Search

Pemphigus is an organ-specific autoimmune disease that targets skin and/or mucous membranes. Our previous study showed that infiltrating lymphocytes in pemphigus vulgaris (PV) lesions produce anti-desmoglein (Dsg) 1/3 antibodies after in vitro culture. In this study, we found diffuse ectopic lymphoid-like structures (ELSs) commonly present in the lesions of both PV and pemphigus foliaceus. Notably, pemphigus lesions contained centroblasts, plasmablasts, and plasma cells, which recapitulated the different stages of B cell differentiation. Elevated mRNA expression levels of the differentiation-related transcription factors BLIMP-1, IRF4, and BCL-6 were observed in pemphigus lesions. Moreover, B cell receptor repertoire analysis revealed the clonal expansion of the lesional B cells. Lesional B cells might recirculate among lesions, lymph nodes, and peripheral blood. Increased mRNA expression levels of multiple chemokines in pemphigus lesions and elevated expression levels of chemokine receptors on lesional B cells were also observed. Collectively, these results show that the ELSs in pemphigus lesions might act as a niche, supporting in situ B cell differentiation and clonal expansion., (Copyright © 2019 The Authors. Published by Elsevier Inc. All rights reserved.)

Paraneoplastic pemphigus (PNP) is an autoimmune blistering disease associated with lymphoproliferative neoplasms and characterized by antibodies against plakins and desmoglein 3 (Dsg3). Anti-Dsg3 antibodies have a primary role in blister formation in PNP. In this study, we used phage display to clone monoclonal anti-Dsg3 antibodies from a PNP patient to further characterize their pathogenicity. We isolated 20 unique Dsg3-reactive mAbs, which we classified into four groups according to the heavy-chain complementarity-determining region 3 (CDR3) region. Genetic analyses demonstrated that three antibody groups used the VH1-46 gene (18 clones) and one group used the VH1-02 gene (2 clones). The results of an in vitro keratinocyte dissociation assay and a human skin organ culture injection assay showed that three antibodies displayed pathogenic activity in blister formation with different potencies. Epitope mapping using domain-swapped Dsg3/Dsg2 showed that these pathogenic mAbs bound Ca(2+)-dependent conformational epitopes in the middle portion of the extracellular region of Dsg3 (EC2 and EC3 domains), in contrast to most previously characterized pathogenic pemphigus vulgaris antibodies, which bound to the EC1 domain of Dsg3. These mAbs reflect the unique polyclonal nature of anti-Dsg3 antibodies in PNP and represent an important tool for detailing the pathophysiological mechanisms of blister formation in PNP.

Epitope spreading is involved in inducing and maintaining self-reactivity. Epitope spreading in pemphigus vulgaris (PV), caused by IgG autoantibodies to desmoglein 3 (Dsg3) and Dsg1, was previously analyzed using Dsg3/Dsg1 extracellular domain-swapped molecules. However, precise identification of the responsible epitopes in each molecule by using only this method was problematic. In this study, we studied epitope spreading in PV by a novel immunoprecipitation-immunoblot method using Dsg3 (or Dsg1)/Dsg2 domain-swapped molecules, which overcomes the problems associated with the previous approaches. We analyzed the antigenic epitopes recognized by 212 sera collected from 53 PV patients at multiple disease stages. The major epitopes were present at the N-terminal region of Dsgs and were unchanged over the course of the disease in both anti-Dsg3 mucosal dominant-type PV and anti-Dsg3/Dsg1 mucocutaneous-type PV. These N-terminal epitopes were calcium dependent. Circulating antibodies in paraneoplastic pemphigus and pemphigus herpetiformis had unique epitope distributions, although the Dsg N-termini still contained the major epitopes. These results suggest that, after onset, intramolecular and intermolecular epitope spreading among extracellular domains on Dsg3 and Dsg1 is rare in PV and has no correlation with disease course.

Pemphigus vulgaris (PV) is an epidermal blistering disorder caused by antibodies directed against the desmosomal cadherin desmoglein-3 (Dsg3). The mechanism by which PV IgG disrupts adhesion is not fully understood. To address this issue, primary human keratinocytes (KCs) and patient IgG were used to define the morphological, biochemical, and functional changes triggered by PV IgG. Three phases of desmosome disassembly were distinguished. Analysis of fixed and living KCs demonstrated that PV IgG cause rapid Dsg3 internalization, which likely originates from a non-junctional pool of Dsg3. Subsequently, Dsg3 and other desmosomal components rearrange into linear arrays that run perpendicular to cell contacts. Dsg3 complexes localized at the cell surface are transported in a retrograde manner along with these arrays before being released into cytoplasmic vesicular compartments. These changes in Dsg3 distribution are followed by depletion of detergent-insoluble Dsg3 pools and by the loss of cell adhesion strength. Importantly, this process of disassembly can be prevented by expressing exogenous Dsg3, thereby driving Dsg3 biosynthesis and desmosome assembly. These data support a model in which PV IgG cause the loss of cell adhesion by altering the dynamics of Dsg3 assembly into desmosomes and the turnover of cell surface pools of Dsg3 through endocytic pathways.

