Selective Inhibition of Histone Deacetylase 2 Silences Progesterone Receptor–Mediated Signaling

Home , Histone deacetylase, Histone deacetylase 2

Research Article

Elona Bic¸aku, Douglas C. Marchion, Morgen L. Schmitt, and Pamela N. Mu¨nster

Department of Interdisciplinary Oncology, Experimental Therapeutics Program, H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida

Abstract and LBH589, convey tamoxifen sensitivity to the ER-negative cell line, MDA-231, by inducing the release of HDAC1 from the ERa Several histone deacetylases (HDAC) are involved in estrogen a receptor (ER)–mediated gene transactivation, and HDAC promoter, thus restoring ER expression (3, 4) or the activation of ERh, as suggested by others (5). In ER-positive cell lines, HDAC inhibitors have been reported to restore sensitivity to a antihormonal therapy. The modulation of ER is the most inhibitors have been shown to repress ER expression (6, 7) and sensitize cells to tamoxifen (5, 8), which may involve the up- promising approach to ER-expressing breast cancers. Recent h studies further suggest a critical role of the progesterone regulation or translocation of ER (9, 10). Although these studies receptor (PR) on ER signaling. Although HDAC inhibitors suggest a benefit of adding an HDAC inhibitor to antiestrogen modulate ER, little is known about their effects on PR. We therapy, the differential effects on ERsuggest that the modulation of ERmay not entirely explain the synergy between HDAC evaluated the roles of specific HDAC isoenzymes and their inhibition on both ER and PR signaling and their importance inhibitors and antiestrogen therapy. in response to endocrine therapy. The roles of individual About 70% of women with breast cancer present with tumors HDAC isoenzymes on ER and PR expression and their that express ERand the majority of those coexpress progesterone functions were evaluated by depletion of select HDAC enzymes receptor (PR; ref. 11). Most of the currently used antihormonal using siRNA or pharmacologic inhibition. Cotreatment of therapies predominantly target ER. Antiprogestins and other breast cancer cell lines with HDAC inhibitors and the modalities targeting PRare now rarely used, not due to inefficacy antiestrogen, tamoxifen, resulted in synergistic antitumor but due to undesirable side effects such as profound weight gain activity with simultaneous depletion of both ER and PR. (12, 13). Although it seems that the coexpression of PRin patients, Selective inhibition of HDAC2, but not HDAC1 or HDAC6, whose tumors express ER, predicts a better outcome, the role of PRas a predictive factor for antihormone therapy remains under was sufficient to potentiate tamoxifen-induced apoptosis in ER/PR-positive cells. Depletion of HDAC1 and HDAC6 was debate (14–16). Furthermore, there are very few reported associated with down-regulation of ER but not PR. Only the modalities to address the potential benefits of simultaneous selective depletion of HDAC2 siRNA down-regulated both ER down-regulation of ERand PR,or the predictive value of the and PR expression, and was sufficient to potentiate tamoxifen. pharmacologic modulation of PRin addition to ER.Although PR Selective depletion of HDAC2 resulted in simultaneous levels are induced by ER, PR may become hormone independent depletion of ER and PR, and potentiated the effects of upon transformation to a more aggressive phenotype (17–19). antihormonal therapy in ER-positive cells. A more effective The role of PRas a target for endocrine therapy is still under debate pharmacologic inhibition of HDAC2 and evaluation of HDAC2 and very little is known about the interaction between HDAC and PR as therapeutic targets or as predictive markers in enzymes and PR. hormonal therapy may be considered when combining HDAC In this study, we show that the selective inhibition of HDAC2 by inhibitors and hormonal therapy. [Cancer Res 2008;68(5):1513–9] siRNAresulted in simultaneous depletion of ERas well as PRand potentiated the antitumor effects of tamoxifen in ER-positive cell lines. Introduction Histone acetyl-transferases and histone deacetylases (HDAC) Materials and Methods play a crucial role in gene regulation. Specific interference with histone acetylation and methylation by drugs has shown to be Chemicals and antibodies. Valproic acid was purchased from Sigma- Aldrich. Vorinostat (suberoylanilide hydroxamic acid) was provided by Aton promising in the treatment and prevention of cancer. Several Pharma. 4-OH-Hydroxytamoxifen (tamoxifen) and MS-275 were purchased reports suggest a potential role of HDACs and their inhibitors in from Calbiochem. ERa, PR, and HER2-neu–directed antibodies were breast cancer. A direct interaction of HDAC1 and estrogen receptor purchased from Santa Cruz Biotechnology, Inc. HDAC1 and HDAC2 (ER) may be involved in the response to antiestrogen therapy antibodies were purchased from Upstate Biotechnology. HDAC6 was (1). Experimental overexpression of HDAC1 was associated with purchased from Santa Cruz Biotechnology, Inc. Glyceraldehyde-3-phosphate repression of ERin MCF-7 cells and reactivation of ERupon dehydrogenase antibody was purchased from Chemicon. exposure to the HDAC inhibitor, trichostatin A (TSA; ref. 2). Further Cell lines. All cell lines were purchased from the American Type Culture reports suggest that the hydroxamic acid HDAC inhibitors, TSA Collection and maintained in DMEM (Fisher Scientific) with 10% heat- inactivated fetal bovine serum (Sigma-Aldrich), 2 mmol/L glutamine, and 50 unit/mL penicillin and 50 Ag/mL streptomycin (Fisher Scientific). Cells

