Epitope Spreading: Lessons From Autoimmune Skin Diseases

Transcription

1 Epitope Spreading: Lessons From Autoimmune Skin Diseases REVIEW Lawrence S. Chan,* Carol J. Vanderlugt, Takashi Hashimoto, Takeji Nishikawa, John J. Zone,** Martin M. Black, Fenella Wojnarowska, Seth R. Stevens, Mei Chen, Janet A. Fairley, David T. Woodley,* Stephen D. Miller, and Kenneth B. Gordon *Medicine Service, Section of Dermatology, Lakeside Division, VA Chicago Health Care System, Chicago, Illinois, U.S.A.; Departments of Dermatology and Microbiology and Immunology, Northwestern University Medical School, Chicago, Illinois, U.S.A.; Department of Dermatology, Kurume University School of Medicine, Kurume, Japan; Department of Dermatology, Keio University School of Medicine, Tokyo, Japan; **Medicine Service, Section of Dermatology, Salt Lake City VA Medical Center, Salt Lake City, Utah, U.S.A.; Department of Dermatopathology, Guy s and St. Thomas Medical and Dental School, London, U.K.; Department of Dermatology, The Oxford Radcliffe Hospital, Oxford, U.K.; Department of Dermatology, Case Western Reserve University School of Medicine, Cleveland, Ohio, U.S.A.; Department of Dermatology, Medical College of Wisconsin, Milwaukee, Wisconsin, U.S.A. Autoimmune diseases are initiated when patients develop aberrant T and/or B cell responses against self proteins. These responses presumably are directed to single immunogenic epitopes on these proteins. Recent data in animal models of autoimmune diseases suggest that the targets of immune responses in autoimmunity do not remain fixed, but can be extended to include other epitopes on the same protein or other proteins in the same tissue, a phenomenon termed epitope spreading. The epitope spreading phenomenon also applies to situations in which tissue damage from a primary inflammatory process causes the release and exposure of a previously sequestered antigen, leading to a secondary autoimmune response against the newly released antigen. In experimental autoimmune animal diseases, epitope spreading seems to have significant physiologic importance in determining the course and duration of disease. In this paper, we review the current concepts in animal models of autoimmune diseases in order to define the epitope spreading phenomenon, and we then propose how this phenomenon might play a significant role in the development and the course of autoimmune skin diseases. Hopefully, an understanding of epitope spreading will help the dermatology community to better understand the pathogenesis of autoimmune skin diseases and to rationally fashion disease-specific immune therapy in the future. Key words: antibody-mediated/autoimmunity/cell-mediated. J Invest Dermatol 110: , 1998 For many years, both clinical and investigative dermatologists have noted the linkage of one disease with another, particularly the clustering of autoimmune diseases. In some instances, an autoimmune disease occurs in association with a second autoimmune disease, whereas in others it occurs in association with another inflammatory disease. Several hypotheses have been suggested to explain such intriguing disease associations. Among these is the hypothesis that autoimmune diseases are genetically determined. Therefore, the occurrence of one autoimmune disease heralds a genetic predisposition and susceptibility to develop another autoimmune disease. While acknowledging the genetic influence on autoimmune diseases, one must also consider other factors that may contribute to autoimmunity because identical genetic traits linked with disease-affected patients may also be present in healthy individuals. The development of cicatricial pemphigoid, a putative autoimmune disease, may serve as an example (Ahmed et al, 1991; Chan et al, 1997). When Stingl and Holubar reported the coexistence of lichen planus and bullous pemphigoid (BP) (Stingl and Holubar, 1975), and we reported three patients who developed BP subsequent to lichen planus (Davis et al, 1991), and five patients who, subsequent to an acute Manuscript received July 28, 1997; revised September 22, 1997; accepted for publication October 14, Reprint requests to: Dr. Lawrence S. Chan, Department of Dermatology, Northwestern University Medical School, Tarry 4 721, Mail Code T225, 300 E. Superior Street, Chicago, IL Abbreviation: NOD, nonobese diabetic. episode of mucosal epithelial injuries due to Stevens Johnson syndrome, developed ocular cicatricial pemphigoid (Chan et al, 1991), the hypothesis that we and Stingl and Holubar put forward was that the basement membrane zone (BMZ) injuries that occurred during the inflammatory process of lichen planus or the Stevens Johnson syndrome may have exposed the hidden antigens to the immune system, and subsequently induced an autoimmune response against the BMZ components. Not until the recent investigations employing animal models of autoimmune diseases, however, has further light been shed on this hypothesis. These animal models have increased our understanding of how a primary autoimmune or inflammatory process can induce a secondary autoimmune response. The autoimmune animal models allow step-by-step dissection of disease processes and have successfully documented the phenomenon of autoimmune reaction to previously sequestered antigens that are released from injured tissue by a primary autoimmune or inflammatory process. This phenomenon has been termed epitope spreading. In this article, we define epitope spreading and provide examples that occur in animal models of autoimmune diseases. Selected examples of human autoimmune skin diseases that are possible manifestations of epitope spreading are illustrated. Because experimental autoimmune diseases cannot be induced in humans for obvious ethical reasons, the examples of epitope spreading in human skin diseases are not documented with the same scientific vigor as the animal models. EPITOPE SPREADING: DEFINITION Epitope spreading can be defined as a specific autoreactive lymphocyte (T or B cell) response to endogenous epitopes, which are distinct X/98/$10.50 Copyright 1998 by The Society for Investigative Dermatology, Inc. 103

