Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/yexcr

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1

0014-4827/$ - see frohttp://dx.doi.org/10.1

Abbreviations: 3-homologous protein

nCorresponding auE-mail address: j

Research Article

Autophagy inhibition induces podocyte apoptosis byactivating the pro-apoptotic pathway of endoplasmicreticulum stress

Li Fang, Xiurong Li, Yuan Luo, Weichun He, Chunsun Dai, Junwei Yangn

Center for Kidney Disease, Second Affiliated Hospital, Nanjing Medical University, Nanjing 210003, Jiangsu Province, China

a r t i c l e i n f o r m a t i o n

Article Chronology:

Received 31 July 2013

Received in revised form10 December 2013Accepted 3 January 2014Available online 11 January 2014

Keywords:

AutophagyApoptosisEndoplasmic reticulum stressPodocyte

nt matter & 2014 Elsevier016/j.yexcr.2014.01.001

MA, 3-methyladenine; S; siRNA, small interferinthor.wyang@njmu.edu.cn (J. Ya

a b s t r a c t

Podocyte apoptosis is a major factor inducing podocyte depletion that predicts the progressivecourse of glomerulosclerosis. However, the molecular mechanisms underlying podocyte apopto-

sis are still not well understood. Autophagy is a lysosomal degradation system involving thedegradation and recycling of obsolete, damaged, or harmful cytoplasmic materials and organelles.Recent advances in the understanding of the molecular processes contributing to autophagy haveprovided insight into the relationship between autophagy and apoptosis. However, their cross-talk remains largely obscure until now. Here, we found that podocytes both in vivo and in vitroalways exhibited high basal levels of autophagy, whereas autophagy inhibition could inducepodocyte apoptosis, suggesting the pro-survival role of autophagy in podocytes. Besides, wefound that autophagy inhibition by 3-methyladenine (3-MA) could induce the activation ofendoplasmic reticulum stress even without any external stimulations, whereas knockdown ofCHOP could effectively improve podocyte apoptosis and down-regulated expression of slit-diaphragm proteins induced by autophagy inhibition. Collectively, this study demonstrated that

autophagy might act as a crucial regulatory mechanism of apoptotic cell death by modulating thebalance between the pro-survival pathway and the pro-apoptotic pathway of endoplasmicreticulum stress, which might provide a novel mechanistic insight into the interface betweenautophagy and apoptosis in the progression of podocyte injury.

& 2014 Elsevier Inc. All rights reserved.

Introduction

Podocyte is the most spectacular cell type whose location,architecture and relevance are unique [1]. In 1989, Fries et al.were the first to suggest that an inability of podocytes for cellreplication in the adult represented the major factor underlyingnephron degeneration [2]. In subsequent years, numerous studies

Inc. All rights reserved.

TZ, streptozotocin; TBGg RNA; ER, endoplasmic r

ng).

have pointed to podocyte depletion as a hallmark of both primaryand secondary forms of glomerulosclerosis [3]. Podocyte apopto-sis, the early glomerular phenotype in the progression of diabeticand non-diabetic renal diseases, has now been suggested tocontribute to progressive podocyte depletion and glomerulo-sclerosis [4]. However, until now, the molecular mechanismsunderlying podocyte apoptosis in chronic progressive glomerular

, tail blood glucose; CHOP, CCAAT-enhancer-binding proteineticulum

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 291

diseases remain largely unknown. Therefore, deeper understand-ing of the cell biology of podocyte will help us to develop rules onhow to protect podocytes from apoptosis and halt the progressionof chronic renal diseases.

Autophagy is a physiologically regulated and evolutionarilyconserved process mostly implicated in the recycling of portionsof cytosol and in the removal of superfluous or damagedorganelles [5]. Because of its internal needs to maintain cellularhomeostasis [6], autophagy is essential for the survival, differ-entiation and development. Dysregulated autophagy has beensuggested to play pathogenic roles in a variety of diseasesincluding cancer, neurodegeneration, diabetes, aging, heart dis-ease and so on [7]. It is believed that autophagy is important incell death decisions and could protect cells by preventing themfrom undergoing apoptosis [8,9]. Nevertheless, autophagy mayalso result in cell death called “autophagic cell death” throughexcessive self-digestion and degradation of essential cellularconstituents under certain circumstances [10,11]. Thus, it seemsthat the relationship between autophagy and apoptosis is com-plex [12].

Our previous studies have demonstrated that high basal levelsof autophagy were essential for podocyte biology and defectiveautophagy happening in the progression of diabetic nephropathymight facilitate podocyte injury [13]. In this study, we furtherdemonstrated that defective autophagy was associated withpodocyte apoptosis and the interplay between autophagy andapoptosis was important in the pathogenesis of diabetic podocyteinjury. Since autophagy played an important role in endoplasmicreticulum quality control [14,15], here we indicated that defectiveautophagy commonly seen in diabetic nephropathy could inducethe activation of endoplasmic reticulum stress. Knockdown ofCHOP expression by siRNA not only prevented autophagyinhibition-induced podocyte apoptosis, but also restored autop-hagy inhibition-induced down-regulated expression of slit-diaphragm proteins. Our studies identified autophagy as a reg-ulatory mechanism of apoptotic cell death by modulating the pro-survival pathway of endoplasmic reticulum stress. Autophagyinhibition in diabetes and aging might induce podocyte apoptosisby activating the pro-apoptotic pathway of endoplasmic reticulumstress.

Materials and methods

Animal model

As described previously [13], Male CD-1 mice weighed �18–22 gwere acquired from the Specific Pathogen-Free (SPF) LaboratoryAnimal Center of Nanjing Medical University. According to theguidelines of the Institutional Animal Care and Use Committee ofthe National Institutes of Health at Nanjing Medical University,animals were treated humanely and rendered diabetic by intra-peritoneal injections of 50 mg/kg streptozotocin daily (STZ, Sigma,St. Louis, MO) for 5 days. For the sham-operated group, normalsaline was administered. Tail blood glucose (TBG) levels weremonitored consecutively and diabetic status was confirmed by themanifestation of weight loss, polyuria, and TBG level greater than500 mg/dl. The diabetic mice were sacrificed at 12-weeks afterthe treatment; serum and urine were collected and the kidneyswere harvested for various analyses.

