCYP2D6 and the mCYP homologues was the likely basis for the hCYP2D6-specific T cell response [75].

Molecular mimicry has not been established as a cause of autoimmune hepatitis in humans [21, 115], but the structural and conformational similarities between environmental pathogens and disease-associated antigens have compelled its consideration [109, 111, 116,117,118,119,120,121,122,123]. Undiscovered pathogens [124], isolated subclinical or antecedent infections [21, 74], and multiple previous infections by diverse viruses mimicking the same self-antigen may contribute to the risk of autoimmune disease [21, 74, 111].

Contributory Effects of Molecular Mimicry

Molecular mimicry may also be a mechanism that accelerates or sustains autoimmune hepatitis [44, 73, 125] (Table 2). The release of neo-antigens from damaged liver tissue [126, 127] or the translocation of bacterial products from a permeable intestine [44, 47, 128,129,130,131,132,133,134,135] could promote molecular mimicry and affect the severity and course of autoimmune hepatitis. Furthermore, molecular mimicry could contribute to the ongoing inflammatory process by expanding the number of targeted epitopes (epitope spread) [136,137,138,139] and the population of memory CD4+ and CD8+ T cells (bystander activation) [140, 141].

Mimicry Between CYP2D6 and Environmental Pathogens

CYP2D6, the principal autoantigen associated with one form of autoimmune hepatitis [109,110,111], has amino acid sequences within its structure that resemble sequences in the hepatitis C virus (HCV) [109, 119, 120], herpes simplex type 1 virus [109, 117], and human T lymphotrophic virus types 1 and 2 [117] (Table 2). A conformational epitope on CYP2D6 in patients with chronic hepatitis C and antibodies to liver kidney microsome type 1 (anti-LKM1) spans the major linear epitope associated with type 2 autoimmune hepatitis [122], and cross-reacting anti-LKM1 have been demonstrated in patients with chronic hepatitis C [122, 142, 143]. Furthermore, a homologous region between human CYP2D6 and protein from Salmonella typhimurium suggests that bacteria could be another environmental source of molecular mimicry [117].

None of the pathogens with amino acid sequences that resemble those of CYP2D6 has been recognized as a cause of type 2 autoimmune hepatitis in humans [21, 115, 144]. The immunogenicity of the shared sequences within the pathogen [145, 146] and the composition of the host-specific microenvironment [21, 76] may be the critical factors that influence the risk of disease.

Mimicry Between Other Immunogenic Antigens and Environmental Pathogens

Molecular mimicry has also been demonstrated between key antigenic targets in type 1 autoimmune hepatitis and environmental pathogens [116, 121, 123]. The deoxyribonucleic acid (DNA) polymerase of the hepatitis B virus (HBV) has structural similarities to smooth muscle and nuclear antigens [116], and homologies have also been described between other regions in these same antigens and amino acid sequences in the polyprotein of HCV [121] (Table 2). Neither virus has been recognized as a cause of type 1 autoimmune hepatitis [21, 144], and the major clinical consequence of the structural homologies may be to lower the diagnostic specificity of smooth muscle antibodies (SMA) and antinuclear antibodies (ANA) [147].

Structural mimicry has also been demonstrated between the immune dominant regions of the SepSecS protein (O-phosphoserine [Sep] transfer RNA:selenocysteine (Sec) transfer RNA synthase) and the PS 120 surface antigen of Rickettsia species [123] (Table 2). The SepSecS protein [148,149,150,151] generates antibodies to soluble liver antigen (anti-SLA) which are highly specific for autoimmune hepatitis [152,153,154]. The clinical impact of the structural homology between the SepSecS protein and the PS 120 surface antigen of Rickettsia species is uncertain [123, 151, 155].

