- Email: [email protected]
Removal of amino-terminal extracellular domains of desmoglein 1 by staphylococcal exfoliative toxin is sufficient to initiate epidermal blister formation
Journal of Dermatological Science 59 (2010) 184–191
Contents lists available at ScienceDirect
Journal of Dermatological Science journal homepage: www.elsevier.com/jds
Removal of amino-terminal extracellular domains of desmoglein 1 by staphylococcal exfoliative toxin is sufficient to initiate epidermal blister formation Koji Nishifuji a,b, Atsushi Shimizu a, Akira Ishiko a, Toshiroh Iwasaki b, Masayuki Amagai a,* a b
Department of Dermatology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan Department of Veterinary Medicine, Tokyo University of Agriculture and Technology, Fuchu, Tokyo, Japan
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 June 2010 Received in revised form 16 July 2010 Accepted 22 July 2010
Background: In both bullous impetigo and staphylococcal scalded-skin syndrome (SSSS), exfoliative toxins (ETs) produced by Staphylococcus aureus cause superficial intraepidermal blisters. ETs are known to cleave specifically a single peptide bond in the extracellular domains 3 and 4 of desmoglein (Dsg) 1. However, the precise mechanisms underlying ET-induced epidermal blister formation remain poorly understood. Objective: To determine whether cleavage of Dsg1 by an ET is sufficient to induce blister formation in vivo or if the subsequent internalization of cleaved Dsg1 or other desmosomal components is required. Methods: Skin samples obtained from neonatal mice injected with ETA were analyzed by time-lapse immunofluorescence and transmission electron microscopy for desmosomal components. Results: Epidermal blister formation was observed as early as 60 min after ETA treatment. At this time, the amino-terminal extracellular domains of Dsg1 disappeared from the surface of keratinocytes, while the cleaved carboxy-terminal domain of Dsg1 (Dsg1-C) as well as the extracellular domains of desmocollin 1 (Dsc1-N) remained on the cell surface. Half-split desmosomes with intracytoplasmic dense plaques and attached tonofilaments were recognized ultrastructurally on the split surface of keratinocytes at 60 min. Subsequent to this, Dsg1-C and Dsc1-N gradually disappeared from the surface layer of keratinocytes. Conclusion: Our findings suggest that the removal of amino-terminal extracellular domains of Dsg1 by ETs is sufficient to initiate epidermal blister formation in bullous impetigo and SSSS. ß 2010 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Desmoglein Desmosome Exfoliative toxin Impetigo Staphylococcal scalded-skin syndrome Staphylococcus
1. Introduction Bullous impetigo and its generalized form, staphylococcal scalded-skin syndrome (SSSS), are highly contagious, bacterial skin diseases. Under these conditions, infection by Staphylococcus aureus causes blisters in the superficial epidermis that are characterized histopathologically as intraepidermal clefts with acantholysis [1–3]. The virulent strains of staphylococci isolated from patients with bullous impetigo or SSSS produce exfoliative toxins (ETs). These are serine protease-like proteins capable of producing intraepidermal blisters with a histopathology identical to bullous impetigo or SSSS when injected into neonatal mice [4–6]. Our recent studies revealed that the two major ETs, ETA and ETB, are serine proteases that hydrolyze specifically only one peptide bond in a calcium-binding
Abbreviations: ET, exfoliative toxin; Dsg, desmoglein; Dsc, desmocollin; SSSS, staphylococcal scalded-skin syndrome. * Corresponding author. Tel.: +81 3 5363 3823; fax: +81 3 3351 6880. E-mail address: [email protected] (M. Amagai).
motif between the extracellular domains 3 and 4 of desmoglein (Dsg) 1, a desmosomal transmembrane constituent that belongs to a cadherin family of cell–cell adhesion molecules [7–9]. Because the amino-terminus of Dsg1 is thought to harbor its binding domain, cleavage of the extracellular region of Dsg1 by an ET may cause loss of its ability to bind [2]. Additionally, ETD (a novel isoform of ET produced by S. aureus originally isolated from patients with deep skin infections) cleaves the same peptide bond in Dsg1 and causes superficial epidermal blisters in neonatal mice [8,10]. Thus, ETs are unique serine proteases that hydrolyze only one peptide bond in a single protein, and are responsible for the formation of intraepidermal blisters in bullous impetigo and SSSS. Despite recent advances in our understanding about the enzymatic characteristics of ETs, the precise mechanisms by which ETs disrupt cell–cell adhesion of keratinocytes and cause acantholysis in vivo remain unclear. In the epidermis, two major desmosomal cadherin-type adhesion molecules, Dsgs and desmocollins (Dscs), are expressed in a differentiation-specific manner and function to maintain cohesion between adjacent keratinocytes [11]. Among the isoforms of desmosomal cadherins, Dsg1 and Dsc1
0923-1811/$36.00 ß 2010 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jdermsci.2010.07.010
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
are expressed predominantly in the spinous and granular layers of epidermis, whereas Dsg3 and Dsc3 are seen in the basal and immediate suprabasal epidermis. Dsg2 and Dsc2 are expressed in the basal epidermis, while Dsg4 is expressed predominantly in anagen hair follicles and just below the cornified layer of the epidermis [12]. While ETs are known to cleave Dsg1, little is known about the fate of desmosomal cadherins during separation of keratinocytes within the superficial epidermis. The aim of this study was to investigate the fate of desmosomal cadherins, especially cleaved Dsg1 and Dsc1, in vivo in ETA-treated mouse epidermis. 2. Materials and methods 2.1. Production and purification of recombinant ETA Recombinant ETA, fused with a His-tag at its carboxyl terminus, was produced in Escherichia coli strain DH10B and purified as described previously [9]. ETA was harvested from the soluble fraction in lysis buffer [20 mM Tris–HCl (pH 8.0), 0.2 M NaCl, 0.2% Triton X-100, and protease inhibitor cocktail (Complete Mini, EDTA-free; Roche Diagnostics, Basel, Switzerland)], purified using the TALON metal affinity resin system (BD Biosciences Clontech, Mountain View, CA) and dialyzed against phosphate-buffered saline (PBS). Protein concentrations were determined by a protein assay (Bio-Rad Laboratories Inc., Richmond, CA). Purified ETA was passed through a 0.22-mm filter (Millipore Corp., Bedford, MA), aliquoted, and stored at 80 8C until required.
185
gel II mini; Daiichi Pure Chemicals, Tokyo, Japan) and transferred to polyvinylidene fluoride membranes (Immobilon–P Transfer Membrane; Millipore Corp.). Intact and cleaved mouse Dsg1 isoforms were detected using an anti-E-tag mAb (1:4000; GE Healthcare, Little Chalfont, Buckinghamshire, UK). 2.5. Neonatal mouse study Neonatal Jcl:ICR mice (<24 h old, weighing 1.6–2.0 g; CLEA Japan, Inc., Tokyo, Japan) were injected subcutaneously with purified ETA (10 mg) in the dorsal neck and thorax. Skin samples were collected from the injection site, in which injected ETA might distribute directly to the epidermis, at 0, 60, 120, and 180 min after ETA injection. Formaldehyde-fixed, paraffin-embedded skin samples were sectioned, stained with hematoxylin and eosin (H&E), and subjected to histopathological analysis. Skin samples were also subjected to immunofluorescence and transmission electron microscopic analyses. Regarding the dose of ETA, in this study we used 10 mg/mouse, which is sufficient to induce blisters alone in upper epidermis of neonatal mice. In some of our previous study, we used 1 mg/mouse, which is not sufficient to cause blisters alone, to analyze the pathogenic activity of anti-Dsg3 monoclonal antibodies [19,20]. 2.6. Immunofluorescence Skin samples obtained from ETA-treated mice were embedded in [(Fig._1)TD$IG] O.C.T. compound (Tissue-Tek; Sakura Finetechnical, Tokyo,
2.2. Production of recombinant mouse Dsg Recombinant, secreted-form extracellular domains of mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4, fused with E- and His-tags on their carboxyl terminus, were produced using the baculovirus expression system as described previously [13–15]. Culture supernatants of High FiveTM insect cells (Invitrogen Corp., Carlsbad, CA) containing recombinant mouse Dsg were collected and stored at 70 8C after removal of cell debris by centrifugation. 2.3. Antibodies Sera from two patients with pemphigus foliaceus (PF) contained IgG that recognizes the extracellular domains of mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4, as determined by immunoprecipitation– immunoblotting using baculovirus recombinant mouse Dsgs. To remove the Dsg4-reacting fraction of IgG, PF sera were applied to a chromatography column (Econo-Column; Bio-Rad Laboratories Inc.) packed with TALON metal affinity resin, along with recombinant mouse Dsg4. Pass-through fractions from the chromatography column were collected and used for immunofluorescence analysis. Serum from a patient with subcorneal pustular dermatosis-type IgA pemphigus contained IgA that recognizes the extracellular domains of human Dsc1, expressed on the surface of COS-7 cells [16]. A mouse monoclonal antibody (mAb; DG3.10) specific for the cytoplasmic domain of three mouse Dsg1 isoforms and Dsg2 (PROGEN, Heidelberg, Germany) [17,18] and a mouse anti-Dsg3 mAb (AK18) that recognizes the extracellular domain of Dsg3 [19] were used for immunofluorescence analysis. 2.4. In vitro digestion of mouse Dsg1 isoforms by ETA Insect culture supernatants containing baculovirus recombinant mouse Dsg1a, Dsg1b, and Dsg1g were incubated in vitro with purified recombinant ETA (1 mg) at 37 8C for 1 h. Samples were extracted with Laemmli sample buffer at 100 8C for 5 min, separated by 10% SDS-polyacrylamide gel electrophoresis (Multi-
Fig. 1. Reactivity of ETA and human pemphigus foliaceus (PF) sera with mouse Dsg1 isoforms. (A) ETA cleaves the extracellular domains of mouse Dsg1a and Dsg1b, but not that of Dsg1g. Arrowhead and arrow indicate intact and cleaved mouse Dsg1 isoforms, respectively. (B) Sera from two patients with PF (PF #1 and #2) that recognize mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4 are applied to a column packed with recombinant mouse Dsg4. Immunoreactivities of pre- and post-column fractions were confirmed by immunoprecipitation and immunoblotting. Note that anti-Dsg4 IgG antibodies are completely absent in both post-column fractions. Bars in the right panels indicate molecular weight standards.