Human renal biopsy specimens

Renal biopsy specimens were obtained from patients undergoingdiagnostic evaluation at the Division of Nephrology, The SecondAffiliated Hospital of Nanjing Medical University. A total of 6subjects were (age range, 29–45) selected from our database withthe criteria of having at least 10 glomeruli in the block availablefor histological sectioning. All biopsy specimens were evaluatedby a single pathologist who was unaware of the results ofmolecular studies. The patients (n¼3) with mild mesangiumproliferative glomerulonephritis whose proteinuria were about0.3270.174 g/24 h were chosen as a control group. The diabeticpatients (n¼3) whose proteinuria were about 5.2171.863 g/24 hwere chosen as a diabetes group.

Cell culture and treatment

The conditionally immortalized mouse podocyte cell line waskindly provided by Dr. Zhihong Liu (Research Institute of Nephrol-ogy, Nanjing General Hospital of Nanjing Military Command,Nanjing, Jiangsu Province, China) while their cell lines wereprovided by Dr. Peter Mundel (Mount Sinai School of Medicine,New York, NY, USA) and described previously [16]. Cells werecultured at the permissive temperature (33 1C) in RPMI-1640medium supplemented with 10% fetal bovine serum (Invitrogen,Grand Island, NY) and recombinant interferon-γ (Invitrogen,Carlsbad, CA, USA). To induce differentiation, podocytes weregrown under non-permissive conditions at 37 1C for 10–14 days inthe absence of interferon-γ. After serum starvation for 16 h, cellswere exposed to the treatment for indicated time periods.

Immunofluorescence staining

Indirect immunofluorescence staining was performed accordingto an established procedure [17]. For immunofluorescence stain-ing of kidney sections, cryosections at 5 μm thickness wereprepared and fixed in cold methanol/acetone (1:1) for 10 min.After being blocked with 2% normal donkey serum in PBS for40 min, the sections were incubated with primary antibodiesagainst LC3 (no. 2775S, Cell Signaling), laminin (L9393, Sigma-Aldrich), podocin (P0372, Sigma-Aldrich), WT1 (sc-192, SantaCruz Biochemical) and cleaved caspase-3 (no. 9661S, Cell Signal-ing), respectively, in PBS containing 1% BSA overnight at 4 1C. As anegative control, the primary antibody was replaced with eithernonimmune mouse or rabbit IgG, corresponding to species of theprimary antibodies. Sections were then washed thoroughly in PBSand incubated with FITC and TRITC-conjugated secondary anti-body (Sigma-Aldrich) at a dilution of 1:500 in PBS containing 1%BSA(bovine serum albumin) in the dark for 1 h. After beingthoroughly washed with PBS, slides were mounted with VECTA-SHIELDs anti-fade mounting media (Vector Laboratories, Burlin-game, CA) and viewed with a Nikon Eclipse 80i Epi-fluorescencemicroscope equipped with a digital camera (DS-Ri1, Nikon). Ineach experimental setting, images were captured with identicallight exposure parameters and aperture settings.Cells cultured on cover slips were washed twice with cold PBS

and fixed with cold methanol/acetone (1:1) for 10 min at �20 1C.Following three extensive washings with PBS, the cellswere blocked with 2% normal donkey serum in PBS buffer for40 min at room temperature. Then the cells were incubated with

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1292

anti-cleaved caspase-3 (no. 9661S, Cell Signaling), anti-GRP78 (sc-13968, Santa Cruz Biochemical) and anti-CHOP (sc-7351, SantaCruz Biochemical) respectively, followed by staining with FITC- orTRITC-conjugated secondary antibodies. Finally, cells were doublestained with 4,6-diamidino-2-phenylindole to visualize thenuclei. Slides were also viewed with a Nikon Eclipse 80i Epi-fluorescence microscope equipped with a digital camera (DS-Ri1,Nikon).

GFP-LC3 transfection

Podocytes were plated at a density of 2�105 on glass cover slipsin six-well plates and cultured up to 70% confluence. Transienttransfections with GFP-LC3 plasmid (kindly provided by Dr.Wenxing Ding, Department of Pharmacology, Toxicology & Ther-apeutics, University of Kansas Medical Center) were carried out byusing Lipofectamine 2000 (Invitrogen, Carlsbad, CA) as permanufacturer0s recommendation. Microphotographs of GFP-LC3fluorescence were obtained with a fluorescence microscope.

Fig. 1 – Podocytes exhibit high basal levels of autophagy in vivo. (immunostained with anti-LC3 antibody (green) to identify autoph(red) to sever as a marker for the basement membrane (B). MerginImmunofluorescence staining shows the occurrence of autophagyidentify autophagosomes (D) and podocin staining to recognize popresented in F. (H–J): Immunofluorescence staining further demonillustrated by colocalization of LC3 staining to identify autophagoMerging the LC3 and podocin staining images is presented in J.

The detection of punctated staining of GFP-LC3 from the diffusestaining indicated the formation of autophagosomes.

Apoptosis assay by flow cytometry

Podocyte apoptosis and necrosis were determined by flowcytometry exactly as recently reported [18]. Briefly, podocyteswere trypsinized and resuspended in 1� binding buffer (AnnexinV-FITC Apoptosis Detection Kit, Sigma). Annexin V staining wasapplied for 10 min at room temperature (see producer protocol)and before analysis 5 μl PI was added. For the assay, 20,000–25,000 cells were analyzed by flow cytometry (BD Biosciences).

Apoptosis detection by fluorescent dye H-33258/PIdouble-staining

Nuclear chromatin morphology was examined by staining withthe fluorescent dye H-33258 (5 ug/ml) as described previously[19]. Briefly, cells cultured on cover slips were washed with

A–C): Frozen sections of kidneys from CD-1 mice wereagosomes (A), followed by staining with anti-laminin antibodyg the LC3 and laminin staining images is presented in C. (D–F):in podocytes, as illustrated by colocalization of LC3 staining todocytes (E). Merging the LC3 and WT1 staining images is

strates that podocytes exhibit high basal levels of autophagy, assomes (H) and podocin staining to recognize podocytes (I).