Molecular Mimicry and Vaccination

Human vaccines have been developed against deleterious environmental pathogens by stimulating the recipient’s immune system to react against critical structural components of the invading organism [156,157,158,159,160,161]. The generation of neutralizing antibodies and the activation of an antigen-specific, adaptive immune response can eliminate the targeted organism [162,163,164,165]. Similarities between peptides generated by the vaccine and normal proteins within the vaccine recipient may generate immune cross-reactivity that is manifested as an autoimmune disease [166,167,168,169,170] (Table 2). Recent experiences in humans after vaccination against the severe acute respiratory syndrome (SARS)-associated coronavirus 2 (SARS-CoV-2) have suggested an etiological connection between the vaccines and autoimmune hepatitis which may relate to molecular mimicry [27,28,29,30,31,32,33,34,35,36,37,38,39,40,41].

Autoimmune Hepatitis-Like Disease After SARS-CoV-2 Vaccination

An acute onset hepatitis has been described in 87 patients within 3–65 days (median 15 days) after vaccination for SARS-CoV-2 [35]. The vaccines mainly contained messenger ribonucleic acid (mRNA) for the spike (S) protein of SARS-CoV-2, and 57% of patients with acute hepatitis had features of autoimmune hepatitis (Table 2). Glucocorticoids had been administered to most patients, and the liver disease had resolved in all treated and untreated patients with one exception (frequency of liver failure and transplantation, 1.1%). Relapse did not occur during an observation period that spanned 44–140 days after treatment withdrawal and 35–172 days after spontaneous resolution.

Similar findings have been reported in two systematic reviews of vaccine-associated liver disease [32, 34] (Table 2). In the largest review, an acute hepatitis with autoimmune features had developed within 7–21 days (median, 14 days) after vaccination in 138 patients; 98% had received mRNA-based vaccines against SARS-CoV-2; most patients had been treated with immunosuppressive drugs; 89% had achieved full recovery; and 2.2% had died [32]. In the smaller review involving 32 patients who had developed an autoimmune hepatitis-like disease after vaccination, 7 improved spontaneously (22%); 24 of the 25 treated patients improved or resolved (96%); and one patient died (3%) [34].

The acute onset of liver injury; its spontaneous resolution in some patients; the near-universal resolution after immunosuppressive therapy; and the absence of relapse after drug withdrawal are atypical features of autoimmune hepatitis [171] (Table 2). The atypical features, however, do not discount the actual or potential occurrence of autoimmune hepatitis after vaccination [34, 172] or the possible role of vaccine-induced molecular mimicry [27,28,29,30,31,32,33, 35, 37,38,39,40,41]. The overall safety and efficacy of vaccination against SARS-CoV-2 have far outweighed the rarity of vaccine-related liver dysfunction, and the recommendation for vaccination in the general population has remained strong [32, 34, 35, 173,174,175].

The apparent rarity of de novo autoimmune hepatitis after coronovirus disease 2019 (COVID-19) [176,177,178] suggests that the vaccine and its delivery system of lipid nanoparticles [158, 161] generate a more intense immunogenic signal than community-acquired infection.

Vaccine-Based Molecular Mimicry and Autoimmunity

Sequence analyses of 20,365 human proteins have identified 3781 proteins that share peptides of at least 6 amino acids with structural proteins of SARS-CoV-2 [179] (Table 2). Pentapeptides are recognized as the minimal immunogenic epitopes [145, 146], and the extensive sharing of hexapeptides between SARS-CoV-2 and human proteins suggests that molecular mimicry after vaccination is common and rarely pathogenic [21, 168, 170]. The immunogenicity and pathogenicity of the molecular mimic may depend on its relevance to a pivotal disease-pertinent self-antigen and the circumstances that affect its processing and presentation by the MHC molecules (Fig. 1).

