Introduction

In subepidermal immunobullous diseases, binding of autoantibodies to structural molecules of the cutaneous basement membrane zone results in impaired epidermal–dermal adhesion (Yancey, 2005; Fassihi et al., 2006). Autoantibodies in bullous pemphigoid (BP), the most frequent autoimmune bullous disorder, target the hemidesmosomal components collagen XVII (BP180) and BP230 (Hofmann et al., 2002; Van den Bergh and Giudice, 2003). Other constituents of the cutaneous basement membrane zone, such as laminin 332 (laminin 5), 64-integrin, and collagen VII, the major component of anchoring fibrils, represent autoantigens of mucous membrane pemphigoid and epidermolysis bullosa acquisita (Hertl, 2000; McMillan et al., 2003). In 1996, anti-p200 pemphigoid, a novel subepidermal immunobullous disorder with circulating IgG₄-antibodies directed against a 200 kDa protein of the dermis was first described (Chen et al., 1996; Zillikens et al., 1996). The clinical spectrum can be extremely variable with regard to morphology and extent (Mascaro et al., 2000; Watanabe et al., 2002; Dilling et al., 2007), but 2D-gel electrophoresis of dermal extracts demonstrated that multiple anti-p200 sera recognized an identical 200 kDa protein, named p200 (Shimanovich et al., 2003).

Ultrastructural studies localized p200 to the lamina lucida–lamina densa interface (Egan et al., 2002) and biochemical characterization of dermal extracts demonstrated that p200 is an acidic, non-collagenous, N-glycosylated protein (Shimanovich et al., 2003). Based on these features and on immunofluorescence (IF) and immunoblot studies, it was predicted that the target is distinct from all major autoantigens of the dermal–epidermal junction (Kawahara et al., 2000; Mascaro et al., 2000; Zillikens et al., 2000; Liu et al., 2003). However, low concentration and poor solubility of p200 have hampered molecular identification with a variety of methods, including N-terminal sequencing and mass spectrometry (Zillikens et al., 1996; Shimanovich et al., 2003). Therefore, the cellular origin and the identity of p200 have yet to be elucidated.

An alternative approach is to test candidate molecules with corresponding biochemical and localization characteristics for reactivity with anti-p200 pemphigoid sera. One such candidate is nidogen-2, a 200 kDa basement membrane protein that – via interaction with collagen I, collagen IV, and perlecan – is implicated in promoting cell adhesion, spreading and stabilization of basement membranes (Kohfeldt et al., 1998). However, p200 might also represent a protein complex formed by stable association of different extracellular matrix proteins, such as the shed collagen XVII-ectodomain and laminin 332-fragments (Tasanen et al., 2004).

In this study, we report three new cases of anti-p200 pemphigoid, which prompted us to investigate the nature of p200 in detergent extracts of cultured epidermal cells and fibroblasts and by IF on normal and genetically altered skin. In addition, we assessed whether the 200 kDa basement membrane constituent nidogen-2 shares identity with the autoantigen of anti-p200 pemphigoid.

Results

Patients P1, P2, and P3 display IF and western blot findings indicating the diagnosis of anti-p200 pemphigoid

Tissue-bound autoantibodies were detected at the basement membrane zone by direct IF in all three patients (Figure 1d). P1 and P3 revealed circulating IgG-autoantibodies binding to the dermal side of the blister by indirect IF on salt-split-skin, serum of P2 was unreactive by indirect IF. Although all patients displayed circulating autoantibodies binding exclusively to a 200 kDa protein in keratinocyte extracts, only serum of P1 labeled p200 in immunoblots of dermal extracts (kindly performed by Drs D. Zillikens and I. Shimanovich, Department of Dermatology, University of Lübeck, Germany).

Figure 1.

Patient P1 displays clinical, histological, and IF findings compatible with anti-p200 pemphigoid. (a and b) Clinically, the patient presented with disseminated small blisters and erosions. (c) Hematoxylin-eosin-stained histology demonstrated subepidermal blistering with a mixed inflammatory infiltrate (bar=100 m) and (d) direct IF using a FITC-labeled antihuman C3 antibody (DAKO) showed linear C3 deposits at the dermal–epidermal junction (bar=50 m).

Full figure and legend (163K)

Circulating autoantibodies in anti-p200 pemphigoid recognize a distinct 200 kDa protein from cultured keratinocytes and fibroblasts

In order to improve the solubility and integrity of basement membrane proteins, we extracted primary and immortalized cultured epidermal cells and fibroblasts with a buffer containing NP-40 as a detergent, EDTA to dissociate Ca⁺⁺-dependent protein–protein interactions, and a broad spectrum of proteinase inhibitors. When the cells were extracted without EDTA, anti-p200 pemphigoid sera showed – in contrast to BP sera – no reactivity in immunoblots. However, in immunoblots with EDTA containing extracts, the sera reacted strongly and consistently with a protein of about 200 kDa derived from primary human keratinocytes (Figure 2a). To confirm the specific reactivity of anti-p200 sera with epidermal cell extracts, we additionally used HaCaT keratinocytes and SCC-25 cells for western blotting (Figure 2b).

Figure 2.