Pemphigus vulgaris (PV) is a potentially fatal blistering disease characterized by autoantibodies against cell surface adhesion proteins desmoglein (Dsg) 3 and Dsg1. Previous studies using phage display to clone Dsg-reactive monoclonal antibodies from a PV patient demonstrated that a limited number of antibody variable region genes encode the autoantibody repertoire, with different genes for pathogenic and non-pathogenic mAbs. Here, we investigated the feasibility of specific autoantibody targeting in pemphigus. We produced rabbit anti-idiotypic antibodies against two pathogenic and two non-pathogenic PV mAbs. Antisera inhibited binding of the immunizing mAb to Dsgs by ELISA as well as pathogenicity against cultured human keratinocytes. Antisera also inhibited other mAbs using the same variable region heavy chain (V(H)) genes, despite different light chains or somatic mutations. Additionally, peptide phage display identified peptide sequences that bound PV mAbs in a V(H)-specific manner. To evaluate the therapeutic potential of V(H) gene-targeted reagents, preimmune sera and antisera were used to adsorb pathogenic antibodies from PV sera. Pooled antisera significantly reduced pathogenic activity from the original PV patient's serum and bound pathogenic antibodies from two other PV sera, suggesting shared autoantibody V(H) gene usage among PV patients. Together, these data suggest novel V(H) gene-targeted approaches toward PV treatment.

Pemphigus vulgaris (PV) is an autoantibody-mediated bullous disease, but the role of natural killer (NK) cells in its pathogenic process has never been examined in detail. Circulating CD56+ CD3- NK cells as well as CD69+-activated NK cells were increased in PV patients compared with healthy controls and patients with other autoantibody-mediated autoimmune diseases, including immune thrombocytopenic purpura and myasthenia gravis. Gene expression analysis of highly purified NK cells demonstrated an increased expression of IL-10 and decreased expression of IL-12Rbeta2, perforin, and granzyme B ex vivo in PV patients versus healthy controls. The NK cells from PV patients also showed impaired signal transducer and activator of transduction4 phosphorylation upon in vitro IL-12 stimulation. Moreover, NK cells from PV patients exhibited reduced IL-10 production in response to in vitro stimulation with IL-2/IL-12. Finally, IL-5 expression in NK cells was exclusively detected ex vivo in PV patients with active disease, and was lost in subsequent analyses performed during disease remission. Together these findings suggest that NK cells contribute to a T helper type 2-biased immune response in PV patients through impaired IL-12 signaling and an upregulation of IL-10 and IL-5.

The Fas/Fas ligand system triggers the extrinsic apoptotic pathway and is involved in several inflammatory conditions, also at the skin level. The Fas/Fas ligand cell death pathway plays a major role in anoikis, a type of apoptosis characterized by cell detachment. As pemphigus is characterized by loss of cell to cell adhesion, we evaluated the role of anoikis and Fas ligand in this bullous disease. We report that, in suprabasal epidermis from perilesional pemphigus skin, most keratinocytes are apoptotic. Moreover, Fas ligand levels are markedly increased in sera from pemphigus patients, whereas they are undetectable in sera from patients undergoing steroid treatment. Sera from untreated patients but not from patients under steroids induce keratinocyte apoptosis. Pemphigus-sera-induced cell death is partially inhibited by pretreatment with anti-Fas ligand antibodies and by incubation with caspase-8 inhibitor Z-IETD-FMK. Finally, caspase-8 is activated in keratinocytes provided with sera from pemphigus patients, whereas cleavage is partially blocked by pretreatment of sera with anti-Fas ligand antibody. These results suggest that increased Fas ligand in pemphigus sera is responsible for keratinocyte apoptosis, which occurs through the activation of a caspase-8-driven extrinsic apoptotic pathway.

