SERCA2 dysfunction in Darier disease causes endoplasmic reticulum stress and impaired cell-to-cell adhesion strength: rescue by Miglustat.

Accepted Article Preview: Published ahead of advance online publication SERCA2 Dysfunction in Darier Disease Causes ER stress and Impaired Cell-to-Cell Adhesion Strength: Rescue by Miglustat Magali Savignac, Marina Simon, Anissa Edir, Laure Guibbal, Alain Hovnanian

Cite this article as: Magali Savignac, Marina Simon, Anissa Edir, Laure Guibbal, Alain Hovnanian, SERCA2 Dysfunction in Darier Disease Causes ER stress and Impaired Cell-to-Cell Adhesion Strength: Rescue by Miglustat, Journal of Investigative Dermatology accepted article preview 3 January 2014; doi: 10.1038/ jid.2014.8. This is a PDF ﬁle of an unedited peer-reviewed manuscript that has been accepted for publication. NPG are providing this early version of the manuscript as a service to our customers. The manuscript will undergo copyediting, typesetting and a proof review before it is published in its ﬁnal form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.

Received 19 July 2013; revised 31 October 2013; accepted 9 December 2013; Accepted article preview online 3 January 2014

© 2014 The Society for Investigative Dermatology

SERCA2 dysfunction in Darier disease causes ER stress and impaired cell-to-cell adhesion strength: rescue by Miglustat

Magali Savignac1,6, Marina Simon2,3,4,6, Anissa Edir2,3, Laure Guibbal2,3,4 and Alain Hovnanian2,3,4,5 1

INSERM, U1043, Toulouse, F-31300 France;

2

INSERM, U781, Paris, F-75015 France;

3

Université Paris Descartes - Sorbonne Paris Cité, Paris, F-75015 France;

4

Institut Imagine, Paris, F-75015 France;

5

Department of Genetics, Necker Hospital, Paris, F-75015 France,

6

These authors contributed equally to this work.

Correspondence to: Alain HOVNANIAN, MD, PhD, CHU Necker-Enfants malades, Department of Genetics, Tour Lavoisier, 3rd floor, 149 rue de Sèvres, 75743 Paris cedex 15, France. E-mail: [email protected], Tel: +33 1 71 19 63 95, Fax: +33 1 71 19 64 20

Short Title : ER stress decreases adhesion in Darier

Abbreviations List NK: normal keratinocytes; DK: Darier keratinocytes; ER: endoplasmic reticulum; DD: Darier disease; PBA: 4-phenylbutyric acid; DP: desmoplakin; Dsg: desmoglein; Dsc: desmocollin; Ca2+: calcium; PG: plakoglobin; ERGIC: ER Golgi intermediaite compartment; TG: thapsigargin; SGPL1: sphingosine lyase; PKC: protein kinase C; [Ca2+]ER: ER Ca2+ concentration.

1 © 2014 The Society for Investigative Dermatology

Abstract

Darier disease (DD) is a severe dominant genetic skin disorder characterized by loss of cell-to-cell adhesion and abnormal keratinization. The defective gene, ATP2A2, encodes SERCA2, a Ca2+-ATPase pump of the endoplasmic reticulum (ER). Here, we show that Darier keratinocytes (DK) display biochemical and morphological hallmarks of constitutive ER stress with increased sensitivity to ER stressors. Desmosome and adherens junctions (AJ) displayed features of immature adhesion complexes: expression of desmosomal cadherins (Desmoglein 3 (Dsg3), Desmocollin 3 (Dsc3)) and desmoplakin (DP) was impaired at the plasma membrane, as well as E-cadherin, -, -, and p120-catenins staining. Dsg3, Dsc3 and E-cadherin showed perinuclear staining and co-immunostaining with ER markers, indicative of ER retention. Consistent with these abnormalities, intercellular adhesion strength was reduced as shown by a dispase mechanical dissociation assay. Exposure of normal keratinocytes to the SERCA2 inhibitor thapsigargin recapitulated these abnormalities, supporting the role of loss of SERCA2 function in impaired desmosome and AJ formation. Remarkably, treatment of DK with the orphan drug Miglustat, a pharmacological chaperone, restored mature AJ and desmosome formation, and improved adhesion strength. These results point to an important contribution of ER stress in DD pathogenesis and provide the basis for future clinical evaluation of Miglustat in Darier patients.


Introduction Darier disease (DD) is a severe dominant genetic skin disease characterized by loss of cell-to-cell adhesion (acantholysis) and abnormal keratinization (Burge and Hovnanian, 2011). We previously identified ATP2A2 as the defective gene in DD (Sakuntabhai et al., 1999). ATP2A2 encodes the sarco/endoplasmic reticulum Ca2+-ATPase isoform 2 (SERCA2), a calcium pump which plays a central role in Ca2+ homeostasis by transporting Ca2+ from the cytosol to the ER lumen to maintain high Ca2+ concentrations in this organel. Functional studies of ATP2A2 mutations identified in DD have shown them to abolish Ca2+ transport by SERCA2, leading to depleted ER Ca2+ stores in patient keratinocytes (reviewed in (Savignac et al., 2011). No animal model is available for DD: atp2a2 knockout is lethal in mice and aged atp2a2+/- mice develop squamous cell carcinomas but no DD like lesions (Liu et al., 2001). Therefore, Darier keratinocytes (DK) represent a unique model to explore the mechanisms underlying DD physiopathology. We and others previously showed defective desmoplakin (DP) trafficking in DK (Celli et al., 2012; Dhitavat et al., 2003; Hobbs et al., 2011), implicating this linker molecule between desmosomes and the keratin filament network, in impaired adhesion in DD. One of the immediate cellular responses of keratinocytes to extracellular Ca2+ increase is the formation of adherens junctions (AJ): E-cadherin associates with