P200 is expressed by keratinocytes and fibroblasts in vitro. (a) The extractability of p200 is dependent on the presence of EDTA in contrast to collagen XVII. Extracts of normal human keratinocytes (KE) were prepared in the presence (+) or absence (-) of EDTA, subjected to 7% SDS-PAGE under reducing conditions and immunoblotted with anti-p200 pemphigoid serum and a well-characterized BP serum containing autoantibodies to 180 kDa-collagen XVII. Migration positions of molecular weight marker (kDa) are shown on the left. (b) p200 is expressed by primary and immortalized keratinocytes. Extracts from normal human keratinocytes (KE), HaCaT keratinocytes (HE), and SCC-25 cells (SE) were assessed for the presence of p200 and the BP-autoantigen collagen XVII. (c) p200 is present in extracts of fibroblasts. Extracts of keratinocytes (KE) and fibroblasts (FE), and conditioned medium of keratinocytes (KM) and fibroblasts (FM) were reacted with anti-p200 sera P1–P3 and a BP serum. Sera of patients P1–P3 recognized p200 in extracts of keratinocytes and fibroblasts (black arrow) and a 170 kDa-band in fibroblast medium (white arrow). In fibroblast medium, some additional weak bands are also visible, the identity and specificity of which remains unknown at present. As expected, the BP serum reacted with 180 kDa-collagen XVII (black arrowhead) and its shed 120 kDa ectodomain (white arrowhead), but did not display immunoreactivity with fibroblast extracts or medium. Additional controls with normal human sera did not show reactivity with keratinocytes or fibroblasts (not shown).

Full figure and legend (63K)

The three new anti-p200 pemphigoid sera recognized p200 not only in keratinocyte extracts but also in extracts of cultured fibroblasts (Figure 2c). To assess secretion of p200 in cell cultures, we processed extracts and media separately for immunoblotting. Although anti-p200 sera did not detect proteins in the medium of epithelial cells, a band of about 170 kDa was labeled in the culture medium of fibroblasts by patients P1, P2, and P3. BP sera, which were used as controls, recognized collagen XVII in extracts of primary keratinocytes and its shed soluble ectodomain in the medium. They were devoid of reactivity with fibroblast extracts or medium (Figure 2c). Normal control sera did not show any specific reactivity.

Interestingly, p200 could only be visualized in immunoblots, if the SDS-PAGE was run under reducing conditions. If dithiothreitol as a reducing agent was omitted, p200 was not detectable.

Immunoblot analysis of human keratinocyte extracts with an SDS-PAGE 4.5–15% gradient gel with anti-p200 sera revealed exclusively a protein band that migrated between BP230 (230 kDa) and collagen XVII (180 kDa), corresponding to an apparent molecular weight of 200 kDa. As shown in Figure 3, this position is clearly distinct from collagen VII, BP230, collagen XVII, and laminin 332. Normal sera showed no specific reactivity.

Figure 3.

Keratinocyte-derived p200 is distinct from the major basement membrane proteins. Keratinocyte extracts were separated on a 4.5–15% gradient gel and transferred to nitrocellulose, which was subsequently incubated with NC1F3 antibodies to collagen VII (290 kDa, lane 1), BP serum containing autoantibodies to BP230 (230 kDa, lane 2) or collagen XVII (180 kDa, lane 3), the 9LN5 antibody to the 3-chain of laminin 332 (165 kDa, lane 4), serum of P1 (lane 5), and a normal human serum (lane 6). The antibodies were visualized by alkaline phosphatase-conjugated anti-rabbit or anti-human IgG. The migration positions of molecular weight markers (kDa) are indicated on the left.

Full figure and legend (28K)

The biochemical properties of keratinocyte-derived p200 are identical with those of dermal p200

As p200 has been characterized so far only in dermal extracts (Shimanovich et al., 2003), we compared the biochemical characteristics of cell culture-derived p200 to those reported for dermis-derived p200 by limited enzyme digestions (Figure 4). As delineated above, only EDTA-containing cell extracts contained the protein of interest. Therefore, before digestion with highly purified bacterial collagenase, a calcium-dependent protease, EDTA was removed from the cell extracts by ethanol precipitation. Similarly to the findings of Shimanovich et al. (2003) on p200 in dermal extracts, p200 from keratinocyte extracts resisted collagenase treatment. Collagen VII, which was used as a positive control, was cleaved to yield the non-collagenous NC1-domain, whereas collagenase digestion of collagen XVII has been previously shown by our laboratory to result in a fragment of about 65 kDa corresponding to the endodomain (Schäcke et al., 1998). Trypsin digestion of keratinocyte extracts completely abolished reactivity with anti-p200 sera. As a positive control, collagen XVII was used, which was cleaved to yield the trypsin resistant triple-helical fragment of about 90 kDa (Franzke et al., 2004). Deglycosylation with N-glycosidase F resulted in a small increase in the electrophoretic mobility of p200, similarly to collagen XVII, indicating the presence of N-linked sugar moieties in both proteins. Comparison of the relative electrophoretic mobility before and after N-glycosidase F digestion suggested that the molecular mass of deglycosylated p200 was about 5,000 Da smaller than that of the cell-derived protein.

Figure 4.

Keratinocyte-derived p200 is an N-glycosylated non-collagenous protein. (a) Collagenase digestion. The untreated (lane 1 and 3) and digested keratinocyte extract (lane 2 and 4) was immunostained with anti-p200 serum P1 (lane 1 and 2) and the NC1F3-antibody to collagen VII (lane 3 and 4). Although p200 was collagenase resistant, collagen VII was cleaved to yield the 145 kDa non-collagenous NC1-domain. Collagenase digestion of collagen XVII has been previously published by us and yielded a fragment of about 65 kDa corresponding to the endodomain (Schäcke et al., 1998). (b and c) Trypsin and N-glycosidase F digestion. Immunoreacitivity of anti-p200 serum P1 (lane 1, before digestion; lane 2, after digestion) was compared with a BP-serum containing autoantibodies against collagen XVII (lane 5, before digestion; lane 6, after digestion). Trypsin digestion of keratinocyte extracts completely abolished reactivity with anti-p200 sera, in contrast to collagen XVII, which was cleaved to the trypsin resistant triple-helical fragment of about 90 kDa (Franzke et al., 2004). Deglycosylation with N-glycosidase F resulted in a small increase in the electrophoretic mobility of p200, similarly to collagen XVII.

