(+)-JJ-74–138 is a Novel Noncompetitive Androgen Receptor Antagonist

This article is featured in Highlights of This Issue, p. 481

Prostate cancer is the second most prevalent cancer and the fifth leading cause of cancer mortality in the world (1). The pathogenesis of prostate cancer is largely driven by the androgen receptor (AR), making AR a favorable therapeutic target (2). Androgen deprivation therapy (ADT) is the current standard treatment method for metastatic prostate cancer, inhibiting AR-mediated signaling which drives prostate cancer growth (3, 4). While most patients initially respond favorably to ADT, most metastatic prostate cancers progress to castration-resistant prostate cancer (CRPC) in about 18 months (5, 6). Although CRPC can grow under ADT, a majority of these cases are still AR-dependent (7). As such, targeting the AR still remains a high priority in CRPC treatment.

AR is a ligand-regulated steroid receptor that consists of three major functional domains known as the N-terminal domain (NTD), DNA binding domain (DBD), and the ligand binding domain (LBD; ref. 8). AR signaling is initiated by the highly specific binding of androgens such as testosterone or a higher affinity ligand, such as DHT, to the LBD of the AR (9). Most AR antagonists have been designed to competitively inhibit the binding of DHT to the LBD, thereby inhibiting the normal biological effect of the AR.

Enzalutamide is a clinically approved second-generation AR antagonist that prolongs the survival of CRPC patients (10). Enzalutamide is prescribed to patients following the failure of ADT and works mechanistically by binding to the LBD and causing a conformational change in AR that prevents the formation of an active transcriptional complex (11). While enzalutamide is initially effective in CRPC patients, resistance eventually occurs. Constitutively active AR splice variants lacking the LBD and AR point mutations in the LBD are a few of the currently understood mechanisms of enzalutamide resistance (12–14). Due to the development of enzalutamide resistance, there is a need to develop new alternative AR antagonists that work independently of the LBD.

Our lab has previously identified a novel AR antagonist IMTPPE. IMTPPE and its analogue JJ-450 are able to inhibit the growth of prostate tumors that express both full-length AR, and those that express AR variants or constructs lacking the LBD (15–17). Because the inhibition is independent of the LBD, these molecules may have a different mechanism of action than enzalutamide and thus could be effective as a next stage of treatment for CRPC. Our lab aimed to develop analogous compounds capable of targeting both full-length AR and variants lacking the LBD that were more potent than JJ-450. (+)-JJ-74–138, a structural analogue of JJ-450, was further evaluated for its ability to noncompetitively bind the AR and inhibit enzalutamide-resistant prostate cancer.

The prostate cancer cell line C4–2 was generously provided by Dr. Leland WK Chung (Cedars-Sinai Medical Center, Los Angeles, California). LNCaP95 (LN95) cell line was kindly provided by Dr. Jun Luo (Johns Hopkins, MD). PC-3 and CWR22Rv1 (22Rv1) prostate cancer cells were obtained through the ATCC. C4–2, PC-3, 22Rv1, and LN95 cells were cultured in RPMI1640 (Corning, Corning, New York) and supplemented with 10% FBS (Atlanta Biologicals) or charcoal-stripped FBS (for LN95 cells). DU145 cells were obtained from ATCC, and cultured in DMEM (Lonza). Media for each cell line was supplemented with 1% l-glutamine and 1% penicillin streptomycin. C4–2, DU145, and 22Rv1 were authenticated in 2016 or 2018 using DNA fingerprinting by examining microsatellite loci in a multiplex PCR reaction (4322288, Applied Biosystems) by the University of Pittsburgh Cell Culture and Cytogenetics Facility. The cell lines were aliquoted and stored in liquid nitrogen after purchase or authentication. We did not perform authentication on PC-3 because the cell line was purchased from ATCC, which authenticated PC-3. LN95 was not authenticated since this cell line is uniquely available from Dr. Jun Luo's team. The cell lines were tested monthly by PCR for Mycoplasma contamination. All experiments were performed using cell lines between passages 4 and 22. (–)-JJ-450, (+)-JJ-450, and (+)-JJ-74–138 were prepared as previously reported (18, 19).

