Loss of conserved Gsdma3 self-regulation causes autophagy and cell ...

Biochem. J. (2015) 468, 325–336

325

doi:10.1042/BJ20150204

Loss of conserved Gsdma3 self-regulation causes autophagy and cell death Peiliang Shi*1 , An Tang*1 , Li Xian*1 , Siyuan Hou*, Dayuan Zou*, Yasu Lv†, Zan Huang*‡, Qinghua Wang*, Anying Song*, Zhaoyu Lin* and Xiang Gao*2 *Key Laboratory of Model Animal for Disease Study of Ministry of Education, Model Animal Research Center, Collaborative Innovation Center of Genetics and Development, Nanjing University, Nanjing 210061, China †Department of Cell Biology and Medical Genetics, Research Center for Molecular Medicine, Institute of Cell Biology, Zhejiang University School of Medicine, Hangzhou 310000, China ‡Jiangsu Province Key Laboratory of Gastrointestinal Nutrition and Animal Health, Nanjing Agriculture University, Nanjing 210095, China

Gasdermin A3 (Gsdma3) was originally identified in association with hair-loss phenotype in mouse mutants. Our previous study found that AE mutant mice, with a Y344H substitution at the C-terminal domain of Gsdma3, display inflammation-dependent alopecia and excoriation [Zhou et al. (2012) Am. J. Pathol. 180, 763–774]. Interestingly, we found that the newly-generated null mutant of Gsdma3 mice did not display the skin dysmorphology, indicating that Gsdma3 is not essential for differentiation of epidermal cells and maintenance of the hair cycle in normal physiological conditions. Consistently, human embryonic kidney (HEK)293 and HaCaT cells transfected with wild-type (WT) Gsdma3 did not show abnormal morphology. However, Gsdma3 Y344H mutation induced autophagy. Gsdma3 N-terminal domain, but not the C-terminal domain, also displayed the similar

pro-autophagic activity. The Gsdma3 Y344H mutant protein and N-terminal domain-induced autophagy was associated with mitochondria and ROS generation. Co-expression of C-terminal domain reversed the cell autophagy induced by N-terminal domain. Moreover, C-terminal domain could be co-precipitated with N-terminal domain. These data indicated that the potential pro-autophagic activity of WT Gsdma3 protein is suppressed through an intramolecular inhibition mechanism. Studies on other members of the GSDM family suggested this mechanism is conserved in several sub-families.

INTRODUCTION

GSDMD. In mouse, there are three homologues in the A cluster [Gasdermin A1-3 (Gsdma1, 2, 3)], four homologues in the C cluster (Gsdmc1, 2, 3, 4), one homologue in the D cluster (Gsdmd) and no counterpart of human GSDMB [8]. GSDMA, GSDMC and GSDMD are all reported to be cancer suppressors, but GSDMB is considered to be an oncogene [9–12]. GSDMA is primarily expressed in the gastrointestinal tract and skin [10]. Reports on the function of GSDMA are very limited. GSDMA acts as a potential cancer suppressor in gastric cancer, is a target of LIM domain only 1 (LMO1) and involved in transforming growth factor (TGF)-β-induced apoptosis [11]. Previously, GWAS (genome-wide association study) and SNP (single nucleotide polymorphism) data have shown that human chromosome 17q21, on which GSDMA is located, is associated with asthma [13,14], type 1 diabetes [15] and other immunerelated diseases [16]. ORMDL3 (orosomucoid like 3)–GsdmA expression affects the secretion level of interleukin (IL)-17 in mononuclear cells, influencing the response of the body to allergens [13]. These reports suggest that GSDMA is an important regulator in immune activation. Gsdma3, one of the GSDMA orthologues in mice, is the only reported orthologue with a skin phenotype. Eight-point mutations have been reported in Gsdma3. They share similar phenotypes, including cicatricial alopecia accompanied by hair follicle differentiation defects, atrophy of the stratum granulosum (SG) and hyperkeratosis [17–23]. Previous reports suggest that Gsdma3 is associated with hair follicle development [21,24],

The skin is one of the largest organs in the human body. The epidermis consists of stratified squamous epithelium and has an outermost cornified layer that forms a protective barrier against the external environment. The skin prevents the loss of water, electrolytes and other substances, while resisting mechanical damage and invasion by pathogenic micro-organisms. There are four layers of epidermis. In decreasing order of depth, they are the stratum basale, stratum spinosum, stratum granulosum (SG) and stratum corneum (SC). The structure provides the most important physical barrier to maintain internal environmental homoeostasis in vertebrates. The function of the barrier relies on the normal terminal differentiation of keratinocytes and the complete tight junction structure [1]. Abnormal differentiation, such as in parakeratosis, hyperkeratosis or keratinocyte death, contributes to the disruption of the skin barrier [2]. Tight junctions contribute to the barrier function by sealing the intercellular space. Tight junction membrane proteins, such as claudin-1, Occludin and zona occludens protein 1 (ZO-1), are concentrated in the SG [3]. Tight junction deficiency in mice causes abnormal SC formation and SC barrier defects [4,5]. Epidermal barrier dysfunction is involved in inflammation and antigen-driven skin diseases, such as psoriasis [6,7]. The GSDM family is a novel gene family containing conserved N- and C-terminal domains. There are four human homologues in the family: GASDERMIN A (GSDMA), GSDMB, GSDMC and

Key words: autophagy, Gasdermin A3 , GSDM family, reactive oxygen species (ROS).

Abbreviations: DFNA5, deafness autosomal dominant non-syndromic sensorineural 5; DOX, doxycycline; ES cell, embryonic stem cell; Gsdm, Gasdermin; GWAS, genome-wide association study; HA, human influenza hemagglutinin; H&E stain, hematoxylin and eosin stain; HEK, human embryonic kidney; HEK293T-N, HEK293T cells with the DOX-inducible N-terminal domain lentivirus; IL , Interleukin; cell; IP, immunoprecipitation; JC-1, mitochondrial membrane potential probe; KO, knockout; LMO1, LIM domain only 1; 3-MA, 3-Methyladenine; NAC, N-cetylcysteine; ORMDL3 , orosomucoid like 3; PINK1, PTEN -induced putative kinase 1; ROS, reactive oxygen species; SC, stratum corneum; SG, stratum granulosum; SNP, Single-nucleotide polymorphism; TGF , transforming growth factor; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling; WT, wild-type; ZO -1, zona occludens protein 1. 1 These authors contributed equally. 2 To whom correspondence should be addressed (email [email protected]). c The Authors Journal compilation c 2015 Biochemical Society