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 293

phosphate-buffered saline, fixed with 3% paraformaldehyde, andfollowed by staining with H-33258 for 10min at 37 1C and thenincubated with 100 ul PI (50 ng/ml) in Dark Place at 4 1C for 20min.Slides were viewed with a Nikon Eclipse 80i Epi-fluorescencemicroscope equipped with a digital camera (DS-Ri1, Nikon).

Western immunoblot analysis

Immortalized mouse podocytes and 3-MA-treated cells werelysed with SDS sample buffer (62.5 mM Tris �HCl, pH 6.8, 2%SDS, 10% glycerol, 50 mM DTT, and 0.1% bromophenol blue).Samples were heated at 100 1C for 5–10 min before loading andwere separated on precast 10 or 5% SDS polyacrylamide gels (Bio-Rad, Hercules, CA). Detection of protein expression by westernblotting was carried out according to the established protocolsdescribed previously [20]. The primary antibodies used were asfollows: Anti-Beclin1 (sc-48381, Santa Cruz Biochemical), anti-Atg12 (sc-68884, Santa Cruz Biochemical), anti-LC3β (no. 2775S,Cell Signaling), anti-cleaved-caspase 3 (no. 9664, Cell Signaling),anti-caspase 3 (no. 9662, Cell Signaling), anti-phospho-eIF2 alpha(sc-101670, Santa Cruz Biochemical), anti-total eIF2 alpha (sc-133132, Santa Cruz Biochemical), anti-caspase 12 (sc-5627, SantaCruz Biochemical), anti-cleaved-caspase 12 (no. 2202S, Cell Sig-naling), anti-CHOP (sc-7351, Santa Cruz Biochemical), anti-podocin (P0372, Sigma-Aldrich) and anti-α-tubulin (T6074,Sigma-Aldrich). Quantification was performed by measurement

Fig. 2 – Defective autophagy is virtually always accompanied by poImmunofluorescence detection of LC3 (green), cleaved caspase-3 (mouse model, suggesting that podocyte apoptosis is accompanied(E–H): the group of 12-week diabetic mice. (I–N) Double immunoflin nondiabetic patients and diabetic patients, further demonstratidefection. (I–K): nondiabetic renal biopsies specimens; (L–N): diab

of the intensity of the signals with the aid of the NationalInstitutes of Health Image software package.

RT-PCR analysis

Total RNA was prepared using a TRIzol RNA isolation systemaccording to the instructions specified by the manufacturer(Invitrogen). The first strand of cDNA was synthesized using2 ug of RNA in 20 ul of reaction buffer using moloney leukemiavirus-RT (Promega, Madison, WI) and random primers at 42 1C for30 min. PCR was performed using a standard PCR kit on 1 ulaliquots of cDNA and HotStarTaq polymerase (Promega) withspecific primer pairs. The sequences of primer pairs were asfollows: WT1 (forward) 50–GGG CAG TTA GTA AAT GGG AAC–30

and (reverse) 50–GGA AAG TGT GCA AAC TGC TTC–30; Nephrin(forward) 50–TTC AGC AAG GAG ACC TTC AAG AA–30 and (reverse)50–GCC CCT CAA TCC ACA GCT T–30; GADPH (forward) 50–CCA TGTTCG TCA TGG GTG TGA ACC A–30 and (reverse) 50–GCC AGT AGAGGC AGG GAT GAT GTT C–30; The PCR products were sizefractionated on a 1.0% agarose gel and detected by NA-green(D0133, Beyotime) staining.

Statistical analysis

Western blotting, RT-PCR, and immunofluorescence staining wereall repeated at least three times independently. For Western blot

docyte apoptosis under diabetic conditions. (A–H):red), and DAPI (blue) of control and diabetic glomerular inby autophagy defection in diabetic mice. (A–D): control group;uorescence staining for LC3 (green) and cleaved caspase-3 (red)ng that podocyte apoptosis is accompanied by autophagyetic renal biopsies specimens.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1294

analysis, quantitation was performed by scanning and analyzingthe intensity of the hybridization signals using the NIH Imaginesoftware. Statistical analysis was performed using the Sigma Statsoftware (Jandel Scientific Software, San Rafael, CA). Comparisonsbetween groups were made using one-way ANOVA, followed byStudent0s t-test. Po0.05 was considered significant.

Results

In vivo, podocytes exhibit high basal levels of autophagy

To explore the role of autophagy in the regulation of podocyteapoptosis, we first used immunofluorescence staining to evaluatethe basal levels of autophagy in normal renal tissue. As shown inFig. 1(A–C), immunofluorescence staining showed that the autop-hagosomes (stained with the anti-LC3 antibody, green) werelocalized on the epithelial side of the glomerular basementmembrane (stained with the anti-Laminin antibody, red) wherethe podocytes line. To further confirm, double immunostainingexperiments was performed with antibodies to the podocytemarkers WT1 and podocin, coupled with the autophagosomemarker LC3. The colocalization of LC3 and WT1/podocin revealedthat podocytes exhibited high basal levels of autophagy in vivo.

Defective autophagy is always accompanied by apoptosisin podocytes under diabetic conditions

Since autophagy and apoptosis have now been shown to beinterconnected by several molecular nodes of crosstalk andpodocyte apoptosis has been indicated as an early phenomenonat the onset of diabetic nephropathy which triggers the initiationof glomerular lesions, to determine whether defective autophagy

Fig. 3 – High levels of autophagy are induced under serum starvatin vitro experiments. (A) Western blot showed the time course ofpodocytes were incubated in 2 ml serum-free culture medium forimmunoblotted with antibodies against Beclin-1, Atg12-5, LC3 andAtg12-5 protein abundances in different times as indicated. Relativα-tubulin. The data are presented as means7S.E. of three experimfluorescence staining of GFP-LC3 under serum starvation. Mouse pThe appearance of punctuated staining (marked by the white arro

is involved in podocyte apoptosis in diabetic nephropathy, weassessed caspase-3 activation and its substrate cleavage byimmunofluorescent microscopy to evaluate podocyte apoptosis.Accompanying a gradually weakening staining for LC3 protein inthe diabetic glomeruli not only in mouse model but also in renalbiopsy specimens of patients (Fig. 2 A, E, I and L), there werealways much more cleaved caspase 3-positive cells (Fig. 2 B, F, Jand M). These findings suggested a correlation between defectiveautophagy and podocyte apoptosis in the development of podo-cyte injury under diabetic conditions.

