Abstract
Although second-line antiandrogen therapy (SAT) is the standard of care in men with castration-resistant prostate cancer (CRPC), resistance inevitably occurs. One major proposed mechanism of resistance to SAT involves the emergence of androgen receptor (AR) splice variant-7, AR-V7. Recently, we developed MTX-23 using the principle of proteolysis targeting chimera (PROTAC) to target both AR-V7 and AR-full length (AR-FL). MTX-23 has been designed to simultaneously bind AR's DNA binding domain (DBD) and the Von Hippel–Lindau (VHL) E3 ubiquitin ligase. Immunoblots demonstrated that MTX-23's degradation concentration 50% (DC50) for AR-V7 and AR-FL was 0.37 and 2 μmol/L, respectively. Further studies revealed that MTX-23 inhibited prostate cancer cellular proliferation and increased apoptosis only in androgen-responsive prostate cancer cells. The antiproliferative effect of MTX-23 was partially reversed when either AR-V7 or AR-FL was overexpressed and was completely abrogated when both were overexpressed. To assess the potential therapeutic value of MTX-23, we next generated 12 human prostate cancer cell lines that are resistant to the four FDA-approved SAT agents—abiraterone, enzalutamide, apalutamide, and darolutamide. When resistant cells were treated with MTX-23, decreased cellular proliferation and reduced tumor growth were observed both in vitro and in mice. These results collectively suggest that MTX-23 is a novel PROTAC small molecule that may be effective against SAT-resistant CRPC by degrading both AR-V7 and AR-FL.
Introduction
Prostate cancer is the most common noncutaneous cancer diagnosed among men and the second leading cause of male cancer deaths in the United States (1). In 2020, approximately 33,330 men are estimated to die from prostate cancer (1). Although radiation and surgery are quite effective for a localized disease, metastatic prostate cancer is essentially incurable (2, 3). In patients with metastatic and recurrent disease, medical castration via androgen deprivation therapy (ADT) is the generally accepted standard of care. Yet, castration-resistant prostate cancer (CRPC) eventually emerges with a median survival time of 18–24 months (4–6). Once CRPC develops, second-line antiandrogen therapy (SAT), immunotherapy, and chemotherapy have demonstrated a modest survival benefit (7–9).
Castration resistance is the uniform clinical feature of a lethal prostate cancer. According to the 2018 National Comprehensive Cancer Network guideline, SAT is the recommended first-line therapy in treating CRPC, especially in men with no definitive evidence of metastasis (stage M0; ref. 10). Although the precise percentage of men with a lethal prostate cancer who are managed with each of the four specific SAT agents—abiraterone, apalutamide, enzalutamide, and darolutamide (11–15)—is uncertain, essentially all men who die from prostate cancer in the United States likely have been treated with at least one of the SAT molecules, due to their proven efficacy and the ease of administration. That being said, the clinical value of SAT is limited due to the inevitable emergence of resistance. For example, the first-in-class abiraterone approved by the FDA in 2010 demonstrated an overall survival benefit of only 3.9 months in men with post-docetaxel CRPC (14). Similarly, the second molecule approved in the SAT class, enzalutamide, increased overall survival in the same group by a mere 4.8 months (16, 17). Therefore, for men with lethal prostate cancer, developing novel therapies for SAT-resistant disease is a priority.
In this context, one major mechanism of escape from SAT is AR-V7 (18–20). AR-V7 is an AR variant with a truncated C-terminal region that contains the ligand-binding domain (LBD; ref. 21). Due to this truncation, AR-V7 is constitutively active. It has been confirmed that AR-V7 expression is induced by ADT and correlates with resistance to both enzalutamide and abiraterone in patients with prostate cancer (22, 23). AR-V7 has also been shown to cause resistance to abiraterone, enzalutamide, and apalutamide (24–28) and predicts poor response clinically to abiraterone and enzalutamide (29). More recently, AR-V7 has been reported to be expressed in approximately 75% of CRPC and increased further with abiraterone or enzalutamide treatment (18). Based on this body of evidence, AR-V7 has been suggested to be a viable therapeutic target in men with SAT-resistant CRPC (30).
As heterobifunctional molecules, proteolysis targeting chimeras (PROTACs) consist of three key structural components: protein-of-interest (POI) ligand, ubiquitin E3 ligase ligand, and linker (31–33). In principle, PROTACs catalyze protein degradation by recruiting ubiquitin E3 ligases to promote polyubiquitination of the POI and its subsequent degradation in the proteasome. Because one PROTAC molecule can degrade multiple POIs, the standard target occupancy pharmacology does not apply and a relatively lower concentration is sufficient to be active (34). Herein, we present the first study on MTX-23, a PROTAC small molecule that degrades both AR-V7 and AR-full length (FL).
