Androgen receptor inhibitors in treating prostate cancer

Abstract

Androgens play an important role in prostate cancer development and progression. Androgen action is mediated through the androgen receptor (AR), a ligand-dependent DNA-binding transcription factor. AR is arguably the most important target for prostate cancer treatment. Current USA Food and Drug Administration (FDA)-approved AR inhibitors target the ligand-binding domain (LBD) and have exhibited efficacy in prostate cancer patients, particularly when used in combination with androgen deprivation therapy. Unfortunately, patients treated with the currently approved AR-targeting agents develop resistance and relapse with castration-resistant prostate cancer (CRPC). The major mechanism leading to CRPC involves reactivation of AR signaling mainly through AR gene amplification, mutation, and/or splice variants. To effectively inhibit the reactivated AR signaling, new approaches to target AR are being actively explored. These new approaches include novel small molecule inhibitors targeting various domains of AR and agents that can degrade AR. The present review provides a summary of the existing FDA-approved AR antagonists and the current development of some of the AR targeting agents.

INTRODUCTION

Prostate cancer incidence rate is the second highest and the disease is a leading cause of cancer death in men worldwide.1 In the USA, prostate cancer is the most frequently diagnosed cancer and the second leading cause of cancer death in men.2 In Asian countries, the incidence rate (age-standardized rate [ASR] of 11.5)3 of prostate cancer is lower than that in North America (ASR of 73.7), but this can be partially attributed to less frequent testing for prostate-specific antigen (PSA), a common biomarker for prostate cancer.4 In a recent report, localized prostate cancer can be effectively treated, with a 10-year survival rate of 100%. However, the 5-year survival rate for patients with distant metastasis is only about 30%.5 A major challenge in prostate cancer research is to enhance the survival of patients with metastatic prostate cancer.

Androgens play a key role in the development and progression of prostate cancer. Androgen action is mediated through the androgen receptor (AR), a member of the nuclear receptor super family. AR consists of the N-terminal domain (NTD), the DNA-binding domain (DBD), the hinge region, and ligand-binding domain (LBD), as shown in Figure 1.6,7,8 Androgen action can be blocked by suppressing androgen production and/or inhibiting AR activation. Androgen deprivation therapy (ADT) was demonstrated to be effective for the treatment of prostate cancer by Huggins and his team more than 80 years ago.9 Surgical castration was initially the main approach of ADT. Subsequently, it was replaced by the treatment with leuprolide, an analog of gonadotropin-releasing hormone (GnRH), also called luteinizing hormone-releasing hormone (LHRH) that was first developed by Tolis et al.10 More recently in 2008, degarelix, a GnRH antagonist, was shown in a phase 3 clinical trial to induce PSA suppression faster than leuprolide.11 First-generation AR antagonists, which target the LBD of AR, were developed to enhance ADT.12 Unfortunately, patients on ADT eventually relapse and develop what was called androgen-insensitive, hormone-insensitive, androgen-resistant, hormone-resistant, or hormone-refractory prostate cancer, which is now called castration-resistant prostate cancer (CRPC).13

Figure 1.

Structure of AR and ARv7. The amino acid residues of various domains and their respective targeting agents are indicated.160,161 U in ARv7 means “unique.” AF-1: activation function-1; AF-2: activation function-2; AR: androgen receptor; ARv7: androgen receptor splice variant 7; DBD: DNA-binding domain; H: hinge region; LBD: ligand-binding domain; NTD: amino-terminal domain; Tau-1: transcription activation unit-1; Tau-5: transcription activation unit-5.