326

P. Shi and others

tumour necrosis factor (TNF)α-induced apoptosis [25], skin inflammation [26,27] and immune-mediated destruction of bulge stem cells [23]. However, the physiological function of Gsdma3 is still unclear. The N- and C-termini of the GSDM family are conserved, but the mutations that result in a skin phenotype are located in the C-terminus of Gsdma3, with six mutants occurring in amino acids 343–348. This high-frequency mutation region is highly conserved in mammals, suggesting that this region plays an important part in the Gsdma3 structure. A leucine zipper sequence is predicted in the C-terminal domain, suggesting that Gsdma3 binds to DNA or other proteins and serves as a transcription or signalling factor. These observations indicate that the C-terminus plays an important role in the protein’s function. In the present study, we showed that the Gsdma3 AE mutation was a gain-of-function mutation and that Gsdma3 KO (knockout) mice did not show visible skin phenotypes. Furthermore, we analysed the functions of the N- and C-terminal domains of Gsdma3 and built a working model of Gsdma3. We also investigated the entire GSDM family to confirm the working model. MATERIALS AND METHODS Mice

The AE strain was generated in an ethyl nitrosourea mutagenesis programme which was reported previously [23]. The C57BL B/6N-Gsdma3tm1 Nju (Gsdma3 KO) mice were obtained from National Database of Mouse Genetic Resources. Gsdma3tm1(KOMP)Vlcg ES (embryonic stem) cells were purchased from Regeneron. Mice were maintained in specific pathogen-free (SPF) animal facility credited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). All animal welfare and experimental procedures are approved by the Animal Care and Use Committee of the Model Animal Research Center, Nanjing University.

Production of recombinant lentiviruses

The pTRIPZ-1–pTRIPZ-2 series plasmids, pVSVG and pSPAX2 plasmids were co-transfected overnight using Lipofectmine 2000 into 293FT cells. The next day, the medium was changed to 10 % FBS DMEM and viruses were harvested after 72 h transfected. The viruses were passed through 0.45-μm filters and stored at − 80 ◦ C. Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, mitochondrial membrane potential and reactive oxygen species detection

Apoptotic cells were detected by in situ cell death assay kit (Roche). Mitochondrial membrane potential was tested by mitochondrial membrane potential probe (JC-1) (Beyotime). Reactive oxygen species (ROS) level was detected by Reactive Oxygen Species Assay Kit (Beyotime). Doxycycline (DOX)induced wtGsdma3 and muGsdma3 stable transfected HEK293T and HaCaT cell lines infected by pTRIPZ lentivirus were used to check the apoptosis induction ability, ROS production and mitochondrial membrane potential of WtGsdma3 and MuGsdma3. DOX (Sigma) was added to the medium at a concentration of 1 μg/ml. Histological analysis

The mice dorsal skin was dissected and fixed overnight in 4 % paraformaldehyde at 4 ◦ C. Fixed tissue was dehydrated in 70 % ethanol and embedded in paraffin (Sigma) for paraffin sections and divided into 7-μm thick sections. Paraffin sections were stained with haematoxylin and eosin (H&E). For frozen sections, paraformaldehyde-fixed sections were soaked in 30 % sucrose PBS, embedded in OCT compound (Leica) and divided into 10μm thick sections.

Plasmid construction

Western blotting

Two kinds of lentivirus plasmids were constructed. pTRIPZ-1 was constructed by inserting Myc-T2A-mCherry sequences into ClaI and EcoRI sites of pTRIPZ plasmid. WtGsdma3 (wildtype, 1–1392) and muGsdma3 (1–1392) were amplified by PCR and cloned into the AgeI and ClaI sites of pTRIPZ-1 plasmid. The encoded protein has the Myc-T2A-mCherry element at the C-terminal. pTRIPZ-2 was constructed by inserting EGFP sequences into ClaI and EcoRI sites of pTRIPZ plasmid. WT (1–1392), MU (1–1392), N (1–786), CWT (787–1392) and CMU (786–1392) cDNA was amplified by PCR and cloned into the AgeI and ClaI sites of pTRIPZ-2 vector. The encoded protein has the EGFP tag at the C-terminal. N (1–786), CWT (787– 1392) and CMU (786–1392) were amplified by PCR and cloned into the ClaI and XbaI sites of the pCS2 + vector. The encoded Gsdma3 protein has the haemagglutinin (HA)-epitope tag at the C-terminal. Human GSDM family (GSDMA-D) and mouse Gsdm family (Gsdma1–2, Gsdmc1–4, Gsdmd) cDNA was amplified by PCR and cloned into the XhoI and EcoRI sites of the pEGFP-N1 vector by using Vazyme c112 one step cloning kit. Primers are listed in Supplementary Table S1.

Cultured cells and dorsal skin proteins were extracted on ice using lysis buffer [1 % NonidetP-40, 150 mM NaCl, 50 mM Tris/HCl (pH 7.4), 1 mM EDTA, 0.25 % Na-deoxycholate, 1 mM PMSF, 1 mM NaF, 1 mM Na3 VO4 and cocktail protein inhibitor). Protein samples were electrophoresed on SDS/PAGE with the appropriate percentage of acrylamide (10 %–15 %), transferred to Hybond-P polyvinylidene difluoride membrane (Amersham Bioscience) and incubated with primary antibodies. The primary antibodies contained rabbit-anti-HA, anti-GFP, anti-Bip, antiMyc antibodies from Santa Cruz Biotechnology (Santa Cruz); anti-LC3 antibody was from Novus (Littleton), mouse-anti-HA antibody was from Sigma; anti-p-STAT3, anti-STAT3, anti-p-p65, anti-p65, anti-Caspase-3, anti-cleaved-Caspase-3, anti-PARP and anti-p-E2F antibodies were from Cell Signaling Technology.

Cell culture and transfection

Human embryonic kidney (HEK)293 cells and HaCaT cells were cultured in Dulbecco’s modified Eagle’s Medium supplemented with 10 % FBS (Invitrogen) with 5 % CO2 at 37 ◦ C. HEK293T cells were transfected with Lipofectmine 2000 regents (Invitrogen). c The Authors Journal compilation c 2015 Biochemical Society

Immunofluorescence of cells and sections

For immunofluorescence of cells, cells were cultured on glass and fixed in 4 % paraformaldehyde for 10 min and then permeabilized in 0.1 % Triton X-100. Cells were blocked with 5 % BSA for 30 min and incubated with primary antibodies. Of sections, frozen sections were used for immunofluorescence. Frozen sections were performed as described above and the programme was same to the above mentioned. The primary antibodies contained anti-mouseK1, anti-K5 and anti-filaggrin antibodies from Convance; antiMyc, anti-GFP antibodies from Santa Cruz Biotechnology (Santa Cruz); anti-LC3 antibody was from Novus.

