Hrk Mediates 2-Methoxyestradiol–Induced Mitochondrial Apoptotic Signaling in Prostate Cancer Cells

Published OnlineFirst April 11, 2013; DOI: 10.1158/1535-7163.MCT-12-1187

Molecular Cancer Cancer Therapeutics Insights Therapeutics

Inik Chang, Shahana Majid, Sharanjot Saini, Mohd S. Zaman, Soichiro Yamamura, Takeshi Chiyomaru, Varahram Shahryari, Shinichiro Fukuhara, Guoren Deng, Rajvir Dahiya, and Yuichiro Tanaka

Abstract Prostate cancer is one of the most prevalent cancers in males and ranks as the second most common cause of cancer-related deaths. 2-methoxyestradiol (2-ME), an endogenous estrogen metabolite, is a promising Jun anticancer agent for various types of cancers. Although 2-ME has been shown to activate c- -NH2-kinase (JNK) and mitochondrial-dependent apoptotic signaling pathways, the underlying mechanisms, including downstream effectors, remain unclear. Here, we report that the human Bcl-2 homology 3 (BH3)-only protein harakiri (Hrk) is a critical effector of 2-ME–induced JNK/mitochondria–dependent apoptosis in prostate cancer cells. Hrk mRNA and protein are preferentially upregulated by 2-ME, and Hrk induction is dependent on the JNK activation of c-Jun. Hrk knockdown prevents 2-ME–mediated apoptosis by attenuating the decrease in mitochondrial membrane potential, subsequent cytochrome c (cyt c)release, and caspase activation. Involvement of the proapoptotic protein Bak in this process suggested the possible interaction between Hrk and Bak. Thus, Hrk activation by 2-ME or its overexpression displaced Bak from the

complex with antiapoptotic protein Bcl-xL, whereas deletion of the Hrk BH3 domain abolished its

interaction with Bcl-xL, reducing the proapoptotic function of Hrk. Finally, Hrk is also involved in the 2-ME–mediated reduction of X-linked inhibitor of apoptosis through Bak activation in prostate cancer cells. Together, our ﬁndings suggest that induction of the BH3-only protein Hrk is a critical step in 2-ME activation of the JNK-induced apoptotic pathway, targeting mitochondria by liberating proapoptotic protein Bak. Mol Cancer Ther; 12(6); 1049–59. 2013 AACR.

Introduction in membrane potential and subsequent release of cyto- c c 2-Methoxyestradiol (2-ME) is produced in vivo by cat- chrome (cyt ) into the cytosol leading to caspase acti- O O vation (4–6). Although the mechanism by which JNK echol- -methyltransferase–mediated -methylation of 2- c hydroxyestradiol (1). Although it is initially considered to mediates cyt release is not fully understood, several be an inactive end product of estrogen metabolism, 2-ME lines of evidence indicate that the Bcl-2 gene family may has emerged as a potential antitumor agent for several be the potential targets of JNK (7). In the unstimulated types of cancer including prostate cancer (2). The molec- state, the antiapoptotic proteins Bcl-2 and Bcl-xL neutral- ular network of 2-ME action is a complex process that ize the function of proapoptotic proteins Bax and Bak, involves various pathways, and despite extensive inves- which constitute the pore or channel that permeabilize tigation, the precise mechanisms for antitumor activity of mitochondria (8). 2-ME–mediated JNK activation regu- 2-ME are not fully elucidated (2, 3). lates the inhibitory function of antiapoptotic proteins by Jun phosphorylation in prostate cancer cells (9, 10). JNK also c- -NH2-kinase (JNK) activation plays an important role in the mitochondrial apoptotic pathway induced by 2- modulate the activities of proapoptotic BH3-only proteins ME (3). Following 2-ME exposure, activated JNK is trans- of the Bcl-2 family, especially Bid, Bim, Bmf, and Bad, at located to the mitochondria, where it initiates a decrease the posttranslational level and by doing so, BH3-only proteins can engage Bax and Bak to release cyt c by inactivating the antiapoptotic proteins (7). Authors' Afﬁliation: Department of Urology, San Francisco Veterans Harakiri (Hrk) belongs to the human BH3-only protein Affairs Medical Center and University of California at San Francisco, San Francisco, California subgroup of the proapoptotic Bcl-2 family and was iden- tiﬁed by its ability to bind Bcl-2 and Bcl-xL in a yeast two- Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). hybrid screen (11). Exogenous expression of Hrk activates cell death and this is repressed by overexpression of Bcl-2 Corresponding Author: Yuichiro Tanaka, Department of Urology, San Francisco Veterans Affairs Medical Center and University of California San and Bcl-xL (11, 12). Bax and mitochondrial protein p32 are Francisco, 4150 Clement Street, San Francisco, CA 94121. Phone: 415- suggested as direct effector molecules for Hrk action (13, 750-2031; Fax: 415-750-6639; E-mail: [email protected] 14). However, despite these studies, the precise molecular doi: 10.1158/1535-7163.MCT-12-1187 mechanism involved in Hrk-mediated cell death remains 2013 American Association for Cancer Research. largely unknown.

www.aacrjournals.org 1049

Chang et al.

