Cross-Resistance Among Next-Generation Antiandrogen Drugs Through the AKR1C3/AR-V7 Axis in Advanced Prostate Cancer Jinge Zhao1, Shu Ning1, Wei Lou1, Joy C

Published OnlineFirst May 19, 2020; DOI: 10.1158/1535-7163.MCT-20-0015

MOLECULAR CANCER THERAPEUTICS | CANCER BIOLOGY AND TRANSLATIONAL STUDIES

Cross-Resistance Among Next-Generation Antiandrogen Drugs Through the AKR1C3/AR-V7 Axis in Advanced Prostate Cancer Jinge Zhao1, Shu Ning1, Wei Lou1, Joy C. Yang1, Cameron M. Armstrong1, Alan P. Lombard1, Leandro S. D'Abronzo1, Christopher P. Evans1,2, Allen C. Gao1,2,3, and Chengfei Liu1,2

ABSTRACT ◥ The next-generation antiandrogen drugs, XTANDI (enzaluta- AKR1C3/AR-V7 axis confers this cross-resistance. Knockdown of mide), ZYTIGA (abiraterone acetate), ERLEADA (apalutamide) AR-V7 in enzalutamide-resistant cells resensitize cells to apaluta- and NUBEQA (darolutamide) extend survival times and improve mide and darolutamide treatment. Furthermore, targeting AKR1C3 quality of life in patients with advanced prostate cancer. Despite resensitizes resistant cells to apalutamide and darolutamide treat- these advances, resistance occurs frequently and there is currently ment through AR-V7 inhibition. Chronic apalutamide treatment in no deﬁnitive cure for castration-resistant prostate cancer. Our C4-2B cells activates the steroid hormone biosynthesis pathway and previous studies identiﬁed that similar mechanisms of resistance increases AKR1C3 expression, which confers resistance to enzalu- to enzalutamide or abiraterone occur following treatment and tamide, abiraterone, and darolutamide. In conclusion, our results cross-resistance exists between these therapies in advanced prostate suggest that apalutamide and darolutamide share similar resistant cancer. Here, we show that enzalutamide- and abiraterone-resistant mechanisms with enzalutamide and abiraterone. The AKR1C3/AR- prostate cancer cells are further cross-resistant to apalutamide and V7 complex confers cross-resistance to second-generation andro- darolutamide. Mechanistically, we have determined that the gen receptor–targeted therapies in advanced prostate cancer.

Introduction olutamide does not penetrate the blood–brain barrier, which lowers the toxicity of this drug. Nevertheless, both apalutamide and darolu- Treatment strategies for castration-resistant prostate cancer tamide still target the full-length AR and mechanisms of resistance and (CRPC) have evolved over the past decade. Before 2010, docetaxel cross-resistance with enzalutamide/abiraterone are largely unknown was the only treatment option for patients with CRPC (1). Following and need to be further investigated. the discovery that patients with CRPC remain dependent on androgen Constitutively active AR variants such as AR-V7 can activate receptor (AR) signaling (2–5), targeting AR became the major strategy distinct transcriptional programs and confer enzalutamide and abir- for the treatment of CRPC. The antiandrogen therapies, including aterone resistance (14–17). In addition to AR-V7, AKR1C3 activation abiraterone, enzalutamide, apalutamide, and darolutamide, signifi- has been shown to be a critical mechanism to promote resistance to cantly improved patient overall survival (OS) rate and are approved by both enzalutamide and abiraterone (18–20). The primary function of the FDA (6–9). However, the length of clinical response to these AKR1C3 is to catalyze androgen synthesis (19, 20). Our previous study treatments is limited and all patients eventually develop resistance. demonstrated a new function of AKR1C3: it can bind with AR-V7 and Thus, it is urgent to understand the underlying mechanisms of the promote its stabilization (18), suggesting that the AKR1C3/AR-V7 axis resistance. plays critical roles in conferring resistance to both enzalutamide and Apalutamide is approved for the treatment of nonmetastatic CRPC abiraterone. However, the role of the AKR1C3/AR-V7 axis in apalu- and metastatic castration-sensitive prostate cancer (mCSPC; ref. 6). A tamide and darolutamide is unknown. missense AR mutant, F877L, is linked to apalutamide resistance in Here, we demonstrate cross-resistance between both apalutamide preclinical models and has been detected in plasma DNA from and darolutamide therapies with enzalutamide and abiraterone treat- apalutamide-treated patients (10). However, recent clinical data sug- ment. Apalutamide and darolutamide target full length AR while gests the F877L and the T878A mutations might not cause either de having no effect on AR-V7 activity. Chronic apalutamide treatment novo or acquired apalutamide resistance (11). Data from phase I/II/III significantly elevated AKR1C3 and AR-V7 expression in prostate clinical trials show that darolutamide has strong activity and minimal cancer cells. Knockdown of AKR1C3 decreased AR-V7 expression toxicity (12, 13). In contrast to enzalutamide and apalutamide, dar- and significantly reversed the cross-resistance to enzalutamide, abiraterone, apalutamide, and darolutamide in resistant cells, suggesting that the AKR1C3/AR-V7 axis plays a critical role in cross-resistance 1 Department of Urologic Surgery, University of California, Davis, Sacramento, among the next-generation antiandrogen drugs. California. 2UC Davis Comprehensive Cancer Center, University of California, Davis, Sacramento, California. 3VA Northern California Health Care System, Sacramento, California. Materials and Methods Corresponding Authors: Chengfei Liu, University of California, Davis, Research 3 Cells lines and tissue culture Bldg, Suite 1300, 4645 2nd Ave, Sacramento, CA 95817. Phone/Fax: 916-703- 5469; E-mail: cffl[email protected]; and Allen C. Gao, E-mail: [email protected] C4-2B and CWR22Rv1 were maintained in RPMI1640 supplemented with 10% FBS, 100 U/mL penicillin, and 0.1 mg/mL streptomycin. Mol Cancer Ther 2020;19:1708–18 293 cells were maintained in DMEM supplemented with 10% FBS, 100 doi: 10.1158/1535-7163.MCT-20-0015 units/mL penicillin, and 0.1 mg/mL streptomycin. All experiments 2020 American Association for Cancer Research. with cell lines were performed within 6 months of receipt from ATCC