Loss of conserved Gsdma3 self-regulation causes autophagy and cell death

327

Immunoprecipitation

Contribution of apoptosis to muGsdma3 -induced cell death

Cell lysates were mixed with GFP-Trap agarose beads (Chromotek) and incubated for 4 h at 4 ◦ C, performed according to the manufacturer’s instructions. The immunoprecipitation (IP) products were tested using Western blotting protocol.

We sought to determine what pathway is important for muGsdma3-induced cell death. At first, we analysed apoptosis in stably transduced, DOX-induced wtGsdma3 and muGsdma3 HEK293T and HaCaT cell lines using the TUNEL assay. The percentage of apoptotic cells was significantly higher in DOXinduced muGsdma3 expressing HEK293T cells (7.8 + − 2.5 %) than in wtGsdma3 cells (0.62 + − 0.26 %; Figures 2A and 2B). Similar results were observed in HaCaT cells (Figure 2C). However, DOX-induced muGsdma3 expression in HEK293T cells caused most of the cell death. Furthermore, there was no significant difference in apoptosis markers in the stablytransduced, DOX-induced wtGsdma3 and muGsdma3 HEK293T and HaCaT cell lines by Western blotting (Figures 2D and 2E). Inflammation and ER-stress also may induce cell death and apoptosis [28,29]. We further tested the levels of inflammation and ER stress in cells. Inflammation and ER stress markers were not activated (Supplementary Figure S2). According to these results, apoptosis did not play an important role in muGsdma3induced cell death.

MTT assay

Cells were plated at 5000 per well in 96-well plate, incubated for 9–12 h after transfection or drugs treatments. Ten microlitres of MTT reagents (Vazyme) was added to the medium and incubated for 3 h. The medium was then removed from the well and 100 μl of DMSO was added (Sigma), the absorbance in each well was measured at 570 nm, including blank and control. The mean value from each treatment was labeled as Vt , the mean value from blank was labeled as Vb , and the mean value from control was labeled as VC . Cell viability was calculated by the formula: Cell viability = (Vt − Vb )/(Vc − Vb ).

Statistical analysis

Data were analysed with the Student’s t test and presented as mean + − S.E.M. P < 0.05 was considered statistically significant.

RESULTS The Gsdam3 AE point mutation caused cell death

AE mice exhibit progressive alopecia and hyperkeratosis phenotypes. Therefore, we generated Gsdma3 KO mice to investigate the function of Gsdma3 in vivo (Supplementary Figure S1). However, Gsdma3 KO homozygotes showed no obvious developmental skin abnormalities (Figure 1A), indicating that the AE mutation (Y344H) was a gain-of-function mutation. To investigate the function of Gsdma3 and the AE mutant, we generated an in vitro system. We constructed stably transduced, DOX-inducible wtGsdma3 and muGsdma3 HEK293T and HaCaT cell lines using a lentivirus carrying cDNA-Myc-T2A-mCherry sequences. DOX (1 mg/ml) was added to the medium to induce wtGsdma3 and muGsdma3 expression. HEK293T cells expressing the mutant cDNA showed obvious abnormal growth and increased cell death compared with the WT cells. Cells shrank, became round, were loosely arranged and adhered poorly after muGsdma3 expression (Figure 1B). DOX-induced wtGsdma3 and muGsdma3 in HaCaT cells showed the same phenotypes (Figure 1B). MuGsdma3-expressing cells showed marked decrease in cell vitality compared with control and wtGsdma3-expressing cells by MTT assay (Figures 1C and 1D). We sought to identify whether the level of expression or the subcellular localization was related to the observed cell death. The expression of the wtGsdma3 and muGsdma3 proteins was examined by Western blotting (Figure 1E). There were no differences between wtGsdma3 and muGsdma3 expressing cells. The subcellular localization of wtGsdma3 and muGsdma3 proteins was observed by confocal microscopy (Figure 1F). The wtGsdma3 and muGsdma3 proteins were both localized in the cytoplasm and not in the nucleus or cytomembrane. According to these results, we found that the Gsdam3 AE point mutation caused abnormal cell morphology and increased cell death in both HEK293 and HaCaT cell lines, which was not due to protein degeneration or changes in subcellular localization.

MuGsdma3 -induced cell death occurred mainly by autophagy

Autophagy is another cause of cell death and the phenotypes observed in autophagic cells, such as rounding and poor adhesion, were similar to those observed in cells expressing muGsdma3. We next tested the expression and lipidation levels of the autophagy marker LC3 in HEK293T and HaCaT cells after DOX-induced wtGsdma3 and muGsdma3 expression. LC3-I is converted into LC3-II through lipidation after autophagy is activated. Western blot analysis of LC3 showed a significant up-regulation of LC3-II in muGsdma3-expressing HEK293T and HaCaT cell lines (Figure 3A). Autophagy was significantly activated in muGsdma3-expressing cells. MuGsdma3-expressing cells showed considerable numbers of LC3 aggregates compared with wtGsdma3 cells through immunostaining, indicating that auto-phagosomes had formed in muGsdma3-expressing cells (Figure 3B). Approximately 56.3 + − 5.4 % of HaCaT cells and 62.7 + − 8.1 % of HEK293T cells expressing muGsdma3 showed LC3 aggregation, proportions that were much higher than in control or wtGsdma3-expressing cells (Figure 3C). We also tested the level of autophagy in vivo. Western blotting showed that LC3-II was significantly increased in the skin of AE compared with WT mice (Figure 3D). Immunohistochemical analysis of the epidermis of WT and AE mice showed high levels of LC3 expression and abnormal aggregation in the upper skin layers of AE mice (Figure 3E). We tested the distribution of tight junction proteins to evaluate the integrity of the skin barrier. We could observe staining of ZO-1, Claudin-1 and Occludin in WT skin in a normal region. The tight junction proteins formed a sharp and complete circle structure in the upper layer and hair follicles of WT skin, but the staining was dispersed in the upper layer and the complete circle structure disappeared in the skin of AE mice (Figure 3F; Supplementary Figure S3). These findings showed that the skin barrier of AE mice was disrupted. The abnormal region was the site of active autophagy, which may be the cause of skin barrier disruption. To further confirm that autophagy was the mechanism of cell death, we treated cells with an autophagy inhibitor, 3-MA. Four hours of treatment with 10 mM 3-Methyladenine (3-MA) significantly inhibited cell death in muGsdma3-expressing HEK293T cells (Figures 3G and 3H). These results demonstrate that autophagy is the main mechanism of muGsdma3-induced cell death. c The Authors Journal compilation c 2015 Biochemical Society

328

Figure 1

P. Shi and others

The Gsdam3 AE point mutation caused cell death

(A) Appearance of WT (left), AE (middle) and KO (right) mice was photographed at 2 months. KO mice showed no visible phenotype. (B) Cell morphology of HEK293T and HaCaT cell lines stably transfected with wtGsdma3 and muGsdma3 after DOX-induced gene expression. The muGsdma3 cell line show significant cell death. Scale bar = 20 μm. (C and D) Cell viability of HEK293T cells (C) and HaCaT cells (D) after 10 h DOX-induced wtGsdma3 and muGsdma3 expression by MTT assay. (E) Western blot analysis of Myc-tagged Gsdma3 expression in DOX-induced HEK293T and HaCaT cells. (E) Immunofluorescence of wtGsdma3 and muGsdma3 (Myc-tag, green) in HEK293T cells, nuclei were stained with DAPI (blue). Scale bar = 5 μm. ***P < 0.001.

