|Home | About | Journals | Submit | Contact Us | Français|
Bullous pemphigoid (BP) is a cutaneous autoimmune inflammatory disease associated with subepidermal blistering and autoantibodies against BP180, a transmembrane collagen and major component of the hemidesmosome. Numerous inflammatory cells infiltrate the upper dermis in BP. IgG autoantibodies in BP fix complement and target multiple BP180 epitopes that are highly clustered within a non-collagen linker domain, termed NC16A. Anti-BP180 antibodies induce BP in mice. In this study, we generated a humanized mouse strain, in which the murine BP180NC14A is replaced with the homologous human BP180NC16A epitope cluster region. We show that the humanized NC16A (NC16A+/+) mice injected with anti-BP180NC16A autoantibodies develop BP-like subepidermal blisters. The F(ab′)2 fragments of pathogenic IgG fail to activate complement cascade and are no longer pathogenic. The NC16A+/+ mice pretreated with mast cell activation blocker or depleting of complement or neutrophils become resistant to BP. These findings suggest that the humoral response in BP critically depends on innate immune system players.
Bullous pemphigoid (BP) is a cutaneous autoimmune disease characterized by subepidermal blisters, a dermal inflammatory infiltrate, and in vivo deposition of autoantibodies and complement components along the basement membrane zone [1–3] BP autoantibodies avidly fix complement in vitro [4–7] These autoantibodies target two hemidesmosomal components: BP230 (also termed BPAG1), a member of the plakin protein family, and BP180 (also termed BPAG2, type XVII collagen) [3, 8–15] BP180 is a transmembrane homo-trimeric glycoprotein with a subunit MW of 180 kDa [3, 16, 17] Its C-terminal ectodomain consists of a long collagenous stretch interrupted and flanked by 16 non-collagen sequences  The membrane-proximal non-collagen linker domain (termed NC16A) harbors multiple epitopes recognized by BP autoantibodies [18, 19] Although the human BP180 shares high overall homology with the murine BP180, the NC16A domain is very poorly conserved in the murine protein (termed NC14A), resulting in a lack of immune-reactivity cross these two species 
A variety of cellular lineages have been identified in these inflammatory infiltrates, including eosinophils, neutrophils, lymphocytes, mast cells, and monocyte/macrophages [3, 21–24] Mast cells found in BP lesions exhibit morphological changes suggesting degranulation [24, 25] Lesional skin in BP patients exhibits several granular proteins derived from leukocytes, such as eosinophil cationic protein, eosinophil major basic protein, and neutrophil-derived myeloperoxidase  Various inflammatory mediators that can activate mast cells or leukocytes have been identified in lesional skin and/or blister fluids of BP patients, including C5a, eosinophilic/neutrophilic chemotactic factors, histamine, leukotrienes, and various cytokines (e.g. interleukin-1, -2, -5, -6, -8, tumor necrosis factors and interferon-γ) [26–33] Several proteinases are also found in BP blister fluid, including plasmin, collagenase, elastase, and 92-kD gelatinase [34–37]
The pathogenicity of anti-BP180 antibodies was initially demonstrated by IgG passive transfer experiment, in which rabbit antibodies against the mBP180NC14A domain, when injected into neonatal mice, induced a blistering skin phenotype that closely resembled human BP at the clinical, histological and immunopathological levels  More recently, Nishie et al showed that anti-BP180 autoantibodies from BP patients also induced BP in the humanized BP180 mice  Clinically, serum levels of BP autoantibodies to NC16A are correlated with the severity of disease [39–41] Thus, there is no doubt that anti-BP180 antibodies are able to induce BP. But, whether pathogenic anti-BP180 autoantibodies act alone or in combination with key innate immune system players to drive the disease remains unresolved.
Our present approach to this important question involved generating a humanized mouse strain in which the murine genomic region encoding the BP180NC14A domain was replaced with the homologous human BP180NC16A sequence. When injected into NC16A+/+ transgenic mice, affinity-purified anti-BP180NC16A autoantibodies from the sera of BP patients induced a BP-like disease. We then determined whether anti-BP180NC16A autoantibody-caused tissue damage at the BMZ requires complement, mast cells and neutrophils.
