Pemphigus vulgaris (PV) is a life-threatening autoimmune blistering disease of skin and mucous membranes (Stanley, 1998). Autoantibodies found in PV patients bind to the cell surface of keratinocytes and cause loss of cell to cell adhesion with resultant blister formation. The autoimmune target in PV was identified to be a desmosomal cadherin, desmoglein (Dsg) 3 (Amagai et al. 1991;Amagai, 1995,1996, 1999). Compelling evidence has been accumulated for the pathogenic roles of IgG autoantibodies against Dsg3 in blister formation in PV (Schiltz (1976;Anhalt et al. 1982;Hashimoto et al. 1983;Amagai et al. 1992;Mahoney et al. 1999).
Characterizing the binding sites of these autoantibodies on the pemphigus antigen or epitope mapping is therefore an essential step to understanding the pathophysiology of blister formation in pemphigus as well as the basic molecular mechanism of cell–cell adhesion mediated by desmogleins. Previously we examined the epitopes of Dsg3 by PV autoantibodies by immunoblot analysis using various segments of Dsg3 expressed by bacteria (Amagai et al. 1992). Only sequential epitopes were characterized with this approach, however, as immunoblotting requires a denaturing process and all bacterial recombinant proteins produced were insoluble in physiologic saline. Subsequently, recombinant Dsg3 (Dsg3-Ig) with proper conformation was produced as a secreted protein by baculovirus expression (Amagai et al. 1994). Dsg3-Ig was shown to be able to immunoadsorb polyclonal autoantibodies from PV sera. Yet, this immunoadsorptive activity was lost once Dsg3-Ig was denatured by chelating Ca2+, treatment with acid or alkaline solution, or boiling (Amagai et al. 1995b). Therefore, conformational epitopes are considered to be important in autoantibody binding in pemphigus (Kowalczyk et al. 1995;Amagai, 1996).
In this study, we attempted to map conformational epitopes of Dsg3 in PV. To achieve this, we first made truncated recombinant Dsg3 from the C-terminus. A large truncation affected the conformation of Dsg3, however. To overcome this problem, we designed domain-swapped molecules between Dsg3 and Dsg1 and produced them by baculovirus expression, which became a valuable tool to characterize the conformational epitopes of Dsg3.
Materials and methods
Human sera
Sera from patients with clinically and histologically typical PV showed positive reactivity against Dsg3 by enzyme-linked immunosorbent assay (ELISA) (Amagai et al. 1999). Among 25 PV sera used in this study, 12 sera were also positive for anti-Dsg1 IgG. Serial serum samples were obtained from four PV patients at different clinical stages.
Plasmid constructs
We have previously constructed secreted forms of recombinant Dsg1 and Dsg3 by baculovirus expression (Amagai et al. 1994,1995a, 1998). These recombinant proteins contain the entire extracellular domain of Dsg1 or Dsg3 fused with the constant region of human IgG1 and His-tag, and are named Dsg1-IgHis and Dsg3-IgHis, respectively (Amagai et al. 1998) (Figure 1). In this study, three variously truncated Dsg3 proteins and four domain-swapped molecules were constructed (Figure 1).
Figure 1.
Molecular structure of recombinant baculoproteins. (A) Full-length constructs. The entire extracellular domains of Dsg3 and Dsg1 were fused with the constant region of IgG1 and His-tag at the C-terminus. (B) Truncated constructs. Truncated Dsg3 molecules from the C-terminus contain amino acid residues of 1–163, 1–276, and 1–397. (C) Domain-swapped constructs. Domain-swapped molecules between Dsg3 and Dsg1 contain 1–403, 1–161, 163–566, and 405–566 residues of Dsg3 and the corresponding Dsg1 for the deleted regions.
Full figure and legend (38K)For truncated series of Dsg3, the extracellular domains of Dsg3 truncated with various lengths from the C-terminus were fused with the constant region of human IgG1 and His-tag (Figure 1). cDNAs encoding various lengths of the extracellular domain of Dsg3 were polymerase chain reaction (PCR) amplified with appropriate primers (Table 1) using pEVmod-PVIg (Amagai et al. 1994) as a template (GenBank accession number M76482). The PCR products were digested with BglII and XhoI, and ligated with BglII/XhoI-cut pEVmod-PVIgHis (Amagai et al. 1998). These constructs were designated Dsg31-163-IgHis, Dsg31-276-IgHis, and Dsg31-397-IgHis. The residue was counted from the amino-terminus of the mature form of Dsg3 and Dsg1.