2 104 CHAN ET AL. THE JOURNAL OF INVESTIGATIVE DERMATOLOGY of clinical relapse, these mice demonstrate proliferative and delayed type hypersensitivity responses to a secondary, subdominant epitope of 14 amino acids (amino acid ) that is distinct and at a distance from the primary immunodominant epitope (McRae et al, 1995). Interestingly, the induction of tolerance to the secondary subdominant epitope prior to clinical relapse can prevent ongoing clinical disease based on autoimmunity to the primary epitope. Thus, not only is epitope spreading within proteolipid protein evident during the course of relapsing experimental autoimmune encephalomyelitis, it also seems to be responsible for the relapsing nature of the disease (Vanderlugt and Miller, 1996). Figure 1. Model of epitope spreading in chronic cell-mediated and antibody-mediated autoimmune diseases. The presentation of the primary (1 ) epitope to the T helper cells (Th) leads to cell- or antibody-mediated tissue injury and release of a secondary (2 ) new antigen (Ag) or 2 epitope of the same Ag. The presentation of the 2 epitope to the naïve Th subsequently leads to the development of cell- or antibody-mediated autoimmunity against the 2 Ag (or epitope). APC, antigen presenting cells (Langerhans cells, macrophages, dendritic cells, B cells); Mac., macrophages; PC, plasma cells; Com., complements; PMN, polymorphic nuclear leukocytes; Eos., eosinophils; Lym. Tiss., lymphoid tissues; Prolif., proliferation; Differ., differentiation; MIP1α, macrophage inflammatory protein 1α; IFNγ, interferon-γ; TNFβ, tumor necrosis factor-β; LT, leukotriene; ( ), help. (Vanderlugt and Miller, 1996.) from and non-cross-reactive with the disease-inducing epitopes, on the (same or different) proteins secondary to the release of such self protein during a chronic autoimmune or inflammatory response (Vanderlugt and Miller, 1996). That means, in essence, that a primary autoimmune or inflammatory process may cause tissue damage in such a manner that certain protein components that are immunologically hidden from the immune system become revealed and evoke a secondary autoimmune response. EPITOPE SPREADING: LESSONS FROM ANIMAL MODELS OF AUTOIMMUNE DISEASES Animal models of autoimmune diseases have provided much of the information available concerning the spreading of antigenic epitopes during the course of autoimmune disease. Unlike human diseases, animal models provide a system in which autoimmune responses can be studied in a controlled fashion from the initiation of the response against self-antigens, even before the clinical symptoms and signs become apparent, and followed at each stage of disease thereafter. In these animal models, epitope spreading can be predominantly B cell (antibody) or T cell mediated. In either case, epitope spreading within the same protein (intramolecular) and/or to other distinct proteins (intermolecular) within the same tissue or protein complex has been observed. Importantly, the data obtained from these animal models indicate that, not only does the phenomenon of epitope spreading exist, it is indeed very important for the clinical manifestations of autoimmune diseases. Figure 1 is a schematic diagram depicting a model of epitope spreading in chronic cell-mediated and antibodymediated autoimmune diseases. Epitope spreading in T cell mediated autoimmune responses within the same protein The most intensively studied of all animal models of autoimmunity is experimental autoimmune encephalomyelitis. In one form of this disease relapsing experimental autoimmune encephalomyelitis relapsing/remitting hind limb paralysis of the animals can be elicited in SJL/J mice after immunization with proteolipid protein in complete Fruend s adjuvant. A disease course that is identical to the disease induced by immunization with whole protein can be induced by immunization with a small domain of the whole protein. This small peptide domain, the so-called immunodominant T cell epitope peptide, consists of only 13 amino acids (amino acid ). As expected, the mice immunized with this peptide developed a strong T cell proliferative response and delayed type hypersensitivity response to the inciting peptide; however, at the time Epitope spreading in T cell mediated autoimmune responses involving different proteins within the same tissue The nonobese diabetic (NOD) mouse is a model of the human disease, in which diabetes mellitus occurs spontaneously in inbred NOD mice. The initial pathology in the pancreas of an NOD mouse is a T cell immune reaction, at µ4 wk of age, against a specific pancreatic β cell protein, glutamic acid decarboxylase; however, there is no clinical manifestations of diabetes at this time. Yet, as this reactivity against β cell continues, new reactions develop against other β cell proteins including carboxypeptidase H, insulin, and heat shock protein 65. It is only after the spread of immune reactivity to the proteins other than the original target (glutamic acid decarboxylase) has occurred that insulitis (massive cellular infiltration of the pancreatic islets) and clinical diabetes develops (Kaufman et al, 1993; Tisch et al, 1993). Thus, as in the case of relapsing experimental autoimmune encephalomyelitis, epitope spreading may play a role in the development of spontaneous diabetes in the NOD mouse model. Epitope spreading in autoantibody-mediated autoimmune response Patients with systemic lupus erythematosus are well known to have autoantibodies directed against multiple epitopes of the Ro/La ribonucleoprotein complex and against spliceosomes. Clinical disease in the form of a systemic lupus erythematosus like syndrome (manifested with proteinuria and cytopenia, and the presence of antinuclear antibodies) can be elicited in mice by immunization of a human spliceosome peptide. In the mice immunized with spliceosome protein, or in mice immunized with the human La protein, the original antibody response against either of the immunized peptides quickly spreads to other regions of these complexes, leading to a multi-epitope immune response (James et al, 1995; Topfer et al, 1995). Thus, the initial singular peptide antibody response spread to other proteins that could account for the strikingly variable antibody responses against numerous proteins commonly seen in human patients with systemic lupus erythematosus. EPITOPE SPREADING: LESSONS FROM HUMAN AUTOIMMUNE SKIN DISEASES Epitope spreading resulted from injuries in autoimmune dermatoses Intramolecular (intra-protein) epitope spreading Pemphigus vulgaris Pemphigus vulgaris is a blistering disease characterized by autoantibodies directed against desmoglein 3, an important structural component of epidermal desmosomes. Clinically, this disease frequently begins with oral lesions and later progresses to diffuse involvement of the skin. Recently, it has been observed that patients with lesions restricted to the mucosa have antibodies that recognize specific epitopes of desmoglein 3. These autoantibodies do not crossreact with the pemphigus foliaceus antigen, desmoglein 1, and are generally not able to induce blister in neonatal mice in passive transfer experiments; however, in patients with mucocutaneous lesions autoantibodies, which recognize a secondary and distinct epitope on desmoglein 3 and are cross-reactive to desmoglein 1, are detected. And these autoantibodies are able to induce blister in neonatal mice. Thus, in patients whose disease has progressed from mucosal pemphigus to mucocutaneous pemphigus, epitope spreading has likely occurred which accounts for the heterogenous antibody response seen in these patients. One could envision that the autoimmune reaction against

