Novel Cell Death Pathways Induced by N-(4-Hydroxyphenyl)Retinamide: Therapeutic Implications

286

Novel cell death pathways induced by N-(4-hydroxyphenyl)retinamide: therapeutic implications

1 1 Roberta Vene`, Giuseppe Arena, membrane potential (#ym), and ATP depletion. We found Alessandro Poggi,2 Cristina D’Arrigo,3 that 4HPR causes lysosomal membrane permeabilization Michele Mormino,3 Douglas M. Noonan,4 and cytosolic relocation of cathepsin D. Pepstatin A Adriana Albini,5 and Francesca Tosetti1 partially rescued cell viability and reduced DNA fragmentation and cytosolic cytochrome c. The antioxidant N- 1Department of Translational Oncology, Molecular Oncology acetylcysteine attenuated cathepsin D relocation into the Laboratory, and 2Immunology Laboratory, National Cancer cytosol, suggesting that lysosomal destabilization is 3 Research Institute (IST); Institute for Macromolecular Studies dependent on elevation of reactive oxygen species and (ISMAC), National Research Council (CNR), Section of Genova, Genoa, Italy; 4Universita`degli Studi dell’Insubria, Varese, Italy; precedes mitochondrial dysfunction. Activation of AKT, and 5IRCCS MultiMedica, Milan, Italy which regulates energy level in the cell, by the retinal survival factor insulin-like growth factor I was impaired and insulin-like growth factor I was ineffective against Abstract ATP and #ym loss in the presence of 4HPR. Lysosomal We previously reported that N-(4-hydroxyphenyl)retina- destabilization, associated with mitochondrial dysfunc- mide (4HPR) inhibits retinoblastoma tumor growth in a tion, was induced by 4HPR also in other cancer cell lines, murine model in vivo and kills Y79 retinoblastoma cells including PC3 prostate adenocarcinoma and the vascular in vitro. In this work, we assayed different cell death– tumor Kaposi sarcoma KS-Imm cells. The novel finding of related parameters, including mitochondrial damage and a lysosome-mediated cell death pathway activated by caspase activation, in Y79 cells exposed to 4HPR. 4HPR 4HPR could have implications at clinical level for the induced cytochrome c release from mitochondria, cas- development of combination chemoprevention and thera- pase-3 activation, and oligonucleosomal DNA fragmenta- py of cancer. [Mol Cancer Ther 2007;6(1):286–98] tion. However, pharmacologic inactivation of caspases by the pan-caspase inhibitor BOC-D-fmk, or specific caspase- Introduction 3 inhibition by Z-DEVD-fmk, was not sufficient to prevent The cancer chemopreventive and therapeutic retinoid N- cell death, as assessed by loss of 3-(4,5-dimethylthiazol-2- (4-hydroxyphenyl)retinamide (4HPR; ref. 1) has extensively yl)-2,5-diphenyltetrazolium bromide reduction, lactate been studied in preclinical models and clinical trials of dehydrogenase release, disruption of mitochondrial trans- cancer prevention due to its favorable toxicity profile in comparison with natural retinoids (2). 4HPR induces cell death in a number of tumor cell types, and different molecular mediators have been identified, suggesting the activation of multiple mechanisms depending on dose and Received 6/12/06; revised 10/4/06; accepted 11/16/06. cell type. In several cellular systems, exposure to 4HPR Grant support: Associazione Italiana per la Ricerca sul Cancro, the Istituto Superiore di Sanita`-Italy USA, Ministero della Sanita`Progetto Finalizzato, results in reactive oxygen species (ROS)–mediated cell the Ministero dell’Istruzione, dell’Universita`e della Ricerca, Progetto death (3, 4). However, ROS-independent mechanisms have Finalizzato and Fondo per gli Investimenti della Ricerca di Base, and the been also reported (5). 4HPR can induce mitochondrial Compagnia di San Paolo. dysfunction and cytochrome c release, leading to caspase- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked dependent apoptosis (6, 7). Other mechanisms include the advertisement in accordance with 18 U.S.C. Section 1734 solely to implication of ceramide/sphingomyelin pathways (8–11), indicate this fact. 12-LOX (8), Bax and/or Bak (7), c-jun NH2-terminal kinase Note: Previous address for A. Albini: Molecular Oncology Laboratory, (12), c-fos (13), GADD153 (14), and caspase-8 (15). Istituto Nazionale per la Ricerca sul Cancro (IST), c/o Centro di Biotecnologie Avanzate, Largo Rosanna Benzi 10, 16132 Genoa, Italy. The identity of the retinoblastoma cell of origin and the The monoclonal antibody H4A3 against human Lamp-1, developed definite molecular events leading to retinal cell transforma- by J.T. August and J.E.K. Hildreth, was obtained from the Developmental tion in humans are still debated (16). Defective cell death– Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Resources and maintained by the regulating pathways in multipotential stem cells of the Department of Biological Sciences, University of Iowa, Iowa City, developing retina lacking the tumor suppressor pRB-1 have IA 52242. been indicated as a cause of retinoblastoma insurgence in Requests for reprints: Adriana Albini, IRCCS MultiMedica, Polo Scientifico humans (17). Current treatment of intraocular retinoblasto- e Tecnologico, Settore Ricerca Oncologica, Via Fantoli 15/16, Milan, Italy. Phone: 02-554061; Fax: 11-39-010573-77231. ma includes a three-drug regimen of high-dose chemother- E-mail: [email protected] apy, in addition to radiotherapy and intensive focal therapy. Copyright C 2007 American Association for Cancer Research. Because of concerns with the systemic toxicity of conven- doi:10.1158/1535-7163.MCT-06-0346 tional chemotherapy agents and the related risk of

