The mechanisms of autoantibody-induced epidermal detachment (acantholysis) in PF/FS remain under intense investigation in different laboratories around the world. The sera of patients with all clinical variants of pemphigus, like that of patients with other autoimmune diseases, contain autoantibodies against multiple self-antigens. Some of the best characterized antigens are expressed on keratinocyte cell-cell adhesion organelles such as the desmosomes, i.e. Dsg1, Dsg3, Dsg4, desmocollins [3
] or interdesmosomal domains of the keratinocyte membrane, i.e. E-cadherin [95
]. In PV, PF and FS, anti-Dsg1 and anti-Dsg3 autoantibodies are pathogenic and serve as diagnostic markers of disease [1
]. Anti-Dsg3 autoantibodies are markers of the mucosal variant of PV; whereas, the mucocutaneous variant of PV is characterized by a mixed population of anti-Dsg1 and anti-Dsg3 autoantibodies [96
]. The pathogenic relevance of other autoantibodies detected in PV, PF or FS sera remains unknown. Given our current understanding, we must conclude that pemphigus is a group of diseases in which desmosomal antigens are targeted by pathogenic and non-pathogenic autoantibodies. The resulting disease phenotypes are epidermal-specific (PF/FS), mucosal-specific (mPV), or epidermal and mucosal specific (mcPV). Only mucosal surfaces that possess squamous epithelium are involved in mPV and mcPV.
Desmosomes mediate cell-cell adhesion and can be found not only in skin, but also in cardiomyocytes, brain arachnoidal cells and lymph node dendritic reticulum cells. Desmosomes are cellular organelles that structurally are formed by a central core and two intracellular plaques. While the core is composed of transmembrane glycoproteins (the desmosomal cadherins), the intracellular plaque contains large molecular weight protein members of the plakin family, i.e. desmoplakins, plakoglobin, and plakophillin [97
]. The desmosomal cadherins comprise Dsg1, Dsg2, Dsg3 and desmocollins. Dsg1 and Dsg3 are restricted to stratified squamous epithelia, while Dsg2 is expressed in all tissues expressing desmosomes. Similar to E cadherin, the extracellular domain of all desmosomal cadherins, except Dsg1, show 5 domains which participate in interactions with desmogleins on opposing cells, thereby establishing adhesion [98
]. Through linker proteins, the intracellular (or cytoplasmic) domains of E cadherin, desmogleins and desmocollins are associated with the keratinocyte cytoskeleton.
The distribution and density of Dsg1 and Dsg3 varies within the layers of the stratified epidermis and mucosal epidermis. As the desmogleins are the target of autoantibodies in pemphigus, their differential distribution is thought to be responsible for the different histological sites of blister formation in PF and PV. This concept is known as the compensation theory [99
]. For example, Dsg1 is expressed throughout the epidermis, but its expression is increased in the superficial epidermis. Dsg3 is expressed in an inverse pattern with increased expression in the basal layers of the epidermis. In mucosal epithelium, Dsg1 is expressed in the superficial squamous epithelium, while Dsg3 is expressed throughout. It is hypothesized that in PF, the anti-Dsg1 autoantibodies target Dsg1, leading to blister formation in the superficial epidermis as Dsg3 expression in basal cells compensates for the loss of Dsg1 mediated adhesion. The mucosa is protected by expression of Dsg3 throughout all layers, and therefore PF patients do not exhibit mucosal involvement. In PV, where the target is Dsg3, blister formation is suprabasilar as superficial epidermal adhesion is maintained by Dsg1. Mucosal involvement is common as there is no expression of Dsg1 in the basal layers of mucosal epithelium to compensate for the loss of Dsg3 function [99
In addition to the distribution of Dsg within the epithelia, i.e. compensation theory, we propose that the epithelial damage in all forms of pemphigus is modulated by the epitope-specificity of the anti-desmoglein autoantibodies. As previously shown, healthy individuals and FS patients during the preclinical stages of FS recognize epitopes on Dsg1 or Dsg3 that are not pathogenic. Once an individual mounts an autoimmune response against epitopes located on the EC1 and EC2 domain of Dsg1 or Dsg3 the autoantibodies lead to pathogenic damage of the epithelia. Importantly, these pathogenic autoantibodies in FS are IgG4 restricted. It is possible that the availability of pathogenically relevant epitopes on Dsg1 and Dsg3 in vivo to react with PV and PF/FS autoantibodies would determine the level of tissue damage. The histological phenotype of pemphigus therefore, would be determined by the fine specificity of the anti-Dsg1 or anti-Dsg3 autoantibodies as well as the distribution of the respective desmosomal antigens in the epithelia.