C4–2, LN95, and 22Rv1 cells were seeded in 24-well plates (1 × 10⁵) in phenol red-free RPMI1640 supplemented with 10% CSS. After 3 days, medium was aspirated and replaced with medium containing of drug was added. Cells were treated with drug or DMSO vehicle for 2 days and the cell medium was collected for each sample. The volume of medium for each sample was then measured and ddH₂O was added to equalize the volume to account for variable evaporation between the wells. Cells were then washed with PBS and detached with Trypsin-EDTA. A Beckman Coulter Particle Counter was used to count the number of cells in each well. A human kallikrein 3/PSA Quantikine ELISA kit (DKK300, R&D Systems) was then used according to the manufacturer's protocol. PSA concentration was measured and normalized to the number of cells in each respective well. Each ELISA experiment was performed at least 2 times with at least three replicates.

Cells at ∼60% confluence were treated for 38 hours with the indicated concentrations of (+)-JJ-74–138 in triplicate. RNA extraction was performed with the E.Z.N.A Total RNA Kit (Omega Bio-Tek, R6834) according to the manufacturer's protocol. RNA quality was assessed using gel electrophoresis. Reverse transcription was performed using the PrimeScript RT Reagent Kit (RR036A, TaKaRa). cDNA was amplified using the ABI StepOnePlus System (Applied Biosystems) using gene specific primers (Supplementary Table S3) and Bullseye Real Time qPCR Master Mix (BEQPCR-R, MidSci). Data were analyzed using the ΔΔC_t method using GAPDH as the housekeeping gene.

C4–2-PSA-rl was generated previously as described (18). The cells were cultured in 10% cFBS, phenol red-free RPMI medium prior to treatment for 24 hours. After treatment, luciferase assays were performed using a Dual-Luciferase Reporter Assay Kit (E1910, Promega) and measured using LMax II Microplate Reader (Molecular Devices) as described previously (15). The firefly luciferase activity was normalized to Renilla luciferase activity.

C4–2 were cultured for 24 hours in 10% charcoal-stripped FBS, phenol red-free RPMI1640 medium and then treated with vehicle, 1 nmol/L R1881, 10 μmol/L (+)-JJ-74–138, or both R1881 and (+)-JJ-74–138 for 2 hours. The chromatin immunoprecipitation (ChIP) assay was performed using a kit (17–295, Sigma-Aldrich) according to the manufacturer's instruction with minor modifications. We used 0.125 mol/L glycine to stop cross-linking using 1% paraformaldehyde and QIAquick PCR purification kit (28104, Qiagen) to purify the DNA following ChIP. PCR primers are shown in Supplementary Table S3 and the analysis was performed as described previously (17).

Cells were lysed with RIPA lysis buffer containing 1X protease inhibitor cocktail and 1 mmol/L PMSF. Nucleocytoplasmic fractionation was performed according to the manufacturer's (G-Biosciences, 786–182) protocol. Protein extracts were boiled in SDS sample buffer and bands were separated by size by SDS-PAGE. Proteins were then transferred to a PVDF membrane, blocked for 1 hour in 5% non-fat milk, and incubated with the following primary antibodies overnight at 4°C: anti-AR rabbit (1:1,000, abcam, ab108341), anti-Lamin B rabbit (1:1,000, CST, 13435S), anti-PARP rabbit (1:1,000, CST, 9532S), or GAPDH-HRP mouse (1:5,000, Santa Cruz Biotechnology Inc, sc-47724). Membranes were washed in tris-buffered saline with 0.1% Tween-20 (TBST) and incubated in 1:5,000 diluted secondary antibody in 5% non-fat milk for 1 hour at room temperature. Membranes were washed 3× in TBST and protein signals were detected with the Chemidoc Imaging System.

LN95 cells were seeded in 12-well plates in phenol red-free media supplemented with 10% CSS. After 3 days, medium was removed and replaced with serum-free RPMI1640 containing 1 nmol/L [³H] labeled DHT and 0, 0.128, 0.064, 0.32, 1.6, 8, and 40 μmol/L of enzalutamide, tamoxifen (Sigma), and (+)-JJ-74–138 and incubated at 37°C for 90 minutes. Cells were washed with PBS and bound ligand was extracted in ethanol for 30 minutes at room temperature. Bound ligand was detected in counts per minute using a scintillation counter (Packer Tri-Carb 1600TR).