3 VOL. 110, NO. 2 FEBRUARY 1998 EPITOPE SPREADING IN CUTANEOUS AUTOIMMUNITY 105 specific desmoglein 3 epitope on mucosae may have induced injuries and exposed a secondary new desmoglein 3 epitope to the autoreactive lymphocytes, leading to autoimmune reaction against the secondary epitope, which is apparently present in the skin and shares significant homology with desmoglein 1. Thus, these data suggest that the evolution of the heterogenous antibody response to desmoglein 3 may account for the progression of disease seen in patients with pemphigus vulgaris (Ding et al, 1997). Epidermolysis bullosa acquisita Patients with epidermolysis bullosa acquisita have IgG autoantibodies against epidermolysis bullosa acquisita antigen (type VII collagen). These autoantibodies target the noncollagenous globular amino terminal domain termed NC1 and do not recognize the collagenous helical portion of the molecule, nor the noncollagenous globular domain (NC2) at the carboxyl terminus (Gammon et al, 1993; Lapiere et al, 1993). As suggested by many examples of autoimmune disease models in animals (Lehmann et al, 1992; McRae et al, 1995), the original development of autoimmunity to type VII collagen in epidermolysis bullosa acquisita patients likely involved primarily or exclusively a single antigenic epitope within the NC1 domain. Nevertheless, Gammon et al (1993), Lapiere et al (1993), and Tanaka et al (1994) have documented that epitope spreading occurs in cases of epidermolysis bullosa acquisita. All three laboratories have reported that the serum from a single epidermolysis bullosa acquisita patient contained IgG autoantibodies that recognized multiple regions of the NC1 domain (Gammon et al, 1993; Lapiere et al, 1993; Tanaka et al, 1994). The mechanism of epitope spreading in this situation may be very similar to that envisioned in animal models. Bullous pemphigoid Assuming that autoimmunity against the BP antigens located in the upper lamina lucida/hemidesmosome also initially involved predominantly or exclusively a single antigenic epitope, studies have suggested that epitope spreading occurs in BP. Using recombinant proteins, synthetic peptides, or blocking methods, Rico et al, Tanaka et al, and Sugi et al have demonstrated that a single BP patient s serum containing IgG autoantibodies that are able to recognize multiple epitopes within the 230 kda glycoprotein designated as BP antigen 1 (Sugi et al, 1989; Rico et al, 1990; Tanaka et al, 1991). Similarly, intraprotein epitope spreading may occur in some cases of autoimmunity against the BP antigen 2, a 180 kda protein with an extracellular collagenous domain (Balding et al, 1996). If the same assumption holds, interprotein epitope spreading also occurred in BP, because many BP patients have IgG autoantibodies against both BP230 and BP180 (Labib et al, 1986) that are products of two completely different genes located on two separated chromosomes. The mechanism of this type of epitope spreading is probably identical or very similar to that described in animal models (Lehmann et al, 1992; McRae et al, 1995). Intermolecular (inter-protein) epitope spreading Pemphigus vulgaris conversion to pemphigus foliaceus One very illustrative example of epitope spreading was reported in patients who initially developed pemphigus vulgaris with typical clinical manifestations, histopathology, immunopathology, and autoantibodies to desmoglein 3; and who later converted to pemphigus foliaceus. This conversion to pemphigus foliaceus was associated with an alteration of the pemphigus vulgaris clinical phenotype towards the phenotype of pemphigus foliaceus with typical clinical manifestations, histopathology, immunopathology, and autoantibodies to desmoglein 1 (Kawana et al, 1994). The interesting and puzzling aspect of these cases is why the disease phenotype was changed from a severer form of disease to a milder form, as a result of epitope spreading. Linear IgA bullous dermatosis with autoantibodies to BP antigen 1 Linear IgA bullous dermatosis is an IgA-mediated blistering skin disease characterized by subepidermal blisters and a neutrophilic infiltration along the skin BMZ (Wojnarowska et al, 1988; Cohen et al, 1997). Although the target antigen in linear IgA bullous dermatosis was initially thought to be a distinct BMZ component (Zone et al, 1990), more recent evidence suggests that the target antigen for these IgA autoantibodies is an epitope within the extracellular portion of the 180 kda BP antigen 2 (Pas et al, 1997). 1,2 It has also been demonstrated that IgG class autoantibodies may be involved (Zone et al, 1994; Chan et al, 1995; Darling et al, 1995). Moreover, it has been shown that IgG or IgA circulating autoantibodies from these patients also recognized the 230 kda BP antigen 1, in addition to the 180 kda BP antigen 2 (Collier et al, 1994; Chan et al, 1995; Darling et al, 1995). This initial autoimmunity to the BP antigen 2 that then expands to involve BP antigen 1 is consistent with the concept of epitope spreading. BP with autoantibodies to another basement membrane antigen p105 BP is characterized histopathologically by subepidermal blister and an inflammtory cell infiltrate along the skin BMZ that is rich in eosinophils (Cohen et al, 1997). The initial injury caused by the subepidermal blistering process may expose other skin BMZ components to autoreactive lymphocytes, and subsequently lead to autoimmune response against another BMZ protein termed p105, located in the lower lamina lucida (Chan et al, 1993; Cotell et al, 1994). Systemic lupus erythematosus associated with epidermolysis bullosa acquisita Systemic lupus erythematosus is a systemic autoimmune disease characterized by the development of autoantibodies against multiple self proteins in various tissues. Some patients with systemic lupus erythematosus have immunoglobulins deposited at the skin BMZ despite the absence of overt skin lesions, i.e., a positive lupus band. Presumably, the immune deposits at the skin BMZ could result in subclinical injury, and expose the anchoring fibril (type VII collagen) to the autoreactive lymphocytes, which ultimately sensitized and invoked autoimmunity to type VII collagen and produced a disease phenotype similar to epidermolysis bullosa acquisita (Gammon et al, 1985; Barton et al, 1986; Yell