AACRJournals.org | 1708

Cross-Resistance in Antiandrogen Drugs

or resuscitation from cryopreservation. C4-2B cells were kindly pro- tion, Sigma-Aldrich). Tubulin was used as loading control. Following vided and authenticated by Dr. Leland Chung, Cedars-Sinai Medical secondary antibody incubation, immunoreactive proteins were visu- Center (Los Angeles, CA). C4-2B neo and C4-2B AKR1C3 cells were alized with an enhanced Chemiluminescence Detection System generated by stable transfection of C4-2B cells with either empty vector (Millipore). pcDNA3.1 or pcDNA3.1 encoding AKR1C3 and were maintained in RPMI1640 medium containing 300 mg/mL G418. C4-2B cells were Luciferase reporter assay chronically exposed to increasing concentrations of apalutamide (5–40 C4-2B cells were transfected with pGL3-PSA6.0-Luc reporters mmol/L) by passage in media containing apalutamide for >12 months along with AR-V7 in charcoal stripped FBS (CS-FBS) conditions. Cell in complete FBS and stored for further analysis. Cells resistant to lysates (35 mL per well) were used for measurement of luciferase apalutamide were referred to as C4-2B APALR (C4-2B apalutamide activity in a luminometer by first mixing the cell lysates (25 mL) with resistant). C4-2B MDVR (C4-2B enzalutamide resistant) and C4-2B 20 mL luciferase assay reagent for measuring firefly luciferase activity AbiR (C4-2B abiraterone resistant) are described previously (15). C4- and subsequently adding 20 mL Stop-Glo reagent for measuring Renilla 2B MDVR, C4-2B AbiR, and C4-2B APALR cells were maintained in luciferase activity. Data were normalized to Renilla luciferase activity. 20 mmol/L enzalutamide containing medium, 10 mmol/L abiraterone acetate containing medium, and 20 mmol/L apalutamide, respectively. Cell growth and survival assay Parental C4-2B cells were passaged alongside the resistant cells as an C4-2B parental, C4-2B AbiR, C4-2B MDVR, C4-2B APALR, or appropriate control. All cells were maintained at 37C in a humidified CWR22Rv1 cells were seeded on 12-well plates at a density of 0.3 105 incubator with 5% carbon dioxide. Enzalutamide, abiraterone acetate, cells/well in RPMI1640 media containing 10% FBS and treated with apalutamide, darolutamide, and indomethacin were purchased from enzalutamide, apalutamide, or darolutamide. Total cell number was Selleck Chemicals. determined after 0, 3, and 5 days. C4-2B MDVR cells were transiently transfected with siRNA targeting AR-V7 or a control siRNA, and then Plasmids and cell transfection treated with apalutamide or darolutamide for 3 days. C4-2B MDVR or For transfection of siRNA, cells were seeded at a density of 0.5 105 C4-2B APALR cells were seeded on 12-well plates at a density of 0.5 cells per well in 12-well plates or 2 105 cells per well in 6-well plates 105 cells/well in RPMI1640 media containing 10% FBS and cotreated and transfected with 20 nmol/L of siRNA targeting the AR-V7 with indomethacin and apalutamide/darolutamide in media contain- sequence (GUAGUUGUGAGUAUCAUGA; Dharmacon), or control ing FBS. The coefficient of drug interaction (CDI; ref. 21) was siRNA (catalog no. 12935300, Invitrogen) using Lipofectamine-iMAX calculated as follows: CDI ¼ AB/(A B). A is growth inhibition (Invitrogen). Briefly, cells were plated 1 day before the transfection. effect of indomethacin treatment. B is growth inhibition effect of siRNA (20 nmol/L) in serum-free Opti-MEM medium was mixed with apalutamide/darolutamide treatment. AB is the growth inhibition same volume of Opti-MEM medium containing Lipofectamine-iMAX effect of combination treatment. The CDIs were analyzed to determine and incubated for 5 minutes at room temperature, prior to be added to the synergism of two drugs combination treatment (CDI value <1, ¼ 1, cells. pcDNA3.1-AR-V7 or AKR1C3 were constructed by PCR ampli- or >1 indicated that the drugs were synergistic, additive, or antago- fication followed by subcloning into the pcDNA3.1 vector. Cells were nistic). C4-2B neo and C4-2B AKR1C3 cells were seeded on 12-well transfected by expressing plasmids for pcDNA3.1 vector, AR-V7, and plates at a density of 0.3 105 cells/well in RPMI1640 media contain- AKR1C3 using Lipofectamine 2000 (Invitrogen). Briefly, cells were ing 10% FBS and treated with different doses of apalutamide or plated 1 day before the transfection. Plasmids (0.5–1 mg) in serum-free darolutamide for 3 days. Total cell number was counted and the cell Opti-MEM medium were mixed with same volume of Opti-MEM survival rate (%) was calculated. Cell survival rate (%) ¼ (treatment medium containing Lipofectamine2000 and incubated for 20 minutes group cell number/control group cell number) 100%. at room temperature, prior to be added to cells. The effect of siRNA- mediated gene silencing and ectopic expression plasmids were exam- Clonogenic assay ined using qRT-PCR or Western blot analysis 2–3 days after trans- C4-2B parental, C4-2B MDVR, C4-2B AbiR, or CWR22Rv1 cells fection. Lentiviral plasmid encoding short hairpin RNA (shRNA) were treated with DMSO, 20 mmol/L enzalutamide, 20 mmol/L targeting AKR1C3 (TRCN0000026561) was purchased from Sigma- apalutamide, or 5 mmol/L darolutamide in media containing Aldrich. pLenti-GFP Lentiviral control vector were used as control. 10% complete FBS. C4-2B APALR cells were treated with 0, 20, or Lentiviral particles were produced in 293T cells after cotransfection of 40 mmol/L indomethacin alone or in combination with 20 mmol/L the lentivirus vectors, psPAX2 and pMD2.G in 10 cm dishes. The apalutamide. All cell lines were plated at equal densities (600 cells/ lentiviral plasmids were kindly provided by Dr. Hong-Wu Chen at well) and all treatments lasted for 14 days. The colonies were rinsed University of California, Davis (Davis, CA). The lentivirus containing with PBS before being stained with 0.5% crystal violet/4% form- medium was collected and cells were infected. aldehyde for 30 minutes and the number of colonies was counted as described previously (22). Protein extraction and Western blotting Whole-cell protein extracts were resolved on SDS-PAGE and Real-time qRT-PCR proteins were transferred to nitrocellulose membranes. After blocking Total RNA was extracted using TRizol Reagent (Invitrogen). cDNA for 1 hour at room temperature in 5% milk in PBS/0.1% Tween-20, was prepared after digestion with RNase-free RQ1 DNase (Promega) membranes were incubated overnight at 4C with the indicated and then subjected to real-time reverse transcription-PCR (RT-PCR) primary antibodies: AR (441, 1:1,000 dilution, Santa Cruz Biotech- using Sso Fast Eva Green Supermix (Bio-Rad) according to the nology, Inc.), AR (N-20, 1:1,000 dilution, Santa Cruz Biotechnology, manufacturer's instructions and as described previously (23). Each Inc.), AR-V7 (AG10008, Mouse mAb, 1:1,000 dilution, Precision reaction was normalized by coamplification of actin. Triplicates of Antibody), AKR1C3 (A6229, 1:1,000 dilution, Sigma-Aldrich), c- samples were run on default settings of a Bio-Rad CFX-96 Real-Time Myc (N262, 1:1,000 dilution, Santa Cruz Biotechnology, Inc.), and Cycler. Primers used for RT-PCR were: AKR1C3: 50-gagaagtaa- Tubulin (T5168, Monoclonal Anti-a-Tubulin antibody, 1:5,000 dilu- agctttggaggtcaca-30 (forward) and 50-caacctgctcctcattattgtataaatga-30

AACRJournals.org Mol Cancer Ther; 19(8) August 2020 1709

Zhao et al.