Paraneoplastic pemphigus sera react with multiple plakin family proteins, among which only envoplakin and periplakin are constantly detected by immunoblotting using normal human epidermal extracts. Using bacterial expression vectors containing polymerase chain reaction-amplified cDNA, we have prepared variously truncated recombinant glutathione-S-transferase-fusion proteins of envoplakin and periplakin, which presented N-terminal, central and C-terminal domains of each protein, as well as the so-called C-terminal homologous domain of envoplakin and the junctional regions of these domains. By immunoblotting using these 11 recombinant proteins, we demonstrated that most of the 26 paraneoplastic pemphigus sera reacted very strongly with multiple recombinant proteins of envoplakin and periplakin, except for the C-terminal homologous domain of periplakin. We also examined the reactivity with these recombinant proteins of other blistering diseases, including pemphigus vulgaris, pemphigus foliaceus, and bullous pemphigoid, and found that a few nonparaneoplastic pemphigus sera showed a weak reactivity with some of the recombinant proteins. Interestingly, some sera showed relatively strong reactivity with the C-terminal homologous domain of periplakin to which paraneoplastic pemphigus sera reacted less frequently. These results indicate that, although nonparaneoplastic pemphigus sera occasionally show a weak reactivity with envoplakin and periplakin, the pathogenicity and the mechanism of antibody production in these cases may be different from those in paraneoplastic pemphigus.

Paraneoplastic pemphigus (PNP) is an autoimmune blistering disease that occurs in association with underlying neoplasms. PNP patients develop characteristic autoantibodies directed against multiple antigens, mostly identified as members of the plakin family of cytoplasmic proteins (desmoplakin I and II, bullous pemphigoid antigen I, envoplakin, and periplakin). HD1/plectin, another member of the plakin family, has not previously been detected in the characteristic PNP antigen complex, which may relate to practical difficulties associated with its large size (molecular weight approximately 500 kDa). In this study, a combination of immunoprecipitation and immunoblot is used to demonstrate that HD1/plectin is also recognized by sera from PNP patients. Thirteen of 16 PNP sera tested were positive for HD1/plectin compared with none of 43 control sera (11 pemphigus vulgaris, 11 pemphigus foliaceus, 11 bullous pemphigoid, and 10 normal individuals). Combined with our recent finding that desmoglein 3 and desmoglein 1 are cell surface target antigens in PNP, this demonstration of plectin/HD1 as another component of the antigen complex in PNP confirms that PNP is an autoimmune disease against desmoglein and plakin family molecules.

Paraneoplastic pemphigus has been associated with both malignancies and multicentric Castleman's disease; the latter is a rare angiolymphoproliferative disorder that has also been linked with human herpesvirus 8 (HHV8) infection. Other diseases definitively associated with HHV8 include Kaposi's sarcoma and primary effusion lymphoma. In a search for additional HHV8-associated diseases, patients with paraneoplastic pemphigus, as well as patients with pemphigus vulgaris and pemphigus foliaceus, were studied. Using an immunofluorescence assay able to specifically detect antibodies directed against lytically induced HHV8 antigens, HHV8 antibodies were not detected in sera from 24 patients with paraneoplastic pemphigus (including 10 with concomitant Castleman's disease) nor from 19 patients with pemphigus vulgaris. Sera from patients with Kaposi's sarcoma and from healthy U.S. blood donors were positive (25 of 26) and negative (none of 20), respectively. In addition, HHV8 DNA was not found in frozen lesional skin of five patients with pemphigus vulgaris and five patients with pemphigus foliaceus by nested polymerase chain reaction (lower limit of detection = 10 copies viral DNA per microg total cellular DNA). Finally, tissue sections of lesional skin from 10 patients with pemphigus vulgaris were negative for HHV8 by in situ hybridization, using probes able to detect both latently and lytically expressed HHV8 genes in Kaposi's sarcoma tissue. In summary, no evidence of HHV8 infection was found in all types of pemphigus using a variety of methods. These findings do not support a general role for HHV8 in skin diseases associated with immunosuppression.

Pemphigus vulgaris (PV) is an autoimmune blistering disease of the skin and is caused by autoantibodies against desmoglein 3 (Dsg3) on epidermal keratinocytes. Because the production of autoantibodies is presumably T cell dependent, Dsg3-specific T cell reactivity was investigated in 14 PV patients and 12 healthy donors. Peripheral blood mononuclear cells from seven PV patients with active disease showed a primary in vitro response to a recombinant protein containing the extracellular portion (EC1-5) of Dsg3, whereas two of seven PV patients in remission or under immunosuppressive treatment exhibited only secondary (2 degrees) or tertiary (3 degrees) T cell responses to Dsg3. T cell responses to Dsg3 were also observed in four of 12 healthy individuals upon 2 degrees or 3 degrees stimulation with Dsg3. Both PV patients and healthy responders were either positive for DRbeta1*0402 -- which is highly prevalent in PV -- or positive for DR11 alleles homologous to DRbeta1*0402. Two CD4+ Dsg3-specific T cell lines and 12 T cell clones from two PV patients and two CD4+ T cell lines and eight T cell clones from two normals were also stimulated by a Dsg3 protein devoid of the EC2-3 (deltaN1), suggesting that epitopes were located in the EC1, EC4, and/or EC5. Using Dsg3 peptides, one immunodominant peptide (residues 161-177) was also identified in the EC2. These observations demonstrate that T cell responses to Dsg3 can be detected in PV patients and in healthy donors carrying major histocompatibility class II alleles identical or similar to those highly prevalent in PV.