[(Fig._2)TD$IG]
186
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Fig. 2. Microscopic blisters developed subsequently to disappearance of the amino-terminal extracellular domains of Dsg1 (Dsg1-N) from keratinocyte surfaces in ETAtreated mice. (A) Time-lapse histopathological and immunofluorescence analysis for Dsg1-N, the intracellular domain of Dsg1 (Dsg1-C), and the extracellular domains of Dsc1 (Dsc1-N) using the skin of ETA-treated mice. Initial separation of keratinocytes began 60 min after ETA injection and extended to intraepidermal blisters over time. Immunofluorescence of Dsg1-N disappeared from keratinocyte surfaces as early as 60 min after ETA injection, while the carboxyl-termini of Dsg1 (Dsg1-C) and Dsc1-N remained on blister cavity-facing keratinocyte surfaces (arrowheads) until 120 min after ETA injection. Dsg1-C and Dsc1-N disappear from the surface of free-floating keratinocytes 180 min after ETA injection. Dashed lines indicate basement membrane zones (BMZ). Bar represents 50 mm. (B) Half-split desmosomes with attached tonofilaments are observable on the surface of acantholytic keratinocytes during early ETA-induced blister formation. Neonatal mouse skin was subjected to transmission electron microscopic analysis 60 min after ETA injection. (a) Intraepidermal separation of keratinocytes in granular layers. Bar represents 2 mm. (b), (d) Separated keratinocytes in the roof (b) and floor (d) of blister cavities retain half-desmosomes (arrows). Bars represent 1 mm in (b) and 2 mm in (d). (c), (e)
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Japan). Non-fixed cryosections (5 mm) were blocked [1% bovine serum albumin (Sigma–Aldrich, St. Louis, MO) and 5% normal goat serum (DAKO, Glostup, Denmark)] in Tris–buffered saline (50 mM; pH7.4) at room temperature for 30 min. For counterstaining the extracellular domains of Dsg1 (Dsg1-N), the intracellular domain of Dsg1 (Dsg1-C), and the extracellular domains of Dsc1 (Dsc1-N), samples were incubated with PF sera (1:100 dilution) and an IgA pemphigus serum (1:40 dilution) at 4 8C overnight and washed in PBS; the secondary antibody (mouse mAb against human IgA, 1:100; Zymed Laboratories, San Francisco, CA) was then added at room temperature for 1 h. All samples were then incubated with a tertiary antibody [a mixture of Alexa Fluor 648 goat antihuman IgG (Invitrogen) and Alexa Fluor 488 anti-mouse IgG] and a quaternary antibody [DG3.10 labeled using the Zenon Alexa Fluor 546 mouse IgG1 labeling kit (1:40)]. For counterstaining Dsg1-C, desmoplakins (DPs), and plakoglobins (PGs), samples were incubated with DG3.10 labeled with a Zenon Alexa Fluor 488 mouse IgG1 labeling kit, 11-5F labeled with a Zenon Alexa Fluor 546 IgG1 labeling kit (1:40), and 11E4 conjugated with an Alexa Fluor 647 conjugation kit (Invitrogen). Following these reactions, samples were fixed in 4% (v/v) paraformaldehyde at room temperature for 10 min and mounted in Mowiol (Calbiochem-Novabiochem Corp., La Jolla, CA). Specimens were examined and images captured and processed under a confocal laser-scanning microscope (LSM510 META; Carl Zeiss Microimaging GmbH, Jena, Germany) using an Argon- or He–Ne laser. Fluorescence intensities of Dsg1-N, Dsg1-C, and Dsc1-N in mouse keratinocytes were quantified using ImageJ 1.42q software (http://rsb.info.nih.gov/ij) as follows. Fifty-four diagonal lines were drawn arbitrarily on 18 keratinocytes in six histopathological sections of the skin collected from two mice at each time point. The distance between two cell membranes along the diagonal line was divided into six areas. Pixel intensity of immunofluorescence along the line was quantified and the mean intensity in each area calculated. Statistical analysis using Dunnett’s test was performed to compare the intensities between area #1 and the other areas examined. Moreover, the percent fluorescence of Dsg1-C and Dsc1N in the cell surface layer to the entire keratinocyte was calculated using the following formula: Percent fluorescence in the surface layer = (summary of the means of intensities in areas #1 and #6)/(summary of the means of intensities from area #1 to area #6) 100. Fluctuation in percent fluorescence was analyzed for Dsg1-N, Dsg1-C, and Dsc1-N. Statistical analysis using Dunnett’s test was performed to compare fluorescence percentages before and after ETA injection. Significance level of the statistical analyses used in this study was defined as P < 0.05. 2.7. Transmission electron microscopy Skin samples obtained from ETA-treated mice were fixed in 2% glutaraldehyde and 1% osmium tetroxide, dehydrated by exposure to increasing ethanol concentrations, and then embedded in Epon 812 resin (Okensyoji, Tokyo, Japan). Resin was polymerized at 55 8C for 72 h. Ultrathin sections, stained with uranyl acetate and lead citrate, were examined under a JEOL-1200 EX transmission electron microscope (JEOL, Tokyo, Japan).
187
of neonatal mice treated with ETA. Neonatal ICR mice were injected subcutaneously with purified ETA [9]; skin samples were collected along the time course and subjected to histopathological as well as immunofluorescence analyses. ETA has been reported to degrade Dsg1a and Dsg1b, but not Dsg1g, in mouse epidermis [17], although the three Dsg1 isoforms are similarly expressed in mouse epidermis [17,18]. We confirmed that ETA selectively cleaves extracellular domains of the mouse Dsg1a, Dsg1b, but not that of Dsg1g, by in vitro digestion of the extracellular domains of recombinant mouse Dsg1 isoforms with ETA (Fig. 1A). For immunofluorescence of the amino-terminal extracellular domains of three Dsg1 isoforms (Dsg1-N), we used the pass-through fractions from PF serum run on a Dsg4 column. These recognize the extracellular domains of Dsg1a, Dsg1b, and Dsg1g, but not that of Dsg4 (Fig. 1B). 3.2. Amino-terminal, but not carboxy-terminal, of Dsg1 disappeared from cell surfaces during the initial phase of blister formation Epidermal blister formation was observed as early as 60 min after ETA injection and progressed with time (Fig. 2A). Dsg1-N disappeared almost totally from the keratinocyte surface at 60 min after ETA injection (Fig. 2A). In contrast, the intracellular carboxyl-termini of Dsg1 (Dsg1-C) and the extracellular domains of Dsc1 (Dsc1-N) remained on the surface of keratinocytes, including those facing toward the blister cavity, until 120 min after ETA injection (Fig. 2A). In some keratinocytes, blister cavity surface-facing Dsg1-C fluorescence intensity was lower than that of cell–cell contact sites. Dsg1-C and Dsc1-N disappeared from the surface of keratinocytes floating freely within the blister cavity 180 min after ETA injection (Fig. 2A). The time courses of Dsg1-N, Dsg1-C, and Dsc1-N staining were essentially identical in six independent samples from two mice injected with ETA. The time courses of DP, PG, and cytoplasmic desmosome component staining were similar and each was detectable on cell surfaces as late as 120 min after ETA injection (data not shown). To determine desmosome structural integrity during the early phase of ETA-mediated keratinocyte separation, skin samples were subjected to transmission electron microscopic analysis. Because we wanted to analyze desmosomes immediately after keratinocyte separation, we used skin samples collected 60 min after injection with ETA (Fig. 2B). At low magnification, intraepidermal separation of keratinocytes in the granular layers of the epidermis was observed. Acantholytic keratinocytes retained half-split desmosomes on their acantholytic surfaces on both the upper and lower surfaces of intraepidermal splits. At higher magnifications, these half-split desmosomes revealed a flocculent material, which may represent desmoglea, on their extracellular surfaces, together with a dense intracytoplasmic plaque with attached tonofilaments. These findings indicate that blister cavity-facing half-split desmosomes are present during the initial phase of ETA-induced blister formation. In contrast, the plasma membranes of separated acantholytic keratinocytes lost their flatness and showed cytoplasmic projections 180 min after ETA treatment (data not shown). Thus, the fate of amino- and carboxy-terminal Dsg1 domains after ETA injection differs, and carboxy-terminal domain of Dsg1 and amino-terminal domain of Dsc1 remained on cell surfaces during the initial phase of ETA-mediated blister formation.