(reverse); AKR1C1/2: 50-ggtcacttcatgcctgtcct-30 (forward) and 50- As shown in Fig. 2A, enzalutamide, apalutamide, and darolutamide actctggtcgatgggaattg-30 (reverse); HSD3B: 50-cgggcccaactcctacaag-30 suppressed both full-length AR and AR-V7 protein expression in C4- (forward) and 50-ttttccagaggctcttcttcgt-30 (reverse); SRD5A1:50-acggg- 2B parental cells. Darolutamide was the most effective drug at catcggtgcttaat-30 (forward) and 50-ccaacagtggcataggctttc-30 (reverse); decreasing full-length AR in these cells compared with enzalutamide HSD17B2: 50-tttgccggagttttgaatgaa-30 (forward) and 50-gcaggttcttcg- and apalutamide. C4-2B MDVR cells expressed higher level of full- caattcct-30 (reverse); and actin: 50-agaactggcccttcttggagg-30 (forward) length AR and AR variants than C4-2B parental cells. Enzalutamide, and 50-gtttttatgttcctctatggg-30 (reverse). apalutamide, and darolutamide had less effect on full-length AR yet no effect on AR-V7 expression in C4-2B MDVR cells. However, total Statistical analysis AR variants' expression was inhibited by enzalutamide and darolu- The data are presented as means SD of the mean from three tamide when detected by pan-AR antibody in C4-2B MDVR cells. independent experiments. Statistical analyses were performed with None of the antiandrogens had effect on full-length AR and AR-V7 SPSS16.0. Differences between individual groups were analyzed by two expression in CWR22Rv1 cells. Promoter activity assays revealed tailed Student t test or one-way ANOVA followed by the Scheffe that enzalutamide, apalutamide, and darolutamide significantly procedure for comparison of means. P < 0.05 was considered statis- suppressed DHT-induced but not AR-V7–induced AR transcrip- tically significant. tional activity (Fig. 2B). Niclosamide, a previously identified AR-V7 inhibitor (15), significantly suppressed both DHT- and AR-V7– induced PSA promoter luciferase activity (Fig. 2B). To further test Results whether AR-V7 overexpression is one of the mechanisms contrib- Enzalutamide/abiraterone-resistant prostate cancer cells are uting to apalutamide and darolutamide cross-resistance, AR-V7 was cross-resistant to apalutamide and darolutamide knocked down in C4-2B MDVR cells followed by treatment with Apalutamide and darolutamide are new-generation antiandrogens apalutamide and darolutamide. As shown in Fig. 2C, AR-V7 knock- that bind the AR with high affinity and inhibit AR transcriptional down reduced cell growth in C4-2B MDVR cells, and apalutamide or activity and nuclear translocation (12, 24). To determine the response darolutamide further inhibited cell growth in AR-V7–knockdown of enzalutamide/abiraterone-resistant cells to apalutamide and dar- groups. We also determined the effects of AR-V7 knockdown by olutamide, C4-2B parental, C4-2B AbiR, C4-2B MDVR, and Western blot analysis. As shown in Fig. 2D, AR-V7 was successfully CWR22Rv1 cells were treated with different doses of enzalutamide, knocked down by AR-V7 siRNA. AR-V7 knockdown significantly apalutamide, or darolutamide. We found that C4-2B MDVR, C4-2B suppressed c-Myc expression, but had no effect on AKR1C3 expres- AbiR, and CWR22Rv1 cells were resistant to all the antiandrogen sion. Similar results were observed in CWR22Rv1 cells showing that treatments. As shown in Fig. 1A,20mmol/L enzalutamide inhibited knockdown of AR-V7 significantly resensitized CWR22Rv1 cells to 41%, 6.3%, 5.5%, and 21%; 40 mmol/L apalutamide inhibited 24%, apalutamide and darolutamide treatment (Fig. 2E). In addition, it 5.8%, 5.7%, and 15%; 10 mmol/L darolutamide inhibited 54%, 26%, reduced c-Myc but no effect on AKR1C3 expression (Fig. 2F). Taken 28%, and 31% of cell growth in C4-2B, C4-2B MDVR, C4-2B AbiR, together, these results suggest that AR-V7 confers apalutamide and and CWR22Rv1 cells, respectively. In C4-2B parental cells, the rank of darolutamide resistance. cell growth inhibition is darolutamide > enzalutamide > apalutamide. Apalutamide is least potent compared with enzalutamide and dar- Targeting AKR1C3 resensitizes resistant cells to apalutamide olutamide in the inhibition of cellular growth of C4-2B parental cells. and darolutamide through AR-V7 inhibition We then determined growth inhibition at different timepoints by these AKR1C3-mediated intracrine androgen biosynthesis is one of the antiandrogens in different cell lines. As shown in Fig. 1B, C4-2B critical mechanisms of enzalutamide resistance (19). Our recent parental cells respond to all the antiandrogen treatments in a time- publication established that AKR1C3 binds to AR-V7 and stabilizes dependent manner; however, C4-2B MDVR, C4-2B AbiR, and the protein level in resistant prostate cancer cells (18). To test whether CWR22Rv1 cells were resistant to all antiandrogen treatments at all the AKR1C3/AR-V7 axis confers cross-resistance to apalutamide and timepoints. At 5 days treatment, 20 mmol/L enzalutamide inhibited darolutamide, we determined whether knockdown of AKR1C3 resen- 44%, 7.3%, 11%, and 19%; 40 mmol/L apalutamide inhibited 23%, 3.0%, sitizes the resistant cells to apalutamide and darolutamide. As shown 2.9%, and 14%; 5 mmol/L darolutamide inhibited 46%, 13.5%, 11%, and in Fig. 3A, knockdown of AKR1C3 by lenti-AKR1C3 shRNA inhibited 13.3% of cell growth in C4-2B, C4-2B MDVR, C4-2B AbiR, and cell growth. AKR1C3 knockdown combined with apalutamide/dar- CWR22Rv1 cells, respectively. These data were confirmed using olutamide further reduced cell number compared with the control colony formation assays. We found that C4-2B parental cells formed group. We further confirmed that knockdown of AKR1C3 signifi- fewer colonies when treated with enzalutamide, apalutamide, or cantly downregulated AR, AR-V7, and c-Myc expression in C4-2B darolutamide, whereas C4-2B MDVR, C4-2B AbiR, and CWR22Rv1 MDVR cells (Fig. 3B). Similar results were found in CWR22Rv1 cells. cells were far less affected by antiandrogen treatment (Fig. 1C). Taken Knockdown of AKR1C3 resensitized CWR22Rv1 cells to apalutamide together, these data suggest enzalutamide/abiraterone-resistant pros- and darolutamide treatment (Fig. 3C), and significantly hampered AR, tate cancer cells are cross-resistant to both apalutamide and AR-V7, and c-Myc expression (Fig. 3D). To further confirm AKR1C3 darolutamide. is involved in apalutamide and darolutamide resistance, we used C4- 2B-AKR1C3 cells to test whether exogenous expression of AKR1C3 AR-V7 confers apalutamide and darolutamide resistance induces apalutamide or darolutamide resistance. As shown in Fig. 3E, Previous studies showed that AR-V7, a truncated variant of the C4-2B-AKR1C3 cells exhibited greater resistance to apalutamide and AR, confers enzalutamide resistance (14, 15). To investigate the darolutamide than C4-2B-neo cells. Western blot analysis showed that potential mechanisms of cross-resistance between enzalutamide, both AR and AR-V7 levels were increased in C4-2B-AKR1C3 cells apalutamide, and darolutamide, we determined levels of full- compared with those in the C4-2B-neo cells (Fig. 3F). Finally, length AR, AR-V7, and AKR1C3 following different antiandrogen AKR1C3 inhibitor, indomethacin, significantly enhanced both apa- treatments in C4-2B parental, C4-2B MDVR, and CWR22Rv1 cells. lutamide and darolutamide treatment in C4-2B MDVR and

1710 Mol Cancer Ther; 19(8) August 2020 MOLECULAR CANCER THERAPEUTICS

Cross-Resistance in Antiandrogen Drugs

A C4-2B parental C4-2B MDVR C4-2B ABIR CWR22Rv1 120 120 120 100 100 100 80 * 80 * 80 60 60 60 * 40 40 40 20 20 20 Cell survival rate (%) Cell survival rate (%) Cell survival rate (%) 0 0 0 0 2.5 5 10 20 0 5 10 20 40 0 1 2.5 5 10 Enzalutamide (μmol/L) Apalutamide (μmol/L) Darolutamide (μmol/L)

B DMSO ENZA APAL DARO

C4-2B C4-2B C4-2B CWR22 ) ) ) ) 80 4 4 4 4 100 100 300 parental MDVR AbiR Rv1 80 80 250 × × × × ( 10 (10 60 200 60 * 60 * 40 150 40 40 100 20 20 20 50 Cell number Cell number ( 10 Cell number Cell number ( 10 0 Cell number 0 0 0 035 035 035 035 Days Days Days Days C Colony DMSO ENZA APAL DARO numbers 100 150 200 250 300 50 0

APAL C4-2B parental

C4-2B * parental * *

APAL C4-2B MDVR C4-2B MDVR

AbiR APAL C4-2B AbiR C4-2B Rv1

Rv1 APAL CWR22 CWR22

Figure 1. Enzalutamide/abiraterone-resistant prostate cancer cells are cross-resistant to apalutamide and darolutamide. A, C4-2B parental, C4-2B AbiR, C4-2B MDVR, and CWR22Rv1 cells were treated with different concentrations of enzalutamide, apalutamide, or darolutamide for 3 days. Total cell numbers were counted and cell survival rate was calculated. B, C4-2B parental, C4-2B MDVR, C4-2B AbiR, and CWR22Rv1 cells were treated with 20 mmol/L enzalutamide, 20 mmol/L apalutamide, or 5 mmol/L darolutamide for 0, 3, and 5 days and total cell numbers were determined. C, The clonogenic ability of C4-2B parental, C4-2B MDVR, C4-2B AbiR, and CWR22Rv1 cells treated with 20 mmol/L enzalutamide, 20 mmol/L apalutamide, or 5 mmol/L darolutamide was analyzed. Abi, abiraterone acetate; APAL, apalutamide; DARO, darolutamide; ENZA, enzalutamide (, P < 0.05).