We have demonstrated previously that pemphigus vulgaris (PV)-IgG induces activation of phospholipase C (PLC), production of inositol 1,4,5-trisphosphate, and a rapid transient increase in [Ca2+]i in cultured human keratinocytes, leading to secretion of plasminogen activator and cell-cell detachment in cell culture. In the current study, to examine the involvement of protein kinase C (PKC) in the mechanism of blister formation in PV, we studied the PV-IgG-induced translocation of PKC isozymes from the cytosol to the particulate/cytoskeleton (p/c) fractions and the activation of PKC in human keratinocytes. Cells cultured in Eagle's minimum essential medium were incubated with PV-IgGs for 30 s, 1 min, 5 min, or 30 min. PV-IgG binding to the cell surface antigen (desmoglein III) induced translocation of PKC-alpha from the cytosol to the p/c fractions within 30 s, with a peak at 1 min that lasted at least 30 min. PKC-delta also was translocated within 1 min and reached a peak at 5 min but was reduced to basal levels at 30 min. Alternatively, PKC-eta translocation to the p/c fraction was induced slowly, taking more than 5 min, and was reduced to approximately half-maximum at 30 min, whereas PKC-zeta translocation reached a maximum at 30 s, rapidly returning to baseline by 5 min after PV-IgG stimulation. The total PKC activity in the p/c fraction also was increased after PV-IgG exposure, peaked at 1 min, and was sustained for at least 30 min. These findings suggest that a unique activation profile of PKC isomers may be involved in mediating the intracellular signaling events induced by PV-IgG binding to desmoglein III in cultured human keratinocytes.

Pemphigus vulgaris and pemphigus foliaceus are characterized by autoantibodies directed against transmembrane glycoproteins of desmosomes. F12, a human monoclonal autoantibody that binds to the desmosomal plaque, recognizes a 180-190-kDa doublet when immunoblotted against bovine tongue epithelium. Because F12 was derived from the peripheral blood lymphocytes of a patient with pemphigus vulgaris, we looked for the presence of anti-180-190-kDa antibodies in pemphigus vulgaris and pemphigus foliaceus serum. By immunoblot analysis, a third of the pemphigus serum contained anti-180-190-kDa antibodies that belonged to IgG subclass 1 or 3, unlike those that recognized desmogleins 1 and 3 (IgG4). By immunoelectron microscopy analysis on human oral mucosa and human skin with mAb to human IgG3, pemphigus serum containing anti-180-190 kDa antibodies recognized desmosomal plaques. The presence of antibodies with F12 properties in pemphigus serum was further demonstrated by a rabbit anti-F12 idiotype antiserum that allowed detection of F12 idiotype in serum with anti-180-190-kDa antibodies. These results indicate that some pemphigus vulgaris and pemphigus foliaceus serums contain antibodies that react with both intra- and extracellular structures of desmosomes and further demonstrate the heterogeneity of the autoimmune response in both types of pemphigus.

In Patients with pemphigus vulgaris, autoantibodies against the desmosomal glycoprotein desmoglein 3 (Dsg3) cause blisters due to loss of keratinocyte cell-cell adhesion in the basal and immediate suprabasal layer of the deeper epidermis, leaving the superficial epidermis intact. Autoantibodies from these patients, however, usually bind to the cell surface of keratinocytes throughout the entire epidermis, as determined by indirect immunofluorescence. To explain this apparent paradox, we immunoadsorbed pemphigus vulgaris sera with the extracellular domains of Dsg3 and desmoglein 1 (Dsg1) produced by insect cells infected with recombinant baculovirus. When adsorbed with extracellular domains of both Dsg3 and Dsg1, these sera no longer stained epidermis, demonstrating that most, if not all, of their cell surface reactivity can be attributed to antibodies against the extracellular domains of these desmogleins. Adsorption with only the Dsg1 extracellular domain left antibodies that stained only the basal and immediate suprabasal layers of the epidermis and immunoprecipitated only Dsg3, not Dsg1, from extracts of cultured cells synthesizing these molecules. In contrast, adsorption with only the Dsg3 extracellular domain left antibodies that stained only the more superficial epidermis and immunoprecipitated only Dsg1. These data localize Dsg3 exactly to the area in the epidermis where blisters occur in pemphigus vulgaris.