3. Results 3.1. Specific cleavage of mouse Dsg1a, Dsg1b, but not Dsg1g, by ETA
3.3. Quantitative analysis of Dsg1 and Dsc1 during early ETA-induced blister formation
To investigate the mechanisms of ET-induced intraepidermal blisters in vivo, we first analyzed the fate of Dsg1 and Dsc1 in the skin
Immunofluorescence analysis showed that Dsg1-N, but not Dsg1-C or Dsc1-N, disappeared from blister-facing cell surfaces
Half-split desmosomes (double arrowheads) with flocculent material on their extracellular surface and intracytoplasmic dense plaques with attached tonofilaments. Bars represent 200 nm.
[(Fig._3)TD$IG]188
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Fig. 3. Quantitative analysis of the fluorescence intensities of Dsg1-N, Dsg1-C, and Dsc1-N. (A) Diagonal lines were drawn arbitrarily and keratinocytes divided into six areas. In non-blistered keratinocytes, areas #1 and #6 included plasma membranes. In blistered keratinocytes, area #1 included plasma membranes facing to the blister cavity, whereas area #6 included those in cell–cell contact sites. Asterisks indicate blister cavity. (B) Comparison of the fluorescent intensities of Dsg1-N, Dsg1-C, and Dsc1-N in six areas. Graph shows the mean SD of pixel intensities of 54 diagonal lines. Blisters began to develop 60 min after ETA injection. Note that the Dsg1-N fluorescence intensity in areas #1 and #6 (including plasma membranes) decreased to similar levels to those detected in areas #2 to #5 (cytoplasmic layers) 60–120 min after ETA injection. Cytoplasmic Dsg1-N
[(Fig._4)TD$IG]
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
189
Fig. 4. Schematic diagram of the fate of desmosomal constituents during ETA-induced blister formation. Intact Dsg1 and Dsc1 mediate cell–cell keratinocyte adhesion prior to ETA treatment (0 min). Blister formation began 60 min after ETA treatment. At this time, Dsg1-N was absent from the surface of keratinocytes, while Dsg1-C and Dsc1-N remained detectable. Dsg1-C and Dsc1-N disappeared from keratinocyte surfaces at 120 and 180 min, respectively, concurrent with blister progression.
during the early phase of ETA-induced blister formation. To analyze the fate of desmosomal cadherins in keratinocytes in a quantitative manner, we calculated pixel intensities of these molecules in six areas of the diagonal lines connecting two plasma membranes (Fig. 3A). In this method, areas #1 and #6 always included plasma membranes of keratinocytes, whereas areas #2 to #5 always included the cytoplasm. In keratinocytes facing the blister cavity, area #1 always included the cell membrane facing the intraepidermal split, while area #6 always included cell–cell contact sites. We found that Dsg1-N fluorescence intensities in
cell-surface layers (areas #1 and #6) decreased to the same level as those in the cytoplasm (areas #2 to #5) as early as 60 min, concurrent with initial blister development (Fig. 3B). Dsg1-C and Dsc1-N intensities were higher in blister cavity-facing cell surface layers (area #1) than in cytoplasmic layers (areas #2 to #5) up to 180 min after ETA injection (Dunnett’s test, P < 0.05, Fig. 3B). No increase in Dsg1-C and Dsc1-N fluorescence intensities in keratinocyte cytoplasm was detected, probably because internalized Dsg1-C and Dsc1-N were removed from tissue sections during immunofluorescence procedures. No significant differences in the
fluorescence intensity decreased at 180 min. Conversely, fluorescence intensities of Dsg1-C and Dsc1-N were consistently higher in keratinocyte surface than in cytoplasmic layers up to 180 min (Dunnett’s test, P < 0.05). (C) Fluctuation in the percent fluorescence of Dsg1-N, Dsg1-C, and Dsc1-N in keratinocyte surface layers. The Dsg1-N percentage 60 min after ETA injection had decreased, while that of Dsg1-C and Dsc1-N remained constant. Data are expressed as means SD.
190
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
fluorescence intensities of Dsg1-C and Dsc1-N between blister cavity-facing cell surfaces (area #1) and cell–cell contact sites (area #6) were detected (P > 0.05; Fig. 3B). Moreover, we assessed fluctuations in the percent fluorescence of Dsg1-N, Dsg1-C, and Dsc1-N in keratinocyte cell-surface layers (Fig. 3C). We found that the Dsg1-N percentage decreased markedly 60 min after injection, concurrent with initial blister formation (Dunnett’s test, P < 0.05), while that of Dsc1-N was not significantly different (P > 0.05). The Dsg1-C fluorescence percentage was slightly lower at time 60 min (P < 0.05). The percentages of Dsg1-C and Dsc1-N decreased gradually as blister formation progressed, up to 180 min after ETA injection. Taken together, these data suggest that Dsg1-N disappeared from the cell surface layer of keratinocytes during the early phase of ETA-induced blister formation. In contrast, the majority of Dsg1-C and Dsc1-N remained on the cell surface layer during the early phase but gradually disappeared during the later phase of ETA-induced blister formation (see schematic drawing in Fig. 4). 4. Discussion In this study, we demonstrated that ETA-induced intraepidermal blisters form subsequent to disappearance of Dsg1-N from the surface of epidermal keratinocytes in vivo. Additionally, our findings suggest that the majority of Dsg1-C and Dsc1-N remained on the blister cavity-facing surface of keratinocytes during the early phase of ETA-induced blister formation. Ultrastructural analysis also revealed split half-desmosomes with intracytoplasmic dense plaques and attached tonofilaments during the early phase of the blister formation, suggesting structural integrity. During the late phase of ETA-induced blister formation, desmosomal cadherins disappeared from the surface layer of acantholytic keratinocytes. Taken together, these findings support our hypothesis that ETA-mediated disappearance of amino-terminal extracellular domains of Dsg1 alone is sufficient to induce initial separation of keratinocytes. Meanwhile, our data do not rule out the possibility that separation of keratinocytes is enhanced by subsequent signaling cascades that include internalization of desmosomal constituents. While the effect of ETB or ETD on the fate of desmosomal cadherins has not been addressed in the present study, it is speculated that these ETs affect desmosomes as does ETA, because all three ETs share identical enzymatic activity to cleave a single peptide bond in Dsg1 efficiently [8,21]. Previous in vitro studies using cell lines expressing ectopic Dsg and/or Dsc revealed that Dsg and Dsc form both homophilic [22] and heterophilic [22,23] complexes. Both Dsg and Dsc are thought to play important roles in cell–cell adhesion, since functional blocking of either one of the molecules causes disruption of cell–cell adhesion both in vitro [24–26] and in vivo [7,9,10,27,28]. In this study, we demonstrated that Dsc1-N remains detectable on blister cavity-facing keratinocyte surfaces in ETA-treated mice. These findings suggest that either the adhesive function of Dsc1 is insufficient to compensate for the loss of that of Dsg1, or that Dsg1 and Dsc1 in combination maintain desmosomal cell–cell adhesion in the superficial epidermis. Alternatively, Dsg1 and Dsc1 may provide heterophilic trans-interaction, and cleavage of Dsg1 by ETs causes disruption of desmosomal cell–cell adhesion. Further studies are necessary to clarify the roles of Dsg1 and Dsc1 in maintaining desmosomal cell–cell adhesion in the epidermis. Although well understood in impetigo, the mechanisms underlying disruption of desmosomal cell–cell adhesion in pemphigus have been widely debated. Two previous reports showed that IgG in serum from patients with PF induces internalization of non-clustered Dsg1, but does not reduce
Dsg1 adhesion in vitro [29,30]. In contrast, an in vitro study using cultured keratinocytes reported that IgG in PF serum does not inhibit homophilic trans-interaction of Dsg1 by steric hindrance. However, whether IgG in PF serum blocks heterophilic interactions of Dsg1 with Dsc1 remains unknown [31]. Conversely, another recent report supported the concept of steric hindrance [32]. Anti-Dsg1 IgG from PF patient serum was found to cause dissociation of keratinocytes by direct inhibition of the trans-interaction site of Dsg, likely because the majority of IgG epitopes are located either within, or close to, the adhesive region of Dsg1, which is recognized by the anti-Dsg1 mAb [32]. Our findings do not shed new light on the pathogenesis of pemphigus. However, our data do suggest that inactivation of the adhesive function of Dsg1, without apparent internalization of desmosomal cadherins, is sufficient to initiate keratinocyte separation in the superficial epidermis. Immunofluorescence staining of keratinocyte surfaces using two PF patient sera, both of which recognize all three mouse Dsg1 isoforms, was absent in ETA-treated mice. ETA-digested mouse Dsg1a and Dsg1b, but not Dsg1g, are similarly expressed on the surface of mouse keratinocytes [17]. Previous study leads a speculation that substitution of Glu-381 to lysine, at the ET cleavage site of Dsg1, in mouse Dsg1g is responsible for substrate’s resistance to cleavage by ETA [17]. Taking our data in combination with previous studies, we postulate that Dsg1g is insufficient to compensate for loss of the adhesive functions of Dsg1a and Dsg1b in superficial mouse epidermis. In humans, Dsg4 is also expressed on the surface of keratinocytes just below the stratum corneum [12]. However, we did not address this in this study because Dsg4 is expressed predominantly in anagen hair bulbs and present at only low levels in the mouse epidermis [23]. Thus, compensation for loss of the adhesive function of Dsg1 beneath the cornified layer of human epidermis by Dsg4 has yet to be elucidated. Collectively, our findings provide not only information on the pathophysiological mechanisms of blister formation in bullous impetigo or SSSS, but also have implications for the role of desmosomal cadherins in maintaining keratinocyte adhesion in the superficial epidermis. Acknowledgments We thank Dr. David R. Garrod for providing the 11-5F mAb. We also thank Ms. Minae Suzuki for help with immunofluorescence, Ms. Hiromi Ito for assistance with animal husbandry, and Ms. Yoshiko Fujii for preparing baculoproteins. Our work was supported by Grants-In-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (to KN, AI, and MA), a Research Fellowship of the Japan Society for the Promotion of Science for Young Scientists (to KN) and a Health and Labour Sciences Research Grant for Research on Measures for Intractable Diseases from Ministry of Health, Labour and Welfare of Japan (to MA). References [1] Amagai M. Desmoglein as a target in autoimmunity and infection. J Am Acad Dermatol 2003;48:244–52. [2] Nishifuji K, Sugai M, Amagai M. Staphylococcal exfoliative toxins: ‘‘molecular scissors’’ of bacteria that attack the cutaneous defense barrier in mammals. J Dermatol Sci 2008;49:21–31. [3] Stanley JR, Amagai M. Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome. N Eng J Med 2006;355:1800–10. [4] Ladhani S, Joannou CL, Lochrie DP, Evans RW, Poston SM. Clinical, microbial, and biochemical aspects of the exfoliative toxins causing staphylococcal scalded-skin syndrome. Clin Microbiol Rev 1999;12:224–42. [5] Melish ME, Glasgow LA. The staphylococcal scalded skin syndrome: development of an experimental model. N Engl J Med 1970;282:1114–9.