In vitro, podocytes exhibit high levels of autophagy underserum starvation

To further confirm the relationship between autophagy andapoptosis, in vitro, we first evaluated the basal autophagic activityof podocytes using western blotting technique. We found that theexpressions of autophagy related proteins such as Beclin-1, Atg12-5 and LC3 were up-regulated distinctly as early as 6 h and peakedat 48 h during serum starvation which was one of the mostfrequently performed procedures in vitro experiments (Fig. 3Aand B). Podocytes transfected with a green fluorescent protein(GFP)-LC3 fusion plasmid was examined by fluorescence micro-scopy, as shown in Fig. 3C and D, the appearance of punctuatedstaining indicated the autophagosome-associated LC3-II. Thesedata suggested that high level of autophagy might be a process bywhich the cells adapt their metabolism situations physiologically.

Inhibition of autophagy by 3-MA induces podocyteapoptosis in vitro

Since previous in vivo data has suggested a correlation betweendefective autophagy and podocyte apoptosis, next, the effects of

ion which is one of the most frequently performed proceduresthe induction of autophagy under serum starvation. Mousevarious periods of time as indicated. Cell lysates wereα-tubulin, respectively. (B) Graphic presentation of Beclin-1 ande Beclin-1 or Atg12-5 level was reported after normalizing withents. nPo0.05 versus controls. (C and D) Representativeodocytes were incubated with 2 ml culture media for 48 h.w) indicates autophagosome-associated LC3-II.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 295

3-methyladenine (3-MA, a well-known autophagy inhibitor) on thepodocytes apoptosis were examined. During the period of autophagyinhibition treated with 3-MA, we found the cells start to float andnot adherent anymore (data not shown). Fig. 4A shows representa-tive flow cytometry data with the abscissa and ordinate representingthe fluorescence intensity of annexin V and PI, respectively. Incontrol group, 6.071.015% of podocytes were annexin V-positive/PI-negative cells, representing early apoptotic podocytes. 3-MA couldincrease apoptosis significantly at 6 h (14.871.201%, Po0.05), 12 h(16.772.014%, Po0.05) and 24 h (24.272.521%, Po0.05). Besides, asshown in Fig. 4(C–F), after treatment with 3-MA for 12 h, about 30%podocytes underwent nuclear condensation and fragmentation asvisualized by H-33258 staining and about 18% podocytes were dead

Fig. 4 – Inhibition of autophagy by 3-MA induces podocyte apoptopodocytes exposed to 3-MA (2 mmol/L) for various periods of timefluorescence intensity of annexin V and propidium iodide (PI), respdeath. Bar graph represents the mean percentages7SD of annexinnPo0.05 versus controls; (C and E) Nuclear condensation of apoptwere incubated with 2 mmol/L 3-MA for 12 h. C: control group; E: 3shown by propidium iodide staining. Podocytes were incubated wgroup; G: Western blot demonstrates the kinetics of pro-caspase-3antibodies against full-length caspase-3, cleaved caspase-3 and α-tfor various periods of time as indicated. (H and I): Representativeand 3-MA treated podocytes. Podocytes were incubated with 2 mm

as visualized by PI staining. Since caspase-3 is a key mediator ofapoptosis and indispensable for apoptotic chromatin condensationand DNA fragmentation in all cell types examined [21], we furtherperformed western blot analysis with cleaved caspase-3 antibodyand caspase-3 antibody to evaluate caspase-3 activation and itssubstrate cleavage. Cleaved caspase-3 is up-regulated after 3-MAtreatment as early as 1 h and sustained until 24 h (Fig. 4G). Asillustrated in Fig. 4H and I, cleaved caspase-3 exhibiting greenfluorescence was also detected in the cytoplasm around nucleolusafter 3-MA treatment. Therefore, it seems that autophagy might becytoprotective, at least under serum starvation, and autophagyinhibition by 3-MA might lead to apoptotic cell death even withoutany additional stimulus.

sis in vitro. A: representative flow cytometry results foras indicated. The abscissa and ordinate represent the

ectively. B: quantitative analysis of 3-MA-induced podocyte cellV-positive/PI-negative (early apoptotic) podocytes (n¼3).

otic cells shown by H33258-PI fluorescent staining. Podocytes-MA treated group; (D and F) DNA staining of apoptotic nucleiith 2 mmol/L 3-MA for 12 h. D: control group; F: 3-MA treatedactivation. Cell lysates were immunoblotted with specificubulin, respectively, after incubation with 2 mmol/L of 3-MAimmunofluorescence staining for cleaved caspase-3 in controlol/L 3-MA for 12 h. H: control group; I: 3-MA treated group.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1296

Inhibition of autophagy by 3-MA induces endoplasmicreticulum stress in podocytes

Although studies have suggested that there was crosstalkbetween the autophagic and apoptotic pathways, the mechanismsmediating the complex counter-regulation of apoptosis andautophagy are not yet fully understood [22]. Recently, growingevidence has indicated that autophagy plays an important role inthe ER quality control system [23–25]. Thus, we reasoned that theactivation of endoplasmic reticulum stress might be involved inthe autophagy inhibition induced apoptosis since the qualitycontrol system was destroyed. With respect to this hypothesis,the phosphorylation of eIF2α, the activation of caspase-12 fromprocaspase-12 and the expression of CHOP were detected withwestern blotting. As shown in Fig. 5A, we observed that CHOPexpression which is typically up-regulated during severe ER stresswas dramatically increased after 6 h. At this time point, moreover,there was also a significant increase in the amount of phosphory-lated eIF2α and cleaved caspase-12. Microscopic analyses alsorevealed that immunofluorescence staining of GRP78 and CHOPwas markedly increased at 24 h after 3-MA treatment (Fig. 5B–M).

