Atg5 Regulates Phenethyl Isothiocyanate–Induced Autophagic and Apoptotic Cell Death in Human Prostate Cancer Cells

Molecular basis for onset and progression of prostate cancer is not fully understood (1), but this malignancy continues to be a leading cause of cancer-related deaths among men in the United States. Prostate cancer is usually diagnosed in the sixth and seventh decades of life, providing a large window of opportunity for intervention to prevent or slow disease progression. Therefore, identification and preclinical/clinical development of novel agents that are nontoxic but can delay onset and/or progression of human prostate cancer is highly desirable. Epidemiologic studies have indicated that dietary intake of cruciferous vegetables may lower the risk of different types of malignancies, including cancer of the prostate (2, 3). Anticarcinogenic effect of cruciferous vegetables is attributed to organic isothiocyanates, which are released on processing (cutting or chewing) of these vegetables (4, 5). Phenethyl isothiocyanate (PEITC) is one such naturally occurring isothiocyanate compound that has received increasing attention due to its cancer chemopreventive effects. For example, PEITC is a potent inhibitor of pulmonary tumorigenesis in rats induced by tobacco-specific carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (6). Prevention of chemical carcinogenesis by PEITC has also been observed against diethylnitrosamine-induced hepatocellular adenomas in C3H mice and N-nitrosobenzylmethylamine–induced esophageal cancer in rats (7, 8). Dietary N-acetylcysteine conjugate of PEITC during the postinitiation phase inhibited benzo[a]pyrene-induced lung adenoma multiplicity in mice (9).

More recent studies have revealed that PEITC treatment suppresses growth of prostate cancer cells in culture and in vivo (10–16). The mechanism of antiproliferative effect of PEITC is not fully understood but known cellular responses to this promising natural product in cultured cancer cells include activation of c-Jun NH₂-terminal kinases and extracellular signal-regulated kinases, activation of G₂-M phase checkpoint, apoptosis induction, suppression of nuclear factor-κB–regulated gene expression, inhibition of epidermal growth factor receptor signaling, repression of androgen receptor expression, and inhibition of cap-dependent translation (10–19). We have also shown previously that PEITC treatment suppresses angiogenesis in vitro and ex vivo at pharmacologically relevant concentrations in association with inhibition of serine-threonine kinase Akt (20).

The Akt-mammalian target of rapamycin (mTOR) signaling pathway has assumed a central role in regulation of autophagy, which is an evolutionarily conserved and dynamic process for bulk degradation of cellular macromolecules and organelles (21–23). Because PEITC treatment inhibits Akt activity in human prostate cancer cells (17, 20), the present study was designed to address the question of whether anticancer effect of PEITC is mediated by autophagic cell death. We now show that PEITC treatment causes Atg5-dependent autophagic as well as apoptotic cell death in human prostate cancer cells.

Reagents. PEITC (purity >99%) was purchased from Sigma-Aldrich. Cell culture reagents and fetal bovine serum (FBS) were purchased from Life Technologies; 3-methyl adenine (3-MA) and acridine orange were from Sigma-Aldrich; rapamycin was from Calbiochem; and a kit for quantification of cytoplasmic histone-associated DNA fragmentation was purchased from Roche Diagnostics. Antibody against microtubule-associated protein 1 light chain 3 (LC3) was from MBL and the antibodies against cleaved caspase-3, Atg5, phospho-(S473)-Akt, total Akt, phospho-(S2448)-mTOR, phospho-(T389)-p70s6k, total p70s6k, and Rheb were from Cell Signaling. The antibody against total mTOR was from Calbiochem.

Cell lines. The LNCaP and PC-3 cell lines were procured from the American Type Culture Collection. Monolayer culture of LNCaP cells was maintained in RPMI 1640 supplemented with 1 mmol/L sodium pyruvate, 10 mmol/L HEPES, 0.2% glucose, 10% (v/v) FBS, and antibiotics. PC-3 cells were cultured in F-12K Nutrient Mixture supplemented with 7% non–heat-inactivated FBS and antibiotics. Normal human prostate epithelial cell line PrEC was purchased from Clonetics and maintained in PrEBM (Cambrex). Stock solution of PEITC was prepared in DMSO and diluted with medium for cellular studies and with PBS for the in vivo experiment. Final concentration of DMSO was <0.1% for cellular studies and 0.1% for the in vivo experiment.