Antibodies to the spike protein of SARS-CoV-2 have had reactivity to human tissue transglutaminase, extractable nuclear antigen, myosin basic protein, mitochondria, and nuclear antigens. These homologies have been expressed in serum as nuclear, actin, and mitochondrial antibodies [180] (Table 2). The serological findings indicate that molecular mimicry after vaccination can be a basis for immune reactivity associated with autoimmune hepatitis. The investigational challenges are to identify the key molecular mimic of an immune dominant epitope that can trigger autoimmune hepatitis and the circumstances that promote its immunogenicity. [21, 22, 170]

Molecular Mimicry and Gut-Derived Microbial Products

The intestinal microbiome is a reservoir of bacterial products and immune cells that have the potential to translocate to extra-intestinal sites and generate an immune response [131, 181]. CD4+ and CD8+ T cells activated by gut-derived peptides may target homologous tissue antigens and contribute to a loss of self-tolerance [44, 182, 183] (Fig. 1). Molecular mimicry based on changes within the intestinal microbiome constitutes a potentially powerful mechanism by which environmental factors could shape the intestinal microbiome.

Diet [25, 134], supplements [24], antibiotics [23], alcohol [23], pollutants [23], and toxins [26] have been associated with the risk of autoimmune hepatitis, and this risk may reflect changes in the intestinal microbiome induced by these diverse environmental factors [25, 26, 42,43,44,45,46,47] (Table 2). Findings in experimental models of autoimmune hepatitis [129, 184] and other autoimmune diseases [185,186,187,188] have supported this premise, albeit there are many other factors outside of molecular mimicry that may be contributory [45,46,47, 135, 189].

Dysbiosis and Disease Severity

Immune tolerances to commensal bacteria within the intestine may be lost by environmentally-induced dysbiosis [190], and toll-like receptors in the intestine may generate pathogenic T and B lymphocytes that are capable of systemic translocation and reactivity against homologous tissue antigens [188, 191] (Table 2). Intestinal dysbiosis has been demonstrated in an experimental model [128] and in patients with autoimmune hepatitis [132,133,134,135]. The key distinguishing features from healthy individuals have been reduced biodiversity of the intestinal microbiome [128, 132,133,134, 192], increased aerobic and facultative anaerobic bacteria [134], increased Veillonella species [132, 134], and decreased Bifidobacterium species [129, 134] (Fig. 1). The composition of the intestinal microbiome has been specific for autoimmune hepatitis when compared to that of healthy individuals and patients with primary biliary cholangitis (PBC) or ulcerative colitis [134].

Changes in the size of certain bacterial populations within the intestine have also been associated with the severity of autoimmune hepatitis [135]. An increased Veillonella population in patients with autoimmune hepatitis has correlated with the serum level of aspartate aminotransferase [132], and a decreased population of Bifidobacterium has been associated with failure to achieve remission during treatment [134] (Table 2). Furthermore, the abundance of Bifidobacterium in the intestinal microbiome has correlated directly with the average protein intake of patients with autoimmune hepatitis [134]. This finding underscores the possibility that dietary effects on the intestinal microbiota might impact on disease severity.

Increased Intestinal Permeability

Alterations in the structural proteins (zona occludens 1 and occludin) that bind intestinal epithelial cells have been described in patients with autoimmune hepatitis [129], and increased intestinal permeability has been demonstrated in experimental models of the disease [193,194,195] (Table 2). Multiple microbial products that are manufactured within the intestinal microbiome may be enhanced by the dysbiosis and account for the barrier disruption [135, 196].

Lipopolysaccharides (LSP) from commensal gram-negative bacteria may increase, and they may enhance intestinal permeability by disrupting signaling pathways that maintain barrier integrity [135, 197] (Fig. 1). The breakdown of indigestible carbohydrates to short chain fatty acids (acetic, proprionic, and butyric acids) may also be impaired by changes in the populations of commensal anaerobic bacteria [198,199,200,201]. Clostridia [201] and Bifidobacterium [202] are key carbohydrate-fermenting, bacterial populations, and reductions in their abundance could reduce concentrations of short chain fatty acids, especially butyrate, within the gut [200].

