Novel Selective Estrogen Mimics for the Treatment of Tamoxifen-Resistant Breast Cancer

Published OnlineFirst September 9, 2014; DOI: 10.1158/1535-7163.MCT-14-0319

Molecular Cancer Small Molecule Therapeutics Therapeutics

Mary Ellen Molloy1, Bethany E. Perez White1, Teshome Gherezghiher2, Bradley T. Michalsen2, Rui Xiong2, Hitisha Patel2, Huiping Zhao1, Philipp Y. Maximov3, V. Craig Jordan3, Gregory R.J. Thatcher2, and Debra A. Tonetti1

Abstract Endocrine-resistant breast cancer is a major clinical obstacle. The use of 17b-estradiol (E2) has reemerged as a potential treatment option following exhaustive use of tamoxifen or aromatase inhibitors, although side effects have hindered its clinical usage. Protein kinase C alpha (PKCa) expression was shown to be a predictor of disease outcome for patients receiving endocrine therapy and may predict a positive response to an estrogenic treatment. Here, we have investigated the use of novel benzothiophene selective estrogen mimics (SEM) as an alternative to E2 for the treatment of tamoxifen-resistant breast cancer. Following in vitro characterization of SEMs, a panel of clinically relevant PKCa-expressing, tamoxifen-resistant models were used to investigate the antitumor effects of these compounds. SEM treatment resulted in growth inhibition and apoptosis of tamoxifen-resistant cell lines in vitro. In vivo SEM treatment induced tumor regression of tamoxifen-resistant T47D:A18/PKCa and T47D:A18-TAM1 tumor models. T47D:A18/PKCa tumor regression was accompanied by translocation of estrogen receptor (ER) a to extranuclear sites, possibly deﬁning a mechanism through which these SEMs initiate tumor regression. SEM treatment did not stimulate growth of E2-dependent T47D:A18/neo tumors. In addition, unlike E2 or tamoxifen, treatment with SEMs did not stimulate uterine weight gain. These ﬁndings suggest the further development of SEMs as a feasible therapeutic strategy for the treatment of endocrine-resistant breast cancer without the side effects associated with E2. Mol Cancer Ther; 13(11); 2515–26. 2014 AACR.

Introduction Protein kinase C alpha (PKCa) belongs to a family of a The selective estrogen receptor modulator (SERM) serine/threonine protein kinases (1, 2). PKC expression tamoxifen is the most widely prescribed endocrine ther- in breast cancer is associated with tamoxifen resistance, apy for the treatment and prevention of breast cancer in poor patient survival, and breast cancer aggressiveness (3–5). To further substantiate these clinical observations, premenopausal women, whereas aromatase inhibitors are a the drug of choice for postmenopausal women. De novo or we reported that ectopic overexpression of PKC in acquired resistance to these endocrine therapies limits the T47D:A18 breast cancer cell line resulted in a hor- their clinical effectiveness, leading to disease progression. mone-independent, tamoxifen-resistant phenotype (6). Interestingly, these tamoxifen-resistant T47D:A18/PKCa As such, there is a clinical need for therapeutic alternatives b in vivo for women who no longer respond to conventional endo- tumors are growth inhibited by 17 -estradiol (E2) crine therapies. (7). Yao and colleagues describe an MCF-7 tumor model in which long-term exposure (5 years) to tamoxifen led to an E2-inhibited phenotype (8) and elevated PKCa expres- 1Department of Biopharmaceutical Sciences, College of Pharmacy, Uni- 2 sion (7). Together these studies provide important ther- versity of Illinois at Chicago, Chicago, Illinois. Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, University of Illinois apeutic implications, suggesting that PKCa expression at Chicago, Chicago, Illinois. 3Department of Oncology, Georgetown may predict both resistance to conventional endocrine University, Lombardi Comprehensive Cancer Center, Washington, District of Columbia. therapies and a predicted response to E2 or estrogen-like compounds. Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). Before the introduction of tamoxifen, patients with breast cancer were treated with high-dose E2 or diethys- Current address for B.E.P. White: Northwestern University, Department of Dermatology, Chicago, IL. tilbesterol. Although, similar response rates were observed (9, 10), tamoxifen treatment became the mainstay Corresponding Author: Debra A. Tonetti, Department of Biopharmaceu- tical Sciences, College of Pharmacy, University of Illinois at Chicago, 833 S. due to a lower incidence of side effects such as nausea, Wood St. (m/c 865), Chicago, IL 60612. Phone: 312-413-1169; Fax: 312- emesis, and edema. Because treatment with E2, diethys- – 996 1698; E-mail: [email protected] tilbesterol, and tamoxifen are now all associated with side doi: 10.1158/1535-7163.MCT-14-0319 effects, including increased risk of thromboembolic dis- 2014 American Association for Cancer Research. orders and unwanted agonist-driven uterine growth, we

www.aacrjournals.org 2515

Molloy et al.