The N-terminal domain of Gsdma3 was not the mutation region and promoted autophagy

The structure of Gsdma3 remains unclear. Secondary structure predictions revealed a propensity for the N-terminal 230 amino acids to form a domain consisting mainly of β-sheets and the 200 C-terminal amino acids to form an α-helix domain, which c The Authors Journal compilation c 2015 Biochemical Society

are connected by a flexible region. All Gsdma3 mutations that cause hair loss were localized to the C-terminus and N-terminalfused 3×Flag-tag could block muGsdma3-induced cell death (Supplementary Figure S4). These findings indicated that the N-terminus is important to Gsdma3 function. Accordingly, we investigated whether the toxicity of the mutant protein was due to the N-terminal domain or the mutations in the C-terminal domain.


Figure 2

329

Contribution of apoptosis to muGsdma3 -induced cell death

(A) Confocal images of HEK293T cell lines stably transfected with wtGsdma3 and muGsdma3 , stained with the in situ cell death detection kit, TM Red. Nuclei were stained with DAPI (blue). Significant more positive TUNEL staining in muGsdma3 expressing cells compared with wtGsdma3 . All cells were harvested after 10 h DOX-induced gene expression. Scale bar = 20 μm. (B) Results of quantification analysis for (A) DOX-induced muGsdma3 cells show more apoptosis compared with wtGsdma3 cells. (C) Result of quantification analysis for TUNEL in HaCaT cells. The result was similar to (B), 7 % positive TUNEL staining was counted in muGsdma3 expressing cells. (D) Western blot analysis of apoptosis signal pathway in DOX-induced wtGsdma3 and muGsdma3 HEK293T cell lines. There are no differences after 10 h DOX-induced gene expression. (E) Western blot analysis of apoptosis signal pathway in DOX-induced wtGsdma3 and muGsdma3 HaCaT cell lines. There are no differences after 2-days of DOX-induced gene expression.

We cloned the N-terminal domain (N, 1–786 bp), the wtGsdma3 C-terminal domain (CWT, 787–1395 bp) and the muGsdma3 Cterminus (CMU, 787–1395 bp, AE mutation) into the DOXinducible lentivirus expression vector, pTRIPZ-2 (Figure 4A). Stably transduced HEK293T cell lines were induced for 10 h with DOX. Expression was evaluated by Western blotting (Figure 4B). Expression of the N-terminal domain and muGsdma3 caused a significant decrease in cell viability in HEK293T cells compared with the other groups (Figure 4C). These results suggested that the N-terminal domain is the functional domain in the Gsdma3 protein and the C-terminal domain may act as a regulatory domain to inhibit the function of the N-terminal domain. We hypothesized that the mutant C-terminal domain lost its binding ability, which resulted in an N-terminal protein gain-of-function. CWT and CMU were separately transfected into HEK293T cells that were stably transduced with the DOX-inducible N-terminal domain lentivirus (HEK293T-N). HEK293T-N cell death was significantly inhibited by the CWT, but not the CMU (Figure 4D). A co-IP assay was used to confirm the binding between the N- and C-terminal domains (Figure 4E). The result showed that the N-terminal domain could bind to the CWT, but not the CMU. Our results suggested that the N-terminal

domain is the functional domain and that the C-terminal domain is an inhibitor of the N-terminal domain.

MuGsdma3 expression decreased mitochondrial activity and increased ROS levels

Mitochondrial dysfunction may result in the accumulation of cytotoxic mediators and cause cell death. Autophagy is a major sensor of ROS signalling in cellular responses. Autophagy was associated with mitochondrial activity [30,31]. We accordingly evaluated the mitochondrial membrane potential of HaCaT cells expressing wtGsdma3 and muGsdma3. Mitochondrial membrane potential was significantly decreased in muGsdma3-expressing HaCaT cells compared with wtGsdma3 cells (Figure 5A). The subcellular localization analysis of muGsdma3 showed cytosolic staining, such as that of wtGsdma3 and a slight colocalization with the mitochondrial dye, Mito-tracker. However, the N-terminal domain of Gsdma3 showed a significant colocalization with the dye. This result suggested that active Gsdma3 binds directly to the mitochondria and reduces the mitochondrial membrane potential (Figure 5B). Because a c The Authors Journal compilation c 2015 Biochemical Society

330

Figure 3

P. Shi and others

MuGsdma3 -induced cell death occurs mainly by autophagy

(A) Western blotting analysis of autophagy marker LC3 in DOX-induced wtGsdma3 and muGsdma3 HEK293T and HaCaT cell lines. MuGsdma3 cell lines show significant up-regulation of class-II LC3 in both HEK293T and HaCat cell lines. (B) Confocal images of wtGsdma3 compared with muGsdma3 stable-transfected HaCaT cells, stained with LC3 antibody. MuGsdma3 -expressing cells show significant aggregations of LC3 compared with wtGsdma3 cells. (C) Results of quantification analysis for (B), 56.3 + − 5.4 % cells showed significant aggregations of LC3. Scale bar in (B) = 5 μm. (D) Western blot analysis of autophagy marker LC3 in dorsal skins of WT and AE mice, AE mice showed significant up-regulation of class-II LC3 compared with WT mice. (E) Skin sections of WT and AE mice were stained with LC3 antibody (green) by immunofluorescence at 1 month. Autophagy is significantly up-regulated in AE mice skin. (F) Skin sections of WT and AE mice were stained with tight junction protein Claudin-1 antibody (green) by immunofluorescence at 1 month. The sharp and complete structure in upper layer and hair follicles of WT skin was disrupted in AE mice skin. Scale bar = 50 μm. (G) Four-hour 10 mM 3-MA treatment could significantly inhibit cell death induced by muGsdma3 . (H) Autophagy marker class-II- LC3 was significant down-regulation after 4 h of 10 mM 3-MA treatment. ***P < 0.001. c The Authors Journal compilation c 2015 Biochemical Society