Full figure and legend (42K)

P200 is not destabilized by lack of collagen XVII, collagen VII, or laminin 332 from the dermal–epidermal junction

By indirect IF on 1 M salt-split-skin, circulating anti-p200 autoantibodies localized to the dermal side of the split at a titer of 1:20, in contrast to BP autoantibodies, which bound to the epidermal side of the blister (Figure 5a and e). A genetic approach was taken for further characterization using skin from epidermolysis bullosa (EB) patients lacking collagen VII, collagen XVII, or laminin 332. In all cases, anti-p200 pemphigoid serum generated a positive signal. In junctional EB (JEB) skin, which has a lamina lucida split, p200 was localized at the dermal side of the blister. Conversely, in dystrophic EB with a sub-lamina densa split, p200 was on the epidermal side, indicating association with the lamina densa of the basement membrane (Figure 5b–d and f–h).

Figure 5.

p200 is distinct from collagen XVII, collagen VII, or laminin 332 and colocalizes with nidogen-2. (a, e, i) For indirect IF, a representative BP serum with autoantibodies to BP230 and collagen XVII, anti-p200 pemphigoid serum P1 and polyclonal nidogen-2 antibodies were incubated with sections of normal human salt-split-skin (SSS), (b, f, j, m) lesional skin of a patient with dystrophic EB Hallopeau–Siemens, (c, g, k, n) lesional skin of a patient with Herlitz JEB, (d, h, l, o) and lesional skin of a patient with non-Herlitz JEB. (m) The absence of collagen VII, (n) laminin 332, and (o) collagen XVII from the skin was verified by antibodies LH 7.2 to the NC1-domain of collagen VII, BM165 to human laminin 3 chain, and ColXVII-NC16a to human collagen XVII, respectively (controls). Immunoreactivity was visualized with FITC-conjugated antihuman IgG₄, anti-rabbit IgG or anti-mouse IgG. BP autoantibodies bind to the epidermal side of the split in all specimens. Anti-p200 serum and nidogen-2 antibodies label the dermal side of the blister in JEB and the epidermal side in dystrophic EB (bar=50 m).

Full figure and legend (142K)

Despite colocalization p200 is distinct from nidogen-2

As nidogen-2 is a known 200 kDa component of the lamina densa of the cutaneous basement membrane, we assessed the precise location of p200 in relation to nidogen-2. As shown in Figure 5, nidogen-2 colocalizes with p200 within the dermal–epidermal junction zone. However, in addition to the junctional staining, nidogen-2 antibody exhibited reactivity with dermal blood vessels. Unfortunately, sera P1, P2, and P3 did not crossreact with normal murine skin, hindering a genetic approach using nidogen-1 or nidogen-2 knockout mouse skin. Electrophoretic co-migration of p200 and nidogen-2 was demonstrated in immunoblots with human fibroblast extracts as antigen (Figure 6a). However, none of the three anti-p200 sera (P1–P3) showed reactivity with recombinant human nidogen-2, indicating that it is distinct from p200 (Figure 6b).

Figure 6.

Nidogen-2 is distinct from p200. (a) Fibroblast extracts (FE) were separated by 7% SDS-PAGE and assayed for immunoreactivity with anti-p200 serum and polyclonal nidogen-2 antibodies (Nd-2). Both antibodies recognize a protein of identical size. (b) To verify whether nidogen-2 is the candidate autoantigen of anti-p200 pemphigoid, 5 g of recombinant human nidogen-2 were immunoblotted with anti-p200 sera. Comparison of the reactivities of anti-p200 sera and nidogen-2 antibodies shows that p200 is distinct from nidogen-2. The migration positions of molecular weight markers (kDa) are indicated on the left.

Full figure and legend (35K)

Discussion

In this study, the diagnosis of anti-p200 pemphigoid in three new patients was based on typical clinical, histological, and IF findings, and on specific reactivity with a 200 kDa-protein in immunoblots (Zillikens et al., 1996). All previous investigations have shown that anti-p200 sera react with dermal extracts. In particular, p200 was usually not detected in cell extracts or conditioned medium of cultured keratinocytes and fibroblasts (Shimanovich et al., 2003). An epidermal cell origin of p200, at least in vitro, is only supported by one report showing immunoreactivity of anti-p200 pemphigoid serum with a 200 kDa-protein in the medium of SCC-25 cells (Zillikens et al., 1996).

The contradiction with the present findings prompted us to investigate the immunoreactive 200 kDa protein in keratinocyte extracts in more detail and to compare it with p200 reported in the literature. First, we analyzed immortalized keratinocyte cell lines (HaCaT keratinocytes and SCC-25 cells) in addition to primary human keratinocytes to confirm the expression of p200 by keratinocytes. Second, thorough investigation of the biochemical properties of keratinocyte-derived p200 using limited proteolysis with collagenase and trypsin, and removal of N-linked sugar moieties showed that, similarly to dermal p200 (Shimanovich et al., 2003), keratinocyte-derived p200 is an N-glycosylated protein devoid of collagenous domains. Thus, it can be assumed that p200 from either source is identical.

The detection of p200 in cell extracts may largely depend on the extraction protocol applied. In contrast to previous reports, we omitted pulse-sonication, but used NP40 as a detergent, and a broad spectrum of protease inhibitors to prevent degradation more efficiently (Schäcke et al., 1998; Shimanovich et al., 2003). Our finding that efficient extraction of p200 depends on the presence of EDTA suggests that p200 forms Ca⁺⁺-dependent protein–protein complexes, similar to the laminin-nidogen complex (Paulsson et al., 1987; Blauvelt et al., 1995), or forms complexes with membrane associated proteins, similar to cadherins (Amagai et al., 1995).