Knockdown of CHOP by siRNA alleviates 3-MA inducedapoptosis

To further confirm the involvement of endoplasmic reticulumstress in autophagy inhibition induced apoptosis, we next inves-tigated whether suppression of CHOP by siRNA could abrogateautophagy inhibition induced apoptosis since CHOP has beenconsidered to play a primary role in ER stress-induced apoptosis[26]. The results showed that the up-regulation of cleaved

Fig. 5 – In vitro, inhibition of autophagy by 3-MA induces endoplablot results showed that 3-MA (2 mmol/L) could induce the phospCHOP/GADD153 in a time-dependent manner in podocytes. (B–G) Iautophagy by 3-MA (2 mmol/L) induced the GRP78 expression in pinhibition of autophagy by 3-MA (2 mmol/L) triggered the nuclear

caspase-3 was effectively abolished in cells transfected with CHOPsiRNA (Fig. 6A and B), whereas treatment with control siRNA hadno effect. Consistent with the results of western blotting, thechanges of H-33258 & PI staining positive-cell number (Fig. 6C–H)and the cleaved caspase-3 fluorescence staining (Fig. 6I–Q) wereidentical. Thus, it appears clear that the pro-apoptotic pathway ofendoplasmic reticulum stress is involved in autophagy inhibitioninduced apoptosis.

Knockdown of CHOP by siRNA restores autophagyinhibition induced down-regulated expressionof slit-diaphragm proteins

We also examined the expression of WT1 and slit-diaphragmproteins. As shown in Fig. 7A and B, representative RT-PCR resultsshowed that the diminished mRNA expressions of Nephrin andWT1 caused by 3-MA could be effectively restored in cellstransfected with CHOP siRNA, whereas treatment with controlsiRNA had no effect. In Fig. 7C and D, western blot analysis alsoshowed that the diminished expressions of podocin caused by3-MA could be restored by knockdown of CHOP. When immuno-fluorescent staining analysis of WT1 was performed, similar resultwas obtained (Fig. 7E–G). Considered together, it suggests thatautophagy may act to achieve a balance between pro-survival andpro-apoptosis of endoplasmic reticulum stress (Fig. 8). Normalautophagy function could contribute to balance the pro-survivalphase and pro-apoptosis phase of endoplasmic reticulum stress.However, defective autophagy under physiological/pathologicalsituations might result in unresolved/amplified endoplasmicreticulum stress that leads to apoptosis and the progression ofpodocyte injury eventually.

smic reticulum stress in podocytes. (A) Representative westernhorylation of eIF2α, the expression of cleaved caspase-12 andmmunofluorescence staining shows that inhibition ofodocytes. (H–M) Immunofluorescence staining shows thattranslocation of CHOP/GADD153 in podocytes.

Fig. 6 – Knockdown of CHOP by siRNA alleviate 3-MA induced apoptosis. (A) Representative western blot analysis shows thekinetics of pro-caspase-3 activation. Podocytes were treated with or without 3-MA (2 mmol/L), transfected with control siRNA orwith CHOP siRNA as indicated. (B) Graphic presentation of the relative abundance of cleaved caspase-3 after normalized by full-length caspase-3. The data are presented as mean7S.E. of three experiments. nPo0.05 versus controls; #Po0.05 versus 3-MAtreated group. (C–H): H-33258 and propidium iodide staining demonstrated the podocyte apoptosis morphologically. Podocyteswere incubated with 2 mmol/L 3-MA for 12 h, transfected without or with CHOP siRNA. (I–Q): Representative immunofluorescencestaining indicated the activation and cleavage of caspase-3. Podocytes were incubated with 2 mmol/L 3-MA for 12 h, transfectedwithout or with CHOP siRNA.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 297

Discussion

Podocytes, the visceral glomerular epithelial cells that lines theouter aspect of the glomerular basement membrane, are highlyspecialized and terminally differentiated cells with limited mitoticcapacity. With their interdigitating foot processes, podocytes formthe glomerular filtration barrier together with glomerularendothelial cells and glomerular basement membrane [1,3]. Giventhe importance of podocytes in establishing the selective perme-ability of the glomerular filtration barrier, podocytopathies havenow been proposed as a crucial hallmark of glomerulosclerosisand are considered as a central problem in the progression ofrenal diseases [27]. Thus, over the past decade, podocyte deple-tion which was caused by detachment from glomerular basementmembrane and/or apoptosis has become the focal point ofresearch in deciphering molecular mechanisms of podocytopa-thies. Since podocyte apoptosis was a major factor inducing

podocyte depletion [28], a comprehensive understanding of themolecular mechanisms in modulating podocyte apoptosis couldprovide new therapeutic targets for prevention and remission ofpodocytopathies.Autophagy, a term which was first coined by Christian de Duve

in 1963, is derived from the Greek and meaning “self-eating”. Sinceautophagy is a eukaryotic, evolutionarily conserved catabolicprocess in which unnecessary or dysfunctional cellular componentsare sequestered in double-membrane vesicles and delivered to thelysosomal machinery for degradation, autophagy is essential forsurvival, differentiation, development and homeostasis of cells[5,29]. Despite having an important role in starvation and stressresponses, the role of autophagy in cellular death and survivalremains controversial [30]. Numerous studies have indicated thatautophagy might function as a pro-survival mechanism whereinautophagy might aid the cell in dealing with stress by executingclearance of misfolded proteins, old or damaged cellular