LN95, 22Rv1, PC-3, DU145 cells were seeded in 24-well plates (50,000 cells/well) in phenol red-free media supplemented with 10% CSS. Cells were allowed to adhere for 2 days, then three replicates were trypsinized and the cells were counted using a Beckman Coulter Particle Counter and considered as day 0. Fresh medium was added to the trypsinized wells to account for variable evaporation in the incubator containing 1.6, 4, or 10 μmol/L (+)-JJ-74–138 or vehicle for counting on day 2, 4, and 6. Three replicates for each concentration of DMSO vehicle or (+)JJ-74–138 were then trypsinized and counted on days 2, 4, and 6. Medium containing (+)-JJ-74–138 or DMSO vehicle was replaced on day 3. To count cells, medium was aspirated, cells were washed with PBS, and cells were detached with Trypsin-EDTA at 37°C. Each well was counted twice with a Beckman Coulter Particle Counter and the average cell number per well was determined. The experiment was replicated two times in each of the cell lines with 10 μmol/L enzalutamide as a control.

Cells were treated in triplicate in 6-well plates for 48 hours with the indicated concentrations of (+)-JJ-74–138. Cells were incubated in 200 nmol/L BrdU (B5002, Sigma) for 30 minutes prior to being fixed with 70% EtOH for 30 minutes on ice. DNA was denatured by incubating in 4N HCl for 30 minutes at room temperature. Cells were incubated in 1:100 FITC conjugated anti-BrdU antibody (347583, BD Biosciences) and incubated at room temperature for 30 minutes. Cells were washed with PBS and incubated with 1 μg/mL propidium iodide (J66584, Alfa Aesar). The cells were acquired using the BD Accuri C6 Plus (BD Biosciences) flow cytometer and analyzed with CFlow Plus software (BD Biosciences).

Full-length recombinant AR was purified by Creative BioMart using a baculovirus system (Creative BioMart). Enzalutamide, (+)-JJ-74–138, and tamoxifen were supplied to Profacgen to perform surface plasmon resonance (SPR) binding (Profacgen). Biotinylated AR protein was dissolved in water and printed onto a bare gold-coated PlexArray Nanocapture Sensor Chip (Plexera Bioscience) in replicate. The buffers and enzalutamide, (+)-JJ-74–138, and tamoxifen samples (100 nmol/L, 200 nmol/L, 400 nmol/L, 800 nmol/L, and 1,600 nmol/L) were injected using a nonpulsatile piston pump into the 30 μL flowcell on the coupling prism at 25°C. The SPRi measurements after binding and washing were recorded in AU (Normalized reflection unit) and performed with PlexArray HT (Plexera Bioscience). The binding signals were analyzed by Data Analysis Module (DAM, Plexera Bioscience) and the average reflectivity variations were plotted as a function of time with GraphPad Prism. The kinetic parameters of binding were performed with BIAevaluation 4.1 software (Biacore, Inc).

Male Crl:SHO-PrkdcscidHrhr mice aged 6 to 8 weeks old were purchased from Charles River Laboratory. Animal care and use were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh. Xenograft tumors were established by gently mixing 22Rv1 (2 × 10⁶ cells) suspended in 180 μL medium with 180 μL of Matrigel (Corning, NY) before subcutaneous inoculation in the right flank region of mice. Tumor size was measured with calipers, and tumor volume was calculated using the modified ellipsoid formula: length × width² × 0.52 (20). Trans-scrotal castration was performed under isoflurane anesthesia with proper aseptic and antiseptic technique when tumor volume reached ∼200 mm³. When tumor volume in castrated mice reached ∼300 mm³, mice were randomized into groups and treated with DMSO vehicle, enzalutamide (i.p. 10 mg/kg body weight), or (+)-JJ-74–138 (i.p. 10 mg/kg body weight) every day for 21 days or when tumor length exceeded 20 mm or tumor volume reached 2,000 mm³. Mice were weighed on days 0, 10, and 17.

Data were expressed as the mean ± SD and were plotted with GraphPad Prism (Version 9; GraphPad Software). The PSA ELISA data, cell proliferation data, qPCR data, and ChIP data were analyzed using the Student t test to determine significance. Statistical significance for Kaplan–Meier analysis was determined using the log-rank (Mantel-Cox) test. *, P < 0.05l; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 were used to denote significance.

