Fig. 14.1
Pemphigus, p38, acantholysis, and apoptosis. (a) Model for the temporal relationship of pemphigus IgG-mediated activation of p38MAPK to blistering and apoptosis. Two peaks of p38 phosphorylation occur subsequent to treatment with pemphigus IgG. Inhibiting the first, but not second, peak of p38 activity blocks blistering. Markers of apoptosis, including caspase-3 cleavage, PARP cleavage, and TUNEL-positive staining, occur subsequent to the second peak of p38MAPK phosphorylation. Blocking the first peak of p38 phosphorylation blocks blistering, indicating a role for the first peak of p38MAPK activation in loss of cell-cell adhesion. In contrast, blocking this second peak of p38 phosphorylation fails to block blistering but blocks increases in apoptotic markers (e.g., caspase-3 cleavage). The second peak of p38MAPK phosphorylation is not part of the mechanism of acantholysis but may represent stress response signaling secondary to acantholysis. (b, c) Molecular model for signaling in pemphigus. (b) Pemphigus IgG binds to dsg and biases the equilibrium of desmosome assembly/disassembly toward disassembly which is linked by an, as yet, undefined mechanism toward activation of p38MAPK. Subsequent p38-dependent alterations in the cell state include RhoA inactivation, dsg endocytosis, HSP27 phosphorylation, keratin intermediate filament retraction, actin, and loss of cell-cell adhesion (acantholysis). (c) A second late peak of p38 activity is observed that is likely a stress response signal induced by loss of cell-cell adhesion and leads to activation of proapoptotic pathways including caspase-3 activation
Data in support of signaling was first provided in a series of experiments undertaken in the early 1990s by Professor Yasuo Kitajima’s research group in Japan. Culturing keratinocytes in the presence of PV sera led to a variety of intracellular events that suggested activation of transmembrane signaling. For example, in 1995, they reported that pemphigus IgG stimulated rapid, transient increases in keratinocyte intracellular calcium and inositol trisphosphate [13]. Subsequent studies by this group implicated signaling by phospholipase C [14] and protein kinase C [15, 28] in pemphigus IgG-mediated blistering.
Changes in cell-cell adhesion often result in changes in cell biology. Reasoning that such changes would likely be transmitted across the desmosome-adhesive junctions themselves, analogous to what had been observed for integrin cell-substrate adhesion molecules and for adherens junctions, it seemed reasonable to hypothesize that changes in desmosome-mediated adhesion similarly triggered cellular signaling events. Toward this goal, a biochemical screen was initially designed to explore the consequences of disrupting desmosome adhesion, using PV IgG as a desmosome-specific reagent because it targets dsg [20]. The immunologic response in mucosal PV is characterized by antibodies to dsg3, whereas mucocutaneous PV is characterized by antibodies to dsg1 as well as dsg3. In these initial series of experiments, IgG purified from mucosal PV patient sera was used, allowing signaling downstream of IgG binding to dsg3 to be investigated. Subsequent series of experiments extended these investigations to dsg1-specific IgG and to IgG purified from mucocutaneous sera, containing dsg1- and dgs3-specific IgG.
Phosphorylation of serine, threonine, and tyrosine residues of proteins is a common regulatory post-translational modification that is central to many signaling pathways. In this initial screen, radioactive phosphate was used because it provided both a sensitive and quantifiable method for detecting signaling in keratinocytes treated with PV IgG. Primary human keratinocytes were grown in culture media supplemented with 32P H3PO4 and treated with IgG fractions purified from PV patient sera. Cells were incubated with different concentrations of purified PV IgG at various time points, and keratinocyte extracts were separated by two-dimensional electrophoresis and analyzed by autoradiography and phosphoimage analysis. Using this approach, several spot patterns, corresponding to phosphoproteins, were identified on two-dimensional gels in which the distribution of charge isoforms and intensity of 32P incorporation were altered when keratinocytes were treated with PV IgG. These spots were cut from the gels, subjected to in situ tryptic digestion, and then analyzed by time-of-flight mass spectroscopy. This led to the identification of the small heat shock protein 27 (HSP27) as a protein whose phosphorylation was altered in cells treated with PV IgG. Using antibodies to both HSP27 and its phosphorylated isoforms, the identity of HSP 27 was confirmed as keratinocytes treated with PV IgG demonstrated increased amounts of phospho-HSP27 when assayed by immunoblot of cell extracts using HSP27 and phospho-HSP27-specific antibodies [20].
