Protein Kinase Cθ Is Required for Autophagy in Response to Stress in the Endoplasmic Reticulum^,

Abstract

Autophagy is an evolutionally conserved process for the bulk degradation of cytoplasmic proteins and organelles. Recent observations indicate that autophagy is induced in response to cellular insults that result in the accumulation of misfolded proteins in the lumen of the endoplasmic reticulum (ER). However, the signaling mechanisms that activate autophagy under these conditions are not understood. Here, we report that ER stress-induced autophagy requires the activation of protein kinase Cθ (PKCθ), a member of the noveltype PKC family. Induction of ER stress by treatment with either thapsigargin or tunicamycin activated autophagy in immortalized hepatocytes as monitored by the conversion LC3-I to LC3-II, clustering of LC3 into dot-like cytoplasmic structures, and electron microscopic detection of autophagosomes. Pharmacological inhibition of PKCθ or small interfering RNA-mediated knockdown of PKCθ prevented the autophagic response to ER stress. Treatment with ER stressors induced PKCθ phosphorylation within the activation loop and localization of phospho-PKCθ to LC3-containing dot structures in the cytoplasm. However, signaling through the known unfolded protein response sensors was not required for PKCθ activation. PKCθ activation and stress-induced autophagy were blocked by chelation of intracellular Ca²⁺ with BAPTA-AM. PKCθ was not activated or required for autophagy in response to amino acid starvation. These observations indicate that Ca²⁺-dependent PKCθ activation is specifically required for autophagy in response to ER stress but not in response to amino acid starvation.

Autophagy is a process by which intracellular material is recycled to supply the cell with nutrients and energy for survival under conditions of severe stress, such as amino acid starvation. Upon nutrient limitation, autophagy is induced through the proliferation of cell membranes that form the autophagosome for the engulfment of large intracellular protein complexes and subcellular organelles. The autophagosome docks and fuses with the lysosome to form the autolysosome that degrades its lumenal contents to recycle amino acids and fatty acids. This “self-eating” process is typically induced under conditions of nutrient starvation. However, autophagy also provides an essential role for the clearance of dead-ended protein aggregates that cannot be degraded by the concerted action of molecular chaperones and the proteasome, such as those that occur in Huntington disease and Parkinson disease, (1-3). Because there is a growing understanding of the significance of autophagy in fundamental pathological conditions ranging from cancer to neurodegeneration, there is enthusiasm for the development of novel therapeutic interventions for these conditions (4). However, although the molecular mechanisms that signal autophagy have been extensively characterized in budding yeast, the process in higher eukaryotes is largely not understood.

Autophagy is induced by a group of evolutionarily conserved autophagy gene-related proteins (ATG proteins) (5). The process is initiated when a class III phosphatidylinositol 3-kinase complex and ATG proteins form an isolation membrane. A ubiquitin-like protein conjugation pathway expands the isolation membrane. LC3 (microtubule-binding protein light chain 3; mammalian homolog of Atg8) is synthesized as an inactive soluble form (LC3-I) that is converted into an active membranous form (LC3-II) by modification with phosphatidylethanolamine. This modification is mediated by the ubiquitin-like conjugation system comprised of the E1-like ATG7 protein and the E2-like ATG3 protein. Phosphatidylethanolamine-modified LC3-II binds to target membranes and, in collaboration with ATG5-ATG12-ATG16 complexes, induces membrane alterations required for autophagosome formation.

The molecular pathways that regulate autophagy are most well understood in the context of nutrient limitation. Autophagy is inhibited by depletion of cellular energy and amino acid levels through the target of rapamycin protein kinase TOR (target of rapamycin) (6-8). TOR activity is inhibited by the AMP-activated protein kinase via a pathway involving the GTPase-activating tuberous sclerosis complex 1/2 and its substrate RHEB (Ras homolog enriched in brain), which is a member of the RAS family of GTP-binding proteins (9). Upon energy depletion, the LKB1 tumor suppressor kinase phosphorylates and activates AMP-activated protein kinase. AMP-activated protein kinase is also activated through the Ca²⁺/calmodulin-dependent protein kinase kinase-β (CaMKK-β)2 in response to Ca²⁺ release from the ER (10-14). Intriguingly, ER-localized BCL2 inhibits autophagy, possibly through reducing Ca²⁺ mobilization from the ER lumen (15, 16).