In previous studies, successful extraction of p200 from the dermis was dependent on high concentrations of reducing agents (Shimanovich et al., 2003). Here, we show that reducing agents are not required for isolation of p200 from cell cultures. However, comparative immunoblot analysis of cell extracts under reducing and non-reducing conditions demonstrated that p200 is only recognized by anti-p200 sera under reducing conditions. This is in line with the hypothesis of Shimanovich et al. (2003) that p200 forms large disulfide-bonded aggregates either as a multimeric protein or together with other components of the basement membrane zone. The disulfide bonds in p200 are either intramolecular, or intermolecular with other proteins extractable with the present protocol, as p200 can be extracted from the cell layer in the absence of reducing agents. It is conceivable that some anti-p200 sera may contain autoantibodies against labile epitopes destroyed or modified by the harsh extraction methods used previously. This could also explain why two of the new anti-p200 pemphigoid patients were devoid of reactivity with p200 from dermal extracts.

Several components of the cutaneous basement membrane zone, including collagen IV and collagen VII, laminin 111 (laminin 1), and fibronectin, are synthesized by keratinocytes and fibroblasts (O'Keefe et al., 1985; Lee and Cho, 2005). Therefore, we determined whether anti-p200 sera displayed immunoreactivity with extracts and media of both cell types. Our data demonstrate that p200 is produced in similar amounts by both, keratinocytes and fibroblasts, in vitro. Furthermore, a protein of approximately 170 kDa was detected by all three sera in conditioned medium of fibroblasts, possibly indicating that p200 is secreted and cleaved in the culture medium. The lack of immunoreactive protein in keratinocyte medium may be due to formation of calcium-dependent insoluble aggregates, as the presence of proteolytic enzymes and potential binding partners varies between keratinocytes and fibroblasts. In general, the in vitro-secretion of p200 would be in accordance with electron microscopy showing that p200 is an extracellular protein (Chen et al., 1996). However, we are aware that the 170-kDa bands detected in fibroblast medium may also represent nonspecific reactivity of polyclonal human sera.

The finding, that p200 is a disulfide-bonded protein, prompted us to assess the relation of p200 with the shed collagen XVII-ectodomain, which extends into the lamina densa (Nonaka et al., 2000), engages in ligand interactions with laminin 332 and could therefore be part of an antigenic protein complex. Along with collagen VII- and laminin 332-deficient skin (Zillikens et al., 2000; Liu et al., 2003), we used skin of a patient with non-Herlitz JEB and defined COL17A1 mutations resulting in the complete loss of collagen XVII, for IF staining with anti-p200 serum. The signals generated by anti-p200 serum indicated that p200 is distinct and functionally independent from collagen XVII, collagen VII, and laminin 332.

In addition to these well-characterized components of the dermal–epidermal junction, nidogen-1 and nidogen-2 are ubiquitous basement membrane proteins involved in stabilization of basement membranes (Salmivirta et al., 2002; Bader et al., 2005). We considered nidogen-2 as a candidate target in anti-p200 pemphigoid, taking into account its predicted molecular weight of 200 kDa, often accompanied by a cleavage product of about 170 kDa (Kohfeldt et al., 1998). Nidogen-2 is physiologically produced by fibroblasts, but synthesis of nidogen-2 can also be observed in keratinocyte cultures under certain conditions (El Ghalbzouri et al., 2005). Immunomapping on normal and EB skin revealed that nidogen-2 indeed colocalized with p200 within the dermal–epidermal junction. However, in contrast to nidogen-2 antibodies, anti-p200 sera do not stain dermal vessel walls (Zillikens et al., 1996). As this discrepancy could have been explained by localization-dependent epitope variation, possibly due to post-translational modifications, we addressed the question whether p200 is identical to nidogen-2 with protein chemical analysis. Sera P1, P2, and P3 did not exhibit reactivity with purified recombinant human nidogen-2, suggesting that the autoantigen of anti-p200 pemphigoid is distinct from nidogen-2. This is in line with the different IF patterns generated with nidogen-2 antibodies and anti-p200 sera.

The molecular identity of p200 remains elusive. IF studies and immunoblotting experiments using purified recombinant or native proteins as well as biochemical analyses of the p200-protein have been employed to rule out a number of basement membrane proteins as candidate autoantigens of anti-p200 pemphigoid (Table 1). Other basement membrane proteins with an approximate molecular weight of 200 kDa comprise 6-integrin, the laminin 1-chain, and the laminin 2-chain. The first mentioned proteins have been previously excluded as autoantigens of anti-p200 pemphigoid due to their epidermal localization or the lack of reactivity of anti-p200 sera with recombinant laminin 111, respectively. In contrast, the laminin 2-chain (190 kDa) is part of laminin 321 (laminin 7) and laminin 521 (laminin 11), both of which are expressed in the upper lamina densa of the epidermal basement membrane, but also in dermal vessels (McMillan et al., 2006). Future studies should address the question, whether the laminin 2-chain is identical to p200.

Table 1 - Candidate autoantigens of anti-p200 pemphigoid.

Full table

In conclusion, this study identifies p200 as an extracellular matrix protein synthesized by keratinocytes and fibroblasts, which is distinct from major basement membrane components, including collagen XVII and nidogen-2. P200 is engaged in calcium-dependent complex formation and can be extracted from cell cultures with detergents and metal chelators. The present protocol for extraction of p200 from cultured skin cells will on one hand enlarge the methods for diagnosing anti-p200 pemphigoid and, on the other hand, also provide helpful knowledge for further identification and functional characterization of the elusive autoantigen in this bullous disorder.