et al, 1995). Epitope spreading resulted from injuries in inflammatory dermatoses and other inflammatory diseases Lichen planus associated with BP Lichen planus is an idiopathic chronic inflammatory skin disease characterized histopathologically by a bandlike dermal infiltrate of predominantly lymphocytes along the skin BMZ and liquefaction degeneration of basal keratinocytes (Toussaint and Kamino, 1997). These pathologic injuries in lichen planus perturb the skin BMZ (Smoller and Glusac, 1994; Marren et al, 1994). Thus, this inflammatory process may expose skin BMZ components to autoreactive lymphocytes that then induce autoimmunity against skin BMZ components, such as the BP antigens (Stingl and Holubar, 1975; Sobel et al, 1976; Hintner et al, 1979). Clinical BP with autoantibodies against the BP antigens have been documented in such patients (Davis et al, 1991; Tamada et al, 1995). Figure 2 illustrates another example: a patient who had a long history of lichen planus with characteristic nonbullous poligonal papules and plaques and then developed a new blistering disease. The new blistering disease had features identical to those of BP as documented by histopathology, linear IgG deposits at the patient s perilesional skin BMZ, and circulating IgG autoantibodies that bound to the epidermal side of salt-split skin and labeled the 180 kda BP antigen 2 (Fig 2). Psoriasis associated with BP and pemphigoid-like blistering disease Psoriasis is a chronic inflammatory and hyperproliferative skin disease characterized histopathologically by dermal and epidermal inflammatory infiltration (Toussaint and Kamino, 1997). These inflammatory cells include neutrophils, activated T cells and macrophages (a professional antigen presenting cell), and perivascular dendritic cells (Stingl et al, 1977; Baadsgaard et al, 1989, 1990; Morganroth et al, 1991). In addition to the inflammatory and hyperproliferative characteristics, psoriasis skin is easily injured with even gentle removal of scale over a psoriasis lesion 1 Zone JJ, Taylor TB, Meyer LJ, Petersen MJ: The 97-kDa linear IgA bullous disease antigen is a portion of the extracellular domain of the 180-kDa bullous pemphigoid antigen. J Invest Dermatol 108:563, 1997 (abstr.) 2 Tran HH, Schumann H, Balding S, Giudice GJ, Bruckner-Tuderman L, Marinkovich MP: LAD-1 is a collagenous component of keratinocyte adhesion complexes which assembles into a high molecular weight complex and which has homology to BP180. J Invest Dermatol 108:586, 1997 (abstr.)

4 106 CHAN ET AL. THE JOURNAL OF INVESTIGATIVE DERMATOLOGY Figure 2. Lichen planus associated with BP. Western blot analyses detected a 180 kda BP antigen recognized by the IgG autoantibodies from a patient affected with lichen planus (lane 2) who subsequently developed a new blistering disease with IgG autoantibodies that label the epidermal side of salt-split skin substrate. Controls include sera from two BP patients (lanes 1 and 3) and serum from a normal volunteer (lane 4). resulting in dermal bleeding, so-called Auspitz sign (Toussaint and Kamino, 1997). Furthermore, neutrophil elastase, identified along the BMZ of psoriatic lesional skin, could play a role in the destruction of the epidermal dermal junction (Glinski et al, 1990). One could envision that the combination of psoriatic injuries caused by chronic inflammation, the trafficking of activated lymphocytes, and the abundance of antigen presenting cells could expose BMZ components to autoreactive lymphocytes and induce autoimmunity against these BMZ components, such as the BP antigens. Such autoimmune reaction may then subsequently lead to autoimmune disease BP. As reported by Grattan, seven patients developed BP after the onset of psoriasis from 1 to 30 y (Grattan, 1985). Kirtschig et al also reported four patients who, after a long history of psoriasis, developed BP with circulating autoantibodies against the BP antigens (Kirtschig et al, 1996). Figure 3 illustrates another clinical example, in which the patient had a long history of psoriasis with characteristic nonbullous sharply demarcated scaly plaques predmoninantly on her extensor surfaces, and then developed a new blistering disease after receiving a course of Psoralen- UVA treatment. The new blistering disease had features identical to those of BP as documented by histopathology, in situ and circulating IgG autoantibodies that bound to the skin BMZ, and circulating IgG autoantibodies that recognized the 230 kda BP antigen 1 (Fig 3). BMZ injuries induced by ultraviolet light in psoriasis patients receiving Psoralen-UVA treatments may promote the exposure of BMZ components to autoreactive lymphocytes (Thomsen and Schmidt, 1976; Abel and Bennett, 1979; Kirtschig et al, 1996). In addition, primary psoriasis has also been linked to the development of a secondary BP-like blistering disease 20 y after the onset of psoriasis. This blistering disease is characterized by IgG autoantibodies targeting a undefined 200 kda lower lamina lucida component (Chen et al, 1996). Stevens Johnson syndrome associated with secondary onset of cicatricial pemphigoid Stevens Johnson syndrome is an acute inflammatory disease involving mucous membranes and skin. Histopathologically, the acute lesion has a mononuclear cell infiltration along the skin BMZ, hydropic degeneration of basal keratinocytes, and subepidermal blisters (Cohen et al, 1997). Chan et al have reported five patients who, after an acute episode of Stevens Johnson syndrome, developed chronic ocular scarring inflammation confirmed to be cicatricial pemphigoid (Chan et al, 1991). In this scenario, one could envision that the initial pathologic injuries associated with florid Stevens Johnson syndrome released skin BMZ components and exposed them to autoreactive lymphocytes and induced autoimmunity against these structural self proteins. Figure 4 illustrates one of these cicatricial pemphigoid patients (Chan et al, 1991). A patient was also reported to develop cicatricial pemphigoid 5 y after the onset of a primary chronic inflammtory skin Figure 3. Psoriasis associated with BP. Clinical observation of new onset of BP in a patient with long history of psoriasis (a). Histopathology confirmed the diagnosis of BP in the blister (curved ) in association with psoriasis plaques (straight ). The diagnosis of BP was further documented by linear IgG deposition in the patient s peri-blister skin BMZ and by patient s circulating IgG autoantibodies that bound to skin BMZ (b) and recognized a 230 kda BP antigen by western blot [c, lane 2 was reacted with this patient s serum, lanes 1 and 3 were reacted with sera from two BP patient controls, normal human serum showed negative reaction (not shown)].