-

catenin in the ER, prior to migrate to the plasma membrane (Perez-Moreno et al., 2003). Then, its extracellular portion interacts with E-cadherin molecules at the surface of neighbouring cells, while its cytoplasmic tail interacts with -,

- and p120-catenins to

form the core adhesive structure of AJ and links E-cadherin to the actin cytoskeleton (Niessen et al., 2011; Pokutta and Weis, 2007). Secondarily, desmosome formation reinforces cell-cell adhesion and plays a crucial role in tissue adhesion strength and resistance to mechanical stress (Getsios et al., 2004). Desmosomes comprise


transmembrane cadherins, desmogleins (Dsg) and desmocollins (Dsc), which interact with members of the armadillo family, including plakoglobin and the plakophilins, which in turn bind to DP, a major plaque protein linked to the keratin intermediate filament cytoskeleton (Garrod and Chidgey, 2008; Green and Simpson, 2007; Jamora and Fuchs, 2002). Desmosomal cadherins are constitutively synthesized and transported to the plasma membrane in response to cell-cell contact, where they stabilize and cluster in conjunction with plaque proteins (Jamora and Fuchs, 2002). ER calcium stores are crucial for post-translational processing, folding, and export of proteins associated to the plasma membrane or secreted (Berridge et al., 2003). Depletion of ER luminal Ca2+ leads to the accumulation of misfolded proteins, preventing their trafficking to the plasma membrane, and induces an ER stress response to rescue protein misfolding (Yoshida, 2007). This response involves PERK (protein kinase-like ER kinase), that phosphorylates eIF2

to attenuate translation; ATF6

(activation of transcription factor 6) that is transported to the Golgi apparatus and cleaved by proteases, resulting in nuclear translocation and enhanced transcription of ER-stress responsive genes; and IRE1 (inositol-requiring transmembrane kinase and endonuclease 1) that splices XBP1 pre-mRNA to mature XBP1 mRNA, leading to the transcription of ER-stress responsive genes. Here, we show that DK display a constitutive ER stress response. We provide evidence that AJ and desmosomes are immature, which results in decreased intercellular adhesion strength. We show that the

-glucosidase inhibitor Miglustat

restores mature AJ and desmosomes in DK and increases adhesion strength.


Results

Darier keratinocytes display a constitutive ER stress response Decreased [Ca2+]ER, a characteristic feature of DK (Dode et al., 2003; Foggia et al., 2006; Miyauchi et al., 2006; Pani et al., 2006), triggers ER stress in many cells (Ashby and Tepikin, 2001; Schroder and Kaufman, 2005). We show that DK in culture display increased eIF2

and IRE1 phosphorylation in the absence of exogenous stressor, as

compared to normal keratinocytes (NK), indicative of a constitutive ER stress (Figure1a and b, Time 0). After exposure to thapsigargin (TG), a highly specific inhibitor of SERCA pumps (Thastrup et al., 1989), phospho-eIF2 and phospho-IRE1 levels were enhanced in DK to a higher extent than in NK (Figure1a and b) (2- to 3-fold in DK and 1.5-fold in NK for phospho-eIF2 ). Consistently, the spliced active form of XBP1 mRNA was increased in DK compared to NK, after TG treatment (Figure1c). Theses ER stress markers were enhanced in DK from the five patients studied (D1 to D5) (Figure1 and data not shown). ER stress has been shown to induce morphological changes of the intracellular compartments in different cell types (Amodio et al., 2009; Schroder and Kaufman, 2005). Immunostaining for specific markers calnexin, ERGIC-53, GM130 and TGN46 showed respectively, the ER, ERGIC (ER Golgi intermediate compartment), cis-Golgi and transGolgi compartments to be strictly localized to a perinuclear zone in NK (Figure1d and FigureS1). In contrast, calnexin staining was diffuse all over the cytoplasm in DK (Figure1d and FigureS1). Specific staining for each post-ER compartment also revealed spots dispersed throughout the cytoplasm, indicating their fragmentation (Figure1d and FigureS1). Quantitative measurements showed a significant increase in volumes of each compartment (Figure1e).


Together, these data show that DK display biochemical and morphological hallmarks of ER stress response under stressor-free conditions.

Ca2+-induced adherens junction formation is impaired in DK The presence of a constitutive ER stress in DK is likely to alter important ER functions, such as folding of secretory and (trans)membrane proteins (reviewed in (Schroder and Kaufman, 2005). Acantholysis being the main histological feature in Darier skin lesions, we first investigated the localization of AJ components, which are addressed to the plasma membrane during the earliest cellular responses to Ca2+ increase leading to cellular adhesion. Under low-calcium conditions (0.06mM), E-cadherin, -, - and p120catenin proteins were diffusely distributed in the cytoplasm in NK and DK (FigureS2). After 4 hours in high-calcium medium (1.2mM), E-cadherin and catenin molecules relocated to the plasma membrane in NK, with a thin staining around the cellular periphery, forming a continuous belt along cell-cell boundaries (Figure2a, 2b and FigureS3a and b). In contrast, a punctuated pattern of all AJ proteins was detected at the cell-cell boundaries of DK, with abnormal organization of E-cadherin in strands arranged perpendicularly or at different angles to the cell edge (Figure2a, 2b and FigureS3a and b). F-actin filaments appeared as thin cortical bundles underneath and parallel to the plasma membrane in NK, whereas they formed thick and disorganized bundles in DK (Figure2b and FigureS3b). Surface biotinylation assays showed decreased Ecadherin expression at the plasma membrane in DK (Figure2c). These findings provide evidence for incorrect localization of AJ components, revealing abnormal formation of these early adhesion structures in DK. Immunoblotting analyses revealed expression levels of each AJ protein similar to controls (data not shown), suggesting that abnormal AJ formation in DK is not a consequence of reduced expression of AJ components. Of note, 4 hours after exposure


to 1.2mM of Ca2+ (time allowing AJ to form), phospho-eIF2 was increased in DK compared to NK (data not shown). Thus, decreased in Ca2+-induced AJ formation correlates with enhanced ER stress response in DK.

