Role of Autophagy in Histone Deacetylase Inhibitor-Induced Apoptotic and Nonapoptotic Cell Death

Role of autophagy in histone deacetylase inhibitor-induced apoptotic and nonapoptotic cell death

Noor Gammoha, Du Lama,1, Cindy Puentea, Ian Ganleyb, Paul A. Marksa,2, and Xuejun Jianga,2

aCell Biology Program, Memorial Sloan Kettering Cancer Center, New York, NY 10065; and bMedical Research Council Protein Phosphorylation Unit, University of Dundee, Dundee DD1 5EH, Scotland, United Kingdom

Contributed by Paul A. Marks, March 16, 2012 (sent for review February 13, 2012) Autophagy is a cellular catabolic pathway by which long-lived membrane vesicle. Nutrient and energy sensing can directly proteins and damaged organelles are targeted for degradation. regulate autophagy by affecting the ULK1 complex, which is Activation of autophagy enhances cellular tolerance to various comprised of the protein kinase ULK1 and its regulators, stresses. Recent studies indicate that a class of anticancer agents, ATG13 and FIP200 (10–12). Under nutrient-rich conditions, histone deacetylase (HDAC) inhibitors, can induce autophagy. One mammalian target of rapamycin (mTOR) directly phosphor- of the HDAC inhibitors, suberoylanilide hydroxamic acid (SAHA), is ylates ULK1 and ATG13 to inhibit the autophagy function of the currently being used for treating cutaneous T-cell lymphoma and ULK1 complex. However, amino acid deprivation inactivates under clinical trials for multiple other cancer types, including mTOR and therefore releases ULK1 from its inhibition. glioblastoma. Here, we show that SAHA increases the expression Downstream of the ULK1 complex, in the heart of the auto- of the autophagic factor LC3, and inhibits the nutrient-sensing phagosome nucleation and elongation, lie two ubiquitin-like kinase mammalian target of rapamycin (mTOR). The inactivation conjugation systems: the ATG12-ATG5 and the LC3-phospha- of mTOR results in the dephosphorylation, and thus activation, of tidylethanolamide (PE) conjugates (13). During autophagy, free the autophagic protein kinase ULK1, which is essential for cytosolic LC3 (termed LC3-I) becomes conjugated to PE autophagy activation during SAHA treatment. Furthermore, we (termed LC3-II). LC3-II is then incorporated to the growing CELL BIOLOGY show that the inhibition of autophagy by RNAi in glioblastoma autophagosome structure that, upon maturation, fuses with the cells results in an increase in SAHA-induced apoptosis. Importantly, lysosome compartment, leading to the degradation of the when apoptosis is pharmacologically blocked, SAHA-induced non- autophagosome content. apoptotic cell death can also be potentiated by autophagy in- The encapsulation and degradation of cytosolic materials by hibition. Overall, our ﬁndings indicate that SAHA activates autophagy aids in the clearance of damaged organelles and autophagy via inhibiting mTOR and up-regulating LC3 expression; misfolded proteins, and thereby plays an important role in the autophagy functions as a prosurvival mechanism to mitigate recycling of macromolecules and energy within the cells. In this SAHA-induced apoptotic and nonapoptotic cell death, suggesting context, autophagy may be regarded as a prosurvival mechanism that targeting autophagy might improve the therapeutic effects (14). Hence, autophagy is frequently activated during nutrient of SAHA. deprivation, hypoxia, and a wide range of anticancer therapy. The activation of autophagy has been frequently shown to inhibit transcription | ATG7 | necrosis the onset of apoptotic and necrotic cell death (15). However, in cases where autophagy may have an additive role in the death istone deacetylase (HDAC) inhibitors emerge as a new class process, autophagy may be regarded as a cell-death mechanism “ ” Hof therapeutic agents with promising outcomes during the (16). Here, excessive self-eating through autophagy may con- treatment of a wide range of cancer types (1). Hematological tribute to cell death by a yet unknown mechanism. Therefore, malignancies appear to be particularly sensitive to HDAC assessing the role of autophagy in a context-dependent manner is inhibitors; however, a number of additional cancer types are crucial, especially when considering whether autophagy-targeting currently being tested for their response to HDAC inhibition can be used during anticancer therapy. therapy. For an example, suberoylanilide hydroxamic acid We have previously shown that SAHA treatment induces (SAHA, vorinostat), which inhibits HDACs 1, 2, 3, and 6, has potent autophagy (8). In this study, we provide insights into the been approved for treatment against cutaneous T-cell lymphoma mechanism by which SAHA induces autophagy. We also show and also has modest effects as a single agent on cancers of the that autophagy-targeting can enhance SAHA-induced apoptotic prostate, ovaries, breast, colorectal, and glioblastoma (2, 3). and nonapoptotic cell death in glioblastoma cells. Although their precise mode of action remains uncertain, Results a number of recent data suggest that HDAC inhibitors may induce apoptotic cell death through both chromatin-dependent SAHA Induces Autophagy and LC3 Transcription in Mouse Embryonic and -independent mechanisms. Fibroblast Cells. To determine the mechanism by which the Treatment with HDAC inhibitors most frequently induces HDAC inhibitor SAHA induces autophagy, we treated mouse apoptosis via the programmed activation of a series of proteases, called caspases (4–6). More recently, HDAC inhibition has been also shown to induce autophagy (7, 8). Unlike apoptosis, the Author contributions: N.G., P.A.M., and X.J. designed research; N.G., D.L., C.P., and I.G. performed research; N.G., P.A.M., and X.J. analyzed data; and N.G., P.A.M., and X.J. wrote contribution of autophagy to cell death remains controversial the paper. and, most likely, context-dependent. Autophagy is a catabolic The authors declare no conﬂict of interest. process by which cytosolic material is targeted for lysosomal 1Present address: Novartis Oncology Global Development, Novartis Pharmaceuticals, Flor- degradation by means of double-membrane cytosolic vesicles, ham Park, NJ 07932. termed autophagosomes (9). The formation of autophagosomes 2To whom correspondence may be addressed. E-mail: [email protected] or marksp@ is orchestrated by upstream signaling molecules, including the mskcc.org. ULK1 and PI3K complexes, which signal to downstream com- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. plexes involved in the nucleation and maturation of the double- 1073/pnas.1204429109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1204429109 PNAS Early Edition | 1of5 Downloaded by guest on September 26, 2021 Fig. 1. SAHA induces autophagy and LC3 up-regulation in MEF cells. (A) MEF cells seeded in six-well dishes were treated with the indicated concentration of SAHA for 8 h. Where indicated, 20 nM Baf A1 was added 2 h before harvesting cells. Cell extracts were analyzed by Western blot using antibodies against the indicated proteins. The accumulation of LC3-II (faster migrating form) relative to LC3-I (slower migrating form) is indicative of the induction of autophagy. (B) MEF cells stably expressing GFP-LC3 grown on glass cover-slips were either left untreated or treated with 5 μMSAHAfor 24 h. Cells were then ﬁxed with 3.7% PFA, pro- cessed for imaging, and visualized under the confocal microscope using the 60× magniﬁcation objective. (C) MEF cells were seeded in six-well dishes and treated with 20 μM of SAHA for 20 h followed by 20 nM Baf A1 for a further 4 h. Cell extracts were − − analyzed by Western blot analysis using the indicated antibodies. (D) Wild-type or ATG3 / MEF cells were treated with the indicated concentrations of SAHA for 24 h. (E) A semiquantitative RT-PCR detecting LC3 expression was performed using RNA extracted from wt MEFs either treated with 10 μM SAHA or left untreated. GAPDH RT-PCR was used as a loading control.