Domain-swapped molecules between Dsg3 and Dsg1 were developed by deleting segmental regions of Dsg3 and substituting the deleted regions by corresponding Dsg1 regions. cDNAs for various regions of Dsg3 and Dsg1 were PCR amplified with appropriate primers (Table 1) using pEVmod-PVIg (Amagai et al. 1994) and pKS-Dsg1 (GenBank accession number X56654), respectively, as templates. The PCR products for Dsg3 were digested with BglII and SpeI, and the PCR products for Dsg1 were digested with SpeI and SalI. These fragments were ligated together with BglII-XhoI-cut pEVmod-Dsg3IgHis. These constructs were designated Dsg31-403/Dsg1-IgHis, Dsg31-161/Dsg1-IgHis, Dsg3163-566/Dsg1-IgHis, and Dsg3405-566/Dsg1-IgHis. Because of the introduction of the SpeI site, the Leu residue of Dsg3 at 162 was changed to Thr in Dsg31-161/Dsg1-IgHis and Dsg3163-566/Dsg1-IgHis and the Asp residue of Dsg3 at 404 was changed to Thr in Dsg31-403/Dsg1-IgHis and Dsg3405-566/Dsg1-IgHis.
Production of baculoproteins
The plasmids were cotransfected with BaculoGold baculovirus DNA (Pharmingen, San Diego, CA) in cultured insect Sf9 cells and recombinant viruses were collected from culture supernatant. A high titer of recombinant baculovirus stock was obtained after several rounds of reamplification. High Five cells cultured in serum-free EX cell 405 medium (JRH Bioscience, Lenexa, KS) were infected with high titer virus stock, and incubated for 3 d. Recombinant proteins were produced in the culture supernatant. The culture supernatant contained approximately 5 
g per ml of recombinant baculoproteins.
Immunoblot analysis
The culture supernatants containing the baculoproteins were size-fractionated by SDS-PAGE and transferred to an Immobilon-P membrane (Millipore, Bedford, MA). To detect the truncated molecules, the membranes were incubated with mouse antihuman Dsg3 monoclonal antibody 5H10 (Proby et al. 2000), and subsequently with 1:1000 dilution of alkaline phosphatase-conjugated goat antimouse IgG antibodies (Zymed Laboratories, San Francisco, CA). To detect the domain-swapped molecules, the membranes were incubated with 1:1000 dilution of alkaline phosphatase-conjugated goat antihuman IgG antibodies (Zymed Laboratories).
Competition ELISA
One microliter of serum was incubated in 200 l of culture supernatant containing various recombinant proteins at 4°C overnight. Immunoblot analysis confirmed that each culture supernatant contained approximately the same amount of recombinant proteins (data not shown). The serum was then subjected to Dsg3 ELISA (Medical & Biological Laboratories, Nagoya, Japan), which measured the reactivity against the entire extracellular domain of Dsg3 (Amagai et al. 1999). When OD450 exceeded 1.2, sera were further diluted until OD450 decreased below 1.2. The competition rate was calculated using the following formula: competition rate (%) = [1 – (ODcompetitor – ODpositive)/(ODnegative – ODpositive)] 
100; ODpositive is an OD obtained with sera incubated with culture supernatant containing Dsg3-IgHis; ODnegative is an OD obtained with sera incubated with culture supernatant of uninfected High Five cells; ODcompetitor is an OD obtained with sera incubated with culture supernatant with an indicated recombinant baculoprotein. Serially diluted purified baculoproteins showed competition rates in a dose-dependent manner. Culture supernatants contained sufficient baculoprotein to achieve the plateau competition level (data not shown).
Absorption experiment using indirect immunofluorescence
One microliter of patients' sera was incubated with 50, 100, and 200 l of culture supernatant containing various baculoproteins at 4°C overnight. These sera were then applied to indirect staining on normal human mucosal epidermis with 1:100 dilution of fluorescein isothiocyanate-conjugated antihuman IgG antibodies (Dako, Copenhagen, Denmark) as a second antibody.