We analyzed the location of binding sites for pemphigus vulgaris (PV) antigen and pemphigus foliaceus (PF) antigen in the human epidermis using serum samples obtained from three patients with PV and three patients with PF. Confocal laser scanning microscopy, immunofluorescent examination of ultrathin cryosections, and immunoperoxidase electron microscopy demonstrated discontinuous dots along the epidermal cell surfaces. Immunogold electron microscopy of ultrathin cryosections showed specific binding of PV and PF autoantibodies only to desmosomes. Post-embedding immunogold electron microscopy using cryofixation and cryosubstitution enabled the whole depth of the epidermis to be examined and the binding of PV and PF autoantibodies to be quantitated by counting gold particles. Both PV and PF autoantibodies bound to all desmosomes in the epidermis, but not to the surface of the non-desmosomal keratinocytes. The majority of auto-antibody binding occurred in the extracellular domain (PV, 62%; PF, 69%). The statistical analysis of two-way analysis of variance regarding the number of gold particles labeling a single desmosome confirmed a significant interaction between subtypes of pemphigus (PV and PF) and the different epidermal cell layers (p < 0.044). The results indicate that the number of gold particles bound to individual desmosomes with PV sera was significantly higher in the lower epidermis than in the upper epidermis, and that of PF sera showed reciprocal pattern. This inversely graded binding pattern suggests heterogeneity of the composition of the desmosomes, which may explain the differences in level of acantholysis between PV and PF.

The target molecule of pemphigus autoantibodies is a transmembrane desmosomal component, desmoglein 3 (Dsg3) in pemphigus vulgaris (PV) and Dsg1 in pemphigus foliaceus (PF). In this study, we examined the effects of calcium and glycosylation on the anti-genicity of the pemphigus antigens and on the generation of conformational epitopes. We used recombinant baculovirus proteins, PVIg and PFIg, which are considered to reflect accurately the native conformation of the extracellular domain of their respective proteins Dsg3 and Dsg1. These baculoproteins could immunoadsorb heterogeneous autoantibodies from the corresponding sera of PV and PF patients, completely blocking indirect immunofluorescence staining of normal human skin. Chelating calcium from the solution containing the baculoproteins using ethylenediaminetetraacetic acid (EDTA) or ethyleneglycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) abolished immunoadsorption by both PVIg and PFIg; however, immunoadsorption by the baculoproteins was restored after dialysis against 1 mM calcium. Nonglycosylated forms of both baculoproteins produced in the presence of tunicamycin retained their immunoadsorptive ability. Furthermore, immunoadsorption by the baculo-proteins was prevented irreversibly by treatment with low pH, high pH, and boiling, but not with the non-ionic detergent Nonidet P-40. These findings indicate that formation of the conformational epitopes on the pemphigus antigens is dependent on calcium but independent of glycosylation, and provide direct evidence that calcium plays an important role in determining the antigenic properties of the pemphigus antigens.

To identify regions of the Desmoglein-1 (Dsg1) extracellular domain that are targeted by pemphigus antisera, cDNA sequences encoding various regions of the Dsg1 extracellular domain were ligated to sequences encoding the E-cadherin extracellular anchor and transmembrane and cytoplasmic domains. These constructs were then expressed in mammalian cells, and pemphigus sera were tested for the ability to recognize the Dsg1 extracellular domains. When analyzed by immunoblot, very few pemphigus foliaceus sera recognized the Dsg1 domains. To determine whether pemphigus sera recognize non-denatured Dsg1 domains, constructs were expressed in cultured cells and tested for reactivity with pemphigus sera using live-cell immunofluorescence. The pemphigus foliaceus sera reacted strongly with the Dsg1 extracellular domain by live-cell immunofluorescence and recognized predominantly the amino-terminal region of the Dsg1 extracellular domain. In addition, sera from patients with pemphigus vulgaris also demonstrated strong reactivity with the Dsg1 extracellular domain when tested using live-cell immunofluorescence. In contrast, sera from normal human controls and sera from bullous pemphigoid patients did not react with the Dsg1 extracellular domain. These data indicate that both pemphigus foliaceus and pemphigus vulgaris sera react with conformationally sensitive epitopes in the amino-terminal region of Dsg1.