All data generated in this study are available upon request from the corresponding author.

The LN95 cell line was developed as a model for CRPC (21) and exhibits resistance to enzalutamide (22, 23). LN95 cells express PSA in androgen-free medium which reflects androgen-independent AR activity (24). Since our goal was to optimize JJ-450 for the inhibition of androgen-independent activation of AR, we synthesized and evaluated various analogues for their ability to inhibit PSA at the protein level in LN95. (+)-JJ-74–138 showed a high efficacy in this screen, so we directly compared the effects of (–)-JJ-450, (+)-JJ-450, and the JJ-450 analogue (+)-JJ-74–138 (chemical structures in Fig. 1A) on AR-target gene PSA expression in LN95 cells cultured in androgen-free medium (Fig. 1B). ELISA analysis of PSA in conditioned medium of LN95 treated with (–)-JJ-450, (+)-JJ-450, or (+)-JJ-74–138 showed that (+)-JJ-74–138 caused more inhibition of PSA expression than (+)-JJ-450, while (–)-JJ-450 only slightly inhibited PSA level. Thus, (+)-JJ-74–138 is a potent analogue of JJ-450 for the inhibition of androgen-independent AR activation in cell lines expressing AR splice variants lacking the LBD.

Since the parental compound IMTPPE inhibited androgen-independent AR nuclear localization (25), we tested if (+)-JJ-74–138 could reduce AR level in the nucleus in CRPC cells. Western blot analysis coupled with nucleocytoplasmic fractionation showed overnight treatment with 10 μmol/L (+)-JJ-74–138 inhibited AR nuclear level in C4–2, LN95, and 22Rv1 (Fig. 1C). As expected, 10 nmol/L synthetic androgen R1881 increased the amount of total AR, particularly the nuclear AR level. As a control, AR mRNA was unaffected by treatment of 10 μmol/L (+)-JJ-74–138 (Supplementary Fig. S1). Also, (+)-JJ-74–138 did not appear to influence the nuclear ARv7 level (Fig. 1C).

In our previous studies, JJ-450 was also able to inhibit AR^F876L, an AR with a point mutation in the LBD causing resistance to enzalutamide (16). To test whether (+)-JJ-74–138 could also inhibit AR in a manner independent of the LBD, we tested the efficacy of (+)-JJ-74–138 in enzalutamide-resistant cells. We measured PSA levels in the medium of the enzalutamide responsive cell line C4–2 and the enzalutamide-resistant cell lines LN95 and 22Rv1 with and without treatment with (+)-JJ-74–138. (+)-JJ-74–138 was able to inhibit PSA production in C4–2, LN95, and 22Rv1 cell lines (Fig. 2A). PSA levels were inhibited by enzalutamide in C4–2 (∼70%, P = 0.003), moderately inhibited in LN95 (∼20%, P = 0.0238), and not inhibited in 22Rv1 (P = 0.869). As expected, the sensitivity of enzalutamide in these cell lines appears to be related to their expression levels of endogenous ARv7 (26). The effect of (+)-JJ-74–138 and enzalutamide on PSA was also determined on a per cell basis which suggested that the inhibition of PSA by (+)-JJ-74–138 was not due to the reduction of the number of cells. Each of these cell lines was also treated with a concentration gradient of (+)-JJ-74–138 to determine the IC₅₀ (Supplementary Table S1). (+)-JJ-74–138 had an IC₅₀ of 1.16 μmol/L in C4–2, 2.95 μmol/L in LN95, and 4.32 μmol/L in 22Rv1. The IC₅₀ value of (+)-JJ-74–138 was lower in the CRPC cell line C4–2 than that in the enzalutamide-resistant cell lines LN95 and 22Rv1. To further demonstrate (+)-JJ-74–138 inhibition of AR, we showed with a quantitative reverse transcription PCR (qRT-PCR) that (+)-JJ-74–138 inhibited mRNA expression of AR downstream genes KLK3 and NKX3–1 in C4–2 and 22Rv1. Furthermore, (+)-JJ-74–138 inhibited ARv7 downstream genes UBE2C and CDC20 (27) in 22Rv1 cells (Fig. 2B). Inhibition of AR activity was further shown in the presence of synthetic androgen R1881, where (+)-JJ-74–138 caused a dose-dependent inhibition of PSA luciferase activity (Fig. 2C) and inhibited liganded AR binding to androgen response elements (ARE) in a C4–2 ChIP assay (Fig. 2D). As a control, we showed that 10 μmol/L (+)-JJ-74–138 did not impact AR nuclear level in the presence of R1881 (Fig. 2E). This suggests that (+)-JJ-74–138 can inhibit liganded AR binding to AREs, in addition to its ability to suppress unliganded AR level in the nucleus.