Recently, it was demonstrated that pharmacological perturbation of ER function induces autophagy (14, 17-20). Conditions that disturb homeostasis in the ER, such as Ca²⁺ depletion from the ER or inhibition of asparagine (N)-linked glycosylation, cause the accumulation of unfolded protein in the ER lumen and activate an adaptive signaling pathway termed the unfolded protein response (UPR). In higher eukaryotes, the ER-resident transmembrane proteins IRE1α, PERK, and ATF6 act as proximal sensors to signal the UPR (21, 22). Upon accumulation of unfolded protein in the ER, the protein kinase PERK phosphorylates the α subunit of eukaryotic translation initiation factor 2 (eIF2α) at Ser⁵¹ to attenuate mRNA translation and thereby reduce the amount of the client proteins translocated into the ER lumen. The PERK-eIF2α pathway also up-regulates amino acid biosynthesis and anti-oxidative stress response genes through promoting preferential translation of ATF4 mRNA. In parallel, cleavage of ATF6α and IRE1α-mediated splicing of X-box-binding protein 1 (Xbp1) mRNA generate two transcription factors that induce expression of genes encoding ER molecular chaperones and ER-associated protein degradation machinery. In this way, the UPR couples the ER protein folding capacity with the protein folding demand. However, chronic unresolved accumulation of unfolded protein in the ER elicits an apoptotic program through ATF4- and ATF6α-mediated transcriptional activation of the CCAAT/enhancer-binding protein homologous protein transcription factor CHOP/GADD153 (23-26). In addition, activated IRE1α leads to phosphorylation c-Jun N-terminal kinase to also contribute to the apoptotic program (27-29).

Presently, there are conflicting reports regarding the requirement for UPR signaling in the autophagic response to ER stress. Ogata et al. (18) reported that ER stress-induced autophagy requires IRE1α-mediated activation of AMP-activated protein kinase N-terminal kinase. In contrast, Kouroku et al. (20) demonstrated that transcriptional up-regulation of ATG12 requires signaling through PERK-eIF2α. Because ER stress also results in Ca²⁺ leak from the ER (30), ER stress-induced autophagy may also be mediated through CaMKK-β-mediated activation of AMP-activated protein kinase to inhibit mTOR kinase (14, 31, 32). Consequently, further studies are required to elucidate the molecular signaling processes by which ER stress induces autophagy.

The ER is the major intracellular Ca²⁺ storage organelle of the cell. Ca²⁺ released from the ER activates members of the protein kinase C (PKC) family (33, 34). PKC isoforms function in multiple cellular processes including cell growth and differentiation, cell cycle control, ion flux, protein secretion, tumorigenesis, and apoptosis (33, 35, 36). Recently, PKCs were suggested to function in ER stress signaling (37-41), although to date there are no studies that demonstrate a requirement for PKC signaling in the autophagic response. In mammals, the PKC family comprises 11 isoforms that are classified into three subgroups based on their domain structures. The classical PKCs (α, βI, βII, and γ) possess both C1 and C2 domains that bind lipid cofactors (diacylglycerol) and Ca²⁺, respectively. In contrast, the novel PKCs (nPKC: δ, ε, η, μ, and θ) lack a C2 domain, and the atypical PKCs (ζ, λ, and ι) lack both C1 and C2 domains (42). In addition to their diverse structures, the different PKC isoforms display different mechanisms of activation and tissue-specific expression (42). Herein, we report that PKCθ, a member of the nPKC family, is a novel factor that mediates Ca²⁺-dependent induction of autophagy in response to ER stress but not in response to amino acid starvation.

EXPERIMENTAL PROCEDURES

Chemical Reagents and Antibodies—Polyclonal antibody against ATG8/LC3 was kindly provided by Dr. Ron R. Kopito (Stanford University). Polyclonal antibodies against PKCθ, phospho-PKCθ (Thr⁵³⁸), PKCδ, phospho-PKCδ (Thr⁵⁰⁵), PKCα, phospho-PKCα/βI (Ser^644/647), and ERK1/2 were purchased from Cell Signaling Inc. (Danvers, MA). Polyclonal antibody against Calnexin was purchased from Stressgen Biotechnologies Inc. (San Diego, CA). Polyclonal antibody against CHOP was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Monoclonal antibody against α-actin was purchased from BD Biosciences (San Jose, CA). Thapsigargin (TG), tunicamycin (TM), bafilomycin A1, and 3-methyladenine (3MA) were purchased from Sigma-Aldrich. Rottlerin and Go6976 were purchased from Calbiochem (La Jolla, CA).

Cell Culture and Stress Treatments—Immortalized hepatocytes were maintained in medium 199 (Invitrogen) supplemented with 15% certified heat-inactivated fetal bovine serum (Invitrogen). The cells were treated with TG or TM to induce ER stress. For transient treatment with TG, the cells were cultured in M199 medium containing 1 μm TG for 3 or 10 min, and then the cells were rinsed three times with PBS and cultured for 6 h in fresh M199 medium in the absence of TG. Where indicated, Go6976, rottlerin, BAPTA-AM, or 3MA was added to cell cultures for 20 min prior to the addition of TG or TM, and these agents were present during the entire time course of the experiment. For amino acid deprivation, the cells were rinsed three times with PBS and cultured in Earl's balanced salt solution in the absence or presence of amino acids (Invitrogen) supplemented with 15% dialyzed fetal bovine serum and 10 nm bafilomycin A1.

Primary Hepatocyte Isolation and Immortalization—Primary murine hepatocytes were isolated and immortalized as previously described (43, 44). Briefly, livers from embryonic day 19 embryos were harvested and mechanically dissociated with a scalpel in 0.25% trypsin (Mg²⁺- and Ca²⁺-free). Hepatocytes were further dissociated from the liver pieces by gentle shaking at 37 °C for 10 min. The liver mass was permitted to settle to replace the trypsin solution with 0.05 mg/ml Liberase Blendzyme 4 solution (Roche Applied Science). After shaking at 37 °C for 20 min, the cells were dispersed by gentle pipetting and centrifuged at 50 × g for 5 min. Hepatocytes were plated onto collagen-coated plates (BD Bioscience) in M199 medium supplemented with 10% heat-inactivated fetal bovine serum and containing penicillin and streptomycin. After overnight incubation, the cultures were rinsed extensively to remove hematopoietic cells and other nonadherent cells.