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191 [6] Vath GM, Earhart CA, Rago JV, Kim MH, Bohach GA, Schlievert PM, et al. The structure of the superantigen exfoliative toxin A suggests a novel regulation as a serine protease. Biochemistry 1997;36:1559–66. [7] Amagai M, Matsuyoshi N, Wang ZH, Andl C, Stanley JR. Toxin in bullous impetigo and staphylococcal scalded-skin syndrome targets desmoglein 1. Nat Med 2000;6:1275–7. [8] Hanakawa Y, Schechter NM, Lin C, Garza L, Li H, Yamaguchi T, et al. Molecular mechanisms of blister formation in bullous impetigo and staphylococcal scalded skin syndrome. J Clin Invest 2002;110:53–60. [9] Amagai M, Yamaguchi T, Hanakawa Y, Nishifuji K, Sugai M, Stanley JR. Staphylococcal exfoliative toxin B specifically cleaves desmoglein 1. J Invest Dermatol 2002;118:845–50. [10] Yamaguchi T, Nishifuji K, Sasaki M, Fudaba Y, Aepfelbacher M, Takata T, et al. Identification of the Staphylococcus aureus etd pathogenicity island which encodes a novel exfoliative toxin, ETD, and EDIN-B. Infect Immun 2002;70: 5835–5845. [11] Yin T, Green KJ. Regulation of desmosome assembly and adhesion. Semin Cell Dev Biol 2004;15:665–77. [12] Bazzi H, Getz A, Mahoney MG, Ishida Yamamoto A, Langbein L, Wahl 3rd JK, et al. Desmoglein 4 is expressed in highly differentiated keratinocytes and trichocytes in human epidermis and hair follicle. Differentiation 2006;74: 129–140. [13] Amagai M, Tsunoda K, Suzuki H, Nishifuji K, Koyasu S, Nishikawa T. Use of autoantigen-knockout mice in developing an active autoimmune disease model for pemphigus. J Clin Invest 2000;105:625–31. [14] Nagasaka T, Nishifuji K, Ota T, Whittock NV, Amagai M. Defining the pathogenic involvement of desmoglein 4 in pemphigus and staphylococcal scalded skin syndrome. J Clin Invest 2004;114:1484–92. [15] Fudaba Y, Nishifuji K, Andresen LO, Yamaguchi T, Komatsuzawa H, Amagai M, Sugai M. Staphylococcus hyicus exfoliative toxins selectively digest porcine desmoglein 1. Microb Pathog 2005;39:171–6. [16] Hashimoto T, Kiyokawa C, Mori O, Miyasato M, Chidgey MA, Garrod DR, et al. Human desmocollin 1 (Dsc1) is an autoantigen for the subcorneal pustular dermatosis type of IgA pemphigus. J Invest Dermatol 1997;109:127–31. [17] Brennan D, Hu Y, Kljuic A, Choi Y, Joubeh S, Bashkin M, et al. Differential structural properties and expression patterns suggest functional significance for multiple mouse desmoglein 1 isoforms. Differentiation 2004; 72:434–49. [18] Mahoney MG, Hu Y, Brennan D, Bazzi H, Christiano AM, Wahl III JK. Delineation of diversified desmoglein distribution in stratified squamous epithelia: implications in diseases. Exp Dermatol 2006;15:101–9. [19] Tsunoda K, Ota T, Aoki M, Yamada T, Nagai T, Nakagawa T, et al. Induction of pemphigus phenotype by a mouse monoclonal antibody against the amino-terminal adhesive interface of desmoglein 3. J Immunol 2003; 170:2170–8. [20] Payne AS, Ishii K, Kacir S, Lin C, Li H, Hanakawa Y, et al. Genetic and functional characterization of human pemphigus vulgaris monoclonal autoantibodies isolated by phage display. J Clin Invest 2005;115:888–99.
191
[21] Hanakawa Y, Schechter NM, Lin C, Nishifuji K, Amagai M, Stanley JR. Enzymatic and molecular characteristics of the efficiency and specificity of exfoliative toxin cleavage of desmoglein 1. J Biol Chem 2004;279:5268–77. [22] Syed SE, Trinnaman B, Martin S, Major S, Hutchinson J, Magee AI. Molecular interactions between desmosomal cadherins. Biochem J 2002;362:317–27. [23] Chitaev NA, Troyanovsky SM. Direct Ca2+-dependent heterophilic interaction between desmosomal cadherins, desmoglein and desmocollin, contributes to cell–cell adhesion. J Cell Biol 1997;138:193–201. [24] Runswick SK, O’Hare MJ, Jones L, Streuli CH, Garrod DR. Desmosomal adhesion regulates epithelial morphogenesis and cell positioning. Nat Cell Biol 2001;3:823–30. [25] Tselepis C, Chidgey M, North A, Garrod D. Desmosomal adhesion inhibits invasive behavior. Proc Natl Acad Sci USA 1998;95:8064–9. [26] Getsios S, Amargo EV, Dusek RL, Ishii K, Sheu L, Godsel LM, et al. Coordinated expression of desmoglein 1 and desmocollin 1 regulates intercellular adhesion. Differentiation 2004;72:419–33. [27] Chidgey M, Brakebusch C, Gustafsson E, Cruchley A, Hail C, Kirk S, et al. Mice lacking desmocollin 1 show epidermal fragility accompanied by barrier defects and abnormal differentiation. J Cell Biol 2001;155:821–32. [28] Amagai M, Hashimoto T, Green KJ, Shimizu N, Nishikawa T. Antigen-specific immunoadsorption of pathogenic autoantibodies in pemphigus foliaceus. J Invest Dermatol 1995;104:895–901. [29] Cirillo N, Gombos F, Lanza A. Changes in desmoglein 1 expression and subcellular localization in cultured keratinocytes subjected to anti-desmoglein 1 pemphigus autoimmunity. J Cell Physiol 2007;210:411–6. [30] Lanza A, De Rosa A, Femiano F, Annese P, Ruocco E, Gombos F, et al. Internalization of non-clustered desmoglein 1 without depletion of desmoglein 1 from adhesion complexes in an experimental model of the autoimmune disease pemphigus foliaceus. Int J Immunopathol Pharmacol 2007;20:355–61. [31] Waschke J, Bruggeman P, Baumgartner W, Zillikens D, Drenckhahn D. Pemphigus foliaceus IgG causes dissociation of desmoglein 1-containing junctions without blocking desmoglein 1 transinteraction. J Clin Invest 2005;115:3157–65. [32] Yokouchi M, Saleh MA, Kuroda K, Hachiya T, Stanley JR, Amagai M, et al. Pathogenic epitopes of autoantibodies in pemphigus reside in the aminoterminal adhesive region of desmogleins which are unmasked by proteolytic processing of prosequence. J Invest Dermatol 2009;129:2156–66. Koji Nishifuji graduated from the United Graduate School of Veterinary Sciences, Gifu University in 2001 and was a research fellow at Department of Dermatology, Keio University School of Medicine from 2001 to 2007. He became an Associate Professor from 2009 at Department of Veterinary Medicine, Tokyo University of Agriculture and Technology. Masayuki Amagai graduated from Keio University School of Medicine in 1985 and did postdoctoral fellowship at Dermatology Branch, National Cancer Institute, National Institutes of Health from 1989 to 1992. He was promoted to Professor and Chair, Department of Dermatology, Keio University in 2005, and vice director of Keio University Hospital in 2007.