Serum samples were collected from three patients with active BP (BP1, BP2, and BP3). Rabbit anti-hBP180NC16A (R594) and rabbit anti-mBP180NC14A (R530) antibodies were generated by our laboratories as described previously  Briefly, New Zealand White rabbits were immunized with the GST-hBP180NC16A or GST-mNC14A fusion protein. IgG fractions from sera of the immunized rabbits were purified using a protein G Sepharose column (Sigma). BP180NC16A-specific autoantibodies from BP patients were purified using a BP180NC16A-glutathione Sepharose column. F(ab′)2 fragments of anti-BP180NC16A autoantibodies were generated by the pepsin digestion method 
To generate the humanized BP180NC16A mice (NC16A+/+), the murine BP180 genomic segment encoding the NC14A domain (exons 17 through 18) were replaced by the human NC16A genomic sequence (exons 18 through 19; see Figure 1). The targeting vector was electroporated into 129/Ola mouse ES cells. After the neo gene was excised from the targeted locus by Flp treatment, neo-minus cells were microinjected into C57BL/6J mouse blastocysts, which were then implanted into pseudopregnant recipients. High-chimera males were mated with B6 females to produce F1 heterozygotes (NC16A+/−). Crossing F1+/− produced F2 homozygotes expressing only mhBP180NC16A hybrid antigen (NC16A+/+). C57BL/6J mice were obtained from Jackson Laboratories. NC16A+/+ and C57BL/6J breeders were hosted at the University of North Carolina-Chapel Hill Animal Facility. Neonatal mice of NC16A+/+ and their littermates and C57BL/6J (24–48 h old with body weight between 1.4–1.6 g) were used for antibody passive transfer experiments. Animal care and animal experiments were in accordance with the Animal Care Committee at the University of North Carolina-Chapel Hill.
A 50 μl aliquot of sterile IgG (control IgG, BP IgG, R594, or R530) in PBS was administered to neonatal mice by i.d. or i.p. injection  The mouse skin was examined 24 h after injection. The extent of cutaneous disease was scored as follows: “(−)”, no detectable skin disease; “1+”, mild erythematous reaction with no evidence of the “epidermal detachment” sign; “2+”, intense erythema and “epidermal detachment” sign involving 10–50% of the epidermis in localized areas; and “3+”, intense erythema with frank “epidermal detachment” sign involving more than 50% of the epidermis.
After clinical examination, the animals were sacrificed, and skin and serum specimens were obtained. The skin sections were used for routine histological examination by light microscopy (H/E staining) to localize the lesional site and neutrophil infiltration, and by toluidine blue staining to quantify mast cells and mast cell degranulation. Direct IF assays were done to detect deposition of human or rabbit IgG and mouse C3 at the basement membrane zone. Myeloperoxidase (MPO) enzymatic assay was performed to quantify the neutrophil accumulation at the skin injection site as described below. The sera of injected animals were tested by indirect IF techniques to determine the titers of anti-BP180 antibodies using humanized NC16A mouse skin cryosections as substrate. Direct and indirect IF studies were performed as previously described  using commercially available FITC-conjugated goat anti-human and goat anti-rabbit IgG (Kirkeggard & Perry Laboratories Inc.). Monospecific goat anti-mouse C3 IgG was purchased from Cappel Laboratories.
Infiltrating neutrophils in the antibody-injected skin were detected by indirect IF using an anti-mouse myeloperoxidase (MPO) antibody (Cell Sciences). Infiltrating eosinophils were identified by an eosinophil-specific histochemical staining  For quantification of neutrophil accumulation in the skin, tissue MPO activity in skin sites of the injected animals was assayed as described [44, 45], using purified MPO as standard. MPO content was expressed as relative MPO activity (OD460nm reading/mg protein of the mouse skin injected with pathogenic anti-mBP180 IgG minus OD460nm reading/mg protein of the mouse skin injected with control IgG). Protein concentrations were determined by the Bio-Rad dye-binding assay using BSA as a standard.
Mast cells and mast cell degranulation in skin samples were quantified by examination of toluidine blue stained sections [46, 47] Total number of mast cells was counted and classified as degranulated (>10% of the granules exhibiting fusion or discharge) or normal in five fields under a light microscope. The results were expressed as percentage of mast cells degranulating (number of degranulating mast cells/total number of mast cells/field).
To block mast cell degranulation, neonatal NC16A+/+ mice were injected i.d. with cromolyn sodium (Sigma, St. Louis, MO) (10 μg/g body weight), a mast cell degranulation inhibitor  To deplete complement, neonatal NC16A+/+ mice were pretreated with cobra venom factor (3 units/100 μl PBS, i.p.)  To deplete circulating neutrophils, neonatal NC16A+/+ mice were pretreated with a polyclonal rabbit anti-murine neutrophil antibody, AI-A31140, which selectively depletes mouse neutrophils in vivo (Accurate Chemical & Scientific Corp.)  The cromolyn-treated mice were then injected i.d with anti-BP180NC16A autoantibodies 90 min later  The complement- and neutrophil-depleted mice were injected i.d with anti-BP180NC16A autoantibodies 12 h later [42, 45] All mice were examined 24 h after pathogenic antibody injection.