Since (+)-JJ-74–138 showed greater efficacy in CRPC cell lines than enzalutamide, (+)-JJ-74–138 is likely to inhibit AR through a mechanism different from that of enzalutamide. To explore the mechanism of its inhibition of AR, we tested to see how (+)-JJ-74–138 would inhibit DHT induced PSA production. C4–2, LN95, and 22Rv1 cells cultured in androgen-free conditions were treated with increasing concentrations of DHT in combination with either (+)-JJ-74–138 or enzalutamide as a control for competitive antagonism (Fig. 3A). In C4–2, different concentrations of enzalutamide and (+)-JJ-74–138 were used to account for the differences in the potency between the two compounds in this cell line. Since LN95 and 22Rv1 are enzalutamide-resistant cell lines (23, 28), 10 μmol/L was used for both (+)-JJ-74–138 and enzalutamide in these cells. In C4–2, LN95, and 22Rv1 cells, enzalutamide acted as expected and was able to dramatically shift the DHT dose response curve to the right, with the EC₅₀ increasing ∼10-fold, while the enzalutamide inhibition of PSA induction was alleviated by high doses of DHT (Supplementary Table S2). In contrast, (+)-JJ-74–138 had little or no effect on the EC₅₀ of DHT induction of PSA while it reduced the maximal PSA induction by high DHT dosages (Supplementary Table S2). This suggests that (+)-JJ-74–138 did not compete with DHT for binding to the LBD, acting as a noncompetitive inhibitor of the AR. In LN95 and 22Rv1 cells, enzalutamide was able to effectively inhibit DHT-stimulated PSA expression (Fig. 3A), although these two cell lines are resistant to enzalutamide.

To further test if (+)-JJ-74–138 acts as a noncompetitive inhibitor, we performed a ligand binding competition assay (Fig. 3B). LN95 cells were incubated with 1 nmol/L [³H] DHT in the presence of increasing concentrations of (+)-JJ-74–138, tamoxifen, or enzalutamide. Tamoxifen was selected to control for nonspecific binding due to its highly hydrophobic structure. (+)-JJ-74–138 at 8 μmol/L or higher caused only ∼10–20% inhibition of DHT binding which is consistent with a noncompetitive antagonism (29). As expected, enzalutamide acted as a competitive antagonist by competing with the [³H] DHT for binding. This indicates that despite LN95 being enzalutamide-resistant, enzalutamide was still able to bind to the AR in LN95 cells.

One major feature of AR antagonist specificity is their ability to inhibit AR-dependent but not AR-independent cell growth (29). To test (+)-JJ-74–138 for this capability, a growth curve was determined for the AR-positive cell lines LN95 (21) and 22Rv1 (30), and the AR-negative cell lines PC-3 (31) and DU145 (32) that were treated with (+)-JJ-74–138 for 6 days. (+)-JJ-74–138 was able to markedly and significantly inhibit cell growth in a dose-dependent manner in both LN95 and 22Rv1 cell lines (Fig. 4A). In contrast, (+)-JJ-74–138 had little or no effect on the growth of PC-3 or DU145 cells (Fig. 4A). PC-3 cell growth was slightly inhibited by 10 μmol/L (+)-JJ-74–138 on day 4 and 6, but the magnitude of this inhibition was much lower compared with the inhibitory effect in LN95 and 22Rv1. (+)-JJ-74–138 had a stronger inhibitory effect on the growth of LN95 cells than in 22Rv1 cells, with 10 μmol/L (+)-JJ-74–138 leading to a ∼3 fold inhibition of cell numbers in LN95, whereas 10 μmol/L (+)-JJ-74–138 showed a ∼2 fold inhibition in 22Rv1. This result was consistent with the IC₅₀ values of PSA inhibition in these two cell lines (Supplementary Table S1). This experiment was repeated with 10 μmol/L enzalutamide and 10 μmol/L (+)-JJ-74–138 for each of the cell lines (Supplementary Fig. S2). In AR negative PC-3 and DU145 cell lines, the minimal inhibitory effect of 10 μmol/L (+)-JJ-74–138 was similar to that of 10 μmol/L enzalutamide. To further explore the effect of (+)-JJ-74–138 on cell growth, a flow cytometry cell-cycle analysis was performed in C4–2, LN95, and 22Rv1 cell lines (Fig. 4B). This analysis indicated that (+)-JJ-74–138 was able to cause G₀–G₁ arrest in a dose-dependent manner. G₀–G₁ arrest is a common feature of anti-androgens in androgen-dependent cell lines (33). Treatment of (+)-JJ-74–138 did not result in cleaved-PARP (Fig. 4C), suggesting that apoptosis induction is not a major mechanism of action for (+)-JJ-74–138.