MolCancerTher2007;6(1).January2007

developing extraocular tumors, the identification of novel concentrations indicated in figure legends. The following molecular targets for the development of more effective and reagents were used at the indicated final concentration: less toxic therapeutic interventions is of great importance to BOC-D-fmk at 10 to 40 Amol/L (stock solution 4 mmol/L in control retinoblastoma progression. Recently, 4HPR has DMSO); Z-DEVD-fmk and Z-LEHD-fmk at 1 to 50 Amol/L been shown to induce cell death in brain tumor cells, (stock solution 5 mmol/L in DMSO); bafilomycin A1 at including glioblastoma and primary meningioma cells (18), 0.25 to 0.5 Amol/L (stock solution 10 mmol/L in DMSO); and has been tested as a chemotherapy drug in gliomas (19). pepstatin A at 200 Amol/L (stock solution 25 mmol/L in We previously reported that inhibition of retinoblastoma DMSO); DL-buthionine-[S,R]-sulfoximine at 150 Amol/L tumor growth by 4HPR in a murine model in vivo is (stock solution 100 mmol/L in water); N-acetylcysteine at associated with an increase in the intracellular levels of 10 mmol/L (stock solution 1 mol/L in culture medium, pH ROS leading to a necrosis-like form of death (20). 7.4); and desferrioxamine mesylate at 200 Amol/L (stock In this work, we show that 4HPR activates caspase- solution 50 mmol/L in water). dependent and caspase-independent cell death pathways CellViability and CytotoxicityAssays through lysosomal membrane permeabilization. We ob- Y79 cell viability was determined by the 3-(4,5-dime- served induction of lysosomal damage also in other cancer thylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) cell lines, including PC3 prostate adenocarcinoma and assay as previously described (20). Cytotoxicity was Kaposi sarcoma KS-Imm cells. Recently, the activation of measured as cell membrane lysis and release of lactate mechanisms of cell death alternative to classic apoptosis by dehydrogenase (LDH) into the culture medium. Y79 cells in cancer therapeutics has been highlighted as a promising suspension were seeded at 1 Â 104 to 1.5 Â 104 per well in tool to overcome drug resistance following conventional poly-D-lysine–coated 96-well microtiter dishes in 100 ALof chemotherapy (21). Because 4HPR as a single agent gave complete medium. The medium was replaced after encouraging results in children affected by neuroblastoma, 24 h with 100 AL of defined N1 medium per well. 4HPR a pediatric tumor of neuroectodermal origin (22), the data dissolved in ethanol (0.1% final ethanol concentration) was shown here could represent an experimental background to added to the cells and incubated for 4 or 24 h at 37jC. The plan novel drug combination strategies for the treatment of LDH assay was carried out with the colorimetric CytoTox pediatric neurogenic tumors. 96 assay kit (Promega, Milan, Italy) following the instructions of the manufacturers. The data obtained were expressed as percent LDH release relative to total LDH in Materials and Methods culture. Absorbance at 570 nm (MTT) or 490 nm (LDH) was Chemicals determined with an automatic microtiter plate reader 4HPR was kindly provided by Dr. James A. Crowell (Molecular Devices Corp., Sunnyvale, CA). (Division of Cancer Prevention, National Cancer Institute, Detection of Apoptosis and Necrosis by Flow Cyto- Bethesda) and Dr. Gregg Bullard (McKessonBio, Rockville, metry, Histone-Associated DNA Fragments, and Cyto- MD). The following reagents were used: the pan-caspase chrome c Release inhibitor Boc-Asp(OMe)-CH2F (BOC-D-fmk), the caspase-3 Fluorescence-activated cell sorting (FACS) analysis was inhibitor Z-Asp(OMe)-Glu(OMe)-Val-Asp(OMe)-CH2F done using an Annexin V-FITC/propidium iodide apopto- (Z-DEVD-fmk), and the caspase-9 inhibitor Z-Leu- sis detection kit (Oncogene Research, Boston, MA) based on Â 5 Glu(OMe)-His-Asp(OMe)-CH2F (Z-LEHD-fmk) were from the method described by Vermes et al. (25), with 5 10 Calbiochem (San Diego, CA); bafilomycin A1, the cathepsin cells from adherent cultures of Y79 cells, as previously D inhibitor pepstatin A, DL-buthionine-[S,R]-sulfoximine, described (20). Analysis was done on 10,000 gated cells N-acetyl-L-cysteine, and desferrioxamine mesylate were using a Coulter Epics XL FACS with excitation set at 488 from Sigma-Aldrich (Milan, Italy). nm and emission at 518 nm (FITC detector) or 620 nm Cell Cultures and Treatments (phycoerythrin fluorescence detector). Human retinoblastoma Y79 cells (ATCC HTB-18) were Apoptotic cell death was determined by an enzyme- propagated in suspension in RPMI 1640 supplemented linked immunoassay kit (Cell Death Detection ELISAPLUS, with 15% heat-inactivated fetal bovine serum, 2 mmol/L Roche, Mannheim, Germany) to detect fragmented DNA glutamine, and 100 mmol/L penicillin/streptomycin (com- and histones (mononucleosomes and oligonucleosomes). 5 plete medium). For adherent cultures (23), cells were Cell lysates, prepared from 5 Â 10 cells seeded on poly-D- seeded on plastic dishes coated with poly-D-lysine (Sigma) lysine–coated 24-well plates and treated with 4HPR in at 5 Ag/cm2 for 24 h, then the medium was replaced by a combination with different inhibitors for 24 h, were chemically defined neuronal medium containing the N1 processed following the instructions of the manufacturers. supplement (transferrin 62.5 nmol/L, selenium 30 nmol/L, Cytochrome c release in the cytosolic fractions isolated insulin 5 Ag/mL; Life Technologies, San Giuliano Milanese, from 5 Â 105 cells was measured with a commercial ELISA Italy; ref. 24). Prostate adenocarcinoma PC3 and DU145 kit (Oncogene Research) following the instructions of the cells and Kaposi sarcoma KS-Imm cells were grown in manufacturers and expressed as nanograms per milliliter. complete RPMI 1640 supplemented with 10% fetal bovine WesternBlotAnalysisofProteins serum. Cells were treated with 4HPR dissolved in ethanol Total cell lysates were prepared by syringe homogeni- in a 10 mmol/L stock solution and used at the final zation of the cells in lysis buffer containing Tris-HCl

Mol Cancer Ther 2007;6(1). January 2007

20 mmol/L (pH 7.5), NaCl 150 mmol/L, EDTA 1 mmol/L, Analytical Sciences, Milan, Italy). The amount of lumines- EGTA 1 mmol/L, Triton X-100 1%, sodium pyrophosphate cence detected was expressed as relative light units. 2.5 mmol/L, glycerol 2-phosphate 1 mmol/L, sodium Because the luminescent assay for caspase-3 also measures orthovanadate 1 mmol/L, leupeptin 1 Ag/mL, phenyl- caspase-7 activity, we analyzed caspase-7 by Western methylsulfonyl fluoride 1 mmol/L. A mitochondrial/ blotting, with no detectable caspase-7 processing in the lysosomal–enriched fraction was isolated from 5 Â 105 presence of 4HPR (data not shown). We then considered cells with a cell fractionation kit (MBL, Watertown, MA) the results obtained with the luminescent assay, or with following the instructions of the manufacturers. The Z-DEVD-fmk, as indicative of caspase-3 activity. All cytosolic fraction isolated after a 10,000 Â g centrifugation samples were assayed in triplicate in at least three was further centrifuged at 30,000 Â g and the supernatant independent experiments. was used for cytosolic cytochrome c and cathepsin D Fluorescence and Confocal Microscopy immunodetection. Protein quantitation in cell lysates and Cells seeded on poly-D-lysine–coated multiwell chamber cytosolic and membrane fractions was done with the DC slides (LabTek, Nunc, Naperville, IL) were cultured for Protein Assay Kit (Bio-Rad, Richmond, CA). Forty micro- 24 h at 37jC, pretreated for 2 h with different inhibitors, grams of total cellular proteins, diluted in reducing SDS and incubated with 4HPR for 1 or 3 h. At the end of the sample buffer, were resolved on 12.5% SDS polyacrylamide incubation, lysosomes were stained for 30 min at 37jC with gels and transferred on nitrocellulose membranes by the acidotropic fluorescent probe LysoTracker Red and Western blotting. The membranes were then incubated mitochondria were stained with MitoTracker Green with antibodies at the appropriate dilutions in 5% pow- (Molecular Probes) diluted in culture medium at 200 dered skim milk dissolved in 25 mmol/L TBS containing nmol/L. Cells were then counterstained with 1 Ag/mL 0.15 mol/L NaCl, 0.1% Tween 20 (BLOTTO), except if 4¶,6-diamidino-2-phenylindole (Sigma-Aldrich) in PBS and stated otherwise. The following antihuman antibodies were analyzed on an IX81 microscopy with the confocal used at the indicated dilution: mouse monoclonal anti– equipment FV500 (Olympus BioSystem). The objectives cathepsin D (Ab-1; Calbiochem), 1:200 diluted in 10% horse used and magnifications were 100Â oil 1.35 numerical serum, TBS-Tween 0.1%; rabbit polyclonal anti–actin aperture, 60Â oil 1.40 numerical aperture, and 40Â oil 1.00 (Sigma), 1:200 in BLOTTO; monoclonal anti–caspase-3 numerical aperture. For the immunofluorescence detection (3G2), polyclonal anti–cleaved poly(ADP-ribose) polymer- of lysosomal cathepsin D, cells previously fixed and ase (Asp214), rabbit polyclonal anti–phospho-AKT-1 permeabilized with cold methanol for 4 min at À20jC (Ser473), AKT, mammalian target of rapamycin (mTOR), and blocked in PBS-10% horse serum for 10 min were phospho-mTOR (Ser2448; Cell Signaling Technology, Bev- incubated for 1 h with anti–cathepsin D monoclonal erly, MA), 1:1,000 in 5% bovine serum albumin, TBS-Tween antibodies (clone Ab-1; Calbiochem) used at 1:100 dilution 0.1%; mouse monoclonal anti–cytochrome c oxidase IV in PBS-1% horse serum. Cells were washed thrice with PBS, (Molecular Probes, Europe BV) and anti–Lamp-1 H4A3 incubated with fluoresceinated secondary antimouse anti- monoclonal antibodies (Developmental Studies Hybridoma bodies (Amersham Pharmacia Biotech) at 1:200 dilution in Bank, University of Iowa, Iowa City, IA), 1:1,000 in PBS-1% horse serum for 30 min, and counterstained with BLOTTO. 4¶,6-diamidino-2-phenylindole (1 Ag/mL) for 5 min. The A mouse monoclonal anti–cytochrome c antibody slides were mounted with ProLong antifade reagent (1:1,000) was provided with the cell fractionation kit for (Molecular Probes) and viewed under a Leica DM L the isolation of mitochondria (MBL) and used at 1:200 epifluorescence microscope at Â40 or Â100 magnification. dilution in BLOTTO. All the antibodies were reacted with Electron Microscopy the nitrocellulose membranes overnight at 4jC. Secondary To analyze the morphology of mitochondria and of the horseradish peroxidase–labeled antirabbit or antimouse intracellular acidic compartment, 106 cells were harvested antibodies (Amersham Pharmacia Biotech, Milan, Italy) and the cell pellet was washed with cacodylate buffer were used at 1:10,000 dilution in BLOTTO. The immuno- 0.1 mol/L (pH 7.5) for 5 min, centrifuged at 5,000 rpm for reaction was revealed by the enhanced chemiluminescence- 5 min, prefixed in Karnovsky’s solution (1% paraformalde- plus detection system (Amersham Pharmacia Biotech). hyde, 2% glutaraldehyde, 2 mmol/L calcium chloride, Caspase-3 ActivityAssay 0.1 mol/L cacodylate buffer, pH 7.5) for 2 h at 4jC, and Protease activity for caspase-3/caspase-7 was measured finally washed with cacodylate buffer. Postfixing was carried with the Caspase-Glo luminescent Assay (Promega) on 4 Â out in 1% osmium tetroxide in cacodylate buffer for 1 h. 104 cells seeded in 96-well white-walled clear-bottom lumin- After dehydration with 70% to 100% alcohol, the cells were ometer microplates. Briefly, Y79 cells were seeded in complete embedded in Poly/Bed 812 resin (Polysciences, Inc., medium and treated with 4HPR in defined medium for 4, 8, Warrington, PA) and polymerized. The ultrathin sections 16, and 24 h, with or without a 2-h preincubation with were stained with 5% uranyl acetate in 50% ethanol followed inhibitors. Ten micrograms of whole-cell lysates were by 0.4% aqueous lead citrate and examined in a Zeiss LEO incubated at room temperature in the dark for 1 h with the 900 electron microscope operating at 80 kV. luminogenic substrate Z-DEVD-aminoluciferin. Following Measurement of Volume of Acidic Compartment incubation, luminescence was measured using a Wallac Triplicate samples of cells (105) were incubated for 1 and 1450 MicroBeta Trilux luminometer (Perkin-Elmer Life and 3 h with 2.5 Amol/L 4HPR or vehicle and loaded with