Skin samples were dissected from the injected sites of test and control animals, and immersed in a fixative of 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2. The tissue blocks were post-fixed with OsO4 and then dehydrated with a series of graded ethanol. Thin sections were contrasted with uranyl acetate and lead citrate and examined with a Hitachi-600 transmission electron microscope at 75 KV 
The data are expressed as mean ± SEM and were analyzed using the Student’s t-test. A p value less than 0.05 was considered significant.
To directly test the hypothesis that anti-BP180NC16A autoantibodies in BP patients are pathogenic and to establish a novel animal model to study immunopathology of BP with patient materials, we created a humanized mouse strain that expresses a hybrid BP180 protein containing the immunodominant human BP180NC16A domain. NC16A+/+ mice developed normally and were fertile. No skin abnormalities at the gross, microscopic and ultrastructural levels were seen up to 12 months of age. PCR and RT-PCR analyses of NC16A+/+ mouse tails detected the hBP180NC16A sequence (Figure 1B and 1C). Indirect IF showed correct expression of hBP180NC16A in basal keratinocytes (Figure 1D). Electron microscopic analysis of the skin revealed no abnormalities in the structure or distribution of hemidesmosome in basal keratinocytes (Figure 1E). The NC16A+/+ mice also reproduced normally with compatible frequency and size of litters as controls. These results showed that the replacement of mBP180NC14A with hBP180NC16A did not impair functions of BP180 protein in the skin.
To test the pathogenicity of anti-human BP180 antibodies, anti-BP180NC16A autoantibodies or rabbit anti-hBP180NC16A IgG (R594) were injected intradermally (i.d.) or intraperitoneally (i.p.) into neonatal NC16A+/+ mice. The injected animals developed skin lesions at 24 h post injection (Figure 2 and Table 1). Direct IF of skin from the diseased mice showed in situ deposition of human IgG (Figure 2b) and murine C3 (Figure 2c) at the basement membrane zone. Histologically, the skin blisters were subepidermal (Figure 2d), and electron microscopic analysis revealed that the lesional site was at the lamina lucida (Figure 2e), exactly the same location as human BP. The diseased skin contained degranulated dermal mast cells (Figure 2f) and a predominant neutrophil infiltrate (Figure 2g), but no eosinophils (Figure 2h). NC16A+/+ mice injected with control human IgG showed no skin lesions (Table 1). Rabbit anti-mBP180NC14A IgG, which is pathogenic in wild-type mice, was no longer pathogenic in the NC16A+/+ mice (Table 1).
Intact pathogenic anti-BP180NC16A IgG autoantibodies activated complement and induced subsequent blistering in the NC16A+/+ mice (Figures 3c and 3d). In contrast, F(ab′)2 fragments of pathogenic anti-BP180NC16A antibodies were able to bind to their target, but failed to activate the complement cascade as evidenced by the lack of C3 deposition at the basement membrane zone (Figure 3e) and induced no skin disease (Figure 3f and Table 1). In addition, the NC16A+/+ mice, pretreated with cobra venom factor to deplete complement, became resistant to the pathogenic effects of anti-BP180NC16A autoantibodies (Table 1). The NC16A+/+ mice injected with pathogenic anti-BP180NC16A autoantibodies showed extensive mast cell degranulation and developed BP (Figures 4c and 4d). In contrast, pretreatment of NC16A+/+ mice with cromolyn, an inhibitor of mast cell activation, inhibited anti-BP180NC16A antibody-induced mast cell degranulation (Figure 4e) and subsequent blistering (Figure 4f). Pathogenic anti-BP180NC16A autoantibodies induced BP in the NC16A+/+ mice pretreated with control rabbit antibody (Figures 5a and 5b). In contrast, anti-BP180NC16A antibodies were no longer pathogenic in neutrophil-deficient NC16A+/+ mice pretreated with neutrophil-depleting rat antibody (Figures 5c and 5d).
MPO activity assays revealed a drastic reduction in neutrophil infiltration in NC16A+/+ mice injected with anti-BP180NC16A F(ab′)2 fragments, anti-BP180NC16A IgG plus cromolyn, or anti-BP180NC16A IgG plus neutrophil-depleting antibody (Figure 6). Taken together, these results demonstrate that subepidermal blistering triggered by pathogenic anti-BP180NC16A antibodies in NC16A+/+ mice requires complement activation, mast cell activation, and neutrophil recruitment.