sought an alternative treatment strategy that would have levels compared with their tamoxifen-sensitive counter- therapeutic efficacy in the tamoxifen-resistant setting. We parts (Supplementary Fig. S1). Before treatment, cell lines have previously reported that tamoxifen-resistant T47D: were cultured in E2-depleted media for 3 days. Cell lines A18/PKCa tumors regress upon treatment with both E2 were routinely tested for mycoplasma contamination and the benzothiophene SERM raloxifene, although the (MycoAlert Mycoplasma Detection Kit, Lonza Ltd. USA). effects of raloxifene did not persist after treatment with- All cell lines were authenticated in April 2014 using short drawal (11). Raloxifene has a favorable antiestrogenic tandem repeat (STR) and ATCC analysis (Promega Cor- profile in the uterus and has proven safety over 15 years poration Core Genomics Facility, UIC). of clinical use in postmenopausal osteoporosis and breast cancer chemoprevention. Synthesis and oral bioavailability of BTC In this study, we tested the in vivo effects of 2 novel Synthesis of both BTC and TTC-352 has been described benzothiophene SEMs, BTC, [2-(4-hydroxyphenyl)ben- previously (13). Dansyl derivatization of BTC was used to zo[b]thiophen-6-ol], and TTC-352, [3-(4-fluorophenyl)-2- increase limits of detection and quantitation for LC/MS- (4-hydroxyphenoxy)benzo[b]thiophen-6-ol], that in con- MS analysis of plasma samples (Supplementary Figs. S2 trast to raloxifene, acted as estrogen agonists in T47D:A18 and S3). Working solutions of BTC and internal standard and MCF-7 cells as reflected by increased cell proliferation (3-Br-BTC) were prepared by serial dilution of 1 mg/mL and estrogen response element (ERE)-luciferase reporter acetonitrile stocks. Calibration standards were prepared by activity. Both of these SEMs induced regression of tamox- spiking BTC or 3-Br-BTC (20 ng/mL) into blank mouse ifen-resistant, hormone-independent T47D:A18/PKCa plasma to give a final concentration range of 5 to 100 ng/mL and T47D:A18-TAM1 xenograft tumors in vivo, but (Supplementary Fig. S4). After addition of cold acetonitrile, remarkably, neither compound was able to support samples were kept at 4C for 2 hours, centrifuged at 10,000 the growth of hormone-dependent, tamoxifen-sensitive rpm for 15 minutes, and supernatants were concentrated T47D:A18/neo tumors. Of particular importance is that under N2 stream. Resulting residues were reconstituted in neither SEM caused any significant increase in the uterine 0.1 mL of 100 mmol/L sodium bicarbonate buffer (pH ¼ weights of treated mice. These data suggest that patients 10.5) and derivatized by addition of 0.1-mL dansyl chloride exhibiting endocrine resistance may respond to ben- (2 mg/mL in acetone) followed by incubation at 60Cfor5 zothiophene SEMs using PKCa as a predictive biomarker. minutes. After removal of solvent, residues were reconstituted in 0.25-mL acetonitrile/water (1:1, v/v) and analyzed Materials and Methods by LC/MS-MS. Reagents BTC was administered in ethanol using a vehicle of per DMSO, E2, and tamoxifen were obtained from Sigma- propylene glycol/carboxymethylcellulose (10 mg/kg os Aldrich. Raloxifene (Evista, Eli Lilly and Company) was ) to ovariectomized 4- to 6-week-old athymic mice n ¼ purchased from the University of Illinois at Chicago (UIC) (Harlan-Sprague-Dawley; 3). Blood samples were Hospital Pharmacy (Chicago, IL). Cell culture reagents collected in EDTA tubes at 20 minutes, 2 hours, and 6 were obtained from Life Technologies. Tissue culture hours after treatment. Plasma was separated from whole plastic ware was purchased from Becton Dickinson. The blood by centrifugation at 4 C. Before analysis, plasma following antibodies were used: rabbit polyclonal PKCa was spiked with internal standard and extracted 3 times (C-20; Santa Cruz Biotechnology), rabbit polyclonal ERa with cold acetonitrile. Recovery of analyte was measured (SP1; Lab Vision), mouse monoclonal b-actin (Sigma), by spiking known amounts of BTC into blank plasma anti-rabbit Alexa Fluor 488 (Life Technologies), and samples. Work up of plasma samples was identical to that anti-mouse Cy3 (Jackson ImmunoResearch Laboratories). described above for standard curve determination. LC/MS-MS analysis was performed using an API Cell culture conditions 3000 (Applied Biosystems) triple quadrupole mass spec- Stable transfectant cell lines T47D:A18/neo and T47D: trometer equipped with Agilent 1200 HPLC (Agilent A18/PKCa were produced and maintained as previously Technologies). Multiple reaction monitoring (MRM) for described (6) in RPMI-1640 (phenol red) supplemented the dissociations of m/z 709 171 and m/z 789 ;171 with 10% FBS containing G418 (500 mg/mL). T47D cells (loss of 5-dimethylaminonaphthalene) was optimized to were initially obtained from ATCC in 1996 and stored measure dansyl-BTC and dansylBr-BTC, respectively during early passage. MCF-7 cells were originally (Supplementary Fig. S5). Separation was performed obtained from the Michigan Cancer Foundation (Detroit, using a Hypersil BDS C18 (2.1 mm 30 mm; 3mm) MI) in 1992 and stored during early passage. MCF-7:5C column (Thermo Quest Corporation) at a flow rate of cells were maintained in E2-depleted media as previously 0.3 mL/min. The elution solvent consisted of water with described (12). The T47D:A18-TAM1 cell line was created 10% MeOH and 0.3% formic acid (A) and MeCN with by maintaining T47D:A18 breast cancer cells long-term 0.3% formic acid (B). The mobile phase was initially held (12 months) in 1 mmol/L of 4-hydroxytamoxifen in E2- at 10% B for 5 minutes, increased to 60% B over 1.5 depleted media. Single-cell clones were derived using the minutes, and then increased to 90% B over 15 minutes, limiting dilution method. All tamoxifen-resistant cell with dansyl-BTC and dansylBr-BTC eluting at 17.8 and lines retain estrogen receptor (ER)a expression at varying 19.7 minutes, respectively (Supplementary Fig. S5).

2516 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

SEMs for the Treatment of Tamoxifen-Resistant Breast Cancer

DNA growth assay metry according to the manufacturer’s instructions. In T47D:A18/neo, T47D:A18/PKCa, and T47D:A18- brief, MCF-7:5C cells were treated with control and test TAM1 cells were maintained in E2-depleted media 3 days compounds for 6 days with treatment renewal on day 3. before plating in 24-well plates (15,000 cells per well). Cells (1 105) were suspended in 1 Annexin V binding Medium containing compound was added the following buffer and were stained with AlexaFluor 488 Annexin V day, and total DNA was determined by incubating cells and propidium iodide (PI). The cells were then analyzed by with Hoechst 33342 cell-permeable dye for 1 hour and FACS using a Gallios Flow Cytometer (Beckman Coulter). reading fluorescence at excitation 355 nm/emission 460 nm on a Perkin Elmer Victor3 V plate reader (Waltham). Animal experiments Treatment medium was changed every 2 to 3 days. T47D:A18/PKCa, T47D:A18/neo, and T47D:A18- TAM1 tumors were established as previously described Proliferation assay (7). E2 was administered via silastic capsules (1.0 cm) Following 3 days of growth in E2-depleted media, 2 105 implanted subcutaneously between the scapulae, produc- cells were seeded into T25 tissue culture flasks. The following a mean serum E2 level of 379.5 pg/mL (11, 15). BTC ing day (day 1) treatment medium was added. Cells were and TTC-352 were administered per os at a dose of 1.5 mg/ counted on day 9 and medium was changed every 3 days. animal daily for 2 weeks as previously described for other SERMs (7). Raloxifene was administered per os at a dose of Western blotting 1.5 mg/animal daily for 2 weeks. Tumor cross-sectional Whole-cell extracts of cultured cells were prepared in area was determined weekly using Vernier calipers and lysis buffer (200 mmol/L Tris, 1% Triton X-100, 5 mmol/L calculated using the formula: length/2 width/2 p. EDTA) with protease and phosphatase inhibitor cocktails Mean tumor area was plotted against time (in weeks) to (1:50, both from Sigma-Aldrich) after scraping from the monitor tumor growth. The mice were sacrificed by CO2 culture plates. Protein concentration was measured using inhalation and cervical dislocation, and tumors and uteri the Bradford method (Bio-Rad). Proteins were separated were excised, cleaned of connective tissue, and immedi- under denaturing conditions and blotted onto nitrocellu- ately weighed. The Animal Care and Use Committee of lose membrane (Bio-Rad) using a wet transfer system the UIC approved all of the procedures involving animals. (Bio-Rad). Images of blots were acquired on a Bio-Rad ChemiDoc System following incubation with SuperSignal Tumor immunofluorescence and confocal West Dura luminol solution (Thermo Fisher Scientific). microscopy Protein bands were quantified using densitometry mea- Tumor sections (4 mm) were cut from paraffin blocks sured in Adobe Photoshop CS4 and normalized to b-actin. and prepared for immunofluorescent staining by depar- affinization and rehydration as previously described (11). Transient transfection and luciferase assays Briefly, antigen retrieval was performed by incubating Cells were transiently transfected by electroporation slides in Tris-EDTA (pH ¼ 9.0) buffer. Slides were blocked with 5 mg ERE-tk-Luc plasmid containing the luciferase with antibody diluent (DAKO) followed by primary anti- reporter gene controlled by a triplet vitellogenin consen- bodies at 1:100 in antibody diluent for 1 hour and then sus ERE (14) and 1 mg pCMVb-galactosidase (b-gal) incubated with secondary antibodies at 1:100 in antibody 0 expressing plasmid. After 24 hours, the cells were treated diluent for 45 minutes followed by 1 mg/mL 4 ,6-diami- and incubated overnight at 37C. Cells were lysed and dino-2-phenylindole (DAPI; DAKO). Confocal analysis luciferase activity and b-gal signals were read by a Mono- was performed with a Zeiss LSM 510 microscope (Carl light 3010 luminometer (Becton Dickinson). Zeiss, Inc.).