The puromycin resistance retroviral vector pBabe encoding SV40 large T antigen (a gift from P. Jat, Ludwig Institute for Cancer Research, London, UK) was transfected into viral PLAT-E packaging cells (kindly provided by D. Fang, University of Missouri, Columbia, MO) at 70% confluency using FuGENE 6. Medium containing virus was harvested and stored at 0 °C. Hepatocytes at 40% confluency were then infected by incubation with polybrene (8 μg/ml)-supplemented virus at 32 °C for 8 h and then transferred to 37 °C for 72 h prior to selection with puromycin (1 μg/ml) for 1 week.

Generation of siRNA Knockdown Stable Cell Lines—siRNA-mediated knockdown was performed with pRNAT6.1/Neo (GeneScript, Inc., Piscataway, NJ). siRNA expression plasmids were constructed according to the manufacturer's instructions. Briefly, 70 bases of siRNA primers were designed by siRNA Target Finder and siRNA Construct Builder (GeneScript, Inc.). The nucleotide sequences of the inserted siRNA were: PKCθ si1 (5′-GATCCCGTTTATCAATGCACTTCTTGTGTTGATATCCGCACAAGAAGTGCATTGATAAATTTTTTCCAAA), PKCθ si2 (5′-GATCCCGTTTATACCACAAAGATTGGCATTGATATCCGTGCCAATCTTTGTGGTATAAATTTTTTCCAAA), and GFP (5′-GATCCCGTAGTTGCCGTCGTCCTTGAAGTTGATATCCGCTTCAAGGACGACGGCAACTATTTTTTCCAAA). A pair of primers was annealed in 1× SSC solution (150 mm sodium chloride, 15 mm sodium citrate) and ligated into the BamHI and HindIII sites of pRNAT-6.1/Neo using the Rapid DNA ligation kit (Roche Applied Science). Purified plasmid DNA was confirmed by DNA sequencing and transfected into the immortalized hepatocytes using FuGENE 6.0 (Roche Applied Science). Immortalized clones were selected in medium containing 400 μg/ml Geneticin (Invitrogen) maintained in medium containing 200 μg/ml Geneticin and analyzed by immunoblotting.

Cell Survival Assay—3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assays were performed using the Cell Titer 96 AQueous One solution reagent (Promega Inc., Madison, WI) as described by the supplier. Briefly, the cells were plated onto 96-well tissue culture plates (100 μl of medium/well) and propagated to ∼50% confluency. The cells were then treated with TG or TM with or without 10 mm 3MA, as described in the figure legends. Cell Titer reagent (10 μl) was added into each well and incubated for 2 h. A_{595 nm} was measured using a VERSA-max microplate reader (Molecular Device, Sunnyvale, CA).

Immunoblotting and Immunofluorescence Analysis—For immunoblotting analysis, the cells were harvested in radioimmune precipitation assay buffer (150 mm NaCl, 1.0% Nonidet P-40, 0.5%DOC, 0.1% SDS, 50 mm Tris-HCl, pH 8.0) containing Complete Mini EDTA-free protease inhibitor mixture (Roche Applied Science) and 50 mm NaF. The protein concentrations were measured using the Dc protein assay system (Bio-Rad). Equal amounts of protein were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes (Bio-Rad). The transferred membranes were incubated with blocking solution (5% bovine serum albumin, 150 mm NaCl, 20 mm Tris-HCl, pH 7.5, 0.1% Tween 20), treated with primary antibodies (diluted with blocking buffer, 1:250 for α-CHOP antibody, 1:5000 for α-actin and α-LC3 antibody, 1:1000 for all other antibodies), and treated with horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG (Promega) as secondary antibodies for detection. Gel images were developed using BioMax MR film (Kodak, Inc., Tokyo, Japan) and an M35A X-Omat processor (Kodak).

For immunofluorescence analysis, the cells were cultured to ∼80% confluency using the LabTek-II Chamber Slide System (Nalge Nunc International Inc., Rochester, NY). After stress treatment, the cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 in PBS. Chamber slides were incubated with blocking solution (2% bovine serum albumin in PBS), treated with primary antibody (1:1000 dilution for all of antibodies except for 1:5000 dilution for α-LC3 antibody), and treated with Alexa Fluor (488 or 594) anti-rabbit IgG (Invitrogen) as the secondary antibody. Microscopic observation was performed using an Olympus BX-51 research microscope (Olympus Imaging America, Center Valley, PA).