Butyrate has strengthened tight junction assembly and improved barrier function in experimental models [203,204,205,206], and its deficiency in autoimmune hepatitis could facilitate translocation of bacterial products and enhance the opportunity for molecular mimicry. A butyrate deficiency could also affect other signaling pathways that modulate immune and inflammatory responses pivotal for the occurrence and maintenance of autoimmune hepatitis [207,208,209,210]. Dietary fiber and butyrate supplementations have increased mucosal junction proteins, elevated the ratio of regulatory T cells (Tregs) to T helper 17 (Th17) cells, and decreased translocation of microbial protein in experimental autoimmune hepatitis [211].

The critical impact of dysbiosis on the integrity of the intestinal mucosal barrier has been demonstrated further by manipulating the gut microbiota. Probiotics are preparations of living enteric commensal micro-organisms that are intended to induce a salutary microbiome [212,213,214]. The administration of probiotics containing species of Bifidobacterium with [195] or without Lactobaccillus [193] has had a broad spectrum of beneficial effects in experimental models of autoimmune hepatitis (Table 2). It has suppressed hepatic inflammation [195], decreased serum aminotransferase levels [195], diminished proliferation of Th17 cells [193, 195], reduced serum endotoxin levels [193], enhanced the abundance of short chain fatty acids [193], inhibited transcription of cytokines [195], and maintained or strengthened the intestinal mucosal barrier [193, 195].

Interactive Modulatory Metabolic Effects

Metabolic products from different bacterial populations within the intestine can become shared energy sources that sustain the composition of the microbiome and the integrity of the intestinal mucosal barrier [42, 199]. The degree of communal interdependence is unclear, but cross-feeding among commensal populations may enhance their resiliency. The acetate produced by Bifidobacterium can be utilized by Fecalibacterium prausnitzii, a dominant intestinal inhabitant, to produce butyrate [215]. The cross-feeding of energy sources may be a mechanism by which Bifidobacterium can up-regulate tight junctions, improve barrier function, reduce endotoxemia, impair proliferation of Th17 cells, and mitigate experimental autoimmune hepatitis [193, 215].

The intestinal microbiota may also produce metabolites that modulate the inflammatory and immune responses. Glycolipids derived from the intestinal bacteria can be presented by the MHC class I-like molecule, CD1d, and activate natural killer T (NKT) cells [216]. NKT cells activated by this mechanism have promoted liver injury in an experimental model of hepatitis [217]. Secondary bile acids (lithocholic acid and deoxycholic acid) derived from the actions of enteric bacteria on the primary bile acids (cholic acid and chenodeoxycholic acid) may counter a pro-inflammatory response by activating the G-protein-coupled bile acid receptor 1 (GPBAR1). GPBAR1 can direct the differentiation of NKT cells to an anti-inflammatory, interleukin 10-producing, subset that ameliorates experimental immune-mediated hepatitis [189].

Molecular Mimicry of Gut-Derived Antigens in Autoimmune Hepatitis

Atypical perinuclear anti-neutrophil cytoplasmic antibodies (pANCA) occur in 49–96% of patients with type 1 autoimmune hepatitis [218,219,220,221,222,223]. They are also found in patients with primary sclerosing cholangitis (PSC) [219, 224, 225] and inflammatory bowel disease [226,227,228]. Atypical pANCA recognize β-tubulin isotype 5 in conjunction with its evolutionary bacterial precursor protein, filamenting temperature-sensitive mutant Z protein (FtsZ) [229] (Table 2). The dual reactivity is dependent on an intact intestinal microbiome in a mouse model of inflammatory bowel disease [229]. The absence of this dual reactivity in germ-free animals supports the premise that gut-derived antigens can escape the intestine and trigger an antibody response that is disease-related [229,230,231]. The high prevalence and high titers of atypical pANCA in some studies of type 1 autoimmune hepatitis [218, 223] compel this consideration, albeit the antibody production may not translate into a pathogenic immune response.