5 VOL. 110, NO. 2 FEBRUARY 1998 EPITOPE SPREADING IN CUTANEOUS AUTOIMMUNITY 107 Figure 4. Stevens Johnson syndrome associated with cicatricial pemphigoid. Histopathology documented the occurence of primary Stevens Johnson syndrome (a) that affected a patient who 6 mo later developed a secondary chronic inflammatory scarring ocular disease (b). At the time of acute episode of Stevens Johnson syndrome, extensive eosinophilic necrosis of keratinocytes, prominent mononuclear cell infiltration, and hydropic degeneration of the basal keratinocytes were observed (a). At the time of chronic ocular disease, biopsy showed linear IgG (c) deposition at the basement membrane. The patient had a high titer (1:2560) of circulating anti-basement membrane IgG autoantibodies that recognized a 120 kda epidermal antigen (Chan et al, 1991). disease, lichen sclerosis et atrophica, a condition characterized by inflammatory cell infiltration and skin BMZ disruption (Marren et al, 1996, 1997). Dermatitis herpetiformis associated with autoantibodies against BP antigens Dermatitis herpetiformis is a pruritic inflammatory blistering skin disease characterized histopathologically by subepidermal separation and a prominent neutrophilic infiltration along the papillary dermis of the skin BMZ (Cohen et al, 1997). Although IgA is deposited high within the papillary dermis juxtaposed to the skin BMZ, circulating autoantibodies to skin BMZ components are not detected. Therefore, it does not fulfill the complete criteria for an autoimmune disease (Rose and Bona, 1993), and, for the purpose of this paper, will be classified as an inflammatory disease. Within the setting of primary and chronic dermatitis herpetiformis, there have been several documented cases of BP autoantibodies developing from 4 mo to 5 y after the onset of dermatitis herpetiformis (Sander et al, 1989; Setterfield et al, 1997). One plausible explanation for these occurrences would be that the neutrophilic inflammation inherent in dermatitis herpetiformis perturbed the BMZ and lead to exposure of BMZ components to autoreactive lymphocytes and autoimmunity against the BP antigens. Discoid lupus erythematosus associated with autoantibodies against BP antigens Discoid lupus erythematosus is a cutaneous form of lupus characterized histopathologically by liquefaction degeneration of basal keratinocytes and a patchy lymphocytic infiltration around the BMZ of adnexal structures (Jaworsky, 1997). Ishikawa et al have documented the occurrence of autoantibodies to BP antigens developing after the onset of discoid lupus erythematosus (Ishikawa et al, 1997). Presumably, the chronic BMZ injury from the initial discoid lupus erythematosus releases and exposes BMZ components to autoreactive lymphocytes and induces autoimmunity against the BP antigens (Marren et al, 1994; Ishikawa et al, 1997). Crohn s disease associated with autoantibodies against type VII collagen Inflammatory bowel disease, especially Crohn s disease, is frequently associated with epidermolysis bullosa acquisita, an autoimmune disease characterized by autoantibodies against the noncollagenous (NC1) domain of type VII collagen (Lapiere et al, 1996). We recently identified the presence of type VII collagen in the intestinal epithelial BMZ. Interestingly, a significant number of patients affected by Crohn s disease had circulating IgG autoantibodies against the NC1 domain of type VII collagen. 3 Moreover, a dog with inflammatory bowel disease also had a high titer of circulating IgG autoantibody against the NC1 domain of type VII collagen. 3 One could envision that the inflammation originally invoked by Crohn s disease could perturb the intestinal epithelial BMZ in such a way that BMZ components are exposed and perhaps altered while in close juxtaposition with antigen presenting cells, and T and B lymphocytes, and this process results in on-going autoimmunity to type VII collagen. Ulcerative colitis associated with linear IgA bullous dermatosis and BP Another inflammatory bowel disease, ulcerative colitis, is known to be associated with linear IgA bullous dermatosis (Smith and Zone, 1996). A recent study of 70 adult patients with linear IgA bullous dermatosis in the U.K. revealed a 7.1% prevalence of ulcerative colitis in the patients affected by linear IgA bullous dermatosis, much higher than the 0.05% prevalance of ulcerative colitis in the general U.K. population (Paige et al, 1997). Paige et al documented eight cases of linear IgA bullous dermatosis preceded by ulcerative colitis by a median of 6.5 y (Paige et al, 1997). Although there is no literature on the presence of BP antigens in human colonic epithelial BMZ, Dr. Giudice had documented the presence of BP antigen 2 in mouse colorectal epithelial BMZ (Dr. George Giudice, Medical College of Wisconsin, personal communication). As in the case of autoantibodies against 3 Chen M, O Toole EA, Sanghavi J, Mahmud N, Kelleher D, Woodley DT: Type VII (anchoring fibril) collagen exists in human intestine and serves as an antigenic target in patients with inflammatory bowel disease. J Invest Dermatol 108:542, 1997 (abstr.)