were incubated in a humidified atmosphere with 5% CO2 at 37jC. Requests for reprints: Pamela N. Mu¨nster, H. Lee Moffitt Cancer Center, 12901 Microarray. Alterations in gene expression induced by valproic acid and Magnolia Drive, SRC 22007, Tampa, FL 33612. Phone: 813-745-8948; Fax: 813-745-1984; vorinostat were evaluated by microarray using Affymetrix U133-plus E-mail: [email protected]. I2008 American Association for Cancer Research. 2.0 Genechips by standard protocols (H. Lee Moffitt Cancer Center and doi:10.1158/0008-5472.CAN-07-2822 Research Institute, Bill and Beverly Young National Functional Genomics www.aacrjournals.org 1513 Cancer Res 2008; 68: (5). March 1, 2008

Figure 1. HDAC inhibitors synergize with tamoxifen in breast cancer cell lines. A, differential expression of HDAC1, 2 and 6, ER, and HER2 in a limited breast cancer cell line panel by Western blot analysis. B, treatment of breast cancer cell lines with different levels of ERand HER2expression with tamoxifen ( tam)inthe presence of the HDAC inhibitor, valproic acid (VPA), resulted in increased cell death compared with tamoxifen alone by MTS assays and potentiation of tamoxifen-induced apoptosis in MCF-7 cells (C) treated for 48 h with three different classes of HDAC inhibitors (MS-275, 2 Amol/L; vorinostat, 1 Amol/L; or valproic acid, 2 mmol/L) either alone or in combination with tamoxifen (10 Amol/L).

Center), with duplicate samples for each condition run on separate 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfo- Genechips. Hybridization to Affymetrix chips was analyzed using Affymetrix phenyl)2H-tetrazolium, inner salt assay. Equal numbers of cells (1 Â 104 Microarray Suite 5.0 software. Signal intensity was scaled to an average cells per well for MCF-7, T-47D, and SKBr-3; 5 Â 103 cells per well for intensity of 500 before comparison analysis. The MAS 5.0 software uses a MDA-231) were cultured in 96-well plates, followed by a 24-hour incubation statistical algorithm to assess increases or decreases in mRNA abundance at 37jC, 5% CO2. Cells were then incubated with 2 mmol/L valproic acid, in a direct comparison between two samples (Statistical Algorithms 10 Amol/L tamoxifen, or the combination for 48 h and evaluated for description document).1 proliferation using the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy- SiRNA. RNA duplexes for HDAC1, HDAC2, and HDAC6 were purchased phenyl)-2-(4-sulfophenyl)2H-tetrazolium, inner salt (MTS) assay (Promega). from Ambion. Cells were transfected by electroporation using the Western blot analysis. Cells were harvested in tissue culture medium Nucleofector transfection kit according to manufacturer’s recommenda- using a Cell Lifter, washed in PBS, and solubilized using SDS lysis buffer tions (Amaxa). Cells (4 Â 106) were suspended in 0.1 mL electroporation [2% SDS, 10% glycerol, and 0.06 mol/L Tris (pH 6.8)]. The protein buffer containing 1 Amol/L siRNA and pulsed with a cell line–specific concentration of the SDS lysates was determined by the bicinchoninic acid protocol as described by the manufacturer. Pulsed cells were resuspended method (Pierce). Proteins (50 Ag) were separated on 8% SDS-PAGE gels and in 0.5 mL complete medium without antibiotics and incubated at 37jCfor transferred to nitrocellulose membranes. Membranes were blocked in tris- 15 min before experimentation. The Silencer Negative Control #2 siRNA buffered saline containing 0.05% Tween 20 (TBST), 5% nonfat milk; and (Ambion), a nonsense siRNA duplex, was used as a control. incubated with primary antibody in TBST, 5% nonfat milk, overnight at Apoptosis assay. Cells (5 Â 105) were treated with drugs (MS-275, 4jC. Membranes were washed thrice for 10 min with TBST and incubated 1 Amol/L; vorinostat, 1 Amol/L; valproic acid, 2 mmol/L; tamoxifen, with the appropriate secondary antibody in TBST, 5% nonfat milk for 10 Amol/L; or combination) for 48 h. Cells were harvested in medium using 60 min at room temperature. Antibody binding was visualized by chemi- a Cell Lifter (Fisher Scientific) and washed in PBS. Cell nuclei were stained luminescence on autoradiography film. with 0.5 Ag/mL bis-benzimide (Hoechst 33258). Apoptosis was scored by the presence of nuclear chromatin condensation and DNA fragmentation, Results and evaluated by fluorescence microscopy. Two hundred cells were counted per experiment in at least six different fields and scored for apoptosis HDAC inhibitors potentiate the antiestrogen, tamoxifen, (apoptotic nuclei/all nuclei Â100). Error bars depict SE. independent of class and structure. Synergistic or additive effects between HDAC inhibitors and tamoxifen have been reported in ER-positive and ER-negative cells (1–10, 20). However, the mechanism(s) by which HDAC inhibitors synergize with anti- 1 http://www.affymetrix.com/support/technical/whitepapers.affx estrogens in ER-positive and ER-negative breast cancer cells