HSP27 is a small-molecular-weight chaperone that oligomerizes into large multimeric structures to facilitate protein refolding when cells are physiologically stressed, for example, when cells are exposed to the thermal stress of a “heat shock.” In addition to this role as a chaperone, HSP27 functions in signaling, where it is downstream of p38 mitogen-activated protein kinase (p38MAPK, p38). Activation of p38 leads to p38-mediated phosphorylation and activation of mitogen-activated protein kinase 2 (MAPKAP2, MK2), which in turn phosphorylates HSP27 [29]. Phosphorylation of HSP27 results in dissociation of large oligomeric HSP27 into smaller multimeric structures [29, 30]. These smaller phosphorylated HSP27 tetramers have been suggested to function in signal transduction and to regulate the cytoskeleton including both actin and intermediate filaments [31–39]. Thus, the observation that PV IgG induced the rapid phosphorylation of HSP27, occurring within minutes of exposure of keratinocytes to PV IgG, led to several predictions that could be tested experimentally. Namely, (1) treatment of keratinocytes with PV IgG should lead to phosphorylation of p38 and dissociation of HSP27 from large oligomers to smaller tetramers, and (2) inhibition of p38 should prevent the phosphorylation of HSP27 and block its dissociation from large oligomers to the smaller tetrameric structures. When tested experimentally, this in fact was the case [20].
During the process of acantholysis, as keratinocytes lose adhesion to adjacent cells, keratin intermediate filaments and actin undergo structural reorganization [17]. In addition to its effect on HSP27, p38 may also directly phosphorylate and regulate keratin intermediate filament structure. The role of HSP27 in regulating the cytoskeleton suggested that activation of p38-MK2-HSP27 signaling could have a functional role in mediating PV IgG-induced changes in the target keratinocytes that contributed to the loss of cell-cell adhesion. A prediction of this hypothesis is that inhibitors of p38 would block PV IgG-induced keratin intermediate filament and actin reorganization in PV IgG-treated cell cultures. Consistent with this hypothesis, p38 inhibitors blocked PV IgG-induced retraction of keratin intermediate filaments as well as actin reorganization [20]. Collectively, these observations suggested that activation of this signaling pathway was not a consequence of PV IgG-induced loss of adhesion, but part of the mechanism by which PV IgG induced loss of adhesion.
If activation of p38 contributed to the mechanism by which PV IgG induced loss of adhesion, then inhibition of p38 should block blistering. This hypothesis was then tested using the PV passive transfer mouse model [40]. Wild-type mice were treated with PV IgG or normal human IgG as a control. Using both immunofluorescence microscopy of skin biopsies as well as immunoblot of skin extracts, increased phosphorylation of p38 and HSP27 was observed in epidermal keratinocytes from PV IgG-treated mice. Either of the two active p38 inhibitors, SB202190 or SB203580, but not the structurally similar inactive analog SB20470, blocked blistering in PV IgG-treated mice. When analyzed by direct immunofluorescence, PV IgG was observed bound to the keratinocytes of PV IgG-treated murine skin in the inhibitor-treated mice indicating that the p38 inhibitors did not block binding of PV IgG to the skin, but inhibited the ability of PV IgG to induce acantholysis through signaling occurring downstream of autoantibody binding. Thus, the p38 inhibitors prevented PV IgG-induced blistering in vivo by blocking mechanistic events within the target keratinocytes that contribute to loss of cell-cell adhesion. Subsequent studies demonstrated that p38 and HSP27 are also activated in PF and that p38 inhibitors similarly block blistering in the PF passive transfer mouse model [41]. Importantly, human PV and PF skin biopsies also show increased phosphorylation of p38 and HSP27 compared to normal controls, indicating that the activation of these signaling molecules first observed in pemphigus tissue culture and animal models also occurs in the human disease [42]. Activation of p38 in pemphigus has subsequently been observed by numerous other investigators [43–47].