MolCancerTher2007;6(1).January2007

LysoTracker Red at 200 nmol/L for 20 min. Cells were then measured MTT reduction and cell membrane damage as washed and suspended in phenol red–free RPMI 1640 measured by LDH release. BOC-D-fmk at 10 Amol/L was supplemented with L-glutamine. Healthy cells were gated unable to block the release of LDH (Fig. 1A) or to rescue the based on size and the fluorescence was measured by flow loss of cell viability (Fig. 1B) caused by 4HPR at the higher cytometry (FACSORT, Becton Dickinson, Buccinasco, doses (2.5–5 Amol/L). Milan, Italy) with CellQuest acquisition and analysis We then investigated whether 4HPR can induce oligo- software. Mean fluorescence intensity was used as the nucleosomal DNA fragmentation that, in the paradigmatic relative value of volume of the acidic compartment as model of apoptotic cell death, is mediated by effector described (26). caspase-3. General caspase inhibition by BOC-D-fmk ROS and Mitochondrial Transmembrane Potential De- significantly, but not completely, inhibited 4HPR-induced tection oligonucleosome release at 1 to 10 Amol/L (Fig. 1C). 5 Y79 cells (5 Â 10 )seededonpoly-D-lysine–coated 24- Similarly, specific inhibition of caspase-3 by Z-DEVD-fmk well plastic dishes were treated with 2.5 Amol/L 4HPR (1–10 Amol/L) only partially prevented DNA fragmenta- for the times indicated in figure legends. Pretreatment of tion (Fig. 1D). BOC-D-fmk at 40 Amol/L and Z-DEVD-fmk the cells with various inhibitors was carried out for at 50 Amol/L were more effective (Fig. 1C and D); however, 2 h before 4HPR addition. Twenty minutes before the at those higher concentrations, methyl ketone peptide end of the treatment, the cells were incubated with inhibitors could nonspecifically inhibit other proteases, A 50 mol/L dichlorofluorescein diacetate (H2DCFDA; including cysteine proteases of the cathepsin family and Molecular Probes) for detection of ROS, or with 200 calpain (27). nmol/L MitoTracker Red chloromethyl-X-rosamine (CMX Ros) or 200 nmol/L 5,5¶,6,6¶-tetrachloro-1,1¶,3,3¶- tetraethylbenzimidazol carbocyanine iodide (JC-1; Mole- cular Probes) to assess the mitochondrial transmembrane Dw potential ( m). The cells were then washed twice with phenol-free HBSS, resuspended in the same medium, andanalyzedwithaflowcytometer(FACScanor FACSORT, Becton Dickinson) or a microplate fluoro- meter (Gemini XPS, Molecular Devices) with excitation set at 488 nm and emission at 530 nm for H2DCFDA, or with excitation at 579 nm and emission at 599 nm for MitoTracker Red CMX Ros. The FACS analysis was done on 10,000 gated cells. ATP Determination ATP was determined luminometrically by the CellTiter- Glo Luminescent Cell Viability assay (Promega) on 4 Â 104 cells as described for the caspase activity assay. ATP concentration in control cells was calculated from a calibration curve and expressed as nanomoles per milli- gram of protein. Statistical Analysis Data are expressed as mean F SD. The statistical significance between two experimental groups was determined by two-tailed, unpaired Student’s t test using the PRISM software. One-way ANOVA followed by Tukey’s test was used in the analysis of three or more groups. P values of <0.05 and <0.005 were considered significant.

Results Cell Death Induced by 4HPRIs Partially Caspase Inde- pendent We previously reported that a mixed form of apoptotic Figure 1. General caspase inhibition by BOC-D-fmk does not prevent cell membrane damage or cell viability in Y79 cells treated with 4HPR for and necrotic cell death, characterized by phosphatidylser- 24 h. A, extracellular LDH release. B, MTT reduction. C, oligonucleosomal ine exposure and loss of cell membrane integrity, occurs DNA fragmentation in Y79 cells treated with 2.5 Amol/L 4HPR for 24 h is within 4 h on 4HPR administration in Y79 cells (20). To partially reduced by a 2-h preincubation with increasing concentrations of BOC investigate the contribution of caspases to 4HPR cytotox- BOC-D-fmk ( ). D, effects of caspase-3 inhibition by Z-DEVD-fmk (DEVD). Representative of three independent experiments run in sextupli- icity, we compared the effect of the irreversible pan-caspase cate. Columns, mean; bars, SD. *, P < 0.05; **, P < 0.005, versus inhibitor BOC-D-fmk on cell viability parameters. We samples treated with 4HPR alone.

Mol Cancer Ther 2007;6(1). January 2007

Figure 2. Caspase-3 is activated in response to 4HPR. A, time course of caspase-3 activation by 2.5 Amol/L 4HPR (HPR). B, preincubation with the caspase-3 inhibitor Z-DEVD-fmk at 10 and 50 Amol/L for 2 h completely abolishes caspase-3 activation by 2.5 Amol/L HPR at 16 h. C, time dependence of procaspase-3 processing in Y79 total cell lysates. Cells were treated with 2.5 Amol/L 4HPR. Appearance of the processed 89-kDa form of poly(ADP-ribose) polymerase (PARP) at the corresponding times is shown. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) immunostaining was done as a protein loading control. D, caspase-3 inhibition by Z-DEVD-fmk does not attenuate cell membrane damage induced by 2.5 Amol/L 4HPR after 24 h of treatment. E, MTT reduction in cells treated as in (D). Representative of three independent experiments run in sextuplicate. Columns, mean; bars, SD. F, cellular morphology of control (CTRL) Y79 cells incubated with vehicle alone or treated with 2.5 Amol/L 4HPR for 24 h (4HPR), and examined by phase-contrast light microscopy (magnification, Â40). Y79 cells grown in monolayer acquire a characteristic rod-like morphology and form ‘‘rosette’’ colonies. 4HPR- treated cells show apoptotic-like cell shrinkage and detachment from the surface, which is not prevented by pretreatment with broad-spectrum or caspase- 3 inhibitors.