Patients with the autoimmune blistering disease pemphigus foliaceus (PF) have circulating autoantibodies directed against the desmosomal cadherin desmoglein 1 (Dsg1). Based on the fact that purified IgG fractions from PF patients induce loss of cell adhesion in organ culture and in a neonatal mouse model, it has been proposed that these anti-Dsg1 antibodies play a pathogenic role in blister formation. To directly address whether antibodies in PF sera specific for the Dsg1 extracellular domain are indeed pathogenic in the disease, PFIg, a chimeric protein containing the entire extracellular domain of human Dsg1 and the constant region of human IgG1, was produced by baculovirus expression. Incubation of PF patients' sera with the PFIg baculoprotein removed the immunoreactivity of autoantibodies against keratinocyte cell surfaces in all 20 PF and eight Brazilian PF patients' sera tested. This adsorption was conformation dependent, because PFIg protein denatured by low pH or heat was no longer able to adsorb the immunoreactivity of PF sera. Furthermore, the incubation with the PFIg baculoprotein eliminated the pathogenic activity of PF patients' sera and prevented gross blister formation in a neonatal mouse model of pemphigus. Anti-Dsg1 antibodies eluted from the PFIg protein column were pathogenic as they resulted in the appearance of gross blisters in neonatal mice with typical histologic findings of PF. These observations indicate that the extracellular domain of Dsg1 expressed by baculovirus is capable of specifically immunoadsorbing pathogenic autoantibodies from PF patients' sera and provide direct evidence that the anti-Dsg1 autoantibodies in PF sera are indeed pathogenic. The availability of this Dsg1 recombinant protein may facilitate the development of antigen-specific plasmapheresis as a novel therapeutic strategy in pemphigus.

We investigated the antigen molecules for six clinically typical cases of paraneoplastic pemphigus (PNP) using immunofluorescence, immunoprecipitation, and immunoblotting. All the PNP sera showed a clear reactivity with transitional epithelia of rat urinary bladder and immunoprecipitated the 250-kD, 230-kD, 210-kD, 190-kD, and 170-kD proteins in various combinations, confirming the diagnosis of PNP. Immunoblot analysis demonstrated slightly different reactivity from that of immunoprecipitation. With immunoblotting of normal human epidermal extract, bovine desmosome preparation, and extract of cultured squamous cell carcinoma cells, all the PNP sera reacted with a characteristic doublet of the 210-kD and 190-kD proteins. However, immunoblotting detected the 250-kD desmoplakin I and the 230-kD bullous pemphigoid antigen less frequently and did not detect the 170-kD protein. Further immunoblot studies indicated that the 210-kD protein is different from desmoplakin II and that the 190-kD protein is most frequently detected by PNP sera. Two of the six PNP sera specifically reacted with the extracellular domain of recombinant pemphigus vulgaris antigen protein, indicating that pemphigus vulgaris antigen may be involved in PNP. In future studies to unravel the complex mechanisms of the PNP antigens, the immunoblot technique may be a useful tool.

It is still unclear what kinds of mechanisms are involved in blister formation after antibodies bind to the antigens in pemphigus and bullous pemphigoid. The effects of IgGs from pemphigus vulgaris, pemphigus foliaceus, and bullous pemphigoid sera on intracellular calcium concentration ([Ca++]i) and inositol 1,4,5-trisphosphate were examined in a human squamous cell carcinoma cell line (DJM-1 cells) and in cultured human keratinocytes to clarify whether signal transduction via calcium is involved. IgGs were purified with protein A affinity column from the sera of five pemphigus vulgaris patients, three pemphigus foliaceus patients, eight bullous pemphigoid patients, and 14 normal volunteers. Keratinocytes were cultured in Eagle's minimum essential medium containing 1.8 mM Ca++ and loaded with fura-2/AM, followed by addition of the IgGs. Subsequently, [Ca++]i was determined by measuring the fluorescence ratio (F340/F360) with videomicroscopy. Pemphigus IgGs (seven of eight cases) induced a rapid and transient increase in [Ca++]i in both the cells, whereas a [Ca++]i increase was caused by very few IgGs from bullous pemphigoid (one of eight cases) and normal sera (two of 14 cases). The pemphigus IgG-induced transient [Ca++]i increase was not affected by chelating extracellular Ca++ with ethyleneglycol-bis(beta-aminoethyl ether)-N,N,N',N'-tetracetic acid. In addition, monoclonal antibodies acid. In addition, monoclonal antibodies against 180-kD and 230-kD antigens did not exert this change. Pemphigus IgGs that caused a [Ca++]i increase induced rapid and transient production of inositol 1,4,5-trisphosphate, peaking at 20 seconds. These findings suggest that IgG from pemphigus induces Ca++ mobilization by inositol 1,4,5-trisphosphate from internal stores, and that mechanisms of antibody-transmitted signaling in pemphigus may differ from those in bullous pemphigoid.