6 108 CHAN ET AL. THE JOURNAL OF INVESTIGATIVE DERMATOLOGY type VII collagen in Crohn s disease, one could envision that the inflammation originally invoked by ulcerative colitis could perturb the intestinal epithelial BMZ in such a manner that BMZ components are exposed while in close juxtaposition with antigen presenting cells, and T and B lymphocytes, and this process thus results in on-going autoimmunity to linear IgA bullous dermatosis antigen (now part of BP antigen 2). In addition, BP has been reported to develop in patients who suffered from ulcerative colitis for a period ranging from 3 to 20 y before the onset of BP (Barth et al, 1988; Harrison et al, 1996). CONCLUSION In this paper, we define the autoimmune phenomenon termed epitope spreading and provide examples of this phenomenon in animal models of autoimmune diseases and selected examples of human autoimmune skin diseases. Whereas the molecular events in which epitope spreading occurs in human autoimmune skin diseases remain to be elucidated, we hypothesize that skin injuries from autoimmune or inflammatory processes could induce autoimmunity to a normally sequestered antigen (or epitope). This hypothesis is consistent with at least two general principles regarding the development of autoimmunity. First, an autoimmune or inflammatory event is capable of bringing all the neccessary components together at a localized tissue site for autoimmunity to be promoted and expanded. These components include: the autoantigen (the skin components: desmosomal or BMZ components), antigen presenting cells (Langerhans cells, macrophages, B cells, and other dendritic cells), and possibly autoreactive lymphocytes, an environment for optimal antigen presentation. Second, the distressed skin cells involved in the primary autoimmune or inflammatory process may have triggered a danger signal. These danger signals are essential to activate the antigen presenting cells that in turn deliver two activating signals, the MHC-peptide and the co-stimulatory molecules, i.e., B7 1 and B7 2, that promote interaction between antigen presenting cells and autoreactive lymphocytes in the induction of autoimmunity (Matzinger and Fuchs, 1996). The contents of this paper are not meant to be inclusive. Nevertheless, the examples provided are sufficient to illustrate how epitope spreading might occur in human autoimmune skin diseases. As it is not possible for ethical reasons to dissect the step-by-step mechanisms of the epitope spreading phenomenon occurring in human autoimmune skin diseases, confirmation will have to come from active animal disease models. It is the hope of the authors that this subject will alert the dermatology community to this interesting autoimmune phenomenon and that this information will ultimately benefit the dermatology community in the understanding and treatment of autoimmune skin diseases. This work is supported by a Clinical Investigator Award (K08 AR01961, National Institutes of Health, Bethesda, MD, L.S. Chan), two Merit Review Research Grants (VA Research Committee, Long Beach, CA, L.S. Chan, J.J. Zone), Research Project Grants (RO1-AR41045 & R01-AR33625, National Institutes of Health, Bethesda, MD, D.T. Woodley), a Training Grant (T32 AR , National Institutes of Health, Bethesda, MD, D.T. Woodley and K.B. Gordon), and a Leader Society Clinical Career Development Award (Dermatology Foundation, Evanston, IL, K.B. Gordon). REFERENCES Abel EA, Bennett A: Bullous pemphigoid: occurrence in psoriasis treated with psoralens plus long wave ultraviolet radiation. Arch Dermatol 115: , 1979 Ahmed AR, Foster S, Zaltas M, Notani G, Awdeh Z, Alper CA, Yunis EJ: Association of DQw7 (DQB1*0301) with ocular cicatricial pemphigoid. Proc Natl Acad Sci USA 88: , 1991 Baadsgaard O, Gupta AK, Taylor RS, Ellis CN, Voorhees JJ, Cooper KD: Psoriatic epidermal cells demonstrate increased numbers and function of non-langerhans antigen-presenting cells. J Invest Dermatol 92: , 1989 Baadsgaard O, Tong P, Elder JT, et al: UM4D4 (Cdw60) T cells are compartmentalized into psoriatic skin and release lymphokines that induce a keratinocyte phenotype expressed in psoriatic lesions. J Invest Dermatol 95: , 1990 Balding SD, Prost C, Diaz LA, Bernard P, Bedane C, Aberdam D, Giudice GJ: Cicatricial pemphigoid autoantibodies react with multiple sites on the BP180 extracellular domain. J Invest Dermatol 106: , 1996 Barth JH, Kelly SE, Wojnarowska F, Savin JA, Whittaker S, Cream JJ, White JE: Pemphigoid and ulcerative colitis. J Am Acad Dermatol 19: , 1988 Barton DD, Fine J-D, Gammon WR, Sams WM: Bullous systemic lupus erythematosus: an unusual clinical course and detectable circulating autoantibodies to the epidermolysis bullosa acquisita antigen. J Am Acad Dermatol 15: , 1986 Chan LS, Soong HK, Foster CS, Hammerberg C, Cooper KD: Ocular cicatricial pemphigoid occurring as a sequela of Stevens Johnson syndrome. JAMA 266: , 1991 Chan LS, Fine J-D, Briggaman RA, Woodley DT, Hammerberg C, Drugge RJ, Cooper KD: Identification and partial