Fig. 7 – Knockdown of CHOP by siRNA restores 3-MA induced down-regulated expression of slit-diaphragm proteins. (A)Representative RT-PCR results show the mRNA levels of Nephrin and WT1 at 6 h after being treated as indicated. (B)Thequantitative data of RT-PCR results. Relative mRNA levels were determined after normalization with GADPH mRNA. The data arepresented as means7S.E. of three experiments. nPo0.05 versus controls; #Po0.05 versus 3-MA treated group. (C) Western blotanalysis shows the protein expression levels of podocin at 12 h after being treated as indicated. Cell lysates were immunoblottedwith Abs against podocin and α-tubulin, respectively. (D) Quantitative determination of podocin protein abundance afternormalization with α-tubulin. Data are presented as mean7SEM of three independent experiments. nPo0.05 versus controls. (E–G) Variable immunofluorescence for WT1 in podocytes which were treated as indicated. E: control group; F: 3-MA treated group; G:CHOP siRNA with 3-MA treated group.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1298

components [29,31–33]. However, at the same time, a large-scaleaccumulation of autophagosomes has also often been observed indying cell, suggesting a possible link between autophagy and celldeath [34–36]. Therefore, the crosstalk between apoptosis andautophagy seems to be complex. Researches investigating thecrosstalk between apoptosis and autophagy might not only enablea complete understanding of the coordinate regulation of autophagyand apoptosis but also present many opportunities for selectivetherapeutic intervention in diseases. However, mechanisms of cross-talk between autophagy and apoptosis in podocytopathies arepoorly understood. In this study, we showed that podocytes bothin vivo and in vitro always exhibited high basal levels of autophagy.What0s more, we provided in vitro and in vivo evidence to supportthat, there was a causal relationship between autophagy inhibitionand podocyte apoptosis. Using the popular inhibitor of the autop-hagic pathway 3-MA, we found that inhibition of autophagy could

directly induce podocyte apoptosis. These data suggested that highlevels of constituent autophagy in non-renewable cells (such aspodocytes) might be important to preserve cell viability andautophagy inhibition was involved in podocyte apoptosis.

Over the last decade, although extensive efforts have beenmade to elucidate the molecular nodes of interface betweenautophagy and apoptosis [12,37], the mechanisms mediating thecomplex counter-regulation of apoptosis and autophagy are notyet fully understood [38]. Multiple direct and indirect mechanisticoverlap and interaction between the apoptosis machinery andautophagy proteins have been described, such as p53 [39], p62[40], Beclin-1 [41], Bcl family proteins [42], FADD [43], ReactiveOxygen Species [44,45] and so on. However, much is stillunknown about the crosstalk between autophagy and apoptosis;therefore, it is likely that future research might be needed toidentify more mechanisms.

Fig. 8 – Autophagy acts to achieve a balance between pro-survival and pro-apoptosis of endoplasmic reticulum stress. The diagramillustrates that normal autophagy function contributes to balance the pro-survival phase and pro-apoptosis phase of endoplasmicreticulum stress. However, autophagy defection might result in unresolved/amplified endoplasmic reticulum stress that leads toapoptosis and the progression of disease eventually.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 299

The endoplasmic reticulum (ER) is the most important organelle forprotein synthesis, folding and maturation in the eukaryotic cell.Intracellular proteins transit through this compartment and onlyproperly folded proteins are allowed to leave the ER while misfoldedproteins are degraded [46]. Under ER stress conditions, too manyunfolded or misfolded proteins are accumulated in the ER lumen thatfinally leads to apoptotic cell death [47]. To maintain fidelity, the ERhas a strict quality-control system for protein proof-reading anddisposal [46,48]. Since growing evidence has suggested that the ERwas the main contributor to the autophagosomes [49–51], autophagywhich was a bulk degradation system that could degrade all forms ofmisfolded proteins has been identified to play an important role inthe ER quality control system. In light of these observations, it isconceivable to speculate that the pro-apoptotic pathway of ER stressmight be involved in autophagy inhibition induced apoptosis sincethe quality-control system was destroyed. Besides, numerous studieshave also proven that ER was involved in calcium homeostasis [52],redox control [44,45,52], the alternative functions of different BCL-2-related proteins [53] and so on. Thus, taken together, ER stress seemsto play a crucial important role in autophagy inhibition inducedapoptosis. Our studies here found that autophagy inhibition couldinduce the activation of ER stress even without any external stimula-tion. To confirm the involvement, we further used the siRNA knock-down technique to suppress the expression of CHOP which has beenconsidered to play a primary role in ER stress-induced apoptosis [26].We found that knockdown of CHOP could effectively improvepodocyte apoptosis and down-regulated expression of slit-diaphragm proteins induced by autophagy inhibition. These resultsin this study suggested that defective autophagy happening in agingand diabetes might directly induce podocyte apoptosis and theprogression of podocytopathies, not just an innocent bystander. SinceER stress has been proven to be involved in autophagy inhibitioninduced apoptosis, it suggested that modulation of ER stress might

present many opportunities for therapeutic interventions in diabeticnephropathy.In summary, we have shown herein that inhibition of autop-

hagy by 3-MA could induce podocyte apoptosis by activating thepro-apoptotic pathway of endoplasmic reticulum stress. Theendoplasmic reticulum stress seems to be an important mechan-ism mediating the cross-regulation of apoptosis and autophagy.Thus, researches elucidating the molecular mechanism of ERstress in the coordinate regulation of autophagy and apoptosiswill not only enable a more complete understanding of thecrosstalk but also unveil novel therapeutic approaches for pre-venting and halting the progression of chronic kidney disease.

Acknowledgments

This work was supported by “973” Science Program of the Ministry ofScience and Technology, China (2011CB504000) to Junwei Yang.The GFP-LC3 plasmid was generously provided by Dr.

Wenxing Ding.

r e f e r e n c e s

[1] H. Pavenstadt, W. Kriz, M. Kretzler, Cell biology of the glomerularpodocyte, Physiol. Rev. 83 (2003) 253–307.

[2] J.W. Fries, D.J. Sandstrom, T.W. Meyer, H.G. Rennke, Glomerularhypertrophy and epithelial cell injury modulate progressiveglomerulosclerosis in the rat, Lab. Invest. 60 (1989) 205–218.

[3] P. Mundel, S.J. Shankland, Podocyte biology and response toinjury, J. Am. Soc. Nephrol. 13 (2002) 3005–3015.