Results
A large truncation of Dsg3 from the C-terminus abolished the proper conformation of the native molecule
A series of truncated Dsg3 from the C-terminus were produced as secreted proteins by baculovirus expression (Figure 1). The approximate molecular weights of Dsg31-163-IgHis, Dsg31-276-IgHis, and Dsg31-397-IgHis were 68 kDa, 77 kDa, and 88 kDa, respectively, as determined by immunoblot analysis (Figure 2).
Figure 2.
Production of recombinant baculoproteins for Dsg3. The baculoproteins were visualized by immunoblot. (A) Truncated constructs of Dsg31-163-IgHis (lane 1), Dsg31-276-IgHis (lane 2), and Dsg31-397-IgHis (lane 3) were detected as 68 kDa, 77 kDa, and 88 kDa, respectively. Molecular mass standards are 215, 122, and 79 kDa from top to bottom. (B) Domain-swapped constructs of Dsg31-403/Dsg1-IgHis (lane 1), Dsg31-161/Dsg1-IgHis (lane 2), Dsg3163-566/Dsg1-IgHis (lane 3), and Dsg3405-566/Dsg1-IgHis (lane 4) were detected as 109 kDa, 112 kDa, 109 kDa (a lower band), and 112 kDa (a lower band), respectively. Molecular mass standards are 203, 116, and 83 kDa from top to bottom.
Full figure and legend (51K)These molecules were used as a competitor for ELISA against the entire extracellular domain of Dsg3 with 5 PV sera (representative data of PV serum #2 are shown in Figure 3a). Dsg31-397-IgHis, which is the least truncated molecule, showed approximately 80%-90% competition, whereas Dsg31-276-IgHis and Dsg31-163-IgHis showed only 5%-15% competition. Our previous findings on sequential epitope mapping by immunoblot analysis with bacterial fusion proteins suggested that the amino-terminal region of Dsg3 contains major immunogenic domains (Amagai et al. 1992). Therefore, we suspected that the large truncation of Dsg3 from the C-terminus may have abolished the proper conformation of the amino-terminal domains of the native Dsg3. To overcome this problem, we attempted to produce domain-swapped molecules between Dsg3 and Dsg1.
Figure 3.
Competition ELISA of PV sera with the truncated and domain-swapped recombinant Dsg3s. Competition rate (%) against Dsg3-IgHis was calculated for each recombinant molecule. The truncated Dsg3 molecules were used as a competitor for PV serum #2 (A). The domain-swapped molecules were used as a competitor for PV serum #2 (B), PV serum #1765 (C), and PV serum #1360 (D). Using the same PV serum, no significant competition was seen with Dsg31-276-IgHis and Dsg31-163-IgHis, whereas significant competition was seen with Dsg31-161/Dsg1-IgHis. In most PV sera, the reactivity against the entire extracellular domain of Dsg3 was significantly competed for by the amino-terminal Dsg3 (Dsg31-161/Dsg1-IgHis) (B, C), whereas in some PV sera the reactivity was competed for by more C-terminal Dsg3 (Dsg3163-566/Dsg1-IgHis) (D).
Full figure and legend (61K)Critical conformational epitopes reside in the amino-terminal residues 1–161 of Dsg3 in PV
A series of domain-swapped molecules between Dsg3 and Dsg1 was produced by baculovirus expression (Figure 1). These were developed by deleting segmental regions of Dsg3 and substituting the deleted regions by corresponding Dsg1 regions. Dsg1 was chosen for the replacement because it has high homology in structure with Dsg3 but distinct epitopes (Ishii et al. 1997). These domain-swapped molecules contain the amino acid residues 1–403, 1–161, 163–566, and 405–566 of Dsg3 (the residue was counted from the amino-terminus of the mature form of Dsg3, EWVKF–). By immunoblot analysis, Dsg31-403/Dsg1-IgHis and Dsg31-161/Dsg1-IgHis were detected as 112 kDa and 109 kDa, respectively, and Dsg3163-566/Dsg1-IgHis and Dsg3405-566/Dsg1-IgHis were detected as doublets of 109 kDa, 112 kDa and 112 kDa, 115 kDa, respectively (Figure 2). These doublet bands were thought to be products with (a lower band) and without (an upper band) proteolytic processing of the prosequence, as previously described (Amagai et al. 1994; 1995a).
