Originally defined in the Drosophila homolog, Armadillo, PG is also called γ[gamma]-catenin and is encoded by a gene in chromosome 17 [28]. Its counterpart is β[beta]-catenin, expressed in adherens junctions. Similar to β[beta]-catenin, PG is characterized by 12 armadillo (arm) repeats, which are ~42 amino acid domains flanked by distinct N- and C- terminal domains, which are less structured than the central region and have been suggested to affect PG function [29, 30]. The central arm region of PG shares 65 % amino acid identity with β[beta]-catenin that associates with adherens junctions. PG can substitute β[beta]-catenin in adherens junctions because both bind E-cadherin with similar affinity. However, PG has higher affinity for desmogleins which would explain the exclusion of β[beta]-catenin from desmosomes [31]. The arm domain of PG associates with both the intracellular domain of desmosomal cadherins and the N-terminus of desmoplakin [32–34]. Thus, PG is a critical linker in desmosomal adhesion.

PKP are members of another subfamily of armadillo proteins. Three desmosomal PKPs (1, 2, and 3) have been described, and the genes encoding them are located on chromosomes 1, 12, and 11, respectively [35]. PKP1 and PKP2 have 2 isoforms (“a” and “b”) that result from alternative splicing [36, 37]. There is a fourth plakophilin: PKP4, also known as p0071, which is highly related to p120ctn and δ[delta]-catenin [38]. PKPs (1–4) have been shown to bind directly to the intracellular domain of desmosomal cadherins. This interaction is mediated by their amino-terminal head domain, and the functions for their armadillo repeats remain unknown. PKPs can also bind PG in order to facilitate clustering of desmosomal cadherins through lateral stabilizing interactions, which increases desmosome strength [32].

Accumulated evidence from several epitope mapping studies indicates that pathogenic pemphigus autoantibodies target the amino-terminal end of Dsg1 and/or Dsg3 ectodomains [65–67]. Data based on the crystal structure of classical cadherins suggest that this N-terminal region harbors the adhesive interface of desmosomal cadherins [21, 68]. In addition, the pathogenicity of pemphigus IgG autoantibodies has been consistently demonstrated by passive transfer studies in neonatal mice since the 1980s [52, 69]. Moreover, not only the whole IgG molecule but also the F(ab)2 and Fab fragments were found to be pathogenic, independently of complement or plasminogen activator [70–73], suggesting that due to their lack of ability to cross-link cell-surface molecules, it is possible that they interfere directly with adhesion. Furthermore, monoclonal antibodies (AK23) derived from a PV mouse model that binds the functionally N-terminal adhesive interface of Dsg3 induced pemphigus vulgaris lesions in mice, whereas monoclonal antibodies recognizing other regions of Dsg3 ectodomain did not cause lesions in mice [74]. Using single-molecule atomic force microscopy (AFM), it has been shown that PV-IgG and AK23 monoclonal antibody directly inhibit Dsg3 homophilic binding under cell-free conditions, suggesting that direct inhibition of Dsg3 binding occurs in PV [75, 76]. Direct blocking of Dsg1 binding was not observed by this technique (AFM); however, keratinocyte dissociation and loss of Dsg1- and Dsg3-coated microspheres to cultured keratinocytes were observed when using laser tweezer trapping [77], indicating that acantholysis may not be solely dependent on direct interference of Dsg1-Dsg1 binding by pemphigus autoantibodies. Recent studies using peptides against the desmoglein adhesive interface as well as tandem peptides (obtained by dimerization of two of the initial peptides) added evidence that direct inhibition contributes to acantholysis when the tandem peptide prevented acantholysis induced by PV-IgG, yet this is not the case when using PF-IgG. Thus, some investigators proposed that PV and PF acantholysis may involve different mechanisms [78].

The direct interference of desmoglein trans interactions by pemphigus autoantibodies has been shown to be insufficient to disrupt keratinocyte adhesion [24]; thus, additional cellular events may be needed to cause blistering. Previous in vitro studies show that PV-IgG induces different events in cultured keratinocytes including a transient increase in intracellular calcium and inositol 1,4,5-triphosphate [79], activation of protein kinase C, and phosphorylation of Dsg-3, which may lead to internalization of Dsg3 from cell surface, therefore depleting Dsg3 from desmosomes [80–85]. Thus, activation of intracellular signaling within the target keratinocyte induced by binding of pemphigus IgG has been proposed to contribute to the loss of cell-cell adhesion. Previous studies have shown that phosphorylation of p38MAPK and HSP25 (the murine homolog of human HSP27) occurs rapidly after exposure of keratinocytes to pemphigus IgG and in the skin of mice treated with pathogenic IgG [86]. Also, phosphorylation of both p38MAPK and HSP27 has been observed in the perilesional epidermis of pemphigus patients [87]. Furthermore, p38MAPK inhibitors block both histological and gross blister formation in the PF passive transfer model [86], suggesting that activation of p38MAPK is an early and key step in PF-IgG-induced acantholysis. p38MAPK signaling has been implicated in other cellular responses such as desmosome assembly, cytoskeleton reorganization, changes of the cell cycle, and apoptosis [88, 89]. Moreover, there is evidence that p38MAPK is involved in keratinocyte apoptosis [90, 91] and that DNA fragmentation and caspase activation are induced in the epidermis of PF-IgG-treated mice [92]. Keratinocyte-derived and local production of apoptotic inducers such as nitric oxide synthase, Fas, and inhibitor Bcl-2 have also been detected in lesional skin of PF patients [93, 94] along with increased levels of Fas ligand in serum of pemphigus patients [95]. Furthermore, a biphasic activation of p38MAPK after the binding of pemphigus IgG has been recently demonstrated where the first activation peak is linked to acantholysis and the second peak coincided with apoptosis, suggesting that apoptosis occurs downstream to acantholysis in pemphigus [90]. However, other studies suggest that apoptosis occurs before acantholysis develops [92]. There is also the hypothesis that apoptotic signaling could precede acantholysis in the absence of apoptotic cell death [89]. More important, caspase inhibitors have been shown to block pemphigus serum-induced keratinocyte apoptosis. Thus, at present, it is not clear which process precedes the other, but there is evidence that both are involved in pemphigus pathogenesis. Another signaling pathway suggested to be involved is through plakoglobin [96]. The observation was made that keratinocytes from plakoglobin-deficient mice were resistant to keratinocyte dissociation induced by PV-IgG, suggesting also that direct inhibition of Dsg binding may not be sufficient to cause acantholysis and that plakoglobin could be part of a complex responsible for transferring the signal upon autoantibody binding from outside into the keratinocyte: “outside-in” signaling [97]. In addition, it has been shown that plakoglobin is involved in c-Myc repression, and c-Myc was also shown to be elevated in keratinocytes exposed to pemphigus autoantibodies [98, 99]. Nonetheless, the role of c-Myc signaling in pemphigus acantholysis remains unclear.