Materials and Methods

Case reports

Patient 1: a 61-year-old male (P1), presented with a 6-week history of a pruriginous vesiculobullous eruption. Clinical examination revealed multiple small, tense blisters, and erosions on the lower back and dorsal aspects of the arms (Figure 1a and b). Large, hemorrhagic bullae at the palms and soles and extensive genital erosions were noted. Histology revealed subepidermal blisters with neutrophilic and eosinophilic infiltrates, and direct IF showed linear IgG and C3 deposits at the dermal–epidermal junction (Figure 1c and d). Circulating autoantibodies binding to the dermal side of 1 M salt-split-skin were detected by indirect IF (titer 1:20) and to a 200-kDa protein by western blot analysis of dermal extracts. The serum was devoid of reactivity with the NC16a-domain of collagen XVII/BP180 by ELISA (Hofmann et al., 2002). These characteristic features confirmed the diagnosis of anti-p200 pemphigoid. Treatment with prednisolone and dapsone led to an improvement of skin lesions. Interestingly, soon after the diagnosis of anti-p200 pemphigoid, the patient developed severe and refractory acquired hemophilia, which was treated by high-dose corticosteroids, protein A-immunoadsorption and rituximab in conventional doses. This therapeutic regimen induced clinical and serological remission of anti-p200 pemphigoid.

Patient 2: a 45-year-old man (P2) was referred to our hospital with a 1-year-history of recurrent pruriginous tense blisters on the right lower leg, which healed without scarring. Otherwise, the dermatological status was unremarkable. Histology showed subepidermal blisters and a predominant neutrophilic infiltrate. Direct IF demonstrated linear C3-deposits at the dermal side after splitting with 1 M NaCl, whereas indirect IF and NC16a-ELISA were negative. The patient's serum recognized a 200-kDa protein by immunoblotting with keratinocyte extracts. The disease remained stable over the past 6 months with topical corticosteroid treatment.

Patient 3: in a young boy (P3) with recurrent blister formation, direct and indirect IF and western blot with keratinocyte extracts suggested the diagnosis of anti-p200 pemphigoid. The clinical features and precise diagnostic workup of this patient will be discussed elsewhere (Trüeb et al., in preparation).

Cell extracts and conditioned medium

Primary human keratinocytes were isolated from skin biopsies and grown in serum-free keratinocyte growth medium supplemented with bovine pituitary extract and epidermal growth factor as described (Schäcke et al., 1998). HaCaT keratinocytes (kindly provided by NE Fusenig, German Cancer Research Center, Heidelberg, Germany) and SCC-25 cells, derived from a well-differentiated tongue squamous cell carcinoma, were grown as described previously (Boukamp et al., 1988; Marinkovich et al., 1992). Fibroblasts were maintained in DMEM supplemented with L-glutamine, 10% fetal calf serum and penicillin/streptomycin (Gibco Invitrogen, Grand Island, NY). Before extraction, 50 g/ml L-ascorbate was added for 48 hours.

For protein extractions, cell layer and media were processed separately as reported (Schäcke et al., 1998). Confluent cell layers were extracted for 30 minutes on ice with 1 ml/75 cm² of a buffer containing 1% NP-40, 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5), 1 mM Pefabloc (Merck, Darmstadt, Germany), and, when appropriate, 4 mM EDTA, 14 g/ml chymostatin, 7 g/ml antipain, 7 g/ml leupeptin, and 14 g/ml pepstatin. The cell lysate was then scraped off with a rubber policeman, centrifuged at 14,000 g for 30 minutes at 4°C and stored at -80°C until use.

After a 48-hour incubation in serum-free medium, the culture supernatant was collected on ice and 1 mM Pefabloc and 4 mM EDTA were added immediately before centrifugation at 13,000 g for 10 minutes at 4°C. Proteins were precipitated with chloroform-methanol, centrifuged at 15,000 g for 10 minutes at 4°C and pellets were dissolved in SDS-PAGE buffer. Protein concentrations were determined using a modified Lowry protocol (DC Protein Assay, Bio-Rad, Munich, Germany) and 30 g/ml of total protein were loaded onto SDS-PAGE.

Proteolytic digests

For collagenase digestion, keratinocyte extracts were precipitated with 70% ethanol and resuspended in collagenase buffer (0.05 M Tris-HCl, 0.65 M NaCl, pH 7.4), 15 mM CaCl₂, 1 mM Pefabloc (Merck, Darmstadt, Germany), and 1.5 mM N-Ethylmaleimide. 1 U of chromatographically purified collagenase form III from Clostridium histolyticum (WAK-Chemie Medical GmbH, Bad Soden, Germany) was added to 35 g of protein. After incubation at 37°C for 4 hours, the reaction was stopped by adding 8 mM EDTA. For trypsin digestion, cell lysates were incubated for 1 minute at 40°C and snap cooled on ice. 1 U of trypsin (Sigma, Saint Louis, MO) was added to 35 g of cellular proteins and digested for 2 minutes at room temperature (Olague-Marchan et al., 2000). The reaction was stopped by adding soybean trypsin inhibitor (Sigma) to a final concentration of 10 g/ml.

For deglycosylation, 35 g of protein extracts were treated with 5% SDS and 0.4 M dithiotreitol for 10 minutes at 60°C before digestion with 5 U PNGase F (New England Biolabs, Ipswich, MA) for 1 hour at 37°C. Undigested and digested samples were subjected to 10% SDS-PAGE and after transfer on nitrocellulose incubated with anti-p200 sera, BP sera or NC1F3 antibodies directed against the NC1-domain of collagen VII (Mecklenbeck et al., 2002).