Materials and Methods
MTX-23 synthesis and cell culture
MTX-23 was synthesized as described in the Supplementary Methods section. Human prostate cancer cell lines LNCaP, 22Rv1, VCaP, PC3, and DU145 were obtained from the ATCC on December 1, 2007, February 9, 2010, March 3, 2008, August 7, 2005, and August 7, 2005, respectively, and maintained in the standard culture media: RPMI-1640 supplemented with 10% fetal bovine serum. Cell line passages used were as follows: LNCaP 30–40, 22Rv1 30–40, VCaP 30–40, PC3 50–60, and DU145 50–60. LNCaP, 22Rv1, and VCaP are androgen-responsive, whereas PC3 and DU145 are not. To establish SAT-resistant prostate cancer cell lines, LNCaP, 22Rv1, and VCaP were treated continuously with 10–50 μmol/L abiraterone (Selleckchem; cat. No. S1123), apalutamide (Selleckchem; cat. No. S2840), darolutamide (Selleckchem; cat. No. S7559), or enzalutamide (Selleckchem; cat. No. S1250). After 3–6 months, viable cells were pooled and designated as LNCaP-AbiR, LNCaP-ApaR, LNCaP-DaroR, LNCaP-EnzR, VCaP-AbiR, VCaP-ApaR, VCaP-DaroR, VCaP-EnzR, 22Rv1-AbiR, 22Rv1-ApaR, 22Rv1-DaroR, and 22Rv1-EnzR. Unless specified, the standard culture media for these SAT-resistant cell lines included 10 μmol/L of their respective SAT. For proteasome inhibitor study, we used MG132 (EMD Millipore; cat. No. 474791) and epoxomicin (Selleckchem; cat. No. S7038). E3 ligase inhibitors were purchased from Tocris (cat. No. 5433 for Heclin, cat. No. 6075 for Nutlin 3a, cat. No. 0652 for Thalidomide, and cat. No. 6156 for VH298). To assess the effect of splicing, transcription inhibitors were used—actinomycin D (ActD; Sigma-Aldrich; cat. No. A9415-2MG), benzimidazole (DRB; Sigma-Aldrich; cat. No. 194123-5G), and trichostatin A (TSA; Selleckchem; cat. No. S1045). After pretreatment of 22Rv1 with 1 μmol/L ActD, 5 μmol/L DRB, or 10 nmol/L TSA for 2 hours, vehicle (DMSO) or 1 μmol/L of MTX-23 was added for 1, 2, 4, 8, and 16 hours. Subsequently, RNA was purified using TRIzol reagent (Thermo Fisher Scientific). We used 18s rRNA as the control. Primer sequences are shown in Supplementary Table S1.
Cell line authentication
Although ATCC provides cell authentication data, we have confirmed the cells by checking their morphology using optical microscopy, establishing baselines for cell proliferation, verifying species of origin using isoenzymology (ATCC, cat. No. 135-XV-10), and characterizing the cell's DNA fingerprint using short tandem repeat profiling (PowerPlex 18D, cat No. DC1802, Promega). Mycoplasma contamination was assessed using a PCR-based detection system (Sigma-Aldrich; cat MP0035-1kit). The last date of cell line authentication was March 1, 2019.
Apoptosis assay
Apoptosis assay was carried out using the Thermo Fisher's ApoDETECT Annexin V–FITC kit (cat. No. 33-1200). The protocol recommended by the vendor was used. Briefly, after treatment with 1 μmol/L MTX-23 for 3 to 24 hours, cells were fixed with 80% ethanol and washed with PBS three times. Then, fixed cells were incubated with Annexin V–FITC in PBS solution for 30 minutes at room temperature. After washing three times with PBS, cells were treated with 300 nmol/L DAPI in PBS for 5 minutes at room temperature. Finally, after washing three times with PBS, mounting solution (cat No. H-1400; Vector Laboratories) was added and visualized using immunofluorescence microscopy.
TUNEL assay was performed using Promega DeadEnd Fluorometric TUNEL system (cat. No. G3250). After MTX-23 treatment and fixation as described in the Annexin V experiment, 100 μL equilibration buffer was incubated for 10 minutes. Then, 50 μL of TdT reaction mix was added and incubated for 60 minutes at 37°C in humidified chamber. Finally, stop solution was added and mounted on slides using mounting medium. To assess nonspecific cytotoxicity, the LDH assay kit was used (Thermo Fisher; cat. No. 88954).