Colony formation assay in Matrigel matrix Statistical analyses Matrigel (Becton Dickinson) was thawed overnight at Statistics were run using GraphPad Prism Version 5.0. 4C. Twelve-well plates were coated with 200 mL Matrigel Statistical analyses used were one-way ANOVA followed t per well and incubated at 37 C for 30 minutes. Cells were by Tukey posttest or the Student test where appropriate. suspended at 5 103 in 400 mL of phenol red–free RPMI- 1640 and spread onto pregelled Matrigel followed by the Results addition of 360 mL of treatment media containing 40 mL PKCa expression correlates with sensitivity to E2 Matrigel. Plates were incubated at 37C for 10 days; PKCa-expression in clinical specimens predicted resis- medium containing 10% Matrigel was replaced to the top tance to tamoxifen (4). The ectopic overexpression of of the Matrigel every 3 days. Colonies were stained with PKCa in T47D cells (T47D:A18/PKCa) led to a tamox- crystal violet on day 10, and each well was counted by ifen-resistant, E2-inhibited phenotype in vivo (7), sug- light microscopy (20). gesting that PKCa may also predict a positive response to estrogenic therapeutic intervention. To derive an inde- Annexin V analysis of apoptosis pendent model of tamoxifen resistance, T47D:A18 cells The AlexaFluor 488 Annexin V/Dead Cell Apoptosis Kit were cultured long-term in the presence of 1 mmol/L (Invitrogen) was used to quantify cell death by ﬂow cyto- 4-hydroxytamoxifen, and several tamoxifen-resistant

www.aacrjournals.org Mol Cancer Ther; 13(11) November 2014 2517

Molloy et al.

50 1.0 Control A Control B Tamoxifen (1.5 mg/day) 4-OHT 100 nmol/L 0.8 40 E2 (capsule)

4 E2 1 nmol/L 30 0.6

20 0.4 RFU x 10

10 Tumor volume 0.2

0 0.0 0246810 0 5 10 105 Week Day Day

C D 15 * *

-Actin * β / α 5 PKC PKCa 0 8 C b-Actin Ca S8 :5 K :A1 W -7 P -TAM1 F / D 8 -7: A18/neo 7 C F M :A18 T4 C D M 47D: 7 T 4 47D:A1 T T

Figure 1. PKCa overexpression in breast cancer cells correlates with sensitivity to E2-induced growth inhibition. A, effect of E2 and 4-hydroxytamoxifen (4-OHT) on T47D:A18-TAM1 proliferation in vitro. DNA assays were performed as described in Materials and Methods. Graph shows mean SEM and is representative of 3 independent experiments. RFU, relative ﬂuorescent units. B, effect of tamoxifen and E2 on T47D:A18-TAM1 xenograft growth. Two mice per group were bilaterally injected with T47D:A18-TAM1 cells and either left untreated or were given oral tamoxifen (1.5 mg/d). Dashed line represents the start of E2 treatment. C, Western blot analysis of PKCa expression in E2-stimulated (T47D:A18/neo, T47D:A18, MCF-7:WS8) and E2-inhibited (T47D:A18/PKCa, T47D:A18-TAM1, and MCF-7:5C) cell lines. D, densitometric quantiﬁcation of 3 Western blot analyses from 3 independent cell lysates. The Student t test was used to compare each E2-inhibited cell line to the E2-stimulated counterpart. , P 0.05.

clones were identiﬁed, including T47D:A18-TAM1. E2-induced growth inhibition and/or be a predictive When evaluated in vitro, T47D:A18-TAM1 cell growth biomarker for E2-induced growth inhibition. was independent of E2 in addition to exhibiting tamoxifen resistance (Fig. 1A). In vivo T47D:A18-TAM1 xeno- SEMs are estrogenic and antiproliferative in grafts were found to act similar to T47D:A18/PKCa endocrine-resistant, PKCa-expressing breast cancer xenografts, growing hormone independently, tamoxi- cell lines in vitro fen-resistant, and regress when treated with E2 (Fig. We have previously reported that the E2 growth- 1B). Similar to the T47D:A18/PKCa cell line, T47D: inhibitory phenotype observed with T47D:A18/PKCa A18-TAM1 cells display increased PKCa expression tumors can be partially recapitulated when these cells compared with the tamoxifen-sensitive parental T47D: are cultured in 3D Matrigel as characterized by inhibi- A18 cell line (Fig. 1C and D, lanes 3 and 4). tion of colony formation (7, 17). In an effort to ﬁnd a The tamoxifen-resistant MCF-7:5C cell clone is growth potential alternative to E2 and to expand on positive inhibited by E2 in vivo and in contrast to T47D:A18/PKCa data obtained in vivo with raloxifene (11), 2 benzothio- and T47D:A18-TAM1 cells, growth is also inhibited by E2 phene SEMs, BTC and TTC-352 (Fig. 2A and B), were in 2-dimensional (2D) culture (12). MCF-7:5C cells display selected from a library of compounds and screened by increased expression of PKCa compared with tamoxifen- DNA growth assay in 2D. sensitive MCF-7:WS8 cells (Fig. 1C and D, lanes 5 and 6). To determine whether BTC and TTC-352 act as estrogen Because all 3 tamoxifen-resistant cell clones display ele- agonists, we treated cells in 2D culture and measured vated PKCa expression compared with the parental lines, DNA content as an index of proliferation. The higher and as PKCa knockdown in T47D:A18/PKCa cells led to concentrations of BTC and TTC-352 (100 nmol/L) stimu- a partial reversal of the E2-inhibited phenotype in vivo lated proliferation of T47D:A18/neo cells comparable (16), these results suggest that PKCa may play a role in with E2 (1 nmol/L; Fig. 2C and D). Both T47D:A18/PKCa

2518 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

SEMs for the Treatment of Tamoxifen-Resistant Breast Cancer

A B

C D 50 50 Control Control E2 1 nmol/L 40 E2 1 nmol/L 40 TTC-352 100 nmol/L

BTC 100 nmol/L 4 4 30 TTC-352 10 nmol/L 30 BTC 10 nmol/L TTC-352 1 nmol/L BTC 1 nmol/L 20 20 RFU x 10 RFU x 10

10 10

0 0 2 4 6 8 2 4 6 8 Day Day

EF60 60 Control Control E2 1 nmol/L E2 1 nmol/L BTC 100 nmol/L TTC-352 100 nmol/L 4 40 4 40 BTC 10 nmol/L TTC-352 10 nmol/L BTC 1 nmol/L TTC-352 1 nmol/L RFU x 10 20 RFU x 10 20

0 0 2 4 6 8 2 4 6 8 Day Day

60 G 80 H Control Control E2 1 nmol/L E2 1 nmol/L 60 TTC-352 100 nmol/L 4 4 BTC 100 nmol/L 40 TTC-352 10 nmol/L BTC 10 nmol/L 40 BTC 1 nmol/L TTC-352 1 nmol/L RFU x 10 RFU x 10 20 20

0 0 2 4 6 8 2 4 6 8 Day Day

Figure 2. Effect of BTC and TTC-352 on proliferation of T47D:A18/neo and T47D:A18/PKCa cells in vitro. Structures of BTC (A) and TTC-352 (B). Effect of BTC and TTC-352 treatment on the growth of T47D:A18/neo cells (C and D, respectively), T47D:A18/PKCa cells (E and F, respectively), and T47D:A18-TAM1 cells (G and H, respectively). DNA assays were performed as described in Materials and Methods. Graphs show mean SEM and are representative of 3 independent experiments. RFU, relative ﬂuorescent units.

www.aacrjournals.org Mol Cancer Ther; 13(11) November 2014 2519

Molloy et al.