Contents lists available at ScienceDirect
Journal of Dermatological Science journal homepage: www.elsevier.com/jds
Removal of amino-terminal extracellular domains of desmoglein 1 by staphylococcal exfoliative toxin is sufficient to initiate epidermal blister formation Koji Nishifuji a,b, Atsushi Shimizu a, Akira Ishiko a, Toshiroh Iwasaki b, Masayuki Amagai a,* a b
Department of Dermatology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan Department of Veterinary Medicine, Tokyo University of Agriculture and Technology, Fuchu, Tokyo, Japan
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 June 2010 Received in revised form 16 July 2010 Accepted 22 July 2010
Background: In both bullous impetigo and staphylococcal scalded-skin syndrome (SSSS), exfoliative toxins (ETs) produced by Staphylococcus aureus cause superficial intraepidermal blisters. ETs are known to cleave specifically a single peptide bond in the extracellular domains 3 and 4 of desmoglein (Dsg) 1. However, the precise mechanisms underlying ET-induced epidermal blister formation remain poorly understood. Objective: To determine whether cleavage of Dsg1 by an ET is sufficient to induce blister formation in vivo or if the subsequent internalization of cleaved Dsg1 or other desmosomal components is required. Methods: Skin samples obtained from neonatal mice injected with ETA were analyzed by time-lapse immunofluorescence and transmission electron microscopy for desmosomal components. Results: Epidermal blister formation was observed as early as 60 min after ETA treatment. At this time, the amino-terminal extracellular domains of Dsg1 disappeared from the surface of keratinocytes, while the cleaved carboxy-terminal domain of Dsg1 (Dsg1-C) as well as the extracellular domains of desmocollin 1 (Dsc1-N) remained on the cell surface. Half-split desmosomes with intracytoplasmic dense plaques and attached tonofilaments were recognized ultrastructurally on the split surface of keratinocytes at 60 min. Subsequent to this, Dsg1-C and Dsc1-N gradually disappeared from the surface layer of keratinocytes. Conclusion: Our findings suggest that the removal of amino-terminal extracellular domains of Dsg1 by ETs is sufficient to initiate epidermal blister formation in bullous impetigo and SSSS. ß 2010 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Desmoglein Desmosome Exfoliative toxin Impetigo Staphylococcal scalded-skin syndrome Staphylococcus
1. Introduction Bullous impetigo and its generalized form, staphylococcal scalded-skin syndrome (SSSS), are highly contagious, bacterial skin diseases. Under these conditions, infection by Staphylococcus aureus causes blisters in the superficial epidermis that are characterized histopathologically as intraepidermal clefts with acantholysis [1–3]. The virulent strains of staphylococci isolated from patients with bullous impetigo or SSSS produce exfoliative toxins (ETs). These are serine protease-like proteins capable of producing intraepidermal blisters with a histopathology identical to bullous impetigo or SSSS when injected into neonatal mice [4–6]. Our recent studies revealed that the two major ETs, ETA and ETB, are serine proteases that hydrolyze specifically only one peptide bond in a calcium-binding
Abbreviations: ET, exfoliative toxin; Dsg, desmoglein; Dsc, desmocollin; SSSS, staphylococcal scalded-skin syndrome. * Corresponding author. Tel.: +81 3 5363 3823; fax: +81 3 3351 6880. E-mail address: [email protected] (M. Amagai).
motif between the extracellular domains 3 and 4 of desmoglein (Dsg) 1, a desmosomal transmembrane constituent that belongs to a cadherin family of cell–cell adhesion molecules [7–9]. Because the amino-terminus of Dsg1 is thought to harbor its binding domain, cleavage of the extracellular region of Dsg1 by an ET may cause loss of its ability to bind [2]. Additionally, ETD (a novel isoform of ET produced by S. aureus originally isolated from patients with deep skin infections) cleaves the same peptide bond in Dsg1 and causes superficial epidermal blisters in neonatal mice [8,10]. Thus, ETs are unique serine proteases that hydrolyze only one peptide bond in a single protein, and are responsible for the formation of intraepidermal blisters in bullous impetigo and SSSS. Despite recent advances in our understanding about the enzymatic characteristics of ETs, the precise mechanisms by which ETs disrupt cell–cell adhesion of keratinocytes and cause acantholysis in vivo remain unclear. In the epidermis, two major desmosomal cadherin-type adhesion molecules, Dsgs and desmocollins (Dscs), are expressed in a differentiation-specific manner and function to maintain cohesion between adjacent keratinocytes [11]. Among the isoforms of desmosomal cadherins, Dsg1 and Dsc1
0923-1811/$36.00 ß 2010 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jdermsci.2010.07.010
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
are expressed predominantly in the spinous and granular layers of epidermis, whereas Dsg3 and Dsc3 are seen in the basal and immediate suprabasal epidermis. Dsg2 and Dsc2 are expressed in the basal epidermis, while Dsg4 is expressed predominantly in anagen hair follicles and just below the cornified layer of the epidermis [12]. While ETs are known to cleave Dsg1, little is known about the fate of desmosomal cadherins during separation of keratinocytes within the superficial epidermis. The aim of this study was to investigate the fate of desmosomal cadherins, especially cleaved Dsg1 and Dsc1, in vivo in ETA-treated mouse epidermis. 2. Materials and methods 2.1. Production and purification of recombinant ETA Recombinant ETA, fused with a His-tag at its carboxyl terminus, was produced in Escherichia coli strain DH10B and purified as described previously [9]. ETA was harvested from the soluble fraction in lysis buffer [20 mM Tris–HCl (pH 8.0), 0.2 M NaCl, 0.2% Triton X-100, and protease inhibitor cocktail (Complete Mini, EDTA-free; Roche Diagnostics, Basel, Switzerland)], purified using the TALON metal affinity resin system (BD Biosciences Clontech, Mountain View, CA) and dialyzed against phosphate-buffered saline (PBS). Protein concentrations were determined by a protein assay (Bio-Rad Laboratories Inc., Richmond, CA). Purified ETA was passed through a 0.22-mm filter (Millipore Corp., Bedford, MA), aliquoted, and stored at 80 8C until required.
185
gel II mini; Daiichi Pure Chemicals, Tokyo, Japan) and transferred to polyvinylidene fluoride membranes (Immobilon–P Transfer Membrane; Millipore Corp.). Intact and cleaved mouse Dsg1 isoforms were detected using an anti-E-tag mAb (1:4000; GE Healthcare, Little Chalfont, Buckinghamshire, UK). 2.5. Neonatal mouse study Neonatal Jcl:ICR mice (<24 h old, weighing 1.6–2.0 g; CLEA Japan, Inc., Tokyo, Japan) were injected subcutaneously with purified ETA (10 mg) in the dorsal neck and thorax. Skin samples were collected from the injection site, in which injected ETA might distribute directly to the epidermis, at 0, 60, 120, and 180 min after ETA injection. Formaldehyde-fixed, paraffin-embedded skin samples were sectioned, stained with hematoxylin and eosin (H&E), and subjected to histopathological analysis. Skin samples were also subjected to immunofluorescence and transmission electron microscopic analyses. Regarding the dose of ETA, in this study we used 10 mg/mouse, which is sufficient to induce blisters alone in upper epidermis of neonatal mice. In some of our previous study, we used 1 mg/mouse, which is not sufficient to cause blisters alone, to analyze the pathogenic activity of anti-Dsg3 monoclonal antibodies [19,20]. 2.6. Immunofluorescence Skin samples obtained from ETA-treated mice were embedded in [(Fig._1)TD$IG] O.C.T. compound (Tissue-Tek; Sakura Finetechnical, Tokyo,
2.2. Production of recombinant mouse Dsg Recombinant, secreted-form extracellular domains of mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4, fused with E- and His-tags on their carboxyl terminus, were produced using the baculovirus expression system as described previously [13–15]. Culture supernatants of High FiveTM insect cells (Invitrogen Corp., Carlsbad, CA) containing recombinant mouse Dsg were collected and stored at 70 8C after removal of cell debris by centrifugation. 2.3. Antibodies Sera from two patients with pemphigus foliaceus (PF) contained IgG that recognizes the extracellular domains of mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4, as determined by immunoprecipitation– immunoblotting using baculovirus recombinant mouse Dsgs. To remove the Dsg4-reacting fraction of IgG, PF sera were applied to a chromatography column (Econo-Column; Bio-Rad Laboratories Inc.) packed with TALON metal affinity resin, along with recombinant mouse Dsg4. Pass-through fractions from the chromatography column were collected and used for immunofluorescence analysis. Serum from a patient with subcorneal pustular dermatosis-type IgA pemphigus contained IgA that recognizes the extracellular domains of human Dsc1, expressed on the surface of COS-7 cells [16]. A mouse monoclonal antibody (mAb; DG3.10) specific for the cytoplasmic domain of three mouse Dsg1 isoforms and Dsg2 (PROGEN, Heidelberg, Germany) [17,18] and a mouse anti-Dsg3 mAb (AK18) that recognizes the extracellular domain of Dsg3 [19] were used for immunofluorescence analysis. 2.4. In vitro digestion of mouse Dsg1 isoforms by ETA Insect culture supernatants containing baculovirus recombinant mouse Dsg1a, Dsg1b, and Dsg1g were incubated in vitro with purified recombinant ETA (1 mg) at 37 8C for 1 h. Samples were extracted with Laemmli sample buffer at 100 8C for 5 min, separated by 10% SDS-polyacrylamide gel electrophoresis (Multi-
Fig. 1. Reactivity of ETA and human pemphigus foliaceus (PF) sera with mouse Dsg1 isoforms. (A) ETA cleaves the extracellular domains of mouse Dsg1a and Dsg1b, but not that of Dsg1g. Arrowhead and arrow indicate intact and cleaved mouse Dsg1 isoforms, respectively. (B) Sera from two patients with PF (PF #1 and #2) that recognize mouse Dsg1a, Dsg1b, Dsg1g, and Dsg4 are applied to a column packed with recombinant mouse Dsg4. Immunoreactivities of pre- and post-column fractions were confirmed by immunoprecipitation and immunoblotting. Note that anti-Dsg4 IgG antibodies are completely absent in both post-column fractions. Bars in the right panels indicate molecular weight standards.