Cancer Res 2008; 68: (5). March 1, 2008 1514 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 2008 American Association for Cancer Research. Effects of HDAC2 Inhibition on ER and PR remain under debate. In ER-positive cells, treatment with an HDAC inhibitors have been reported to down-regulate ER, whereas in inhibitor, such as TSA, vorinostat, LBH529, and MS-275, has been ER-negative cells, HDAC inhibitors, such as the hydroxamic acids, shown to reduce the expression of ERa while increasing the TSA, and LBH589 may up-regulate the expression of ER. To rule sensitivity of cancer cells to tamoxifen. It has further been proposed out a differential effect on ERexpression depending on the type of that select HDAC inhibitors may reactivate the expression of ERa HDAC inhibitor, the ER-positive MCF-7 cells were exposed to or ERh in ER-negative cell lines, thereby increasing the sensitivity of various HDAC inhibitors representing three structurally different these cells to an antiestrogen (4, 5, 9, 21). classes at clinically relevant concentrations, including the fatty The effects of HDAC inhibitors on tamoxifen-induced cell acids (valproic acid), the hydroxamic acids (vorinostat), and the growth arrest and cell death were evaluated in cells with different amides (MS-275; refs. 22–24). Tamoxifen was used as a positive levels of ERand HER2expression. Figure 1 A shows the expression control as its effects on ERprotein expression are well described. of ERa and HER2 in a panel of breast cancer cell lines as well as Treatment with all three HDAC inhibitors as well as tamoxifen the expression of several known relevant targets of HDAC resulted in depletion of ERprotein in the ER-positiveMCF-7 cells inhibitors: HDAC1, HDAC2, and HDAC6. The cell lines, SKBr-3 (Fig. 2A). The effects of HDAC inhibitors on ERwere then evaluated (HER2+, ERÀ), MCF-7 (HER2-, ER+), BT-474 (HER2+, ER+), in cell lines with various degrees of ERexpression using the HDAC MDA-231 (HER2À,ERÀ), MDA-361 (HER2+, ER+), and T-47D inhibitor, valproic acid. As seen in Fig. 2B, ERwas down-regulated (HER2À, ER+) were evaluated for cell survival after simultaneous in all examined ER-positive cell lines. In contrast, protein treatment with 0, 0.25, 0.5, 1, or 2 mmol/L of the HDAC inhibitor, expression of ERwas observed neither at baseline nor after valproic acid, and 10 Amol/L of the antiestrogen, tamoxifen, for treatment with valproic acid at 0.5 or 2 or 5 mmol/L, or with 48 h by MTS assays (Fig. 1B). An enhancement of the antipro- vorinostat 1 Amol/L in the ER-negative cell lines, MDA-231 and liferative effects of tamoxifen was observed in all examined cell SKBr-3 cells (Fig. 2B; data not shown). lines irrespective of their ERand HER2status, suggesting that ER To further delineate the effects of HDAC inhibitors on ER, HDAC and HER2 may not be sole contributors to the synergy. These inhibitor-induced changes in ERmRNAlevels were analyzed by effects were synergistic by isobologram analyses (data not shown). microarray analysis. Exposure to an HDAC inhibitor for 48 h was To determine whether the synergistic activity between an HDAC associated with a down-regulation of ERa mRNA levels compared inhibitor and an antiestrogen was dependent on the structural type with baseline in the ER-positive MCF-7 cell line. The reduction in of HDAC inhibitor, MCF-7 breast cancer cells were treated with ERa expression was seen with 1 Amol/L vorinostat [47% (46–48%)] 10 Amol/L tamoxifen for 48 h in the presence of the benzamide, as well as with 2 mmol/L valproic acid [60% (58–62%); Fig. 2C]. MS-275 (2 Amol/L); the hydroxamic acid, vorinostat (1 Amol/L); or In contrast, baseline levels of ERa mRNA expression were low in the short-chain fatty acid, valproic acid (2 mmol/L), and evaluated MDA-231, and at the concentrations used to elicit synergy with for apoptotic nuclei (Fig. 1C). A greater than additive or synergistic tamoxifen, no changes in ERa mRNA expression were observed interaction was seen with all of the HDAC inhibitors when added with either valproic acid or vorinostat (Fig. 2C). In addition, no to tamoxifen. Similar to the MTS assays, HDAC inhibitors significant changes in ERh mRNA expression were seen in either sensitized all evaluated cell lines to tamoxifen-induced apoptosis, MCF-7 or MDA-231 cells after exposure to the respective HDAC independent of their ERor HER2status. inhibitors (Fig. 2C). The effects of different classes of HDAC inhibitors on ER. The synergistic interaction between tamoxifen and HDAC Although synergistic effects of HDAC inhibitors and tamoxifen inhibitors involves down-regulation of PR. The synergistic have been described in ER-positive and ER-negative cells (1–10, 20), interaction between HDAC inhibitors and tamoxifen in ER-positive differential effects on ERwere noted. In ER-positivecells, HDAC as well as ER-negative tumor cells in the absence of an effect of ER