A B C

250 150 150 * 200 * * 100 100 150 *** *** *** 100 *** 50 50 *** *** 50 Colony formation (%) Colony formation (%)

0.0 0.0 Colony formation (%) 0.0 Control E2 BTC TTC-352 Control E2 BTC TTC-352 Control E2 BTC TTC-352

D E F 100 5 *** 1.5 *** *** 4 75 ** ** 1.0 *** 3 50 *** *** 2 0.5 Cell number Cell number * 25 1 * Apoptotic cells (%)

(normalized to control) (normalized to control) ** 0 0.0 0

Control Control Control

E2 (1 nmol/L) E2 (1 nmol/L) BTC (1 nmol/L) BTC (1 nmol/L) E2 (1 nmol/L) BTC (10 nmol/L) BTC (10 nmol/L) BTC (100 nmol/L) BTC (100 nmol/L) TTC-352 (1 nmol/L) TTC-352 (1 nmol/L) BTC (100 nmol/L) TTC-352 (10 nmol/L) TTC-352 (10 nmol/L) TTC-352 (100 nmol/L) TTC-352 (100 nmol/L) TTC-352 (100 nmol/L)

Figure 3. BTC and TTC-352 inhibit T47D:A18/PKCa and T47D:A18-TAM1 colony formation in 3D Matrigel and induce apoptosis in MCF-7:5C cells in vitro. Colonies were established as described in Materials and Methods and treated for 10 days [Control (0.1% DMSO), E2 1 nmol/L, BTC 100 nmol/L, TTC-352 100 nmol/L]. A, T47D:A18/neo. B, T47D:A18/PKCa. C, T47D:A18-TAM1. Proliferation assay was performed following 9 days of treatment. D, MCF-7:WS8. E, MCF-7:5C. F, apoptosis was assessed in MCF-7:5C cells following 6 days of treatment as described in Materials and Methods. , P 0.05 versus DMSO; , P 0.01 versus DMSO; , P 0.001 versus DMSO. Graph shows mean SEM of 3 independent experiments.

and T47D:A18-TAM1 cells proliferated equally in the pres- TTC-352 treatment resulted in an increase in ERa tran- ence of E2, BTC, and TTC-352 (1–100 nmol/L; Fig. 2E–H). scriptional activity that was concentration-dependent To determine whether BTC and TTC-352 mirror the (Fig. 4A–C). BTC acted as a full agonist with respect to antiproliferative effect of E2 on T47D:A18/PKCa cells in ERE-luciferase induction in all cell lines, whereas TTC-352 3D culture, colony formation in Matrigel was examined. appears to act as a partial ER agonist yielding a significant BTC and TTC-352 treatment resulted in increased T47D: but partial induction at 100 nmol/L concentration. Taken A18/neo colony formation (Fig. 3A) while significantly together, our in vitro findings suggest that SEMs exhibit preventing T47D:A18/PKCa and T47D:A18-TAM1 colo- estrogenic activity and mimic the inhibitory activity of E2 ny formation in 3D (Fig. 3B and C). in MCF-7:5C cells in 2D culture and in T47D:A18/PKCa To confirm the inhibitory activity of SEMs in a third in and T47D:A18-TAM1 cells in 3D culture. vitro model of E2-induced growth inhibition, we determined the effects in tamoxifen-resistant MCF-7:5C cells: Bioavailability of BTC and benzothiophene SERMs MCF-7:5C cells are growth inhibited by E2 in 2D culture In humans, the absolute bioavailability of orally admin- (12). Parental hormone-dependent MCF-7:WS8 cells were istered raloxifene is reported as 2%, with oral clearance of growth stimulated when subjected to BTC and TTC-352 44 L/kg/h (18, 19). Desmethylarzoxifene is a more potent treatment (Fig. 3D); however, BTC and TTC-352 (100 estrogen antagonist in the breast and maintains estrogen nmol/L) significantly inhibited the growth of MCF-7:5C agonist actions in bone tissues, however, it also has poor cells over 9 days (Fig. 3E). In addition, apoptosis signif- bioavailability (20). Arzoxifene was designed as a des- icantly increased in these cells 6 days posttreatment with methylarzoxifene prodrug to overcome the problems E2, BTC, and TTC-352 (Fig. 3F). associated with low bioavailability (21–24). As part of a To confirm the estrogenic activity of BTC and TTC-352, comparative study of biologic activity of desmethylarzox- we examined transcriptional activation of ERa using an ifene, arzoxifene, and F-DMA in juvenile female rats, ERE-luciferase reporter construct (14). In T47D:A18/neo, metabolism was assessed by quantification of remaining T47D:A18/PKCa, and T47D:A18-TAM1 cells, BTC and drug in plasma after 3 days of drug administration:

2520 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

SEMs for the Treatment of Tamoxifen-Resistant Breast Cancer

A B C 250 150 150

200 *** 100 *** 100 *** 150 ** ** 100 50 50 * 50

ERE-luciferase activity 0 ERE-luciferase activity 0 ERE-luciferase activity 0 E2 DMSO DMSO DMSO

E2 1 nmol/L E2 1 nmol/L BTC 1 nmol/L BTC 1 nmol/L BTC 1 nmol/L BTC 10 nmol/L BTC 10 nmol/L BTC 10 nmol/L BTC 100 nmol/L BTC 100 nmol/L BTC 100 nmol/L TTC-352 1 nmol/L TTC-352 1 nmol/L TTC-352 1 nmol/L TTC-352 10 nmol/L TTC-352 10 nmol/L TTC-352 10 nmol/L TTC-352 100 nmol/L TTC-352 100 nmol/L TTC-352 100 nmol/L