These domain-swapped molecules showed a quite different competition pattern from that with the truncated Dsg3. Using the same PV serum, Dsg31-403/Dsg1-IgHis and Dsg31-161/Dsg1-IgHis showed 96% and 88% competition against the reactivity of the entire extracellular domains of Dsg3, whereas Dsg31-276-IgHis and Dsg31-163-IgHis showed only 11% and 7% competition (compare Figure 3a,b). Dsg3163-566/Dsg1-IgHis and Dsg3405-566/Dsg1-IgHis showed only 4% and no competition. Therefore, we assumed that these domain-swapped molecules have retained the proper conformation of the native Dsg3 and we used them for further analyses of the conformational epitope mapping of Dsg3.
As a detection assay, we also performed immunofluorescence using normal human skin as a substrate. We found that ELISA was more sensitive and quantitative, however, for evaluating the competition rate by these molecules (data not shown). Therefore, we used ELISA for further analyses.
PV serum #1765 showed almost complete competition by Dsg31-403/Dsg1-IgHis and Dsg31-161/Dsg1-IgHis, but essentially no competition was seen with either Dsg3163-566/Dsg1-IgHis or Dsg3405-566/Dsg1-IgHis. This competition pattern indicates that the major epitopes for PV serum #1765 are located on the amino-terminal residues 1–161. In contrast, PV serum #1360 showed only 52% competition by Dsg31-403/Dsg1-IgHis and essentially no competition by Dsg31-161/Dsg1-IgHis, whereas 60% and 35% competition was seen with Dsg3163-566/Dsg1-IgHis and Dsg3405-566/Dsg1-IgHis, respectively. This competition pattern indicates that there are no apparent epitopes in the amino-terminal regions (1–161), but they are located in the more C-terminal regions (163–566).
To obtain a general picture for the epitopes of Dsg3 in PV, we put all the competition data for 25 PV sera on one graph (Figure 4). The 25 PV sera were ordered according to the competition rate by Dsg31-161/Dsg1-IgHis and appeared in the same order in each graph. Considering more than 50% competition as significant, Dsg31-403/Dsg1-IgHis and Dsg31-161/Dsg1-IgHis showed significant competition in 24 (96%) and 18 (72%), respectively, of 25 PV sera. In contrast, Dsg3163-566/Dsg1-IgHis and Dsg3405-566/Dsg1-IgHis demonstrated significant competition only in four (16%) and one (4%) of them. There was a tendency for PV sera with less competition by Dsg31-161/Dsg1-IgHis to show more competition by Dsg3163-566/Dsg1-IgHis.
Figure 4.
The amino-terminal residues 1–161 express major conformational epitopes of Dsg3 in PV. PV sera were competed with Dsg31-403/Dsg1-IgHis (A), Dsg31-161/Dsg1-IgHis (B), Dsg3163-566/Dsg1-IgHis (C), and Dsg3405-566/Dsg1-IgHis (D). The 25 PV sera tested appeared in the same order in each graph.
Full figure and legend (67K)These findings indicate that the amino-terminal residues 1–161 express major conformational epitopes of Dsg3 in PV, and that the residues 163–403 contain some minor epitopes, whereas the residues 405–566 do not contain any critical epitopes.
No apparent epitope change during the time course in PV
To evaluate the change in conformational epitopes of Dsg3 along the time course, we compared the competition pattern by these swapping molecules at different time points in four PV patients. Sera were taken at active stages and in remission. All of these four cases had positive anti-Dsg3 IgG detected by ELISA even in remission. The duration of the tested sera for cases 1–4 was 17, 26, 28, and 31 mo, respectively. A representative case of PV, case 1 (Figure 5), showed the same competition pattern throughout the time course (repeated measures analysis of variance, p = 0.523), in this case demonstrating significant competition by both Dsg31-403/Dsg1-IgHis and Dsg31-161/Dsg1-IgHis during active stages and remission. We did not find any significant difference in the competition pattern in three other PV patients tested as well (p > 0.2), suggesting that apparent intramolecular epitope spreading along the disease course is not a common event in PV.
Figure 5.
No apparent change in competition pattern during the time course was seen in a representative case of PV, case 1. The competition pattern by various domain-swapped molecules was compared at the initial stage (A), intermediate stage (B), and in remission (C).