Multiple investigations showed that AR is the key driver of CRPC, including those that are resistant to the current USA Food and Drug Administration (FDA)-approved AR targeting agents.14,15,16,17,18,19 Thus, new agents targeting AR, including those targeting the NTD and DBD of AR (Figure 1), are being developed with the intention to overcome the resistance of CRPC to the current AR-targeting medications. Multiple review articles describing AR-targeting agents have been published over the years since antiandrogens were first developed.20,21,22,23,24,25,26 This article is intended to provide a review on AR-targeting agents, which is undoubtedly not comprehensive due to the very large number of AR-targeting agents in the literature. In addition to summarizing the FDA-approved AR-targeting agents, this review will describe the recent development of AR inhibitors and degraders that can potentially be effective on CRPC tumors resistant to the current AR-targeted therapies.

LBD-TARGETING AR ANTAGONISTS

Steroidal AR antagonists

Steroidal AR antagonists are a class of molecules that target the AR through competition with androgen binding to the LBD domain. Steroidal AR antagonists have the structures derived from steroids that contain three interconnected cyclohexane rings and one cyclopentane ring arranged in a particular pattern. The functional groups and side chains of steroids that are attached to this core structure give steroids unique and specific biological functions. Steroidal AR antagonists share this steroidal structure and are specifically designed to bind to the LBD of AR and compete with the binding of the endogenous ligand testosterone or dihydrotestosterone (DHT). Cyproterone acetate is a widely available steroidal AR antagonist that works in this manner.27 Unfortunately, steroidal AR antagonists have been found to have cross-reactivity with other steroid hormones and to have progestational and antigonadotropic effects.28 These effects are likely caused by their steroidal structures.

First-generation non-steroidal LBD-targeting AR antagonists

Although steroidal AR antagonists competitively inhibit DHT binding to the LBD, they are not the only class of molecule that does this. The different classes of LBD-binding AR antagonists are summarized in Supplementary Table 1. In contrast to the steroidal class of AR antagonists, nonsteroidal AR antagonists are also able to competitively bind to the LBD of AR, but have a structure unrelated to steroids. Therefore, nonsteroidal AR antagonists have not had the same negative side effects of steroidal antiandrogens and are more commonly used in prostate cancer treatment.

Supplementary Table 1.

Ligand-binding domain-targeting androgen receptor antagonists/inhibitors

Compound	Structure	Description	IC50	Stage of development	Reference
Cyproterone acetate		Steroidal antagonist	4.4 nmol l−1 (3H labeled DHT)	Not clinically approved in US	PMID: 1825143 PMID: 16206830
Flutamide		First-generation nonsteroidal antagonist	50 nmol l−1 (3H labeled testosterone)	FDA approved	PMID: 788293 PMID: 3956856
Bicalutamide		First-generation nonsteroidal antagonist	160 nmol l−1 (LNCaP)	FDA approved	PMID: 8560677 PMID: 19359544
Nilutamide		First-generation nonsteroidal antagonist	412 nmol l−1 (mouse mammary carcinoma cells)	FDA approved	PMID: 8997470 PMID: 9111629
Enzalutamide		Second-generation nonsteroidal antagonist	21.4 nmol l−1 (LNCaP)	FDA approved	PMID: 19359544
Apalutamide		Second-generation nonsteroidal antagonist	16 nmol l−1 (LNCaP)	FDA approved	PMID: 31150574 PMID: 22266222
Dirlotapides		Second-generation nonsteroidal antagonist	26 nmol l−1 (AR-HEK293 transactivation)	FDA approved	PMID: 30763142 PMID: 26137992
Proxalutamide		Second-generation nonsteroidal antagonist	32 nmol l−1 (competitive binding assay)	Phase 2/3 trial	https://doi.org/10.1158/1538-7445.AM2014-614 PMID: 32460179
TRC-253		Nonsteroidal antagonist	197 nmol l−1 (LNCaP)	Phase 1/2A trial	PMID: 33649102
CPPI		Nonsteroidal antagonist	4.3 µmol l−1 (C4-2)	Preclinical	PMID: 27187604
EPPI		Nonsteroidal antagonist	2.3 µmol l−1 (C4-2)	Preclinical	PMID: 27187604
Galeterone		LBD inhibitor	6 µmol l−1 (LNCaP); 3.2 µmol l−1 (LAPC-4)	Failed Phase 3 trial	PMID: 21729712
MDG506		AF-2 inhibitor	26.3 µmol l−1 (AR TR-FRET assay)	Preclinical	PMID: 22280402
SPC002		AF-2 inhibitor	24 µmol l−1 (AR TR-FRET assay)	Preclinical	PMID: 25233256
LHJ-647		AF-2 inhibitor	~1 µmol l−1 (LNCaP)	Preclinical	PMID: 29050220
IMB-A6		AF-2 inhibitor	1–10 µmol l−1 (LNCaP)	Preclinical	PMID: 30555332
T1-12		AF-2 inhibitor	1.42 µmol l−1 (LNCaP)	Preclinical	PMID: 35077161
DIM20		LBD dimerization inhibitor	1.3 µmol l−1 (ClARE)	Preclinical	PMID: 36218067
DIM20.39		LBD dimerization inhibitor	5.1 µmol l−1 (ClARE)	Preclinical	PMID: 36218067