embryonic fibroblast (MEF) cells with various concentrations of induction of autophagy by monitoring LC3 conversion. As shown SAHA and assayed the expression of the autophagy marker, in Fig. 2A, unlike wild-type MEFs that contained an intact LC3, by both Western blot and microscopy. As shown in Fig. 1A, autophagy response (Fig. 2A, Upper), LC3 conversion was de- treatment of MEF cells with SAHA resulted in the accumulation fective in ULK1/2 DKO MEFs (Fig. 2A, Lower), indicating that of LC3-II (faster migrating form); additional treatment of cells SAHA induces autophagy in a ULK1-dependent mechanism. with the lysosomal inhibitor Bafilomycin A1 (Baf A1) caused Previous reports suggest that the autophagy function of the a further increase of LC3-II level, demonstrating that SAHA ULK1 complex is suppressed by the nutrient-sensing kinase induces a full flux of autophagy, resulting in the lysosomal deg- mTOR (10, 12) and mTOR does so by directly phosphorylating radation of LC3-II. Induction of autophagy by SAHA was fur- ULK1 and its regulator ATG13. This finding prompted us to ther confirmed by imaging GFP-tagged LC3 expressed in MEF examine whether SAHA treatment can cause the inactivation of cells (Fig. 1B). In untreated cells, which mostly express LC3-I, mTOR and thereby activation of the ULK1 complex. Indeed, as GFP-LC3 showed a cytosolic, diffused localization (Fig. 1B, shown in Fig. 2B, two well-known mTOR substrates, p70S6K and Left). However, SAHA treatment resulted in the relocalization 4EBP, are dephosphorylated upon SAHA treatment in a similar of GFP-LC3 into punctate structures corresponding to auto- manner to that upon amino acid starvation. Importantly, in cells phagosomes (Fig. 1B, Right). Furthermore, to test whether treated with SAHA, ULK1 was dephosphorylated and thus ac- SAHA treatment results in the induction of a functional auto- tivated, as monitored by its faster mobility in SDS/PAGE. These phagic flux whereby the autophagosome content is targeted for results indicate that SAHA treatment can induce autophagy via lysosomal degradation, we monitored the degradation of relieving the ULK1 complex from inhibition by mTOR- a known autophagy substrate, p62 (17). As can be seen from Fig. mediated phosphorylation. 1C, p62 was dramatically destabilized in the presence of SAHA. This decrease in p62 levels was reverted by Baf A1, suggesting Autophagy Inhibition Accelerates Apoptotic Cell Death in T98G Cells. that SAHA can induce lysosome-mediated degradation of To study the role of autophagy in SAHA-induced cancer cell autophagy substrates, and thus a functional autophagic flux. death, we used the glioblastoma cell line, T98G, as a model. First, Overall, this molecular analysis confirms that SAHA can trigger we show that SAHA can induce autophagy, measured by LC3-I to robust autophagy in MEF cells. LC3-II conversion, in these cells (Fig. 3A). This conversion is Upon treatment of MEF cells with SAHA for 24 h, we noticed a dramatic increase in the unconjugated form of LC3, LC3-I (Fig. 1D). This increase is the case even in autophagy-defective − − ATG3 / MEFs. Because HDAC inhibitors were previously shown to impact gene expression, we tested whether the effect of SAHA treatment on LC3 levels was transcriptional. We performed a semiquantitative RT-PCR analysis on RNA extracted from wild-type MEFs that were left untreated or treated with 10 μM of SAHA for 18 h. As shown in Fig. 1E, SAHA treatment indeed increased LC3 transcript levels.