CWR22Rv1 cells in a time-dependent manner (Fig. 3G). At 5 days gistic interactions (CDI < 1). In light of these results, our data indicated treatment, the CDIs of 20 mmol/L indomethacin with 5 mmol/L that the AKR1C3/AR-V7 axis is one of the major mechanisms apalutamide or 20 mmol/L darolutamide combination treatment were conferring enzalutamide/apalutamide/darolutamide cross-resistance 0.56 and 0.64, respectively, in C4-2B MDVR cells, suggesting syner- in advanced prostate cancer.

AACRJournals.org Mol Cancer Ther; 19(8) August 2020 1711

Zhao et al.

A B 20 Figure 2. 16 AR-V7 confers apalutamide and darolutamide C4-2B C4-2B CWR 12 resistance. A, C4-2B parental, C4-2B MDVR, and m parental MDVR 22Rv1 RLU 8 CWR22Rv1 cells were treated with 20 mol/L * * m 4 * * enzalutamide, 20 mol/L apalutamide, or 5 * mmol/L darolutamide for 3 days. Whole-cell 0 * lysates were collected and subjected to West- APAL APAL DMSO ENZA APAL DARO DMSO ENZA DARO DARO ENZA DMSO ern blot analysis. B, C4-2B cells were trans- DMSODMSO AR-V7 fected with PSA luciferase promoter in CS-FBS

DARO 5 µmol/L NIC 0.5 µmol/L condition for 24 hours, followed by treatment ENZA APAL20 µmol/L 20 µmol/LDARODARO 10 µmol/L 20 µmol/L AR-FL with 1 nmol/L DHT with or without 20 mmol/L Short exposure 1 nmol/L DHT AR-Vs enzalutamide, 20 mmol/L apalutamide, 5, 10, or 7 n.s. 20 mmol/L darolutamide, or 0.5 mmol/L niclo- 6 samide overnight. Whole-cell lysates were col- AR-FL Long exposure 5 lected and subjected to luciferase assay (top). 4 C4-2B cells were cotransfected with PSA lucif- AR-Vs RLU 3 2 * erase promoter and AR-V7 in CS-FBS conditions AKR1C3 1 for 24 hours, followed by treatment with 20 0 mmol/L enzalutamide, 20 mmol/L apalutamide, Tubulin 5, 10, or 20 mmol/L darolutamide, or 0.5 mmol/L µmol/Lµ µmol/L DMSODMSO mol/L niclosamide overnight. Whole-cell lysates were collected and subjected to luciferase assay (bot- µ DAROµ 5 NIC 0.5 µmol/L ENZAAPAL 20 mol/L 20 mol/LDARODARO 10 20 tom). C, C4-2B MDVR cells were transiently AR-V7 transfected with control siRNA or AR-V7 siRNA. C D Following transfection, cells were treated with Control AR-V7 20 mmol/L apalutamide or 5 mmol/L daroluta- C4-2B MDVR siRNA siRNA mide and cell numbers were determined after 40 5 days. D, Whole-cell lysates were collected and subjected to Western blot analysis. E, ) APAL DARO APAL DMSO DARO DMSO 4 CWR22Rv1 cells were transiently transfected 0 30 1 AR-V7

× with control siRNA or AR-V7 siRNA. Following DMSO * transfection, cells were treated with 20 mmol/L 20 APAL 20 μmol/L AR-FL apalutamide or 5 mmol/L darolutamide and DARO 5 μmol/L AR-Vs cell numbers were determined after 5 days. F, 10 Whole-cell lysates were collected and subjected c-Myc to Western blot analysis. Abi, abiraterone ace- Cell number ( 0 AKR1C3 tate; APAL, apalutamide; DARO, darolutamide; ENZA, enzalutamide; NIC, niclosamide; n.s., sicontrol siAR-V7 Tubulin not signiﬁcant (, P < 0.05).

Control AR-V7 siRNA siRNA E CWR22Rv1 F 200 APAL ) APAL DARO DMSO DARO DMSO 4 150 AR-V7 × AR-FL DMSO 100 * APAL 20 μmol/L AR-Vs number ( 10 DARO 5 μmol/L 50 c-Myc Cell AKR1C3 0 sicontrol siAR-V7 Tubulin

Chronic apalutamide treatment in C4-2B cells upregulates the growth (Fig. 4B) and colony formation assays (Fig. 4C and D) steroid biosynthesis pathway conﬁrmed that C4-2B APALR cells exhibited robust resistance not Apalutamide is a new next-generation antiandrogen with a molec- only to apalutamide but also to darolutamide, enzalutamide, and ular structure very similar to enzalutamide (6). To investigate abiraterone. We then examined the transcriptional alterations of apalutamide resistance, we have successfully generated a C4-2B steroid hormone biosynthesis–related genes in C4-2B APALR cells apalutamide–resistant cell line (C4-2B APALR). As shown compared with the parental cells. The steroid hormone biosynthesis– in Fig. 4A, C4-2B parental cells were sensitive to apalutamide related genes were signiﬁcantly upregulated in C4-2B APALR cells treatment in a dose-dependent manner; however, C4-2B APALR compared with C4-2B parental cells (Fig. 4E). Results from Western cells were resistant to the treatment. To test whether cross-resistance blot analysis indicated that the protein levels of both AKR1C3 exists between apalutamide and other androgen signal-targeting and AR-V7 were overexpressed in C4-2B APALR cells. However, agents in this cell model, C4-2B APALR cells were treated with these levels were lower than those in C4-2B MDVR cells (Fig. 4F). apalutamide, darolutamide, enzalutamide, and abiraterone. Both cell These results strongly imply that the AKR1C3/AR-V7 axis is critical

1712 Mol Cancer Ther; 19(8) August 2020 MOLECULAR CANCER THERAPEUTICS

Cross-Resistance in Antiandrogen Drugs

Figure 3. A C4-2B MDVR B Lenti Lenti 25 shcontrol shAKR1C3 AKR1C3 confers apalutamide and darolutamide * resistance through AR-V7 regulation. A and B, C4- ) 2B MDVR cells were infected with lenti-control 4 20 APAL DARO APAL DMSO DARO DMSO shRNA or lenti-AKR1C3 shRNA and then treated with × DMSO 20 mmol/L apalutamide or 5 mmol/L darolutamide 15 * AR-V7 APAL 20 μmol/L and cell numbers were determined after 5 days. 10 AR-FL Whole-cell lysates were collected and subjected to DARO 5 μmol/L AR-Vs Western blot analysis. C and D, CWR22Rv1 cells were 5 infected with lenti-control shRNA or lenti-AKR1C3 Cell number ( 10 c-Myc shRNA and then treated with 20 mmol/L apalutamide 0 or 5 mmol/L darolutamide and cell numbers were Lenti Lenti AKR1C3 shAKR1C3 determined after 5 days. Whole-cell lysates were shcontrol Tubulin collected and subjected to Western blot analysis. E and F, C4-2B neo or C4-2B AKR1C3 cells were treated C D Lenti Lenti CWR22Rv1 shcontrol shAKR1C3 with different concentrations of apalutamide or dar- 80 olutamide for 3 days, total cell numbers were * )

4 70 counted, and cell survival rate (%) was calculated.