To understand the complexity of the humoral autoimmune response in all forms of pemphigus we hypothesize that the autoantibody response against “unique epitopes” on Dsg1 and Dsg3 in humans is associated with a spectrum of clinical phenotypes (). They include (i) normal individuals who exhibit normal skin and have non-pathogenic anti-Dsg1/Dsg3 autoantibodies, (ii) PV patients with pure mucosal disease and pathogenic anti-Dsg3 autoantibodies (oral PV), (iii) PV patients exhibiting combined mucosal and cutaneous disease with pathogenic anti-Dsg3 and anti-Dsg1 autoantibodies (mucocutaneous PV). The subset of pemphigus showing a pure cutaneous blistering eruption and having pathogenic anti-Dsg1 autoantibodies includes PF/FS patients. Importantly, our studies on FS in Brazil are allowing us to investigate the transition of individuals from a preclinical state (possessing non-pathogenic anti-Dsg1 autoantibodies) to individuals with classic FS, exhibiting pathogenic anti-Dsg1 autoantibodies. Moreover, we have recently described normal individuals possessing non-pathogenic anti-Dsg3 autoantibodies and a subset of patients with an endemic form of mucocutaneous PV [101
]. We believe that the clinical phenotypes of all forms of pemphigus may result, aside from the Dsg1/Dsg3 distribution in the target tissues, from the epitope specificity of the autoantibodies harbored in normal individuals and patients, and this in turn, from the immunogenetic make up of the host.
Diversity of anti-desmoglein autoantibodies present in normal individuals and patients exhibiting the different disease phenotypes of pemphigus.
It has been proposed that anti-desmosomal antibodies induce acantholysis by steric hindrance, where autoantibody binding directly blocks Dsg adhesive interactions on opposing cells [102
]. Steric hindrance is supported by the compensation theory as well as by findings in staphylococcal scalded skin syndrome, in which an exfoliative toxin cleaves Dsg1 within the extracellular domains and leads to an identical histological phenotype as PF [103
]. In addition, Dsg3 knockout mice display blister formation as seen in PV [104
]. Thus, it seems that impairing the function of these desmogleins at the extracellular level leads to histological features resembling acantholysis triggered by binding of anti-Dsg1 or anti-Dsg3 autoantibodies to their target antigen. The detachment process is complement-independent and can be triggered by Fab fragments of the pathogenic autoantibodies [105
]. Waschke et al [108
], utilizing single molecule-based micromechanical laser tweezers and atomic force microscopy, have probed Dsg1 and Dsg3 ectodomain transinteractions. They have suggested that PV anti-Dsg3 autoantibodies impair the adhesive function of Dsg3 and thus trigger acantholysis; whereas, PF autoantibodies trigger acantholysis by decreasing the amounts of Dsg1 available on the surface of keratinocytes secondary to intracellular signaling. These studies need to be extended and reproduced by other laboratories. The fate of the anti-Dsg1 or anti-Dsg3 autoantibodies following the initial binding to the ectodomain of the antigens has been studied at the ultrastructural level by Patel el al. in 1984 [109
] and recently by Delva et al. [110
] in cell culture. They demonstrated that PV IgG molecules are internalized via a clathrin and dynamin-independent pathway to form endocytic vesicles that fuse with lysosomes. This process is associated with widening of the intercellular spaces. Pharmacological inhibition of the internalization process may provide new therapeutic approaches for pemphigus acantholysis. In addition to these findings, other investigators [111
] have demonstrated that following the binding of PV IgG to keratinocytes surfaces, Dsg3 is depleted from desmosomes, which in turn may be relevant in triggering acantholysis.