Desmosome formation and intercellular adhesion strength are impaired in DK Previously, we and others showed impaired DP trafficking in DK and in NK treated by pharmacological or genetic inhibitors of SERCA2 in culture (Celli et al., 2012; Dhitavat et al., 2003; Hobbs et al., 2011). Here, we confirm that DP staining, in DK, is thick with DP particles farther apart from one another and abnormally organized in strands perpendicular to the cell edge (Figure 2d and Figure S3c). Moreover, Dsg3 and Dsc3, which are expressed in the basal and supra-basal layers of the epidermis, where acantholytic lesions occur in DD skin, displayed a thicker and punctuated staining pattern underneath the plasma membrane in DK compared to NK (Figure2d and FigureS3c). Noticeably, a perinuclear staining of Dsg3 and Dsc3 was seen in DK suggesting their retention in the ER (Figure2d and FigureS3c). Moreover, dispase dissociation assay showed that a greater number of fragments (3-5 fold increase) was counted in DK compared to NK demonstrating impaired intercellular adhesive strength in DK (Figure2e).

Thapsigargin inhibits desmosome and adherens junction formation in NK To test whether SERCA2 loss of function is sufficient to induce defective AJ and desmosomes assembly, we treated NK with TG to specifically inhibit SERCA pumps. TG induced an enlargement of the ER compartment with a two-fold increase of ER volume in TG-treated NK compared to untreated NK (Figure3a). Immunofluorescence analysis of AJ and desmosome components in TG-pretreated NK cultured in 1.2mM Ca2+ revealed similar abnormalities to those in DK: E-cadherin, catenins, DP and Dsg3 were


organized into strands arranged at an angle with the cell edge, with a punctuated staining at the cell-cell boundaries (Figure3b-c and FigureS4), and the F-actin network appeared disorganized in the cytosol (Figure3b and FigureS4a). Dsc3 staining was lost in TG-treated NK compared to NK (data not shown), as previously described (Hakuno et al., 2000). These results were consistent with ER spreading and defective AJ and desmosomes formation observed in DK being a consequence of SERCA2 loss of function.

β-catenin is able to localize to the plasma membrane in DK Synthesized E-cadherin requires -catenin binding to facilitate its transport out of the ER to the plasma membrane (Chen et al., 1999). To investigate whether the defect in AJ formation is caused by lack of

-catenin recruitment to the plasma membrane, we

expressed an E-cadherin mutant targeted to the plasma membrane independently of catenin binding (Fukumoto et al., 2008). NK and DK were transduced with viral particles (MSCV-EcadM-GFP) encoding this E-cadherin mutant, under low-calcium conditions to prevent plasma membrane localization of endogenous E-cadherin. We observed in both cell types, a thin, pericellular staining of E-cadherin mutant at the plasma membrane (FigureS5a), in addition to a cytosolic fraction as previously reported in (Fukumoto et al., 2008). Interestingly, -catenin re-localized to the plasma membrane in DK as observed in NK (FigureS5b), suggesting that expression of the E-cadherin mutant stabilized

-

catenin expression (FigureS5b versus FigureS5d), as previously described by Wheelock’s group (Fukumoto et al., 2008). NK and DK transduced with a control MSCVIRES-GFP vector, expressing GFP alone, showed neither E-cadherin staining, nor

-

catenin re-localization to the plasma membrane under low-calcium conditions


(FigureS5c and d). These results excluded an intrinsic defect of

-catenin in AJ

formation in DK.

E-cadherin and desmosomal cadherins are retained in the ER in DK and in TGtreated NK Using an antibody against the intracellular region of E-cadherin, we confirmed the thick and punctuated staining of this molecule at the plasma membrane in DK and in TGtreated NK compared to NK (Figure4a and FigureS6a). In addition, we observed perinuclear staining of E-cadherin in DK and in TG-treated NK as compared to NK (Figure4a and FigureS6a), as seen for Dsg3 and Dsc3 (Figure2c and FigureS3c), suggesting intracellular retention of these molecules. Co-immunofluorescence of E-cadherin, Dsg3 or Dsc3 with calnexin, an ER marker demonstrated a significant increase of co-localisation signals in DK compared to NK (Figure4). Similar results for E-cadherin and Dsg3 were obtained in TG-treated NK compared to NK (FigureS6). E-cadherin co-localized with GM130, a marker of cis-Golgi (data not shown) to a lesser extent than calnexin in DK and TG-treated NK.

Miglustat restores mature adherens junction, desmosome formation and intercellular adhesion strength in DK The pharmacologic chaperone Miglustat (N-butyldeoxynojirimycin, Zavesca®) is a

-

glucosidase inhibitor, which has been proposed for clinical use in cystic fibrosis due to its capacity to correct defective trafficking of F508-CFTR (Norez et al., 2006). Remarkably, treatment of DK with Miglustat restored a narrow, intense, continuous pericellular staining of all AJ and desmosomal proteins in DK (Figure5a and b and FigureS7a-d), similar to NK (Figure2). In contrast, its inactive analogue (NB-DGJ) had no effect on E-


cadherin localization (Figure5c), supporting Miglustat specificity. Surface biotinylation assays confirmed the presence of a higher amount of E-cadherin at the plasma membrane following Miglustat treatment (Figure5d and FigureS7eS7e). Moreover, dispase dissociation assay showed that Miglustat-treated DK were less susceptible to mechanical dissociation than untreated DK (Figure5e). Altogether, these results provided evidence that Miglustat restored desmosome and AJ formation, and improves intercellular adhesion strength in DK. The effect of Miglustat was observed reproducibly in keratinocytes from the five DD patients studied.