Figure 2. HDAC inhibitors down-regulate the expression of ERa in ER-positive cell lines. A, treatment of the ER-positive cell line MCF-7 with the structurally distinct HDAC inhibitors, valproic acid, vorinostat (Vor), and MS-275 for 48 h led to reduced ERa protein levels. As a positive control, cells were also treated with tamoxifen. B, effects of valproic acid (2 mmol/L) on ERa protein expression in ER-positive cells (MCF-7 and T47D) and ER-negative cells (MD-231 and SKBr-3). C, mean fluorescence signal intensity (MAS5 generated) of ERa and ERh mRNA in two replicates of MCF-7 and MDA-231 cells after exposure to either vehicle (control), 1Amol/L vorinostat, or 2 mmol/L valproic acid for 48 h. Columns, mean; bars, SE. Analyses by Student’s t test suggested a significant change in the expression of ERa after treatment with vorinostat (P = 0.005) and valproic acid (P = 0.015) in MCF-7 cells but no significant changes on ERa in the MDA-231 cells or on ERh in either cell line. www.aacrjournals.org 1515 Cancer Res 2008; 68: (5). March 1, 2008

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 2008 American Association for Cancer Research. Cancer Research expression in the ER-negative tumor cells suggested the possible vorinostat affect HDAC6 at clinically relevant concentrations. involvement of an additional or an alternative effect on a The selective depletion of HDAC1, 2, and 6 were all associated with tamoxifen-regulated protein such as the PR. Exposure of cells to the reduced expression of ERin T47D cells (Fig. 4 A). In contrast, an HDAC inhibitor was associated with a down-regulation of PR siRNA depletion of HDAC2 resulted in the down-regulation of ER (Fig. 3). The HDAC inhibitor effects on PRwere not linked to a as well as both isoforms of PR, PR-A, and PR-B (Fig. 4A). The effects certain class of HDAC inhibitors but occurred with both a fatty acid of select HDAC depletion on tamoxifen-induced apoptosis and (valproic acid) and a hydroxamic acid (vorinostat; Fig. 3A and B). growth were evaluated after selective depletion of each, HDAC1, The effects of valproic acid and vorinostat on PRwere observed HDAC2, or HDAC6. In the absence of tamoxifen, no significant at the gene transcript level as well as at the protein level. effects on growth by MTS assay or apoptosis were observed with Furthermore, the simultaneous exposure of cells to valproic acid depletion of either three HDAC enzymes (data not shown; Fig. 4B). and tamoxifen was associated with a more pronounced effect on However, although the effects of HDAC1 and HDAC6 depletion PR. As seen in Fig. 3C, the combination of tamoxifen and valproic showed no additional effects, the selective depletion of HDAC2 was acid affected both isoforms of PR, PR-A, and PR-B (25). In contrast, associated with a significant enhancement of tamoxifen-induced no effects on PRmRNAwere observed in the MDA-231 cells; apoptosis (Fig. 4B). The effects of HDAC2 depletion by siRNA on however, these cells showed minimal expression of PR(Fig. 3 A). ERand PRwere similar to those seen with 2 mmol/L valproic acid Furthermore, compared with the protein expression shown for (Fig. 4C). No effects of valproic acid on HDAC1 and HDAC2 protein T47D and MCF-7, protein expression in SKBR-3 and MDA-231 was expression were observed (Fig. 4C). Selective depletion of HDAC1 only minimally detectable (Fig. 3B). These cells are typically and HDAC2 did not result in potentiation of tamoxifen in the considered PR-negative. ER- and PR-negative cell line, MDA-231 (Fig. 4D). Expression of Cooperative down-regulation of ER and PR is mediated ERand PRin MDA-231 cells are considered absent, and treatment through HDAC2. Although there are several HDAC inhibitors of these cells with an HDAC inhibitor did not affect ERor PR under development, there is currently no consensus on whether mRNA or protein expression (Figs. 2B and C;3A and B). the development of isotype-specific HDAC inhibitors will result in more effective drugs for the treatment of any one particular or several tumor types, or whether the expression of select HDAC Discussion enzymes may serve as a prognostic or predictive factor. To The potential role of HDAC inhibitors to enhance the efficacy determine a potential role of select HDAC enzymes in the synergy of tamoxifen by modulation of ERhas been described previously between an HDAC inhibitor and tamoxifen, the known relevant (2–10). In this study, we present the putative role of the PRin the targets of HDAC inhibitors, HDAC1, HDAC2, and HDAC6 were synergistic interaction between HDAC inhibitors and antiestrogen selectively depleted using siRNA. therapy in ER- or PR-expressing cells. Similar to the findings HDAC6 is not a target of valproic acid when used at physiologic reported by others, we found that the HDAC inhibitor effects on concentrations (26), whereas the hydroxamic acids such as tamoxifen-induced apoptosis (Fig. 1) were associated with a