Full figure and legend (50K)Discussion
In this study, a novel strategy based on domain-swapped molecules was employed to map regions within Dsg3 that constitute the conformational epitopes for PV autoantibodies. This approach has several advantages over previous approaches. Unlike in epitope mapping with recombinant bacterial proteins (Amagai et al. 1992), Dsg3 candidate regions were grafted onto a Dsg1 framework and produced by baculovirus expression to ensure the native conformation. Dsg1 was most appropriate as a backbone structure for small Dsg3 regions because Dsg1 has high homology in structure with Dsg3 (Amagai et al. 1991) and there is no significant cross-reactivity to Dsg1 by anti-Dsg3 IgG in PV sera (Ishii et al. 1997). In addition, the resulting domain-swapped molecules were used as competitors for ELISA against the entire extracellular domain of Dsg3, which allowed us to measure autoantibodies against specific regions on Dsg3 in a quantitative fashion.
A similar approach with domain-swapped molecules was taken to map binding specificity of E-cadherin and P-cadherin (Nose et al. 1990). Several domain-swapped molecules between E-cadherin and P-cadherin were constructed and transfected on mouse fibroblastic L cells. Investigation of adhesive selectivity of the transfected cells indicated that the amino-terminal 113 amino acid region is essential to determine the binding specificity. This approach with domain-swapped molecules proved to be useful with various molecules in characterizing specific functional domains on the molecule or mapping conformational epitopes. To perform epitope mapping for Goodpasture autoimmune disease, immunoreactive 
3(IV) collagen chain was swapped with nonimmunoreactive 1(IV) collagen chain (Netzer et al. 1999). To characterize epitopes for thyroid stimulating autoantibodies in Graves' disease and blocking autoantibodies in Hashimoto's disease, swapped molecules between thyrotropin receptor and lutropin-choriogonadotropin receptor were constructed (Tahara et al. 1997). To map conformational surface epitopes on trypanosome, a variant-specific surface glycoprotein was swapped with another antigenically distinct variant surface glycoprotein (Hsia et al. 1996).
Using the domain-swapped molecules between Dsg3 and Dsg1, it was shown from our study that the amino-terminal residues 1–161 contain the major determinant of conformational epitopes of Dsg3 in PV. Dsg3 and Dsg1, like other typical cadherins, have five tandemly repeated extracellular domains (EC1-EC5) (Figure 1). EC1-EC4 domains consist of so-called cadherin repeat containing calcium binding motifs and show significant similarities between Dsg3 and Dsg1 (Amagai et al. 1991). EC5 shows no significant homologies between the two molecules. The amino-terminal residues 1–161 correspond to EC1 and part of the EC2 domain. Although the precise functional domain mapping for Dsg remains to be elucidated, the three-dimensional molecular structure of the amino-terminal classic cadherin has become available by multidimensional heteronuclear magnetic resonance spectroscopy and radiographic characterization of the crystal structure (Overduin et al. 1995;Shapiro et al. 1995;Nagar et al. 1996). These observations propose a cell-adhesion zipper model for the structural basis of cell adhesion mediated by cadherins. In this model, the EC1 domain contains binding regions that directly interact with an opposing EC1 domain on another cell. If we can apply this model of classic cadherins to Dsg, our findings suggest that most autoantibodies against Dsg3 raised in patients with PV are directed against the binding surface of Dsg3 and that these autoantibodies may directly interfere with the adhesive function of Dsg3 with a resultant blister formation.
Recent studies have suggested that the target of immune responses in autoimmunity does not remain fixed but can be extended to other epitopes on the same or different proteins in the same tissue (Vanderlugt & Miller, 1996). It is hypothesized that epitopes secondary to the release of self-antigens during a chronic autoimmune or inflammatory response develop. In pemphigus, intermolecular epitope spreading that accompanies the phenotypic shift between PV and pemphigus foliaceus occurs as a rare event (Iwatsuki et al. 1991;Kawana et al. 1994;Ishii et al. 2000). In this study we looked for the possibility of intramolecular epitope spreading during the course of the disease. We did not find any change in the pattern of competition ELISA in the four patients with PV tested, however, suggesting that intramolecular epitope spreading along the disease course is not a common event (Figure 5). More precise epitope mapping needs to be performed to confirm this speculation.