Immunoblot analyses

As antigens, extracts of primary and HaCaT keratinocytes, SCC cell lines and fibroblasts, and recombinant human nidogen-2 were employed. When necessary, the protein extracts were concentrated by precipitation with 70% ethanol at -20°C overnight before separation by SDS-PAGE (7% or 4.5–15% gradient gels) under non-reducing or reducing conditions. Recombinant human nidogen-2 was produced in human EBNA-293 cells and purified as described (Kohfeldt et al., 1998). For immunoblotting, 0.25–5 g of recombinant protein were separated on a 7% SDS-PAGE gel under reducing conditions. Human sera were diluted 1:20–1:100 and the anti-nidogen-2 antiserum 1:2,000 (Kohfeldt et al., 1998). As controls, antibody NC1F3 to collagen VII at a dilution of 1:1,000 (Mecklenbeck et al., 2002) and antibody 9LN5 to the 3-chain of laminin 332 at a dilution of 1:2,000 were used. Incubations with primary antibodies were overnight at 4°C and with alkaline phosphatase-labeled-anti-human or anti-rabbit IgG (Sigma) diluted 1:30,000 and 1:10,000, respectively, were for 2 hours. Immunoreactivity was visualized using NBT/BCIP (Roth, Karlsruhe, Germany) as a chromophore.

Indirect IF

Immunomapping of the dermal–epidermal junction was performed on 8 m cryosections of salt-split normal human skin, skin of EB patients with defined mutations and skin of wild type and nidogen knockout mice. Anti-p200 sera, a BP-serum reactive with collagen XVII and BP230, and normal human serum were used in a 1:20 dilution. As controls, the following antibodies were employed: anti-nidogen-2 antibodies (diluted 1:100) (Kohfeldt et al., 1998), ColXVII-NC16a to human collagen XVII (Schumann et al., 2000), BM165 to human laminin 3 chain (a kind gift from Dr P. Rousselle, Lyon, France (Rousselle and Aumailley, 1994)), and 6F12 to human laminin 3 chain (a kind gift from Dr R.E. Burgeson, Cutaneous Biology Research Center, Harvard Medical School, Cambridge, MA, USA), GB3 to laminin 2 chain (a generous gift of Dr G. Meneguzzi, Nice, France (Meneguzzi et al., 1992)), LH 7.2 to the NC1-domain of collagen VII (Sigma) and antibodies raised against the triple helical domain of collagen VII (Hammami-Hauasli et al., 1997). Immunoreactivity was detected using FITC-conjugated monoclonal antibodies to human IgG₄ at a dilution of 1:60 (clone HP-6025, Sigma), FITC-conjugated anti-rabbit IgG diluted 1:30 or FITC-conjugated anti-mouse IgG diluted 1:40 (both DAKO, Glostrup, Denmark).

Skin specimens with deficiency of a defined protein were obtained from patients with non-Herlitz JEB with a homozygous deletion 2560–2563delAATT in the COL17A1 gene (Vaisanen et al., 2005), Herlitz JEB with a homozygous R42X-mutation in the LAMB3 gene (Kivirikko et al., 1996), dystrophic EB Hallopeau–Siemens with compound heterozygous mutations c.425A>G and c.5261dupC in the COL7A1 gene (Gardella et al., 1996; Kern et al., 2006), or from knockout mice with targeted ablation of nidogen-1 or nidogen-2 (Murshed et al., 2000; Schymeinsky et al., 2002). The complete absence of the mutated proteins was verified using the antibodies listed above. This study was approved by the medical ethical committee of the University Medical Center Freiburg, Germany. Written informed consent was obtained from all patients whose material was used in this study, in adherence to the Declaration of Helsinki Principles.

Conflict of Interest

The authors state no conflict of interest.