Discussion This study demonstrates that DK display a constitutive ER stress response associated with defective AJ and desmosome

formation,

and that treatment

with the

pharmacological chaperone Miglustat restores both mature AJ and functional desmosomes resulting in increased intercellular adhesion strength. [Ca2+]ER plays a key role in ER-related functions (Federovitch et al., 2005), ER remodelling (Sammels et al., 2010), post-ER compartment architecture (Amodio et al., 2009), and ER-stress response (Schroder and Kaufman, 2005). Here, we show that loss of SERCA2 function from one mutated ATP2A2 allele in DK, which has previously be shown to induce a chronic reduction of ER Ca2+ stores in DK (Foggia et al., 2006; Pani et al., 2006), is sufficient to induce a constitutive ER stress under basal conditions, which is aggravated under stressor conditions. This is consistent with a previous report describing enhanced expression of the ER stress molecules calreticulin, CHOP and Bip in a DD patient keratinocytes in culture following TG-induction (Onozuka et al., 2006). Depletion of Ca2+ ER stores in other cell types has been shown to induce similar changes characteristic of ER-stress (Shaffer et al., 2004). Recently, another mechanism by which the aggregation of mutated SERCA2 pumps elicit ER stress has been proposed in DD (Wang et al., 2011). However, these results were obtained after high expression of transfected mutated SERCA2 pumps, a model which is distinct from DK in which the expression of the endogenous mutated SERCA2 allele is often reduced (data not shown and (Foggia et al., 2006), with no evidence for protein aggregation. The demonstration of ER stress in DK has important implications for the understanding of DD pathogenesis. In particular, DD runs a chronic relapsing course, with exacerbations by UVB irradiation, heat, friction and infections. The triggering factors could act as stressors by exacerbating constitutive ER stress. Consistent with this notion, UVB irradiation reduces ATP2A2 mRNA levels (Mayuzumi et al., 2005) and


induces ER stress response (Mera et al., 2010). Also, the presence of rounded keratinocytes (corps ronds), which is a hallmark of DD and is thought to correspond to apoptotic cells, may result from uncontrolled ER stress response leading to apoptosis. Finally, ER stress induced by Ca2+ depletion has been shown to control differentiation through XBP1 activation in TG-treated normal keratinocytes (Celli et al., 2011), suggesting that excessive ER stress could contribute to abnormal differentiation in DD. Desmosomal defects have long been described in DD, for which acantholysis leading to suprabasal clefts is a diagnostic feature (Burge and Hovnanian, 2011; Burge and Garrod, 1991). Impaired DP trafficking has been reported in DK, in TG- and SERCA2 siRNA-treated NK (Celli et al., 2012; Dhitavat et al., 2003; (Hobbs et al., 2011). Here we show that Dsg3 and Dsc3, which cooperate to form the adhesive interface in basal and supra-basal layers of the epidermis where cleavage occurs in DD (Garrod and Chidgey, 2008), also displayed an abnormal pattern suggestive of immature desmosomes. In contrast, Dsg1 and Dsc1, which are expressed in more differentiated layers of the epidermis, showed normal staining at the plasma membrane in DK, while Dsg2 and Dsc2 displayed an intermediate pattern compared to Dsg3 (data not shown). As AJ formation is one of the earliest events following the increase in [Ca2+]extra (Wheelock and Johnson, 2003), we investigated the expression of AJ components in DK. The specific pattern of distribution of E-cadherin in two rows of puncta at the edges of cell-cell contacts with actin filaments approaching initial cell contacts perpendicularly in DK, corresponds to the first step of AJ formation and is typical of immature AJ (Adams et al., 1998; Adams et al., 1996), providing evidence that AJ formation is also affected in DK. The observation that TG-treated NK recapitulated immature AJ and impaired desmosome formation (Celli et al., 2012; Dhitavat et al., 2003; Stuart et al., 1996) indicates that loss of SERCA2 function is sufficient to cause abnormal intercellular


junctions. Moreover, we show that intercellular adhesion strength is decreased in DK, a key feature likely to underlie acantholysis in DD patients. These results were obtained after 24 hours in high-calcium medium, demonstrating that AJ and desmosomes in DK are unable to mature even after prolonged exposure to calcium. Several lines of evidence support the conclusion that immature AJ and desmosome formation in DK is caused by ER retention leading to abnormal trafficking of E-cadherin, Dsg3 and Dsc3 to the plasma membrane: i. protein expression levels are comparable between DK and NK despite reduced staining at the plasma membrane; ii. E-cadherin, Dsg3 and Dsc3 show intracellular retention with co-staining with ER components; iii. endogenous -catenin properly relocates to the plasma membrane in the presence of an E-cadherin mutant targeted to the membrane, thus excluding an intrinsic defect of

-catenin; iv. treatment of NK with TG recapitulates ER retention of

these components, suggesting a mechanism involving ER stress and unfolded protein response (UPR); v. the effectiveness of Miglustat to reverse AJ and desmosome components trafficking abnormalities and to restore cell-cell adhesion. Importantly, primary keratinocytes from all Darier patients studied showed constitutive ER stress, defective AJ and desmosome formation with impaired adhesion strength and displayed improvement after Miglustat treatment. These observations suggest a shared disease mechanism in DD. Recently, Mauro’s group has studied the role of the shingolipid signaling pathway in TG-treated and in SERCA2b siRNA-treated keratinocytes. Sphingosine levels were enhanced, shingosine kinase 1 expression was decreased and inhibition of sphingosine lyase reversed abnormal E-cadherin and DP localization (Celli et al., 2012). Another study by Green’s group has shown defective PKC

activity in a TG-treated cell line,

resulting in decreased intercellular adhesion strength (Hobbs et al., 2011). Although the


mechanisms by which loss of SERCA2 function leads to abnormal sphingolipid pathway and defective PKC

activity remain to be determined, it is interestingly to note that

sphingosine has been reported as being a potent inhibitor of PKC (Hannun and Bell, 1989). These results reveal pleiomorphic effects of loss of SERCA2 function on epithelial cells homeostasis and further illustrate the complexity of intercrosses between different calcium signalling pathways. The current treatment for DD is not satisfactory, and relies mainly on the use of topical or oral retinoids, which are often not well tolerated and not fully effective (Burge, 1999). Miglustat, a pharmacological chaperone known to allow proper localization of mutated CFTR (Norez et al., 2006), restored mature AJ and desmosome formation and increased intercellular adhesion strength in DK. In cystic fibrosis, Miglustat prevents the interaction between