It has been suggested that Hrk downregulation con- cDNAs were amplified with the TaqMan Gene Expres- tributes to the development and progression of cancers sion Assays and TaqMan Fast Universal PCR Master Mix (15). Hrk expression is frequently lost in colorectal and using the 7500 Fast Real-Time PCR System (Applied gastric cancer (16), glioblastoma (17), primary central Biosystems). To investigate the expression of genes key nervous system lymphoma (18), and prostate cancer to apoptosis, the Human Apoptosis RT2 Profiler PCR (19) due to aberrant methylation of its promoter region. Array (SABiosciences) were used as per manufacturer’s Hrk inactivation is also inversely correlated with apopto- instructions. sis indices in tumors (17–19). Hrk is located in chromosome 12q13 (20) where loss of heterozygosity is often Western blot analysis observed in several types of human tumors (21–23). Whole-cell extracts were prepared using radioimmu- Moreover, Hrk overexpression suppresses cell growth in noprecipitation assay buffer (Thermo Scientific) contain- prostate, breast, and ovarian cancer cells (24). Despite its ing protease inhibitor cocktails (Roche). For subcellular potential significance, functional role of Hrk has not been fractionation, cytosolic and membrane proteins were iso- defined in cancer. lated with a Subcellular Protein Fractionation Kit for In this study, we found that 2-ME induces Hrk in a Cultured Cells (Thermo Scientific). Immunoblotting was JNK-dependent manner in prostate cancer cells. We also carried out according to standard protocols with antibo- show that Hrk is involved in the 2-ME–induced apoptotic dies against Hrk (Sigma), cyt c, phospho-JNK, JNK, c-Jun, pathway by activating caspase through Bak-mediated Bcl-xL, Bcl-2, Bak, X-linked inhibitor of apoptosis (XIAP; cyt c release. Therefore Hrk, a newly identified target of Cell Signaling Technology), phospho-c-Jun (Ser 63/73), 2-ME action, is a critical downstream effector of JNK- JNK1 (Santa Cruz Biotechnology), and FLAG (OriGene). dependent mitochondrial apoptotic signaling pathway Antibodies against glyceraldehyde-3-phosphate dehy- of 2-ME in prostate cancer cells. drogenase (GAPDH) and b-actin (Santa Cruz) were used to confirm equal loading. Materials and Methods Cell lines and reagents Caspase enzymatic activity assay Human prostate carcinoma cell lines (LNCaP, PC-3, and Caspase-3 activation was measured using a Caspase-3 DU-145) were obtained from the American Type Culture Assay Kit as described by the manufacturer’s instructions Collection. All cell lines were authenticated at ATCC before (BD Biosciences). purchase by standard DNA typing. Eagle’s Minimum Essential Medium (EMEM), RPMI-1640, Opti-MEM, and Transfection penicillin/streptomycin mixtures were obtained from Cells were transfected with pCMV6-ENTRY vector the University of California San Francisco Cell Culture expressing the C-terminally FLAG-tagged human Hrk Facility. FBS was a product of Atlanta Biologicals. 2-ME cDNA and empty pCMV6-ENTRY vector as a control and SP600125 were obtained from Sigma and chemical (OriGene) using Fugene HD Transfection Reagent structures are shown in Fig. 6. (Roche) according to the manufacturer’s protocol. For small interfering RNA (siRNA) transfection, siRNA Cell culture duplexes (20 nmol/L) or universal scrambled negative The LNCaP and PC-3 cell lines were grown in RPMI- control (OriGene) was transfected using Lipofectamine 1640. The DU-145 cell line was cultured in EMEM. All 2000 Transfection Reagent (Invitrogen) as described by culture medium contained 10% FBS and 100 mg/mL the manufacturer’s instructions. Target specificity and penicillin/streptomycin mixture. All cell lines were main- knockdown efficiency were evaluated by real-time (RT) tained at 37C in a humidified atmosphere composed of PCR with 3 different sets of siRNA duplexes at different 5% CO2/95% air. concentrations.

Apoptosis assay Immunoprecipitation Cells were stained with an Annexin V-ﬂuorescein iso- Cell lysates were prepared in lysis buffer (20 mmol/L thiocyanate (FITC)/7-amino-actinomycin D (7-AAD; BD Tris-pH 7.4, 135 mmol/L NaCl, 1.5 mmol/L MgCl2, Biosciences) as described by the manufacturer and ana- 1 mmol/L EGTA, 10% glycerol) containing 1% CHAPS lyzed by a Cell Lab Quanta SC MPL (Beckman Coulter). (Cell Signaling Technology), supplemented with protease Both early (Annexin V-positive, 7-AAD–negative) and inhibitor cocktails (Roche). Immunoprecipitation was late (Annexin V-positive, 7-AAD–positive) apoptotic cells conducted with a Pierce Direct IP Kit (Pierce) according were included in cell death determinations. to the manufacturer’s instructions.

Quantitative RT-PCR Assessment of mitochondrial membrane potential Total RNA was isolated using the RNeasy Mini Kit 2-ME–Treated cells were stained with CMXRos (Mito- (Qiagen) and was converted into cDNA by using the tracker Red; Molecular Probes) in PBS for 20 minutes at iScript cDNA Synthesis Kit (Bio-Rad) according to the 37C, and analyzed by a Cell Lab Quanta SC MPL (Beck- manufacturer’s instructions. To assess gene expression, man Coulter).

1050 Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics

Hrk Is a Novel Target of 2-ME in Prostate Cancer

Statistical analysis Table 1. Summary of apoptosis-related genes Values are presented as the mean SEM based on results identified as significantly altered in LNCaP cells obtained from at least 3 independent experiments. Statis- tical significance was evaluated by conducting a 2-tailed by 2-ME treatment for 12 hours unpaired Student t test using GraphPad PRISM Software. A P < 0.05 was regarded as statistically significant. Gene namea Fold changeb P TP53 9.3 <0.01 Results Hrk 6.7 <0.01 Hrk is induced by 2-ME in prostate cancer cells DAPK1 3.4 <0.01 Dose-dependent apoptosis assays show that 2-ME was TNF 3.0 <0.01 effective at a concentration of 1 mmol/L (Fig. 1A), and TNFSF8 2.4 <0.01 modest degrees of apoptotic cell death were noted after CASP1 2.3 <0.05 12 hours of 2-ME exposure reaching maximal levels after BIRC8 0.5 <0.05 72 hours (Fig. 1B). On the basis of these results, andro- IGF1R 0.4 <0.05 gen-dependent LNCaP and androgen-independent PC-3 XIAP 0.4 <0.01 cells and 1 mmol/Lof2-MEwereusedinsubsequent PYCARD 0.2 <0.01 experiments. aGenes identified as increased or decreased by 2-fold or To identify 2-ME signaling pathways, we looked for more are listed. changes in gene expression in LNCaP cells exposed to 2- bFold change represents the ratio of signal in 2-ME–treated ME using the Human Apoptosis RT2 Profiler PCR Array. LNCaP cells relative to dimethyl sulfoxide-treated LNCaP Among several affected genes, we detected an increase of cells for each primer set. TP53 gene expression similar to a previous report showing its induction by 2-ME (25). Also, Hrk was induced 6.7- fold over controls (Table 1). A significant increase in Hrk We also evaluated whether mRNA expression of other mRNA expression was detected as early as 1 hour after 2- BH3-only proteins such as Bim, Puma, and Noxa are ME exposure in both LNCaP and PC-3 cells (Fig. 1C). An influenced by 2-ME. Induction of Puma and Noxa was increase in Hrk protein expression was also noted after 6- at modest levels compared with Hrk in LNCaP cells hour exposure of 2-ME reaching maximum levels after 36 (Supplementary Fig. S1). As Puma and Noxa are p53- hours in LNCaP and 24 hours in PC-3 cells (Fig. 1D). dependent genes, we examined induction of these genes