Figure 3. Synergistic modulation of PRlevels by HDAC inhibitors and tamoxifen. A, mean fluorescence signal intensity (MAS5 generated) indicating relative mRNA expression of the PRin MCF-7 and MDA-231 cells by microarray after a 48-h exposure to vehicle ( control), 1 Amol/L vorinostat, or 2 mmol/L valproic acid. Two biological replicates for each condition were run on two separate Genechips. Columns, mean; bars, SE. B, Western blot analysis showing PRprotein levels in MCF-7 (PR-B), T47D (PR-A and PR-B), MDA-231 (PR-B), and SKBr-3 (PR-B) cells in the presence and absence of 2 mmol/L valproic acid or 1 Amol/L vorinostat for 48 h. Top, relative expression of PRin these cell lines. C, enhanced depletion of PRprotein levels by valproic acid was seen in cells cotreated with tamoxifen (10 Amol/L).

Cancer Res 2008; 68: (5). March 1, 2008 1516 www.aacrjournals.org

Figure 4. Role of HDAC enzymes in the synergy between valproic acid and tamoxifen in ER-positive cell lines. A, selective depletion of HDAC2 but not HDAC1 or HDAC6 by siRNA in T47D cells resulted in reduced expression of both the ERa and the PR, and sensitized cells to apoptosis induced by 10 Amol/L tamoxifen (B). C, modulation of ERa and PRexpression after siRNA depletion of HDAC2 was comparable with the effects on ERand PRby a 48-h valproic acid (2 mmol/L) treatment of MCF-7 cells. No effects of valproic acid on HDAC1 and HDAC2 were seen. D, HDAC2 depletion by siRNA potentiated the apoptosis induced by 10 Amol/L tamoxifen in MCF-7 and T47D cells but not MDA-231 cells. No sensitization to tamoxifen was observed when cells were depleted of HDAC1. A nonsilencing siRNA was used as control.