Figure 4. BTC and TTC-352 induce ERa transcriptional activity in T47D:A18/neo, T47D:A18/PKCa, and T47D:A18-TAM1 cell lines. A, T47D:A18/neo. B, T47D: A18/PKCa. C, T47D:A18-TAM1 cell lines. Data are expressed normalized to E2 (100%). , P 0.05 versus DMSO; , P 0.01 versus DMSO; , P 0.001 versus DMSO. Graph shows mean SEM of 3 independent experiments. desmethylarzoxifene was not observed above detection E2 capsule (9 mice), oral raloxifene 1.5 mg/d (9 mice), limits, whereas arzoxifene and F-DMA were easily quan- oral BTC 1.5 mg/d (9 mice), or oral TTC-352 1.5 mg/d (4 tifiable (25–27). BTC is the benzothiophene core of des- mice). Following 2 weeks, all treatments significantly methylarzoxifene and raloxifene, whereas TTC-352 bears reduced tumor volume compared with nontreated con- structural similarity with F-DMA. Consequently, we were trols (P < 0.05; Fig. 5A): BTC (88%) and TTC-352 (70%) concerned that the bioavailability of BTC would be too elicited a decrease in tumor volume that exceeded the low for study in vivo (24, 25, 27, 28). Therefore, plasma regression elicited by raloxifene (50%). Most notably, levels of BTC were measured after oral administration to regression induced by E2, BTC, and TTC-352 was sus- ovariectomized athymic nude mice. Because BTC ionizes tained for at least 4 weeks after drug withdrawal, in poorly by electrospray ionization-mass spectrometry contrast to raloxifene (Fig. 5A). (ESI-MS), a chemical derivatization method was devel- We previously reported that regression of T47D:A18/ oped using tandem mass spectroscopic MRM analysis. PKCa tumors, induced by both raloxifene and E2, is Drug was detected in plasma at a peak concentration of 40 accompanied by the exit of ERa from the nucleus suggest- ng/mL at 30 minutes, which decreased to 10 ng/mL after ing the possible mechanistic involvement of extranuclear 6 hours (Supplementary Fig. S6). Given the observation of ERa (11, 17). We show here that tumor regression follow- BTC well above detection limits in mouse plasma, it was ing treatment with BTC and TTC-352 is also accompanied decided to compare the effects of BTC and TTC-352 with by exit of ERa from the nucleus (Fig. 5B). These results raloxifene and E2 in the T47D:A18/PKCa xenograft mod- suggest that the mechanism of regression by benzothio- el using oral delivery. phene SEMs is similar to that of E2 and may involve extranuclear ERa. SEMs cause regression of hormone-independent, To test the effects of E2 and SEMs on T47D:A18-TAM1 tamoxifen-resistant xenografts xenograft tumors, tumors were established in absence of We recently reported that raloxifene treatment results treatment, and at 10 weeks, 11 athymic nude mice were in significant regression of tamoxifen-resistant T47D: randomized into 4 treatment groups: untreated control A18/PKCa tumors. However, raloxifene did not yield arm (2 mice), E2 capsule (3 mice), oral BTC 1.5 mg/d (3 complete tumor regression and importantly was without mice), or oral TTC-352 1.5 mg/d (3 mice). Following 2 activity in 3D cultures (11). In contrast to raloxifene, the weeks of daily BTC and TTC-352 treatment, we observed SEMs, BTC and TTC-352, are ER agonists in 2D cultures significant T47D:A18-TAM1 tumor regression (Fig. 5C). (Fig. 2) and are able to inhibit both T47D:A18/PKCa and Therefore, the SEMs BTC and TTC-352 which are weak T47D:A18-TAM1 colony formation in 3D (Fig. 3B and C). agonists in vitro show similar efficacy to E2 causing To determine whether these compounds could induce significant tumor regression in 2 tamoxifen-resistant T47D:A18/PKCa tumor regression, T47D:A18/PKCa breast cancer xenografts. cells were injected into 40 athymic mice and tumors were allowed to grow in the absence of any treatment for 7 SEMs do not support growth of hormone-dependent weeks reaching a mean tumor size of 0.5 cm2 (100%). At xenografts that time, the mice were randomized to 5 treatment The ER-positive, hormone-dependent T47D:A18/neo groups: continued no treatment (control arm, 9 mice), breast cancer cell line requires E2 for growth in vitro and

www.aacrjournals.org Mol Cancer Ther; 13(11) November 2014 2521

Molloy et al.

A 150 Control E2 (capsule) 100 RAL (1.5 mg/day) BTC (1.5 mg/day) TTC-352 (1.5 mg/day) 50 * * * Tumor volume (% change) 0 * 8 8 10 20 30 40 50 Day B Figure 5. BTC and TTC-352 inhibit Control E2 RAL BTC TTC-352 T47D:A18/PKCa and T47D:A18- TAM1 xenograft tumors. A, BTC and TTC-352 treatment result in regression of T47D:A18/PKCa ERa tumors. Graph shows percentage of tumor regression (100%; 0.5 cm2). Dotted line indicates when treatment was ended. Arrow designates where tumors from B DAPI were obtained. , P 0.001 versus control. Graph shows mean SEM. B, ERa localization in T47D:A18/PKCa tumors by immunoﬂuorescence staining. Total magniﬁcation, 630.C, Merge T47D:A18-TAM1 tumors regress when treated with BTC and TTC- 352. Dotted lines represent start of treatment. RAL, raloxifene. C 0.5 Control E2 (capsule) 0.4 BTC (1.5 mg/day) TTC-352 (1.5 mg/day) 0.3

0.2

Tumor volume 0.1

0.0 0 2 4 6 8 10 11 12 Week

in vivo (6, 7). Because BTC and TTC-352 support the growth treatment groups (Fig. 6A). Thus, the growth of T47D:A18/ of T47D:A18/neo cells in 2D (Fig. 1A and B) and 3D neo cells seen at higher concentrations of BTC and TTC-352 cultures (Fig. 3A), we next asked whether these SEMs in vitro (Fig. 2C and D) was not recapitulated in vivo. would sustain the growth of T47D:A18/neo tumors in vivo. Importantly, the dose of SEMs capable of causing robust T47D:A18/neo cells were bilaterally injected into 20 athy- regression of T47D:A18/PKCa and T47D:A18-TAM1 mic mice and divided into 6 treatment groups: 3 nontreated tumors did not stimulate T47D:A18/neo tumor growth. controls, 3 E2 capsule, 3 oral tamoxifen 1.5 mg/d, 3 oral In addition, no significant weight loss was observed over raloxifene 1.5 mg/d, 4 oral BTC 1.5 mg/d, or 4 oral TTC-352 the 7-week treatment period (Fig. 6B). Therefore, SEMs are 1.5 mg/d. Following 7 weeks of treatment, mice treated capable of causing regression of endocrine-resistant breast with E2, as expected, harbored T47D:A18/neo tumors that cancer xenografts, and importantly they are unable to reached an average size of about 0.35 cm2 (100%), whereas support the growth of hormone-dependent xenografts in no significant tumor growth was observed in the other vivo. This result is significant as it suggests that SEMs will

2522 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

SEMs for the Treatment of Tamoxifen-Resistant Breast Cancer

treated with raloxifene, which is known to be without A uterotrophic actions. In contrast, and as expected, tamox- 150 Control ifen and E2 caused signiﬁcant increases in uterine weight. E2 (capsule) 100 *** TAM (1.5 mg/day) RAL (1.5 mg/day) BTC (1.5 mg/day) Discussion 50 TTC-352 (1.5 mg/day)

(% change) Before the introduction of tamoxifen, the ER agonists, Tumor volume E2 and diethystilbesterol, were recognized to have clinical 0 2 4 6 8 utility in the treatment of breast cancer (9, 34). Discovery Week of alternative ER agonists that might mimic the anticancer B effects of E2 and diethystilbesterol, without the unaccept- 30 Control able side effects, is a reasonable and novel strategy. To E2 (capsule) identify such SEMs, we chose to take advantage of the 20 TAM (1.5 mg/day) RAL (1.5 mg/day) knowledge that the growth of several ER-positive, but BTC (1.5 mg/day) endocrine-resistant, breast cancer cell lines is inhibited 10 TTC-352 (1.5 mg/day) by E2. These tamoxifen-resistant cell lines overexpress Body weight (g) PKCa, the role of which in the antiproliferative actions of 0 0 2 4 6 8 E2 is not yet deﬁned. Using 3 such cell lines (T47D:A18/ Week PKCa, T47D:A18-TAM1, MCF-7:5C), 2 novel SEMs were identiﬁed. C 6 BTC and TTC-352 showed estrogenic activity in vitro in *** *** 2D cultures with respect to proliferation and ERE-medi- 4 ated transcription, although (i) both SEMs were less potent than E2 and (ii) TTC-352 displayed partial agonist 2 properties. In 2D cultures of tamoxifen-resistant MCF-