A number of events have been shown to be downstream of p38 activity in the cascade of events that contribute to loss of cell-cell adhesion. For example, PV autoantibody binding to cell surface dsg3 results in depletion of dsg3 from the keratinocyte cell membrane [48]. Kowalczyk and co-workers have demonstrated that dsg3 is endocytosed, exiting from the cell membrane in a clathrin- and dynamin-independent process [49, 50]. Detailed cell biologic experiments by this group have demonstrated co-localization of internalized dsg3 into early endosomes. Inhibition of p38 blocks PV IgG-induced dsg3 endocytosis indicating that PV IgG-induced depletion of dsg3 is also mediated downstream of p38 activity [51]. Similarly, PV IgG induces structural alterations to both the keratin intermediate filament and actin cytoskeletons, and these are also blocked by p38 inhibition [20, 52]. In part, this may be mediated by HSP27; however, Waschke and co-workers have provided additional evidence implicating RhoA regulation of actin in PV IgG-mediated loss of adhesion and that RhoA activity is also downstream of p38 [47]. Another second messenger that regulates events downstream of PV IgG binding to keratinocytes is cyclic adenosine monophosphate (cAMP) [53]. Intracellular cAMP levels are increased by treatment of keratinocytes with PV IgG. Increased cAMP appears to be in part mediated via protein kinase A (PKA) activity. The increase in cAMP appears to be a protective compensatory mechanism as increased cAMP partially attenuates the acantholytic effects of PV IgG. Pharmacologic elevation of cAMP by forskolin or by the β (beta)-receptor agonist isoproterenol blocked loss of intercellular adhesion, depletion of cellular dsg3, and pemphigus IgG-induced cytoskeletal changes in cultured keratinocytes. cAMP may be working through p38 since elevating cAMP levels blocked PV IgG increased p38 activity in both in vitro and in vivo models.
Collectively, these observations provided the mechanistic rationale to explore the use of p38 inhibitors for treating patients and led to the design and implementation of a multicenter open-label clinical trial of a proprietary small-molecular-weight oral p38 inhibitor for the treatment of PV [54]. This non-placebo-controlled trial was successful in enrolling 15 patients with PV. A variety of assessment tools were utilized to measure disease activity including the recently developed PDAI [55, 56]. Outcome measures included decreased disease activity as well as the ability to taper systemic corticosteroids during the 12-week course of the trial. Unfortunately, systemic toxicities and off-target effects of the investigational drug used in this trial limited its dosing, and the trial was terminated prior to completion. Targeting p38 for PV will require the development of a new generation of safer and more specific inhibitors.
The presence or absence of adhesion requires a major shift in the biology of a cell. For example, proliferating cells come in contact and form adhesive junctions with one another; they typically will stop proliferating. This biological principle is known as contact inhibition and is a regulatory feature typically lost during malignant transformation. Analogously, depending on the context, the loss of cell adhesion may result in cells undergoing apoptosis, or alternatively, in the context of development or wound healing, loss of adhesion could trigger migration and proliferation. Therefore, it should not be surprising to find a complex interaction between structures that regulate and mediate cell adhesion with cellular systems that regulate such basic processes as cell growth, migration, and apoptosis. Consistent with this, additional keratinocyte signaling proteins have been identified in pemphigus. For example, plakoglobin, a component of both adherens junctions and desmosomes, appears to have an essential role in pemphigus acantholysis. Mueller’s group has shown that cells lacking plakoglobin are resistant to the acantholytic effects of PV IgG [17, 57]. Additionally, they have suggested a role for c-myc. Other signaling proteins that have been suggested to contribute to the mechanism of acantholysis include:
14.3 Apoptosis
Apoptosis is a type of cell death characterized at the morphological level by chromatin condensation, nuclear fragmentation, and plasma membrane blebbing. The apoptotic machinery centers on the activity of various members of the caspase family. The intrinsic apoptotic pathway is mostly triggered by UV radiation and is mediated in mitochondria via the release of cytochrome c that complexes with apoptotic protease activating factor (Apaf-1) that recruits and activates the initiator caspase-9. Caspase-9 in turn activates caspases 3/7 leading to DNA fragmentation. The extrinsic apoptotic pathway is triggered by death ligand binding to death receptors. For example, Fas ligand (FasL) assembles with the adaptor protein FADD in a complex known as DISC that in turn recruits the initiator caspase-8, with subsequent activation of the effector caspases 3/7 followed by DNA fragmentation [66].