Figure 4

331

The N-terminal domain of Gsdma3 was not the mutation region and promoted autophagy

(A) Graphical representation of the insert cDNA. (B) Western blot analysis of selected Gsdma3 constructs after 10 h DOX-induced expression in HEK293T cells. (C) Cell morphology of the stable-transfected HEK293T cell lines after 10 h DOX-induced related cDNA expression. Significant cell death was observed in MU- and N-expressing HEK293T cells. (D) Cell viability of HEK293T cells after 10 h DOX-induced related cDNA expression. (E) Western blotting analysis of autophagy marker LC3 in DOX-induced related cDNA expression in HEK293T cell lines. MU and N cell lines show significant up-regulation of class-II LC3. (F) Cell viability of HEK293T cells co-expressing N and CWT/CMU cDNA. CWT expression in DOX-induced N expressing cells could inhibit cell death. (G) Co-IP assay was preformed to identify the interaction between N and CWT/CMU. CWT could be pulled down by N, but CMU did not have this ability. ***P < 0.001. c The Authors Journal compilation c 2015 Biochemical Society

332

Figure 5

P. Shi and others

MuGsdma3 expression decreased mitochondrial activities and increased ROS level

(A) JC-1 was used to detect mitochondrial membrane potential. MuGsdma3 expressing HaCaT cells showed a significant decrease of mitochondrial membrane potential. (B) Immunofluorescence of Gsdma3 (green) and mitochondria (red). Mitochondria were stained using Mito-tracker. Significant co-localization of Gsdma3 and mitochondria was observed in muGsdma3 and N expressing cells (arrows). Scale bar = 5 μm. (C) ROS level was significantly increased in muGsdma3 expressing HEK293T cells. The superoxide indicator dihydroethidium was used to measure ROS. Green fluorescence intensity was measured as ROS levels. (D) ROS level was also significantly increased in muGsdma3 expressing HaCaT cells. (E) Cell viability after 10 h DOX-induced wtGsdma3 and muGsdma3 expression in HEK293T cells cultured with antioxidant (0.1 mM NAC). NAC can rescue cell death induced by muGsdma3 expression. (F) Western blot analysis of autophagy marker LC3 in DOX-induced wtGsdma3 and muGsdma3 HEK293T cell lines with 0.1 mM NAC. ***P < 0.001.

lower mitochondrial membrane potential would lead to an increase in ROS levels [32], we evaluated the ROS levels in our stably transduced, DOX-inducible cells using the superoxide indicator dihydroethidium. The green fluorescence signal of dihydroethidium was significantly elevated in both c The Authors Journal compilation c 2015 Biochemical Society

muGsdma3-expressing HEK293T and HaCaT cells compared with wtGsdma3-expressing cells (Figures 5C and 5D). Nacetylcysteine (NAC), an antioxidant, was used to suppress ROS levels. Cell viability measurements showed that NAC rescued muGsdma3-induced cell death (Figure 5E). Western blot analysis


of the autophagy marker LC3 in NAC-treated, DOX-induced wtGsdma3 and muGsdma3 HEK293T cell lines indicated that NAC suppressed muGsdma3-induced autophagy (Figure 5F). Together, these results suggested that muGsdma3 disrupted the mitochondria and induced high levels of ROS. ROS was one reason for autophagy. GSDM family members function similarly to induce cell death

Members of the GSDM family have conserved N- and Cterminal sequences (Figure 6A), but their function is unclear. We constructed a working model of Gsdma3 self-regulation and sought to determine whether the mechanism was conserved across family members by performing similar functions and sharing common regulation. We cloned human GSDMA-D and entire mouse Gsdma-d full-length cDNAs, N- and Cterminal domains into the pEGFP-N1 plasmid. Plasmids were transfected separately into 293T cells. Western blotting showed that GSDM family proteins were expressed normally in 293T cells (Figure 6C; Supplementary Figure S5A). After 12 h, we found that the N-terminal domains of GSDM family members had a distinct function in inducing cell death, with the exception of GSDMB (Figure 6B and 6E; Supplementary Figure S5C). We evaluated the levels of LC3 and determined that the human and mouse N-terminal domains of GSDMA and GSDMD showed strong autophagic activities (Figure 6D; Supplementary Figure S5B). We further sought to determine whether GSDM family members have structural and functional similarities. N-terminal domain-induced cell death should be suppressed by the C-terminal domains of other family members if there is some similarity in the GSDM family. We transfected the GSDMA-D C-terminal domains into HEK293T-N cells and added DOX to induce the expression of the Gsdma3 N-terminal domain. Gsdma3 N-terminal domaininduced cell death was blocked by the C-termini of the other family members (Figure 6E). We also transfected GSDMA-D N-terminal domains into HEK293T–CWT and added DOX to induce Gsdma3 C-terminal domain expression (Figure 6F). Cell death inhibition was also observed. We speculated that the Nand C-terminal domains of the GSDM family share a conserved functional domain and control cell death through binding between the N-and C-terminal domains. DISCUSSION

Gsdma3 is specifically expressed in the upper layer of skin, where keratinization occurs. However, it is not clear whether Gsdma3 function is involved in the regulation of keratinization. Keratinization is a process in which stem cells gradually differentiate into the dead, flattened cell remnants in the uppermost layer of corneocytes [33]. In this process, keratinocytes gradually lose mitochondria, nuclei and other organelles [34,35]. We report in the present study that the N-terminus and AE mutant proteins induced decreases in mitochondrial activity and autophagy, similar to the process of keratinization. However, the Gsdma3 null mice did not show visible skin phenotypes, suggesting that the autophagy ‘activation’ induced by N-terminus domain may require additional physiological or pathological signals. We speculate that Gsdma3 is probably a stress-associated activator to maintain the normal skin structure in response to the environment. Indeed, derangements of autophagy have been reported in many autoimmune skin disorders and infectious skin diseases [36]. Many reports show that mutation of Gsdma3 leads