SAHA Induces Autophagy by Suppressing mTOR and Activating the ULK1 Complex. Although transcriptional up-regulation of LC3 can contribute to SAHA-induced autophagy, it is nevertheless in- sufﬁcient for autophagy initiation, which requires activation of upstream signaling to induce the conversion of LC3-I to LC3-II. Fig. 2. SAHA triggers autophagy by suppressing mTOR and activating The ULK1 complex has been shown to be the most upstream ULK1. (A) ULK1/2 DKO or MEFs with intact ULK1/2 expression (wild-type, wt component in the core autophagy pathway and is essential for MEFs) were treated with the indicated concentrations of SAHA for 24 h. Baf autophagy induced by various conditions (18, 19). Therefore, we A1 (20 nM) was included 2 h before harvesting and cell lysates were ana- ﬁ lyzed by Western blot using the indicated antibodies. (B) MEF cells were rst tested whether ULK1 activity is essential for SAHA-medi- either left untreated, amino acid-starved for 2 h, or treated with 5 μMof ated autophagy. For this purpose, we used MEF cells with ge- SAHA for 16 h. Cell extracts were then analyzed by Western blot analysis netic deletions of both ULK1 and its functional homolog ULK2 using antibodies against the proteins indicated. A change in migration of (ULK1/2 double knockout, DKO) (20). We treated both ULK1/ ULK1, 4EBP, and p70S6K to faster migrating forms is indicative of protein 2 DKO MEFs and wild-type MEFs with SAHA and assessed the dephosphorylation and thus mTOR inactivation.