F508-CFTR and calnexin, a Ca2+-dependent ER chaperone

protein, enabling a functional F508-CFTR to reach the plasma membrane (Norez et al., 2006). Binding of glycosylated proteins to calnexin is a general mechanism involved in the quality-control system for protein folding in the ER. Since Miglustat did not reduce the ER stress response (ER expansion and eif2a phosphorylation were unchanged, data not shown), it is thus possible that calnexin, or another Ca2+-dependent lectin molecule, is involved in the quality control of E-cadherin, Dsg3, Dsc3. Screening and testing drugs allowing adhesion molecules to escape from UPR becomes a possible field of investigation for the development of treatment for DD. In conclusion, this work identifies Ca2+-depletion induced ER stress as an important determinant of DD pathogenesis and suggests classifying DD as an ER stressrelated disease. The observation that Miglustat is able to restore proper localization to the plasma membrane of non-mutated proteins retained in the ER supports a misfolding mechanism and provides a therapeutic option. Miglustat is an orally active orphan drug


(Zavzesca®) already successfully used in clinical trials for Gaucher disease and Cystic Fibrosis (Cox et al., 2003; Norez et al., 2006). Our findings therefore provide the basis for future clinical evaluation of Miglustat in Darier patients.

Materials and methods DD patients and molecular diagnostic. All clinical samples were obtained with informed written consent and in adherence to the Declaration of Helsinki guidelines. Genomic DNA was isolated from venous blood. The five Darier patients (D1 to D5) were screened for mutations in the ATP2A2 gene by sequencing the 21 exons and flanking splice sites of ATP2A2. ATP2A2 mutations identified at the heterozygous state are recapitulated in Table S1 and supplementary data. In this study, keratinocytes from these five DD patients were compared to normal keratinocytes from six matched healthy controls (NK).

Cell Culture. Biopsies from Darier patients were taken from unaffected (perilesional) skin of the groin. Control biopsies of normal skin (surgical skin margins from similar location) were obtained from five age- and sex-matched controls. 4-mm punch biopsies were explanted to isolate primary keratinocytes. Keratinocytes were grown in Green medium on a feeder layer of lethally irradiated mouse 3T3-J2 fibroblasts as described previously (Barrandon and Green, 1987; Rochat et al., 1994). For experiments, keratinocytes were grown in 0.06 mM CaCl2 EpiLife medium (Invitrogen) to 60-80% confluence and then incubated in 1.2 mM CaCl2 EpiLife medium for 4 and 6 hours to induce AJ and desmosomes formation, respectively. Experiments were performed at comparable passages between NK and DK.


Treatments with drugs. For ER stress induction, keratinocytes grown in 1.2 mM Ca2+ medium were incubated with thapsigargin (TG, Sigma) at 1µM for different time periods. For SERCA pump inhibition in NK, keratinocytes were pre-treated with 100nM of TG for 30 min in 0.06 mM Ca2+ medium. Medium was removed, and cells were incubated in 1.2 mM Ca2+ medium for an additional 4 hours to induce the formation of cell-cell contacts. For chaperone molecules treatment, keratinocytes at confluency were pre-treated with Miglustat and its inactive imino-sugar analogue (NB-DGJ, 500µM) for 24h before changing the medium for 1.2 mM Ca2+ medium. Results shown in this article were obtained with 500µM Miglustat, a concentration at which all cells present a polygonal phenotype, with no sign of toxicity. Increased plasma membrane localization of E-cadherin was also observed at lower concentration of Miglustat (100µM) although in a limited number of cells.

Immunofluorescence. Keratinocytes were cultured on coverslips, fixed with 4% paraformaldehyde and permeabilized with 0.2% TritonX-100. For DP staining, keratinocytes were fixed and permeabilized in ice-cold methanol. Cells were stained with primary antibodies (Table S2). Coverslips were examined using a Zeiss LSM 710 confocal microscope with a 40 Plan-Apochromat objective (1.3 oil) or a 63 PlanApochromat objective (1.4 oil), together with a 2

zoom. In figures, Z-stacks were

processed as maximal projections except for co-localization with calnexin. The quantification of intracellular compartment volumes and co-localization is performed as described in Supplementary materials and methods.

Western-Blot analyses. Western-blots were performed as reported in (Savignac et al., 2010).


RT-PCR. RNA was isolated from keratinocytes using TRIzol (Invitrogen) and reverse transcribed using hexamer primers and Moloney murine leukemia virus RT (Invitrogen). PCR amplification was performed using primers listed in TableS3.

Dispase mechanical dissociation assay. Keratinocytes were seeded in triplicate onto 6-well plates and cultured in medium containing 0.06mM calcium. Twenty four hours after increasing of extracellular Ca2+ concentration, cells were subjected to a dispase mechanical dissociation assay as described by (Ishii et al., 2005). Fragments were counted using a dissecting microscope (Leica MZ6) and final images were generated using Adobe Photoshop.

Surface biotinylation assay. Keratinocytes were seeded onto 6-well plates and cultured in medium containing 0.06mM Calcium. Four hours after Ca2+ concentration increase, keratinocytes were biotinylated as described in (Brittain et al., 2011).