AF-2: activation function-2; AR: androgen receptor; DHT: dihydrotestosterone; IC₅₀: half-maximal inhibitory concentration; LBD: ligand-binding domain; TR-FRET: time-resolved fluorescence resonance energy transfer; FDA: Food and Drug Administration

Since the recognition that AR could be targeted through the LBD using nonsteroidal structures, there has been a concerted effort to develop new AR antagonists. Flutamide, the first nonsteroidal AR antagonist, was developed in 1967, tested in patients in the 1970s, and approved by the FDA in 1989.29 It was originally used in combination with GnRH agonists.30 Although its efficacy was demonstrated as a combination therapy, the results were less conclusive when flutamide was used as a monotherapy.31,32 As such, flutamide is not currently listed as a monotherapeutic drug for metastatic prostate cancer.

Bicalutamide and nilutamide are additional first-generation AR antagonists that were subsequently developed. Bicalutamide, in particular, has been widely used and investigated in prostate cancer patients, with 50 mg of bicalutamide once a day showing comparable efficacy to 250 mg of flutamide three times a day.33 This increased efficacy is likely due to bicalutamide exhibiting 2–4 times greater affinity than flutamide for binding the LBD of AR.34 Bicalutamide antagonizes AR by binding to an additional site adjacent to the binding site of androgen.35 Moreover, contrary to flutamide, bicalutamide showed convincing evidence of efficacy as a monotherapy.36 Despite these promising results, patients receiving bicalutamide alone had lower survival rates than patients receiving surgical or medical castration.37 Importantly, additional evidence began to elucidate the potential failures of bicalutamide treatment. Under androgen ablation conditions, bicalutamide could transition from an AR antagonist to a partial AR agonist.38 The implications of this finding indicated that in certain patients, bicalutamide could be contributing to prostate cancer tumor growth and progression. As such, better AR antagonists known as second-generation AR antagonists were developed.

Second-generation LBD-targeting AR antagonists: enzalutamide, apalutamide, darolutamide, and proxalutamide

Since AR was established as an important drug target for inhibiting prostate cancer growth and progression, finding more potent LBD-targeting agents with no agonistic activity was a major goal. From a screening of structural analogs of a high-affinity AR agonist, the second-generation AR antagonist enzalutamide was discovered.39 Enzalutamide exhibited excellent pharmacokinetic properties including high oral availability40 and long serum half-life.41 Furthermore, enzalutamide has very high affinity to the LBD of AR; specifically, enzalutamide has a roughly 4-fold higher affinity for the AR than bicalutamide. Unlike other first-generation AR antagonists such as bicalutamide, enzalutamide did not exhibit detectable agonist behavior under androgen ablation conditions.39 Moreover, enzalutamide retarded AR nuclear localization and inhibited binding of the hormone-bound AR complex to DNA. These properties coupled with successful clinical trials demonstrating the antitumor activity of enzalutamide led to its FDA approval for CRPC and metastatic CRPC (mCRPC).18,19,42