[(Fig._2)TD$IG]
186
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Fig. 2. Microscopic blisters developed subsequently to disappearance of the amino-terminal extracellular domains of Dsg1 (Dsg1-N) from keratinocyte surfaces in ETAtreated mice. (A) Time-lapse histopathological and immunofluorescence analysis for Dsg1-N, the intracellular domain of Dsg1 (Dsg1-C), and the extracellular domains of Dsc1 (Dsc1-N) using the skin of ETA-treated mice. Initial separation of keratinocytes began 60 min after ETA injection and extended to intraepidermal blisters over time. Immunofluorescence of Dsg1-N disappeared from keratinocyte surfaces as early as 60 min after ETA injection, while the carboxyl-termini of Dsg1 (Dsg1-C) and Dsc1-N remained on blister cavity-facing keratinocyte surfaces (arrowheads) until 120 min after ETA injection. Dsg1-C and Dsc1-N disappear from the surface of free-floating keratinocytes 180 min after ETA injection. Dashed lines indicate basement membrane zones (BMZ). Bar represents 50 mm. (B) Half-split desmosomes with attached tonofilaments are observable on the surface of acantholytic keratinocytes during early ETA-induced blister formation. Neonatal mouse skin was subjected to transmission electron microscopic analysis 60 min after ETA injection. (a) Intraepidermal separation of keratinocytes in granular layers. Bar represents 2 mm. (b), (d) Separated keratinocytes in the roof (b) and floor (d) of blister cavities retain half-desmosomes (arrows). Bars represent 1 mm in (b) and 2 mm in (d). (c), (e)
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Japan). Non-fixed cryosections (5 mm) were blocked [1% bovine serum albumin (Sigma–Aldrich, St. Louis, MO) and 5% normal goat serum (DAKO, Glostup, Denmark)] in Tris–buffered saline (50 mM; pH7.4) at room temperature for 30 min. For counterstaining the extracellular domains of Dsg1 (Dsg1-N), the intracellular domain of Dsg1 (Dsg1-C), and the extracellular domains of Dsc1 (Dsc1-N), samples were incubated with PF sera (1:100 dilution) and an IgA pemphigus serum (1:40 dilution) at 4 8C overnight and washed in PBS; the secondary antibody (mouse mAb against human IgA, 1:100; Zymed Laboratories, San Francisco, CA) was then added at room temperature for 1 h. All samples were then incubated with a tertiary antibody [a mixture of Alexa Fluor 648 goat antihuman IgG (Invitrogen) and Alexa Fluor 488 anti-mouse IgG] and a quaternary antibody [DG3.10 labeled using the Zenon Alexa Fluor 546 mouse IgG1 labeling kit (1:40)]. For counterstaining Dsg1-C, desmoplakins (DPs), and plakoglobins (PGs), samples were incubated with DG3.10 labeled with a Zenon Alexa Fluor 488 mouse IgG1 labeling kit, 11-5F labeled with a Zenon Alexa Fluor 546 IgG1 labeling kit (1:40), and 11E4 conjugated with an Alexa Fluor 647 conjugation kit (Invitrogen). Following these reactions, samples were fixed in 4% (v/v) paraformaldehyde at room temperature for 10 min and mounted in Mowiol (Calbiochem-Novabiochem Corp., La Jolla, CA). Specimens were examined and images captured and processed under a confocal laser-scanning microscope (LSM510 META; Carl Zeiss Microimaging GmbH, Jena, Germany) using an Argon- or He–Ne laser. Fluorescence intensities of Dsg1-N, Dsg1-C, and Dsc1-N in mouse keratinocytes were quantified using ImageJ 1.42q software (http://rsb.info.nih.gov/ij) as follows. Fifty-four diagonal lines were drawn arbitrarily on 18 keratinocytes in six histopathological sections of the skin collected from two mice at each time point. The distance between two cell membranes along the diagonal line was divided into six areas. Pixel intensity of immunofluorescence along the line was quantified and the mean intensity in each area calculated. Statistical analysis using Dunnett’s test was performed to compare the intensities between area #1 and the other areas examined. Moreover, the percent fluorescence of Dsg1-C and Dsc1N in the cell surface layer to the entire keratinocyte was calculated using the following formula: Percent fluorescence in the surface layer = (summary of the means of intensities in areas #1 and #6)/(summary of the means of intensities from area #1 to area #6) 100. Fluctuation in percent fluorescence was analyzed for Dsg1-N, Dsg1-C, and Dsc1-N. Statistical analysis using Dunnett’s test was performed to compare fluorescence percentages before and after ETA injection. Significance level of the statistical analyses used in this study was defined as P < 0.05. 2.7. Transmission electron microscopy Skin samples obtained from ETA-treated mice were fixed in 2% glutaraldehyde and 1% osmium tetroxide, dehydrated by exposure to increasing ethanol concentrations, and then embedded in Epon 812 resin (Okensyoji, Tokyo, Japan). Resin was polymerized at 55 8C for 72 h. Ultrathin sections, stained with uranyl acetate and lead citrate, were examined under a JEOL-1200 EX transmission electron microscope (JEOL, Tokyo, Japan).
187
of neonatal mice treated with ETA. Neonatal ICR mice were injected subcutaneously with purified ETA [9]; skin samples were collected along the time course and subjected to histopathological as well as immunofluorescence analyses. ETA has been reported to degrade Dsg1a and Dsg1b, but not Dsg1g, in mouse epidermis [17], although the three Dsg1 isoforms are similarly expressed in mouse epidermis [17,18]. We confirmed that ETA selectively cleaves extracellular domains of the mouse Dsg1a, Dsg1b, but not that of Dsg1g, by in vitro digestion of the extracellular domains of recombinant mouse Dsg1 isoforms with ETA (Fig. 1A). For immunofluorescence of the amino-terminal extracellular domains of three Dsg1 isoforms (Dsg1-N), we used the pass-through fractions from PF serum run on a Dsg4 column. These recognize the extracellular domains of Dsg1a, Dsg1b, and Dsg1g, but not that of Dsg4 (Fig. 1B). 3.2. Amino-terminal, but not carboxy-terminal, of Dsg1 disappeared from cell surfaces during the initial phase of blister formation Epidermal blister formation was observed as early as 60 min after ETA injection and progressed with time (Fig. 2A). Dsg1-N disappeared almost totally from the keratinocyte surface at 60 min after ETA injection (Fig. 2A). In contrast, the intracellular carboxyl-termini of Dsg1 (Dsg1-C) and the extracellular domains of Dsc1 (Dsc1-N) remained on the surface of keratinocytes, including those facing toward the blister cavity, until 120 min after ETA injection (Fig. 2A). In some keratinocytes, blister cavity surface-facing Dsg1-C fluorescence intensity was lower than that of cell–cell contact sites. Dsg1-C and Dsc1-N disappeared from the surface of keratinocytes floating freely within the blister cavity 180 min after ETA injection (Fig. 2A). The time courses of Dsg1-N, Dsg1-C, and Dsc1-N staining were essentially identical in six independent samples from two mice injected with ETA. The time courses of DP, PG, and cytoplasmic desmosome component staining were similar and each was detectable on cell surfaces as late as 120 min after ETA injection (data not shown). To determine desmosome structural integrity during the early phase of ETA-mediated keratinocyte separation, skin samples were subjected to transmission electron microscopic analysis. Because we wanted to analyze desmosomes immediately after keratinocyte separation, we used skin samples collected 60 min after injection with ETA (Fig. 2B). At low magnification, intraepidermal separation of keratinocytes in the granular layers of the epidermis was observed. Acantholytic keratinocytes retained half-split desmosomes on their acantholytic surfaces on both the upper and lower surfaces of intraepidermal splits. At higher magnifications, these half-split desmosomes revealed a flocculent material, which may represent desmoglea, on their extracellular surfaces, together with a dense intracytoplasmic plaque with attached tonofilaments. These findings indicate that blister cavity-facing half-split desmosomes are present during the initial phase of ETA-induced blister formation. In contrast, the plasma membranes of separated acantholytic keratinocytes lost their flatness and showed cytoplasmic projections 180 min after ETA treatment (data not shown). Thus, the fate of amino- and carboxy-terminal Dsg1 domains after ETA injection differs, and carboxy-terminal domain of Dsg1 and amino-terminal domain of Dsc1 remained on cell surfaces during the initial phase of ETA-mediated blister formation.