The above evidence supported that (+)-JJ-74–138 noncompetitively inhibited CRPC and enzalutamide-resistant cell lines and had specificity for AR-dependent cell lines. However, these experiments did not resolve whether (+)-JJ-74–138 could bind to AR directly or indirectly. To test the potential direct binding and the kinetic binding parameters of (+)-JJ-74–138 to AR, full-length recombinant AR was purified using a baculovirus system (Fig. 5A), and binding was measured using SPR. By using AR protein on a chip and plotting the average reflectivity variations of collimated light reflecting off the chip as a function of time, the kinetic binding parameter K_D for enzalutamide, (+)-JJ-74–138, and tamoxifen was calculated (Fig. 5B). As expected, enzalutamide had a strong association as reflected by the low K_D, indicating that the SPR experiment was successful. (+)-JJ-74–138 exhibited a K_D of 4.29 μmol/L, indicating that (+)-JJ-74–138 could bind directly to the AR. The K_D value was also similar to the IC₅₀ value from the PSA ELISA assay (Supplementary Table S1), indicating that direct binding to AR was likely a major mechanism of action for (+)-JJ-74–138. The K_D of (+)-JJ-74–138 was roughly 10-fold lower than the value measured for the highly hydrophobic molecule tamoxifen, thus suggesting that the binding of (+)-JJ-74–138 was not due to strictly hydrophobic interactions. As expected, tamoxifen did not affect PSA expression in LN95 cells. (Supplementary Fig. S3).

After showing in vitro efficacy, our group wanted to explore whether (+)-JJ-74–138 would also inhibit enzalutamide-resistant 22Rv1 xenograft tumor growth in vivo. Treatment of xenografts with (+)-JJ-74–138 for 3 weeks did not cause a significant change in body weight of the mice compared with vehicle control, suggesting no toxicity of (+)-JJ-74–138 at 10 mg/kg. (Fig. 6A). Tumor volume in xenograft tumors treated with (+)-JJ-74–138 was significantly reduced on days 10 and 13 compared with the control group (Fig. 6B). Consistent with previous studies (17, 34), the tumor growth rate in mice treated with enzalutamide was similar to the control group through 13 days. Since the 22Rv1 xenograft tumor growth is rapid (Fig. 6B), two animals from both the control and the enzalutamide treatment group had to be euthanized at day 10. Hence, the data presented included the tumor volume through 13 days with the last observation carried forward (Fig. 6B). Although this experiment was not intended to determine survival, ad hoc analysis identified a significantly increased survival for (+)-JJ-74–138-treated mice compared with enzalutamide (P < 0.05) (Fig. 6C). Enzalutamide-treated mice survival did not differ from the vehicle control. This trend suggested that (+)-JJ-74–138 could increase the mean survival time of 22Rv1 xenograft mice. This animal data was in agreement with the in vitro inhibition of 22Rv1 proliferation and PSA, indicating that (+)-JJ-74–138 may prove effective in treating enzalutamide-resistant prostate cancer.