body weight (g) 7:5C cells, both SEMs and E2 inhibited growth and Uterine weight (mg)/ 0 induced apoptosis (Fig. 3E and F, respectively). BTC and NTC E2 TAM RAL BTC TTC-352 TTC-352 mimicked E2 and inhibited tamoxifen-resistant T47D:A18/PKCa and T47D:A18-TAM1 colony formation a Figure 6. BTC and TTC-352 have no effect on T47D:A18/neo tumor (Fig. 3B and C). Regression of T47D:A18/PKC tumors by growth, body weight, or uterine weights of ovariectomized mice. A, BTC BTC and TTC-352 was accompanied by exit of ERa from and TTC-352 do not result in growth of T47D:A18/neo tumors. B, body the nucleus to extranuclear sites, again mimicking the weights of treated mice from A. C, uterine weights from mice in A. Weights actions of E2 (Fig. 5B) and hinting at a role for extranuclear P are reported as uterine weight (mg)/body weight (g). , 0.001 versus ERa in BTC- and TTC-352–induced tumor regression. control. Graphs show mean SEM. RAL, raloxifene; TAM, tamoxifen. Withdrawal of SEM treatment did not lead to relapse and regrowth. E2 and SEM treatment also induced regression not stimulate the growth of any endocrine-sensitive breast of established T47D:A18-TAM1 tumors in a xenograft cancer cells that may be present in heterogeneous tumors. model (Fig. 5C). These xenograft models which represent both exogenous and endogenous PKCa expression are SEMs are without uterotrophic actions in athymic similarly inhibited by SEMs. mice We have observed that translocation of ERa from the E2 has a proliferative effect on the endometrium result- nucleus to cytoplasm is a common feature of treatments ing in an increase in uterine weight. Tamoxifen has an that cause regression of T47D:A18/PKCa tumors, but not estrogenic effect on endometrial growth, which leads to an those that are ineffective, that is, tamoxifen (11). The increased risk of developing endometrial cancer (29). In similarity with the diarylthiohydantoin antiandrogens ovariectomized rats at a minimally effective dose, ralox- (e.g., enzalutamide, ARN509, RD162) that cause a similar ifene did not increase uterine weight in contrast to E2 and translocation of the androgen receptor (AR) in prostate tamoxifen and at doses up to 10 mg/kg/d did not increase cancer cells is of interest, in particular because this feature luminal epithelial cell thickness (30–33). Mindful of the is seen as a clinical advantage over older antiandrogens in vitro estrogen agonist actions of BTC and TTC-352 (35, 36). Enzalutamide is currently approved for the treat- described above, we sought to compare the effects on ment of castration-resistant prostate cancer. We have uterine weight of BTC, TTC-352, raloxifene, tamoxifen, previously reported an increased physical interaction of and E2. Following 7 weeks of treatment, the uteri from the ERa and caveolin-1 (cav-1) in E2-induced tumor regres- ovariectomized mice used in the T47D:A18/neo xenograft sion (11), suggesting that cav-1 may be responsible for the study (Fig. 6A) were excised and weighed. Interestingly, transport of ERa to extranuclear sites following E2 or SEM BTC and TTC-352 caused no signiﬁcant increase in the treatment. Cav-1 serves as a scaffold protein recruiting uterine weights of the treated mice (Fig. 6C). The SEM- signaling molecules, including ERa, to the plasma mem- treated mice were indistinguishable from those that were brane to form a signalosome complex. Contrary to the

www.aacrjournals.org Mol Cancer Ther; 13(11) November 2014 2523

Molloy et al.

growth-promoting effects and activation of kinase cas- exhibiting endocrine therapy–resistant breast cancers (40, cades generally associated with extranuclear ERa and E2 41). Clinical trials have demonstrated the efficacy of E2 in treatment, we have previously reported a downregulation this setting (38, 39). In fact, a long-term follow-up study and inactivation of Akt following E2 treatment in tumors indicated a survival advantage for patients treated with overexpressing PKCa (17). We hypothesize that the over- the synthetic estrogen diethystilbesterol compared with expression of PKCa in the presence of E2 or SEMs may patients treated with tamoxifen (42). The basis for the lead to a modified signalosome complex in the cytoplasm, clinical use of estrogens is supported by a number of altering the canonical effects of E2 thus leading to apo- preclinical laboratory models (7, 8, 12, 43–48). The ability ptosis. We are currently overexpressing kinase dead to predict a patient’s response to therapy before treatment PKCa mutants in T47D cells, as it is likely that the kinase would be a very attractive clinical option. Patients pre- activity of PKCa contributes to the increased sensitivity to senting with tumors overexpressing PKCa would likely estrogenic compounds in endocrine-resistant breast can- benefit from an E2-like therapy, of which there are cur- cer cells. rently few options. Use of tamoxifen in breast cancer therapy is currently In the present study, we sought to identify possible recommended for at least 5 years. Although only used alternative therapeutic options for tamoxifen-resistant following tamoxifen treatment failure, there is a potential breast cancers. As the presence of PKCa dictated an risk with the use of E2 or SEMs in promoting tumorigen- enhanced estrogenic response to BTC and TTC-352 as esis of endocrine-dependent neoplasms. Although E2 well as a tumor regressing phenotype, these compounds induced growth of hormone-dependent, tamoxifen-sen- may have potential clinical value in the endocrine-resis- sitive, parental T47D:A18/neo tumors in vivo, neither BTC tant setting. Our findings support the use of SEMs in nor TTC-352 supported T47D:A18/neo tumor growth patients who no longer respond to conventional endo- (Fig. 6A). Furthermore, in contrast to E2 and tamoxifen, crine therapies and whose tumors overexpress PKCa. neither BTC nor TTC-352 treatment resulted in an increase Importantly, treatment with BTC and TTC-352 had min- in the uterine weight of mice (Fig. 6C), indicating that BTC imal effects on proliferation within the uteri of mice in and TTC-352 act in vivo as SEMs with selective E2-like vivo, suggesting that the estrogenic effects of these agents activity in endocrine-resistant mammary tumors. It is are specific to the breast. Both BTC and TTC-352 are fascinating that structurally related compounds, various- potential alternatives to E2 treatment and represent chem- ly showing classical ERa antagonist activity (raloxifene), ical probes and lead compounds for further optimization or classical agonist activity (BTC, TTC-352) should elicit toward new treatment options in the management of the same tumor regressing actions in T47D:A18/PKCa endocrine-resistant breast cancer. and T47D:A18-TAM1 xenografts, although the failure of fl raloxifene-induced regression to persist after drug with- Disclosure of Potential Con icts of Interest No potential conflicts of interest were disclosed. drawal is noted. That the estrogen agonists, BTC and TTC- 352, did not stimulate growth of estrogen-sensitive T47D: Authors' Contributions A18/neo xenografts or uterine tissues is most simply Conception and design: M.E. Molloy, B.E.P. White, T. Gherezghiher, rationalized by the relatively low potency of these ago- V.C. Jordan, G.R.J. Thatcher, D.A. Tonetti Development of methodology: M.E. Molloy, B.E.P. White, T. Gherezghi- nists, again indicating involvement of a pathway that is her, H. Zhao, V.C. Jordan, D.A. Tonetti not simply classically mediated by ERa in T47D:A18/ Acquisition of data (provided animals, acquired and managed patients, PKCa and T47D:A18-TAM1 xenografts. provided facilities, etc.): M.E. Molloy, B.E.P. White, B.T. Michalsen, R. Xiong, H. Patel, H. Zhao, V.C. Jordan, D.A. Tonetti Resistance to endocrine therapies is a major obstacle Analysis and interpretation of data (e.g., statistical analysis, biostatis- encountered in the clinical setting. Currently, there is a tics, computational analysis): M.E. Molloy, B.E.P. White, T. Gherezghiher, lack of effective therapeutic options for women who no H. Zhao, V.C. Jordan, D.A. Tonetti Writing, review, and/or revision of the manuscript: M.E. Molloy, longer respond to conventional, endocrine therapy. The T. Gherezghiher, V.C. Jordan, G.R.J. Thatcher, D.A. Tonetti combination of the aromatase inhibitor exemestane with Administrative, technical, or material support (i.e., reporting or orga- nizing data, constructing databases): P.Y Maximov, V.C. Jordan the mTOR inhibitor everolimus improved progression- Study supervision: V.C. Jordan, G.R.J. Thatcher, D.A. Tonetti free survival for patients with endocrine-resistant breast cancer; however, an increased toxicity profile was Acknowledgments observed in patients who took the combined treatment The authors acknowledge the Histopathology Core and the Confocal Microscopy Facility of the Research Resources Center at the University of (37). There is a clinical need to provide endocrine therapy– Illinois at Chicago for providing services and expertise. resistant patients with effective and safe alternatives. Our findings and those of others suggest that PKCa expression Grant Support is a predictive marker of disease outcome for patients on These studies were supported by NCI/NIH RO1 CA122914 (to D.A. a Tonetti) and R01 102590 to (G.R.J. Thatcher). endocrine therapy (3–5). Further PKC expression may The costs of publication of this article were defrayed in part by the predict a positive response to E2 or an E2-like compound payment of page charges. This article must therefore be hereby marked (7). E2 has clinical efficacy (9, 34, 38, 39), but due to advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. unfavorable side effects, it is no longer used for treatment. Recently, the use of E2 or an E2-like compound has Received April 10, 2014; revised August 27, 2014; accepted September 3, reemerged as a possible treatment strategy for patients 2014; published OnlineFirst September 9, 2014.