333

to skin inflammation, but the mechanisms remain unknown. We found an enhanced autophagy activity in mutant mice skin, along with inflammation. Most studies suggested activating the autophagy pathway suppresses the inflammation [37,38]. However, we believe that the uncontrolled continuous autophagy in Gsdma3 mutant mice may cause cell death, which in turn activates the inflammation. In future studies, it will be interesting to screen the endogenous signal that may ‘open-up’ Gsdma3 protein and ‘activate’ its N-terminal domain function. If so, the activation of keratinocytes autophagy will be regulated according to the environmental or physiological conditions. Additional studies will be required to examine the skin phenotype of Gsdma3 null mice under different environmental insults, including UV radiation, pathogens, mechanical stresses and toxic chemicals. We also observed the disruptive skin barrier in mutant mice, along with the defective tight junctions between keratinocytes. Skin barrier defects also promote allergen sensitization and several diseases are associated with skin barrier defects, such as atopic dermatitis, Netherton’s syndrome and psoriasis [1,39,40]. We suspect that skin barrier defects may also contribute to inflammation. However, the connection between autophagy induction and skin barrier defect needs to be established in a future study. The mutation or loss of the C-terminal domain of Gsdma3 results in abnormal protein aggregates in the mitochondria, a significant decline in mitochondrial activity and increase in ROS level. The precise mechanism by which N-terminal domain is recruited to mitochondria remains unclear. Using tandem affinity purification, Zanon et al. [41] reported Parkin might interact directly with GSDMA (human form of Gsdma3). When mitochondria are damaged, Parkin can bind with PTEN-induced putative kinase 1 (PINK1) and be recruited to mitochondrial outer membrane and initiate the mitophagy process [42,43]. So it will be crucial to test whether GSDMA can regulate this process. Our results showed that the Gsdma3 C-terminus bound to the Nterminal domain to inhibit autophagy. The AE mutant C-terminus lost the ability to inhibit the N-terminus. The mutation region in Gsdma3 (amino acids 343–359) is highly conserved in the GSDM family and we propose that this region is central to the function of the C-terminal domain in mediating the connection between the N- and C-terminal domains. Another homologue, deafness autosomal dominant non-syndromic sensorineural 5 (DFNA5) and GSDM share the conserved DFNA5–GSDM domain region (the N-terminal domain). Previous reports suggest that DFNA5 consists of two globular domains separated by a hinge region; the N-terminal domains display cell death induction activity, whereas the second domain might serve as a regulatory domain [44,45]. That report and the results presented in the present study ascribe a similar structure and conserved function to the GSDM family. Gsdma1 KO mice also show no visible phenotypes and overexpression of Gsdma1 in mouse skin induced hyperplasia and skin inflammation [46] similar to that in Gsdma3 KO and mutant mice. These observations indicate that GSDM family proteins have similar functions. Our study has shown that Gsdma3 is a potential pro-autophagic gene in skin. By screening the entire GSDM family, we also confirmed that the N-terminus of GSDMfamily proteins, with the exception of GSDMB, has the ability to induce cell death and autophagy and that GSDMA and GSDMD are the most potent in inducing autophagy. We propose that GSDM family proteins perform similar functions, based on their common structure. In conclusion, we have found that the N-terminal domain of Gsdma3 showed strong pro-autophagic activity in vivo and in vitro. Moreover, the C-terminus functions as a regulatory c The Authors Journal compilation c 2015 Biochemical Society

334

Figure 6

P. Shi and others

GSDM family members function similarly to induce cell death

(A) Phylogeny of the GSDM family members in human (hosa), mouse (mumu), rat (rano) and pan troglodytes (patr). (B) Cell morphology of HEK293T cell lines transfected with human full-length GSDM , N- and C-terminal domain after 12 h. N-terminal domain of GSDMA , GSDM C and GSDM D showed significant cell death. Scale bar = 20 μm. (C) Western blotting analysis of the expression levels of human full-length GSDM N- and C-terminal domains after 12 h. (D) Western blotting analysis of autophagy marker LC3 in HEK293T cells transfected with human full-length GSDM N- and C-terminal domains after 12 h. Class-II LC3 was significant up-regulated in cells transfected with GSDMA /GSDM C/GSDM D N-terminal domains. (E) Cell viability of HEK293T cells transfected with human full-length GSDM N- and C-terminal domains after 12 h. N-terminal domain of GSDMA , GSDM C and GSDM D could induce cell death in HEK293T cells. (F) Gsdma3 N-terminal-induced cell death could be blocked by C-terminal domains of GSDMA /GSDM B/GSDM C/GSDM D. (G) N-terminal domains of GSDMA /GSDM B/GSDM C/GSDM D-induced cell death could be blocked by C-terminal domain of Gsdma3 . *P < 0.05, **P < 0.01 and ***P < 0.001. c The Authors Journal compilation c 2015 Biochemical Society


domain and inhibited the function of the N-terminal domain. This self-regulation mechanism was conserved in the GSDM family. Our study advances the understanding of the GSDM family and hints at its physiological function.

AUTHOR CONTRIBUTION Peiliang Shi, An Tang, Li Xian, Siyuan Hou, Dayuan Zou, Yasu Lv, Zan Huang, Qinghua Wang and Anying Song performed experiments, analysed data and prepared figures. Peiliang Shi and Zhaoyu Lin contributed to the experimental design and drafted the paper. Xiang Gao designed the study, edited and revised the paper. All authors interpreted data and approved the final paper.

ACKNOWLEDGEMENTS We thank Dr Yue Zhou for the preliminary work of the present study and Dr Minsheng Zhu, Dr Shuai Chen and Dr Ying Cao for their useful suggestions.

FUNDING This work was supported by the State Key Development Program for Basic Research of China [grant number 2011C13944104]; and the National Key Technology R&D Program [grant number 2011BAI15B02].