2of5 | www.pnas.org/cgi/doi/10.1073/pnas.1204429109 Gammoh et al. Downloaded by guest on September 26, 2021 dependent upon intact autophagy machinery and is impeded upon survival during SAHA and zVAD treatment. These results shRNA-mediated knockdown of ATG7. Furthermore, we show demonstrate that autophagy acts as a protective mechanism to that GFP-LC3 localizes to punctate structures, autophagosomes, reduce SAHA-induced nonapoptotic cell death. in the presence of SAHA (Fig. 3B), thus further supporting that SAHA treatment induces autophagy in T98G cells. Subsequently, Discussion to investigate the role of autophagy during apoptotic cell death In this study, we investigate the mechanisms by which the induced by SAHA, we used an in vitro assay to measure caspase-3 HDAC inhibitor SAHA induces autophagy and the role of activity, and compared the induced apoptosis in T98G cells har- autophagy in cancer cell survival during SAHA treatment. We boring control shRNA or shRNA against ATG7. Caspase-3 ac- found that SAHA can enhance autophagy by influencing the tivity was greatly enhanced at various concentrations of SAHA in autophagy pathway at various points. Furthermore, we provide ATG7 knockdown cells compared with control knockdown (Fig. evidence that autophagy inhibition results in a higher level of 4A). This increase in caspase-3 activity is specific to caspase acti- apoptotic and nonapoptotic cell death in response to SAHA vation and can be suppressed by cotreating cells with the pan- treatment. Overall, our study suggests that targeting autoph- caspase inhibitor, zVAD. We also confirmed that ATG7 knock- agy during cancer treatment with SAHA may augment down increases caspase-3 activation in the presence of SAHA by therapeutic effects. Western blot as indicated by the accumulation of the active, Mechanistically, SAHA regulates autophagy by both inducing cleaved form of caspase-3 in these cells (Fig. 4B). Furthermore, we LC3 expression transcriptionally and inactivating mTOR, which measured cell viability during SAHA treatment, and compared might be indirect or not at all related to transcription. Previously, control and ATG7 knockdown cells. Consistently, ATG7 knock- HDAC inhibitors were shown to both increase and decrease the down resulted in a decrease in cell survival during SAHA treat- transcriptional expression of certain genes. HDAC inhibitors can ment compared with control knockdown cells (Fig. 4C), which induce cell-cycle arrest by increasing the expression of the cell- correlates with an increase in cell detachment (Fig. 4D). Overall, cycle inhibitor p21, and decreasing the expression of cyclin- these results indicate that autophagy protects T98G cells from dependent kinases required for cell cycle progression (21). In apoptotic cell death and enhances cell survival in the presence addition, HDAC inhibition resulted in impaired DNA repair of SAHA. during cytotoxic damage at least in part by downregulating the expression of double-stranded break repair machinery including Autophagy Suppression Decreases Cell Survival in the Absence of Ku70, RAD51, and BRCA1/2 (22, 23). In the case of LC3, we Apoptosis. We have previously shown that SAHA can also induce observed a transcription-dependent increase in LC3 levels in- CELL BIOLOGY nonapoptotic cell death when apoptosis is blocked (8). In our duced by SAHA. During persistent autophagy, LC3 protein assays, cells treated with zVAD, which completely abolished levels may drop as LC3 is conjugated to autophagosome mem- caspase activation (Fig. 4A), still underwent autophagy induced brane, which in turn is targeted for degradation upon fusion with by SAHA (Fig. 5A). To test whether autophagy may cause an the lysosome compartment. Therefore, up-regulating LC3 tran- increased cell survival in the absence of apoptotic cell death, we scription during autophagy becomes an important mechanism to measured cell survival in cells cotreated with SAHA and zVAD. avoid the exhaustion of the pathway during prolonged treatment. As shown in Fig. 5B, ATG7 knockdown, which blocks autophagy On the other hand, although transcriptional up-regulation of at least in part (Fig. 5A), results in a decreased cell survival during LC3 may be important to ensure persisting autophagy, it is not fi SAHA and zVAD cotreatment. This effect correlates with an suf cient to induce autophagy. Indeed, our results indicate that increase in cell detachment of ATG7 knockdown cells compared SAHA can cause the inactivation of mTOR, thus inducing with control knockdown (Fig. 5C). In addition, the clonogenic autophagy through a ULK1 complex-dependent mechanism survival of cells treated with SAHA and zVAD was reduced in (Fig. 2). The exact mechanism by which mTOR is inactivated by ATG7 knockdown cells compared with control knockdown cells HDAC inhibition remain to be unraveled. SAHA does not have (Fig. 5D), further suggesting that ATG7 knockdown reduces cell an effect on mTOR levels or acetylation, using the conditions empolyed in this study (Fig. S1). Therefore, it is not clear whether SAHA inactivates mTOR via inhibition of histone deacetylation and thus transcription of certain genes or via inhibition of deacetylation of a nonhistone protein that is involved in mTOR regulation. As with many other anticancer therapies, SAHA-induced autophagy appears to act as a prosurvival mechanism to coun- teract the cytotoxic activity of SAHA. Autophagy may delay the onset of apoptosis during SAHA treatment through various mechanisms including clearance of reactive oxygen species that are generated during SAHA treatment (24), clearance of p62- containing protein aggregates, which may accumulate during HDAC inhibition, and clearance of damaged mitochondria (25). Abrogated apoptosis frequently contributes to chemotherapy resistance. SAHA was previously shown to induce nonapoptotic cell death in cells deficient in apoptotic machinery or cotreated with zVAD (8, 26). The nature of this type of nonapoptotic cell death triggered by SAHA remains unresolved. Our preliminary Fig. 3. SAHA induces autophagy in T98G glioblastoma cells. (A) T98G studies suggest that inhibiting RIP1 kinase, a key player in death glioblastoma cell lines expressing either control shRNA or shRNA targeting receptor-mediated necrosis by necrostatin A (27), does not in- ATG7 sequences were treated with SAHA using the indicated concentrations hibit cell death triggered by SAHA and zVAD cotreatment (Fig. and times. Cell lysates were then subjected to Western blot analysis using the following antibodies: anti-ATG7 to confirm knockdown efficiency, anti- S2). However, this result does not rule out necrosis, as RIP1- LC3 to analyze autophagy induction and anti-actin as a loading control. (B) independent necrosis has been previously reported (28). T98G stably expressing GFP-LC3 were grown on glass cover-slips and treated A wide range of targeted cancer therapeutic agents, such as with 10 μM SAHA for the indicated times. Slides were then fixed with PFA mTOR inhibitors, AKT inhibitors, and proteasome inhibitors and analyzed by confocal microscopy using a 20× magnification objective. were shown to induce autophagy in cancer cells as conventional