Figure 1. 2-ME induces apoptosis and Hrk activation. A and B, 2-ME induces apoptotic cell death. LNCaP (*), PC-3 (&), and DU-145 (~) cells were treated with indicated concentration (A) or 1 mmol/L (B) of 2- ME. Cell death was determined by double staining with Annexin V-FITC and 7-AAD at 48 hours (A) or various time points (B) after 2-ME treatment. C, induction of Hrk mRNA in cells treated with 2-ME for the indicated time interval. Hrk expression determined by quantitative RT-PCR is normalized to GAPDH and values are presented as fold increase relative to Hrk expression in dimethyl sulfoxide (DMSO)-treated cells (0 hours). , P < 0.01, , P < 0.001 compared to 0 hours. D, upregulation of Hrk protein expression as determined by Western blot analysis in cells treated with 2-ME for the indicated time interval.

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1051

Chang et al.

in PC-3 and DU-145 cells, which have mutated p53. Unlike Although JNK regulation of Hrk levels has been LNCaP cells, which have wild-type p53, Puma and Noxa studied in cultured neurons and pancreatic b-cells were not increased by 2-ME in PC-3 and DU-145 cells. Bim (26–28), it is unknown in cancers. Therefore, we exam- was induced only in PC-3 cells (Supplementary Fig. S1). ined whether JNK activation is required for Hrk upregulation in prostate cancer cells using JNK inhibitors. JNK signaling induces Hrk in prostate cancer cells SP600125, a potent JNK inhibitor, suppressed Hrk 2-ME caused an increase in the level of phospho-JNK induction by 2-ME (Fig. 2B). As a further test of spec- without detectable changes in total JNK levels (Fig. 2A). ificity, we conducted selective knockdown of the JNK This was followed by c-Jun phosphorylation, paralleling pathway using a siRNA. JNK1 siRNA significantly the increase in JNK phosphorylation. The c-Jun protein reduced the level of endogenous JNK1 protein (Fig. level was significantly increased, the probable conse- 2C) and resulted in a dramatic decrease in Hrk mRNA quence of autoregulation (Fig. 2A). induction (Fig. 2D). In addition, c-Jun siRNA efficiently

Figure 2. JNK activation of c-Jun mediates Hrk upregulation. A, JNK-mediated c-Jun activation by 2-ME in prostate cancer cells. Protein expression was determined by Western blot analysis in cells treated with 2-ME for the indicated time interval. B, JNK inhibitor, SP600125 attenuates Hrk upregulation. Hrk expression was determined by quantitative RT-PCR in cells exposed to 2-ME in the absence or presence of 10 mmol/L SP600125 for 24 hours. Hrk expression is normalized to GAPDH and values are presented as fold increase relative to gene expression in untreated cells. , P < 0.001. C, suppression of endogenous JNK1 by siRNA knockdown. JNK1 expression was determined by Western blot analysis in cells transfected with nonspeciﬁc (NS) control or JNK1 siRNA for 24 hours. D, JNK knockdown represses Hrk induction. Hrk expression was determined by quantitative RT-PCR in cells transfected with NS or JNK1 siRNA for 24 hours and then treated with 2-ME for 12 hours. Hrk expression was normalized to GAPDH and values are presented as fold increase relative to gene expression in NS siRNA-transfected cells. P < 0.001. E, suppression of endogenous c-Jun by siRNA knockdown. C-jun expression was determined by Western blot analysis in cells transfected with NS or c-Jun siRNA for 24 hours. F, c-Jun knockdown represses Hrk induction. Hrk expression was determined by quantitative RT-PCR in cells transfected with NS or c-Jun siRNA for 24 hours and then treated with 2-ME for 12 hours. Hrk expression was normalized to GAPDH and values are presented as fold increase relative to gene expression in NS siRNA-transfected cells. , P < 0.05; , P < 0.01.