Additional second-generation AR antagonists have been developed. Proxalutamide is one of these newer antagonists that was selected for its roughly 3-fold higher binding affinity for AR than enzalutamide coupled with its ability to downregulate AR protein levels.43,44 A clinical trial is currently ongoing to assess the efficacy of proxalutamide in patients with mCRPC. Apalutamide is another AR antagonist that was developed to better inhibit AR activity. Apalutamide has a near identical backbone structure as enzalutamide and thus has very similar effects. Importantly, apalutamide has lower permeability of the blood–brain barrier compared to enzalutamide.45 This lower blood–brain permeability should reduce some of the central nervous system side effects patients experienced with enzalutamide. In clinical trials, apalutamide improved the survival of patients with both CRPC and mCRPC and received FDA approval for these diseases.46,47 Darolutamide is another FDA-approved second-generation AR antagonist. It has a structure that is distinct from the structures of enzalutamide and apalutamide and only negligibly crosses the blood–brain barrier.48 Darolutamide continues to show antagonism in AR mutants that are traditionally resistant to other first- and second-generation AR antagonists.49

Two meta-analyses comparing the efficacies of enzalutamide, apalutamide, and darolutamide looked at the outcomes in patients exhibiting nonmetastatic castration-resistant prostate cancer (nmCRPC). These meta-analyses concluded that apalutamide and enzalutamide treatment increased metastasis-free survival longer than darolutamide for these patients. Although darolutamide was less effective, it did still show efficacy compared to placebo, and patients receiving darolutamide had less adverse events than patients receiving apalutamide or enzalutamide.50,51 Taken together, these findings indicate that all three drugs may be appropriate for individualized treatment strategies.

Patients undergoing treatment with AR antagonists, particularly second-generation AR antagonists, do have survival benefits compared to patients receiving a placebo. Importantly, after prolonged use of these antagonists, patients inevitably develop resistance to these therapies. The major mechanism of the resistance is due to the reactivation of AR in the CRPC cells, even in the presence of AR antagonists. There are a variety of different mechanisms through which AR develops resistance to LBD-targeting agents. These mechanisms include mutations in the LBD of AR, AR overexpression, and AR splice variants lacking the LBD.52 While different combinations or new formulations of LBD-targeting agents may be able to overcome some of this resistance, AR splice variants without the LBD are not targetable by these antagonists. As such, there has been a concentrated effort in the field of prostate cancer research to target AR in alternative ways.

AR antagonists targeting activation function-2 (AF-2)

The ligand binding pocket is only a portion of the AR ligand binding domain. The binding of agonists and antagonists triggers conformational changes in the AR that change its function. One important region of the AR that undergoes a conformational change is the AF-2. The AF-2 is located in the LBD of AR and has been shown to be an important site for regulating interactions between the amino-terminal/carboxy-terminal (N/C) domains and the binding of important cofactors.53,54,55 During LBD dimerization, the AF-2 motif remains solvent-exposed and accessible, further emphasizing its role in binding to key coregulators.56 N/C interactions and these cofactors have also been implicated in vitro for regulating AR stability and increasing AR DNA-binding affinity.57 Despite these findings, the importance of these N/C interactions for normal AR function is debated. For instance, mice with mutated 23FQNLF27, an important amino acid motif implicated in N/C interactions, showed normal male development.58 Regardless of the debate on the importance of N/C interactions, small molecules have been developed to directly target the AF-2 of AR. An estrogen receptor AF-2 antagonist was modified for specificity to the AF-2 of AR. These modifications led to a class of inhibitors that interact with the AF-2 of AR and can prevent transcriptional activity of AR.59 Since the efficacy of targeting AF-2 was demonstrated, there has been an effort to develop more potent AF-2 inhibitors. A virtual docking screening identified several molecules shown to bind to the AF-2.60 More recent developments have identified the molecules MDG506,61 SPC002,62 LHJ-647,63 IMB-A6,64 and T1-12.65 These compounds show the promise of targeting alternative regions on the LBD of AR.