The major antibody binding regions of desmoglein 1 (Dsg1) in pemphigus foliaceus and pemphigus vulgaris were examined using cDNA-encoded fusion proteins combining glutathione S-transferase with various domains of bovine Dsg1, namely, the extracellular regions EC1-2, EC3-5, EC1-5, and the entire intracellular region IC. In immunoblot analyses using these fusion proteins, EC1-2, as well as EC1-5, which comprises EC1-2, were recognized by 50% of the sporadic pemphigus foliaceus sera and 45% of Brazilian pemphigus foliaceus sera that reacted with Dsg1 in immunoblotting of bovine desmosome preparations. None of these fusion proteins reacted with any sera of pemphigus vulgaris. None of these sera showed reactivity with EC3-5. In contrast, the IC domain was recognized by 91% of pemphigus vulgaris sera reactive with Dsg1 in bovine desmosome preparations, and by certain pemphigus foliaceus and Brazilian pemphigus foliaceus sera. These results indicate that major epitopes of Dsg1 recognized by pemphigus foliaceus and Brazilian pemphigus foliaceus sera are located in the extracellular amino-terminal domain EC1-2, and that sera of the Dsg1-positive pemphigus vulgaris contain antibodies against the intracellular domain, which may not play a pathogenic role. Possible reasons for this selectivity of antigen binding site are discussed.

It has previously been demonstrated that sera from endemic and nonendemic pemphigus foliaceus patients recognize three immunoreactive fragments of 80, 62, and 45 kilodaltons (kD) from extracts of the envelope fraction of human and bovine epidermis. These polypeptides are also immunoprecipitated by approximately 50% of pemphigus vulgaris sera, but are unreactive with sera from bullous pemphigoid patients or normal controls. The 80-kDa antigen has been shown to be a glycoprotein with N-linked oligosaccharides. Complete removal of the carbohydrate moieties produced a 76-kD polypeptide that continued to react with pemphigus foliaceus autoantibodies in a Ca(++)-dependent manner. To further characterize this antigen/antibody system, the 80-kD pemphigus foliaceus antigen solubilized from a bovine epidermal envelope extract was purified by affinity chromatography using a pemphigus foliaceus patient's immunoglobulin (Ig)G immobilized on agarose. After elution with 0.2 M glycine/HCl, pH 2.8, 5 mM ethylene diaminetetraacetic acid, the polypeptide was mixed with a small amount of 125I-labeled 80-kD antigen, added as a tracer, fractionated by sodium dodecylsulfate-polyacrylamide gel electrophoresis, and electrotransferred onto a polyvinylidene difluoride membrane. The 80-kD band detected by amido black staining and autoradiography was excised and characterized by amino acid sequence analysis. The resulting sequence, EXIKFAAAXREGEXNSKRNPIA, matched perfectly with the N-terminal 22 amino acids of the mature form of bovine desmoglein 1. These findings demonstrate that the 80-kD bovine autoantibody-reactive polypeptide is the glycosylated ectodomain of desmoglein 1, which may contain epitopes recognized by pathogenic autoantibodies.

We recently reported the production of a human monoclonal antibody (MoAb) derived from a patient with pemphigus vulgaris (PV) that binds to the keratinocyte membrane and reacts with a 185-kD polypeptide by immunoblot analysis. We have since examined the tissue specificity of that MoAb, F12. By indirect immunofluorescence (IIF), F12 stained both the cell membrane and the basement membrane zone of stratified squamous epithelia. Moreover, MoAb F12 stained other epithelial tissues, such as urinary bladder, small bowel, thymus, and liver, and non-epithelial tissues, such as myocardium. Indirect immunoelectron microscopy (IIEM) analysis showed that MoAb F12 bound to a component common to desmosomal and hemidesmosomal plaques and to zona adherens-type junctions between hepatocytes and bile duct cells. Inhibition experiments were then performed with sera from patients with pemphigus vulgaris, pemphigus foliaceus, paraneoplastic pemphigus, or bullous pemphigoid. Three sera blocked F12 reactivity; two were from paraneoplastic pemphigus patients and the other was from the pemphigus vulgaris patient whose peripheral blood lymphocytes were used to make F12. All these sera recognized a 185-kD band that co-migrated with the polypeptide labeled by MoAb F12 on immunoblots. In addition, the IIF and IIEM staining patterns of MoAb F12 were similar to those observed with sera from two patients with paraneoplastic pemphigus. These observations suggest a relationship between MoAb F12 and the autoimmune response characterizing paraneoplastic pemphigus patients' sera.