Gammoh et al. PNAS Early Edition | 3of5 Downloaded by guest on September 26, 2021 Fig. 4. ATG7 knockdown in T98G cells increases SAHA- induced apoptotic cell death. (A) An in vitro caspase-3 assay performed on cell lysates from T98G cells expressing control or ATG7 shRNA. Cells were treated with various concentrations of SAHA in the presence or absence of the pan-caspase inhibitor, zVAD, at 10 μM. The y axis values correspond to RFU as a measure of caspase activity using a fluorescence-based caspase-3 substrate. Error bars correspond to SEM values of at least three independent assays. (B) Western blot analysis to detect expression of the active cleaved caspase-3 levels. Extracts are taken from T98G cells expressing control or ATG7 shRNA left untreated or treated with SAHA for 24 h. (C) Cell survival assay using Resazurin dye. T98G cells expressing control or ATG7 shRNA were seeded in a 96-well plate at 500 cells per well and treated with 10 μM SAHA for the indicated times. Resazurin dye was then added to each well and plate and was further incubated for 3 h before measurement of the Resazurin fluorescence. Cell numbers were then deduced from a standard curve and survival was measured relative to untreated samples, which corresponded to 100% survival. Error bars correspond to SEM values of at least three independent assays. (D) T98G cells expressing control or ATG7 shRNA were seeded in a six-well plate, left untreated, or treated with SAHA for 48 h. Phase-contrast images of cells were then captured under Nikon fluorescent scope using a 20× magnification objective.