Transient transfections
One microgram of a plasmid containing cDNA of AR-V7 (cat. No. 64636; Addgene) or AR-FL (cat. No. sc114220, Origene) was transfected into indicated prostate cancer cell lines on a 6-well plate. Three microliters of Lipofectamine 3000 (Thermo Fisher, cat. No. L3000015) was used for each transfection.
As for the knockdown study, siRNA was used (Sigma-Aldrich). siRNA (500 ng) with 1-μL Lipofectamine 3000 (Thermo Fisher) in 24 wells was used. After transfection of siRNA for AR-FL and/or AR-V7, 22Rv1-EnzR cells were cultured for 6 days. Media were changed after 3 days. siRNA sequences are: 5′GAAGGCCAGUUGUAUGGAC3′ (AR-FL) and 5′GACCAGACCCUGAAGAAAG3′ (AR-V7).
Immunoblot analysis
Prostate cancer cells were collected and lysed with the lysis buffer (20 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, 1 mmol/L Na2EDTA, 1 mmol/L EGTA, 1% Triton, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L beta-glycerophosphate, 1 mmol/L Na3VO4, and 1 μg/mL leupeptin) containing 1 mmol/L phenylmethylsulfonyl fluoride (PMSF). Cell lysates were then centrifuged, and protein in the supernatant was quantified. After separating 25–50 μg protein using SDS-PAGE, samples were incubated with AR-V7 (Fisher Scientific; cat No. NC0752138), AR-V567 (Abcam; cat. No. ab200827), glucocorticoid receptor (GR; Thermo Fisher; cat. No. MA1-510; progesterone receptor (PR; Thermo Fisher, cat. No. MA1-410), ERα (Santa Cruz Biotechnology; cat. No. sc-8002), AR-FL (Santa Cruz Biotechnology; cat. No. sc-7305), ubiquitin (Cell Signaling Technology; cat No. 3933S) or β-actin (Sigma-Aldrich; cat. No. A5441) antibodies. For AR-V7, AR-FL, PR, GR, and ERα immunoblots, primary antibody was diluted 1:1,000 in 5% skim milk. For the β-actin immunoblot, 1:10,000 diluted primary antibody was used. All membranes were incubated overnight at 4°C. Following the incubation with appropriate secondary antibody, immunoblot was analyzed using SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher).
In vivo study
All mice and experimental procedures were conducted using protocols approved by and in accordance with the Rutgers University Institutional Animal Care and Use Committee approval (PROTO 999900168) and the NIH Guide for the Care and Use of Laboratory Animals. For anesthesia, 3% isofluorane gas inhalation was used. Nu/nu mice were purchased from The Jackson Laboratory. Where indicated, surgical castration was carried out via bilateral orchiectomy. Tumor size was measured using calipers, and tumor volume was calculated using the formula: tumor volume = length × width2/0.361.
To explore the therapeutic potential of MTX-23, 22Rv1, 22Rv1-EnzR, VCaP, and VCaP-EnzR were injected into nu/nu immunodeficient mice. When tumors reached an average size of 3 mm in diameter, all animals were surgically castrated and divided into four groups of five mice each. Mice were treated daily with MTX-23 with or without enzalutamide via the indicated route (intratumoral, intraperitoneal, or oral) for 5 to 6 weeks. All animals were sacrificed by CO2 asphyxiation at the end of the study, and tumors were harvested and analyzed.
Statistical analysis
Statistical significance was calculated using the Student t test for paired comparisons of experimental groups and, where appropriate, by Wilcoxon rank-sum test, and by two-way ANOVA. In vitro experiments were repeated a minimum of three times. Continuous variables were expressed as mean ± standard deviation (SD).