2524 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

SEMs for the Treatment of Tamoxifen-Resistant Breast Cancer

References 1. Dempsey EC, Newton AC, Mochly-Rosen D, Fields AP, Reyland chemoprevention of experimental breast cancer. Cancer Res 2001;61: ME, Insel PA, et al. Protein kinaseCisozymesandtheregulationof 8412–5. diverse cell responses. Am J Physiol Lung Cell Mol Physiol 2000; 22. Burke TW, Walker CL. Arzoxifene as therapy for endometrial cancer. 279:L429–38. Gynecol Oncol 2003;90:S40–6. 2. Mackay HJ, Twelves CJ. Protein kinase C: a target for anticancer 23. Buzdar A, O'Shaughnessy JA, Booser DJ, Pippen JE Jr, Jones SE, drugs? Endocr Relat Cancer 2003;10:389–96. Munster PN, et al. Phase II, randomized, double-blind study of two 3. Assender JW, Gee JMW, Lewis I, Ellis IO, Robertson JFR, Nicholson RI. dose levels of arzoxifene in patients with locally advanced or meta- Protein kinase C isoform expression as a predictor of disease outcome static breast cancer. J Clin Oncol 2003;21:1007–14. on endocrine therapy in breast cancer. J Clin Pathol 2007;60:1216–21. 24. Liu H, Liu J, van Breemen RB, Thatcher GRJ, Bolton JL. Bioactivation 4. Tonetti DA, Morrow M, Kidwai N, Gupta A, Badve S. Elevated protein of the selective estrogen receptor modulator desmethylated arzox- kinase C alpha expression may be predictive of tamoxifen treatment ifene to quinoids: 40-fluoro substitution prevents quinoid formation. failure. Br J Cancer 2003;88:1400–2. Chem Res Toxicol 2005;18:162–73. 5. Lonne G, Cornmark L, Zahirovic I, Landberg G, Jirstrom K, Larsson C. 25. Yu B, Dietz BM, Dunlap T, Kastrati I, Lantvit DD, Overk CR, et al. PKCalpha expression is a marker for breast cancer aggressiveness. Structural modulation of reactivity/activity in design of improved ben- Mol Cancer 2010;9:76. zothiophene selective estrogen receptor modulators: induction of 6. Tonetti DA, Chisamore MJ, Grdina W, Schurz H, Jordan VC. Stable chemopreventive mechanisms. Mol Cancer Ther 2007;6:2418–28. transfection of protein kinase C alpha cDNA in hormone-dependent 26. Overk CR, Peng KW, Asghodom RT, Kastrati I, Lantvit DD, Qin Z, et al. breast cancer cell lines. Br J Cancer 2000;83:782–91. Structure-activity relationships for a family of benzothiophene selec- 7. Chisamore MJ, Ahmed Y, Bentrem DJ, Jordan VC, Tonetti DA. Novel tive estrogen receptor modulators including raloxifene and arzoxifene. antitumor effect of estradiol in athymic ice injected with a T47D breast ChemMedChem 2007;2:1520–6. cancer cell line overexpressing protein kinase C alpha. Clin Cancer Res 27. Qin Z, Kastrati I, Ashgodom RT, Lantvit DD, Overk CR, Choi Y, et al. 2001;7:3156–65. Structural modulation of oxidative metabolism in design of improved 8. Yao K, Lee E-S, Bentrem DJ, England G, Schafer JIM, ORegan RM, benzothiophene selective estrogen receptor modulators. Drug Metab et al. Antitumor action of physiological estradiol on tamoxifen-stimu- Dispos 2009;37:161–9. lated breast tumors grown in athymic mice. Clin Cancer Res 2000;6: 28. Liu H, Bolton JL, Thatcher GRJ. Chemical modification modulates 2028–36. estrogenic activity, oxidative reactivity, and metabolic stability in 40F- 9. Ingle JN, Ahmann DL, Green SJ, Edmonson JH, Bisel HF, Kvols LK, DMA, a new benzothiophene selective estrogen receptor modulator. et al. Randomized clinical trial of diethylstilbestrol versus tamoxifen in Chem Res Toxicol 2006;19:779–87. postmenopausal women with advanced breast cancer. N Engl J Med 29. Fisher B, Costantino JP, Redmond CK, Fisher ER, Wickerham DL, 1981;304:16–21. Cronin WM, et al. Endometrial cancer in tamoxifen-treated breast 10. Cole M, Jones C, Todd I. A new anti-oestrogenic agent in late breast cancer patients: findings from the National Surgical Adjuvant Breast cancer: an early clinical appraisal of ICI46474. Br J Cancer 1971; and Bowel Project (NSABP) B-14. J Natl Cancer Inst 1994;86:527–37. 25:270. 30. Grese TA, Sluka JP, Bryant HU, Cullinan GJ, Glasebrook AL, Jones CD, 11. Perez White B, Molloy M, Zhao H, Zhang Y, Tonetti D. Extranuclear et al. Molecular determinants of tissue selectivity in estrogen receptor ERalpha is associated with regression of T47D PKCalpha-overexpres- modulators. Proc Natl Acad Sci U S A 1997;94:14105–10. sing, tamoxifen-resistant breast cancer. Mol Cancer 2013;12:34. 31. Grese TA, Cho S, Finley DR, Godfrey AG, Jones CD, Lugar CW III, et al. 12. Lewis JS, Osipo C, Meeke K, Jordan VC. Estrogen-induced apoptosis Structure-activity relationships of selective estrogen receptor modu- in a breast cancer model resistant to long-term estrogen withdrawal. lators: modifications to the 2-arylbenzothiophene core of raloxifene. J Steroid Biochem Mol Biol 2005;94:131–41. J Med Chem 1997;40:146–67. 13. Abdelhamid R, Luo J, Vandevrede L, Kundu I, Michalsen B, Litosh VA, 32. Black LJ, Sato M, Rowley ER, Magee DE, Bekele A, Williams DC, et al. et al. Benzothiophene selective estrogen receptor modulators provide Raloxifene (LY139481 HCI) prevents bone loss and reduces serum neuroprotection by a novel GPR30-dependent mechanism. ACS cholesterol without causing uterine hypertrophy in ovariectomized Chem Neurosci 2011;2:256–68. rats. J Clin Invest 1994;93:63–9. 14. Catherino WH, Jordan VC. Increasing the number of tandem estrogen 33. Fuchs-Young R, Glasebrook AL, Short LL, Draper MW, Rippy MK, Cole response elements increases the estrogenic activity of a tamoxifen HW, et al. Raloxifene is a tissue-selective agonist/antagonist that func- analogue. Cancer Lett 1995;92:39–47. tions through the estrogen receptor. Ann N Y Acad Sci 1995;761:355–60. 15. O'Regan RM, Cisneros A, MacGregor JL, Muenzner HD, Assikis VJ, 34. Kennedy BJ. Massive estrogen administration in premenopausal Piette M, et al. Effects of the antiestrogens tamoxifen, toremifene, and women with metastatic breast cancer. Cancer 1962;15:641–8. ICI 182,780 on endometrial cancer growth. J Natl Cancer Inst 35. Tran C, Ouk S, Clegg NJ, Chen Y, Watson PA, Arora V, et al. Devel- 1998;90:1552–8. opment of a second-generation antiandrogen for treatment of 16. Lin X, Yu Y, Zhao H, Zhang Y, Manela J, Tonetti DA. Overexpression of advanced prostate cancer. Science 2009;324:787–90. PKCa is required to impart estradiol inhibition and tamoxifen-resis- 36. Clegg NJ, Wongvipat J, Joseph JD, Tran C, Ouk S, Dilhas A, et al. ARN- tance in a T47D human breast cancer tumor model. Carcinogenesis 509: A novel antiandrogen for prostate cancer treatment. Cancer Res 2006;27:1538–46. 2012;72:1494–503. 17. Zhang Y, Zhao H, Asztalos S, Chisamore M, Sitabkhan Y, Tonetti DA. 37. Baselga J, Campone M, Piccart M, Burris HA, Rugo HS, Sahmoud T, Estradiol-induced regression in T47D:A18/PKCa tumors requires the et al. Everolimus in postmenopausal hormone-receptor-positive estrogen receptor and interaction with the extracellular matrix. Mol advanced breast cancer. N Engl J Med 2012;366:520–9. Cancer Res 2009;7:498–510. 38. Ellis MJ, Gao F, Dehdashti F, Jeffe DB, Marcom PK, Carey LA, et al. 18. Snyder KR, Sparano N, Malinowski JM. Raloxifene hydrochloride. Am Lower-dose vs high-dose oral estradiol therapy of hormone receptor- J Health Syst Pharm 2000;57:1669–75; quiz 1676–8. positive, aromatase inhibitor-resistant advanced breast cancer. JAMA 19. Hochner-Celnikier D. Pharmacokinetics of raloxifene and its clinical 2009;302:774–80. application. Eur J Obstet Gynecol Reprod Biol 1999;85:23–9. 39. Lønning PE, Taylor PD, Anker G, Iddon J, Wie L, Jørgensen L-M, et al. 20. Palkowitz AD, Glasebrook AL, Thrasher KJ, Hauser KL, Short LL, High-dose estrogen treatment in postmenopausal breast cancer Phillips DL, et al. Discovery and synthesis of [6-hydroxy-3-[4-[2-(1- patients heavily exposed to endocrine therapy. Breast Cancer Res piperidinyl)ethoxy]phenoxy]-2-(4-hydroxyphenyl)]benzo[b]thiophene: Treat 2001;67:111–6. a novel, highly potent, selective estrogen receptor modulator. J Med 40. Ingle J. Estrogen as therapy for breast cancer. Breast Cancer Res Chem 1997;40:1407–16. 2002;4:133–6. 21. Suh N, Glasebrook AL, Palkowitz AD, Bryant HU, Burris LL, Starling JJ, 41. Jordan VC, Obiorah I, Fan P, Kim HR, Ariazi E, Cunliffe H, et al. et al. Arzoxifene, a new selective estrogen receptor modulator for The St. Gallen Prize Lecture 2011: Evolution of long-term adjuvant