1052 Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics

Hrk Is a Novel Target of 2-ME in Prostate Cancer

suppressed endogenous c-Jun (Fig. 2E) and led to inhi- results led us to examine whether Hrk affected the asso- bition of Hrk mRNA induction in prostate cancer cells ciation between Bak and Bcl-xL. (Fig. 2F). These results indicate that Hrk is regulated by While the levels of Hrk interacting with Bcl-xL were JNK signaling in prostate cancer cells. increased by 2-ME, the level of Bak associated with Bcl- xL was diminished (Fig. 4E). This change is abolished by Hrk upregulation is required for 2-ME–induced Hrk knockdown (Fig. 4F). We subsequently evaluated apoptotic cell death whether overexpression of Hrk affects the interaction of To investigate functional significance of Hrk in 2-ME– Bak with Bcl-xL. Consistent with 2-ME treatment, the mediated apoptosis of prostate cancer cells, we conducted level of Bak associated with Bcl-xL was reduced in Hrk knockdown experiments using Hrk-specific siRNA. transfectants compared with the vector control (Fig. 4G). Transfection with 2 different Hrk siRNAs resulted in a Together, these data show that Hrk contributes to 2-ME– dramatic reduction in endogenous levels of Hrk mRNA mediated apoptosis by liberating Bak from the complex (Fig. 3A) and suppressed 2-ME–mediated induction of with Bcl-xL. Hrk mRNA and protein compared with a control siRNA BH3 domain of BH3-only proteins mediates its associ- (Fig. 3A and B). Hrk knockdown did not affect cell via- ation with antiapoptotic proteins and is required for cell bility under basal conditions, but significantly reduced death induction (30). To determine whether the BH3 the cytotoxic effect of 2-ME (Fig. 3C and Supplementary domain of Hrk is critical for 2-ME–mediated prostate Fig. S2). Also Hrk siRNA-1, which had a more significant cancer cell apoptosis, we engineered a mutant form of knockdown effect than siRNA-2, caused less 2-ME– Hrk lacking the BH3 domain (Hrk DBH3). Substantial induced cell death than siRNA-2 (Fig. 3A–C). amounts of both wild-type and mutant Hrk proteins were As 2-ME potently induces apoptosis through the mito- expressed in transiently transfected prostate cancer cells chondrial apoptotic pathway (29) and Hrk is predomi- (Supplementary Fig. S7A). Absence of the BH3 domain nantly located in the mitochondria (Supplementary dramatically reduced the ability of Hrk to induce apopto- Fig. S3), we examined the effect of Hrk knockdown on tic cell death (Supplementary Fig. S7B) and to interact with the mitochondrial apoptotic events triggered by 2-ME. Bcl-xL (Supplementary Fig. S7C). As a consequence, there As shown in Fig. 3D, 2-ME caused a significant increase was no change in the level of Bak associated with Bcl-xL by of CMXRos-negative cells indicating the reduction of Hrk DBH3 overexpression (Supplementary Fig. S7D). mitochondrial membrane potential but Hrk knockdown These results indicate that the BH3 domain-mediated reduced this change. In a parallel experiment, no differ- interaction with Bcl-xL is critical for the proapoptotic ence was observed in total cell number by 2-ME treat- activity of Hrk in prostate cancer cells. ment (Supplementary Fig. S4). This result excludes the possibility that the decrease in CMXRos-stained cells is Hrk-mediated Bak activation represses XIAP levels not a secondary result caused by a reduction in mito- To investigate additional downstream target molecules chondrial number but rather the result of a decrease in of Hrk activation, we evaluated gene mRNA levels altered membrane potential following 2-ME exposure. Hrk by 2-ME (Table 1) in cells overexpressing Hrk (Fig. 5A). knockdown also diminished 2-ME–induced cyt c release From the genes analyzed, we found that only XIAP mRNA into the cytosol from mitochondria (Fig. 3E) and subse- was significantly decreased by 40% compared with cells quent caspase-3 activation (Fig. 3F). Collectively, these transfected with vector control (Fig. 5B). We also verified results indicate that Hrk plays an important functional the decrease in XIAP mRNA expression by 2-ME (Fig. 5C). role in 2-ME–induced prostate cancer cell apoptosis by Furthermore, XIAP protein expression was significantly targeting mitochondria. reduced by both 2-ME exposure (Fig. 5D) and Hrk overexpression (Fig. 5E). Next, we evaluated XIAP levels in Hrk triggers apoptotic cell death by displacing Bak cells transfected with Bak siRNA after 2-ME treatment to assess whether Hrk activation of Bak is required for the from Bcl-xL sequestration regulation of XIAP level. As shown in Fig. 5F, Bak knock- Hrk physically interacts with Bcl-2 and Bcl-xL, but not Bax or Bak (11). Hrk-induced cell death is repressed by down abolished the reduction of XIAP protein in cells treated with 2-ME. Therefore, these data indicate that Hrk overexpression of Bcl-2 and Bcl-xL (11, 12). Therefore, we hypothesized that Hrk triggers apoptosis by neutralizing participates in the 2-ME-mediated downregulation of XIAP by activating Bak in prostate cancer cells. Bcl-2 and/or Bcl-xL rather than directly activating Bax and/or Bak. Reduction of 2-ME–induced cell death with knockdown of Bak (Fig. 4A and B) but not Bax (Supple- Discussion mentary Fig. S5) indicates that Bak activation is a 2-ME In this study, we identified the BH3-only protein, Hrk, downstream apoptotic signal. Moreover, Bak knockdown as a new 2-ME target gene, acting downstream of the JNK diminished 2-ME–induced cyt c release into cytosol signaling pathway in prostate cancer cells. Hrk is a par- (Fig. 4C) and subsequent caspase-3 activation (Fig. 4D) ticularly interesting candidate because: (i) it belongs to the consistent with the results obtained by Hrk knockdown proapoptotic Bcl-2 gene family (11, 12), (ii) it is down- (Fig. 3E and F). In prostate cancer cells, Bak binds with Bcl- regulated in prostate cancer and loss of Hrk expression is xL, but not with Bcl-2 (Supplementary Fig. S6). These closely associated with decreased apoptosis in high-grade

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1053

Chang et al.

Figure 3. Hrk knockdown prevents 2-ME-mediated apoptotic cell death. A, suppression of the 2-ME–mediated Hrk induction by Hrk siRNA. Hrk expression determined by quantitative RT-PCR in cells transfected with NS or Hrk siRNAs for 24 hours and then treated with 2-ME for 12 hours. Hrk expression is normalized to GAPDH and values are presented as fold increase relative to Hrk expression in NS siRNA-transfected cells. , P < 0.05; , P < 0.01; , P < 0.001. B, suppression of the 2-ME–mediated Hrk protein induction by Hrk siRNA. Hrk expression was determined by Western blot analysis in cells transfected with NS control or Hrk siRNAs for 24 hours and then treated with 2-ME for 12 hours. C, Hrk knockdown prevents 2-ME–mediated apoptotic cell death. Cells were transfected with NS or Hrk siRNAs for 24 hours and then treated with 2-ME for 48 hours. The number of apoptotic cells was determined by double staining with Annexin V-FITC and 7-AAD. , P < 0.01. D, Hrk knockdown prevents 2-ME–mediated decrease in mitochondrial membrane potential. Cells were transfected with NS or Hrk siRNA-1 for 24 hours and then treated with 2-ME for 6 hours. Reduction in mitochondrial membrane potential was determined by monitoring uptake of CMXRos. , P < 0.05; , P < 0.01. E, Hrk knockdown represses 2-ME–mediated cyt c release. Cells were transfected with NS or Hrk siRNA-1 for 24 hours and then treated with 2-ME for 12 hours. Lysates of cytosolic (C) and membrane (M) fractions were subjected to Western blot analysis. F, Hrk knockdown impedes 2-ME–mediated caspase activation. Caspase enzymatic activity was assessed with whole-cell lysates from cells transfected with NS or Hrk siRNA for 24 hours and then treated with 2-ME for 12 hours. , P < 0.01.