Several lines of evidence indicate that apoptosis is involved in the pathomechanisms of pemphigus [67]. The first observations of apoptotic mechanisms operating in pemphigus were reported about 10 years ago and mostly overlooked [68]. Recently, a genome-wide study using single-nucleotide polymorphism (SNP) in a subpopulation of pemphigus patients has revealed a significant association of a genomic segment on chromosome 8q11.23 that spans the ST18 gene, with pemphigus [69]. Pemphigus is more common in certain ethnic groups, such as Ashkenazi Jews and those of Mediterranean origin. In particular, it is between 3-fold and 45-fold more prevalent in patients of Jewish descent, as compared with other populations [70]. Sarig and co-workers have nicely shown a significant association of ST18 gene in a Jewish and an Egyptian cohort of pemphigus patients, but not in a German group, suggesting that ST18 gene may predispose to pemphigus in a population-specific manner. Interestingly, the ST18 gene regulates apoptosis [71] and is overexpressed in non-lesional epidermis of patients with pemphigus [69]. This pattern of expression well correlates with previous observations demonstrating TUNEL-positive cells in perilesional pemphigus skin [68, 72], which has been interpreted as indicating that keratinocyte apoptosis occurs before cell-to-cell detachment (acantholysis). More recent work demonstrates the presence of TUNEL-positive cells both in the floor and in the roof of the blisters [73]. Interestingly, TUNEL reactivity that co-localized with nuclear fragmentation was detected in close vicinity to intraepidermal blisters by Schmidt and co-workers who consider apoptosis in pemphigus as a secondary phenomenon [74, 75].
Nevertheless, more specific apoptotic markers are detected in lesional pemphigus skin before acantholysis [76] (CP, personal observations), and PV IgG induces keratinocyte apoptosis in vitro [77] and cell detachment by enrichment in caspase 8 and activation of caspase 3 [78]. On the other hand, caspase inhibitors prevent PV IgG-induced acantholysis in vitro [79]. The critical role of caspases in the pathogenesis of pemphigus is further confirmed by the use of caspase 3/7 as well as by pan-caspase inhibitors that protect mice from developing blisters induced in the passive transfer pemphigus mouse model by PF IgG [80]. More recently, Pacheco-Tovar and co-workers demonstrated that PV IgG induces apoptosis in perilesional keratinocytes before acantholysis, while the pan-caspase inhibitor prevents blister formation [81]. In addition to autoantibodies, pemphigus serum contains non-IgG substances that can cause a sharp reduction of keratinocyte viability and weaken intercellular adhesion strength [82]. The serum factors implicated in pemphigus pathomechanisms include tumor necrosis factor (TNF) alpha [83], nitric oxide [84], kallikreins [85], and antimitochondrial antibodies [46]. The latter factors are pathogenic and induce the release of cytochrome c, suggesting the involvement of the intrinsic apoptotic pathway in pemphigus. This is consistent with the recent observation that IVIg, an effective treatment in pemphigus, upregulates serum level of the inhibitor-of-apoptosis proteins (IAPs) that in turn block the caspase 9/mitochondrial pathway [86]. Yet, a body of experimental evidence points to the Fas/FasL system as a key pathway in the pathogenesis of pemphigus [87]. Indeed, FasL and Fas are significantly elevated in sera from pemphigus patients [68, 88]. Original immunohistochemical studies revealed Fas receptor on keratinocyte membranes of pemphigus lesional epidermis [89]. FasL contained in pemphigus sera induces apoptosis in normal human keratinocytes, via caspase-8 activation [68]. In addition, PV IgG treatment induces an mRNA upregulation of proapoptotic molecules in keratinocytes, including FasL [90], and the secretion from cells of soluble FasL. This is not just a mere upregulation, as PV IgG treatment induces a co-aggregation of FasL and Fas receptor with caspase-8 with consequent DISC formation [78]. FasL acts synergistically with PV IgG and TNF alpha in the induction of acantholysis in an organ culture model of pemphigus [91]. Neutralization of FasL prevents caspase-8 activation and decreases the number of apoptotic cells after treatment of keratinocytes with pemphigus sera [68]. Anti-FasL neutralizing antibody protects dsgs from caspase-induced cleavage, thus preventing PV IgG-induced acantholysis. Moreover, anti-FasL inhibits the PV IgG-induced activation of caspase 8 [67].