depletion of ERmRNAand protein expression in ER-positive increased HDAC1 mRNA and protein expression were linked to cells (Fig. 2). However, we found minimal or no changes in ER better outcome, whereas HDAC3 expression did not seem to affect expression in ER-negative cells at concentrations of valproic acid overall or disease-free survival (27, 28). However, the reports were or vorinostat used to elicit a synergistic interaction with tamoxifen, equivocal in whether the expression of HDAC1 mRNA was an suggesting that at least for the ER-negative cell lines, modulation independent prognostic indicator or linked to other features (27). of ERalone may not completely explain the synergy between an Increased expression of HDAC6 has been associated with improved HDAC inhibitor and tamoxifen. Our data suggest that in ER- disease-free and overall survival in patients with hormone-sensitive positive cell lines, exposure to an HDAC inhibitor was associated breast tumors treated with tamoxifen (20). A direct interaction of with a depletion of PRexpression. The effects on PRwere further HDAC1 and the ERhas been implicated in the response to enhanced in the presence of tamoxifen. HDAC inhibitors were antiestrogen therapy (1). Genetic overexpression of HDAC1 in associated with a depletion of both ERand PRmRNAas well as MCF-7 cells has been associated with repression of ERa, with protein levels (Fig. 3). Valproic acid affected both PRisoforms in subsequent reactivation upon exposure to the HDAC inhibitor, TSA T47-D cells. (2). Other investigators have proposed that the hydroxamic acid Our data further suggested that the effects of valproic acid on ER HDAC inhibitors, TSA and LBH589, may convey sensitivity to and PRcould also be elicited by the selective down-regulation of tamoxifen in the ER-negative cell line, MDA-231, by inducing HDAC2 (Fig. 4). Depletion of HDAC1, HDAC2, or HDAC6 by siRNA release of HDAC1 from the ERa promoter, thus, restoring ERa resulted in depletion of ER, but only the selective depletion of expression (3, 4). Although HDAC2 has been associated with HDAC2 resulted in reduced expression of both ERand PR.The poorer outcomes in colon cancer and may be overexpressed in APC selective depletion of HDAC2 but not HDAC1 or HDAC6 was gene–related colon polyps, the role of HDAC2 in breast cancer has further associated with enhanced tamoxifen-induced apoptosis, not been determined (29). similar to the effects seen with valproic acid in cells expressing In contrast to reports by other investigators, the concentrations either ERor PR(Fig. 4). Several investigators have reported a link and the agents used to potentiate tamoxifen in our studies, valproic between HDAC enzymes and hormonal regulation of breast cancer. acid and vorinostat, were not associated with an increase in ER HDAC1 mRNA levels by reverse transcription-PCR as well as protein expression as reported for the hydroxamic acids, TSA and HDAC1 and HDAC3 protein expression by immunohistochemistry LBH589, in ER-negative tumor cells (3, 4). The induction of ER were statistically associated with smaller, ER- and PR-positive reported in these studies may be due to a more potent inhibition tumors, whereas HDAC1 mRNA, but not HDAC1 protein expres- of HDAC6 induced by the use of these compounds (TSA and sion, was linked to node-negative tumors (27, 28). In addition, LBH589). HDAC inhibitor treatment resulted in depletion of ERin www.aacrjournals.org 1517 Cancer Res 2008; 68: (5). March 1, 2008

ER-positive cell lines; however, although the selective depletion of HDAC inhibitors on PRmay open several new venues for drug HDAC1 and HDAC6 was associated with a decrease in ERprotein development. expression, this was not sufficient to potentiate tamoxifen. The Despite the link between HDAC enzymes and hormone receptor depletion of HDAC2 by siRNA was sufficient to reduce ER and PR expression, to date, no studies have determined the relevance of expression and to potentiate the apoptosis induced by tamoxifen individual HDAC enzymes as targets for HDAC inhibitors. Our data to a similar degree as valproic acid and vorinostat. These findings suggest that the selective depletion of HDAC2 by siRNA potentiated suggest that for the interaction between HDAC inhibitors and the effects of tamoxifen, whereas the selective depletion of HDAC1 antiestrogens, HDAC2 may be the relevant target. Selective and HDAC6 did not affect the apoptosis or the growth arrest depletion of HDAC1, HDAC2, or HDAC6 did not potentiate the induced by tamoxifen in ER- and/or PR-expressing breast cancer effects of tamoxifen in the ER- and PR-negative cell line, MDA231. cells. The effects on apoptosis were similar to those seen with Furthermore, HDAC inhibitors did not affect either ERor PRmRNA 2 mmol/L valproic acid. In contrast to reports by other or protein expression in these cells, which are commonly investigators, no effects of valproic acid on HDAC2 protein considered ER- and PR-negative. The synergy seen with tamoxifen expression were seen at the concentrations used in the MCF-7 and HDAC inhibitors in these cells are likely due to a different and T47D cells (30, 31). mechanism than the synergy seen in the ER- or PR-positive cell lines In summary, although HDAC inhibitors seem to have antitumor and may require the inhibition of another select HDAC enzyme or activity as single agents in hematologic malignancies and select the simultaneous interference of several HDAC enzymes. solid tumors, their efficacy in breast cancer may be limited. The role of HDAC2 rather than HDAC1 or HDAC6 is further However, HDAC inhibitors may potentiate the effects of tamoxifen supported by the findings that not only the hydroxamic acids but in ER- or PR-positive breast cancers by modulation of not only ER also the more class I–specific HDAC inhibitors such as valproic but also PRsignaling mediated by HDAC2. Given the emerging acid and MS-275 potentiated tamoxifen. Although vorinostat has development of HDAC inhibitors and the still controversial role of been associated with inhibition of HDAC6 at the concentrations PRas a predictive marker of response, these studies may provide used for the presented studies, valproic acid requires supra- insight into the mechanism by which HDAC inhibitors potentiate therapeutic doses (>15 mmol/L) for HDAC6 inhibition (26). This is antiestrogen therapy and may provide guidance in the further further supported by the absence of a-tubulin acetylation seen with development of more selective means to inhibit HDAC2. valproic acid in this study (data not shown). Our data suggest that the effects of HDAC inhibitors on the potentiation of tamoxifen- Acknowledgments induced apoptosis are predominantly mediated through HDAC2 Received 7/24/2007; revised 11/9/2007; accepted 1/3/2008. and involve the presence and modulation of PR. The role of PR in Grant support: Department of Defense Concept Award (W81XWH-06-1-0622). breast cancer as a prognostic and predictive marker remains under The costs of publication of this article were defrayed in part by the payment of page debate, and therapies targeting PRare not commonly used. Several charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. studies suggest that tumors that express both ERand PRmay We thank Anita Bruce and Connie Schmitt for editorial assistance, and Bill and respond better to antiestrogen therapy; hence, the effects of Beverly Young National Functional Genomics Center for their contribution.