These data indicated that DNA fragmentation induced We observed a remarkable difference in the modulation by 4HPR is partially caspase independent. We tried to of the cell viability parameters by using low(1 Amol/L) or directly inactivate caspase-3 with small interfering RNA in higher (2.5–5 Amol/L) doses of 4HPR, which correspond to Y79 cells using different standard transfection methods average peak plasma levels in patients in chemopreventive (electroporation, lipofection); however, none of the meth- (2) or therapeutic (22) clinical trials, respectively. The data ods used proved effective in our system, giving very low obtained with BOC-D-fmk, in particular, suggested the transfection efficiency (<5%). We then used different induction of a form of cell death progressively independent experimental approaches in parallel to assess the role of of caspase activation with higher doses of 4HPR (Fig. 1A caspase-3 in response to 4HPR. We verified caspase-3 and B). activation in Y79 cells treated with 2.5 Amol/L 4HPR Because the induction of nonapoptotic cell death by over time (Fig. 2). Maximal activity was detectable at 4HPR in retinoblastoma cells could potentially over- 24 h (Fig. 2A) and, as expected, it was totally abolished by come resistance to common chemotherapeutic agents, we 10 Amol/L Z-DEVD-fmk (Fig. 2B). Immunodetection by subsequently focused on the minimum dose capable of Western blotting of the 89-kDa fragment of poly(ADP- inducing apparently caspase-independent cell killing ribose) polymerase generated by caspase-3 cleavage and (2.5 Amol/L). the decrease of the full-length 35-kDa proform of caspase- Cathepsin D Is Involved in 4HPR-Induced Cell Death 3 confirmed the activation of the enzyme (Fig. 2C). Elevation of ROS has been shown to play a critical However, inhibition of caspase-3 activity by Z-DEVD- role in 4HPR-induced necrosis-like cell death in differ- fmk (1–50 Amol/L) had no effect on LDH release or loss ent tumor cell types (9), including Y79 cells (20). of MTT reduction induced by 4HPR (Fig. 2D and E). Because ROS-induced lysosomal rupture (28, 29) deter- Microscopic inspection confirmed apoptotic-like cell mines loss of plasma membrane integrity (26), which shrinkage and detachment from the surface caused by was a sign of cell death induced by 4HPR (Fig. 1), we 4HPR at 2.5 Amol/L, which were not prevented by considered the involvement of lysosomes in 4HPR pretreatment with BOC-D-fmk or Z-DEVD-fmk cytotoxicity. The cathepsin D inhibitor pepstatin A was (10 Amol/L; Fig. 2F). as effective as BOC-D-fmk in rescuing cell viability, as Taken together, these data indicate that caspase-3 activity measured by LDH release (Fig. 3A) and MTT reduction and DNA fragmentation do not significantly contribute to (Fig. 3B), and caused a significant decrease of 4HPR- 4HPR cytotoxicity in Y79 cells. induced DNA fragmentation (Fig. 3C). On the contrary,

MolCancerTher2007;6(1).January2007

inhibition of cysteine proteases by E64 (10 Amol/L) or probably as a consequence of cell membrane rupture and calpain by Z-LLL-fmk (10 Amol/L) was ineffective (data leakage into the extracellular medium (33). not shown). To obtain a more general protection of To define whether caspase activity affects lysosomal lysosomal structural integrity, we tried to overexpress permeabilization, we probed cathepsin D released into the heat shock protein 70, which prevents lysosomal cytosolic fractions of cells treated with caspase inhibitors. membrane permeabilization (30), in Y79 cells. Again, Specific caspase-9 and caspase-3 inhibition by Z-LEHD- the lowefficiency of transfection and failure of selection fmk or Z-DEVD-fmk (10 Amol/L) did not significantly due to the high mortality of transfected Y79 cells did not affect the amount of cytosolic cathepsin D (Fig. 3E), indi- allowus to use genetic methods to prevent lysosomal cating that caspase-9 and caspase-3 do not influence damage. We then took advantage of the iron chelator lysosomal destabilization, and this event probably precedes desferrioxamine, which, on internalization by endocyto- the activation of the mitochondrial apoptotic pathway. sis and selective accumulation into lysosomes, protects 4HPR Induces Lysosomal Destabilization in Retino- cells from cell death mediated by oxidative stress blastoma Cells (31, 32). Interestingly, desferrioxamine mesylate pro- To confirm the involvement of lysosomal damage in tected against 4HPR-induced cytolysis, loss of MTT 4HPR-induced cell death, we analyzed cytosolic cathepsin reduction, and DNA fragmentation more effectively D release in Y79 cells by immunofluorescence detection. than pepstatin A (Fig. 3A–C). Cathepsin D immunostaining revealed a granular pattern Cytosolic relocation of cathepsin D was detected into in control Y79 cells, consistent with a lysosomal distribu- the cytosol of 4HPR-treated cells by Western blot analysis tion (Fig. 4). Cathepsin D immunostaining after 1 or 3 h of at 3 h on 4HPR administration (Fig. 3D and E), confirming treatment with 4HPR showed a more diffuse distribution the involvement of lysosomal destabilization in this para- indicative of a cytosolic relocation of the enzyme. A more digm of cell death. diffuse staining of cathepsin D was also observed with Cytosolic cytochrome c, indicative of mitochondrial bafilomycin A1, a specific inhibitor of vacuolar Na+/H+ membrane permeabilization, was also detectable. Cytosolic ATPase, which can induce lysosomal destabilization, used cytochrome c and cathepsin D declined with time (Fig. 3D), as a control (Fig. 4).

Figure 3. Inhibition of cathepsin D and attenuation of lysosomal damage protect Y79 cells against 4HPR cytotoxicity. A, LDH release induced by 2.5 Amol/L 4HPR at 24 h was measured in cells pretreated with pepstatin A (pepA; 100 Amol/L) or desferrioxamine mesylate (DFX; 200 Amol/L) for 2 h before 4HPR addition. BOC-D-fmk (10 Amol/L) was included to compare the effect of caspase inhibition. B, MTT reduction was rescued close to control levels by desferrioxamine mesylate. C, DNA fragmentation was attenuated by pepstatin A and by desferrioxamine mesylate. Columns, means of three independent experiments run in sextuplicate; bars, SD. *, P < 0.05; **, P < 0.005. D, Western blot analysis of cytochrome c (cyt c) and cathepsin D (CathD) in the cytosolic fractions isolated from Y79 cells treated with 4HPR for 3 h. Immunostaining of the integral mitochondrial and lysosomal proteins cytochrome c oxidase IV (COXIV) and Lamp-1, respectively, was used to exclude membrane contamination in the cytosolic fractions. E, inhibition of caspase-9 by Z-LEHD-fmk (LEHD;10Amol/L) or caspase-3 by Z-DEVD-fmk (10 Amol/L) does not significantly affect cytosolic relocation of cathepsin D induced by 4HPR. Quantification by image analysis was carried out with the NIH Image 1.4 software; columns, average of three experiments; bars, SD.