Transmission electron microscopy. Transmission electron microscopy was performed as described by us previously (24). Briefly, LNCaP or PC-3 cells (2 × 10⁵) were seeded in six-well plates and allowed to attach by overnight incubation. The cells were treated with either DMSO (final concentration, <0.1%) or 5 μmol/L PEITC for 6 or 9 h at 37°C. Cells were fixed in ice-cold 2.5% electron microscopy grade glutaraldehyde [in 0.1 mol/L PBS (pH 7.3)], rinsed with PBS, postfixed in 1% osmium tetroxide with 0.1% potassium ferricyanide, dehydrated through a graded series of ethanol (30–90%), and embedded in Epon. Semithin sections (300 nm) were cut using a Reichart Ultracut, stained with 0.5% toluidine blue, and examined under a light microscope. Ultrathin sections (65 nm) were stained with 2% uranyl acetate and Reynold's lead citrate and examined using JEOL 1210 transmission electron microscope at ×5,000 and ×30,000 magnifications.

Detection of acidic vesicular organelles. Cells (1 × 10⁵) were plated on coverslips in 12-well plates and allowed to attach by overnight incubation. Following treatment with DMSO (control) or PEITC, cells were stained with 1 μg/mL acridine orange in PBS for 15 min, washed with PBS, and examined under a Leica fluorescence microscope at ×100 magnification.

Immunofluorescence microscopy for LC3 localization. Cells (1 × 10⁵) were grown on coverslips in 12-well plates and allowed to attach by overnight incubation. Cells were then exposed to the desired concentration of PEITC or DMSO (control) for specified time periods at 37°C. The cells were washed with PBS and fixed in 2% paraformaldehyde overnight at 4°C. The cells were permeabilized with 0.1% Triton X-100 for 15 min at room temperature, washed with PBS, and blocked with bovine serum albumin (BSA) buffer (PBS containing 0.5% BSA and 0.15% glycine) for 1 h at room temperature. The cells were then treated with the anti-LC3 antibody (1:50 dilution in BSA buffer) for 1 h at room temperature followed by incubation with 2 μg/mL Alexa Fluor 488–conjugated goat anti-rabbit secondary antibody (Molecular Probes) for 1 h at room temperature. The cells with punctate pattern of LC3 localization were visualized under a Leica fluorescence microscope.

Immunoblotting. Cells were treated with the desired concentrations of PEITC or DMSO (control) for specified time periods and lysed as described by us previously (25). Lysate proteins were resolved by SDS-PAGE and transferred onto polyvinylidene fluoride membrane. Immunoblotting was performed as described by us previously (24, 25).

Transient transfection. PC-3 or LNCaP cells (1 × 10⁵) were plated in six-well plates, allowed to attach, and transiently transfected with pCMV6 vector encoding constitutively active Akt-1 (kindly provided by Dr. Daniel Altschuler, University of Pittsburgh, Pittsburgh, PA), pcDNA3.1 vector encoding Rheb, or corresponding empty vector using Fugene6 transfection reagent. Twenty-four hours after transfection, the cells were treated with DMSO (control) or 5 μmol/L PEITC for 9 h. The cells were then processed for immunofluorescence microscopy for LC3 or immunoblotting as described above.

RNA interference of Atg5. PC-3 or LNCaP cells (10⁵ in six-well plates) were transfected at 30% to 50% confluency with either 100 nmol/L Atg5-targeted small interfering RNA (siRNA; pool of three target-specific 20- to 25-nucleotide siRNAs; Santa Cruz Biotechnology) or a control nonspecific siRNA (Qiagen). Twenty-four hours after transfection, cells were exposed to 5 μmol/L PEITC or DMSO and the incubation was continued for an additional 9 h. Cells were then collected and processed for immunoblotting of Atg5, LC3, and caspase-3, analysis of cytoplasmic histone-associated apoptotic DNA fragmentation, or immunofluorescence microscopy for LC3 localization.