Electron Microscopy—For transmission electron microscopy, cell monolayers were rinsed in serum-free medium and then fixed for 1 h at room temperature in 2.5% glutaraldehyde in 0.1 m Sorensen's buffer, pH 7.4. Following a buffer rinse, the cells were post-fixed for 15 min in 1% osmium tetroxide in the same buffer. The cells were then rinsed in double-distilled water, en bloc stained for 1 h with a saturated, aqueous solution of uranyl acetate, scraped from the culture dishes, and collected by centrifugation in Eppendorf tubes. For each subsequent step, the cells were resuspended in the next reagent and then centrifuged. The cells were dehydrated rapidly in a graded series of ethanol, infiltrated and embedded in Epon, and polymerized. Ultra-thin sections were collected onto copper grids and poststained with uranyl acetate and lead citrate. The sections were viewed on a Philips CM100 at 60 kV. The images were recorded digitally using a Hamamatsu ORCA-HR digital camera system with AMT software (Advanced Microscopy Techniques Corp., Danvers, MA).

Subcellular Fractionation—TG-treated immortalized hepatocytes were harvested by scraping in PBS and collected by centrifugation at 600 × g for 5 min. After the supernatant was removed, the cells were resuspended in three volumes of 1× hypotonic buffer (10 mm HEPES, pH 7.8, 1 mm EGTA, and 25 mm potassium chloride) and then incubated on ice for 20 min. The cell suspensions were centrifuged at 600 × g for 5 min, and the pellets were resuspended in two volumes of 1× isotonic buffer (1× hypotonic buffer with 250 mm sucrose) and subjected to Dounce homogenization. The homogenates were centrifuged at 1000 × g for 10 min to remove the cell debris, and the supernatants were centrifuged at 12,000 × g for 15 min to obtain the post-mitochondrial supernatant fractions. The supernatants were centrifuged at 100,000 × g for 60 min to isolate the supernatant cytosolic fractions. Then pellets were resuspended in isotonic buffer and centrifuged at 100,000 × g for 60 min to obtain the microsomal pellet fractions. Equal percentages of the cytosolic fractions and microsomal fractions were analyzed by Western blotting. All of the procedures for the cells fractionation were performed on ice with Complete-mini EDTA-free protease inhibitor mixture (Roche Applied Science) and 50 mm sodium fluoride.

RESULTS

The Autophagic Response Is Induced by ER Stress—Recently, it was demonstrated that ER stress induces autophagy in Saccharomyces cerevisiae and mammalian cells (17-20). In mammalian cells, autophagy is detected by phosphatidylethanolamine addition to LC3 and its relocalization to membranes (45). LC3-I and LC3-II are distinguishable by SDS-PAGE because of increased mobility of LC3-II that occurs as a consequence of increased hydrophobicity caused by lipidation. In addition, the distribution of LC3 changes from a diffuse cytoplasmic localization to dot-like structures that represent the autophagosome. We measured LC3 modification and localization in immortalized hepatocytes in response to the sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase inhibitor TG. Treatment with 1 μm TG for 8 h caused conversion of LC3-I to LC3-II and LC3 localization to dot structures in the cytoplasm (Fig. 1, A-C). Treatment with 5 mm 3MA, an inhibitor of class III phosphatidylinositol 3-kinase (encoded by a single gene, the mammalian homolog of VPS34) that prevents autophagosome formation, partially inhibited the TG-dependent production of LC3-II and LC3 localization to dot structures (Fig. 1, A-C). Ultrastructural analysis by transmission electron microscopy detected the presence of numerous double membrane-enclosed vesicles that contained membranous and cytoplasmic material in TG-treated cells (Fig. 1D, panels b-d). These structures are characteristic of autophagosomes. In addition, TG treatment altered the structure of the ER where it became distended with a greater amount of electron dense material (Fig. 1D, panels a and b). This morphology is typical of cells in which unfolded proteins accumulate in the ER lumen (46). Finally, vacuole-like structures that apparently contained fragments of ER were observed in TG-treated cells (Fig. 1D, panels e and f). These vacuole-like structures suggested that portions of the ER were engulfed and degraded by the autolysosome. These structures are reminiscent of structures recently described in S. cerevisiae that occur in response to ER stress (17, 47).

FIGURE 1.

ER stress induces autophagy in immortalized hepatocytes. Immortalized hepatocytes were treated with TG (1 μm) in the presence or absence of 3MA (5 mm) for 8 h and analyzed for LC3-II production (A) and LC3-containing dot structures (B). C, the results from B were quantified by counting the number of cells with LC3 dots relative to the total number of observed cells. The number of LC3-containing dot structures observed in response to amino acid (AA) starvation for 90 min in the presence or absence of 3MA (5 mm) was quantified as described under “Experimental Procedures.” D, transmission electron microscopy micrographs of control and TG-treated immortalized hepatocytes are shown. Compared with the ER observed in untreated cells (panel a, arrows), the ER in cells treated with TG (1 μm for 8 h) had greater electron density (panel b, arrows). Autophagosomes were frequently observed in the TG-treated cells (panels b-d, arrowheads). Vacuoles with fragments of ER were also observed only in the TG-treated cells (panels e and f). NT, no treatment.