Results
MTX-23 is a novel PROTAC small molecule that degrades both AR-V7 and AR-FL
To develop PROTACs that target AR-V7, we started with molecules that have been reported to bind to the DBD of AR-FL (35, 36), because AR-V7 is an AR splice variant that lacks the entire LBD and contains only the N-terminal regulatory domain and DBD (21, 22). In the class of PROTACs exemplified by MTX-23, potential AR DBD binding molecules were linked to E3 CRL2VHL compounds, also called VHL ligands (35, 37, 38). Potential linkers with varying lengths and conformational flexibility were analyzed. This effort led to the development of our lead molecule MTX-23 (Fig. 1A). When the AR-V7–expressing prostate cancer cell line 22Rv1 was treated with MTX-23 for 24 hours, immunoblot demonstrated decreased AR-V7 and AR-FL protein levels in 22Rv1 cells starting at concentration as low as 0.1 μmol/L and 1 μmol/L, respectively (Fig. 1B). MTX-23's effect on AR-V7 and AR-FL protein levels were concentration-dependent and AR-specific as there was no visible effect on the GR, PR a and b, and estrogen receptor alpha (ER) expression levels. The concentration of MTX-23 at which 50% of AR-V7 and AR-FL degraded in 24 hours (DC50) were 0.37 and 2 μmol/L, respectively (Fig. 1C). Consistent with the concept that MTX-23 binds to AR's DBD, we also observed that MTX-23 decreases the protein levels of another AR splice variant, AR-V567 (Supplementary Fig. S1). ARV-567 mRNA level was not affected. To determine whether MTX-23 affects AR splicing, we measured AR-FL and AR-V7 mRNA levels after treating cells with three different transcription inhibitors—actinomycin D (ActD), benzimidazole (DRB), and trichostatin A (TSA; ref. 39). The results demonstrated that MTX-23 had no significant impact on AR-FL and AR-V7 mRNA levels when new transcription was blocked, suggesting that MTX-23 does not affect AR splicing (Supplementary Fig. S2). Therefore, MTX-23 is a novel PROTAC small molecule that degrades both AR-V7 and AR-FL. However, MTX-23's degradation effect was more efficient against AR-V7 when compared with AR-FL.
Screening of MTX-23. A, MTX-23 chemical structure. B, Immunoblot for AR-V7 and AR-FL in 22Rv1 cells. MTX-23 treatment for 24 hours decreased AR-V7 protein levels starting at 0.01 μmol/L while AR-FL was affected at 1 μmol/L or higher concentrations. GR, PR, and ERα (ER) protein levels were not affected by MTX-23. C, DC50 of MTX-23 for AR-V7 and AR-FL. Quantitation of three immunoblots demonstrated MTX-23's DC50 for AR-V7 and AR-FL to be 0.37 and 2 μmol/L. D, Cell proliferation assay. 22Rv1 was treated for 6 days with increasing concentrations of MTX-23, MTX-23′s DNA binding motif containing the linker (C1) and VHL E3 recruiting motif containing the linker (C2). Only MTX-23 inhibited the proliferation of 22Rv1 starting at 0.01 μmol/L in a concentration-dependent manner. E, Effect of MTX-23 on AR-positive and -negative human prostate cancer cell lines. MTX-23 effect is specific for androgen-responsive prostate cancer cell lines, 22Rv1, LNCaP, and VCaP. AR-negative cell lines, DU145 and PC3, were not significantly affected by MTX-23 at the indicated concentrations. F, MTX-23's inhibitory effect is partially reversed when AR-V7 or AR-FL were transfected. Complete abrogation of MTX-23's effect was seen when both AR-V7 and AR-FL were cotransfected. *, P < 0.05.
Along with degrading AR-V7 and AR-FL, MTX-23 decreased the cell count of 22Rv1 in a concentration-dependent manner (Fig. 1D). As controls, DBD binding motif with linker and CRL2VHL with linker were used (C1 and C2, respectively). At 1 μmol/L of MTX-23, 22Rv1 cell count was approximately 40% of the control. In additional human prostate cancer cell lines, MTX-23 only inhibited the proliferation of androgen-responsive cells (22Rv1, LNCaP, and VCaP; Fig. 1E). The AR-negative cell lines PC3 and DU145 were not affected by MTX-23. Upon transfecting AR-V7 or AR-FL into 22Rv1 cells, there was a partial resistance to MTX-23's inhibitory effect (Fig. 1F). However, with the transfection of both AR-V7 and AR-FL, MTX-23's effect was completely abrogated. In addition, MTX-23 10 μmol/L (concentration sufficient for degrading both AR-FL and AR-V7) was more effective than knocking down AR-FL and AR-V7 individually using siRNA (Supplementary Fig. S3). These initial results suggest that MTX-23 inhibits prostate cancer cellular proliferation by degrading both AR-V7 and AR-FL.
MTX-23 induces apoptosis
After treating 22Rv1 cells with 1 μmol/L MTX-23, Annexin V assay was carried out to assess the effect on apoptosis. Representative result is shown in Fig. 2A. Starting approximately 3 hours after MTX-23 treatment, Annexin V staining increased. This result was confirmed by the TUNEL assay (Fig. 2B). As a negative control, the DBD binding motif with linker C1 was used. Using the lactate dehydrogenase (LDH) assay, we observed that MTX-23 up to 20 μmol/L had no nonspecific cytotoxic effect in 22Rv1 cells after 6 days of treatment (Fig. 2C).