In addition to the above observations supporting a role for apoptosis in the pathogenesis of blistering in pemphigus, time course studies in both in vitro and in vivo pemphigus models suggest that apoptosis may not be a primary event, but rather a secondary event that may contribute to blistering by sensitizing cells to the acantholytic effects of pemphigus IgG [52]. In pemphigus tissue culture and mouse models, two peaks of p38 activation have been reported, (1) an early peak that occurs within minutes of exposure to PV IgG and is activated by binding of PV IgG to dsg and (2) a second late peak occurring hours after exposure to PV IgG that represents stress response signaling secondary to PV IgG-induced adhesive changes. Administration of p38 inhibitors before pemphigus IgG injection blocked both the early and late peaks of p38 phosphorylation and blister formation; however, administration of the inhibitor after the first peak, but before the second peak of p38 activity, blocked only the later peak of p38 activation and failed to block blistering. Examination of the temporal relationship of p38 phosphorylation and apoptosis showed that apoptosis occurs at or after the second peak of p38 activation. The time course of p38 activation and apoptotic markers as well as the ability of inhibitors of p38 to block activation of the proapoptotic proteinase caspase-3 suggest that activation of apoptosis is downstream to, and a consequence of, p38 activation in pemphigus acantholysis. Furthermore, these observations suggest that the earlier peak of p38 activation is part of the mechanism leading to acantholysis; whereas, the later peak of p38 and apoptosis may not be essential for acantholysis.
The observation that the activation of proapoptotic pathways is a late event and may not be essential for blistering in pemphigus does not exclude the possibility that activation of components of apoptotic signaling, including caspase family member proteinases, could augment the blistering response as downstream effects of p38 activation. This hypothesis is supported by the observations that (1) in staphylococcal scalded skin syndrome (SSSS), direct proteolysis of the dsg1 ectodomain induces loss of cell-cell adhesion [92] and (2) caspase inhibitors block pemphigus IgG-induced acantholysis in the passive transfer mouse model [80]. Components of the desmosome including dsg3, dsg1, plakoglobin, and desmoplakin [93, 94], as well as intermediate filaments [95, 96], have all been shown to undergo caspase-dependent cleavage suggesting that caspase-dependent proteolysis may disrupt these structures, weakening cell adhesion and augmenting the pathogenic response in pemphigus. Although induction of apoptosis may be a secondary response in the mechanism of acantholysis, blocking caspase-dependent dsg degradation may augment cell-cell adhesion and decrease keratinocyte sensitivity to the acantholytic effects of pathogenic pemphigus IgG.
As we have reviewed, several hypotheses have been proposed for the mechanism by which pemphigus IgG induced loss of adhesion, including steric hindrance, activation of proteolysis, and signaling. These different hypotheses are not mutually exclusive, and data generated over the past two decades supports a mechanism in which all may contribute to acantholysis. For example, PV IgG likely acts as competitive inhibitors for endogenous homophilic and/or heterophilic dsg binding at the EC1/EC2 extracellular domains. The disruption of endogenous dsg molecular interactions (e.g., steric hindrance) activates a variety of intracellular signaling events, including p38 and HSP27 phosphorylation that lead to dsg endocytosis and reorganization of the keratin intermediate filament and actin cytoskeletons (e.g., signaling). As a result of changes in the cellular adhesive structures, some keratinocytes undergo apoptosis, activating and releasing proteolytic enzymes including caspases and metalloproteinases that have the potential to digest dsg and other components of the desmosome adhesion complex on adjacent cells, further weakening and/or disrupting cell-cell adhesion. Importantly, targeting many of these components can antagonize the acantholytic effects of PV IgG providing multiple potential druggable targets for blocking blistering in this life-threatening autoimmune disease.
Acknowledgments
This work was supported in part by National Institutes of Health Grant RO1 AI49427 (to DSR).
References
2.
3.
4.
5.
6.
Amagai M, Klaus-Kovtun V, Stanley JR. Autoantibodies against a novel epithelial cadherin in pemphigus vulgaris, a disease of cell adhesion. Cell. 1991;67(5):869–77.CrossRefPubMed
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