A strategy using domain-swapped molecules between Dsg3 and Dsg1 is useful for conformational epitope mapping in pemphigus, and may be applicable to various types of autoimmune disease with conformation-dependent epitopes. Using these molecules, amino-terminal residues 1–161 of Dsg3 were found to be the major immunogenic epitope recognized by autoantibodies in PV sera. This region is considered to also include a region essential for cell-cell adhesion in cadherins. These findings provide new knowledge to elucidate the molecular mechanism of the adhesion function of Dsg3 and the pathophysiologic mechanism of blister formation in pemphigus.
References
| 1. | Amagai M. Adhesion molecules I. Keratinocyte–keratinocyte interactions, cadherins and pemphigus. J Invest Dermatol (1995) 104: 146–152. | Article | PubMed | ISI | ChemPort | |
| 2. | Amagai M. Autoimmunity against desmosomal cadherins in pemphigus. J Dermatol Sci (1999) 20: 92–102 10.1016/S0923-1811(99)00016-X. | Article | PubMed | ISI | ChemPort | |
| 3. | Amagai M. Pemphigus: autoimmunity to epidermal cell adhesion molecules. Adv Dermatol (1996) 11: 319–352. | PubMed | ChemPort | |
| 4. | Amagai M, Hashimoto T, Green KJ, Shimizu N & Nishikawa T. Antigen-specific immunoadsorption of pathogenic autoantibodies in pemphigus foliaceus. J Invest Dermatol (1995a) 104: 895–901. | Article | PubMed | ISI | ChemPort | |
| 5. | Amagai M, Hashimoto T, Shimizu N & Nishikawa T. Absorption of pathogenic autoantibodies by the extracellular domain of pemphigus vulgaris antigen (Dsg3) produced by baculovirus. J Clin Invest (1994) 94: 59–67. | PubMed | ISI | ChemPort | |
| 6. | Amagai M, Ishii K, Hashimoto T, Gamou S, Shimizu N & Nishikawa T. Conformational epitopes of pemphigus antigens (Dsg1 and Dsg3) are calcium dependent and glycosylation independent. J Invest Dermatol (1995b) 105: 243–247. | Article | PubMed | ISI | ChemPort | |
| 7. | Amagai M, Karpati S, Prussick R, Klaus-Kovtun V & Stanley JR. Autoantibodies against the amino-terminal cadherin-like binding domain of pemphigus vulgaris antigen are pathogenic. J Clin Invest (1992) 90: 919–926. | PubMed | ISI | ChemPort | |
| 8. | Amagai M, Klaus-Kovtun V & Stanley JR. Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion. Cell (1991) 67: 869–877. | Article | PubMed | ISI | ChemPort | |
| 9. | Amagai M, Komai A & Hashimoto T et al. Usefulness of enzyme-linked immunosorbent assay (ELISA) using recombinant desmogleins 1 and 3 for serodiagnosis of pemphigus. Br J Dermatol (1999) 140: 351–357 10.1046/j.1365-2133.1999.02752.x. | Article | PubMed | ISI | ChemPort | |
| 10. | Amagai M, Nishikawa T, Nousari HC, Anhalt GJ & Hashimoto T. Antibodies against desmoglein 3 (pemphigus vulgaris antigen) are present in sera from patients with paraneoplastic pemphigus and cause acantholysis in vivo in neonatal mice. J Clin Invest (1998) 102: 775–782. | PubMed | ISI | ChemPort | |
| 11. | Anhalt GJ, Labib RS, Voorhees JJ, Beals TF & Diaz LA. Induction of pemphigus in neonatal mice by passive transfer of IgG from patients with the disease. N Engl J Med (1982) 306: 1189–1196. | PubMed | ISI | ChemPort | |
| 12. | Hashimoto K, Shafran KM, Webber PS, Lazarus GS & Singer KH. Anti-cell surface pemphigus autoantibody stimulates plasminogen activator activity of human epidermal cells. J Exp Med (1983) 157: 259–272. | Article | PubMed | ISI | ChemPort | |
| 13. | Hsia R, Beals T & Boothroyd JC. Use of chimeric recombinant polypeptides to analyse conformational, surface epitopes on trypanosome variant surface glycoprotein. Mol Microbiol (1996) 19: 53–63. | Article | PubMed | ISI | ChemPort | |