MTX-23 stimulates apoptosis in 22Rv1 cells. A, 22Rv1 was incubated with 1 μmol/L of MTX-23, and Annexin V was measured using fluorescence microscopy at indicated time points. Indicating apoptosis, Annexin V was detected as early as 3 hours after MTX-23 exposure. B, TUNEL assay was performed. C1 (AR DBD binding motif with linker) is a negative control. Again, MTX-23 treatment induced apoptosis. C, MTX-23 has no detectable nonspecific cytotoxic activity. Following MTX-23 treatment for 24 hours, there was no significant increase in LDH activity in LNCaP, 22Rv1, and VCaP.
MTX-23 induces poly-ubiquitination of AR-V7 and AR-FL
When 22Rv1 was pretreated for 2 hours with the proteasome inhibitors MG-132 and epoxomicin prior to treatment with MTX-23 at 1 μmol/L for 24 hours, AR-V7 and AR-FL degradation was completely blocked (Fig. 3A). Subsequently, the following E3 ubiquitin ligase inhibitors were examined: VH 298 (VHL inhibitor), heclin (HECT inhibitor), nutlin 3a (MDM2 inhibitor), and thalidomide (cereblon inhibitor; Fig. 3B). When 22Rv1 cells were pretreated with each of these molecules for 2 hours prior to incubation with 1 μmol/L MTX-23, only VH298 pretreatment inhibited AR-V7 and AR-FL degradation. To determine the ubiquitination status, 22Rv1 was pretreated with MG132 for 2 hours prior to adding MTX-23 (Fig. 3C). Immunoblot analysis demonstrated polyubiquitination of AR-V7 starting at 6 hours of MTX-23 treatment. AR-FL polyubiquitination was also detected but occurred significantly later at 24 hours after MTX-23 treatment. Collectively, these results suggest that MTX-23 stimulates AR-V7 and AR-FL ubiquitination and degradation specifically via the VHL E3 ligase/proteasome axis.
Mechanism of MTX-23 for cell proliferation inhibition. A, Effect of proteasome inhibitors. Reduced AR-V7 and AR-FL protein levels following MTX-23 1 μmol/L treatment for 24 hours was completely blocked by 2 hours of pretreatment with the proteasome inhibitors, MG132 or epoxomicin. B, Among various E3 ligase-specific inhibitors, VH298 (VHL E3 ligase-specific inhibitor) pretreatment for 2 hours inhibited the reduction of AR-V7 and AR-FL protein levels by MTX-23 1 μmol/L after 24 hours. Heclin, Nutlin 3a, and thalidomide E3 ligase inhibitors had no effect on MTX-23–stimulated AR-V7 protein degradation. C, Ubiquitination of AR-FL and AR-V7 following MTX-23 treatment in 22Rv1. Cells were pretreated with MG132 for 2 hours prior to MTX-23 exposure. At the indicated time point after adding MTX-23 1 μmol/L, cell lysate was harvested and immunoprecipitated using primary antibody against ubiquitin. Then, immunoblot for AR-V7 or AR-FL was carried out. Polyubiquitination of AR-V7 was detected starting at 6 hours of MTX-23 treatment. In contrast, polyubiquitination of AR-FL was detected at 24 hours of MTX-23 treatment.
MTX-23 is effective against SAT-resistant prostate cancer cell lines
To assess the potential therapeutic role of MTX-23, we generated 12 human prostate cancer cell lines that are resistant to the four FDA-approved SAT agents: abiraterone, apalutamide, enzalutamide, and darolutamide. Specifically, LNCaP, VCaP, and 22Rv1 cells were cultured with each of the SAT agents for three to six months until resistance emerged. Resulting cells were designated LNCaP-AbiR, LNCaP-ApalR, LNCaP-DarolR, LNCaP-EnzR, VcaP-AbiR, VcaP-ApalR, VcaP-DarolR, VcaP-EnzR, 22Rv1-AbiR, 22Rv1-ApalR, 22Rv1-DarolR, and 22Rv1-EnzR. Quantitative PCR and immunoblot demonstrated that all 12 SAT-resistant prostate cancer cell lines express slightly decreased mRNA and protein levels of AR-FL (Supplementary Fig. S4). On the other hand, AR-V7 mRNA and proteins levels were uniformly higher in these SAT-resistant cells. All 12 cell lines' cell counts decreased when cultured with 1 μmol/L MTX-23 for 6 days (Fig. 4A). Immunoblot assays with 22Rv1 cells demonstrated that MTX-23 decreased the protein levels of both AR-V7 and AR-FL in the parental and SAT-resistant cells (Fig. 4B). Results of cell lines derived from VCaP and LNCaP are shown in Supplementary Figs. S2 and S3, respectively. Again, in all cell lines, the effect on the protein levels of AR-V7 was greater than that on AR-FL. As negative controls, the AR-negative PC3 and DU145 cells were used. As predicted based on the mechanism of action of PROTAC, MTX-23 had no major effect on the mRNA levels of AR-FL and AR-V7 (Fig. 4C and D; Supplementary Figs. S5 and S6).