prostate tumors (19), (iii) its expression is regulated by Puma, and Noxa, 2-ME–mediated Hrk induction is not JNK (Fig. 2) which is a downstream molecule of 2-ME affected by androgen dependency or p53 status of cells (3, 26), (iv) unlike other BH3-only proteins such as Bim, (Supplementary Fig. S1).

1054 Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics

Hrk Is a Novel Target of 2-ME in Prostate Cancer

Figure 4. Hrk disrupts the association of Bak with Bcl-xL. A, suppression of endogenous Bak by siRNA knockdown. Bak expression was determined by Western blot analysis in cells transfected with nonspeciﬁc (NS) control or Bak siRNA for 24 hours. B, Bak knockdown prevents 2-ME–mediated apoptotic cell death. Cells were transfected with NS or Bak siRNA for 24 hours and then treated with 2-ME for 48 hours. The number of apoptotic cells was determined by double staining with Annexin V-FITC and 7-AAD. , P < 0.05. C, Bak knockdown represses 2-ME–mediated cyt c release. Cells were transfected with NS or Bak siRNA for 24 hours and then treated with 2-ME for 12 hours. Lysates of cytosolic (C) and membrane (M) fractions were subjected to Western blot analysis. D, Bak knockdown impedes 2-ME–mediated caspase activation. Caspase enzymatic activity was assessed with whole-cell lysates from cells transfected with NS or Hrk siRNA for 24 hours and then treated with 2-ME for 12 hours. , P < 0.05; , P < 0.01. E,

2-ME–mediated Hrk activation attenuates Bak association with Bcl-xL. Lysates from cells treated with DMSO or 2-ME were immunoprecipitated with anti-Bcl-xL or control IgG antibody and immunoblotted with anti-Bak or -Hrk antibody. Immunoblots of whole-cell lysates with anti-Bak antibody represents equal amount of Bak expression in cells. F, Hrk knockdown prevents disruption of the interaction between Bak and Bcl-xL.Cellswere transfected with Hrk siRNA-1 and then treated with DMSO or 2-ME for 12 hours. Lysates were immunoprecipitated with anti-Bcl-xL or control IgG antibody and immunoblotted with anti-Bak antibody. G, Hrk overexpression diminishes Bak association with Bcl-xL. Cells were transiently transfected with either empty vector (pCMV6) or vector containing Hrk cDNA (Hrk) for 24 hours. Lysates were immunoprecipitated with anti-Bcl-xL or control IgG antibody and immunoblotted with anti-Bak antibody. Immunoblots of whole-cell lysates with anti-FLAG and -Bak antibodies represent Hrk overexpression and equal amount of Bak expression in cells, respectively.

Although it has been well described that JNK activation other investigators (31). Therefore, DAPK1 and Smad7 is involved in the 2-ME–mediated cytotoxic effect in may be downstream signaling components of 2-ME that cancer cells, the upstream components regulating JNK activate the JNK cascade. Further investigation regarding activation are still unknown. Davoodpour and Landstrom the interaction between DAPK1 and Smad7 will be need- suggest that Smad7, which is stabilized by 2-ME, is ed to delineate the precise molecular mechanisms of 2- required for 2-ME–induced JNK activation (6). In addi- ME–mediated JNK activation. tion, death-associated protein kinase 1 (DAPK1), which JNK activation occurs rapidly following 2-ME treat- was induced by 2-ME in our study (Table 1) also regulates ment and plays an important role in the 2-ME–mediated JNK signaling through protein kinase D1 as described by mitochondrial apoptotic pathway (3). Although JNK-

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1055

Chang et al.

Figure 5. Hrk overexpression represses XIAP levels. A, Hrk overexpression in LNCaP cells. Hrk expression was determined by Western blot analysis in cells transiently transfected with either empty vector (pCMV6) or vector containing Hrk cDNA (Hrk) for 24 hours. B, XIAP expression is decreased in Hrk overexpressing LNCaP cells. Gene expression was determined by quantitative RT-PCR in cells transiently transfected with either empty vector (pCMV6) or vector containing Hrk cDNA (Hrk) for 24 hours. Gene expression is normalized to GAPDH and values are presented as fold change relative to gene expression in pCMV6-transfected cells. , P < 0.001. C, 2-ME reduces endogenous XIAP levels. XIAP mRNA levels determined by quantitative RT-PCR in LNCaP cells treated with 2-ME for 18 hours. XIAP expression is normalized to GAPDH and values are presented as fold change relative to expression in DMSO-treated cells. , P < 0.001. D and E, decrease in XIAP protein levels by 2-ME (D) and Hrk overexpression (E). XIAP protein level was determined by Western blot analysis in cells treated with 2-ME (D) or transiently transfected with either empty vector (pCMV6) or vector containing Hrk cDNA (Hrk) for 24 h (E). F, Bak knockdown reverses 2-ME–mediated XIAP reduction. XIAP expression was determined by Western blot analysis in cells transfected with NS or Bak siRNA for 24 hours and then treated with 2-ME for 12 hours.