Statistical analysis. Student’s unpaired two-tailed t-test was used for statistical analyses. * P < 0.05, ** P < 0.01 and *** P < 0.001.

Study approval. This study was conducted in compliance with the ethical principles of the Declaration of Helsinki and was approved by the medical ethical committee of the Purpan Hospital in Toulouse. All patients gave informed consent.


Conflict of interest The authors state no conflict of interest.

Acknowledgments We thank Lucette Pelletier and Jose Mejia (INSERM U1043) for helpful discussion and critical readings, Sophie Allart and Daniel Sapede from the imaging facility (IFR150, Institut Claude de Préval, Toulouse, F-31300 France), M.J. Wheelock for the kind gift of constitutive membrane E-cadherin plasmid and F. Becq for the kind gift of Miglustat. This work was supported by the “Programme National de Recherche en Dermatologie” (PNRderm), the “Société Française de Dermatologie” (SFD) and the “Fondation pour la Recherche Médicale” (FRM).


References Adams, C.L., Chen, Y.T., Smith, S.J. and Nelson, W.J. (1998) Mechanisms of epithelial cell-cell adhesion and cell compaction revealed by high-resolution tracking of Ecadherin-green fluorescent protein. J Cell Biol, 142, 1105-1119. Adams, C.L., Nelson, W.J. and Smith, S.J. (1996) Quantitative analysis of cadherincatenin-actin reorganization during development of cell-cell adhesion. J Cell Biol, 135, 1899-1911. Amodio, G., Renna, M., Paladino, S., Venturi, C., Tacchetti, C., Moltedo, O., Franceschelli, S., Mallardo, M., Bonatti, S. and Remondelli, P. (2009) Endoplasmic reticulum stress reduces the export from the ER and alters the architecture of post-ER compartments. Int J Biochem Cell Biol, 41, 2511-2521. Ashby, M.C. and Tepikin, A.V. (2001) ER calcium and the functions of intracellular organelles. Semin Cell Dev Biol, 12, 11-17. Barrandon, Y. and Green, H. (1987) Cell migration is essential for sustained growth of keratinocyte colonies: the roles of transforming growth factor-alpha and epidermal growth factor. Cell, 50, 1131-1137. Berridge, M.J., Bootman, M.D. and Roderick, H.L. (2003) Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol, 4, 517-529. Brittain, J.M., Chen, L., Wilson, S.M., Brustovetsky, T., Gao, X., Ashpole, N.M., Molosh, A.I., You, H., Hudmon, A., Shekhar, A., White, F.A., Zamponi, G.W., Brustovetsky, N., Chen, J. and Khanna, R. (2011) Neuroprotection against traumatic brain injury by a peptide derived from the collapsin response mediator protein 2 (CRMP2). J Biol Chem, 286, 37778-37792. Burge, S. (1999) Management of Darier's disease. Clin Exp Dermatol, 24, 53-56.


Burge, S. and Hovnanian, A. ( 2011) Acantholytic Disorders of the Skin. Chapter 51, 550-556, in Fitzpatrick’s Dermatology in General Medecine 2011, 8th edition, , vol. 1, McGrawHill Burge, S.M. and Garrod, D.R. (1991) An immunohistological study of desmosomes in Darier's disease and Hailey-Hailey disease. Br J Dermatol, 124, 242-251. Celli, A., Mackenzie, D.S., Crumrine, D.S., Tu, C.L., Hupe, M., Bikle, D.D., Elias, P.M. and Mauro, T.M. (2011) Endoplasmic reticulum Ca2+ depletion activates XBP1 and controls terminal differentiation in keratinocytes and epidermis. Br J Dermatol, 164, 16-25. Celli, A., Mackenzie, D.S., Zhai, Y., Tu, C.L., Bikle, D.D., Holleran, W.M., Uchida, Y. and Mauro, T.M. (2012) SERCA2-controlled Ca(2)-dependent keratinocyte adhesion and differentiation is mediated via the sphingolipid pathway: a therapeutic target for Darier's disease. J Invest Dermatol, 132, 1188-1195. Chen, Y.T., Stewart, D.B. and Nelson, W.J. (1999) Coupling assembly of the Ecadherin/beta-catenin complex to efficient endoplasmic reticulum exit and basallateral membrane targeting of E-cadherin in polarized MDCK cells. J Cell Biol, 144, 687-699. Cox, T.M., Aerts, J.M., Andria, G., Beck, M., Belmatoug, N., Bembi, B., Chertkoff, R., Vom Dahl, S., Elstein, D., Erikson, A., Giralt, M., Heitner, R., Hollak, C., Hrebicek, M., Lewis, S., Mehta, A., Pastores, G.M., Rolfs, A., Miranda, M.C. and Zimran, A. (2003) The role of the iminosugar N-butyldeoxynojirimycin (miglustat) in the management of type I (non-neuronopathic) Gaucher disease: a position statement. J Inherit Metab Dis, 26, 513-526. Dhitavat, J., Cobbold, C., Leslie, N., Burge, S. and Hovnanian, A. (2003) Impaired trafficking of the desmoplakins in cultured Darier's disease keratinocytes. J Invest Dermatol, 121, 1349-1355.