Effect of MTX-23 on SAT-resistant prostate cancer cell lines. LNCaP, VCaP, and 22Rv1 were treated with the FDA-approved SAT agents, abiraterone, apalutamide, enzalutamide, and darolutamide for three to six months to establish resistant cells. A, MTX-23 1 μmol/L treatment for 6 days decreased the proliferation all SAT-resistant prostate cancer cells. However, AR-negative prostate cancer cell lines, PC3 and DU145, were not affected by MTX-23. B, AR-FL and AR-V7 protein levels were decreased by MTX-23 1 μmol/L treatment in all four SAT-resistant 22Rv1 cells. The effect on AR-V7 was more efficient than that on AR-FL. PC3 cells do not express AR-FL and AR-V7. Con = 22Rv1 parental line. C, AR-FL mRNA expression levels were not affected by MTX-23 1 μmol/L after 24 hours in SAT-resistant 22Rv1 cells. D, AR-V7 mRNA expression levels were not affected by MTX-23 treatment.
MTX-23 inhibits enzalutamide-resistant prostate cancer growth in mice
In vivo effects of MTX-23 were assessed in mice using enzalutamide-resistant prostate cancer xenografts. We first established 22Rv1-EnzR tumor xenografts in nu/nu mice. Upon prostate cancer xenograft formation with the average diameter of 3 mm, mice were randomized into control or treatment group. Control mice were treated with 100-μL vehicle containing 10 mg/mL enzalutamide (n = 5). The treatment group was injected with 100 μL containing 2.5 mg/mL MTX-23 and 10 mg/mL enzalutamide daily directly into tumors (n = 5) and tumor volumes were followed. The results demonstrated that mice treated with MTX-23 had a significantly smaller tumor volume at the end of 5 weeks (Fig. 5A). During the treatment period, weight of the mice did not significantly change (Fig. 5B), suggesting that MTX-23 is well tolerated in mice. When tumors were harvested and analyzed at the end of the study period, immunoblot demonstrated a significant decrease in AR-V7 and AR-FL protein levels in all treated tumors (Fig. 5C). Similar results were obtained when MTX-23 was injected intratumorally into VCaP-EnzR xenografts (Supplementary Fig. S7).
MTX-23 inhibits enzalutamide-resistant 22Rv1 (22Rv1-EnzR) tumor growth in mice. Nu/nu mice were inoculated with 22Rv1-EnzR subcutaneously. When tumors reached an average size of 3 mm, animals were divided into two groups of five each. Mice in the treatment group received daily intratumoral injection of 100 μL of 2.5 mg/mL MTX-23 and 10 mg/mL enzalutamide. Control group received the vehicle/enzalutamide injection only. A, Over a 5-week period, MTX-23 treatment significantly decreased tumor growth. B, During the same treatment period, MTX-23 treatment had no significant impact on body weight. C, Immunoblot demonstrated that AR-V7 and AR-FL protein levels significantly decreased in MTX-23–treated tumors.
To test whether MTX-23 is active systemically, we carried out an identical study as above except for delivering MTX-23 intraperitoneally (IP) or orally. Enzalutamide-resistant prostate cancer tumor xenografts were established in 12 mice, and three each was assigned to the following four groups: control IP (vehicle), control oral (vehicle), 8.3 mg/kg IP, and 8.3 mg/kg oral. Because these cells are resistant to enzalutamide, all animals were administered enzalutamide. The results demonstrated that MTX-23 was effective when administered IP (Fig. 6A). Again, no significant change in weight was detected following MTX-23 treatment (Fig. 6B). Immunoblot showed that IP administration of MTX-23 decreased AR-V7 and AR-FL protein levels in EnzR prostate cancer tumors (Fig. 6C). Similar results were found when MTX-23 was delivered orally (PO; Fig. 7).
Intraperitoneal (IP) injection of MTX-23 inhibits 22Rv1-EnzR tumor growth in mice. Nu/nu mice were inoculated with 22Rv1-EnzR subcutaneously. When tumors reached an average diameter of 3 mm, animals were divided into two groups of three mice each. Mice in the first group received daily 8.3 mg/kg MTX-23 and 10 mg/kg enzalutamide IP injection. Control group received the vehicle/enzalutamide injection only. A, Over a 4-week period, IP injection of MTX-23 significantly decreased tumor growth. B, During the same treatment period, IP administration of MTX-23 had no significant impact on body weight. C, Immunoblot demonstrated that AR-V7 and AR-FL protein levels significantly decreased in MTX-23–treated tumors.