induced cyt c release is a critical step for downstream addition to Hrk, 2-ME upregulates other BH3-only proapoptotic signaling, the mechanism is unclear. 2-ME– teins Bim, Puma, and Noxa (Supplementary Fig. S1). Bim induced JNK phosphorylation and subsequent translo- and Puma can directly activate Bax and Bak, and they also cation to the mitochondria was proposed as a possible sequester all antiapoptotic Bcl-2 family members (8). explanation (4). On the basis of the ﬁndings that JNK Noxa neutralizes only Mcl-1 and A1 and promotes Mcl- mitochondrial signaling modulated the activities of 1 degradation (32). Despite the idea that Hrk might pro- BH3-only proteins at the posttranslational level in other mote cell death by inhibiting the protective activity of Bcl- cell types by different apoptotic stimuli (7), it is highly 2 and Bcl-xL (11), the precise mechanism of Hrk has not possible that BH3-only proteins might also be involved been extensively studied. The results of this study suggest in 2-ME activation of JNK-mediated cyt c release. In this a mechanism for Hrk-induced apoptotic cell death by study, we have shown that the BH3-only protein Hrk is regulating the interaction of Bak with Bcl-xL. In untreated a key molecule that links JNK signaling to mitochondria prostate cancer cells, Bcl-xL sequesters active forms of Bak; targeted by the interaction of Bak with Bcl-xL to induce upon 2-ME treatment, transcriptionally activated Hrk c apoptosis accompanied by cytosolic release of cyt . displaces Bak from the heterodimer with Bcl-xL, thus How BH3-only proteins trigger the activation of Bax or releasing Bak (Fig. 4E and G). This promotes Bak oligo- Bak has been a central issue in the regulation of apoptosis merization on the mitochondrial membrane and results in by the Bcl-2 gene family (8). BH3-only proteins could cyt c release followed by induction of apoptosis. Thus, our promote activation of Bax or Bak via their ability to data support the indirect activation or displacement mod- inactivate antiapoptotic Bcl-2 family members. Alterna- el to explain how Hrk activates Bak in addition to the tively, Bax or Bak may be activated via direct association results showing no association of Bim and Puma with Bak with certain BH3-only proteins. We observed that in after 2-ME treatment in prostate cancer cells (I. Chang;

1056 Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics

Hrk Is a Novel Target of 2-ME in Prostate Cancer

teins Bcl-2, Bcl-xL, and Mcl-1 are significantly expressed and confer resistance to apoptotic signaling in prostate cancer cells (36). Hrk depletion causes partial reduction of 2-ME–induced apoptosis (Fig. 3C). This can be explained by the incomplete knockdown of Hrk but also indicates the possibility that Hrk activation is not entirely responsible for induction of apoptosis and other mechanisms are involved. Although 2-ME causes diverse apoptotic signaling, it is an interesting possibility that both Hrk and Noxa are involved in 2-ME–mediated prostate cancer cell death. XIAP is often overexpressed in malignant cells and elevated levels of XIAP increase resistance to apoptosis (37). It has been shown that 2-ME reduces the levels of XIAP in prostate cancer cells (38, 39). However, the underlying mechanism is largely unknown. In this Figure 6. Chemical structures of 2-ME (A) and SP600125 (B). study, we found that Bak activation is essential for 2- ME–mediated XIAP reduction and Hrk plays an important role in the process by Bak activation and subse- unpublished data). However, this does not entirely rule quent cyt c release. It has been reported that 2-ME also out the possibility that the direct activation model might induces a second mitochondrial-derived activator of also be involved in this mechanism. For instance, Bim and caspase (Smac) (4). Therefore, in addition to cyt c-medi- Puma upregulated by 2-ME may independently cause Bax ated caspase activation, the release of Smac from mito- or Bak activation. It is also possible that Bim and Puma chondria presumably mediated by Bak activation may may interact with Bax instead of Bak. Kuwana and col- facilitate the processing of caspase-3 by competitive leagues observed that Hrk seemed to activate Bax directly binding to XIAP and liberate the bound caspases. The although it was modest and the significance of the acti- increased caspase activation could further inactivate vation was uncertain (33). Therefore, whether 2-ME– XIAP by proteolysis. mediated induction of BH3-only proteins activates Bak It is well known that antiapoptotic proteins Bcl-2 and or Bax directly or indirectly remains a matter of discussion Bcl-xL are substantially expressed and confer resistance to and needs further clarification. apoptotic signaling in prostate cancer cells (36). 2-ME On the basis of previous findings, Hrk may regulate activation of JNK phosphorylates Bcl-2 and Bcl-xL to apoptotic cell death via several different mechanisms. suppress their antiapoptotic function (9, 10). Aside from Rizvi and colleagues reported that C6 ceramide induces this mechanism, Hrk induction is a newly discovered not only Hrk expression through JNK phosphorylation inhibitor of Bcl-xL and an initiator of apoptosis in human but also its translocation to mitochondria where Hrk prostate cancer cells. As Hrk expression is downregulated interacts with the mitochondrial protein p32 and the in prostate cancers through promoter methylation (19), a BH3-only protein Bad. These protein interactions even- combination of 2-ME with a DNA methylation inhibitor tually lead to mitochondria dysfunction and cell death in may enhance the Hrk-mediated cell death pathway. human corneal stromal fibroblasts (34). It will therefore According to Lin and colleagues, Hrk is the most effective be of interest to determine whether 2-ME also induces suppressor of cancer cell growth compared with p53, translocation of Hrk to mitochondria to interact with p32 BRAC1, and PTEN in prostate, breast, and ovarian cancer and Bad, and plays a role in Hrk proapoptotic activity in cells which are estrogen-associated cancers expressing prostate cancer cells. Unlike Bim, Puma, and Bid, other high levels of Bcl-2 and Bcl-xL (24). Moreover, breast and BH3-only proteins selectively neutralize only a subset of ovarian cancer cells are very sensitive to 2-ME (2). There- antiapoptotic Bcl-2 proteins. Thus, a combination of fore, it would be of interest to determine whether Hrk is several BH3-only proteins binding to complementary involved in the 2-ME–mediated apoptotic pathway in subsets is required to promote apoptosis; for example, these types of cancer. Bad binding to Bcl-2, Bcl-xL, and Bcl-w, plus Noxa In summary, we have identified Hrk as a critical medi- binding to Mcl-1 and A1. In vitro competitive binding ator of 2-ME–induced apoptosis in prostate cancer cells. assays have shown that Hrk has high binding affinity to Our study shows that Hrk links 2-ME activation of JNK Bcl-w and A1 in addition to Bcl-xL but not to Mcl-1 (35). signaling to the mitochondrial apoptotic pathway and Although binding to Bcl-2 is controversial (35), Hrk was contributes to 2-ME–mediated XIAP reduction by releas- originally identified as a Bcl-2 binding protein (11) and ing Bak from the complex with Bcl-xL. This study provides we have also verified its interaction with Bcl-2 in prostate useful insights into the molecular mechanisms underly- cancer cells (I. Chang; unpublished data). Therefore, in ing Hrk-mediated apoptosis and also into the future addition to Bad, Hrk and Noxa can completely neutralize development of 2-ME–based therapeutic strategies for antiapoptotic proteins. Intriguingly, antiapoptotic pro- prostate cancer treatment.