References

1. Amagai M, Ishii K, Hashimoto T, Gamou S, Shimizu N, Nishikawa T (1995) Conformational epitopes of pemphigus antigens (Dsg1 and Dsg3) are calcium dependent and glycosylation independent. J Invest Dermatol 105:243–247 | Article | PubMed | ISI | ChemPort |
2. Bader BL, Smyth N, Nedbal S, Miosge N, Baranowsky A, Mokkapati S et al. (2005) Compound genetic ablation of nidogen 1 and 2 causes basement membrane defects and perinatal lethality in mice. Mol Cell Biol 25:6846–6856 | Article | PubMed | ISI | ChemPort |
3. Blauvelt A, Katz SI, Udey MC (1995) Human Langerhans cells express E-cadherin. J Invest Dermatol 104:293–296 | Article | PubMed | ISI | ChemPort |
4. Boukamp P, Petrussevska RT, Breitkreutz D, Hornung J, Markham A, Fusenig NE (1988) Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line. J Cell Biol 106:761–771 | Article | PubMed | ISI | ChemPort |
5. Chen KR, Shimizu S, Miyakawa S, Ishiko A, Shimizu H, Hashimoto T (1996) 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 134:340–346 | Article | PubMed | ISI | ChemPort |
6. Dilling A, Rose C, Hashimoto T, Zillikens D, Shimanovich I (2007) Anti-p200 pemphigoid: a novel autoimmune subepidermal blistering disease. J Dermatol 34:1–8 | Article | PubMed | ISI |
7. Egan CA, Yee C, Zillikens D, Yancey KB (2002) Anti-p200 pemphigoid: diagnosis and treatment of a case presenting as an inflammatory subepidermal blistering disease. J Am Acad Dermatol 46:786–789 | Article | PubMed | ISI |
8. El Ghalbzouri A, Jonkman MF, Dijkman R, Ponec M (2005) Basement membrane reconstruction in human skin equivalents is regulated by fibroblasts and/or exogenously activated keratinocytes. J Invest Dermatol 124:79–86 | Article | PubMed | ISI | ChemPort |
9. Fassihi H, Wong T, Wessagowit V, McGrath JA, Mellerio JE (2006) Target proteins in inherited and acquired blistering skin disorders. Clin Exp Dermatol 31:252–259 | Article | PubMed | ISI | ChemPort |
10. Franzke CW, Tasanen K, Borradori L, Huotari V, Bruckner-Tuderman L (2004) Shedding of collagen XVII/BP180: structural motifs influence cleavage from cell surface. J Biol Chem 279:24521–24529 | Article | PubMed | ISI | ChemPort |
11. Gardella R, Belletti L, Zoppi N, Marini D, Barlati S, Colombi M (1996) Identification of two splicing mutations in the collagen type VII gene (COL7A1) of a patient affected by the localisata variant of recessive dystrophic epidermolysis bullosa. Am J Hum Genet 59:292–300 | PubMed | ISI | ChemPort |
12. Hammami-Hauasli N, Kalinke DU, Schumann H, Kalinke U, Pontz BF, Anton-Lamprecht I et al. (1997) A combination of a common splice site mutation and a frameshift mutation in the COL7A1 gene: absence of functional collagen VII in keratinocytes and skin. J Invest Dermatol 109:384–389 | Article | PubMed | ChemPort |
13. Hertl M (2000) Humoral and cellular autoimmunity in autoimmune bullous skin disorders. Int Arch Allergy Immunol 122:91–100 | Article | PubMed | ISI | ChemPort |
14. Hofmann SC, Thoma-Uszynski S, Hunziker T, Bernard P, Koebnick C, Stauber A et al. (2002) Severity and phenotype of bullous pemphigoid relate to autoantibody profile against the NH2- and COOH-terminal regions of the BP180 ectodomain. J Invest Dermatol 119:1065–1073 | Article | PubMed | ISI | ChemPort |
15. Kawahara Y, Zillikens D, Yancey KB, Marinkovich MP, Nie Z, Hashimoto T et al. (2000) Subepidermal blistering disease with autoantibodies against a novel dermal 200-kDa antigen. J Dermatol Sci 23:93–102 | Article | PubMed | ISI | ChemPort |
16. Kern JS, Kohlhase J, Bruckner-Tuderman L, Has C (2006) Expanding the COL7A1 mutation database: novel and recurrent mutations and unusual genotype-phenotype constellations in 41 patients with dystrophic epidermolysis bullosa. J Invest Dermatol 126:1006–1012 | Article | PubMed | ISI | ChemPort |
17. Kivirikko S, McGrath JA, Pulkkinen L, Uitto J, Christiano AM (1996) Mutational hotspots in the LAMB3 gene in the lethal (Herlitz) type of junctional epidermolysis bullosa. Hum Mol Genet 5:231–237 | Article | PubMed | ISI | ChemPort |
18. Kohfeldt E, Sasaki T, Gohring W, Timpl R (1998) Nidogen-2: a new basement membrane protein with diverse binding properties. J Mol Biol 282:99–109 | Article | PubMed | ISI | ChemPort |
19. Lee DY, Cho KH (2005) The effects of epidermal keratinocytes and dermal fibroblasts on the formation of cutaneous basement membrane in three-dimensional culture systems. Arch Dermatol Res 296:296–302 | Article | PubMed | ISI |
20. Liu Y, Shimizu H, Hashimoto T (2003) Immunofluorescence studies using skin sections of recessive dystrophic epidermolysis bullosa patients indicated that the antigen of anti-p200 pemphigoid is not a fragment of type VII collagen. J Dermatol Sci 32:125–129 | Article | PubMed | ISI | ChemPort |
21. Marinkovich MP, Lunstrum GP, Keene DR, Burgeson RE (1992) The dermal–epidermal junction of human skin contains a novel laminin variant. J Cell Biol 119:695–703 | Article | PubMed | ISI | ChemPort |
22. Mascaro JM Jr, Zillikens D, Giudice GJ, Caux F, Fleming MG, Katz HM et al. (2000) A subepidermal bullous eruption associated with IgG autoantibodies to a 200 kd dermal antigen: the first case report from the United States. J Am Acad Dermatol 42:309–315 | Article | PubMed | ISI |
23. McMillan JR, Akiyama M, Nakamura H, Shimizu H (2006) Colocalization of multiple laminin isoforms predominantly beneath hemidesmosomes in the upper lamina densa of the epidermal basement membrane. J Histochem Cytochem 54:109–118 | Article | PubMed | ISI | ChemPort |
24. McMillan JR, Akiyama M, Shimizu H (2003) Epidermal basement membrane zone components: ultrastructural distribution and molecular interactions. J Dermatol Sci 31:169–177 | Article | PubMed | ISI | ChemPort |