60 APAL APAL DARO DMSO DARO DMSO Whole-cell lysates were collected and subjected to × DMSO Western blot analysis. G, C4-2B MDVR and 50 µ AR-V7 CWR22Rv1 cells were treated with 20 mmol/L apa- 40 APAL 20 mol/L AR-FL lutamide or 5 mmol/L darolutamide or 20 mmol/L 30 * DARO 5 µmol/L AR-Vs indomethacin, alone or in combination, total cell 20 numbers were determined on 3 and 5 days. APAL, Cell number ( 10 10 c-Myc apalutamide; DARO, darolutamide; INDO, indomethacin (,P< 0.05). 0 AKR1C3 Lenti Lenti shcontrol shAKR1C3 Tubulin E 120 C4-2B neo C4-2B AKR1C3 F 100 * * * * * 80 AR-V7 60 AR-FL 40 AR-Vs ell survival rate (%)

C 20 AKR1C3 0 DMSO 5 10 20 20 40 Tubulin DARO (µmol/L) APAL (µmol/L) G C4-2B MDVR CWR22Rv1 ) ) 4 60 DMSO 4 250 DMSO 0 × × 50 200 APAL ( 1 APAL 40 * * DARO 150 DARO 30 INDO INDO 20 100 * * number 10 APAL+INDO 50 APAL+INDO Cell number ( 10 0 DARO+INDO Cell 0 DARO+INDO 035 035 Days Days in conferring cross-resistance between apalutamide and other in C4-2B MDVR cells (18). Finally, C4-2B APALR cells were treated antiandrogens. with the AKR1C3 inhibitor, indomethacin, or apalutamide, alone and in combination. Indomethacin effectively resensitized C4-2B APALR Targeting AKR1C3 resensitize apalutamide-resistant cells to cells to apalutamide treatment (Fig. 5C), the CDIs of 20 mmol/L apalutamide and enzalutamide apalutamide with 20 mmol/L indomethacin or 40 mmol/L indometh- To confirm the contribution of the AKR1C3/AR-V7 axis in cross- acin combination treatment were 0.61 and 0.63, respectively, suggest- resistance between enzalutamide and apalutamide, C4-2B APALR ing drug synergism (CDI < 1). The results were confirmed by clono- cells were treated with enzalutamide or apalutamide with or without genic assay. Apalutamide (20 mmol/L) has no effects on colony the knockdown of AKR1C3 by lenti-AKR1C3 shRNA (Fig. 5A). formation in C4-2B APALR cells, single indomethacin treatment Growth of C4-2B APALR cells was largely inhibited by AKR1C3 reduced the colony formation, and combination treatment signifi- knockdown. AKR1C3 knockdown in combination with enzalutamide/ cantly suppressed colony size and numbers (Fig. 5D and E). Collec- apalutamide treatment further reduced cell growth. The knockdown of tively, our data suggest that targeting AKR1C3 in apalutamide- AKR1C3 was accompanied with reduced expression levels of AR-FL, resistant cells could efficiently reverse cross-resistance between apa- AR-V7, and c-Myc (Fig. 5B), which is consistent with what is observed lutamide and enzalutamide.

AACRJournals.org Mol Cancer Ther; 19(8) August 2020 1713

Zhao et al.

A B 120 140 C4-2B parental C4-2B APALR 100 120 80 * 100 * * 60 80 * C4-2B parental 60 40 * C4-2B APALR 40 20 Cell survival rate (%) 20 0 Cell survival rate (%) 02.55 102040 0 Apalutamide (μmol/L) DMSO APAL DARO ENZA ABI D C DMSO APAL DARO ENZA ABI 120 C4-2B parental C4-2B APALR 100 80 C4-2B

parental 60 * * 40 inhibition rate * * y 20

C4-2B 0 Colon APALR DMSO APAL DARO ENZA ABI

E F 5 * * 4 * C4-2B-MDVR C4-2B-APALR 3 C4-2B * C4-2B 2 * AKR1C3 APALR expression 1 AR-V7 Relative mRNA 0 AR-FL AKR1C1/2 AKR1C3 HSD3B SRD5A1 HSD17B2 Tubulin

Figure 4. Chronic apalutamide treatment in C4-2B cells activates the steroid biosynthesis pathway through AKR1C3 upregulation. A, C4-2B parental and C4-2B APALR cells were treated with different concentration of apalutamide for 3 days, total cell numbers were determined, and cell survival rate was calculated. B, C4-2B parental and C4-2B APALR cells were treated with 20 mmol/L apalutamide, 5 mmol/L darolutamide, 20 mmol/L enzalutamide, or 10 mmol/L abiraterone and cell numbers were determined after 3 days. C and D, The clonogenic ability of C4-2B parental and C4-2B APALR cells treated with 20 mmol/L apalutamide, 5 mmol/L darolutamide, 20 mmol/L enzalutamide, or 10 mmol/L abiraterone was analyzed. E, The genes involved in steroid hormone biosynthesis pathway were determined by qRT-PCR. F, Whole-cell lysates were collected from C4-2B parental, C4-2B MDVR, and C4-2B ARALR cells and subjected to Western blot analysis. ABI, abiraterone acetate; APAL, apalutamide; DARO, darolutamide; ENZA, enzalutamide (, P < 0.05).

Discussion overcome resistance. In addition, more therapeutic options become available with additional possibilities of diverse sequential treatments. Our study provides the first experimental evidence to show that Hence, identifying the key factors responsible for the cross-resistance cross-resistance exists between enzalutamide/abiraterone and the among different therapies is essential for determining the optimal newly FDA-approved apalutamide and darolutamide. We demon- long-term sequential treatment schemes for patients with metastatic strated that AKR1C3/AR-V7 axis plays a critical role in cross- castration-resistant prostate cancer (mCRPC). resistance among the next-generation antiandrogen drugs. Results The mainstay treatments for advanced prostate cancer consist of two from this study and our previous publications (25) complete a major parts, that is, AR-targeting therapy and cytotoxic chemotherapy. blueprint regarding the cross-resistance. These findings strengthened Clinical data have implied the presence of cross-resistance between the concept that because AR-targeting drugs share similar action of enzalutamide and abiraterone, while docetaxel-based chemotherapy is mechanisms, there is nonnegligible cross-resistance among them. still effective for patients with mCRPC after the failure of enzalutamide The past decade has witnessed the remarkable development of the or abiraterone and vice versa (26–29). These phenomena were validated next-generation androgen-targeting agents in the treatment of in vitro by our previous work using the drug-resistant cell line models we advanced prostate cancer, which vastly improved the survival out- have developed (25). Apalutamide and darolutamide were recently comes of patients with prostate cancer (6–9). However, challenges approved for the treatment of nonmetastatic (M0) CRPC in clinical remain that after a period of efficient treatment, tumor cells inevitably settings and significantly extendedthe metastasis-free survival and OS in become resistant. Thus, unveiling the mechanism underlying drug patients with CRPC over placebo (6, 13). Recently, apalutamide was also resistance is of great importance to optimize treatment decisions and approved by the FDA to treat patient with mCSPC due to the fact that

1714 Mol Cancer Ther; 19(8) August 2020 MOLECULAR CANCER THERAPEUTICS

Cross-Resistance in Antiandrogen Drugs

ABshcontrol shAKR1C3 C C4-2B APALR * )

4 70 * APAL APAL ENZA DMSO ENZA 60 DMSO ×

) 50 4 50 * DMSO AR-V7 * 40 40 × ENZA 30 30 * AR-FL 20 APAL 10

20 shAKR1C3 AR-Vs Cell number ( 10 0 * ENZA+shAKR1C3 10 c-Myc Cell number ( 10 APAL+shAKR1C3 0 035 AKR1C3 Days Tubulin