AR antagonists targeting LBD-dimerization

While DBD dimerization has been long known to play an important role in AR transactivation, recent data suggests that LBD dimerization also regulates AR activity.66 Proper AR dimerization is partially regulated from regions outside both the ligand-binding pocket and the AF-2, opening unique targetable regions on the LBD for antagonists. A new binding site known as the dimerization inhibiting molecules (DIM) pocket was identified as the interface where dimerization occurs. Two molecules known as DIM20 and DIM20.39 were identified based on a virtual screening and shown to antagonize this region and inhibit the proliferation of AR-positive cell lines. Consistent with being outside the ligand-binding pocket, DIM20 and DIM20.39 were shown not to compete with DHT binding and were synergistic with the ligand-binding pocket inhibitor enzalutamide. Interestingly, DIM20 and DIM20.39 showed preferential inhibition on the AR, minor inhibition on the estrogen receptor (ER), and moderate inhibition on the progesterone receptor (PR), suggesting the potential importance of dimerization on other nuclear receptors.67 These compounds show the promise of targeting LBD dimerization.

LBD-targeting AR degraders

Since AR must be localized in the nucleus to function as a transcription factor, a high-throughput screening was developed by Johnston et al.68 to screen for AR antagonists that could cause preferential cytoplasmic localization of AR. From this screening of 219 055 molecules, 3 molecules were identified that cause cytoplasmic localization of AR. Two of these identified small molecules known as CPPI and EPPI were structurally similar and were able to inhibit CRPC cell growth both in culture and in immunodeficient mice.68,69 Further study of these compounds showed that CPPI is a competitive AR antagonist that selectively causes AR polyubiquitination and degradation in the nucleus. The unique mechanism of CPPI as an LBD-targeting AR degrader makes it a promising molecule for continued study.70

Galeterone is a small molecule that utilizes a variety of mechanisms to antagonize AR. These mechanisms include the inhibition of 17α-hydroxylase/C17,20-lyase (CYP17), an important enzyme for androgen biosynthesis, AR antagonism, and the promotion of AR and AR variant degradation. Galeterone is expected to bind in the ligand-binding pocket in the LBD of AR because in LNCaP and VCaP cells, its effect can be outcompeted by DHT, a classic feature of a competitive antagonist.71 Since galeterone was shown to degrade both AR and AR variants, its degradation properties are not mediated purely through direct binding to the LBD. Further evaluation showed that galeterone degrades AR through a proteasomal pathway. Specifically, galeterone was shown to bind to two deubiquitinating enzymes, ubiquitin-specific protease 12 (USP12), and ubiquitin-specific protease 46 (USP46),72 thus promoting degradation. The variety of mechanisms that target the AR made galeterone a promising molecule for the treatment of CRPC. Unfortunately, a phase 3 clinical trial of enzalutamide versus galeterone had to be prematurely ended since the men receiving galeterone had rapid disease progression.73

TRC-253, also known as JNJ-63576253, is a competitive binding inhibitor that was further developed due to its efficacy in the F877L AR mutation that provides resistance to second-generation AR antagonists.74,75 TRC-253 has a high affinity for AR with a half-maximal inhibitory concentration (IC₅₀) of 6.9 nmol l⁻¹ of DHT-binding inhibition. In vitro and in vivo testing in the LNCaP model, both supported the efficacy of TRC-253.76 TRC-253 was tested in a phase 1/2A clinical trial that demonstrated that TRC-253 could be well tolerated (NCT02987829). Future development and testing of this molecule are needed to further evaluate its efficacy in patients.