Prospect of Molecular Mimicry as a Cause of Autoimmune Hepatitis

The prospect that molecular mimicry of environmental antigens will emerge as a valid cause of autoimmune hepatitis depends on the demonstration of its immunogenicity and pathogenicity. Examinations of 30 viral proteomes [232] and 40 bacterial proteomes [233] for similarities to the human proteome have disclosed massive sharing of amino acid motifs at the pentapeptide level and higher. The ubiquity of molecular mimicry between pathogenic viral and bacterial species and normal tissue proteins indicates that molecular mimicry by itself is insufficient to cause autoimmune disease [179, 232,233,234].

Future investigations must evaluate diverse environmental antigens in addition to pathogens as candidates for molecular mimicry [131, 188, 235,236,237,238]. They must demonstrate the immunogenicity [21] and pathogenicity [188] of the implicated homologues, and they must determine the factors that influence their potency as triggering peptides. The intestinal microbiome must also be rigorously assessed as a source of molecular mimicries that can initiate, enhance, and sustain a deleterious autoreactive immune response [44, 73, 131, 188, 238, 239]. Demonstration of the pathogenic nature of the molecular mimicries will require a humanized animal model of autoimmune hepatitis that expresses human leukocyte antigen (HLA) risk alleles.

Discovering Diverse Environmental Homologues and Demonstrating Immunogenicity

Pentapeptides are the shortest functional units recognized by the immune system [145, 146]. The search for immunogenic and pathogenic molecular mimics in autoimmune hepatitis must focus on identifying peptides in environmental antigens that not only mimic the immune dominant epitopes of CYP2D6 but also have sufficient size to be immunogenic [109, 111, 120] (Table 3). CYP2D6 is the pivotal antigen for the discovery of disease-relevant molecular mimics in the environment because of its high specificity for type 2 autoimmune hepatitis [109, 111, 120]. Type 2 autoimmune hepatitis, however, is less common in North American adult patients than type 1 autoimmune hepatitis [110, 147, 153, 240]. The key antigen associated with type 1 autoimmune hepatitis remains unclear, and this deficiency limits discovery of its disease-relevant molecular mimics within the environment.

Table 3 Establishing environmentally-induced molecular mimicry as cause of autoimmune hepatitis

The epitope of CYP2D6 which spans the amino acid (aa) region, aa193-212, is a prime immune dominant epitope associated with antibodies to liver kidney microsome type 1 (anti-LKM1) and type 2 autoimmune hepatitis [111]. The CYP2D6 epitope spanning the amino acid sequence, aa316 and 327, is expressed on the molecular surface, and its surface location may enhance its candidacy as another immune dominant epitope of CYP2D6 [120]. A 33 amino acid segment of the CYP2D6 molecule contains a shorter segment of 8 amino acids which resembles the herpes simplex type 1 virus [109], and it is another epitope of CYP2D6 that could be used to probe for homologies in diverse environmental antigens. Homologues of each immune dominant epitope of CYP2D6 can be sought using established databases (Genbank, Protein Information Resource [PIR], SWISS-PROT, or the International Nucleotide Sequence Database Collaboration [INSCD]) [109, 111, 241].

The immunogenicity of the implicated homologue can be demonstrated by detecting its antibodies in sera from patients with type 2 autoimmune hepatitis. Pre-absorption of the sera with LKM1 enhances its specificity for the candidate homologue [109, 242]. Immunogenic peptide mimics of the pivotal CYP2D6 epitopes could also be assessed by using peptide libraries of environmental antigens expressed on a microarray and probed with the pre-absorbed serum [242,243,244,245]. The identification of disease-pertinent peptide mimics in diverse environmental antigens (foodstuffs, vaccines, toxins, and micro-organisms [235,236,237, 246]) would support the causal role of molecular mimicry in type 2 autoimmune hepatitis and encourage discovery of the immune dominant epitopes in type 1 disease.