Mol Cancer Ther 2007;6(1). January 2007

ment correlated with an increased volume of the acidic compartment, which reflects the size of lysosomes (26), as determined by LysoTracker Red uptake and quantitative evaluation by flowcytometry (Fig. 5B). Interestingly, increased volume of the acidic compartment has been indicated as a determinant of plasma membrane disruption in necrotic cell death (26). Transmission electron micrographs showed intense cytosolic vacuolization with the appearance of clusters of large electron lucent organelles into the cytosol of cells as early as 1 h after 4HPR addition, together with decreased nuclear volume, dilated cisternae of the nuclear envelope, and swollen mitochondria (Fig. 6). Taken together, these data suggest that lysosomal destabilization is an early event and accompanies mitochondrial damage in 4HPR-induced cell death. N-Acetylcysteine Reduces Cathepsin D and Cyto- chrome c Release into the Cytosol of Y79 Cells Because 4HPR is a pro-oxidant drug, the antioxidant N- acetylcysteine can reduce ROS generation and cytotoxicity induced by 4HPR (9, 20). Here we monitored ROS increase in 4HPR-treated cells with time, in the presence or absence of N-acetylcysteine (Fig. 7A). Elevation of ROS, as detected by the increase of H2DCFDA fluorescence, was an early event, doubling within 15 min of exposure to the drug (average of 2.1 F 0.7-fold increase over the control). The ROS content reached a maximum at f6 h of treatment (5.2 F 0.5-fold over the control) and declined at 16 h. We then verified whether 4HPR-induced lysosomal Figure 4. 4HPR causes cathepsin D relocation and lysosomal damage in 4HPR-treated cells. Immunofluorescence staining of Y79 cells with anti – permeabilization was ROS dependent. cathepsin D antibodies was done in Y79 cells treated with 2.5 Amol/L Western blot analysis showed that cytosolic cathepsin 4HPR for 1 or 3 h. Bafilomycin A1 (BAF A1), a specific inhibitor of the D released after 3 h of exposure to 4HPR is significantly vacuolar proton pump and inducer of lysosomal destabilization, was used as a positive control. Y79 cells were counterstained with 4¶,6-diamidino-2- reduced in cells pretreated with N-acetylcysteine for phenylindole (DAPI;1Ag/mL). The punctated green fluorescence is 2 h (Fig. 7B). N-Acetylcysteine also decreased cytochrome reduced by 4HPR after 1 h of treatment. c release from mitochondria (Fig. 7B and E) and oligonucleosomal DNA fragmentation (not shown). To probe lysosomal and mitochondrial integrity, we The attenuating effect of antioxidants on lysosomal stained Y79 cells with the fluorescent dyes LysoTracker perturbation was confirmed by immunofluorescence detec- Red, a lysosomotropic and acidotropic reagent, and tion of cathepsin D. N-Acetylcysteine pretreatment par- MitoTracker Green. Confocal microscopy analysis showed tially preserved the punctated lysosomal distribution of enlargement of the acidic compartment, as indicated by a cathepsin D, which was lost in 4HPR-treated cells (Fig. 7C). more intense staining with LysoTracker Red, and localiza- We then investigated whether lysosomal injury in tion of the fluorescent dye into large, clustered vesicles response to 4HPR could affect mitochondrial function detectable as early as 1 h after 4HPR administration and integrity. Protection of lysosomal integrity by desfer- (Fig. 5A). This was paralleled by decreased MitoTracker rioxamine mesylate significantly reduced ROS increase in Green staining and mitochondrial stretching, suggestive of cells at 15 min (not shown) and 2 h on 4HPR administration impaired mitochondrial function. A similar perturbation of (average inhibition at 2 h, 52.8 F 3.9%; Fig. 7D), indicating lysosomes and of the mitochondrial network by 4HPR was that lysosomal labilization caused by 4HPR can in turn detectable in other cancer cell lines including prostate affect mitochondrial ROS production. Pepstatin A, on the adenocarcinoma PC3 and Kaposi sarcoma KS-Imm cells contrary, did not influence the amount of intracellular ROS (see Supplementary Fig. S1A and B, respectively)6 and in (Fig. 7D). DU145 cells (data not shown), suggesting a more general A cytochrome c ELISA assay done on the cytosolic effect of 4HPR on lysosomal stability. Lysosomal enlarge- fractions isolated from Y79 cells pretreated with N- acetylcysteine (10 mmol/L) or pepstatin A (100 Amol/L) and then with 4HPR (2.5 Amol/L) for 3 h showed that both 6 Supplementary material for this article is available at Molecular Cancer agents can decrease the release of cytochrome c in 4HPR- Therapeutics Online (http://mct.aacrjournals.org/). treated cells (Fig. 7E).

MolCancerTher2007;6(1).January2007

Collectively, these data suggest that attenuation of the indicated by an increased fluorescence associated with pro-oxidant effects of 4HPR and ROS generation can CMX Ros retention, temporally related with initial ROS reduce mitochondrial and lysosomal damage induced by elevation, which has been reported to be an early and 4HPR, and that the release of lysosomal proteases can reversible step in the activation of cell death (38); Dw f F influence the integrity of mitochondria in Y79 cells. following this, m was reduced to 50% (50.2 4HPR Induces ATP Depletion and Irreversible Loss of 10.7%) of controls after 6 h and remained stable on these #ym levels through 24 h (Fig. 8B). A similar late decrease of Dw Energy level, as indicated by ATP concentration, can m resulted by measuring voltage-dependent J aggre- switch cell death from apoptosis to necrosis (34). Because gate (JC-1) fluorescence at 16 h after 4HPR administration 4HPR induced early signs of apoptosis, rapidly followed by (Fig. 8C). N-Acetylcysteine, which was able to prevent plasma membrane disruption, we measured intracellular mitochondrial permeabilization, cytochrome c release, ATP levels with time in cells treated with 4HPR. ATP levels and ROS elevation occurring within 3 h of exposure to were maintained for the first 4 h, declined starting at 8 h 4HPR (Fig. 7A, B, and E), was ineffective against the Dw (data not shown), and reached f50% (41.3 F 8.3%) of subsequent loss of m (Fig. 8D). Pretreatment of the control at 24 h (Fig. 8A; average ATP concentration in cells with caspase inhibitors, desferrioxamine mesylate, or control cells was 1.2 F 0.3 nmol/mg of protein). Desferriox- pepstatin A gave results similar to those observed for Dw amine mesylate per se caused a decline in ATP level in ATP depletion, as none of the drugs affected loss of m control cells. Interestingly, iron chelators, which are neuro- induced by 4HPR (data not shown). protective (35, 36), have been shown to influence energy 4HPR Compromises Insulin-like Growth Factor metabolism by decreasing ATP formation via oxidative I ^ Stimulated AKT Survival Pathway phosphorylation while increasing glucose consumption Growth factor stimulation of the serine/threonine kinase (37). However, N-acetylcysteine, pepstatin A, and desfer- AKT protects cells against cell death by maintaining ATP rioxamine mesylate were not able to significantly rescue levels (39) and preventing mitochondrial damage, includ- ATP loss caused by 4HPR (Fig. 8A). ing mitochondrial permeability transition (40). We then To correlate ATP depletion with bioenergetic parame- assessed whether activation of AKT by the insulin-like ters indicative of mitochondrial dysfunction, we mea- growth factor I (IGF-I), which is a major survival factor in Dw Dw sured the m. We stained the cells with the fluorescent the retina (41), was able to rescue ATP loss and m dye MitoTracker Red CMX Ros (200 nmol/L) and disruption induced by 4HPR. evaluated total fluorescence intensity by cytofluorimetric We incubated the cells with IGF-I (100 ng/mL) for 2 h and analysis. 4HPR at 2.5 Amol/L caused transient mitochon- then with 2.5 Amol/L 4HPR for 24 h. ATP levels in cells drial hyperpolarization within 15 min of treatment, as treated with 4HPR, as expected, were remarkably lower

Figure 5. 4HPR induces enlargement of lysosomes and increase of volume of the acidic compartment at early times of treatment. A, confocal microscopy analysis of lysosomal and mitochondrial morphology was done in Y79 cells treated for 1 or 3 h and stained with MitoTracker Green (MTG) and LysoTracker Red (LTR; 200 nmol/L). Intense staining with LysoTracker Red in cells treated with 4HPR was localized in large vesicles. B, volume of the acidic compartment (VAC) was measured by LysoTracker Red uptake followed by flow cytometric analysis. Columns, mean fluorescence intensity of the cells calculated from two independent experiments run in triplicate; bars, SD.