3. Results 3.1. Specific cleavage of mouse Dsg1a, Dsg1b, but not Dsg1g, by ETA
3.3. Quantitative analysis of Dsg1 and Dsc1 during early ETA-induced blister formation
To investigate the mechanisms of ET-induced intraepidermal blisters in vivo, we first analyzed the fate of Dsg1 and Dsc1 in the skin
Immunofluorescence analysis showed that Dsg1-N, but not Dsg1-C or Dsc1-N, disappeared from blister-facing cell surfaces
Half-split desmosomes (double arrowheads) with flocculent material on their extracellular surface and intracytoplasmic dense plaques with attached tonofilaments. Bars represent 200 nm.
[(Fig._3)TD$IG]188
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
Fig. 3. Quantitative analysis of the fluorescence intensities of Dsg1-N, Dsg1-C, and Dsc1-N. (A) Diagonal lines were drawn arbitrarily and keratinocytes divided into six areas. In non-blistered keratinocytes, areas #1 and #6 included plasma membranes. In blistered keratinocytes, area #1 included plasma membranes facing to the blister cavity, whereas area #6 included those in cell–cell contact sites. Asterisks indicate blister cavity. (B) Comparison of the fluorescent intensities of Dsg1-N, Dsg1-C, and Dsc1-N in six areas. Graph shows the mean SD of pixel intensities of 54 diagonal lines. Blisters began to develop 60 min after ETA injection. Note that the Dsg1-N fluorescence intensity in areas #1 and #6 (including plasma membranes) decreased to similar levels to those detected in areas #2 to #5 (cytoplasmic layers) 60–120 min after ETA injection. Cytoplasmic Dsg1-N
[(Fig._4)TD$IG]
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
189
Fig. 4. Schematic diagram of the fate of desmosomal constituents during ETA-induced blister formation. Intact Dsg1 and Dsc1 mediate cell–cell keratinocyte adhesion prior to ETA treatment (0 min). Blister formation began 60 min after ETA treatment. At this time, Dsg1-N was absent from the surface of keratinocytes, while Dsg1-C and Dsc1-N remained detectable. Dsg1-C and Dsc1-N disappeared from keratinocyte surfaces at 120 and 180 min, respectively, concurrent with blister progression.
during the early phase of ETA-induced blister formation. To analyze the fate of desmosomal cadherins in keratinocytes in a quantitative manner, we calculated pixel intensities of these molecules in six areas of the diagonal lines connecting two plasma membranes (Fig. 3A). In this method, areas #1 and #6 always included plasma membranes of keratinocytes, whereas areas #2 to #5 always included the cytoplasm. In keratinocytes facing the blister cavity, area #1 always included the cell membrane facing the intraepidermal split, while area #6 always included cell–cell contact sites. We found that Dsg1-N fluorescence intensities in
cell-surface layers (areas #1 and #6) decreased to the same level as those in the cytoplasm (areas #2 to #5) as early as 60 min, concurrent with initial blister development (Fig. 3B). Dsg1-C and Dsc1-N intensities were higher in blister cavity-facing cell surface layers (area #1) than in cytoplasmic layers (areas #2 to #5) up to 180 min after ETA injection (Dunnett’s test, P < 0.05, Fig. 3B). No increase in Dsg1-C and Dsc1-N fluorescence intensities in keratinocyte cytoplasm was detected, probably because internalized Dsg1-C and Dsc1-N were removed from tissue sections during immunofluorescence procedures. No significant differences in the
fluorescence intensity decreased at 180 min. Conversely, fluorescence intensities of Dsg1-C and Dsc1-N were consistently higher in keratinocyte surface than in cytoplasmic layers up to 180 min (Dunnett’s test, P < 0.05). (C) Fluctuation in the percent fluorescence of Dsg1-N, Dsg1-C, and Dsc1-N in keratinocyte surface layers. The Dsg1-N percentage 60 min after ETA injection had decreased, while that of Dsg1-C and Dsc1-N remained constant. Data are expressed as means SD.
190
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191
fluorescence intensities of Dsg1-C and Dsc1-N between blister cavity-facing cell surfaces (area #1) and cell–cell contact sites (area #6) were detected (P > 0.05; Fig. 3B). Moreover, we assessed fluctuations in the percent fluorescence of Dsg1-N, Dsg1-C, and Dsc1-N in keratinocyte cell-surface layers (Fig. 3C). We found that the Dsg1-N percentage decreased markedly 60 min after injection, concurrent with initial blister formation (Dunnett’s test, P < 0.05), while that of Dsc1-N was not significantly different (P > 0.05). The Dsg1-C fluorescence percentage was slightly lower at time 60 min (P < 0.05). The percentages of Dsg1-C and Dsc1-N decreased gradually as blister formation progressed, up to 180 min after ETA injection. Taken together, these data suggest that Dsg1-N disappeared from the cell surface layer of keratinocytes during the early phase of ETA-induced blister formation. In contrast, the majority of Dsg1-C and Dsc1-N remained on the cell surface layer during the early phase but gradually disappeared during the later phase of ETA-induced blister formation (see schematic drawing in Fig. 4). 4. Discussion In this study, we demonstrated that ETA-induced intraepidermal blisters form subsequent to disappearance of Dsg1-N from the surface of epidermal keratinocytes in vivo. Additionally, our findings suggest that the majority of Dsg1-C and Dsc1-N remained on the blister cavity-facing surface of keratinocytes during the early phase of ETA-induced blister formation. Ultrastructural analysis also revealed split half-desmosomes with intracytoplasmic dense plaques and attached tonofilaments during the early phase of the blister formation, suggesting structural integrity. During the late phase of ETA-induced blister formation, desmosomal cadherins disappeared from the surface layer of acantholytic keratinocytes. Taken together, these findings support our hypothesis that ETA-mediated disappearance of amino-terminal extracellular domains of Dsg1 alone is sufficient to induce initial separation of keratinocytes. Meanwhile, our data do not rule out the possibility that separation of keratinocytes is enhanced by subsequent signaling cascades that include internalization of desmosomal constituents. While the effect of ETB or ETD on the fate of desmosomal cadherins has not been addressed in the present study, it is speculated that these ETs affect desmosomes as does ETA, because all three ETs share identical enzymatic activity to cleave a single peptide bond in Dsg1 efficiently [8,21]. Previous in vitro studies using cell lines expressing ectopic Dsg and/or Dsc revealed that Dsg and Dsc form both homophilic [22] and heterophilic [22,23] complexes. Both Dsg and Dsc are thought to play important roles in cell–cell adhesion, since functional blocking of either one of the molecules causes disruption of cell–cell adhesion both in vitro [24–26] and in vivo [7,9,10,27,28]. In this study, we demonstrated that Dsc1-N remains detectable on blister cavity-facing keratinocyte surfaces in ETA-treated mice. These findings suggest that either the adhesive function of Dsc1 is insufficient to compensate for the loss of that of Dsg1, or that Dsg1 and Dsc1 in combination maintain desmosomal cell–cell adhesion in the superficial epidermis. Alternatively, Dsg1 and Dsc1 may provide heterophilic trans-interaction, and cleavage of Dsg1 by ETs causes disruption of desmosomal cell–cell adhesion. Further studies are necessary to clarify the roles of Dsg1 and Dsc1 in maintaining desmosomal cell–cell adhesion in the epidermis. Although well understood in impetigo, the mechanisms underlying disruption of desmosomal cell–cell adhesion in pemphigus have been widely debated. Two previous reports showed that IgG in serum from patients with PF induces internalization of non-clustered Dsg1, but does not reduce
Dsg1 adhesion in vitro [29,30]. In contrast, an in vitro study using cultured keratinocytes reported that IgG in PF serum does not inhibit homophilic trans-interaction of Dsg1 by steric hindrance. However, whether IgG in PF serum blocks heterophilic interactions of Dsg1 with Dsc1 remains unknown [31]. Conversely, another recent report supported the concept of steric hindrance [32]. Anti-Dsg1 IgG from PF patient serum was found to cause dissociation of keratinocytes by direct inhibition of the trans-interaction site of Dsg, likely because the majority of IgG epitopes are located either within, or close to, the adhesive region of Dsg1, which is recognized by the anti-Dsg1 mAb [32]. Our findings do not shed new light on the pathogenesis of pemphigus. However, our data do suggest that inactivation of the adhesive function of Dsg1, without apparent internalization of desmosomal cadherins, is sufficient to initiate keratinocyte separation in the superficial epidermis. Immunofluorescence staining of keratinocyte surfaces using two PF patient sera, both of which recognize all three mouse Dsg1 isoforms, was absent in ETA-treated mice. ETA-digested mouse Dsg1a and Dsg1b, but not Dsg1g, are similarly expressed on the surface of mouse keratinocytes [17]. Previous study leads a speculation that substitution of Glu-381 to lysine, at the ET cleavage site of Dsg1, in mouse Dsg1g is responsible for substrate’s resistance to cleavage by ETA [17]. Taking our data in combination with previous studies, we postulate that Dsg1g is insufficient to compensate for loss of the adhesive functions of Dsg1a and Dsg1b in superficial mouse epidermis. In humans, Dsg4 is also expressed on the surface of keratinocytes just below the stratum corneum [12]. However, we did not address this in this study because Dsg4 is expressed predominantly in anagen hair bulbs and present at only low levels in the mouse epidermis [23]. Thus, compensation for loss of the adhesive function of Dsg1 beneath the cornified layer of human epidermis by Dsg4 has yet to be elucidated. Collectively, our findings provide not only information on the pathophysiological mechanisms of blister formation in bullous impetigo or SSSS, but also have implications for the role of desmosomal cadherins in maintaining keratinocyte adhesion in the superficial epidermis. Acknowledgments We thank Dr. David R. Garrod for providing the 11-5F mAb. We also thank Ms. Minae Suzuki for help with immunofluorescence, Ms. Hiromi Ito for assistance with animal husbandry, and Ms. Yoshiko Fujii for preparing baculoproteins. Our work was supported by Grants-In-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (to KN, AI, and MA), a Research Fellowship of the Japan Society for the Promotion of Science for Young Scientists (to KN) and a Health and Labour Sciences Research Grant for Research on Measures for Intractable Diseases from Ministry of Health, Labour and Welfare of Japan (to MA). References [1] Amagai M. Desmoglein as a target in autoimmunity and infection. J Am Acad Dermatol 2003;48:244–52. [2] Nishifuji K, Sugai M, Amagai M. Staphylococcal exfoliative toxins: ‘‘molecular scissors’’ of bacteria that attack the cutaneous defense barrier in mammals. J Dermatol Sci 2008;49:21–31. [3] Stanley JR, Amagai M. Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome. N Eng J Med 2006;355:1800–10. [4] Ladhani S, Joannou CL, Lochrie DP, Evans RW, Poston SM. Clinical, microbial, and biochemical aspects of the exfoliative toxins causing staphylococcal scalded-skin syndrome. Clin Microbiol Rev 1999;12:224–42. [5] Melish ME, Glasgow LA. The staphylococcal scalded skin syndrome: development of an experimental model. N Engl J Med 1970;282:1114–9.