References 8. Hirokawa Y, Arnold M, Nakajima H, Zalcberg J, Maruta in premenopausal breast cancer patients. Clin Cancer H. Signal therapy of breast cancers by the HDAC Res 2006;12:4614–8. 1. Varshochi R, Halim F, Sunters A, et al. ICI182,780 inhibitor FK228 that blocks the activation of PAK1 and 16. Dowsett M, Houghton J, Iden C, et al. Benefit from induces p21Waf1 gene transcription through releasing abrogates the tamoxifen-resistance. Cancer Biol Ther adjuvant tamoxifen therapy in primary breast cancer a histone deacetylase 1 and estrogen receptor from Sp1 2005;4:956–60. patients according oestrogen receptor, progesterone sites to induce cell cycle arrest in MCF-7 breast cancer 9. Duong V, Licznar A, Margueron R, et al. ERa and ERh receptor, EGF receptor and HER2 status. Ann Oncol cell line. J Biol Chem 2005;280:3185–96. expression and transcriptional activity are differentially 2006;17:818–26. 2. Kawai H, Li H, Avraham S, Jiang S, Avraham HK. regulated by HDAC inhibitors. Oncogene 2006;25: 17. Lin VC, Eng AS, Hen NE, Ng EH, Chowdhury SH. Overexpression of histone deacetylase HDAC1 modu- 1799–806. Effect of progesterone on the invasive properties and lates breast cancer progression by negative regulation of 10. Jang ER, Lim SJ, Lee ES, et al. The histone deacetylase tumor growth of progesterone receptor-transfected estrogen receptor a. Int J Cancer 2003;107:353–8. inhibitor trichostatin A sensitizes estrogen receptor a- breast cancer cells MDA-MB-231. Clin Cancer Res 3. Yang X, Phillips DL, Ferguson AT, Nelson WG, Herman negative breast cancer cells to tamoxifen. Oncogene 2001;7:2880–6. JG, Davidson NE. Synergistic activation of functional 2004;23:1724–36. 18. McGowan EM, Saad S, Bendall LJ, Bradstock KF, estrogen receptor (ER)- a by DNA methyltransferase 11. Pichon MF, Broet P, Magdelenat H, et al. Prognostic Clarke CL. Effect of progesterone receptor a predom- and histone deacetylase inhibition in human ER-a- value of steroid receptors after long-term follow-up of inance on breast cancer cell migration into bone negative breast cancer cells. Cancer Res 2001;61:7025–9. 2257 operable breast cancers. Br J Cancer 1996;73: marrow fibroblasts. Breast Cancer Res Treat 2004;83: 4. Zhou Q, Atadja P, Davidson NE. Histone deacetylase 1545–51. 211–20. inhibitor LBH589 reactivates silenced estrogen receptor 12. Kaufmann M, Bajetta E, Dirix LY, et al. Exemestane 19. Kato S, Pinto M, Carvajal A, et al. Progesterone a (ER) gene expression without loss of DNA hyper- is superior to megestrol acetate after tamoxifen failure increases tissue factor gene expression, procoagulant methylation. Cancer Biol Ther 2007;6:64–9. in postmenopausal women with advanced breast activity, and invasion in the breast cancer cell line ZR- 5. Hodges-Gallagher L, Valentine CD, Bader SE, Kushner cancer: results of a phase III randomized double- 75–1. J Clin Endocrinol Metab 2005;90:1181–8. PJ. Inhibition of histone deacetylase enhances the anti- blind trial. The Exemestane Study Group. J Clin Oncol 20. Saji S, Kawakami M, Hayashi S, et al. Significance of proliferative action of antiestrogens on breast cancer 2000;18:1399–411. HDAC6 regulation via estrogen signaling for cell cells and blocks tamoxifen-induced proliferation of 13. Stuart NS, Warwick J, Blackledge GR, et al. A motility and prognosis in estrogen receptor-positive uterine cells. Breast Cancer Res Treat 2006;105:297–309. randomised phase III cross-over study of tamoxifen breast cancer. Oncogene 2005;24:4531–9. 6. Reid G, Metivier R, Lin CY, et al. Multiple mechanisms versus megestrol acetate in advanced and recurrent 21. Sharma D, Saxena NK, Davidson NE, Vertino PM. induce transcriptional silencing of a subset of genes, breast cancer. Eur J Cancer 1996;32A:1888–92. Restoration of tamoxifen sensitivity in estrogen recep- including oestrogen receptor a, in response to deacety- 14. Yu KD, Liu GY, Di GH, et al. Progesterone receptor tor-negative breast cancer cells: tamoxifen-bound reac- lase inhibition by valproic acid and trichostatin A. status provides predictive value for adjuvant endocrine tivated ERrecruits distinctive corepressor complexes. Oncogene 2005;24:4894–907. therapy in older estrogen receptor-positive breast Cancer Res 2006;66:6370–8. 7. Rocha W, Sanchez R, Deschenes J, et al. Opposite cancer patients. Breast 2007;16:307–15. 22. Kelly WK, O’Connor OA, Krug LM, et al. Phase I study effects of histone deacetylase inhibitors on glucocorti- 15. Stendahl M, Ryden L, Nordenskjold B, Jonsson PE, of an oral histone deacetylase inhibitor, suberoylanilide coid and estrogen signaling in human endometrial Landberg G, Jirstrom K. High progesterone receptor hydroxamic acid, in patients with advanced cancer. J Ishikawa cells. Mol Pharmacol 2005;68:1852–62. expression correlates to the effect of adjuvant tamoxifen Clin Oncol 2005;23:3923–31.