Autophagy Is Required for Survival in Response to ER Stress—We next determined whether cells require autophagy to survive ER stress. MTT reduction analysis revealed that treatment with either TG or TM, an inhibitor of N-linked glycosylation, reduced viability to ∼65% compared with untreated cells (Fig. 2A). Treatment with 3MA to inhibit autophagy further reduced viability to ∼40%, suggesting that autophagy does increase viability of ER-stressed cells. To determine whether ER stress-induced autophagy may be a consequence of ER stress-induced apoptosis, the cells were transiently treated with TG for 10 min and then cultured in the absence of TG for 6 h. Cell viability was significantly improved in the cells transiently treated with TG for 10 min compared with cells continuously treated with TG for 6 h (Fig. 2B). However, both conditions significantly induced the UPR, monitored by either Xbp1 mRNA splicing or CHOP expression (Fig. 2C). Both the 10 min transient TG treatment, as well as the 6-h continuous TG treatment, produced significant numbers of cells with LC3-containing dot structures (Fig. 2, D and E). In contrast, significantly fewer dot structures were observed in cells that were transiently treated with TG for only 3 min (Fig. 2E). Because the 10-min transient TG treatment induced autophagy without significant cell death, the majority of our experiments were performed using cells treated transiently with TG for 10 min.

FIGURE 2.

ER stress-induced autophagy is required for survival. A, immortalized hepatocytes were treated with TG (12.5 nm) or TM (250 ng/ml) for 24 or 43 h, respectively, in the presence or absence of 10 mm 3MA, and survival was measured by MTT assay. The results are normalized to untreated cells. B-E, cells were treated with TG (1 μm) either transiently for 10 min with a 3-h (B) or 6-h recovery time in media lacking TG or continuously in media containing TG for 3 h (B) or 6 h. B, cell viability was measured by MTT assay. C, total RNA was prepared for analysis by reverse transcription-PCR for Xbp1 mRNA splicing and protein extracts were analyzed by Western immunoblot for CHOP and α-actin as a loading control. D and E, cells were analyzed by immunofluorescence using anti-LC3 antibody and dot structures were quantified. For E, the cells were also treated transiently with TG for 3 min and allowed to recover in media lacking TG for 6 h. NT, no treatment.

Autophagy Induced by ER Stress Requires PKCθ—The requirement for PKC activity in ER stress-induced autophagy was studied using specific PKC inhibitors. Go6976 inhibits classical PKCs (IC₅₀ = 2.3 and 6.2 nm for PKCα and PKCβI, respectively), and rottlerin inhibits nPKCs (IC₅₀ = 3-6 μm for PKCδ and PKCθ, 60-80 μm for PKCε). We found that 20 μm rottlerin significantly blocked formation of LC3-containing dot structures in response to TG treatment, whereas 1 μm Go6976 had no effect (Fig. 3, A and B). This result suggested that PKCδ and/or PKCθ might be required for ER stress-induced autophagy.

FIGURE 3.

nPKC is required for autophagy in response to ER stress. Immortalized hepatocytes were treated transiently with TG (1 μm for 10 min) and then incubated in complete M199 medium lacking TG for 6 h. Where indicated, either Go6976 (1 μm) or rottlerin (20 μm) was present during the time course of the experiment as described under “Experimental Procedures.” A and B, after 6 h, the cells were stained with anti-LC3 antibody for analysis by immunofluorescence microscopy and dot structures were quantified. C, immortalized hepatocytes that express the indicated siRNAs were treated transiently with TG for 10 min and propagated 6 h in complete medium lacking TG for Western blot analysis to monitor LC3 conversion, total PKCθ, and phosphorylated (Thr⁵³⁸) PKCθ, PKCδ and CHOP as described under “Experimental Procedures.” D and E, LC3 immunofluorescence was performed to analyze and quantify LC3-containing dots. NT, no treatment.

To elucidate whether ER stress-induced autophagy requires PKCθ, we established stable clones of hepatocytes that express siRNAs to knock down expression of PKCθ. To minimize potential off target effects of the siRNA, we targeted two different regions within PKCθ mRNA. Control cells were derived that express siRNA specific to GFP. The expression of PKCθ mRNA in cells transfected with PKCθ siRNA was reduced to ∼10% of that in cells with the control GFP siRNA, whereas the expression of PKCδ mRNA was not changed (data not shown). Expression of PCKθ protein was reduced ∼2-3-fold in the PCKθ knockdown cells (Fig. 3C). Where TG treatment induced the conversion of LC3-I to LC3-II and formation of LC3-containing dot structures in cells that express the control GFP siRNA, these processes were significantly reduced in cells expressing either of the PKCθ siRNAs (Fig. 3, C-E, and data not shown). In contrast, knockdown of PKCθ did not significantly affect UPR activation monitored by CHOP expression (Fig. 3C). These observations indicate that PKCθ is required for ER stress-induced autophagy.