K. Nishifuji et al. / Journal of Dermatological Science 59 (2010) 184–191 [6] Vath GM, Earhart CA, Rago JV, Kim MH, Bohach GA, Schlievert PM, et al. The structure of the superantigen exfoliative toxin A suggests a novel regulation as a serine protease. Biochemistry 1997;36:1559–66. [7] Amagai M, Matsuyoshi N, Wang ZH, Andl C, Stanley JR. Toxin in bullous impetigo and staphylococcal scalded-skin syndrome targets desmoglein 1. Nat Med 2000;6:1275–7. [8] Hanakawa Y, Schechter NM, Lin C, Garza L, Li H, Yamaguchi T, et al. Molecular mechanisms of blister formation in bullous impetigo and staphylococcal scalded skin syndrome. J Clin Invest 2002;110:53–60. [9] Amagai M, Yamaguchi T, Hanakawa Y, Nishifuji K, Sugai M, Stanley JR. Staphylococcal exfoliative toxin B specifically cleaves desmoglein 1. J Invest Dermatol 2002;118:845–50. [10] Yamaguchi T, Nishifuji K, Sasaki M, Fudaba Y, Aepfelbacher M, Takata T, et al. Identification of the Staphylococcus aureus etd pathogenicity island which encodes a novel exfoliative toxin, ETD, and EDIN-B. Infect Immun 2002;70: 5835–5845. [11] Yin T, Green KJ. Regulation of desmosome assembly and adhesion. Semin Cell Dev Biol 2004;15:665–77. [12] Bazzi H, Getz A, Mahoney MG, Ishida Yamamoto A, Langbein L, Wahl 3rd JK, et al. Desmoglein 4 is expressed in highly differentiated keratinocytes and trichocytes in human epidermis and hair follicle. Differentiation 2006;74: 129–140. [13] Amagai M, Tsunoda K, Suzuki H, Nishifuji K, Koyasu S, Nishikawa T. Use of autoantigen-knockout mice in developing an active autoimmune disease model for pemphigus. J Clin Invest 2000;105:625–31. [14] Nagasaka T, Nishifuji K, Ota T, Whittock NV, Amagai M. Defining the pathogenic involvement of desmoglein 4 in pemphigus and staphylococcal scalded skin syndrome. J Clin Invest 2004;114:1484–92. [15] Fudaba Y, Nishifuji K, Andresen LO, Yamaguchi T, Komatsuzawa H, Amagai M, Sugai M. Staphylococcus hyicus exfoliative toxins selectively digest porcine desmoglein 1. Microb Pathog 2005;39:171–6. [16] Hashimoto T, Kiyokawa C, Mori O, Miyasato M, Chidgey MA, Garrod DR, et al. Human desmocollin 1 (Dsc1) is an autoantigen for the subcorneal pustular dermatosis type of IgA pemphigus. J Invest Dermatol 1997;109:127–31. [17] Brennan D, Hu Y, Kljuic A, Choi Y, Joubeh S, Bashkin M, et al. Differential structural properties and expression patterns suggest functional significance for multiple mouse desmoglein 1 isoforms. Differentiation 2004; 72:434–49. [18] Mahoney MG, Hu Y, Brennan D, Bazzi H, Christiano AM, Wahl III JK. Delineation of diversified desmoglein distribution in stratified squamous epithelia: implications in diseases. Exp Dermatol 2006;15:101–9. [19] Tsunoda K, Ota T, Aoki M, Yamada T, Nagai T, Nakagawa T, et al. Induction of pemphigus phenotype by a mouse monoclonal antibody against the amino-terminal adhesive interface of desmoglein 3. J Immunol 2003; 170:2170–8. [20] Payne AS, Ishii K, Kacir S, Lin C, Li H, Hanakawa Y, et al. Genetic and functional characterization of human pemphigus vulgaris monoclonal autoantibodies isolated by phage display. J Clin Invest 2005;115:888–99.
191
[21] Hanakawa Y, Schechter NM, Lin C, Nishifuji K, Amagai M, Stanley JR. Enzymatic and molecular characteristics of the efficiency and specificity of exfoliative toxin cleavage of desmoglein 1. J Biol Chem 2004;279:5268–77. [22] Syed SE, Trinnaman B, Martin S, Major S, Hutchinson J, Magee AI. Molecular interactions between desmosomal cadherins. Biochem J 2002;362:317–27. [23] Chitaev NA, Troyanovsky SM. Direct Ca2+-dependent heterophilic interaction between desmosomal cadherins, desmoglein and desmocollin, contributes to cell–cell adhesion. J Cell Biol 1997;138:193–201. [24] Runswick SK, O’Hare MJ, Jones L, Streuli CH, Garrod DR. Desmosomal adhesion regulates epithelial morphogenesis and cell positioning. Nat Cell Biol 2001;3:823–30. [25] Tselepis C, Chidgey M, North A, Garrod D. Desmosomal adhesion inhibits invasive behavior. Proc Natl Acad Sci USA 1998;95:8064–9. [26] Getsios S, Amargo EV, Dusek RL, Ishii K, Sheu L, Godsel LM, et al. Coordinated expression of desmoglein 1 and desmocollin 1 regulates intercellular adhesion. Differentiation 2004;72:419–33. [27] Chidgey M, Brakebusch C, Gustafsson E, Cruchley A, Hail C, Kirk S, et al. Mice lacking desmocollin 1 show epidermal fragility accompanied by barrier defects and abnormal differentiation. J Cell Biol 2001;155:821–32. [28] Amagai M, Hashimoto T, Green KJ, Shimizu N, Nishikawa T. Antigen-specific immunoadsorption of pathogenic autoantibodies in pemphigus foliaceus. J Invest Dermatol 1995;104:895–901. [29] Cirillo N, Gombos F, Lanza A. Changes in desmoglein 1 expression and subcellular localization in cultured keratinocytes subjected to anti-desmoglein 1 pemphigus autoimmunity. J Cell Physiol 2007;210:411–6. [30] Lanza A, De Rosa A, Femiano F, Annese P, Ruocco E, Gombos F, et al. Internalization of non-clustered desmoglein 1 without depletion of desmoglein 1 from adhesion complexes in an experimental model of the autoimmune disease pemphigus foliaceus. Int J Immunopathol Pharmacol 2007;20:355–61. [31] Waschke J, Bruggeman P, Baumgartner W, Zillikens D, Drenckhahn D. Pemphigus foliaceus IgG causes dissociation of desmoglein 1-containing junctions without blocking desmoglein 1 transinteraction. J Clin Invest 2005;115:3157–65. [32] Yokouchi M, Saleh MA, Kuroda K, Hachiya T, Stanley JR, Amagai M, et al. Pathogenic epitopes of autoantibodies in pemphigus reside in the aminoterminal adhesive region of desmogleins which are unmasked by proteolytic processing of prosequence. J Invest Dermatol 2009;129:2156–66. Koji Nishifuji graduated from the United Graduate School of Veterinary Sciences, Gifu University in 2001 and was a research fellow at Department of Dermatology, Keio University School of Medicine from 2001 to 2007. He became an Associate Professor from 2009 at Department of Veterinary Medicine, Tokyo University of Agriculture and Technology. Masayuki Amagai graduated from Keio University School of Medicine in 1985 and did postdoctoral fellowship at Dermatology Branch, National Cancer Institute, National Institutes of Health from 1989 to 1992. He was promoted to Professor and Chair, Department of Dermatology, Keio University in 2005, and vice director of Keio University Hospital in 2007.