Xenograft study. Male nude mice (6–8 wk old) were divided into two groups of five mice per group. The experimental mice were gavaged with 9 μmol PEITC in 0.1 mL PBS containing 0.1% DMSO five times per week, whereas the control mice received 0.1 mL vehicle. Exponentially growing PC-3 cells (3 × 10⁶) were injected s.c. onto the right flank of each mouse above the hind limb. Treatment was started on the day of tumor cell injection and continued until the termination of the experiment (38 d after tumor cell injection). Tumor volume was measured as described by us previously (26). Tumor tissues were dissected and used for immunohistochemistry or immunoblotting. A portion of tumor tissue from control and PEITC-treated mice was fixed in 10% neutral-buffered formalin, dehydrated, embedded in paraffin, and sectioned at 4 to 5 μm thickness. Apoptotic bodies in tumor sections were visualized by terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assay using ApopTag Plus Peroxidase In Situ Apoptosis kit (Chemicon International) according to the manufacturer's instructions. Representative tumor sections from control and PEITC-treated mice were fixed in acetone for 10 min at 4°C. After endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 15 min, the sections were treated with normal serum for 20 min. The sections were then incubated with anti-LC3 antibody (1:100 dilution) for 1 h at room temperature, washed with TBS, and treated with biotinylated anti-rabbit IgG for 30 min at room temperature. Characteristic brown color was developed by incubation with 3,3′-diaminobenzidine. At least three nonoverlapping images from each tissue section were captured using a camera mounted onto the microscope.

Assessment of autophagy in Caenorhabditis elegans. The DA2123 strain (adIs2122[lgg-1:GFP rol-6(df)]) was generously provided by Dr. Chanhee Kang (University of Texas Southwestern, Dallas, TX). Adult worms were transferred to plates spotted with PEITC (10 or 25 μmol/L) or DMSO (control). The plates were incubated at 20°C overnight. The worms were mounted in M9 medium containing 10 mmol/L sodium azide and photographed using an Olympus BX51 upright microscope at 40× objective lens magnification.

PEITC treatment caused autophagy in prostate cancer cells. Figure 1A depicts effect of PEITC treatment on ultrastructures of LNCaP and PC-3 cells. The DMSO-treated control LNCaP and PC-3 cells exhibited normal complement of healthy looking mitochondria. Exposure of LNCaP and PC-3 cells to 5 μmol/L PEITC for 6 hours (Fig. 1A) or 9 hours (data not shown) resulted in appearance of membranous vacuoles resembling autophagosomes (Fig. 1A,, arrows), which were rare in DMSO-treated controls. We confirmed autophagic response to PEITC by analysis of acidic vesicular organelles (AVO) and processing and recruitment of LC3, which are hallmarks of autophagy (27–29). As can be seen in Fig. 1B, the DMSO-treated control LNCaP and PC-3 cells primarily exhibited green fluorescence. On the other hand, treatment of LNCaP and PC-3 cells with rapamycin (positive control) and PEITC resulted in formation of yellow-orange AVOs. The LC3 protein (18 kDa) is cleaved by autophagic stimuli to a 16-kDa intermediate (LC3-II) that localizes to the autophagosomes (29). Recruitment of LC3-II to the autophagosomes is characterized by punctate pattern of its localization. The DMSO-treated control LNCaP cells (9-hour treatment) exhibited diffuse and weak LC3-associated green fluorescence (Fig. 1C). On the other hand, the LNCaP cells treated for 9 hours with 5 μmol/L PEITC exhibited punctate pattern of LC3 immunostaining (Fig. 1C). A marked increase in fraction of cells with punctate pattern of LC3 over DMSO-treated control was also observed in LNCaP and PC-3 cells treated for 6 hours with 2.5 and 5 of μmol/L PEITC (results not shown). Punctate pattern of LC3 localization was not readily apparent by 5 μmol/L PEITC treatment for 6 hours (Fig. 1C) or 9 hours (results not shown) in a normal prostate epithelial cell line PrEC. Treatment of LNCaP and PC-3 cells with 2.5 and 5 μmol/L of PEITC for 6 and 9 hours resulted in cleavage of LC3 (Fig. 1D). The control PrEC cells (DMSO treated) exhibited higher levels of basal cleaved LC3. The PEITC-mediated cleavage of LC3 was much less pronounced in the PrEC cell line compared with prostate cancer cells (Fig. 1D). Collectively, these observations indicated that PEITC treatment caused autophagy in human prostate cancer cells.