DEIndomethacin (μmol/L) 0 20 40 * 350 * 300 250 * * 200 150 DMSO 100

Colony numbers 50 0 APAL

Figure 5. Targeting AKR1C3 resensitizes apalutamide-resistant cells to apalutamide and enzalutamide. A, C4-2B APALR cells were treated with 20 mmol/L enzalutamide, 20 mmol/L apalutamide, or infected with lenti-AKR1C3 shRNA, alone or in combination. Total cell numbers were determined on 3 and 5 days. B, C4-2B APALR cells were infected with lenti-control shRNA or lenti-AKR1C3 shRNA and then treated with 20 mmol/L enzalutamide or 20 mmol/L apalutamide. After 5 days, whole-cell lysates were collected and subjected to Western blot analysis. C, C4-2B APALR cells treated with 0, 20, and 40 mmol/L indomethacin with or without 20 mmol/L apalutamide and cell numbers were determined after 5 days. D and E, The clonogenic ability of C4-2B ARALR cells treated with 0, 20, and 40 mmol/L indomethacin with or without 20 mmol/L apalutamide was analyzed. APAL, apalutamide; ENZA, enzalutamide; INDO, indomethacin (,P< 0.05). the drug was able to prolong the radiographic progression-free survival However, both enzalutamide- and abiraterone-resistant cell models are and OS in the setting of this patient population (30). Darolutamide still cross-resistant to apalutamide and darolutamide treatment; and expresses better in vitro activity than other antiandrogens, which is importantly, acquired apalutamide-resistant prostate cancer cells are consistent with our current findings. Darolutamide shows higher cross-resistant to all other antiandrogens. Because apalutamide currently binding affinity to the AR than enzalutamide and apalutamide is approved to treat patient with mCSPC with androgen deprivation in vitro and can target AR mutants, such as F877L and W742L, which therapy (ADT), the fact we should not overlook is after these patients drives other antiandrogens resistance (31). Nevertheless, a fact that become resistant to apalutamide, they are expected to be resistant to should be noticed is that darolutamide has worse pharmacokinetic other antiandrogens as well, such as enzalutamide, darolutamide, and parameters in vivo compared with enzalutamide and apaluta- abiraterone based on our preclinical data. mide (32, 33). In the recently published phase III clinical trial, the Both AR-V7 and AKR1C3 play important roles in leading disease darolutamide oral dose is 1,200 mg per day, which is 7.5-fold and 5-fold progression and conferring drug-resistant prostate cancer higher dose than the enzalutamide or apalutamide oral doses, respec- cells (14–19, 23, 36–38). AR-V7 is a truncated AR variant that lacks tively (NCT02799602). Although darolutamide was approved in the the ligand-binding domain and is able to directly translocate into the treatment of patients with prostate cancer, to address its bioavailability is nucleus without activation by androgen (39). Thus, current antian- still a necessary topic in future research. drogen treatment cannot target AR-V7, and the tumor cells harboring Apalutamide has a very similar chemical structure to enzalutamide AR-V7 may escape full-length AR-targeting therapies and survive. andRD162. Compared with RD162, apalutamide only differs by a carbon Moreover, detection of AR-V7 in circulating tumor cells is strongly atom replaced with a nitrogen atom in one of the phenyl rings (4, 24). On correlated to enzalutamide and abiraterone resistance clinically (14). the basis of the similar structures of apalutamide to enzalutamide, we In our previous studies, we have shown that niclosamide, an AR-V7 postulated that both drugs might share the same resistant mechanisms. inhibitor, significantly reverses drug resistance in enzalutamide and Intriguingly, darolutamide has the potential to overcome enzalutamide abiraterone therapy (15, 23, 36). A clinical trial using niclosamide in resistance by targeting AR mutants, such as F877L, which uses enzalu- combination with abiraterone to treat patient with enzalutamide- tamide or apalutamide as an AR agonist instead of antagonist (34, 35). In resistant prostate cancer is currently ongoing (NCT02807805). How- our study, we found that darolutamide is more effective than enzaluta- ever, opposition to AR-V7's role in the development of antiandrogen- mide and apalutamide at the same dose in C4-2B parental cells in vitro. resistant CRPC is surging, and the exact position of AR variants in

AACRJournals.org Mol Cancer Ther; 19(8) August 2020 1715

Zhao et al.

contributing to therapeutic resistance is in debates (40). By using According to the National Comprehensive Cancer Network guide- reagents that can bind to AR but not AR-V7 and induce AR degra- lines (49), among the four next-generation AR-targeting agents, dation, Steven and colleagues showed that AR instead of AR-V7 is the enzalutamide has the widest scope of application and could be used prominent driver of enzalutamide resistance (41), which challenged in mCSPC, M0 CRPC, or mCRPC stage. Abiraterone is approved in the role of AR-V7 in enzalutamide resistance. In addition, previous patients with mCSPC and mCRPC, whereas apalutamide is approved study indicated AR-V7 remains dependent on full-length AR in gain of in patients with mCSPC and M0 CRPC. So far, darolutamide is only function (42). Intriguingly, AR-V7 can form a heterodimer with full- approved in M0 CRPC stage. However, several phase III trials length AR in controlling gene regulation. Targeting AR might block (NCT02799602 and NCT02489318) are now ongoing and the appli- the downstream effector genes of the AR/AR-V7 dimers. Above all, the cation spectrum of these drugs will certainly enlarge in the near future. complex role of AR-V7 in the development of drug-resistant CRPC Thus, our study provides important preclinical clues about the values remains to be revealed in future research. of AKR1C3 and AR-V7 in distinguishing the optimal candidates of On the other hand, AKR1C3, a multifunctional enzyme, induces these AR-targeting treatments from the potential nonresponders. In enzalutamide and abiraterone resistance by converting the weak addition, we showed that the combination of antiandrogens and the androgens androstenedione and 5a-androstenedione to the more AKR1C3 inhibitor indomethacin effectively reverses the drug resis- active androgens testosterone and DHT (19, 23). Various inhibitors tance, which may be a new treatment strategy to maximize the have been developed to target AKR1C3 activation in xenograft tumor therapeutic efficacy. models (43), including indomethacin (44), GTx-560 (45), and Collectively, our study provides the information for clinicians to ASP9521 (46), however, most untested in the clinical settings so far. have a better understanding about the mechanisms underlying drug Thus, there is a great unmet need to engage AKR1C3 inhibitors for resistance of different next-generation AR-targeting drugs. The results clinical trial initiation. Previously, we have reported that indometh- demonstrated that enzalutamide- and abiraterone-resistant cells are acin, a commonly used NSAID, could efficiently inhibit AKR1C3 cross-resistant to apalutamide and darolutamide, and that AKR1C3/ activity both in vivo and in vitro (23). Given the promising effect of AR-V7 plays a critical role in conferring the cross-resistance among indomethacin, a phase I/II trial (NCT02935205) aiming at exploring these next-generation antiandrogen drugs. its effect in treating patients with mCRPC in combination with enzalutamide is ongoing. Moreover, we recently found that over- Disclosure of Potential Conflicts of Interest expression of AKR1C3 increases AR-V7 expression (18). This finding C.P. Evans reports receiving speakers bureau honoraria from Astellas, Janssen, was supported clinically by a study where positive AKR1C3 expression Ferring International, and Sanofi. No potential conflicts of interest were disclosed by from prostate biopsy samples was associated with higher positivity of the other authors. P ¼ AR-V7 [7/38 (18.4%) vs. 4/76 (5.3%); 0.025; ref. 47]. Consistent Authors’ Contributions fi with our previous ndings in C4-2B MDVR cells, here we showed that Conception and design: C.P. Evans, A.C. Gao, C. Liu inhibition of AKR1C3 decreases AR and AR-V7 levels in C4-2B Development of methodology: A.C. Gao, C. Liu APALR cells. AKR1C3 has been shown to bind to AR-FL as a Acquisition of data (provided animals, acquired and managed patients, provided coactivator in a previous study (45). Our subsequent study found facilities, etc.): J. Zhao, S. Ning, J.C. Yang, C.M. Armstrong, A.P. Lombard, A.C. Gao, that AKR1C3 binds with AR-V7 and promotes AR-V7 protein sta- C. Liu bilization via the ubiquitin-mediated proteasome pathway. Notably, Analysis and interpretation of data (e.g., statistical analysis, biostatistics, fi computational analysis): J. Zhao, J.C. Yang, A.P. Lombard, C.P. Evans, A.C. Gao, we found that overexpression of AKR1C3 signi cantly stabilizes both C. Liu AR-FL and AR-V7 proteins. Inhibition of AKR1C3 by indomethacin Writing, review, and/or revision of the manuscript: J. Zhao, J.C. Yang, C.P. Evans, reduced both AR-FL and AR-V7 protein half-lives, which could be A.C. Gao, C. Liu reversed by proteasome inhibitor MG132 (18). Thus, AKR1C3 may Administrative, technical, or material support (i.e., reporting or organizing data, stabilize AR similar to AR-V7 via the same mechanism through constructing databases): J. Zhao, W. Lou, J.C. Yang, L.S. D'Abronzo, A.C. Gao, C. Liu ubiquitin–proteasome system regulation. In addition, we previously Study supervision: A.C. Gao, C. Liu showed that knockdown of AR-V7 in C4-2B MDVR cells decreases the Acknowledgments fi level of c-Myc. In this study, this phenomenon was con rmed in C4-2B This work was supported, in part, by grants NIH/NCI CA168601, CA179970, and APALR cells. c-Myc is an important proto-oncogene in prostate cancer DOD PC150229, the U.S. Department of Veterans Affairs, Office of Research & that can deregulate the AR and AR variant signaling pathway. c-Myc Development BL&D grant number I01BX0002653 (to A.C. Gao), and a Research activates AR target genes and enhances AR and AR variant protein Career Scientist Award (to A.C. Gao). A.C. Gao is also a Research Career Scientist at stability (48). Here, our data showed that c-Myc is also a downstream VA Northern California Health Care System, Mather, California. target of AR-V7, which suggests a possible reciprocal regulation of AR- The costs of publication of this article were defrayed in part by the payment of page V7 and c-Myc. Knockdown of AKR1C3 suppresses both AR-V7 and c- charges. This article must therefore be hereby marked advertisement in accordance Myc expression in resistant cells. However, the interplay among with 18 U.S.C. Section 1734 solely to indicate this fact. AKR1C3, AR-V7, and c-Myc needs to be further investigated. Overall, our findings indicate that AKR1C3/AR-V7 axis plays an important Received January 8, 2020; revised April 8, 2020; accepted May 14, 2020; role in controlling antiandrogen resistance. published first May 19, 2020.