chemotherapies (29). The exact role of autophagy in these For Western blot analysis, the following antibodies were used: anti-LC3 treatments may vary, depending on the contexts of individual (Sigma; #L7543); anti-Actin (Sigma; #A5316); anti-ULK1 (Sigma; #A7481); tumors and treatments. As suggested by this study, when treating anti-caspase-3 (Cell Signaling; clone 8G10); anti-S6K (Cell Signaling; total resistant tumors, such as recurrent glioblastoma (30–32), autophagy- #2708, phospho #9205); anti-4-EBP (Cell Signaling; #9452); anti-ATG7 (Santa Cruz; clone H300); anti-p62/SQSTM1 (MBL, #PM045B). targeting might prove to be a unique combinational therapy that can potentiate the anticancer effect of SAHA treatment. SAHA Treatment. MEF and T98G cells were seeded at 3 × 10⁵ per well in a six- well dish. Twenty-four hours later, fresh medium containing the indicated Materials and Methods concentration of SAHA was added on the cells. Treatment with zVAD along −/− Cell Culture. Wild-type MEF cells, ULK1/2 DKO MEFs, ATG3 knockout (ATG3 ) with SAHA was included as indicated at a final concentration of 10 μM. The MEFs, and T98G glioblastoma cell line were cultured in DMEM supplemented lysosome degradation inhibitor, Baf A1, was added 2 h before harvest at with 10% (vol/vol) FBS, L-Glutamine (2 mM), pencillin (10 Units/mL), and a final concentration of 20 nM. streptomycin (0.1 mg/mL). Cells stably expressing GFP-LC3 were generated by retroviral infection using pBabe-GFP-LC3 followed by blasticidin selection. Microscopy. For microscopy, T98G or MEF cells stably expressing GFP-LC3 were For amino acid starvation experiments, cells were grown in DMEM lacking grown on glass cover-slips in a six-well plate. Twenty-four hours later, SAHA amino acids and serum for 2 h before harvest, as described previously (33). treatment was applied as indicated. Cover-slips were then fixed with 3.7% (vol/vol) paraformaldehyde in 20 mM Hepes pH 7.5 for 30 min at room Reagents and Antibodies. The following drugs were dissolved in DMSO: SAHA temperature. Cover-slips were then mounted on microscope slides and pro- (Chemietek), zVAD-FMK (Enzo Life Sciences), Baf A1 (Sigma). cessed using Nikon Confocal Microscope using the 20× or 60× magnification

Fig. 5. ATG7 knockdown decreases cell survival in the absence of apoptosis. (A) Western blot analysis of T98G cells expressing control or ATG7 shRNA. Cells were left untreated or treated as indicated with a combination of SAHA and zVAD at 10 μM each for 48 h. (B) Cell survival assay using Resazurin dye as in Fig. 4C except that SAHA concentration was used at 20 μM and zVAD treatment at 10 μM was included in all samples. Error bars correspond to SEM values of at least three independent assays. (C) Cell detachment was analyzed as in Fig. 4D (20× magniﬁcation), except that 10 μM zVAD was included along with SAHA treatment. (D) Colony formation assay of T98G cells treated with 20 μM SAHA and 10 μM zVAD for 48 h. Cells were further grown for 2 wk in fresh media before ﬁxation, staining and counting of colonies.

4of5 | www.pnas.org/cgi/doi/10.1073/pnas.1204429109 Gammoh et al. Downloaded by guest on September 26, 2021 objectives. For cell death imaging, T98G cells were grown in six-well plates TECAN at excitation and emission 485 nm and 535 nm, respectively. Average for 24 h before treatment with SAHA, as indicated. Cell detachment was relative ﬂuorescence units (RFU) at 60 min were plotted and SEM of at least × ﬁ then acquired using a Nikon Fluorescent scope (20 magni cation objective). three independent assays.

RNAi. RNAi-mediated knockdown was performed using a doxycycline in- Resazurin Cell Survival Assay. T98G cells were seeded in 96-well plates (500 fi ducible pTRIPz vector (Thermo Scienti c) expressing shRNAmir against ATG7 cells per well). Twenty-four hours later, treatment with SAHA with or without or nontarget control (ATG7 KD or control KD). The sequences targeted by the zVAD was initiated as indicated. The Resazurin dye was added to each well shRNAmir were as follows, ATG7 KD: 5′ CCAGTTCAGAGCTAAATAATA; according to the manufacturer’s instruction (R&D Systems). Three hours control KD: 5′ CAACAAGATGAAGAGCACCAA. Lentiviruses harboring the fl fl knockdown sequences were used to infect T98G cells. Cells when then se- later, uorescence in plate was measured using a TECAN uorescence plate lected with puromycin for at least 3 d, followed by addition of doxycycline to reader with excitation and emission of 535 nm and 590 nm, respectively. Cell induce shRNA and red fluorescent protein expression. Cell sorting by FACS numbers were derived from a standard curve and cell survival calculated as was performed to select the 20% highest red fluorescent protein-expressing a percentage of mock treatment, which corresponded to 100% survival. population, which was further propagated and tested for knockdown. Error bars correspond to SEM of at least three independent assays.