REFERENCES 1 De Benedetto, A., Kubo, A. and Beck, L.A. (2012) Skin barrier disruption: a requirement for allergen sensitization? J. Invest. Dermatol. 132, 949–963 CrossRef PubMed 2 Elias, P.M. (2005) Stratum corneum defensive functions: an integrated view. J. Invest. Dermatol. 125, 183–200 PubMed 3 Morita, K., Itoh, M., Saitou, M., Ando-Akatsuka, Y., Furuse, M., Yoneda, K., Imamura, S., Fujimoto, K. and Tsukita, S. (1998) Subcellular distribution of tight junction-associated proteins (occludin, ZO-1, ZO-2) in rodent skin. J. Invest. Dermatol. 110, 862–866 CrossRef PubMed 4 Furuse, M., Hata, M., Furuse, K., Yoshida, Y., Haratake, A., Sugitani, Y., Noda, T., Kubo, A. and Tsukita, S. (2002) Claudin-based tight junctions are crucial for the mammalian epidermal barrier: a lesson from claudin-1-deficient mice. J. Cell Biol. 156, 1099–1111 CrossRef PubMed 5 Igawa, S., Kishibe, M., Murakami, M., Honma, M., Takahashi, H., Iizuka, H. and Ishida-Yamamoto, A. (2011) Tight junctions in the stratum corneum explain spatial differences in corneodesmosome degradation. Exp. Dermatol. 20, 53–57 CrossRef PubMed 6 Kirschner, N., Poetzl, C., von den Driesch, P., Wladykowski, E., Moll, I., Behne, M.J. and Brandner, J.M. (2009) Alteration of tight junction proteins is an early event in psoriasis: putative involvement of proinflammatory cytokines. Am. J. Pathol. 175, 1095–1106 CrossRef PubMed 7 Ye, L., Lv, C., Man, G., Song, S., Elias, P.M. and Man, M.Q. (2014) Abnormal epidermal barrier recovery in uninvolved skin supports the notion of an epidermal pathogenesis of psoriasis. J. Invest. Dermatol. 134, 2843–2846 CrossRef PubMed 8 Tamura, M., Tanaka, S., Fujii, T., Aoki, A., Komiyama, H., Ezawa, K., Sumiyama, K., Sagai, T. and Shiroishi, T. (2007) Members of a novel gene family, Gsdm, are expressed exclusively in the epithelium of the skin and gastrointestinal tract in a highly tissue-specific manner. Genomics 89, 618–629 CrossRef PubMed 9 Komiyama, H., Aoki, A., Tanaka, S., Maekawa, H., Kato, Y., Wada, R., Maekawa, T., Tamura, M. and Shiroishi, T. (2010) Alu-derived cis-element regulates tumorigenesis-dependent gastric expression of GASDERMIN B (GSDMB). Genes Genetic Syst. 85, 75–83 CrossRef 10 Saeki, N., Usui, T., Aoyagi, K., Kim, D.H., Sato, M., Mabuchi, T., Yanagihara, K., Ogawa, K., Sakamoto, H., Yoshida, T. and Sasaki, H. (2009) Distinctive expression and function of four GSDM family genes (GSDMA-D) in normal and malignant upper gastrointestinal epithelium. Genes Chromosomes Cancer 48, 261–271 CrossRef PubMed 11 Saeki, N., Kim, D.H., Usui, T., Aoyagi, K., Tatsuta, T., Aoki, K., Yanagihara, K., Tamura, M., Mizushima, H., Sakamoto, H. et al. (2007) GASDERMIN, suppressed frequently in gastric cancer, is a target of LMO1 in TGF-beta-dependent apoptotic signalling. Oncogene 26, 6488–6498 CrossRef PubMed 12 Saeki, N., Kuwahara, Y., Sasaki, H., Satoh, H. and Shiroishi, T. (2000) Gasdermin (Gsdm) localizing to mouse chromosome 11 is predominantly expressed in upper gastrointestinal tract but significantly suppressed in human gastric cancer cells. Mamm. Genome 11, 718–724 CrossRef PubMed

335

13 Lluis, A., Schedel, M., Liu, J., Illi, S., Depner, M., von Mutius, E., Kabesch, M. and Schaub, B. (2011) Asthma-associated polymorphisms in 17q21 influence cord blood ORMDL3 and GSDMA gene expression and IL-17 secretion. J. Allergy Clin. Immunol. 127, 1587–1594 e1586 CrossRef 14 Yu, J., Kang, M.J., Kim, B.J., Kwon, J.W., Song, Y.H., Choi, W.A., Shin, Y.J. and Hong, S.J. (2011) Polymorphisms in GSDMA and GSDMB are associated with asthma susceptibility, atopy and BHR. Pediatr. Pulmonol. 46, 701–708 CrossRef PubMed 15 Saleh, N.M., Raj, S.M., Smyth, D.J., Wallace, C., Howson, J.M., Bell, L., Walker, N.M., Stevens, H.E. and Todd, J.A. (2011) Genetic association analyses of atopic illness and proinflammatory cytokine genes with type 1 diabetes. Diabetes Metab. Res. Rev. 27, 838–843 CrossRef PubMed 16 Halapi, E., Gudbjartsson, D.F., Jonsdottir, G.M., Bjornsdottir, U.S., Thorleifsson, G., Helgadottir, H., Williams, C., Koppelman, G.H., Heinzmann, A., Boezen, H.M. et al. (2010) A sequence variant on 17q21 is associated with age at onset and severity of asthma. Eur. J. Hum. Genet. 18, 902–908 CrossRef PubMed 17 Sato, H., Koide, T., Masuya, H., Wakana, S., Sagai, T., Umezawa, A., Ishiguro, S., Tamai, M. and Shiroishi, T. (1998) A new mutation Rim3 resembling Re(den) is mapped close to retinoic acid receptor alpha (Rara) gene on mouse chromosome 11. Mamm. Genome 9, 20–25 CrossRef PubMed 18 Porter, R.M., Jahoda, C.A., Lunny, D.P., Henderson, G., Ross, J., McLean, W.H., Whittock, N.V., Wilson, N.J., Reichelt, J., Magin, T.M. and Lane, E.B. (2002) Defolliculated (dfl): a dominant mouse mutation leading to poor sebaceous gland differentiation and total elimination of pelage follicles. J. Invest. Dermatol. 119, 32–37 CrossRef PubMed 19 Runkel, F., Marquardt, A., Stoeger, C., Kochmann, E., Simon, D., Kohnke, B., Korthaus, D., Wattler, F., Fuchs, H., Hrabe de Angelis, M. et al. (2004) The dominant alopecia phenotypes Bareskin, Rex-denuded, and Reduced Coat 2 are caused by mutations in gasdermin 3. Genomics 84, 824–835 CrossRef PubMed 20 Wood, G.A., Flenniken, A., Osborne, L., Fleming, C., Vukobradovic, I., Morikawa, L., Xu, Q., Porter, R., Adamson, S.L., Rossant, J. and McKerlie, C. (2005) Two mouse mutations mapped to chromosome 11 with differing morphologies but similar progressive inflammatory alopecia. Exp. Dermatol. 14, 373–379 CrossRef PubMed 21 Tanaka, S., Tamura, M., Aoki, A., Fujii, T., Komiyama, H., Sagai, T. and Shiroishi, T. (2007) A new Gsdma3 mutation affecting anagen phase of first hair cycle. Biochem. Biophys. Res. Commun. 359, 902–907 CrossRef PubMed 22 Kumar, S., Rathkolb, B., Budde, B.S., Nurnberg, P., de Angelis, M.H., Aigner, B. and Schneider, M.R. (2012) Gsdma3(I359N) is a novel ENU-induced mutant mouse line for studying the function of Gasdermin A3 in the hair follicle and epidermis. J. Dermatol. Sci. 67, 190–192 CrossRef PubMed 23 Zhou, Y., Jiang, X., Gu, P., Chen, W., Zeng, X. and Gao, X. (2012) Gsdma3 mutation causes bulge stem cell depletion and alopecia mediated by skin inflammation. Am. J. Pathol. 180, 763–774 CrossRef PubMed 24 Lunny, D.P., Weed, E., Nolan, P.M., Marquardt, A., Augustin, M. and Porter, R.M. (2005) Mutations in gasdermin 3 cause aberrant differentiation of the hair follicle and sebaceous gland. J. Invest. Dermatol. 124, 615–621 CrossRef PubMed 25 Lei, M., Bai, X., Yang, T., Lai, X., Qiu, W., Yang, L. and Lian, X. (2012) Gsdma3 is a new factor needed for TNF-alpha-mediated apoptosis signal pathway in mouse skin keratinocytes. Histochem. Cell Biol. 138, 385–396 CrossRef PubMed 26 Zulfakar, M.H., Alex, A., Povazay, B., Drexler, W., Thomas, C.P., Porter, R.M. and Heard, C.M. (2011) In vivo response of GsdmA3Dfl/ + mice to topically applied anti-psoriatic agents: effects on epidermal thickness, as determined by optical coherence tomography and H&E staining. Exp. Dermatol. 20, 269–272 CrossRef PubMed 27 Ruge, F., Glavini, A., Gallimore, A.M., Richards, H.E., Thomas, C.P., O’Donnell, V.B., Philpott, M.P. and Porter, R.M. (2011) Delineating immune-mediated mechanisms underlying hair follicle destruction in the mouse mutant defolliculated. J. Invest. Dermatol. 131, 572–579 CrossRef PubMed 28 Kaltschmidt, B., Kaltschmidt, C., Hofmann, T.G., Hehner, S.P., Droge, W. and Schmitz, M.L. (2000) The pro- or anti-apoptotic function of NF-kappaB is determined by the nature of the apoptotic stimulus. Eur. J. Biochem. 267, 3828–3835 CrossRef PubMed 29 Szegezdi, E., Logue, S.E., Gorman, A.M. and Samali, A. (2006) Mediators of endoplasmic reticulum stress-induced apoptosis. EMBO Rep. 7, 880–885 CrossRef PubMed 30 Okamoto, K. and Kondo-Okamoto, N. (2012) Mitochondria and autophagy: critical interplay between the two homeostats. Biochim. Biophys. Acta 1820, 595–600 CrossRef PubMed 31 Lee, J., Giordano, S. and Zhang, J. (2012) Autophagy, mitochondria and oxidative stress: cross-talk and redox signalling. Biochem. J. 441, 523–540 CrossRef PubMed 32 Kirkinezos, I.G. and Moraes, C.T. (2001) Reactive oxygen species and mitochondrial diseases. Semin. Cell Dev. Biol. 12, 449–457 CrossRef PubMed 33 Lippens, S., Denecker, G., Ovaere, P., Vandenabeele, P. and Declercq, W. (2005) Death penalty for keratinocytes: apoptosis versus cornification. Cell Death Differ 12 (Suppl 2), 1497–1508 CrossRef PubMed 34 Candi, E., Schmidt, R. and Melino, G. (2005) The cornified envelope: a model of cell death in the skin. Nat. Rev. Mol. Cell Biol. 6, 328–340 CrossRef PubMed c The Authors Journal compilation c 2015 Biochemical Society