| 14. | Ishii K, Amagai M & Hall RP et al. Characterization of autoantibodies in pemphigus using antigen-specific ELISAs with baculovirus expressed recombinant desmogleins. J Immunol (1997) 159: 2010–2017. | PubMed | ISI | ChemPort | |
| 15. | Ishii K, Amagai M, Ohata Y, Shimizu H, Hashimoto T, Ohya K & Nishikawa T. Development of pemphigus vulgaris in a patient with pemphigus foliaceus: confirmation of anti-desmoglein antibody profile shift by ELISA. J Am Acad Dermatol (2000) 42: 859–861. | Article | PubMed | ISI | ChemPort | |
| 16. | Iwatsuki K, Takigawa M, Hashimoto T, Nishikawa T & Yamada M. Can pemphigus vulgaris become pemphigus foliaceus? J Am Acad Dermatol (1991) 25: 797–800. | PubMed | ISI | ChemPort | |
| 17. | 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 (1994) 130: 1534–1538. | Article | PubMed | ISI | ChemPort | |
| 18. | Kowalczyk AP, Anderson JE, Borgwardt JE, Hashimoto T, Stanley JR & Green KJ. Pemphigus sera recognize conformationally sensitive epitopes in the amino-terminal region of desmoglein-1. J Invest Dermatol (1995) 105: 147–152. | Article | PubMed | ISI | ChemPort | |
| 19. | Mahoney MG, Wang Z, Rothenberger KL, Koch PJ, Amagai M & Stanley JR. Explanation for the clinical and microscopic localization of lesions in pemphigus foliaceus and vulgaris. J Clin Invest (1999) 103: 461–468. | PubMed | ISI | ChemPort | |
| 20. | Nagar B, Overduin M, Ikura M & Rini JM. Structural basis of calcium-induced E-cadherin rigidification and dimerization. Nature (1996) 380: 360–364. | Article | PubMed | ISI | ChemPort | |
| 21. | Netzer K, Leinonen A & Boutaud A et al. The Goodpasture autoantigen. Mapping of the major conformational epitope (s) of 3 (IV) collagen to residues 17–31 and 127–141 of the NC1 domain. J Biol Chem (1999) 274: 11267–11274. | Article | PubMed | ISI | ChemPort | |
| 22. | Nose A, Tsuji K & Takeichi M. Localization of specificity determining sites in cadherin cell adhesion molecules. Cell (1990) 61: 147–155. | Article | PubMed | ISI | ChemPort | |
| 23. | Overduin M, Harvey TS, Bagby S, Tong KI, Yau P, Takeichi M & Ikura M. Solution structure of the epithelial cadherin domain responsible for selective cell adhesion. Science (1995) 267: 386–389. | PubMed | ISI | ChemPort | |
| 24. | Proby CM, Ohta T & Suzuki H et al. Development of chimeric molecules for recognition and targeting of antigen-specific B cells in pemphigus vulgaris. Br J Dermatol (2000) 142 in press. | Article | PubMed | ChemPort | |
| 25. | Schiltz JR & Michel B. Production of epidermal acantholysis in normal human skin in vitro by the IgG fraction from pemphigus serum. J Invest Dermatol (1976) 67: 254–260. | Article | PubMed | ISI | ChemPort | |
| 26. | Shapiro L, Fannon AM & Kwong PD et al. Structural basis of cell-cell adhesion by cadherins. Nature (1995) 373: 327–337. |
| 27. | Stanley JR. Pemphigus. In: Freedberg, IM, Eisen, AZ, Wolff, K, Austen, KF, Goldsmith, LA, Katz, SI & Fitzpatrick, TB, edsDermatology in General Medicine (1998) New York: McGraw-Hill pp 654–666. |
| 28. | Tahara K, Ishikawa N & Yamamoto K et al. Epitopes for thyroid stimulating and blocking autoantibodies on the extracellular domain of the human thyrotropin receptor. Thyroid (1997) 7: 867–877. | PubMed | ISI | ChemPort | |
| 29. | Vanderlugt CJ & Miller SD. Epitope spreading. Curr Opin Immunol (1996) 8: 831–836. | Article | PubMed | ISI | ChemPort | |
Acknowledgments
We thank Ms. Minae Suzuki for the immunofluorescence work. This work was supported by Health Sciences Research Grants for Research on Specific Diseases from the Ministry of Health and Welfare and Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan.