25. Mecklenbeck S, Compton SH, Mejia JE, Cervini R, Hovnanian A, Bruckner-Tuderman L et al. (2002) A microinjected COL7A1-PAC vector restores synthesis of intact procollagen VII in a dystrophic epidermolysis bullosa keratinocyte cell line. Hum Gene Ther 13:1655–1662 | Article | PubMed | ISI | ChemPort |
26. Meneguzzi G, Marinkovich MP, Aberdam D, Pisani A, Burgeson R, Ortonne JP (1992) Kalinin is abnormally expressed in epithelial basement membranes of Herlitz's junctional epidermolysis bullosa patients. Exp Dermatol 1:221–229 | Article | PubMed | ChemPort |
27. Morris SD, Mallipeddi R, Oyama N, Gratian MJ, Harman KE, Bhogal BS et al (2002) Psoriasis bullosa acquisita. Clin Exp Dermatol 27:665–669 | Article | PubMed | ISI | ChemPort |
28. Murshed M, Smyth N, Miosge N, Karolat J, Krieg T, Paulsson M et al. (2000) The absence of nidogen 1 does not affect murine basement membrane formation. Mol Cell Biol 20:7007–7012 | Article | PubMed | ISI | ChemPort |
29. Nonaka S, Ishiko A, Masunaga T, Akiyama M, Owaribe K, Shimizu H et al. (2000) The extracellular domain of BPAG2 has a loop structure in the carboxy terminal flexible tail in vivo. J Invest Dermatol 115:889–892 | Article | PubMed | ISI | ChemPort |
30. O'Keefe EJ, Payne RE Jr, Russell N, Woodley DT (1985) Spreading and enhanced motility of human keratinocytes on fibronectin. J Invest Dermatol 85:125–130 | Article | PubMed | ChemPort |
31. Olague-Marchan M, Twining SS, Hacker MK, McGrath JA, Diaz LA, Giudice GJ (2000) A disease-associated glycine substitution in BP180 (type XVII collagen) leads to a local destabilization of the major collagen triple helix. Matrix Biol 19:223–233 | Article | PubMed | ISI | ChemPort |
32. Paulsson M, Aumailley M, Deutzmann R, Timpl R, Beck K, Engel J (1987) Laminin-nidogen complex. Extraction with chelating agents and structural characterization. Eur J Biochem 166:11–19 | Article | PubMed | ISI | ChemPort |
33. Rousselle P, Aumailley M (1994) Kalinin is more efficient than laminin in promoting adhesion of primary keratinocytes and some other epithelial cells and has a different requirement for integrin receptors. J Cell Biol 125:205–214 | Article | PubMed | ISI | ChemPort |
34. Salmivirta K, Talts JF, Olsson M, Sasaki T, Timpl R, Ekblom P (2002) Binding of mouse nidogen-2 to basement membrane components and cells and its expression in embryonic and adult tissues suggest complementary functions of the two nidogens. Exp Cell Res 279:188–201 | Article | PubMed | ISI | ChemPort |
35. Schäcke H, Schumann H, Hammami-Hauasli N, Raghunath M, Bruckner-Tuderman L (1998) Two forms of collagen XVII in keratinocytes. A full-length transmembrane protein and a soluble ectodomain. J Biol Chem 273:25937–25943 | Article | PubMed | ISI | ChemPort |
36. Schumann H, Baetge J, Tasanen K, Wojnarowska F, Schacke H, Zillikens D et al. (2000) The shed ectodomain of collagen XVII/BP180 is targeted by autoantibodies in different blistering skin diseases. Am J Pathol 156:685–695 | PubMed | ISI | ChemPort |
37. Schymeinsky J, Nedbal S, Miosge N, Poschl E, Rao C, Beier DR et al. (2002) Gene structure and functional analysis of the mouse nidogen-2 gene: nidogen-2 is not essential for basement membrane formation in mice. Mol Cell Biol 22:6820–6830 | Article | PubMed | ISI | ChemPort |
38. Shimanovich I, Hirako Y, Sitaru C, Hashimoto T, Brocker EB, Butt E et al. (2003) The autoantigen of anti-p200 pemphigoid is an acidic noncollagenous N-linked glycoprotein of the cutaneous basement membrane. J Invest Dermatol 121:1402–1408 | Article | PubMed | ISI | ChemPort |
39. Tasanen K, Tunggal L, Chometon G, Bruckner-Tuderman L, Aumailley M (2004) Keratinocytes from patients lacking collagen XVII display a migratory phenotype. Am J Pathol 164:2027–2038 | PubMed | ISI | ChemPort |
40. Vaisanen L, Has C, Franzke C, Hurskainen T, Tuomi ML, Bruckner-Tuderman L et al. (2005) Molecular mechanisms of junctional epidermolysis bullosa: Col 15 domain mutations decrease the thermal stability of collagen XVII. J Invest Dermatol 125:1112–1118 | Article | PubMed | ISI | ChemPort |
41. Van den Bergh F, Giudice GJ (2003) BP180 (type XVII collagen) and its role in cutaneous biology and disease. Adv Dermatol 19:37–71 | PubMed |
42. Watanabe M, Tsunoda T, Tagami H (2002) A subepidermal blistering dermatosis associated with coexistent IgG and IgA anti-dermal basement membrane zone antibodies; demonstration of IgG antibodies reactive against a 200-kDa dermal antigen. Eur J Dermatol 12:603–606 | PubMed | ISI | ChemPort |
43. Yancey KB (2005) The pathophysiology of autoimmune blistering diseases. J Clin Invest 115:825–828 | Article | PubMed | ISI | ChemPort |
44. Zillikens D, Ishiko A, Jonkman MF, Chimanovitch I, Shimizu H, Hashimoto T et al. (2000) Autoantibodies in anti-p200 pemphigoid stain skin lacking laminin 5 and type VII collagen. Br J Dermatol 143:1043–1049 | Article | PubMed | ISI | ChemPort |
45. Zillikens D, Kawahara Y, Ishiko A, Shimizu H, Mayer J, Rank CV et al. (1996) A novel subepidermal blistering disease with autoantibodies to a 200-kDa antigen of the basement membrane zone. J Invest Dermatol 106:1333–1338 | Article | PubMed | ChemPort |

Acknowledgments

We thank Käthe Thoma and Annegret Bedorf for expert technical assistance and Drs D. Zillikens and I. Shimanovich, Department of Dermatology, University of Lübeck, Germany, for performing western blot studies with dermal extracts. This work was supported in part by Grant Br1475/6-3 from the German Research Foundation, DFG, to L.B-T and by a research fellowship from the German Dermatological Society, DDG, to S.H.