Since human and murine BP180 antigens are not immune cross-reactive, a humanized mouse strain expressing the BP immunodominant epitopes, BP180NC16A domain, is needed to study the humoral immune response in BP disease immunopathology. In this work, we created such a mouse strain (NC16A+/+) and demonstrated that anti-BP180NC16A autoantibodies in BP are indeed pathogenic. These findings are in agreement with a recent report by Nishie et al , in which anti-BP180 autoantibodies induce BP in mice humanized with whole BP180 instead of only the NC16A domain.
The NC16A humanized mouse model also sheds light on the structure/function of BP180. This transmembrane collagen is a key component of the junctional adhesion complex, or hemidesmosome, and is thought to function in cell-matrix adhesion [3, 48] The homologous human BP180NC16A and murine BP180NC14A domains are not cross-reactive immunologically and exhibit only 59% identity and 71% similarity, compared with the relatively high overall homology exhibited by the human and murine BP180 proteins (81% identity, 86% similarity)  Nevertheless, the humanized NC16A mice are phenotypically normal and develop normal hemidesmosomes at both the microscopic and ultrastuctural levels. Our findings suggest that domain-specific replacement of mBP180NC14A with hBP180NC16A does not impair the expression and function of this protein in vivo, indicating that this model system might well prove useful in future investigations into the biological roles of BP180. In addition, domain-specific humanization could be a useful approach for a human protein that may lead to a dominant negative phenotype when introduced into mice as a whole molecule.
We previously showed that an intact complement cascade, degranulation of resident mast cells, and activation of infiltrating neutrophils are required in the dermal-epidermal separation induced by experimentally generated rabbit anti-mBP180 antibodies [42, 45, 46, 49] While the rabbit anti-mBP180 IgG passive transfer model has provided invaluable insight to the key steps in BP disease development, it has been debated whether the rabbit anti-mBP180 IgG faithfully mirrors effector functions of pathogenic anti-BP180 autoantibodies since rabbit IgG is a single isotype while the BP autoantibodies are a mixture of IgG1, IgG3 and IgG4. Our present study offers us direct evidence that the key innate immune system players, complement and inflammatory cells, also play a critical role in BP autoantibody-mediated subepidermal blistering. Involvement of these same components in the immuno-pathogenic response to BP autoantibodies has also been supported by the results of previous clinical and in vitro studies. Intact and degranulating mast cells, eosinophils, and neutrophils are seen in the dermis, suggesting these cells have been activated locally  In vitro dermal-epidermal junction separation of human skin sections induced by human BP autoantibodies depends on complement and leukocytes (predominantly neutrophils) [50, 51] Human BP blister fluids contain high levels of both neutrophil elastase and gelatinase B, but BP180 degradation by blister fluid depends on neutrophil elastase activity [52, 53] Thus, the present findings provide important information concerning potential therapeutic targets for BP.
While human BP and humanized murine BP closely mimic each other at the clinical, histological, and immunological levels, the IgG passive transfer models do not reflect the large number of eosinophils typically found in the inflammatory infiltrate of human BP lesional skin. Therefore, the question of the role of eosinophils in BP cannot be addressed in our humanized BP180NC16A mouse model.
In conclusion, this study provides direct evidence that anti-BP180NC16A autoantibodies are pathogenic and their tissue injury activity depends on innate immune responses. This novel experimental mouse model will be useful for dissecting the immunopathology of BP, and will aid in the future investigation of some important questions which could not be addressed before. These include the role of different isotypes and IgG subclasses of anti-BP180 autoantibodies since BP180-specific IgA, IgE, IgG1, IgG3, and IgG4 are present in BP [39, 54, 55] This humanized mouse strain may also provide an in vivo system to elucidate the pathogenic mechanisms of other subepidermal blistering diseases with an autoimmune response to BP180, such as herpes gestationis, mucous membrane pemphigoid, linear IgA bullous dermatosis, and lichen planus pemphigoides [18, 56–58]
We thank Dr. Pamela Groben for routine histology, Shiliang Wang and Sarah Rice for their excellent technical assistance, and Eric Bankaitis for assistance of manuscript preparation. We are indebted to Dr. Lowell Goldsmith and Lisa Leighty for their critical reading of the manuscript. This work was supported in part by U.S. Public Health Service NIH grants AI40768 and AI61430 (Z. L.), AR052109 and AR053313 (N.L.), AR32599 and AR32081 (L.A.D.).
Disclosures: The authors have no financial conflict of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.