RT-PCR. MEF cells were harvest by directly lysing in RLT lysing reagent fol- Colony Formation Assay. T98G cells were seeded in 6-cm dishes, 24 h later cells lowed by RNA extraction using RNeasy (Qiagen) according to the manu- were treated with SAHA (20 μM) and zVAD (10 μM) for a further 48 h. ’ μ facturer s instruction (Qiagen). Reverse transcription using 1 g of RNA as Inhibitors were then washed off and cells were grown in fresh growth media a template was used to produce cDNA using iScript cDNA synthesis kit for 2 wk before ﬁxation with 10% (vol/vol) formaldehyde and staining with (BioRad) according to the manufacturer’s instructions. PCR was then per- Giemsa stain (Sigma) to visualize cell colonies. formed to amplify LC3 and GAPDH sequences.

ACKNOWLEDGMENTS. We thank Dr. Craig Thompson for the ULK1/2 double- Caspase Assay. Floating and attached cells (collected by trypsinization) were − − knockout cells and Dr. Masaaki Komatsu for ATG3 / cells. This work was lysed. Twenty micrograms of cell lysate was then mixed with caspase-3 supported in part by a Goodwin Experimental Therapeutic Center fund, and fi μ (Apopain) substrate Rhodamine 110 (AnaSpec; nal concentration, 1.5 M) National Institutes of Health U54CA137788/U54CA132378 and R01CA113890 on ice in a final volume of 20 μL. Samples were then immediately loaded on funds (to X.J.); the David Koch Foundation and the Cap Cure Foundation (to 96-well plates and fluorescence was measured at 30 °C for 180 min using P.A.M.); and National Institutes of Health Fellowship 1F32CA162691 (to N.G.).

1. Khan O, La Thangue NB (2012) HDAC inhibitors in cancer biology: Emerging mech- 18. Itakura E, Mizushima N (2010) Characterization of autophagosome formation site by