Finding new AR antagonists that can treat CRPC resistant to enzalutamide is of clinical importance. We previously evaluated the AR inhibition of our compounds using a PSA promoter/enhancer driven luciferase assay in the presence of 1 nmol/L R1881 and identified JJ-450 as a potent analogue of IMPTTE, a novel AR antagonist identified from a high throughput screen (25). Further studies showed that JJ-450 could be resolved, with (–)-JJ-450 being more potent than (+)-JJ-450 in the inhibition of androgen-induced AR activity. RNA sequencing analysis showed that (–)-JJ-450 and enzalutamide had a very similar effect on the expression of the androgen-induced transcriptome, suggesting that the mechanism of (–)-JJ-450 may be related to that of enzalutamide. Compared with enzalutamide and (–)-JJ-450, (+)-JJ-450 had different effect on androgen-induced transcriptome, suggesting that (+)-JJ-450 is likely to inhibit AR differently from enzalutamide and that JJ-450 could bind to AR and inhibit AR with mutations in the LBD or lacking LBD (17). Here, we have characterized a structural analogue of JJ-450, (+)-JJ-74–138, that is capable of noncompetitively inhibiting AR. (+)-JJ-74–138 was chosen for this study because of its higher potency than JJ-450 in the LN95 cell line. Through treatment with different concentrations of (+)-JJ-74–138, we showed that (+)-JJ-74–138 was capable of inhibiting PSA at the protein level in both enzalutamide-responsive and enzalutamide-resistant AR-positive cell lines. Like its parental compound IMTPPE, (+)-JJ-74–138 reduced AR level in the nucleus. Further studies showed (+)-JJ-74–138 inhibition of transcriptional activity of full-length AR and ARv7 as well as the inhibition of liganded AR binding to AREs. (+)-JJ-74–138 noncompetitively inhibited the AR by preventing the maximal DHT induced PSA response. (+)-JJ-74–138 can cause G₀–G₁ phase cell-cycle arrest and does not appear to induce apoptosis. The selective inhibition of cell growth in AR-positive cell lines, but not AR-negative cell lines, suggested that (+)-JJ-74–138 interacted directly with AR. SPR experiments provided more direct evidence for (+)-JJ-74–138 binding to AR, indicating that the inhibitory effect we observed was through its affinity to AR. Furthermore, (+)-JJ-74–138 was well tolerated in mice and could reduce 22Rv1 xenograft tumor growth, demonstrating the potential efficacy of (+)-JJ-74–138 in vivo.

Figure 4 showed (+)-JJ-74–138 had a small, but statistically significant, inhibition of PC3 cell growth at day 4 and day 6. This finding suggests that high dose (+)-JJ-74–138 could have off-target effects. However, these off-target effects appeared to be weak compared with its AR-targeting inhibition. Thus, there should be an excellent therapeutic window for this compound, which is consistent with the finding that this compound did not affect animal body weight or behavior.

IMTPPE was identified in a high throughput screen for its ability to prevent AR nuclear localization (25), and (+)-JJ-74–138 retained this ability. (+)-JJ-74–138 reduced nuclear AR level in CRPC cell lines, including C4–2, LN95 and 22Rv1. A recent paper by our group showed that unliganded AR is predominantly degraded in the nucleus (35). (+)-JJ-74–138 treatment is likely to cause more effective AR degradation in the nucleus of CRPC cells. Future studies should explore the mechanisms responsible for the (+)-JJ-74–138 inhibition of AR nuclear level.

The exact binding site of (+)-JJ-74–138 on AR remains to be identified; however, our results suggest that (+)-JJ-74–138 does not compete for binding at the LBD and can inhibit ARv7 downstream genes. Currently, there are no clinically approved drugs that specifically bind to a non-LBD region of AR. Other regions such as the NTD are critically important for proper AR function, and AR splice variants lacking the LBD are upregulated during the progression of prostate cancer (36–38). The noncompetitive inhibitory nature of (+)-JJ-74–138 indicates that, unlike existing therapies, it could be targeting non-LBD regions of AR and could provide a strategy following the failure of existing AR antagonists. Due to its different mechanism, (+)-JJ-74–138 may be also able to act synergistically with LBD-targeting agents. Further studies should be conducted to explore this potential synergism.