Oral administration of MTX-23 inhibits 22Rv1-EnzR tumor growth in mice. Nu/nu mice were inoculated with 22Rv1-EnzR subcutaneously. When tumors reached an average diameter of 3 mm, animals were divided into two groups of three each. Mice in the treatment group received daily oral administration of 8.3 mg/kg MTX-23 and 10 mg/kg enzalutamide. Control group received the vehicle/enzalutamide only. A, Over a 4-week period, oral administration of MTX-23 significantly decreased tumor growth. B, During the same treatment period, oral MTX-23 treatment had no significant impact on body weight. C, Immunoblot demonstrated that AR-V7 and AR-FL protein levels significantly decreased in MTX-23–treated tumors.
Discussion
In the present study, we have demonstrated that MTX-23 is a novel PROTAC small molecule that degrades both AR-V7 and AR-FL. In tissue culture, MTX-23 inhibited the proliferation of only androgen-responsive prostate cancer cell lines. The underlying mechanism for such inhibitory effect was apoptosis mediated via the degradation of both AR-V7 and AR-FL. Importantly, MTX-23 had a significant antiproliferative activity in SAT-resistant prostate cancer cells. Collectively, these results suggest that MTX-23 may be a novel treatment option in men with CRPC that have progressed on current standard of care.
Although MTX-23 degraded both AR-V7 and AR-FL, the degradation effect was more efficient for AR-V7 than AR-FL (DC50 of 0.37 vs. 2 μmol/L). This was a surprising observation because MTX-23 has been designed to target the DBD of AR. Accordingly, MTX-23 should theoretically degrade both AR-V7 and AR-FL with a similar efficacy. At the present time, we do not have a clear explanation for this result. One plausible possibility is that the retained LBD, or proteins that bind to the LBD, in AR-FL may physically interfere with MTX-23's access to the DBD. Currently, further studies are underway to assess how MTX-23 interacts with AR-V7 and AR-FL.
An alternative explanation for the preferential effect of MTX-23 on AR-V7 over AR-FL is that the biological consequence of MTX-23-mediated ubiquitination may be different between AR-V7 and AR-FL. To date, at least three functional ubiquitination sites within AR have been reported. Of these, two are ubiquitinated by RNF6 E3 ligase and located at the c-terminal LBD (40), while one is ubiquitinated by MDM2 E3 ligase and located at the N-terminal regulatory domain's lysine residue K311 (41). Because the N-terminal regulatory domain is present in both AR-FL and most AR splice variants including AR-V7, K311 ubiquitination should theoretically occur in both AR-V7 and AR-FL. Yet, it has been reported that K311 was ubiquitinated only in AR-FL and not AR-V7. Therefore, it is possible that MTX-23 may be more efficient at ubiquitinating AR-V7 than AR-FL. A more detailed study on MTX-23–mediated ubiquitination is necessary to verify this concept.
The time difference in ubiquitination between AR-V7 and AR-FL provides the initial clue as to the specific mechanism for MTX-23's more efficient degradation effect on AR-V7. Specifically, AR-V7 polyubiquitination was detected starting at 6 hours whereas AR-FL polyubiquitination at 24 hours. Such delay is consistent with the concept of lower MTX-23 binding affinity for AR-FL than AR-V7. Indeed, near identical ubiquitination pattern suggests that MTX-23 ubiquitinates the same lysine residues in AR-V7 and AR-FL. Additional studies such as surface plasmon resonance and bioluminescence resonance energy transfer will be necessary to assess the binding affinity of MTX-23 for AR-V7 and AR-FL in detail.
Although MTX-23 was more efficient in targeting AR-V7 than AR-FL, the degradation effect of AR-FL is likely relevant and important. Specifically, MTX-23 stimulated apoptosis within 3 hours. Given that SAT-resistant prostate cancer cells express both AR-V7 and AR-FL, degrading AR-V7 exclusively should not result in apoptosis. Indeed, it has been recently reported that a PROTAC-mediated degradation of AR-FL alone even in the presence of AR-V7 was sufficient for inducing apoptosis (42). In this regard, we have also observed that the effect of MTX-23 is completely blocked only when both AR-V7 and AR-FL are overexpressed.
It should be noted that MTX-23 was effective again all SAT-resistant prostate cancer cell lines. Yet, the fundamental mechanism of action of abiraterone is different from that of apalutamide, enzalutamide, and darolutamide. Specifically, abiraterone inhibits androgen synthesis, whereas the latter three block the binding of androgen to AR (43). Nevertheless, the observation that AR-V7 expression levels increased universally in SAT-resistant cells confirms the importance of AR variants as an escape mechanism from the “super” suppression of androgen signaling. In this regard, MTX-23's potential therapeutic role may not be limited by prior treatment, but rather by the status of AR-V7 and AR-FL expression.