Cancer Res 2008; 68: (5). March 1, 2008 1518 www.aacrjournals.org

23. Kelly WK, Richon VM, O’Connor O, et al. Phase I progesterone receptor forms A and B. EMBO J 1990; 29. Zhu P, Martin E, Mengwasser J, Schlag P, Janssen KP, clinical trial of histone deacetylase inhibitor: suberoy- 9:1603–14. Gottlicher M. Induction of HDAC2 expression upon loss lanilide hydroxamic acid administered intravenously. 26. Gurvich N, Tsygankova OM, Meinkoth JL, Klein PS. of APC in colorectal tumorigenesis. Cancer Cell 2004;5: Clin Cancer Res 2003;9:3578–88. Histone deacetylase is a target of valproic acid-mediated 455–63. 24. Munster P, Marchion D, Bicaku E, et al. Phase I trial cellular differentiation. Cancer Res 2004;64:1079–86. 30. Kramer OH, Zhu P, Ostendorff HP, et al. The histone of histone deacetylase inhibition by valproic acid 27. Zhang Z, Yamashita H, Toyama T, et al. Quantitation deacetylase inhibitor valproic acid selectively induces followed by the topoisomerase II inhibitor epirubicin of HDAC1 mRNA Expression in Invasive Carcinoma of proteasomal degradation of HDAC2. EMBO J 2003;22: in advanced solid tumors: a clinical and translational the Breast. Breast Cancer Res Treat 2005;94:11–6. 3411–20. study. J Clin Oncol 2007;25:1979–85. 28. Krusche CA, Wulfing P, Kersting C, et al. Histone 31. Atmaca A, Al-Batran SE, Maurer A, et al. Valproic 25. Kastner P, Krust A, Turcotte B, et al. Two distinct deacetylase-1 and -3 protein expression in human breast acid (VPA) in patients with refractory advanced cancer: estrogen-regulated promoters generate transcripts cancer: a tissue microarray analysis. Breast Cancer Res a dose escalating phase I clinical trial. Br J Cancer 2007; encoding the two functionally different human Treat 2005;90:15–23. 97:177–82.

www.aacrjournals.org 1519 Cancer Res 2008; 68: (5). March 1, 2008

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 2008 American Association for Cancer Research. Selective Inhibition of Histone Deacetylase 2 Silences Progesterone Receptor−Mediated Signaling

Elona Biçaku, Douglas C. Marchion, Morgen L. Schmitt, et al.

Cancer Res 2008;68:1513-1519.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/68/5/1513

Cited articles This article cites 31 articles, 12 of which you can access for free at: http://cancerres.aacrjournals.org/content/68/5/1513.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/68/5/1513.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://cancerres.aacrjournals.org/content/68/5/1513. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.