Defining Conditions That Enhance Immunogenicity

The ubiquity of molecular mimicry between foreign and self-proteins and the relative rarity of autoimmune diseases [233] have driven the search for factors that enhance the causative impact of molecular mimicry [21]. The factors that have been associated with the loss of self-tolerance (bystander activation of pre-primed T cells, memory cells and APCs [73, 140, 141]; epitope spread to tissue antigens of progressively lower homology [136,137,138,139, 247]; the release of neo-antigens from damaged tissue [126, 127]; and the production of superantigens that directly activate T lymphocytes [248, 249]) intensify the immune response [6], but they are not the primary causes of it. Adjuvant mechanisms that can particularly enhance the immunogenicity of molecular mimics would strengthen their candidacy as primary interfaces between the environment and autoimmune hepatitis (Table 3).

Molecular Chaperones and Peptide Immunogenicity

Molecular chaperones are highly conserved proteins that are present in all cellular compartments [250]. They stabilize the processing, maturation, and activity of intracellular proteins by aiding in their folding, assembly, and trafficking [250,251,252]. Heat shock proteins (HSPs) are molecular chaperones that reside constitutively within the cytoplasm, and they can be induced by oxidative stress, nitrosative stress, inflammation, or infection [253, 254]. HSPs can also affect the immunogenicity of intra- and extracellular peptides [255,256,257].

HSPs can bind to intracellular antigens containing both MHC class I and MHC class II epitopes, and as molecular chaperones, they can facilitate cross-presentation of acquired antigens by dendritic cells [257] (Fig. 1). HSPs may be adjuvants that could strengthen the immunogenicity of an environmental peptide that mimics a self-antigen (Table 3). The investigation of HSPs as modulators of the immune response to foreign antigens may strengthen the candidacy of molecular mimicry as a causative factor in autoimmune hepatitis. It may also suggest HSP-directed therapeutic strategies for future study [254, 258,259,260].

Other molecular chaperones that could affect the immunogenicity of environmental antigens are tapasin [261, 262] and the transporter associated with antigen processing (TAP)-binding protein (TAPBPR) [263,264,265,266]. Both chaperone molecules facilitate the selection, processing, and editing of high affinity peptides for MHC class I molecules. The lysosome-associated membrane protein-2 isoform (LAMP-2a) promotes the presentation of peptides by MHC class II molecules [267, 268], and the human leukocyte antigen (HLA)-B-associated transcript 3 (Bat3) protects autoreactive T cell responses by preventing T cell exhaustion [269, 270]. Chaperone molecules that contribute to the selection, editing, and presentation of peptides and the preservation of the autoreactive T cell response are factors that can affect the immunogenicity of environmental antigens. Their roles are unassessed in autoimmune hepatitis.

Demonstrating Pathogenicity of the Implicated Homologue

The implicated homologue must generate T and B cell reactivity to the principal antigen associated with autoimmune hepatitis (CYP2D6) [271], and the cross-reactivity must produce the disease in a genetically predisposed experimental model [188, 272]. The homologue must be non-identical to CYP2D6, and the experimental model must be susceptible to the disease in order to mitigate immune tolerance [75, 188]. The pathogenicity of a homologue derived from the intestinal microbiome has been elegantly demonstrated in patients with antiphospholipid syndrome, [188] and a similar methodology can be applied in patients with autoimmune hepatitis.

Immunogenic epitopes of CYP2D6 can be used to search for mimics within the intestinal microbiome by examining available databases, including resources developed by the Human Microbiome Project Consortium [273, 274] (Table 3). The enteric commensal bacterial species expressing the critical homologue can be identified, sought in the fecal specimens of patients with type 2 autoimmune hepatitis, and assessed for pathogenicity by evaluating T and B cell cross-reactivity between the epitopes of CYP2D6 and the molecular mimics expressed in the bacterial species. The methodology can be expanded to identify and assess homologues derived from other environmental sources.