www.aacrjournals.org Mol Cancer Ther; 13(11) November 2014 2525

Molloy et al.

anti-hormone therapy: consequences and opportunities. The 45. Liu H, Lee E-S, Gajdos C, Pearce ST, Chen B, Osipo C, et al. Apoptotic Breast 2011;20 Suppl 3:S1–11. action of 17b-estradiol in raloxifene-resistant MCF-7 cells in vitro and 42. Peethambaram PP, Ingle JN, Suman VJ, Hartmann LC, Loprinzi CL. in vivo. J Natl Cancer Inst 2003;95:1586–97. Randomized trial of diethylstilbestrol vs. tamoxifen in postmenopausal 46. Santen RJ, Song RX, Zhang Z, Yue W, Kumar R. Adaptive hypersen- women with metastatic breast cancer. An updated analysis. Breast sitivity to estrogen. Clin Cancer Res 2004;10:337s-45s. Cancer Res Treat 1999;54:117–22. 47. Song RX-D, Mor G, Naftolin F, McPherson RA, Song J, Zhang Z, et al. 43. Osipo C, Gajdos C, Cheng D, Jordan VC. Reversal of tamoxifen Effect of long-term estrogen deprivation on apoptotic responses of breast resistant breast cancer by low dose estrogen therapy. J Steroid cancer cells to 17beta-estradiol. J Natl Cancer Inst 2001;93:1714–23. Biochem Mol Biol 2005;93:249–56. 48. Shim W-S, Conaway M, Masamura S, Yue W, Wang J-P, Kumar R, 44. Osipo C, Gajdos C, Liu H, Chen B, Jordan VC. Paradoxical action of et al. Estradiol hypersensitivity and mitogen-activated protein kinase fulvestrant in estradiol-induced regression of tamoxifen-stimulated expression in long-term estrogen deprived human breast cancer cells breast cancer. J Natl Cancer Inst 2003;95:1597–608. in vivo. Endocrinology 2000;141:396–405.

2526 Mol Cancer Ther; 13(11) November 2014 Molecular Cancer Therapeutics

Novel Selective Estrogen Mimics for the Treatment of Tamoxifen-Resistant Breast Cancer

Mary Ellen Molloy, Bethany E. Perez White, Teshome Gherezghiher, et al.

Mol Cancer Ther 2014;13:2515-2526. Published OnlineFirst September 9, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-14-0319

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2014/09/09/1535-7163.MCT-14-0319.DC1

Cited articles This article cites 48 articles, 13 of which you can access for free at: http://mct.aacrjournals.org/content/13/11/2515.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/13/11/2515.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/13/11/2515. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.