Mol Cancer Ther 2007;6(1). January 2007

Figure 6. 4HPR induces cytoplasmic vacuolization and mitochondrial and nuclear morphologic alterations related with cell death. Top, electron microphotographs showing the ultra- structure of Y79 cells treated with 4HPR (2.5 Amol/L, 3 h). The appearance of numerous electrolucent organelles in the cytoplasm, swollen mitochondria, and dilated cisternae of the nuclear envelope can be seen. Bottom, low- magnification images of cells treated with 4HPR showing nuclear volume decrease and intense cytoplasmic vacuolization. M, mitochondria; L, lysosomes; V, vacuoles.

than in controls (40.5 F 5.3%; Fig. 9A). ATP content in control The molecular mechanisms identified in this work cells treated with IGF-I alone was increased by f2-fold suggest that oxidative stress triggered by 4HPR operates (2.3 F 1.1). However, IGF-I was unable to prevent ATP in parallel on lysosomal perturbation and mitochondrial Dw depletion (Fig. 9A) or m loss (Fig. 9B) in cells treated with dysfunction, both leading to a form of necrotic cell death. In 4HPR. this model, early events occurring within 4 h on 4HPR We then verified whether activation of AKT by IGF-I was administration include ROS production, phosphatidylser- impaired in Y79 cells treated with 4HPR. IGF-I induced ine extrusion (20), mitochondrial outer membrane perme- rapid phosphorylation of AKT at Ser473 and of mTOR at abilization, release of cytochrome c into the cytosol, and cell Ser2448, which was sustained over time (Fig. 9C). Phos- membrane damage. Other events, including ATP and Dw phorylation of AKT was almost completely suppressed by disruption of m, indicative of irreversible mitochondrial 4HPR at 24 h. mTOR phosphorylation was decreased but dysfunction, occur later in the process (16–24 h on 4HPR was less affected. These data suggest that, in the presence addition). Several lines of evidence indicate that a caspase- of 4HPR, the AKT-mediated survival pathway is compro- independent pathway is activated in response to 4HPR in mised, and this could account for the lack of protection by Y79 cells. First, we observed that general caspase inactiva- IGF-I against 4HPR cytotoxicity. tion by BOC-D-fmk and specific caspase-3 inactivation by lowdoses of Z-DEVD-fmk had limited activity in oligonucleosomal DNA fragmentation. Most importantly, caspase Discussion inhibitors were ineffective in preventing loss of cell Due to the increasing clinical application of 4HPR as a viability caused by 4HPR. cancer chemopreventive and therapeutic agent, it is Based on early studies showing lysosomal labilization in important to understand in detail howthis retinoid response to oxidative stress (42), increasing evidence promotes cell death in tumor cells. The induction of a establishes lysosomal proteases, in particular cathepsin mixed form of apoptotic and necrotic cell death, as a D and B, as central regulators of programmed cell death consequence of oxidative stress, has previously been in response to oxidative stress (28, 29). Indeed, we detec- described in tumor cells derived by neurogenic tumors ted biochemical and morphologic signs of lysosomal (9, 20). destabilization, including cathepsin D translocation into

MolCancerTher2007;6(1).January2007

the cytosol, enlargement of lysosomes, and increased mediated events that render lysosomes particularly vul- volume of the acidic compartment, as early as 1 h on nerable to oxidative stress (42). Further, iron chelators have 4HPR administration. In addition, the cathepsin D been shown to exert antioxidant and neuroprotective inhibitor pepstatin A reduced the amount of cytosolic activity (36) through induction of the transcription factor cytochrome c and, accordingly, was as effective as BOC- hypoxia-inducible transcription factor 1a, which increases D-fmk in protecting against 4HPR-induced DNA frag- glucose uptake (36). mentation. Therefore, more than one mechanism might contribute to Because pepstatin A offered only partial protection desferrioxamine mesylate–induced preservation of cell against loss of cell viability, indicating the involvement of viability in 4HPR-treated Y79 cells. lysosomal proteases other than cathepsin D (43), we took We found that ROS elevation affects lysosomal stability advantage of the iron chelator desferrioxamine mesylate to and that, conversely, preservation of lysosomal integrity attain structural protection of lysosomes. Desferrioxamine reduces 4HPR-induced oxidative stress. ROS elevation, in mesylate, in fact, has been shown to specifically accumulate fact, was the earliest measurable event after 4HPR admi- in lysosomes by fluid phase endocytosis and to prevent nistration (doubling within 15 min); however, sustained lysosome-mediated cell death induced by oxidant chal- ROS production continued over time (maximal increase lenge (31, 32). Notably, desferrioxamine mesylate was the after 6 h) after cytochrome c release (detectable at 3 h). most effective protective agent against 4HPR cytotoxicity. N-Acetylcysteine not only decreased the amount of cyto- Desferrioxamine mesylate chelates lysosomal iron, which solic cytochrome c but also reduced cytosolic cathepsin D, represents the major intracellular iron pool, thus inhibiting suggesting that preservation of a reducing environment and iron-dependent Fenton-type reactions in hydrogen peroxide– inhibition of oxidative stress attenuate both mitochondrial

Figure 7. ROS generation by 4HPR detected by flow cytometric analysis of H2DCFDA fluorescence. A, pretreatment with N-acetylcysteine (NAC; 10 mmol/L) for 2 h decreases ROS production at early times. Cells treated for 6 h with DL-buthionine-[S,R]-sulfoximine (BSO; 100 Amol/L) and hydrogen peroxide (H2O2; 100 Amol/L) were used as positive controls. Columns, mean fluorescence intensity of the cells calculated from two independent experiments run in triplicate; bars, SD. B, N-acetylcysteine reduces cytosolic cathepsin D and cytochrome c in 4HPR-treated cells. A membrane fraction enriched in mitochondria and lysosomes (m/l) and a cytosolic fraction (cyt) were isolated from Y79 cells treated with 2.5 Amol/L 4HPR. Cytochrome c oxidase IV immunostaining was used as a loading control for the membrane fractions and to exclude membrane contamination in the cytosolic fractions. C, N-acetylcysteine partially preserves the punctate cathepsin D staining pattern in Y79 cells treated with 4HPR for 3 h. D, N-acetylcysteine and desferrioxamine mesylate (200 Amol/L), but not pepstatin A, attenuate ROS increase induced by 4HPR. Y79 cells were treated with 2.5 Amol/L 4HPR for 2 h and processed as described in (A). Fluorescence intensity was detected with a fluorescence microtiter plate reader with excitation set at 488 nm and emission set at 518 nm. E, pepstatin A and N-acetylcysteine reduce the release of cytochrome c into the cytosolic fractions of Y79 cells treated with 2.5 Amol/L 4HPR for 3 h, as determined by ELISA assay. Columns, mean of the data collected from two independent experiments run in triplicate; bars, SD.

Mol Cancer Ther 2007;6(1). January 2007

and lysosomal membrane permeabilization in 4HPR- release (47). However, blocking several early events, treated cells. Further, the results of the present study including ROS production, cytochrome c and cathepsin D indicate that lysosomal destabilization precedes evident release by N-acetylcysteine, or even preservation of cell mitochondrial dysfunction, in agreement with other mod- membrane integrity by desferrioxamine mesylate, could Dw els of lysosome-dependent cell death (44, 45). In fact, we not delay or prevent the later disruption of m and ATP observed cytosolic relocation of cathepsin D and morpho- depletion induced by 4HPR. logic signs of lysosomal destabilization in Y79 cells after Interestingly, our data showthat stimulation of the AKT Dw 1 h of 4HPR administration, whereas loss of m and ATP survival pathway by IGF-I is compromised in Y79 cells Dw depletion occurred at later times (6–16 h). exposed to 4HPR at a time coincident with m and ATP In an attempt to order the sequence of events activated by loss. 4HPR, we speculate that, because lysosomes are acidic Because AKT and its substrate mTOR are prominent organelles that are very sensitive to oxidative stress (46), regulators of energetic metabolism by modulating ATP initial elevation of ROS of mitochondrial origin affects levels in the cell (39), we speculate that 4HPR interferes lysosomal integrity before sustained mitochondrial damage with the molecular machinery in the control of protection occurs. Thus, ROS produced at mitochondrial level might from cell death through energy regulation in the cell. induce limited lysosomal damage, possibly by reacting In summary, the biochemical features of cell death with lysosomal iron, and leakage of proteases, which induced by 4HPR in our system depict a complex pathway generates a positive feedback loop enhancing subsequent distinct from classic caspase-dependent apoptosis. mitochondrial dysfunction, ROS production, and further The involvement of intracellular organelles other than lysosomal damage. This model is supported by our data mitochondria in retinoid-mediated cell death, at least in showing a reduction by 50% of ROS content in 4HPR- some cell types, has previously been reported. For example, treated cells where oxidative stress was attenuated by a lysosomal pathway has been implicated in cell death desferrioxamine mesylate. induced by the synthetic retinoid CD437 (48). In this work, Our data showthat early mitochondrial outer membrane we documented that the induction of lysosomal destabili- permeabilization and release of cytochrome c occur several zation by 4HPR occurs in several cancer cell lines, in hours before irreversible mitochondrial dysfunction, indi- addition to Y79 cells. These data emphasize the opportu- Dw cated by dissipation of m and ATP loss, which is in nity to explore alternative pathways of cell death in future agreement with previous studies showing that mitochon- studies on retinoids to optimize the therapeutic use of these drial functions can be maintained after cytochrome c compounds.