Dode, L., Andersen, J.P., Leslie, N., Dhitavat, J., Vilsen, B. and Hovnanian, A. (2003) Dissection of the functional differences between sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA) 1 and 2 isoforms and characterization of Darier disease (SERCA2) mutants by steady-state and transient kinetic analyses. J Biol Chem, 278, 47877-47889. Federovitch, C.M., Ron, D. and Hampton, R.Y. (2005) The dynamic ER: experimental approaches and current questions. Curr Opin Cell Biol, 17, 409-414. Foggia, L., Aronchik, I., Aberg, K., Brown, B., Hovnanian, A. and Mauro, T.M. (2006) Activity of the hSPCA1 Golgi Ca2+ pump is essential for Ca2+-mediated Ca2+ response and cell viability in Darier disease. J Cell Sci, 119, 671-679. Fukumoto, Y., Shintani, Y., Reynolds, A.B., Johnson, K.R. and Wheelock, M.J. (2008) The regulatory or phosphorylation domain of p120 catenin controls E-cadherin dynamics at the plasma membrane. Exp Cell Res, 314, 52-67. Garrod, D. and Chidgey, M. (2008) Desmosome structure, composition and function. Biochim Biophys Acta, 1778, 572-587. Getsios, S., Huen, A.C. and Green, K.J. (2004) Working out the strength and flexibility of desmosomes. Nat Rev Mol Cell Biol, 5, 271-281. Green, K.J. and Simpson, C.L. (2007) Desmosomes: new perspectives on a classic. J Invest Dermatol, 127, 2499-2515. Hakuno, M., Shimizu, H., Akiyama, M., Amagai, M., Wahl, J.K., Wheelock, M.J. and Nishikawa, T. (2000) Dissociation of intra- and extracellular domains of desmosomal cadherins and E-cadherin in Hailey-Hailey disease and Darier's disease. Br J Dermatol, 142, 702-711. Hannun, Y.A. and Bell, R.M. (1989) Regulation of protein kinase C by sphingosine and lysosphingolipids. Clin Chim Acta, 185, 333-345.


Hobbs, R.P., Amargo, E.V., Somasundaram, A., Simpson, C.L., Prakriya, M., Denning, M.F. and Green, K.J. (2011) The calcium ATPase SERCA2 regulates desmoplakin dynamics and intercellular adhesive strength through modulation of PKC alpha; signaling. Faseb J, 25, 990-1001. Ishii, K., Harada, R., Matsuo, I., Shirakata, Y., Hashimoto, K. and Amagai, M. (2005) In vitro keratinocyte dissociation assay for evaluation of the pathogenicity of antidesmoglein 3 IgG autoantibodies in pemphigus vulgaris. J Invest Dermatol, 124, 939-946. Jamora, C. and Fuchs, E. (2002) Intercellular adhesion, signalling and the cytoskeleton. Nat Cell Biol, 4, E101-108. Liu, L.H., Boivin, G.P., Prasad, V., Periasamy, M. and Shull, G.E. (2001) Squamous cell tumors in mice heterozygous for a null allele of Atp2a2, encoding the sarco(endo)plasmic reticulum Ca2+-ATPase isoform 2 Ca2+ pump. J Biol Chem, 276, 26737-26740. Mayuzumi, N., Ikeda, S., Kawada, H., Fan, P.S. and Ogawa, H. (2005) Effects of ultraviolet B irradiation, proinflammatory cytokines and raised extracellular calcium concentration on the expression of ATP2A2 and ATP2C1. Br J Dermatol, 152, 697-701. Mera, K., Kawahara, K., Tada, K., Kawai, K., Hashiguchi, T., Maruyama, I. and Kanekura, T. (2010) ER signaling is activated to protect human HaCaT keratinocytes from ER stress induced by environmental doses of UVB. Biochem Biophys Res Commun, 397, 350-354. Miyauchi, Y., Daiho, T., Yamasaki, K., Takahashi, H., Ishida-Yamamoto, A., Danko, S., Suzuki, H. and Iizuka, H. (2006) Comprehensive analysis of expression and function of 51 sarco(endo)plasmic reticulum Ca2+-ATPase mutants associated with Darier disease. J Biol Chem, 281, 22882-22895.


Niessen, C.M., Leckband, D. and Yap, A.S. (2011) Tissue organization by cadherin adhesion

molecules:

dynamic

molecular

and

cellular

mechanisms

of

morphogenetic regulation. Physiol Rev, 91, 691-731. Norez, C., Noel, S., Wilke, M., Bijvelds, M., Jorna, H., Melin, P., DeJonge, H. and Becq, F. (2006) Rescue of functional delF508-CFTR channels in cystic fibrosis epithelial cells by the alpha-glucosidase inhibitor miglustat. FEBS Lett, 580, 2081-2086. Onozuka, T., Sawamura, D., Goto, M., Yokota, K. and Shimizu, H. (2006) Possible role of endoplasmic reticulum stress in the pathogenesis of Darier's disease. J Dermatol Sci, 41, 217-220. Pani, B., Cornatzer, E., Cornatzer, W., Shin, D.M., Pittelkow, M.R., Hovnanian, A., Ambudkar, I.S. and Singh, B.B. (2006) Up-regulation of transient receptor potential canonical 1 (TRPC1) following sarco(endo)plasmic reticulum Ca2+ ATPase 2 gene silencing promotes cell survival: a potential role for TRPC1 in Darier's disease. Mol Biol Cell, 17, 4446-4458. Perez-Moreno, M., Jamora, C. and Fuchs, E. (2003) Sticky business: orchestrating cellular signals at adherens junctions. Cell, 112, 535-548. Pokutta, S. and Weis, W.I. (2007) Structure and mechanism of cadherins and catenins in cell-cell contacts. Annu Rev Cell Dev Biol, 23, 237-261. Rochat, A., Kobayashi, K. and Barrandon, Y. (1994) Location of stem cells of human hair follicles by clonal analysis. Cell, 76, 1063-1073. Sakuntabhai, A., Ruiz-Perez, V., Carter, S., Jacobsen, N., Burge, S., Monk, S., Smith, M., Munro, C.S., O'Donovan, M., Craddock, N., Kucherlapati, R., Rees, J.L., Owen, M., Lathrop, G.M., Monaco, A.P., Strachan, T. and Hovnanian, A. (1999) Mutations in ATP2A2, encoding a Ca2+ pump, cause Darier disease. Nat Genet, 21, 271-277.