As potential therapeutic option for treating SAT-resistant CRPC, three PROTACs—ARD-61, ARD-69, and ARCC-4—have been reported to date (42, 44, 45). Most recently, ARCC-4 has been further optimized to ARV-110 and has entered the phase I clinical trial (NCT03888612). In this regard, it should be noted that MTX-23 exhibits antitumorigenic effect in the low micromolar range, whereas the aforementioned agents are effective at high nanomolar concentrations. Currently, we have begun exploring various modifications of MTX-23 to enhance its potency. Notwithstanding, we posit that MTX-23 is a reasonable molecule for further studies with the goal of clinical translation for the following three reasons. First, although ARD-61, ARD-69, and ARCC-4/ARV-110 target AR-FL only, MTX-23 is unique in that AR-V7 is degraded efficiently. Therefore, we propose that MTX-23 will likely be complementary to ARD-61 and ARCC-4/ARV-110 and enhance their therapeutic effect. Second, in our assays, IC50 concentration for MTX-23 is significantly lower and was in the 1 μmol/L range when compared with that of the FDA-approved SAT agents abiraterone, enzalutamide, apalutamide, and darolutamide which were in the 10 μmol/L range. Third, a relatively higher dose of MTX-23 compared with ARD-61, ARD-69, and ARCC-4 should be well tolerated because the degradation effect was more profound with AR-V7 when compared with AR-FL. Because normal cells do not express AR-V7, MTX-23's toxicity profile is expected to be very favorable. In support of this concept, the in vivo mouse studies demonstrated that MTX-23 is well tolerated at the therapeutic dosage in treating EnzR prostate cancer tumors. Notwithstanding, because no formal toxicity study has been carried out, the stability in body weight during MTX-23 treatment is only suggestive of MTX-23's safety.
Because degradation of AR variants is another form of inhibiting androgen signaling pathway, it is likely that MTX-23 will not be curative against SAT-resistant CRPC. Indeed, just as AR-variant expression emerged prominently with SAT, prostate cancer cells resistant to MTX-23 are inevitable. In addition, we would like to point out that MTX-23 will be ineffective against the AR-negative lethal neuroendocrine-like prostate cancer. Currently, studies are underway to assess MTX-23 resistance.
In conclusion, MTX-23 is a novel PROTAC molecule that degrades both AR-V7 and AR-FL. Due to the preferential effect on AR-V7 over AR-FL, MTX-23 should have a favorable efficacy-to-toxicity ratio. Future studies will focus on how MTX-23 interacts with AR-V7 and AR-FL as well as establishing essential pharmacokinetics in preparation for a clinical trial in men with SAT-resistant CRPC.
Authors’ Disclosures
J. Desantis reports a patent for US 16/876949 pending. H.E. Sabaawy reports grants from NCI outside the submitted work; Dr. Sabaawy is the scientific founder of Celvive, Inc. R.J. Vaz reports a patent for 16/876949 pending. I.Y. Kim reports other from Montelino Therapeutics during the conduct of the study; grants and personal fees from Janssen and personal fees from Bayer and Sanofi-Aventis outside the submitted work. No disclosures were reported by the other authors.
Authors' Contributions
G.T. Lee: Data curation, formal analysis, investigation, methodology, writing–original draft, and project administration. N. Nagaya: Data curation, methodology, writing–review and editing. J. Desantis: Data curation, methodology, writing–review and editing. K. Madura: Supervision, methodology, writing–review and editing. H.E. Sabaawy: Supervision, methodology, writing–review and editing. W.-J. Kim: Supervision, methodology, writing–review and editing. R.J. Vaz: Conceptualization, resources, supervision, methodology, writing–review and editing. G. Cruciani: Conceptualization, resources, formal analysis, supervision, methodology, writing–review and editing. I.Y. Kim: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, methodology, writing–original draft, project administration, writing–review and editing.
Acknowledgments
This work was supported in part by Montelino Therapeutics, the cancer center grant from the NCI (grant P30CA072720), and generous support from the Marion and Norman Tanzman Charitable Foundation and Mr. Malcolm Wernik.
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.
Footnotes
Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).
Mol Cancer Ther 2021;20:490–9
- Received May 17, 2020.
- Revision received August 29, 2020.
- Accepted November 30, 2020.
- Published first December 4, 2020.
- ©2020 American Association for Cancer Research.
References
Volume 20, Issue 3