PEITC-mediated autophagy correlated with suppression of activating phosphorylations of Akt and mTOR. The protein kinase mTOR has emerged as a key negative regulator of autophagy (21–23, 28). The activity of mTOR is regulated by Akt-mediated phosphorylation of tuberin (30). To gain insight into the mechanism of PEITC-mediated autophagy in our model, initially we determined the effect of PEITC treatment on activating phosphorylations of Akt (S473) and mTOR (S2448) using LNCaP and PC-3 cells (Fig. 2A). We found that PEITC treatment decreased S473 phosphorylation of Akt in both LNCaP and PC-3 cells, albeit more efficiently in the PC-3 cell line (Fig. 2A). The level of total Akt protein was only modestly decreased (20–30% decrease compared with control) on treatment of LNCaP and PC-3 cells with PEITC. The levels of S2448-phosphorylated mTOR and phosphorylation of mTOR downstream target p70s6k (T389) were also decreased on treatment of LNCaP and PC-3 cells with PEITC (Fig. 2A). The PEITC treatment transiently decreased the level of mTOR protein in the LNCaP but not in the PC-3 cell line.

To determine functional significance of Akt suppression in autophagic response to PEITC, we determined the effect of ectopic expression of constitutively active Akt (CA-Akt) on PEITC-mediated cleavage and recruitment of LC3 using PC-3 and LNCaP cells. Transient transfection of PC-3 cells with 3 μg CA-Akt plasmid DNA resulted in ∼1.8-fold increase in the levels of S473-phosphorylated Akt relative to empty vector–transfected control cells (Fig. 2B). The PEITC-mediated cleavage of LC3 was ∼1.7-fold higher in the empty vector–transfected control cells compared with CA-Akt–overexpressing cells (Fig. 2B). Overexpression of CA-Akt also conferred partial protection against PEITC-mediated increase in fraction of cells with punctate LC3 over DMSO-treated control in both PC-3 and LNCaP cell lines (Fig. 2C and D).

PEITC-induced autophagy was marginally affected by overexpression of mTOR regulator Rheb. Because CA-Akt overexpression resulted in only partial protection against PEITC-induced autophagy, we carried out additional experiments involving ectopic expression of Rheb to further define the role of mTOR in autophagic response to PEITC. The Rheb activates mTOR by antagonizing its endogenous inhibitor FKBP38 (31). Ectopic expression of Rheb in PC-3 cells resulted in activation of mTOR as evidenced by ∼3.1-fold increase in T389 phosphorylation of its downstream target p70s6k (Supplementary Fig. S1). The Rheb-mediated activation of mTOR conferred marginal protection against PEITC-mediated cleavage of LC3 (Supplementary Fig. S1). These results indicate that suppression of Akt/mTOR cannot fully explain PEITC-mediated autophagy.

Atg5 knockdown conferred significant protection against PEITC-mediated autophagy. Atg5 is required for formation of autophagosomes and Atg5-deficient mouse embryonic stem cells display significantly diminished number of autophagic vesicles (32). Treatment of control nonspecific siRNA-transfected PC-3 cells with 5 μmol/L PEITC for 9 hours resulted in modest induction of Atg5-12 complex (Fig. 3A). Atg5-12 was not detectable in PC-3 cells transiently transfected with a pool of three Atg5-targeted siRNAs, confirming knockdown of this protein (Fig. 3A). The cleavage of LC3 resulting from a 9-hour exposure to 5 μmol/L PEITC was suppressed by ∼50% in PC-3 cells with knockdown of Atg5 protein (Fig. 3A). The level of Atg5 was reduced by ∼80% in LNCaP cells transiently transfected with the Atg5-targeted siRNA compared with nonspecific siRNA-transfected LNCaP cells (Fig. 3B). The PEITC-mediated cleavage of LC3 was ∼2.2-fold higher in nonspecific siRNA-transfected LNCaP cells than in the LNCaP cells transfected with the Atg5-specific siRNA (Fig. 3B). In addition, PEITC treatment (5 μmol/L, 9 hours) caused a significant increase in fraction of cells with punctate LC3 over DMSO-treated control in control siRNA-transfected cells, which was not observed in cells transfected with the Atg5-targeted siRNA (Fig. 3C). Together, these results indicated that the PEITC-induced autophagy was regulated by Atg5.