[4] M. Schiffer, M. Bitzer, I.S. Roberts, J.B. Kopp, P. ten Dijke, P.Mundel, E.P. Bottinger, Apoptosis in podocytes induced by TGF-beta and Smad7, J. Clin. Invest. 108 (2001) 807–816.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1300

[5] B. Levine, R. Ranganathan, Autophagy: snapshot of the network,Nature 466 (2010) 38–40.

[6] N. Mizushima, M. Komatsu, Autophagy: renovation of cells andtissues, Cell 147 (2011) 728–741.

[7] K.M. Lucin, T. Wyss-Coray, Targeting autophagy for diseasetherapy, Nat. Biotechnol. 31 (2013) 322–323.

[8] C. He, H. Zhu, H. Li, M.H. Zou, Z. Xie, Dissociation of Bcl-2-Beclin1complex by activated AMPK enhances cardiac autophagy andprotects against cardiomyocyte apoptosis in diabetes, Diabetes62 (2013) 1270–1281.

[9] R.H. Bhogal, C.J. Weston, S.M. Curbishley, D.H. Adams, S.C. Afford,Autophagy: a cyto-protective mechanism which prevents pri-mary human hepatocyte apoptosis during oxidative stress,Autophagy 8 (2012) 545–558.

[10] I.P. Nezis, B.V. Shravage, A.P. Sagona, T. Johansen, E.H. Baehrecke,H. Stenmark, Autophagy as a trigger for cell death: autophagicdegradation of inhibitor of apoptosis dBruce controls DNAfragmentation during late oogenesis in Drosophila, Autophagy 6(2010) 1214–1215.

[11] P. Sansanwal, B. Yen, W.A. Gahl, Y. Ma, L. Ying, L.J. Wong, M.M.Sarwal, Mitochondrial autophagy promotes cellular injury innephropathic cystinosis, J. Am. Soc. Nephrol. 21 (2010) 272–283.

[12] M.C. Maiuri, E. Zalckvar, A. Kimchi, G. Kroemer, Self-eating andself-killing: crosstalk between autophagy and apoptosis, Nat. Rev.Mol. Cell Biol. 8 (2007) 741–752.

[13] L. Fang, Y. Zhou, H. Cao, P. Wen, L. Jiang, W. He, C. Dai, J. Yang,Autophagy attenuates diabetic glomerular damage through pro-tection of hyperglycemia-induced podocyte injury, PLoS One 8(2013) e60546.

[14] T. Yorimitsu, D.J. Klionsky, Eating the endoplasmic reticulum: qualitycontrol by autophagy, Trends. Cell Biol. 17 (2007) 279–285.

[15] K.B. Kruse, J.L. Brodsky, A.A. McCracken, Autophagy: an ERprotein quality control process, Autophagy 2 (2006) 135–137.

[16] P. Mundel, J. Reiser, A. Zuniga Mejia Borja, H. Pavenstadt, G.R.Davidson, W. Kriz, R. Zeller, Rearrangements of the cytoskeletonand cell contacts induce process formation during differentiationof conditionally immortalized mouse podocyte cell lines, Exp.Cell Res. 236 (1997) 248–258.

[17] J. Yang, Y. Liu, Blockage of tubular epithelial to myofibroblasttransition by hepatocyte growth factor prevents renal interstitialfibrosis, J. Am. Soc. Nephrol. 13 (2002) 96–107.

[18] J. Sieber, M.T. Lindenmeyer, K. Kampe, K.N. Campbell, C.D. Cohen,H. Hopfer, P. Mundel, A.W. Jehle, Regulation of podocyte survivaland endoplasmic reticulum stress by fatty acids, Am. J. Physiol.Renal. Physiol. 299 (2010) F821–F829.

[19] C. Dai, J. Yang, Y. Liu, Transforming growth factor-beta1 potentiatesrenal tubular epithelial cell death by a mechanism independent ofSmad signaling, J. Biol. Chem. 278 (2003) 12537–12545.

[20] J. Yang, Y. Liu, Dissection of key events in tubular epithelial tomyofibroblast transition and its implications in renal interstitialfibrosis, Am. J. Pathol. 159 (2001) 1465–1475.

[21] A.G. Porter, R.U. Janicke, Emerging roles of caspase-3 in apopto-sis, Cell Death Differ. 6 (1999) 99–104.

[22] B. Loos, A.M. Engelbrecht, R.A. Lockshin, D.J. Klionsky, Z. Zakeri,The variability of autophagy and cell death susceptibility:unanswered questions, Autophagy 9 (9) (2013) 1270–1285.

[23] W.X. Ding, H.M. Ni, W. Gao, T. Yoshimori, D.B. Stolz, D. Ron, X.M.Yin, Linking of autophagy to ubiquitin-proteasome system isimportant for the regulation of endoplasmic reticulum stress andcell viability, Am. J. Pathol. 171 (2007) 513–524.

[24] E. Fujita, Y. Kouroku, A. Isoai, H. Kumagai, A. Misutani, C.Matsuda, Y.K. Hayashi, T. Momoi, Two endoplasmic reticulum-associated degradation (ERAD) systems for the novel variant ofthe mutant dysferlin: ubiquitin/proteasome ERAD(I) and autop-hagy/lysosome ERAD(II), Hum. Mol. Genet. 16 (2007) 618–629.

[25] R. Sheng, X.Q. Liu, L.S. Zhang, B. Gao, R. Han, Y.Q. Wu, X.Y. Zhang,Z.H. Qin, Autophagy regulates endoplasmic reticulum stress inischemic preconditioning, Autophagy 8 (2012) 310–325.

[26] K. Zhang, R.J. Kaufman, Identification and characterization ofendoplasmic reticulum stress-induced apoptosis in vivo, Meth-ods Enzymol. 442 (2008) 395–419.

[27] G. Wolf, S. Chen, F.N. Ziyadeh, From the periphery of theglomerular capillary wall toward the center of disease: podocyteinjury comes of age in diabetic nephropathy, Diabetes 54 (2005)1626–1634.