Figure 8. ATP depletion and disruption of Dw m, indicative of irreversible mitochondrial damage, are late events in 4HPR-treated cells. A, a late decline of ATP levels in Y79 cells treated with 2.5 Amol/L 4HPR for 16 h cannot be prevented by pretreatment with N-acetylcysteine (10 mmol/L), pepstatin A (100 Amol/L), or desferrioxamine mesylate (200 Amol/L) for 2 h. B, Y79 cells were stained with the Dw m sensitive probe MitoTracker CMX Ros (200 nmol/L) and analyzed by flow cytometry. Col- umns, mean of data collected from two independent experiments run in triplicate; bars, SD. C, top, bivariate cytofluorimetric analysis of Y79 cells treated with 2.5 Amol/L 4HPR for 16 h on addition of the potentiometric probe JC-1 (10 Ag/mL). The red/green fluorescence ratio of the reading at 590 nm (aggregate) to the reading at 530 nm (monomer) was calculated and expressed as percent relative to controls. Bot- tom, in situ evidence of Dw m disruption indicated by the spectral red-to-green shift of JC-1. Magnification, Â40. D, N -acetylcysteine (10 mmol/L) cannot prevent late mitochondrial membrane depolarization induced by 4HPR. Representative cytofluorimetric histograms showing the attenuating effect of N-acetylcysteine on Dw m low peak increase (plain numbers in quadrants), but not on loss of total cell- associated CMX Ros fluorescence, expressed as a percentage relative to controls (bold numbers in quadrants), in Y79 cells treated with 4HPR for 16 h.

MolCancerTher2007;6(1).January2007

Figure 9. Energy depletion is not prevented by IGF-I in cells treated with 4HPR. A, ATP levels were determined in cells treated with IGF-I (100 ng/mL) for 2 h and with 4HPR (2.5 Amol/L) for 24 h. B, Dw m disruption is also unaffected by IGF-I. C, IGF-I induces rapid phosphorylation of AKT at Ser473 and of mTOR at Ser2448, which is sustained over time. The 24-h time point from two independent samples was loaded. Phosphorylation of AKT is almost completely suppressed in the presence of 4HPR at 24 h. Poly(ADP-ribose) polymerase cleavage was followed to detect signs of cell death. Extracellular signal – regulated kinase 1/2 (ERK1/2) and glyceraldehyde-3-phosphate dehydrogenase immunostaining was done as a protein loading control.

Recently, the discovery of novel caspase-independent Acknowledgments forms of cell death as potential targets to overcome We thank David Schubert for critical review of the manuscript, Francesco impairment of apoptosis in cancer cells opened new Procopio and Stefania Martini for assistance with FACS analysis, and Dr. Anna Rapetti for expert secretarial assistance and preparation of the perspectives for the improvement of cancer treatment (21). manuscript. 4HPR gave significant results in experimental and clinical trials of cancer chemoprevention (2). Despite the limited References efficacy displayed in experimental and clinical contexts of 1. Swanson BN, Zaharevitz DW, Sporn MB. Pharmacokinetics of N-(4- late-stage disease, 4HPR as a single agent has progressed hydroxyphenyl)-all-trans-retinamide in rats. Drug Metab Dispos 1980;8: into clinical evaluation as a cancer therapy drug in 168 – 72. advanced tumors. The knowledge that different cell death 2. Veronesi U, Mariani L, Decensi A, et al. Fifteen-year results of a randomized phase III trial of fenretinide to prevent second breast cancer. paradigms can be triggered in a dose- and cell type–specific Ann Oncol 2006;4:4. manner could be of importance to reduce the risk of inef- 3. Suzuki S, Higuchi M, Proske RJ, et al. Implication of mitochondria- ficacy or resistance due to the administration of suboptimal derived reactive oxygen species, cytochrome c and caspase-3 in N- (hydroxyphenil)retinamide-induced apoptosis in cervical carcinoma cells. doses of the drug. The present study shows that 4HPR can Oncogene 1999;18:6380 – 7. activate caspase-independent pathways at concentrations 4. Sun SY, Li W, Yue P, et al. Mediation of N-(4-hydoxyphenyl)retina- (2.5–5 Amol/L) that are more elevated than chemopreven- mide-induced apoptosis in human cancer cells by different mechanisms. tive doses (1 Amol/L). These concentrations are, however, Cancer Res 1999;59:2493 – 8. clinically achievable, as shown by the good tolerability of 5. DiPietrantonio AM, Hsieh TC, Juan G, et al. Fenretinide-induced caspase 3 activity involves increased protein stability in a mechanism therapeutic doses of 4HPR administered to children affected distinct from reactive oxygen species elevation. Cancer Res 2000;60: by neuroblastoma in a phase I clinical trial (average peak 4331 – 5. plasma concentration of 12.9 Amol/L, comparable to 6. Hail NJ, Lotan R. Mitochondrial respiration is uniquely associated with cytotoxic doses in neuroblastoma cell lines; ref. 22). the prooxidant and apoptotic effects of N-(4-hydroxyphenyl)retinamide. J Biol Chem 2001;276:45614 – 21. The data shown can also contribute to the clinical 7. Boya P, Morales MC, Gonzalez-Polo RA, et al. The chemopreventive improvement of 4HPR in combination with anticancer agent N-(4-hydroxyphenyl)retinamide induces apoptosis through a mito- agents acting through nonoverlapping mechanisms (e.g., chondrial pathway regulated by proteins from the Bcl-2 family. Oncogene tyrosine kinase receptor inhibitors specific for tissue- 2003;22:6220 – 30. 8. Lovat PE, Di Sano F, Corazzari M, et al. Gangliosides link the acidic relevant growth factors that can protect cancer cells from sphingomyelinase-mediated induction of ceramide to 12-lipoxygenase- caspase-dependent apoptosis). The experimental evidence dependent apoptosis of neuroblastoma in response to fenretinide. J Natl of multiple programs integrating and converging to final Cancer Inst 2004;96:1288 – 99. cell demise (21) provides, in fact, the rationale for 9. Maurer BJ, Metelitsa LS, Seeger RC, Cabot MC, Reynolds CP. Increase of ceramide and induction of mixed apoptosis/necrosis by N-(4-hydrox- combination regimens at the clinical level to control cancer yphenyl)-retinamide in neuroblastoma cell lines. J Natl Cancer Inst 1999; development and progression (49). 91:1138 – 46.