anisms and clinical applications. Immunol Cell Biol 90:85–94. a hierarchical analysis of mammalian Atg proteins. Autophagy 6:764–776. CELL BIOLOGY 2. Galanis E, et al. (2009) Phase II trial of vorinostat in recurrent glioblastoma multi- 19. Chan EY, Kir S, Tooze SA (2007) siRNA screening of the kinome identifies ULK1 as forme: A north central cancer treatment group study. J Clin Oncol 27:2052–2058. a multidomain modulator of autophagy. J Biol Chem 282:25464–25474. 3. Marks PA, Breslow R (2007) Dimethyl sulfoxide to vorinostat: Development of this 20. Cheong H, Lindsten T, Wu J, Lu C, Thompson CB (2011) Ammonia-induced autophagy histone deacetylase inhibitor as an anticancer drug. Nat Biotechnol 25:84–90. is independent of ULK1/ULK2 kinases. Proc Natl Acad Sci USA 108:11121–11126. 4. Rosato RR, Almenara JA, Dai Y, Grant S (2003) Simultaneous activation of the intrinsic 21. Lin YC, et al. (2008) Statins increase p21 through inhibition of histone deacetylase and extrinsic pathways by histone deacetylase (HDAC) inhibitors and tumor necrosis activity and release of promoter-associated HDAC1/2. Cancer Res 68:2375–2383. factor-related apoptosis-inducing ligand (TRAIL) synergistically induces mitochondrial 22. Lee JH, Choy ML, Ngo L, Foster SS, Marks PA (2010) Histone deacetylase inhibitor – damage and apoptosis in human leukemia cells. Mol Cancer Ther 2:1273 1284. induces DNA damage, which normal but not transformed cells can repair. Proc Natl 5. Emanuele S, Lauricella M, Tesoriere G (2008) Histone deacetylase inhibitors: Apoptotic Acad Sci USA 107:14639–14644. – effects and clinical implications (Review). Int J Oncol 33:637 646. 23. Adimoolam S, et al. (2007) HDAC inhibitor PCI-24781 decreases RAD51 expression and 6. Zhang J, et al. (2011) Mule determines the apoptotic response to HDAC inhibitors by inhibits homologous recombination. Proc Natl Acad Sci USA 104:19482–19487. – targeted ubiquitination and destruction of HDAC2. Genes Dev 25:2610 2618. 24. Ungerstedt JS, et al. (2005) Role of thioredoxin in the response of normal and 7. Robert T, et al. (2011) HDACs link the DNA damage response, processing of double- transformed cells to histone deacetylase inhibitors. Proc Natl Acad Sci USA 102: strand breaks and autophagy. Nature 471:74–79. 673–678. 8. Shao Y, Gao Z, Marks PA, Jiang X (2004) Apoptotic and autophagic cell death induced 25. Ellis L, et al. (2009) The histone deacetylase inhibitors LAQ824 and LBH589 do not by histone deacetylase inhibitors. Proc Natl Acad Sci USA 101:18030–18035. require death receptor signaling or a functional apoptosome to mediate tumor cell 9. Mizushima N, Komatsu M (2011) Autophagy: Renovation of cells and tissues. Cell 147: death or therapeutic efficacy. Blood 114:380–393. 728–741. 26. Yamamoto S, et al. (2008) Suberoylanilide hydroxamic acid (SAHA) induces apoptosis 10. Ganley IG, et al. (2009) ULK1.ATG13.FIP200 complex mediates mTOR signaling and is or autophagy-associated cell death in chondrosarcoma cell lines. Anticancer Res essential for autophagy. J Biol Chem 284:12297–12305. 28(3A):1585–1591. 11. Kamada Y, et al. (2000) Tor-mediated induction of autophagy via an Apg1 protein 27. Degterev A, et al. (2008) Identification of RIP1 kinase as a specific cellular target of kinase complex. J Cell Biol 150:1507–1513. – 12. Hosokawa N, et al. (2009) Nutrient-dependent mTORC1 association with the ULK1- necrostatins. Nat Chem Biol 4:313 321. Atg13-FIP200 complex required for autophagy. Mol Biol Cell 20:1981–1991. 28. Upton JW, Kaiser WJ, Mocarski ES (2010) Virus inhibition of RIP3-dependent necrosis. – 13. Geng J, Klionsky DJ (2008) The Atg8 and Atg12 ubiquitin-like conjugation systems in Cell Host Microbe 7:302 313. macroautophagy. ‘Protein modifications: Beyond the usual suspects’ review series. 29. Janku F, McConkey DJ, Hong DS, Kurzrock R (2011) Autophagy as a target for anti- – EMBO Rep 9:859–864. cancer therapy. Nat Rev Clin Oncol 8:528 539. 14. Mathew R, Karantza-Wadsworth V, White E (2007) Role of autophagy in cancer. Nat 30. Kanzawa T, et al. (2004) Role of autophagy in temozolomide-induced cytotoxicity for Rev Cancer 7:961–967. malignant glioma cells. Cell Death Differ 11:448–457. 15. Duprez L, Wirawan E, Vanden Berghe T, Vandenabeele P (2009) Major cell death 31. Fan QW, Weiss WA (2011) Autophagy and Akt promote survival in glioma. Autoph- pathways at a glance. Microbes Infect 11:1050–1062. agy 7:536–538. 16. Kroemer G, Mariño G, Levine B (2010) Autophagy and the integrated stress response. 32. Fan QW, et al. (2010) Akt and autophagy cooperate to promote survival of drug- Mol Cell 40:280–293. resistant glioma. Sci Signal 3:ra81. 17. Bjørkøy G, et al. (2005) p62/SQSTM1 forms protein aggregates degraded by au- 33. Ganley IG, Wong PM, Gammoh N, Jiang X (2011) Distinct autophagosomal-lysosomal tophagy and has a protective effect on huntingtin-induced cell death. J Cell Biol 171: fusion mechanism revealed by thapsigargin-induced autophagy arrest. Mol Cell 42: 603–614. 731–743.

Gammoh et al. PNAS Early Edition | 5of5 Downloaded by guest on September 26, 2021