References 1. Tannock IF, de Wit R, Berry WR, Horti J, Pluzanska A, Chi KN, et al. Docetaxel 3. Holzbeierlein J, Lal P, LaTulippe E, Smith A, Satagopan J, plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. Zhang L, et al. Gene expression analysis of human prostate N Engl J Med 2004;351:1502–12. carcinoma during hormonal therapy identiﬁes androgen-responsive 2. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, et al. Molecular genes and mechanisms of therapy resistance. Am J Pathol 2004; determinants of resistance to antiandrogen therapy. Nat Med 2004;10:33–9. 164:217–27.

1716 Mol Cancer Ther; 19(8) August 2020 MOLECULAR CANCER THERAPEUTICS

Cross-Resistance in Antiandrogen Drugs

4. Tran C, Ouk S, Clegg NJ, Chen Y, Watson PA, Arora V, et al. Development of a ar-targeting therapies in advanced prostate cancer. Mol Cancer Ther 2018;17: second-generation antiandrogen for treatment of advanced prostate cancer. 2197–205. Science 2009;324:787–90. 26. Bianchini D, Lorente D, Rodriguez-Vida A, Omlin A, Pezaro C, Ferraldeschi R, 5. Scher HI, Sawyers CL. Biology of progressive, castration-resistant prostate et al. Antitumour activity of enzalutamide (MDV3100) in patients with met- cancer: directed therapies targeting the androgen-receptor signaling axis. astatic castration-resistant prostate cancer (CRPC) pre-treated with docetaxel J Clin Oncol 2005;23:8253–61. and abiraterone. Eur J Cancer 2014;50:78–84. 6. Smith MR, Saad F, Chowdhury S, Oudard S, Hadaschik BA, Graff JN, et al. 27. Loriot Y, Bianchini D, Ileana E, Sandhu S, Patrikidou A, Pezaro C, et al. Apalutamide treatment and metastasis-free survival in prostate cancer. N Engl J Antitumour activity of abiraterone acetate against metastatic castration- Med 2018;378:1408–18. resistant prostate cancer progressing after docetaxel and enzalutamide 7. de Bono JS, Logothetis CJ, Molina A, Fizazi K, North S, Chu L, et al. Abiraterone (MDV3100). Ann Oncol 2013;24:1807–126. and increased survival in metastatic prostate cancer. N Engl J Med 2011;364: 28. Noonan KL, North S, Bitting RL, Armstrong AJ, Ellard SL, Chi KN. Clinical 1995–2005. activity of abiraterone acetate in patients with metastatic castration-resistant 8. Scher HI, Fizazi K, Saad F, Taplin ME, Sternberg CN, Miller K, et al. Increased prostate cancer progressing after enzalutamide. Ann Oncol 2013;24:1802–7. survival with enzalutamide in prostate cancer after chemotherapy. N Engl J Med 29. Schrader AJ, Boegemann M, Ohlmann CH, Schnoeller TJ, Krabbe LM, Hajili T, 2012;367:1187–97. et al. Enzalutamide in castration-resistant prostate cancer patients progressing 9. Shore N, Zurth C, Fricke R, Gieschen H, Graudenz K, Koskinen M, et al. after docetaxel and abiraterone. Eur Urol 2014;65:30–6. Evaluation of clinically relevant drug-drug interactions and population phar- 30. Chi KN, Agarwal N, Bjartell A, Chung BH, Pereira de Santana Gomes AJ, Given macokinetics of darolutamide in patients with nonmetastatic castration- R, et al. Apalutamide for metastatic, castration-sensitive prostate cancer. N Engl J resistant prostate cancer: results of pre-speciﬁed and post hoc analyses of the Med 2019;381:13–24. phase III ARAMIS trial. Target Oncol 2019;14:527–39. 31. Moilanen AM, Riikonen R, Oksala R, Ravanti L, Aho E, Wohlfahrt G, et al. 10. Joseph JD, Lu N, Qian J, Sensintaffar J, Shao G, Brigham D, et al. A clinically Discovery of ODM-201, a new-generation androgen receptor inhibitor targeting relevant androgen receptor mutation confers resistance to second-generation resistance mechanisms to androgen signaling-directed prostate cancer therapies. antiandrogens enzalutamide and ARN-509. Cancer Discov 2013;3:1020–9. Sci Rep 2015;5:12007. 11. Rathkopf DE, Smith MR, Ryan CJ, Berry WR, Shore ND, Liu G, et al. Androgen 32. Sulochana SP, Saini NK, Daram P, Polina SB, Mullangi R. Validation of an LC- receptor mutations in patients with castration-resistant prostate cancer treated MS/MS method for simultaneous quantitation of enzalutamide, N-desmethy- with apalutamide. Ann Oncol 2017;28:2264–71. lenzalutamide, apalutamide, darolutamide and ORM-15341 in mice plasma and 12. Fizazi K, Massard C, Bono P, Jones R, Kataja V, James N, et al. Activity and safety its application to a mice pharmacokinetic study. J Pharm Biomed Anal 2018;156: of ODM-201 in patients with progressive metastatic castration-resistant prostate 170–80. cancer (ARADES): an open-label phase 1 dose-escalation and randomised phase 33. Fizazi K, Smith MR, Tombal B. Clinical development of darolutamide: a novel 2 dose expansion trial. Lancet Oncol 2014;15:975–85. androgen receptor antagonist for the treatment of prostate cancer. 13. Fizazi K, Shore N, Tammela TL, Ulys A, Vjaters E, Polyakov S, et al. Darolu- Clin Genitourin Cancer 2018;16:332–40. tamide in nonmetastatic, castration-resistant prostate cancer. N Engl J Med 34. Prekovic S, van Royen ME, Voet AR, Geverts B, Houtman R, Melchers D, et al. 2019;380:1235–46. The effect of F877L and T878A mutations on androgen receptor response to 14. Antonarakis ES, Lu C, Wang H, Luber B, Nakazawa M, Roeser JC, et al. AR-V7 enzalutamide. Mol Cancer Ther 2016;15:1702–12. and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med 35. Korpal M, Korn JM, Gao X, Rakiec DP, Ruddy DA, Doshi S, et al. An F876L 2014;371:1028–38. mutation in androgen receptor confers genetic and phenotypic resistance to 15. Liu C, Lou W, Zhu Y, Nadiminty N, Schwartz CT, Evans CP, et al. Niclosamide MDV3100 (enzalutamide). Cancer Discov 2013;3:1030–43. inhibits androgen receptor variants expression and overcomes enzalutamide 36. Liu C, Armstrong C, Zhu Y, Lou W, Gao AC. Niclosamide enhances abiraterone resistance in castration-resistant prostate cancer. Clin Cancer Res 2014;20: treatment via inhibition of androgen receptor variants in castration resistant 3198–210. prostate cancer. Oncotarget 2016;7:32210–20. 