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1057

Chang et al.

Disclosure of Potential Conﬂicts of Interest Acknowledgments No potential conﬂicts of interest were disclosed. The authors thank Dr. Roger Erickson for critical reading of the manuscript. Authors' Contributions Conception and design: Y. Tanaka, I. Chang, R. Dahiya Development of methodology: Y. Tanaka, I. Chang, S. Majid, G. Deng, R. Dahiya Grant Support Acquisition of data (provided animals, acquired and managed patients, This study was supported by the Veterans Affairs Merit Review grant provided facilities, etc.): Y. Tanaka, I. Chang, V. Shahryari, S. Fukuhara and Research Enhancement Award Program (Y. Tanaka). Analysis and interpretation of data (e.g., statistical analysis, biostatis- The costs of publication of this article were defrayed in part by the tics, computational analysis): Y. Tanaka, I. Chang, S. Majid, S. Saini payment of page charges. This article must therefore be hereby marked Writing, review, and/or revision of the manuscript: Y. Tanaka, I. Chang, advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate M.S. Zaman, R. Dahiya this fact. Administrative, technical, or material support (i.e., reporting or orga- nizing data, constructing databases): Y. Tanaka, S. Saini, S. Yamamura, T. Chiyomaru, R. Dahiya Received December 11, 2012; revised March 26, 2013; accepted April 5, Study supervision: Y. Tanaka, R. Dahiya 2013; published OnlineFirst April 11, 2013.