Mol Cancer Ther 2007;6(1). January 2007

10. O’Donnell PH, Guo WX, Reynolds CP, Maurer BJ. N-(4-Hydroxyphe- protein 70 promotes cell survival by inhibiting lysosomal membrane nyl)retinamide increases ceramide and is cytotoxic to acute lymphoblastic permeabilization. J Exp Med 2004;200:425 – 35. leukemia cell lines, but not to non-malignant lymphocytes. Leukemia 31. Persson HL, Yu Z, Tirosh O, Eaton JW, Brunk UT. Prevention of 2002;16:902 – 10. oxidant-induced cell death by lysosomotropic iron chelators. Free Radic 11. Erdreich-Epstein A, Tran L, Bowman N, et al. Ceramide signaling in Biol Med 2003;34:1295 – 305. fenretinide-induced endothelial cell apoptosis. J Biol Chem 2002;277: 32. Doulias PT, Christoforidis S, Brunk UT, Galaris D. Endosomal and 49531 – 7. lysosomal effects of desferrioxamine: protection of HeLa cells from 12. Chen YR, Zhou G, Tan TH. c-Jun N-terminal kinase mediates hydrogen peroxide-induced DNA damage and induction of cell-cycle apoptotic signaling induced by N-(4-hydroxyphenyl)retinamide. Mol arrest. Free Radic Biol Med 2003;35:719 – 28. Pharmacol 1999;56:1271 – 9. 33. Jemmerson R, LaPlante B, Treeful A. Release of intact, monomeric 13. Appierto V, Villani MG, Cavadini E, et al. Involvement of c-Fos in cytochrome c from apoptotic and necrotic cells. Cell Death Differ 2002;9: fenretinide-induced apoptosis in human ovarian carcinoma cells. Cell 538 – 48. Death Differ 2004;11:270 – 9. 34. Nicotera P, Leist M, Manzo L. Neuronal cell death: a demise with 14. Lovat PE, Oliverio S, Corazzari M, et al. Induction of GADD153 and different shapes. Trends Pharmacol Sci 1999;20:46 – 51. Bak: novel molecular targets of fenretinide-induced apoptosis of neuro- 35. Zaman K, Ryu H, Hall D, et al. Protection from oxidative stress- blastoma. Cancer Lett 2003;197:157 – 63. induced apoptosis in cortical neuronal cultures by iron chelators is 15. Cuello M, Coats AO, Darko I, et al. N-(4-Hydroxyphenyl) retinamide associated with enhanced DNA binding of hypoxia-inducible factor-1 and (4HPR) enhances TRAIL-mediated apoptosis through enhancement of a ATF-1/CREB and increased expression of glycolytic enzymes, p21(waf1/ mitochondrial-dependent amplification loop in ovarian cancer cell lines. cip1), and erythropoietin. J Neurosci 1999;19:9821 – 30. Cell Death Differ 2004;11:527 – 41. 36. Soucek T, Cumming R, Dargusch R, Maher P, Schubert D. The 16. Dyer MA, Bremner R. The search for the retinoblastoma cell of origin. regulation of glucose metabolism by HIF-1 mediates a neuroprotective Nat Rev Cancer 2005;5:91 – 101. response to amyloid h peptide. Neuron 2003;39:43 – 56. 17. DiCiommo D, Gallie BL, Bremner R. Retinoblastoma: the disease, gene 37. Oexle H, Gnaiger E, Weiss G. Iron-dependent changes in cellular and protein provide critical leads to understand cancer. Semin Cancer Biol energy metabolism: influence on citric acid cycle and oxidative phosphor- 2000;10:255 – 69. ylation. Biochim Biophys Acta 1999;1413:99 – 107. 18. Puduvalli VK, Li JT, Chen L, McCutcheon IE. Induction of apoptosis 38. Perl A, Gergely P, Jr., Nagy G, Koncz A, Banki K. Mitochondrial in primary meningioma cultures by fenretinide. Cancer Res 2005;65: hyperpolarization: a checkpoint of T-cell life, death and autoimmunity. 1547 – 53. Trends Immunol 2004;25:360 – 7. 19. Puduvalli VK, Yung WK, Hess KR, et al. Phase II study of fenretinide 39. Hahn-Windgassen A, Nogueira V, Chen CC, et al. Akt activates the (NSC 374551) in adults with recurrent malignant gliomas: A North mammalian target of rapamycin by regulating cellular ATP level and AMPK American Brain Tumor Consortium study. J Clin Oncol 2004;22:4282 – 9. activity. J Biol Chem 2005;280:32081 – 9. Epub 2005 Jul 15. 20. Tosetti F, Vene R, Arena G, et al. N-(4-Hydroxyphenyl)retinamide 40. Miyamoto S, Howes AL, Adams JW, Dorn GW II, Brown JH. Ca2+ inhibits retinoblastoma growth through reactive oxygen species-mediated dysregulation induces mitochondrial depolarization and apoptosis: role of cell death. Mol Pharmacol 2003;63:565 – 73. Na+/Ca2+ exchanger and AKT. J Biol Chem 2005;280:38505 – 12. 21. Mathiasen IS, Jaattela M. Triggering caspase-independent cell death Epub 2005 Aug 1. to combat cancer. Trends Mol Med 2002;8:212 – 20. 41. Seigel GM, Chiu L, Paxhia A. Inhibition of neuroretinal cell death by 22. Garaventa A, Luksch R, Lo Piccolo MS, et al. Phase I trial and insulin-like growth factor-1 and its analogs. Mol Vis 2000;6:157 – 63. pharmacokinetics of fenretinide in children with neuroblastoma. Clin 42. Zdolsek J, Zhang H, Roberg K, Brunk U. H2O2-mediated damage to Cancer Res 2003;9:2032 – 9. lysosomal membranes of J-774 cells. Free Radic Res Commun 1993;18: 23. Fassina G, Aluigi M, Gentleman S, et al. The cAMP analog 8-Cl-cAMP 71 – 85. inhibits growth, induces differentiation and apoptosis in retinoblastoma 43. Cirman T, Oresic K, Mazovec GD, et al. Selective disruption of cells. Int J Cancer 1997;72:1088 – 94. lysosomes in HeLa cells triggers apoptosis mediated by cleavage of Bid by 24. Bottenstein J. Advances in vertebrate cell culture methods. Science multiple papain-like lysosomal cathepsins. J Biol Chem 2004;279:3578 – 1988;239:G42, G8. 87. Epub 2003 Oct 27. 25. Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. A novel 44. Johansson AC, Steen H, Ollinger K, Roberg K. Cathepsin D mediates c assay for apoptosis. Flow cytometric detection of phosphatidylserine cytochrome release and caspase activation in human fibroblast expression on early apoptotic cells using fluorescein labelled Annexin V. apoptosis induced by staurosporine. Cell Death Differ 2003;10:1253 – 9. J Immunol Methods 1995;184:39 – 51. 45. Jaattela M, Cande C, Kroemer G. Lysosomes and mitochondria in the 26. Ono K, Kim SO, Han J. Susceptibility of lysosomes to rupture is a commitment to apoptosis: a potential role for cathepsin D and AIF. Cell determinant for plasma membrane disruption in tumor necrosis factor a- Death Differ 2004;11:135 – 6. induced cell death. Mol Cell Biol 2003;23:665 – 76. 46. Antunes F, Cadenas E, Brunk UT. Apoptosis induced by exposure to a 27. Schotte P, Declercq W, Van Huffel S, Vandenabeele P, Beyaert R. low steady-state concentration of H2O2 is a consequence of lysosomal Non-specific effects of methyl ketone peptide inhibitors of caspases. FEBS rupture. Biochem J 2001;356:549 – 55. Lett 1999;442:117 – 21. 47. Waterhouse NJ, Goldstein JC, von Ahsen O, et al. Cytochrome c 28. Kagedal K, Johansson U, Ollinger K. The lysosomal protease maintains mitochondrial transmembrane potential and ATP generation cathepsin D mediates apoptosis induced by oxidative stress. FASEB J after outer mitochondrial membrane permeabilization during the apoptotic 2001;15:1592 – 4. process. J Cell Biol 2001;153:319 – 28. 29. Raymond MA, Mollica L, Vigneault N, et al. Blockade of the apoptotic 48. Zang Y, Beard RL, Chandraratna RA, Kang JX. Evidence of a machinery by cyclosporin A redirects cell death toward necrosis in arterial lysosomal pathway for apoptosis induced by the synthetic retinoid endothelial cells: regulation by reactive oxygen species and cathepsin D. CD437 in human leukemia HL-60 cells. Cell Death Differ 2001;8:477 – 85. FASEB J 2003;17:515 – 7. Epub 2003 Jan 2. 49. Sporn MB, Liby KT. Cancer chemoprevention: scientific promise, 30. Nylandsted J, Gyrd-Hansen M, Danielewicz A, et al. Heat shock clinical uncertainty. Nat Clin Pract Oncol 2005;2:518 – 25.

MolCancerTher2007;6(1).January2007

Downloaded from mct.aacrjournals.org on September 26, 2021. © 2007 American Association for Cancer Research. Novel cell death pathways induced by N -(4-hydroxyphenyl)retinamide: therapeutic implications

Roberta Venè, Giuseppe Arena, Alessandro Poggi, et al.

Mol Cancer Ther 2007;6:286-298.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/6/1/286

Cited articles This article cites 48 articles, 17 of which you can access for free at: http://mct.aacrjournals.org/content/6/1/286.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/6/1/286.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/6/1/286. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.