Focusing on the Intestinal Microbiome

Gut-derived homologous peptides must translocate from the intestine to extra-intestinal sites to generate pathogenic, cross-reactive, T and B cell responses [188]. The genetic and biological factors that contribute to these responses must be clarified in autoimmune hepatitis [231] (Table 3). The enteric milieu of short chain fatty acids [198,199,200,201], LSP [197], glycolipids [217], and bile acids [189, 209] must be characterized to define the conditions for increased intestinal permeability. Furthermore, the principal factors that alter or restore the function of the intestinal epithelial cells must be determined [129, 275]. The investigational emphasis must shift from measuring the size of the enteric populations in autoimmune hepatitis to assessing their function and interactivity [47]. The pathogenic sequence from environmental factors to intestinal dysbiosis to a disease-producing immune response by molecular mimicry awaits experimental validation [214, 275, 276].

Prospective Management Strategies

The feasible management strategies for pathogenic molecular mimicry include life-style changes [277], dietary adjustments or supplements [204, 211, 231, 277,278,279,280], improved vaccine production [161], and targeted manipulation of the intestinal microbiome [44, 47, 135, 214, 231, 281] (Table 3). Diets emphasizing fiber [211] and favoring the intestinal production of short chain fatty acids [280] and butyrate [203,204,205, 211] may preserve or strengthen the intestinal mucosal barrier. Probiotics [195, 213, 278], pharmacological mucosal protectors [275, 282, 283], and fecal microbiota transplantation (FMT) [284, 285] are other interventions that could mitigate the impact of gut-derived molecular mimicry [193, 195]. Studies that clarify the principal environmental sources for pathogenic molecular mimicry will direct the evaluation of appropriate management strategies.

Molecular Mimicry and Epigenetic Changes

Environmental factors may influence the causality of autoimmune hepatitis outside of molecular mimicry [52, 53, 55, 286]. Epigenetic changes induced by the environment can alter the transcriptional activity of immune regulatory genes without changing the nucleotide sequence of DNA [54, 287, 288]. Induced alterations in the methylation, acetylation, and phosphorylation of histones within the nucleosomes can reversibly up- or downregulate gene expression and influence the immune response [48, 53, 55].

Unlike molecular mimicry which is antigen-dependent, epigenetic changes can be induced by non-antigenic environmental factors that affect the activity of enzymes that modify histone structure [48]. Nutritional deficiencies [51], alcohol [289], tobacco [290], pollutants [291, 292], and psychological stress [50, 293] are environmental factors that may affect disease susceptibility by altering pivotal gene transcription rather than mimicking a self-antigen. Non-antigenic environmental stimuli may also change the intestinal microbiome and indirectly promote intestinal dysbiosis and gut-derived, antigen-induced, molecular mimicry [23, 26, 42, 47].

Molecular mimicry and epigenetic changes may contribute independently to the causality of autoimmune hepatitis depending on the nature of the environmental stimulus. Future investigations must establish the pathogenicity and relevance of molecular mimicry compared to epigenetic mechanisms. Therapeutic interventions could emerge that are mechanism-dependent.

Conclusions

Molecular mimicry may translate diverse environmental antigens into autoimmune hepatitis by increasing the avidity of immune cells for multiple homologous tissue antigens. Vaccinations and the intestinal microbiome may be prime environmental sources for peptides that mimic immune dominant epitopes of key self-antigens associated with autoimmune hepatitis. Disease-pertinent peptide mimics can be sought among diverse environmental antigens by searching for homologues of the known immune dominant epitopes of CYP2D6. Immune dominant epitopes of key antigens associated with type 1 autoimmune hepatitis await clarification. The conditions required for immunogenicity of the mimicking epitope must be clarified and include chaperone molecules, such as HSPs. Translocated gut-derived bacterial antigens should be evaluated as molecular mimics, and the conditions for increased intestinal permeability clarified. The implicated environmental homologue must be shown to be pathogenic by generating T and B cell reactivity to CYP2D6 and producing experimental autoimmune hepatitis. The relative contributions of molecular mimicry and epigenetic alterations to the causality of autoimmune hepatitis should be determined, and management strategies evaluated that are appropriate for the principal pathogenic mechanism.