References 1. Guldberg HC, Marsden CA. Catechol-O-methyl transferase: pharma- 18. Nakamura M, Ishida E, Shimada K, Nakase H, Sakaki T, Konishi N. cological aspects and physiological role. Pharmacol Rev 1975;27: Defective expression of HRK is associated with promoter methylation 135–206. in primary central nervous system lymphomas. Oncology 2006;70: 2. Mueck AO, Seeger H. 2-Methoxyestradiol–biology and mechanism of 212–21. action. Steroids 2010;75:625–31. 19. Higuchi T, Nakamura M, Shimada K, Ishida E, Hirao K, Konishi N. HRK 3. Mooberry SL. New insights into 2-methoxyestradiol, a promising inactivation associated with promoter methylation and LOH in prostate antiangiogenic and antitumor agent. Curr Opin Oncol 2003;15:425–30. cancer. Prostate 2008;68:105–13. 4. Chauhan D, Li G, Hideshima T, Podar K, Mitsiades C, Mitsiades N, 20. Muravenko OV, Kashuba VI, Kutsenko AS, Gizatullin RZ, Protopopov et al. JNK-dependent release of mitochondrial protein, Smac, dur- AI, Kvasha SM, et al. Human HRK gene maps to position 12q13.1. ing apoptosis in multiple myeloma (MM) cells. J Biol Chem 2003; Chromosome Res 2000;8:656. 278:17593–6. 21. Wistuba II, Tang M, Maitra A, Alvarez H, Troncoso P, Pimentel F, et al. 5. Gao N, Rahmani M, Dent P, Grant S. 2-Methoxyestradiol-induced Genome-wide allelotyping analysis reveals multiple sites of allelic loss apoptosis in human leukemia cells proceeds through a reactive oxy- in gallbladder carcinoma. Cancer Res 2001;61:3795–800. gen species and Akt-dependent process. Oncogene 2005;24: 22. Shiseki M, Kohno T, Adachi J, Okazaki T, Otsuka T, Mizoguchi H, et al. 3797–809. Comparative allelotype of early and advanced stage non-small cell 6. Davoodpour P, Landstrom M. 2-Methoxyestradiol-induced apoptosis lung carcinomas. Genes Chromosomes Cancer 1996;17:71–7. in prostate cancer cells requires Smad7. J Biol Chem 2005;280: 23. Kim HS, Woo DK, Bae SI, Kim YI, Kim WH. Allelotype of the adenoma- 14773–9. carcinoma sequence of the stomach. Cancer Detect Prev 2001;25: 7. Dhanasekaran DN, Reddy EP. JNK signaling in apoptosis. Oncogene 237–44. 2008;27:6245–51. 24. Lin J, Page C, Jin X, Sethi AO, Patel R, Nunez G. Suppression activity of 8. Chipuk JE, Green DR. How do BCL-2 proteins induce mitochondrial pro-apoptotic gene products in cancer cells, a potential application for outer membrane permeabilization?Trends Cell Biol 2008;18:157–64. cancer gene therapy. Anticancer Res 2001;21:831–9. 9. Bu S, Blaukat A, Fu X, Heldin NE, Landstrom M. Mechanisms for 2- 25. Shimada K, Nakamura M, Ishida E, Kishi M, Konishi N. Roles of p38- methoxyestradiol-induced apoptosis of prostate cancer cells. FEBS and c-jun NH2-terminal kinase-mediated pathways in 2-methoxyes- Lett 2002;531:141–51. tradiol-induced p53 induction and apoptosis. Carcinogenesis 2003; 10. Basu A, Haldar S. Identiﬁcation of a novel Bcl-xL phosphorylation site 24:1067–75. regulating the sensitivity of taxol- or 2-methoxyestradiol-induced 26. Ma C, Ying C, Yuan Z, Song B, Li D, Liu Y, et al. dp5/HRK is a c-Jun apoptosis. FEBS Lett 2003;538:41–7. target gene and required for apoptosis induced by potassium 11. Inohara N, Ding L, Chen S, Nunez G. harakiri, a novel regulator of cell deprivation in cerebellar granule neurons. J Biol Chem 2007;282: death, encodes a protein that activates apoptosis and interacts selec- 30901–9. tively with survival-promoting proteins Bcl-2 and Bcl-X(L). EMBO J 27. Gurzov EN, Ortis F, Cunha DA, Gosset G, Li M, Cardozo AK, et al. 1997;16:1686–94. Signaling by IL-1betaþIFN-gamma and ER stress converge on DP5/ 12. Imaizumi K, Tsuda M, Imai Y, Wanaka A, Takagi T, Tohyama M. Hrk activation: a novel mechanism for pancreatic beta-cell apoptosis. Molecular cloning of a novel polypeptide, DP5, induced during pro- Cell Death Differ 2009;16:1539–50. grammed neuronal death. J Biol Chem 1997;272:18842–8. 28. Young JE, Garden GA, Martinez RA, Tanaka F, Sandoval CM, Smith 13. Harris CA, Johnson EM Jr. BH3-only Bcl-2 family members are AC, et al. Polyglutamine-expanded androgen receptor truncation coordinately regulated by the JNK pathway and require Bax to induce fragments activate a Bax-dependent apoptotic cascade mediated by apoptosis in neurons. J Biol Chem 2001;276:37754–60. DP5/Hrk. J Neurosci 2009;29:1987–97. 14. Sunayama J, Ando Y, Itoh N, Tomiyama A, Sakurada K, Sugiyama A, 29. Qanungo S, Basu A, Das M, Haldar S. 2-Methoxyestradiol induces et al. Physical and functional interaction between BH3-only protein Hrk mitochondria dependent apoptotic signaling in pancreatic cancer and mitochondrial pore-forming protein p32. Cell Death Differ 2004;11: cells. Oncogene 2002;21:4149–57. 771–81. 30. Willis SN, Adams JM. Life in the balance: how BH3-only proteins 15. Nakamura M, Shimada K, Konishi N. The role of HRK gene in human induce apoptosis. Curr Opin Cell Biol 2005;17:617–25. cancer. Oncogene 2008;27 Suppl 1:S105–13. 31. Eisenberg-Lerner A, Kimchi A. DAP kinase regulates JNK signaling by 16. Obata T, Toyota M, Satoh A, Sasaki Y, Ogi K, Akino K, et al. Identi- binding and activating protein kinase D under oxidative stress. Cell ﬁcation of HRK as a target of epigenetic inactivation in colorectal and Death Differ 2007;14:1908–15. gastric cancer. Clin Cancer Res 2003;9:6410–8. 32. Willis SN, Chen L, Dewson G, Wei A, Naik E, Fletcher JI, et al. 17. Nakamura M, Ishida E, Shimada K, Nakase H, Sakaki T, Konishi N. Proapoptotic Bak is sequestered by Mcl-1 and Bcl-xL, but not Bcl- Frequent HRK inactivation associated with low apoptotic index in 2, until displaced by BH3-only proteins. Genes Dev 2005;19: secondary glioblastomas. Acta Neuropathol 2005;110:402–10. 1294–305.

1058 Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics

Hrk Is a Novel Target of 2-ME in Prostate Cancer

33. Kuwana T, Bouchier-Hayes L, Chipuk JE, Bonzon C, Sullivan BA, and mcl-1 expression in prostate cancers. Am J Pathol 1996;148: Green DR, et al. BH3 domains of BH3-only proteins differentially 1567–76. regulate Bax-mediated mitochondrial membrane permeabilization 37. Devi GR. XIAP as target for therapeutic apoptosis in prostate cancer. both directly and indirectly. Mol Cell 2005;17:525–35. Drug News Perspect 2004;17:127–34. 34. Rizvi F, Heimann T, Herrnreiter A, O'Brien WJ. Mitochondrial dysfunc- 38. Foster PA, Ho YT, Newman SP, Leese MP, Potter BV, Reed MJ, tion links ceramide activated HRK expression and cell death. PLoS et al. STX140 and STX641 cause apoptosis via the intrinsic mito- ONE 2011;6:e18137. chondrial pathway and down-regulate survivin and XIAP expression 35. Chen L, Willis SN, Wei A, Smith BJ, Fletcher JI, Hinds MG, et al. Differential in ovarian and prostate cancer cells. Anticancer Res 2009;29: targeting of prosurvival Bcl-2 proteins by their BH3-only ligands allows 3751–7. complementary apoptotic function. Mol Cell 2005;17:393–403. 39. Perez-Stable C. 2-Methoxyestradiol and paclitaxel have similar effects 36. Krajewska M, Krajewski S, Epstein JI, Shabaik A, Sauvageot J, on the cell cycle and induction of apoptosis in prostate cancer cells. Song K, et al. Immunohistochemical analysis of bcl-2, bax, bcl-X, Cancer Lett 2006;231:49–64.

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1059

Hrk Mediates 2-Methoxyestradiol−Induced Mitochondrial Apoptotic Signaling in Prostate Cancer Cells

Inik Chang, Shahana Majid, Sharanjot Saini, et al.

Mol Cancer Ther 2013;12:1049-1059. Published OnlineFirst April 11, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-12-1187

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2013/04/11/1535-7163.MCT-12-1187.DC1

Cited articles This article cites 39 articles, 12 of which you can access for free at: http://mct.aacrjournals.org/content/12/6/1049.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/12/6/1049.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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/12/6/1049. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.