[28] K. Susztak, A.C. Raff, M. Schiffer, E.P. Bottinger, Glucose-inducedreactive oxygen species cause apoptosis of podocytes andpodocyte depletion at the onset of diabetic nephropathy, Dia-betes 55 (2006) 225–233.

[29] N. Mizushima, B. Levine, A.M. Cuervo, D.J. Klionsky, Autophagyfights disease through cellular self-digestion, Nature 451 (2008)1069–1075.

[30] G. Kroemer, B. Levine, Autophagic cell death: the story of amisnomer, Nat. Rev. Mol. Cell Biol. 9 (2008) 1004–1010.

[31] R. Sheng, L.S. Zhang, R. Han, X.Q. Liu, B. Gao, Z.H. Qin, Autophagyactivation is associated with neuroprotection in a rat model offocal cerebral ischemic preconditioning, Autophagy 6 (2010)482–494.

[32] H.D. Xu, D. Wu, J.H. Gu, J.B. Ge, J.C. Wu, R. Han, Z.Q. Liang, Z.H.Qin, The pro-survival role of autophagy depends on Bcl-2 undernutrition stress conditions, PLoS One 8 (2013) e63232.

[33] Y.T. Wu, H.L. Tan, Q. Huang, C.N. Ong, H.M. Shen, Activationof the PI3K-Akt-mTOR signaling pathway promotes necroticcell death via suppression of autophagy, Autophagy 5 (2009)824–834.

[34] Y. Tsujimoto, S. Shimizu, Another way to die: autophagicprogrammed cell death, Cell Death Differ. 12 (Suppl. 2) (2005)S1528–S1534.

[35] S. Pattingre, A. Tassa, X. Qu, R. Garuti, X.H. Liang, N. Mizushima,M. Packer, M.D. Schneider, B. Levine, Bcl-2 antiapoptoticproteins inhibit Beclin 1-dependent autophagy, Cell 122 (2005)927–939.

[36] R.C. Scott, G. Juhasz, T.P. Neufeld, Direct induction of autophagyby Atg1 inhibits cell growth and induces apoptotic cell death,Curr. Biol. 17 (2007) 1–11.

[37] A.D. Rubinstein, A. Kimchi, Life in the balance – a mechanisticview of the crosstalk between autophagy and apoptosis, J. CellSci. 125 (2012) 5259–5268.

[38] C. Gordy, Y.W. He, The crosstalk between autophagy and apop-tosis: where does this lead? Protein Cell 3 (2012) 17–27.

[39] D. Crighton, S. Wilkinson, J. O0Prey, N. Syed, P. Smith, P.R.Harrison, M. Gasco, O. Garrone, T. Crook, K.M. Ryan, DRAM, ap53-induced modulator of autophagy, is critical for apoptosis,Cell 126 (2006) 121–134.

[40] J. Moscat, M.T. Diaz-Meco, p62 At the crossroads of autophagy,apoptosis, and cancer, Cell 137 (2009) 1001–1004.

[41] J. Wang, Beclin 1 bridges autophagy, apoptosis and differentia-tion, Autophagy 4 (2008) 947–948.

[42] B. Levine, S. Sinha, G. Kroemer, Bcl-2 family members: dualregulators of apoptosis and autophagy, Autophagy 4 (2008) 600–606.

[43] J. Thorburn, F. Moore, A. Rao, W.W. Barclay, L.R. Thomas, K.W.Grant, S.D. Cramer, A. Thorburn, Selective inactivation of a Fas-associated death domain protein (FADD)-dependent apoptosisand autophagy pathway in immortal epithelial cells, Mol. Biol.Cell 16 (2005) 1189–1199.

[44] T. Ma, J. Zhu, X. Chen, D. Zha, P.C. Singhal, G. Ding, High glucoseinduces autophagy in podocytes, Exp. Cell Res. 319 (2013) 779–789.

[45] A. Yadav, S. Vallabu, S. Arora, P. Tandon, D. Slahan, S. Teichberg, P.C. Singhal, ANG II promotes autophagy in podocytes, Am. J.Physiol. Cell Physiol. 299 (2010) C488–C496.

[46] A. Buchberger, B. Bukau, T. Sommer, Protein quality control in thecytosol and the endoplasmic reticulum: brothers in arms, Mol.Cell 40 (2010) 238–252.

[47] M. Kitamura, Endoplasmic reticulum stress and unfolded proteinresponse in renal pathophysiology: Janus faces, Am. J. Physiol.Renal Physiol. 295 (2008) F323–F334.

E X P E R I M E N T A L C E L L R E S E A R C H 3 2 2 ( 2 0 1 4 ) 2 9 0 – 3 0 1 301

[48] L. Ellgaard, A. Helenius, Quality control in the endoplasmicreticulum, Nat. Rev. Mol. Cell Biol. 4 (2003) 181–191.

[49] M. Hamasaki, N. Furuta, A. Matsuda, A. Nezu, A. Yamamoto, N.Fujita, H. Oomori, T. Noda, T. Haraguchi, Y. Hiraoka, A. Amano, T.Yoshimori, Autophagosomes form at ER-mitochondria contactsites, Nature 495 (2013) 389–393.

[50] M. Hayashi-Nishino, N. Fujita, T. Noda, A. Yamaguchi, T. Yoshi-mori, A. Yamamoto, A subdomain of the endoplasmic reticulumforms a cradle for autophagosome formation, Nat. Cell Biol. 11(2009) 1433–1437.

[51] P. Yla-Anttila, H. Vihinen, E. Jokitalo, E.L. Eskelinen, 3D tomo-graphy reveals connections between the phagophore and endo-plasmic reticulum, Autophagy 5 (2009) 1180–1185.

[52] A. Gorlach, P. Klappa, T. Kietzmann, The endoplasmic reticulum:folding, calcium homeostasis, signaling, and redox control,Antioxid. Redox Signal. 8 (2006) 1391–1418.

[53] D. Rojas-Rivera, B. Caballero, S. Zamorano, F. Lisbona, C. Hetz,Alternative functions of the BCL-2 protein family at theendoplasmic reticulum, Adv. Exp. Med. Biol. 687 (2010)33–47.