Autophagy was not a protective mechanism against PEITC-induced apoptosis. Autophagy represents a protective mechanism against apoptotic cell death under starvation as well as contributes to resistance against therapy-induced apoptosis in cancer cells (21, 22, 33–35). At the same time, autophagy has been suggested to be a form of cell death, referred to as type II cell death, distinct from apoptosis (type I cell death) for various antineoplastic agents (36, 37). We proceeded to determine the role of autophagy in growth suppression and apoptosis induction by PEITC using autophagy inhibitor 3-MA. The cytoplasmic histone-associated apoptotic DNA fragmentation, which is a widely accepted technique for quantitation of apoptotic cell death, resulting from exposure of LNCaP (Fig. 4A) and PC-3 (results not shown) to 5 μmol/L PEITC for 16 hours was partially but statistically significantly attenuated in the presence of 3-MA. Consistent with these results, the 3-MA treatment conferred partial yet marked protection against PEITC-mediated cleavage of procaspase-3 in LNCaP (Fig. 4B) and PC-3 cells (results not shown). Together, these results indicated that PEITC can trigger both apoptosis (type I cell death) and autophagy (type II cell death) in prostate cancer cells.

To further define the relationship between autophagy and apoptosis in our model, we determined the effect of Atg5 knockdown on PEITC-mediated apoptosis. Consistent with the results using 3-MA (Fig. 4A and B), the cytoplasmic histone-associated DNA fragmentation (Fig. 4C) and cleavage of procaspase-3 (Fig. 4D) resulting from PEITC exposure (5 μmol/L, 9 hours) was significantly attenuated by knockdown of Atg5 protein level in both PC-3 and LNCaP cells.

Oral administration of PEITC inhibited PC-3 xenograft growth in association with increased apoptosis and LC3 cleavage in the tumor. We designed animal experiments to test whether PEITC treatment caused autophagy in vivo. The average tumor volume on the day of sacrifice (38 days after tumor cell implantation) in mice gavaged with 9 μmol PEITC was lower by ∼48% compared with the vehicle-treated control mice (Fig. 5A). The body weights of the control and PEITC-treated mice did not differ significantly throughout the experimental protocol (results not shown). The H&E staining revealed a marked decrease in nuclear to cytoplasmic ratio in the tumors from PEITC-treated mice relative to control tumors, indicating suppression of cell proliferation by PEITC administration (Fig. 5A,, right). The fraction of brown-color TUNEL-positive apoptotic bodies was statistically significantly higher in tumor sections from PEITC-treated mice compared with control tumors (Fig. 5B). Moreover, the tumors from PEITC-treated mice exhibited increased expression (Fig. 5C) as well as cleavage of LC3 (Fig. 5D). For example, cleaved LC3-II protein was detectable in each tumor sample from the PEITC treatment group but not in each tumor from control mice (Fig. 5D). These results indicated that the PEITC-mediated suppression of PC-3 xenograft growth in vivo was accompanied by apoptosis as well as autophagy.

PEITC treatment caused autophagy in C. elegans. To test whether PEITC treatment caused autophagy in an intact organism, we treated the adult nematode C. elegans with PEITC for 24 hours and monitored autophagy by following the cellular localization of the lgg-1:GFP fusion protein expressed by the DA2123 strain (38). The lgg-1 is the worm ortholog of yeast Apg8p and mammalian LC3. The lgg-1:GFP fusion protein has been shown to undergo redistribution from diffuse cytoplasmic localization to form punctate foci representing preautophagosomal and autophagosomal structures (39). Previous work has shown redistribution of lgg-1:GFP in response to starvation, caloric restriction, and reductions in daf-2 insulin/insulin-like growth factor-I receptor signaling (38–40). As can be seen in Fig. 6, treatment of DA2123 with either 10 or 25 μmol/L PEITC resulted in the formation of punctate lgg-1:GFP foci in the intestine of worm in contrast to the diffuse GFP expression seen in control worms. Results in C. elegans provided added in vivo evidence for autophagy induction by PEITC in an intact organism.