characterization of a novel 105-kDalton lower lamina lucida autoantigen associated with a novel immune-mediated subepidermal blistering disease. J Invest Dermatol 101: , 1993 Chan LS, Traczyk T, Taylor TB, Eramo LR, Woodley DT, Zone JJ: Linear IgA bullous dermatosis: characterization of a subset of patients with concurrent IgA and IgG anti-basement membrane autoantibodies. Arch Dermatol 131: , 1995 Chan LS, Hammerberg C, Cooper KD: Significantly increased of HLA-DQB1*0301 allele in patients with ocular cicatricial pemphigoid. J Invest Dermatol 108: , 1997 Chen KR, Shimizu S, Miyakawa S, Ishiko A, Shimizu H, Hashimoto T: Coexistence of psoriasis and an unusual IgG-mediated subepidermal bullous dermatosis: identification of a novel 200-kDa lower lamina lucida target antigen. Br J Dermatol 134: , 1996 Cohen LM, Skopicki DK, Harrist TJ, Clark WH: Non-infectious vesiculobullous and vesiculopustular diseases. In: Elder D, Elenitsas R, Jaworsky C, Johnson B (eds). Lever s Histopathology of the Skin, 8th edn. Lippincott-Raven, Philadelphia, 1997, pp Collier P, Wojnarowska F, Allen J, Kirtschig G: Molecular overlap of the IgA target antigens in the subepidermal blistering diseases. Dermatol 189 (Suppl. 1): , 1994 Cotell SL, Lapiere JC, Chen JD, Iwasaki T, Krusinski PA, Chan LS, Woodley DT: A novel 105-kDa lamina lucida autoantigen: association with bullous pemphigoid. J Invest Dermatol 103:78 83, 1994 Darling TN, Cardenas AA, Beard JS, Sau P, Yee CL, Zone JJ, Yancey KB: A child with antibodies targeting both linear IgA bullous dermatosis and bullous pemphigoid antigens. Arch Dermatol 131: , 1995 Davis AL, Bhogal BS, Whitehead P, Frith P, Murdoch ME, Leigh IM, Wojnarowska F: Lichen planus pemphigoides: its relationship to bullous pemphigoid. Br J Dermatol 125: , 1991 Ding X, Aoki V, Mascaro JM, Lopez-Swiderski A, Diaz LA, Fairley JA: Mucosal and mucocutaneous (generalized) pemphigus vulgaris show distinct autoantibody profiles. J Invest Dermatol 109: , 1997 Gammon WR, Woodley DT, Dole KC, Briggaman RA: Evidence that anti-basement membrane zone antibodies in bullous eruptions of systemic lupus erythematosus recognize epidermolysis bullosa acquisita antigen. J Invest Dermatol 84: , 1985 Gammon WR, Murrell DF, Jenison MW, et al: Autoantibodies to type VII collagen recognize epitopes in a fibronectin-like region of the noncollagenous (NC1) domain. J Invest Dermatol 100: , 1993 Glinski W, Jarzabek-Chorzelska M, Kuligowski M, Pierozynska-Dubowska M, Glinska- Ferenz M, Jablonska S: Basement membrane zone as a target for human neutrophil elastase in psoriasis. Arch Derm Res 282: , 1990 Grattan CEH: Evidence of an association between bullous pemphigoid and psoriasis. Br J Dermatol 113: , 1985 Harrison PV, Blewitt RW, Allen J, et al: Bullous pemphigoid and ulcerative colitis: a report of two cases and description of immunoblot findings. Br J Dermatol 134: , 1996 Hintner H, Tappeiner G, Honigsmann H, et al: Lichen planus and bullous pemphigoid. Acta Derm Venereol (Stockh) 59 (Suppl. 85):71 76, 1979 Ishikawa O, Zaw KK, Miyachi Y, Hashimoto T, Tanaka T: The presence of antibasement membrane zone antibodies in the sera of patients with non-bullous lupus erythematosus. Br J Dermatol 136: , 1997 James JA, Gross T, Scofield RH, Harley JB: Immunoglobulin epitope spreading and autoimmune disease after peptide immunization: Sm B/B -derived PPPGMRPP and PPPGIRGP induce splicesome autoimmunity. J Exp Med 181: , 1995 Jaworsky C: Connective tissue diseases. In: Elder D, Elenitsas R, Jaworsky C, Johnson B (eds). Lever s Histopathology of the Skin, 8th edn. Lippincott-Raven, Philadelphia, 1997, pp Kaufman DL, Clare-Salzer M, Tian J, et al: Spontaneous loss of T cell tolerance to glutamic acid decarboxylase in murine insulin-dependent diabetes. Nature 336:69 72, 1993 Kawana S, Hashimoto T, Nishikawa T, Nishiyama S: Changes in clinical features, histologic findings, and antigen profiles with development of pemphigus foliaceus from pemphigus vulgaris. Arch Dermatol 130: , 1994 Kirtschig G, Chow ETY, Venning VA, Wojnarowska FT: Acquired subepidermal bullous diseases associated with psoriasis: a clinical, immunopathological and immunogenetic study. Br J Dermatol 135: , 1996 Labib RS, Anhalt GJ, Patel HP, Mutasim DF, Diaz LA: Molecular heterogeneity of the bullous pemphigoid antigens as detected by immunoblotting. J Immunol 136: , 1986 Lapiere J-C, Woodley DT, Parente MG, Iwasaki T, Wynn KC, Christiano AM, Uitto J: Epitope mapping of type VII collagen. Identification of discrete peptide sequences recognized by sera from patients with acquired epidermolysis bullosa. J Clin Invest 92: , 1993 Lapiere J-C, Chan LS, Woodley DT: Epidermolysis bullosa acquisita. In: Arndt KA, LeBoit PE, Robinson JK, Wintroub BU (eds). Cutaneous Medicine and Surgery. WB Saunders, Philadelphia, 1996, pp Lehmann PV, Forsthuber T, Miller A, Sercarz EE: Spreading of T-cell autoimmunity to cryptic determinants of an autoantigen. Nature 358: , 1992 Marren P, De Berker D, Wojnarowska F, Davies MG, Leigh IM, Burge SM: The dermo epidermal interface in lichen planus and lupus erythematosus. Europ J Dermatol 4:58 62, 1994