ER Stress Induces Ca²⁺-dependent Phosphorylation and Localization of PKCθ to LC3-containing Dot Structures—PKCθ requires phosphorylation within the activation loop to elicit protein kinase activity and for proper intracellular localization (48-50). Immunoblot analysis demonstrated that TG treatment specifically increased PKCθ phosphorylation at Thr⁵³⁸ within the activation loop, whereas phosphorylation observed at known activating sites in PKCα/βII (Thr^638/641), PKCδ (Thr⁵⁰⁵), or PKCδ (Ser⁶⁴³) (48, 49, 51-53) was not significantly altered (Fig. 4A). Significantly less PKCθ Thr⁵³⁸ phosphorylation was observed upon TG treatment of the PKCθ knockdown cells (Fig. 3C). In addition, rottlerin treatment significantly reduced TG-induced formation of LC3-containing dots structures and also PKCθ Thr⁵³⁸ phosphorylation in immortalized hepatocytes, as well as in murine embryonic fibroblasts (supplemental Fig. S1). Therefore, the requirement for PKCθ for TG-induced autophagy does not appear to be restricted to immortalized hepatocytes. Finally, transient overexpression of T538A mutant PKCθ partially prevented ER stress-induced formation of LC3-containing dots in immortalized hepatocytes (supplemental Fig. S2). Taken together, these results support the hypothesis that phosphorylation at Thr⁵³⁸ in PKCθ contributes to ER stress-induced autophagy.

FIGURE 4.

PKCθ is required for autophagy induction in response to ER stress. Immortalized hepatocytes were subjected to 10 min of transient TG (1 μm) treatment and then incubated for 2 or 6 h in complete medium. Immunoblotting was performed to detect phosphorylated PKCθ, PKCδ and PKCα/βI (A), and immunofluorescence was performed to localize phosphorylated PKCθ (B). C, immortalized hepatocytes were transiently transfected with an EGFP-LC3 expression vector and subsequently treated with TG for 8 h. The cells were stained with primary antibody against LC3 or phospho-PKCθ (Thr⁵³⁹), and then anti-rabbit IgG was conjugated with rhodamine as a secondary antibody. D, immortalized hepatocytes were treated with TG for 6 h and then harvested for subcellular fractionation. Membrane and cytosolic fractions were subjected to Western blot analysis to detect LC3-II, phospho-PKCθ (Thr⁵³⁸), calnexin (membrane marker), and ERK1/2 kinase (cytosolic marker). NT, no treatment.

The intracellular localization of phosphorylated PKCθ in response to ER stress was determined by confocal immunofluorescence microscopy and subcellular fractionation. Upon TG treatment, phosphorylated PKCθ (Thr⁵³⁸) was localized to dot structures in the cytoplasm (Fig. 4B). To analyze whether these dot structures may represent autophagosomes, the cells were transiently transfected with an EGFP-LC3 expression vector and subsequently treated with TG. The cells were then fixed analyzed by immunofluorescence for detection of LC3 and phospho-PKCθ (Thr⁵³⁸). The results demonstrated that both phospho-PKCθ and LC3 colocalized with the EGFP-LC3-containing dot structures in the cytoplasm (Fig. 4C).

The localization of phosphorylated PKCθ was also studied by cellular subfractionation. The majority of phospho-PKCθ (Thr⁵³⁸) was detected in the membrane fraction with the ER membrane marker calnexin (Fig. 4D). Therefore, both the immunofluorescence data and cellular subfractionation experiments indicate that ER stress induces localization of phospho-PKCθ with LC3-II in dot structures.

Upon treatment with either TG or TM, the kinetics of PKCθ phosphorylation correlated with the processing of LC3-I to LC3-II and the formation of LC3-containing dots (Fig. 5, A and B). However, the conversion of LC3-I to LC3-II, formation of LC3-containing dots, and phosphorylation of PKCθ occurred within 1-2 h after TG treatment and at 8-12 h after TM treatment. In contrast, the induction of the UPR, monitored by Xbp1 mRNA splicing, occurred within 30 min after treatment with either TG or TM (Fig. 5C). TG induces ER stress through inhibition of the sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase and depletion of ER lumenal Ca²⁺ that is required for ER protein folding and chaperone functions. In contrast, TM inhibits N-linked glycosylation that interferes with protein folding and, after prolonged treatment, induces Ca²⁺ leak from the ER (30). Therefore, it is possible that the different kinetics observed for TG-induced and TM-induced PKCθ phosphorylation, LC3-II conversion, and LC3 dot structure formation may be a consequence of different rates of Ca²⁺ release from the ER.

FIGURE 5.

PKCθ activation coincides with ER stress-induced autophagy. The kinetics of PKCθ phosphorylation and conversion of LC3-I to LC3-II were monitored in immortalized hepatocytes treated continuously for increasing periods of time with TG (1 μm) or TM (10 μg/ml). A, cell extracts were prepared for Western blot analysis for LC3-I and LC3-II, and total versus phosphorylated PKCθ. B, cells were prepared for immunofluorescence analysis for LC3, and dot structures were quantified. C, in parallel, UPR induction was monitored by analysis of Xbp1 mRNA splicing as described under “Experimental Procedures.” NT, no treatment.