The present study shows that dietary cancer chemopreventive agent PEITC causes autophagy in cultured human prostate cancer cells. Furthermore, oral administration of PEITC inhibits growth of PC-3 xenograft in vivo in association with induction and cleavage of LC3. The autophagic response to PEITC is rapid and evident at low micromolar concentrations. The PEITC concentrations required to elicit autophagic response in cultured prostate cancer cells (2.5–5 μmol/L) and in PC-3 xenograft in vivo (9 μmol/d PEITC) are well within the range that can be generated through dietary intake of cruciferous vegetables (41, 42). We also found that a normal prostate epithelial cell line (PrEC) is markedly more resistant to PEITC-mediated cleavage and recruitment of LC3. Recent studies have pointed toward an important role of p53 tumor suppressor in regulation of autophagy (43). Our data suggest that p53 is not necessary for PEITC-induced autophagy because this response is observed in both wild-type p53-expressing LNCaP cell line and p53-null PC-3 cells.

Molecular mechanism for autophagy induction in mammalian cells is still not fully understood but at least 30 autophagy-related genes have been identified in yeast (44). The mTOR has emerged as a key negative regulator of autophagy in yeast and possibly in mammalian cells (21–23, 45). Results of the present study indicate that although PEITC treatment suppresses Akt-mTOR signaling axis, this mechanism cannot fully explain PEITC-mediated autophagic response.

The connection between apoptotic (type I cell death) and autophagic cell death (type II cell death) in the context of cancer-relevant stimuli is still unresolved, but autophagy seems to protect against apoptosis in some models. For example, inhibition of autophagy by chloroquine has been shown to increase activity of the alkylating agent cyclophosphamide in a Myc-driven model of lymphoma (34). Furthermore, 3-MA and chloroquine have been shown to synergistically augment the proapoptotic response to a histone deacetylase inhibitor (46). The results of the present study clearly indicate that autophagy is not a protective mechanism against PEITC-induced apoptosis at least in LNCaP and PC-3 cells and that PEITC-mediated autophagy and apoptosis both contribute to PEITC-mediated suppression of prostate cancer cell growth.

The process of formation of autophagosomes is regulated by multiple proteins, including Atg5. Previous studies have shown that enforced expression of Atg5 not only promotes autophagy but also enhances susceptibility toward apoptotic stimuli irrespective of the cell type (47). Apoptosis was associated with calpain-mediated cleavage of Atg5, resulting in translocation of truncated Atg5 to mitochondria for its association with Bcl-xL (47). We found that Atg5 protein status significantly affects autophagic response to PEITC. Atg5 knockdown confers marked protection against PEITC-mediated cleavage of LC3 in both PC-3 and LNCaP cell lines. PEITC-mediated apoptotic DNA fragmentation in PC-3 and LNCaP cells is also significantly suppressed by knockdown of Atg5 protein level. These results not only show a critical role of Atg5 protein in regulation of PEITC-mediated autophagy in prostate cancer cells but also provide additional evidence to indicate that autophagy and apoptosis are interrelated in cancer cell growth suppression by PEITC. The mechanism by which Atg5 connects autophagic and apoptotic responses to PEITC is not yet clear.

In conclusion, the present study indicates that PEITC treatment selectively triggers autophagy in prostate cancer cells and that Atg5 is critically involved in regulation of PEITC-mediated autophagic and apoptotic cell death.

No potential conflicts of interest were disclosed.

Grant support: USPHS grants CA101753 and CA115498 (S.V. Singh) awarded by the National Cancer Institute and the grant K08 AG028977 (A.L. Fisher) awarded by the National Institute of Aging.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.