16. Hu R, Lu C, Mostaghel EA, Yegnasubramanian S, Gurel M, Tannahill C, et al. 37. Mostaghel EA, Marck BT, Plymate SR, Vessella RL, Balk S, Matsumoto AM, et al. Distinct transcriptional programs mediated by the ligand-dependent full-length Resistance to CYP17A1 inhibition with abiraterone in castration-resistant androgen receptor and its splice variants in castration-resistant prostate cancer. prostate cancer: induction of steroidogenesis and androgen receptor splice Cancer Res 2012;72:3457–62. variants. Clin Cancer Res 2011;17:5913–25. 17. Li Y, Chan SC, Brand LJ, Hwang TH, Silverstein KA, Dehm SM. Androgen 38. Schrader AJ, Schrader MG, Cronauer MV. Re: androgen receptor splice variants receptor splice variants mediate enzalutamide resistance in castration-resistant mediate enzalutamide resistance in castration-resistant prostate cancer cell lines. prostate cancer cell lines. Cancer Res 2013;73:483–9. Eur Urol 2013;64:169–70. 18. Liu C, Yang JC, Armstrong CM, Lou W, Liu L, Qiu X, et al. AKR1C3 promotes 39. Chan SC, Li Y, Dehm SM. Androgen receptor splice variants activate androgen AR-V7 protein stabilization and confers resistance to AR-targeted therapies in receptor target genes and support aberrant prostate cancer cell growth inde- advanced prostate cancer. Mol Cancer Ther 2019;18:1875–86. pendent of canonical androgen receptor nuclear localization signal. J Biol Chem 19. Liu C, Lou W, Zhu Y, Yang JC, Nadiminty N, Gaikwad NW, et al. Intracrine 2012;287:19736–49. androgens and AKR1C3 activation confer resistance to enzalutamide in prostate 40. Luo J, Attard G, Balk SP, Bevan C, Burnstein K, Cato L, et al. Role of androgen cancer. Cancer Res 2015;75:1413–22. receptor variants in prostate cancer: report from the 2017 mission androgen 20. Liu C, Armstrong CM, Lou W, Lombard A, Evans CP, Gao AC. Inhibition of receptor variants meeting. Eur Urol 2018;73:715–23. AKR1C3 activation overcomes resistance to abiraterone in advanced prostate 41. Kregel S, Wang C, Han X, Xiao L, Fernandez-Salas E, Bawa P, et al. Androgen cancer. Mol Cancer Ther 2017;16:35–44. receptor degraders overcome common resistance mechanisms developed during 21. Zhao Y, Gao JL, Ji JW, Gao M, Yin QS, Qiu QL, et al. Cytotoxicity enhancement in prostate cancer treatment. Neoplasia 2020;22:111–9. MDA-MB-231 cells by the combination treatment of tetrahydropalmatine and 42. Watson PA, Chen YF, Balbas MD, Wongvipat J, Socci ND, Viale A, et al. berberine derived from Corydalis Yanhusuo W. T. Wang. J Intercult Ethno- Constitutively active androgen receptor splice variants expressed in castration- pharmacol 2014;3:68–72. resistant prostate cancer require full-length androgen receptor. Proc Natl Acad 22. Liu C, Zhu Y, Lou W, Nadiminty N, Chen X, Zhou Q, et al. Functional p53 Sci U S A 2010;107:16759–65. determines docetaxel sensitivity in prostate cancer cells. Prostate 2013;73: 43. Penning TM. Aldo-keto reductase (AKR) 1C3 inhibitors: a patent review. 418–27. Expert Opin Ther Pat 2017;27:1329–40. 23. Liu C, Armstrong CM, Lou W, Lombard AP, Cucchiara V, Gu X, et al. 44. Flanagan JU, Yosaatmadja Y, Teague RM, Chai MZ, Turnbull AP, Squire CJ. Niclosamide and bicalutamide combination treatment overcomes enzaluta- Crystal structures of three classes of non-steroidal anti-inﬂammatory drugs in mide- and bicalutamide-resistant prostate cancer. Mol Cancer Ther 2017;16: complex with aldo-keto reductase 1C3. PLoS One 2012;7:e43965. 1521–30. 45. Yepuru M, Wu Z, Kulkarni A, Yin F, Barrett CM, Kim J, et al. Steroidogenic 24. Clegg NJ, Wongvipat J, Joseph JD, Tran C, Ouk S, Dilhas A, et al. ARN-509: a enzyme AKR1C3 is a novel androgen receptor-selective coactivator that pro- novel antiandrogen for prostate cancer treatment. Cancer Res 2012;72:1494–503. motes prostate cancer growth. Clin Cancer Res 2013;19:5613–25. 25. Lombard AP, Liu L, Cucchiara V, Liu C, Armstrong CM, Zhao R, et al. Intra 46. Kikuchi A, Furutani T, Azami H, Watanabe K, Niimi T, Kamiyama Y, et al. versus inter cross-resistance determines treatment sequence between taxane and In vitro and in vivo characterisation of ASP9521: a novel, selective, orally

AACRJournals.org Mol Cancer Ther; 19(8) August 2020 1717

Zhao et al.

bioavailable inhibitor of 17beta-hydroxysteroid dehydrogenase type 5 48. Bai S, Cao S, Jin L, Kobelski M, Schouest B, Wang X, et al. A positive role of c-Myc (17betaHSD5; AKR1C3). Invest New Drugs 2014;32:860–70. in regulating androgen receptor and its splice variants in prostate cancer. 47. Zhao J, Zhang M, Liu J, Liu Z, Shen P, Nie L, et al. AKR1C3 expression in primary Oncogene 2019;38:4977–89. lesion rebiopsy at the time of metastatic castration-resistant prostate cancer is 49. Sarwar M, Semenas J, Miftakhova R, Simoulis A, Robinson B, Gjorloff Wingren strongly associated with poor efﬁcacy of abiraterone as a ﬁrst-line therapy. A, et al. Targeted suppression of AR-V7 using PIP5K1alpha inhibitor overcomes Prostate 2019;79:1553–62. enzalutamide resistance in prostate cancer cells. Oncotarget 2016;7:63065–81.

1718 Mol Cancer Ther; 19(8) August 2020 MOLECULAR CANCER THERAPEUTICS

Cross-Resistance Among Next-Generation Antiandrogen Drugs Through the AKR1C3/AR-V7 Axis in Advanced Prostate Cancer

Jinge Zhao, Shu Ning, Wei Lou, et al.

Mol Cancer Ther 2020;19:1708-1718. Published OnlineFirst May 19, 2020.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-20-0015

Cited articles This article cites 48 articles, 18 of which you can access for free at: http://mct.aacrjournals.org/content/19/8/1708.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/19/8/1708.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/19/8/1708. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.