7 VOL. 110, NO. 2 FEBRUARY 1998 EPITOPE SPREADING IN CUTANEOUS AUTOIMMUNITY 109 Marren P, Neild V, Frith P, Wojnarowska F: Cicatricial pemphigoid and lichen sclerosis. J Europ Acad Dermatol Venerol 7:71 74, 1996 Marren P, Dean D, Charnock M, Wojnarowska F: The basement membrane zone in lichen sclerosis: an immunohistochemical study. Br J Dermatol 136: , 1997 Matzinger P, Fuchs EJ: Beyond self and non-self : immunity is a conversation, not a war. J NIH Res 8:36 39, 1996 McRae BL, Vanderlugt CL, Dal Canto MC, Miller SD: Functional evidence for epitope spreading in the relapsing pathology of experimental autoimmune encephalomyelitis. J Exp Med 182:75 85, 1995 Morganroth GS, Chan LS, Weinstein GD, Voorhees JJ, Cooper KD: Proliferating cells in psoriatic dermis are comprised primarily of T cells, endothelial cells, and factor XIIIa perivascular dendritic cells. J Invest Dermatol 96: , 1991 Paige DG, Leonard JN, Wojnarowska F, Fry L: Linear IgA disease and ulcerative colitis. Br J Dermatol 136: , 1997 Pas HH, Kloosterhuis GJ, Heeres K, van der Meer JB, Jonkman MF: Bullous pemphigoid and linear IgA dermatosis sera recognize a similar 120-kDa keratinocyte collagenous glycoprotein with antigenic cross-reactivity to BP180. J Invest Dermatol 108: , 1997 Rico MJ, Korman NJ, Stanley JR, Tanaka T, Hall RP: IgG antibodies from patients with bullous pemphigoid bind to localized epitopes on synthetic peptides encoded by bullous pemphigoid antigen cdna. J Immunol 145: , 1990 Rose NR, Bona C: Defining criteria for autoimmune disease (Witebsky s postulates revisited). Immunol Today 14: , 1993 Sander HM, Uyz MMP, Peters MS: Bullous pemphigoid and dermatitis herpetiformis. mixed bullous disease or coexistence of two separate entities. J Cutan Pathol 16: , 1989 Setterfield J, Bhogal B, Black MM, McGibbon DH: Dermatitis herpetiformis and bullous pemphigoid: a developing association confirmed by immunoelectronmicroscopy. Br J Dermatol 136: , 1997 Smith EP, Zone JJ: Linear IgA bullous dermatosis. In: Arndt KA, LeBoit PE, Robinson JK, Wintroub BU (eds). Cutaneous Medicine and Surgery. WB Saunders, Philadelphia, 1996, pp Smoller BR, Glusac EJ: Immunofluorescent analysis of the basement membrane zone in lichen planus suggests destruction of the lamina lucida in bullous lesions. J Cutaneous Pathol 21: , 1994 Sobel S, Miller R, Shatin H: Lichen planus pemphigoides. Arch Dermatol 112: , 1976 Stingl G, Holubar K: Coexistence of lichen planus and bullous pemphigoid. Br J Dermatol 93: , 1975 Stingl G, Wolff K, Diem E, Baumgartner G, Knapp W: In situ identification of lymphoreticular cells in benign and malignant infiltrates by membrane receptor sites. J Invest Dermatol 69: , 1977 Sugi T, Hashimoto T, Hibi T, Nishikawa T: Production of human monoclonal antibasement membrane zone (BMZ) antibodies from a patient with bullous pemphigoid (BP) by Epstein Barr virus transformation. J Clin Invest 84: , 1989 Tamada Y, Yokochi K, Nitta Y, Ikeya T, Hara K, Owaribe K: Lichen planus pemphigoides: identification of 180 kd hemidesmosome antigen. J Am Acad Dermatol 32: , 1995 Tanaka M, Hashimoto T, Amagai M, et al: Characterization of bullous pemphigoid antibodies by use of recombinant bullous pemphigoid antigen proteins. J Invest Dermatol 97: , 1991 Tanaka T, Furukawa F, Imamura S: Epitope mapping for epidermolysis bullosa acquisita autoantibody by molecularly cloned cdna for type VII collagen. J Invest Dermatol 102: , 1994 Thomsen K, Schmidt J: PUVA-induced bullous pemphigoid. Br J Dermatol 95: , 1976 Tisch R, Yang XD, Singer SM, Liblau RS, Fugger L, McDevitt HO: Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. 336:72 75, 1993 Topfer F, Gordon T, McCluskey J: Intra- and intermolecular spreading of autoimmunity involving the nuclear self-antigens La (SS-B) and Ro (SS-A). Proc Natl Acad Sci USA 92: , 1995 Toussaint S, Kamino H: Noninfectious erythematous, papular, and squamous diseases. In: Elder D, Elenitsas R, Jaworsky C, Johnson B (eds). Lever s Histopathology of the Skin, 8th edn. Lippincott-Raven, Philadelphia, 1997, pp Vanderlugt CJ, Miller SD: Epitope spreading. Current Opinion Immunol 8: , 1996 Wojnarowska F, Marsden R, Bhogal B, Black M: Chronic bullous disease of childhood, childhood cicatricial pemphigoid and linear IgA disease of adults: a comparative study demonstrating clinical and immunopathological overlap. J Am Acad Dermatol 19: , 1988 Yell JA, Allen J, Wojnarowska F, Kirtschig G, Burge SM: Bullous systemic lupus erythematosus: revised criteria for diagnosis. Br J Dermatol 132: , 1995 Zone JJ, Taylor TB, Kadunce DP, Meyer LJ: Identification of the cutaneous basement membrane zone antigen and isolation of antibody in linear immunoglobulin A bullous dermatosis. J Clin Invest 85: , 1990 Zone JJ, Smith EP, Powell D, Taylor TB, Smith JB, Meyer LJ: Antigenic specificity of antibodies from patients with linear basement membrane deposition of IgA. Dermatol 189 (Suppl. 1):64 66, 1994