Sammels, E., Parys, J.B., Missiaen, L., De Smedt, H. and Bultynck, G. (2010) Intracellular Ca2+ storage in health and disease: a dynamic equilibrium. Cell Calcium, 47, 297-314. Savignac, M., Edir, A., Simon, M. and Hovnanian, A. (2011) Darier disease : a disease model of impaired calcium homeostasis in the skin. Biochim Biophys Acta, 1813, 1111-1117. Savignac, M., Mellstrom, B., Bebin, A.G., Oliveros, J.C., Delpy, L., Pinaud, E. and Naranjo, J.R. (2010) Increased B cell proliferation and reduced Ig production in DREAM transgenic mice. J Immunol, 185, 7527-7536. Schroder, M. and Kaufman, R.J. (2005) The mammalian unfolded protein response. Annu Rev Biochem, 74, 739-789. Shaffer, A.L., Shapiro-Shelef, M., Iwakoshi, N.N., Lee, A.H., Qian, S.B., Zhao, H., Yu, X., Yang, L., Tan, B.K., Rosenwald, A., Hurt, E.M., Petroulakis, E., Sonenberg, N., Yewdell, J.W., Calame, K., Glimcher, L.H. and Staudt, L.M. (2004) XBP1, downstream of Blimp-1, expands the secretory apparatus and other organelles, and increases protein synthesis in plasma cell differentiation. Immunity, 21, 8193. Stuart, R.O., Sun, A., Bush, K.T. and Nigam, S.K. (1996) Dependence of epithelial intercellular junction biogenesis on thapsigargin-sensitive intracellular calcium stores. J Biol Chem, 271, 13636-13641. Thastrup, O., Dawson, A.P., Scharff, O., Foder, B., Cullen, P.J., Drobak, B.K., Bjerrum, P.J., Christensen, S.B. and Hanley, M.R. (1989) Thapsigargin, a novel molecular probe for studying intracellular calcium release and storage. Agents Actions, 27, 17-23.


Wang, Y., Bruce, A.T., Tu, C., Ma, K., Zeng, L., Zheng, P. and Liu, Y. (2011) Protein aggregation of SERCA2 mutants associated with Darier disease elicits ER stress and apoptosis in keratinocytes. J Cell Sci, 124, 3568-3580. Wheelock, M.J. and Johnson, K.R. (2003) Cadherins as modulators of cellular phenotype. Annu Rev Cell Dev Biol, 19, 207-235. Yoshida, H. (2007) ER stress and diseases. Febs J, 274, 630-658.


Figure legends Figure 1: ER stress response is increased in DK in basal conditions and after induction by thapsigargin. Keratinocytes were grown in 1.2mM Ca2+ and exposed to thapsigargin (TG, 1µM). Immunoblotting analysis using phospho-eIF2 and total-eIF2 (a), phospho-IRE1 and total-IRE1 (b). RT-PCR for Xbp1 mRNA (c). PGK was used as a loading control. a to c are representative of three independent experiments performed with five DK. (d) Immunofluoresence of Calnexin, ERGIC-53, GM130 and TGN46 in keratinocytes grown in 1.2mM Ca2+ for 4 hours. Scale bar = 20µm. (e) Quantification of intracellular compartment volumes. Numbers in brackets correspond to the numbers of cells analyzed for quantification. **P

Endoplasmic reticulum stress and endothelial dysfunction.

SERCA2 Haploinsufficiency in a Mouse Model of Darier Disease Causes a Selective Predisposition to Heart Failure.

Endoplasmic reticulum dysfunction in Alzheimer's disease.

Endoplasmic Reticulum Stress and Microvascular Endothelial Dysfunction in Diabetes.

Mechanism of endoplasmic reticulum stress-induced vascular endothelial dysfunction.

Rescue of Glaucomatous Neurodegeneration by Differentially Modulating Neuronal Endoplasmic Reticulum Stress Molecules.

Binge ethanol exposure causes endoplasmic reticulum stress, oxidative stress and tissue injury in the pancreas.

A small deletion in SERPINC1 causes type I antithrombin deficiency by promoting endoplasmic reticulum stress.

Efavirenz Causes Oxidative Stress, Endoplasmic Reticulum Stress, and Autophagy in Endothelial Cells.

Endoplasmic reticulum stress and cancer.

Endoplasmic reticulum stress in amelogenesis imperfecta and phenotypic rescue using 4-phenylbutyrate.

Endoplasmic reticulum stress in malignancy.

Miltirone exhibits antileukemic activity by ROS-mediated endoplasmic reticulum stress and mitochondrial dysfunction pathways.

MicroRNA-204 promotes vascular endoplasmic reticulum stress and endothelial dysfunction by targeting Sirtuin1.

Impaired oxidative endoplasmic reticulum stress response caused by deficiency of thyroid hormone receptor α.

The role of endoplasmic reticulum stress in endothelial dysfunction induced by homocysteine thiolactone.

Nephrin missense mutations: induction of endoplasmic reticulum stress and cell surface rescue by reduction in chaperone interactions.

Endoplasmic Reticulum Oxidative Stress Triggers Tgf-Beta-Dependent Muscle Dysfunction by Accelerating Ascorbic Acid Turnover.

Reduced endoplasmic reticulum stress-induced apoptosis and impaired unfolded protein response in TRPC3-deficient M1 macrophages.

Acetic Acid Causes Endoplasmic Reticulum Stress and Induces the Unfolded Protein Response in Saccharomyces cerevisiae.

Endoplasmic Reticulum Stress Links Oxidative Stress to Impaired Pancreatic Beta-Cell Function Caused by Human Oxidized LDL.

Oxidative and endoplasmic reticulum stress in β-cell dysfunction in diabetes.

Targeting endoplasmic reticulum stress in insulin resistance.

Cancer Microenvironment and Endoplasmic Reticulum Stress Response.