To test the requirement for an increase in cytosolic Ca²⁺, we studied the effect of the membrane-permeable intracellular Ca²⁺ chelator BAPTA-AM. BAPTA-AM treatment prevented the formation of LC3-containing dot structures (Fig. 6A), partially inhibited the conversion of LC3-I to LC3-II (Fig. 6B), and also significantly reduced PKCθ phosphorylation and localization to dot structures (Fig. 6, B and C) in response to TG treatment. These observations indicate that Ca²⁺ leak from the ER is required for PKCθ activation and localization to dot structures, as well as ER stress-induced autophagy.

FIGURE 6.

Calcium mediates PKCθ activation and autophagy in response to ER stress. Cells were treated continuously with TG (1 μm) in the presence or absence of BAPTA-AM (20 μm) for 6 h (A and B) or 8 h (C). A, cells were analyzed by immunofluorescence for LC3, and dot structures were quantified. B, Western blot analysis was performed for total PKCθ, phosphorylated PKCθ, and LC3. C, cells were analyzed by immunofluorescence for LC3 as described under “Experimental Procedures.” NT, no treatment.

PKCθ Activation Is Specifically Required for Autophagy in Response to ER Stress but Not in Response to Amino Acid Starvation—The signaling pathways that induce autophagy in mammalian cells have been most extensively characterized by analysis of the response to nutrition starvation. During nutrient starvation, mTOR kinase phosphorylates ATG13 to form an autophagy regulatory complex that is essential to initiate the autophagic response (54, 55). Therefore, we examined the requirement for PKCθ in the autophagic response to nutrition starvation. Phosphorylation of PKCθ (Thr⁵³⁸) was not observed in response to amino acid starvation conditions that induce autophagy (Fig. 7A). In addition, the autophagic response to amino acid starvation was not reduced by siRNA-mediated knockdown of PKCθ (Fig. 7, B and C). A marker for mTOR kinase activation is phosphorylation of p70 S6 kinase (p70 S6K) at Thr³⁸⁹. In response to amino acid starvation, p70 S6K was immediately dephosphorylated within 30 min, consistent with inactivation of mTOR kinase. In contrast, p70 S6K phosphorylation was not reduced upon treatment with TG (Fig. 7D). These observations indicate that ER stress-induced autophagy is executed in an mTOR-independent manner and that PKCθ is specifically required for ER stress-induced autophagy.

FIGURE 7.

PKCθ is not required for autophagy in response to amino acid starvation. The cells were subjected to transient 10 m in treatment with TG (1 μm) and then incubated for increasing periods of time in complete medium. In parallel, the cells were deprived of amino acids (AA) as described under “Experimental Procedures.” A, phosphorylation of PKCθ (Thr⁵³⁸) was monitored by Western blot analysis. B and C, autophagy in control GFP and PKCθ knockdown cells was analyzed by immunofluorescence microscopy for LC3 and quantification. D, cells were treated as above with TG or by amino acid deprivation for the indicated times. Cell extracts were prepared for Western blot analysis to detect phosphorylation of p70 S6 kinase (Thr³⁸⁹). NT, no treatment.

UPR Signaling Is Not Required for PKCθ Phosphorylation in Response to ER Stress—Previous reports have indicated that the IRE1α and PERK-eIF2α UPR pathways participate in the regulation of ER stress-induced autophagy. We have tested the requirements for UPR signaling in PKCθ activation by analysis of immortalized hepatocytes that harbor IRE1α deletion (supplemental Fig. S3), ATF6α deletion (24), or S51A knock-in mutation at the PERK phosphorylation site in eIF2α (63). The results show that ER stress-induced autophagy was significantly reduced in hepatocytes derived from S51A eIF2α knock-in mutant mice and partially reduced in hepatocytes lacking IRE1α (Fig. 8, A and B), consistent with previous findings (18, 20). In contrast, ER stress-induced autophagy was not significantly affected in hepatocytes deleted in ATF6α (Fig. 8C). However, hepatocytes from IRE1α-null or ATF6α-null mice, as well as hepatocytes from S51A knock-in mutant eIF2α mice, all exhibit ER stress-induced PKCθ phosphorylation. Therefore, PKCθ activation does not require signaling through the IRE1α, ATF6α, or PERK/eIF2α UPR pathways.

FIGURE 8.

PKCθ-mediated autophagy does not require UPR signaling. Ire1α^-/- (A), S51A eIF2α (B), and Atf6^-/- (C) mutant immortalized hepatocytes and their respective litter-matched control wild-type (Wt) cells were subject to transient TG treatment (1 μm 10 min) and cultured in TG-free medium for the indicated periods. The cell lysates were prepared and analyzed by Western blotting with PKCθ, phospho-PKCθ (Thr⁵³⁸), and LC3 antibodies (upper panels). In parallel, immunostaining was performed to quantify the percentage of cells containing LC3 dots (lower panels).