Several studies have identified signaling capabilities of the desmogleins, which may also play a role in autoantibody induced acantholysis. It was initially observed that keratinocyte cell lines treated with PV sera show a transient increase in intracellular calcium and inositol 1,4,5-triphosphate, supporting the idea that autoantibody binding to desmogleins initiates intracellular signaling events [113
]. These investigators extended these studies by demonstrating that PV IgG induces phosphorylation of Dsg3 and subsequent dissociation from plakoglobin in cultured primary keratinocytes, thus triggering acantholysis [114
]. This conclusion was reasonable since it was well known that phosphorylation of the classical cadherins at serine and tyrosine residues, regulates cell adhesion [115
]. It has been proposed that the plakoglobin linker protein plays a crucial role in the signaling process mediated by desmosomes and consequently in pemphigus acantholysis. Caldelari et al. [117
] showed that keratinocytes from plakoglobin knockout mice were unresponsive to the pathogenic effects of PV IgG on cell adhesion compared to keratinocytes from wild type mice. They hypothesize that the plakoglobin/c-Myc proto-oncogene axis may be relevant to the pathogenesis of pemphigus acantholysis [118
Members of our own group have shown that primary human keratinocytes treated with PV IgG show a time and dose-dependent increase in levels of phosphorylated p38MAPK and heat shock protein HSP27 [119
]. Both p38MAPK and HSP27 are known to be important in regulating cytoskeletal components such as actin and keratin intermediate filaments. Further studies have shown that inhibitors of MAPK signaling not only blocked phosphorylation of HSP27 following PV IgG stimulation of cultured primary keratinocytes, but also prevented keratin filament retraction, an early change in the cytoskeleton associated with acantholysis [119
]. These findings were recapitulated in the PV passive transfer murine model [120
]. Subsequent studies have identified increased phosphorylation of p38MAPK and HSP27 in skin biopsies of patients with PV and PF compared to that of controls [121
]. Collectively, these data have identified an important signaling pathway in the pathogenesis of PV and PF. Studies are underway to test whether p38MAPK inhibitors might be useful in treating clinical disease.
In addition to the signaling pathways described above, in vitro cell culture studies have shown that exposure of keratinocytes to PV IgG induce apoptosis in these cells [122
]. The pro-apoptotic changes are evident by various measurements, including annexin V binding, Hoechst 33342 staining, TUNEL labeling, DNA laddering, oligonucleosome formation, caspase activation, up-regulation of pro-apoptotic proteins (Fas, FasL, Bax, p53), and down-regulation of anti-apoptotic proteins such as Bcl-2 and FLIP-l. It is speculated that keratinocyte apoptosis results in pemphigus acantholysis [122
While much in vitro data has been published on the induction of apoptosis by PV IgG, little is known about the possible apoptotic effect of PF IgG. Our group [128
] has evaluated the role of the biochemical response of apoptosis in PF using the IgG passive transfer mouse model of the disease [2
]. We found TUNEL positive epidermal cells and increased cytosolic oligonucleosomes in epidermal cells of the diseased mice. A time course study revealed that TUNEL-positive epidermal cells appear prior to intraepidermal blisters. Western blot analysis showed that the pro-apoptotic factor Bax was upregulated at the earlier time points (2 and 4 h) while the anti-apoptotic factor Bcl-xl was downregulated at the later time points (6, 8, and 20 h) post PF IgG injection. Correspondingly, the active forms of caspase-3 and -6 were detected at the later time period (6, 8, and 20 h). Administration of Ac-DEVD-cmk, a peptide-based caspase-3 inhibitor, protected mice from developing intraepidermal blisters and clinical disease induced by PF IgG. The same protective effect was also observed for a broad-spectrum caspase inhibitor Bok-D-fmk. Collectively, the findings of this study suggest that some biochemical events of apoptosis are provoked in epidermal cells by PF autoantibodies and caspase-3 activation may contribute to acantholytic process and disease pathogenesis.
This observation may suggest that activation of caspase leads to epidermal cell injury and dysfunction, which in turn is manifested as intraepidermal blisters. Acantholysis may occur before or without the end-point of epidermal cell death. We hypothesize that activated caspase-3 is involved in the process of acantholysis and/or intraepidermal blister formation, not through epidermal cell death, but through the proteolytic cleavage of a structural protein(s) involved in epidermal cell-cell adhesion. Possible candidates may include components of the desmosome, adherens junction, and cytoskeleton. Many of cell adhesion molecules are caspase substrates [129
] and are degraded during apoptosis, such as Dsg1 [130
], Dsg3, plakoglobin, plakophillin, plakin proteins [132
], E-cadherin [135
], and β-catenin [137
]. Interestingly, it has been shown that shedding of ectodomain of Dsg1 or Dsg3 during the process of apoptosis is inhibited by caspase inhibitors [130
]. Further studies are required to identify the potential targets for caspase-3 that may be responsible for PF blistering.
Impairment of Dsg1/Dsg3 adhesive function by autoantibody induced steric hindrance or by binding of autoantibodies triggering signaling and apoptosis that lead to acantholysis are not mutually exclusive (). Continued research is aimed at further delineating the exact role of these multistep processes. Certainly, it is an exciting area of investigation and likely to yield potential novel treatment options.
Figure 5 Molecular mechanisms of acantholysis in pemphigus. Desmoglein 1 and 3 are linked to the intermediate filament keratin network of the keratinocyte whereas E-cadherin links to the actin cytoskeleton. Pathogenic autoantibodies in PV and PF bind the ectodomain (more ...)