Developing non-LBD targeting AR antagonists is currently an important direction in prostate cancer research. Different non-LBD targeting AR antagonists may have different mechanisms of action and could be used more effectively in diverse subpopulations of patients with prostate cancer. Selective AR degraders (SARDs) such as UT-65 and UT-155 are a class of molecules that bind to both the LBD and the NTD of AR (39). EPI-001 and its analogues bind to NTD of AR and are effective in inhibiting AR splice variants (40). (+)-JJ-74–138 provides an alternative to EPI compounds for certain types of prostate cancers. For example, 22Rv1 tumors are resistant to EPI-001 but sensitive to (+)-JJ-74–138 (41). Other molecules such as pyrvinium pamoate, VPC-14228, and VPC-14449 are AR antagonists that target the DBD of AR (29, 42). It will be important to further understand and develop these non-LBD targeting AR antagonists.

Our studies indicate that androgen-stimulated PSA production in 22Rv1 and LN95 cell lines is enzalutamide-sensitive, although the PSA production in androgen-free medium was enzalutamide-resistant. This is consistent with findings in the literature with different methodologies (27, 43), and suggests that AR and ARv7 in LN95 and 22Rv1 cells are able to drive PSA expression under androgen-free conditions (44). However, the full-length AR could be activated by androgens and the androgen-induced PSA expression can be inhibited by enzalutamide, indicating that full-length AR in LN95 and 22Rv1 cells can interact with both androgens and antiandrogens.

Overall, our studies indicate that (+)-JJ-74–138 is a promising novel noncompetitive AR antagonist for treating enzalutamide-resistant CRPC, with improved potency compared with its parental compound JJ-450. While the exact binding site of (+)-JJ-74–138 remains to be elucidated, our studies suggest that this compound directly binds to a non-LBD region of the AR. On the basis of the current studies, (+)-JJ-74–138 inhibits AR through the following mechanisms: (+)-JJ-138–138 causes preferential nuclear degradation of unliganded full-length AR, decreases liganded AR binding to AREs, and inhibits both AR and ARv7 transcriptional activity. Future work should elucidate a more detailed mechanism of inhibition, and analogues of (+)-JJ-74–138 should be further explored to maximize the efficacy of these compounds.

J.B. Nelson reports a patent for “Small Molecules Targeting Androgen Receptor Nuclear Localization and/or Level in Prostate Cancer”; Zhou Wang, Joel B. Nelson, Minh Mindy Bao Nguyen, John S. Lazo, Paul A. Johnston, Peter Wipf–Patent No. 9708276, issued July 18, 2017 issued to UPMC Enterprises. P. Wipf reports personal fees from UPMC Enterprises outside the submitted work; in addition, P. Wipf has a patent for US Patent App. 17/089519 pending, a patent for US Patent 10980806 issued, a patent for US Patent 10544110 issued, a patent for US Patent 10882834 issued, a patent for US Patent App. 17/095653 pending, a patent for US Patent 10004730 pending, a patent for US Patent 9981974 pending, and a patent for US Patent 9708276 pending. Z. Wang reports grants from US Department of Defense; and grants from NIH during the conduct of the study; personal fees from Venable LLP outside the submitted work; in addition, Z. Wang has a patent for US 10004730 issued. No disclosures were reported by the other authors.

R.N. Cole: Conceptualization, data curation, formal analysis, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing. W. Chen: Formal analysis, investigation. L.E. Pascal: Formal analysis, writing–review and editing. J.B. Nelson: Formal analysis. P. Wipf: Resources, formal analysis. Z. Wang: Conceptualization, resources, formal analysis, supervision, funding acquisition, methodology, project administration, writing–review and editing.

We would like to thank Leland W.K. Chung (Cedars-Sinai, Los Angeles, California) for C4-2 cells, Jun Luo (John Hopkins, Baltimore, Maryland) for LNCaP95 cells, Jianhua Zhou for technical assistance, and Taber S. Maskrey for supervising the scale-up of (–)-JJ-450, (+)-JJ-450, (–)-JJ-74-138, and (+)-JJ-74-138. This work was funded in part by DOD Award W81XWH-16-1-0659 and by NIH grants R01 CA186780 (to Z Wang) and R50 CA211242 (to LEP) as well as by the Department of Urology, University of Pittsburgh. This project used the UPMC Hillman Cancer Center Animal Facility, and the Tissue and Research Pathology Services/Pitt Biospecimen Core, which were supported in part by NCI award P30 CA047904 and the Senior Vice Chancellor's Office at the University of Pittsburgh.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.