336

P. Shi and others

35 Mascia, F., Denning, M., Kopan, R. and Yuspa, S.H. (2012) The black box illuminated: signals and signaling. J. Invest. Dermatol. 132, 811–819 CrossRef PubMed 36 Yu, T., Zuber, J. and Li, J. (2015) Targeting autophagy in skin diseases. J. Mol. Med. 93, 31–38 CrossRef PubMed 37 Levine, B., Mizushima, N. and Virgin, H.W. (2011) Autophagy in immunity and inflammation. Nature 469, 323–335 CrossRef PubMed 38 Deretic, V., Saitoh, T. and Akira, S. (2013) Autophagy in infection, inflammation and immunity. Nat. Rev. Immunol. 13, 722–737 CrossRef PubMed 39 Cork, M.J., Danby, S.G., Vasilopoulos, Y., Hadgraft, J., Lane, M.E., Moustafa, M., Guy, R.H., Macgowan, A.L., Tazi-Ahnini, R. and Ward, S.J. (2009) Epidermal barrier dysfunction in atopic dermatitis. J. Invest. Dermatol. 129, 1892–1908 CrossRef PubMed 40 Guttman-Yassky, E., Nograles, K.E. and Krueger, J.G. (2011) Contrasting pathogenesis of atopic dermatitis and psoriasis–part I: clinical and pathologic concepts. J. Allergy Clin. Immunol. 127, 1110–1118 CrossRef PubMed 41 Zanon, A., Rakovic, A., Blankenburg, H., Doncheva, N.T., Schwienbacher, C., Serafin, A., Alexa, A., Weichenberger, C.X., Albrecht, M., Klein, C. et al. (2013) Profiling of Parkin-binding partners using tandem affinity purification. PLoS One 8, e78648 CrossRef PubMed Received 18 February 2015/25 March 2015; accepted 31 March 2015 Published as BJ Immediate Publication 31 March 2015, doi:10.1042/BJ20150204

c The Authors Journal compilation c 2015 Biochemical Society

42 Eiyama, A. and Okamoto, K. (2015) PINK1/Parkin-mediated mitophagy in mammalian cells. Curr. Opin. Cell Biol. 33C, 95–101 CrossRef 43 Youle, R.J. and Narendra, D.P. (2011) Mechanisms of mitophagy. Nat. Rev. Mol. Cell Biol. 12, 9–14 CrossRef PubMed 44 Van Rossom, S., Op de Beeck, K., Franssens, V., Swinnen, E., Schepers, A., Ghillebert, R., Caldara, M., Van Camp, G. and Winderickx, J. (2012) The splicing mutant of the human tumor suppressor protein DFNA5 induces programmed cell death when expressed in the yeast Saccharomyces cerevisiae . Front. Oncol. 2, 77 CrossRef PubMed 45 Op de Beeck, K., Van Camp, G., Thys, S., Cools, N., Callebaut, I., Vrijens, K., Van Nassauw, L., Van Tendeloo, V.F., Timmermans, J.P. and Van Laer, L. (2011) The DFNA5 gene, responsible for hearing loss and involved in cancer, encodes a novel apoptosis-inducing protein. Eur. J. Hum. Genet. 19, 965–973 CrossRef PubMed 46 Tanaka, S., Mizushina, Y., Kato, Y., Tamura, M. and Shiroishi, T. (2013) Functional conservation of Gsdma cluster genes specifically duplicated in the mouse genome. G3 3, 1843–1850 CrossRef PubMed