Recently, it has been shown that desmoglein, pemphigus foliaceus target antigen, and a 130-kD pemphigus vulgaris antigen belong to the cadherin family of cell adhesion molecules. We tried to determine whether desmocollins I/II, other cadherin-like transmembranous glycoproteins present in desmosomes, are also recognized by pemphigus autoantibodies of the IgG class. We examined 16 pemphigus vulgaris sera, 15 pemphigus foliaceus sera, 15 Brazilian pemphigus foliaceus sera, five bullous pemphigoid sera, and 65 normal sera. Four (25%) pemphigus vulgaris sera, one (7%) pemphigus foliaceus serum, eight (53%) Brazilian pemphigus foliaceus sera, and three (5%) normal sera reacted with desmocollins I/II on immunoblots of bovine desmosome preparation. The affinity-purified desmocollins I/II pemphigus autoantibodies were shown to bind the epidermal cell surface by indirect immunofluorescence. Immunoblot analysis revealed one pemphigus vulgaris serum, one Brazilian pemphigus foliaceus serum, and one normal serum recognizing a recombinant protein produced by a desmocollin cDNA clone. Moreover, immunoblot analysis of reactivity of a Brazilian pemphigus foliaceus serum with recombinant proteins produced by deletion mutants of the desmocollin cDNA clone showed that the extracellular portion of desmocollin is immunogenic in this pemphigus patient. We conclude that desmocollins I/II are recognized by certain sera from patients with various types of pemphigus, particularly Brazilian pemphigus foliaceus. However, the significance of this reactivity remains to be defined.

The indirect immunofluorescent (IF) reactivity of the pemphigus antibodies in sera of 21 cases of pemphigus vulgaris (PV), 15 cases of pemphigus foliaceus (PF), and 14 cases of Brazilian PF (BPF) was compared on 2 substrates, notably monkey esophagus (ME) sections and guinea pig esophagus (GPE) sections. The IF reactions of the pemphigus antibodies of PV could be distinguished from those of PF or BPF by differences in their reactivity on ME and GPE sections in 98% of the cases examined in this study. In most cases, the pemphigus antibodies of PV cases gave higher titers and stronger IF staining reactions on ME sections, while those of PF and BPF cases gave stronger reactions on GPE sections. In addition, most (13 of 21) PV sera react with the lowest 3-4 cell layers of ME sections, while most (13 of 15) PF sera failed to do so but did react with the upper layers of the sections. Importantly, in 8 of the 50 cases examined by IF, the choice of substrate affected the detectability of the pemphigus antibodies, i.e., 4 of 15 PF and 2 of 14 BPF sera reacted only with GPE and 2 of 21 PV sera reacted only on ME. These research findings point to the need for an evaluation of the combined use of ME and GPE in routine diagnostic studies of pemphigus antibodies.

We have identified a novel IgG antikeratin autoantibody in the serum of a Brazilian pemphigus foliaceus patient (Cascas-42). This antibody is specific for the 59 kD acidic murine keratin and its 56.5 kD human counterpart (Moll's catalogue #10), and is distinct from the pemphigus antibody system. Antikeratin autoantibodies present in the Cascas-42 serum were purified by affinity chromatography with a 59 kD murine keratin-agarose column (IAP-Cascas-42 antibodies). The specificity of the IAP-Cascas-42 antibodies was tested by indirect immunofluorescence and immunoelectron microscopy against epidermal cryosections, trypsin-dissociated keratinocytes, and epidermal cell cultures. The serum was also tested with extracts from unlabeled and surface 125I-labeled keratinocytes (Iodo-Gen method) by immunoblot analysis of one- and two-dimensional polyacrylamide gel electrophoresis. The IAP-Cascas-42 antibodies bind the intercellular spaces of murine epidermis, and the cell surfaces of viable, dissociated murine keratinocytes, as well as murine epidermal cells in culture by immunofluorescence and immunoelectron microscopy. These autoantibodies did not stain cytoplasmic keratins and did not react with parallel human epidermal substrates. The Cascas-42 serum identified the 59 kD murine acidic keratin and its 56.5 kD human counterpart in epidermal extracts by two-dimensional polyacrylamide gel electrophoresis and immunoblot analysis. In addition, surface radioiodination of viable murine keratinocytes selectively labeled the 59 kD keratin suggesting that a domain of this molecule is exposed on the cell surface. The 125I-labeled 59 kD keratin was also recognized by the Cascas-42 serum by immunoblotting and autoradiography. These studies suggest that in murine epidermis, the 59 kD keratin is a transmembrane protein with an extracellular domain recognized by the IAP-Cascas-42 antibodies.