DISCUSSION

In this study, we demonstrated that PKCθ is specifically activated and required for autophagy in response to ER stress but not for autophagy in response to amino acid starvation. This conclusion is supported by the following observations: 1) ER stress induced by TG or TM coordinately induced phosphorylation of PKCθ within the activation loop at Thr⁵³⁸, conversion of LC3-I to LC3-II, and formation of LC3-containing dot structures in the cytoplasm, a typical marker of autophagy; 2) the nPKC inhibitor rottlerin and PKCθ siRNA-mediated knockdown significantly blocked autophagy in response ER stress; 3) ER stress was associated with localization of phosphorylated PKCθ to LC3-containing dot structures in the cytoplasm; 4) an increase in cytosolic Ca²⁺ was required for ER stress-induced PKCθ phosphorylation and localization to dot structures, as well as for LC3-I conversion to LC3-II and formation of LC3-containing dot structures; 5) PKCθ was not phosphorylated in response to amino acid starvation; and 6) amino acid starvation-induced autophagy was not blocked by PKCθ knockdown.

PKCθ was originally identified as an essential factor for T cell receptor/CD3-induced T cell activation. During T cell stimulation, PKCθ is activated by phospholipase Cγ1 and phosphatidylinositol 3-kinase, which are downstream of T cell receptor/CD3 and the Src/Syc family of protein-tyrosine kinases. PKCθ regulates the activation of activator protein-1, NF-κB, and cAMP response element-binding protein to stimulate interleukin-2 gene expression (56). However, very little is known about the role of PKCθ in other tissues and cell types. PKCθ was reported to inhibit insulin receptor signaling (57), although analysis of Pkcθ^-/- mice has yielded conflicting reports on the role of PKCθ in development of insulin resistance (58, 59). Recently, ER stress was also reported to cause insulin resistance in liver tissue (60, 61). Future studies are required to determine whether PKCθ activation couples ER homeostasis with insulin receptor signaling.

The precise mechanism by which PKCθ is activated in response to ER stress remains unknown. Ca²⁺ chelation by BAPTA-AM treatment inhibited both PKCθ phosphorylation and localization to dot structures in the cytosol, as well as autophagy in response to ER stress, suggesting that Ca²⁺ leak from the stressed ER plays a fundamental role in PKCθ activation. Because PKCθ is a PKC isoform that is not directly regulated by Ca²⁺, it is likely that Ca²⁺ plays an indirect role in PKCθ activation (Fig. 9). In preliminary studies we found that the phospholipase C inhibitor U73122 partially inhibited autophagy in response to ER stress (data not shown). Previously, the expression of a set of genes was shown to respond to altered ER Ca²⁺ homeostasis in a phospholipase C-dependent manner (62). Future studies should determine whether a Ca²⁺-dependent phospholipase C pathway activates PKCθ in response to ER stress.

FIGURE 9.

Model depicting the role of PKCθ in ER stress-induced autophagy. Disruption of ER homeostasis results in Ca²⁺ leak into the cytosol. The increase in cytosolic Ca²⁺ induces phosphorylation within the activation loop of PKCθ. PKCθ phosphorylation is inhibited by BAPTA-AM. Phosphorylated PKCθ translocates with LC3 to dot structures in the cytoplasm. PKCθ is required for ER stress-induced LC3 conversion and autophagy.

Our findings support the notion that PKCθ activation and autophagy in response to ER stress is independent from the mTOR kinase signal transduction pathway. This observation conflicts with recent findings by Hoyer-Hansen et al. (14) that concluded autophagy in response to TG-treatment occurs through inactivation of mTOR kinase that is mediated by CaMKKβ/AMP-activated protein kinase-dependent activation of tuberous sclerosis complex 2. It is possible that the discrepancy results from differences in the experimental conditions and/or cell types analyzed. Whereas Hoyer-Hansen et al. studied the autophagic response in MCF-7S1 breast carcinoma cells treated with 100 nm TG for 24 h, we analyzed responses in hepatocytes after a shorter time period (6-8 h) using 1 μm TG. Although we have not detected expression of PKCθ in MCF7 cells (data not shown), we have shown that PKCθ phosphorylation is required for ER stress-induced autophagy in another cell type, mouse embryonic fibroblasts. Therefore, we do not believe our findings are restricted to immortalized hepatocytes. Although we did not observe dephosphorylation of p70 S6K after 6-8 h of TG treatment, we did detect p70 S6K dephosphorylation after 12-16 h in hepatocytes (data not shown). It is possible that a PKCθ-dependent pathway is required for autophagy in response to acute severe ER stress, whereas an mTOR-dependent pathway is required for autophagy in response to chronic ER stress, possibly related to secondary effects of ER stress on nutrient metabolism.

Recently, ER stress-induced autophagy was independently reported to require either the IRE1α/c-Jun N-terminal kinase UPR subpathway or the PERK/eIF2α UPR subpathway (18, 20). We have observed that mutation at the PERK phosphorylation site in eIF2α or deletion of IRE1α did not interfere with PCKθ phosphorylation, although they did reduce ER stress-induced autophagy, consistent with previous findings (18, 20). In addition, both eIF2α phosphorylation and Xbp1 mRNA splicing were intact after TG treatment in PKCθ knockdown cells (data not shown). Therefore, it is unlikely that PKCθ is required for either IRE1α or PERK signaling. Finally, deletion of ATF6α did not reduce ER stress-induced autophagy or PKCθ activation. Therefore, UPR signaling appears to be independent of the PKCθ requirement for ER stress-induced autophagy. Studies are underway to identify upstream regulators and downstream targets of PKCθ activation that occur in response to ER stress.