NTD-TARGETING AR ANTAGONISTS

Constitutively active AR splice variants lacking the LBD are a factor that can drive the resistance to FDA-approved AR antagonists in prostate cancer patients.77,78,79 Although the presence of AR variants has been shown in prostate cancer disease progression, some reports suggest that it is elevated full-length AR expression that is correlated with the presence of AR variants and not AR variant expression itself that is driving disease progression.80,81,82 Despite this ongoing debate, it is thought that NTD-targeting agents will be effective since they can target both full-length AR and AR variants. While the NTD of AR is intrinsically disordered and intrinsically disordered domains were thought to be not druggable targets, several small molecules have been discovered that specifically target the NTD of AR and show efficacy. These molecules are summarized in Supplementary Table 2 and discussed below.

Supplementary Table 2.

N-terminal domain-targeting androgen receptor antagonists/inhibitors

Compound	Structure	Description	IC50	Stage of development	Reference
IMTPPE		AR and AR variant inhibitor	1.8 µmol l−1 (C4-2)	Preclinical	PMID: 27187604
JJ-450		AR and AR variant inhibitor	2.7 µmol l−1 (C4-2)	Preclinical	PMID: 27563404
(+)-JJ-74-138		AR and AR variant inhibitor	1.2 µmol l−1 (C4-2); 4.3 µmol l−1 (22Rv1)	Preclinical	PMID: 35058329
EPI-001		NTD-binding inhibitor and AR variant inhibitor	~6 µmol l−1 (AR-NTD transactivation); 17.1 µmol l−1 (ARv7 activity)	Preclinical	PMID: 20541699 PMID: 34298700
EPI-002		NTD-binding inhibitor and AR variant inhibitor	7.4 µmol l−1 (LNCaP); ~20 µmol l−1 (LN95)	Preclinical	PMID: 23722902 https://www.essapharma.com/wp-content/uploads/IND-Candidate-EPI-7386-is-an-N-terminal-Domain-Androgen-Receptor-Inhibitor-for-the-Treatment-of-Prostate-Cancer.pdf
EPI-7386		NTD-binding inhibitor and AR variant inhibitor	421 nmol l−1 (LNCaP); 3.7 µmol l−1 (LN95)	Phase 1/2 clinical trial	https://www.annalsofoncology.org/article/S0923-7534 (20) 40655-6/fulltext https://www.essapharma.com/wp-content/uploads/IND-Candidate-EPI-7386-is-an-N-terminal-Domain-Androgen-Receptor-Inhibitor-for-the-Treatment-of-Prostate-Cancer.pdf
SINT1		AF-1 binding inhibitor and AR variant inhibitor	10.7 µmol l−1 (LNCaP)	Preclinical	PMID: 18834139 PMID: 27576691
Niphatenone B		AR and AR variant inhibitor	~6 µmol l−1 (LNCaP)	Preclinical	PMID: 25268119
UT-155		AF-1 binding AR inhibitor and AR variant inhibitor	78.4 nmol l−1 (AR transfected HEK-293)	Preclinical	PMID: 28978635
UT-34		AR and AR variant degrader	211.7 nmol l−1 (AR transfected COS7)	Preclinical	PMID: 31481513
Mahanine		AR and AR variant degrader	Not reported	Preclinical	PMID: 24258347
Cinobufagin- 3-acetate		AR and AR variant inhibitor	Not reported	Preclinical	PMID: 30410767
JN018		AR NTD inhibitor and AR variant inhibitor	Not reported	Preclinical	PMID: 30565725
QW07		AR and AR variant inhibitor	4.9 µmol l−1 (LNCaP); 7.5 µmol l−1 (22Rv1)	Preclinical	PMID: 32002708
VPC-220010		AR and AR variant inhibitor	2.7 µmol l−1 (ARv7 activity); 0.7 µmol l−1 (LNCaP)	Preclinical	PMID: 34298700
SC428		AR and AR variant inhibitor	1.4 µmol l−1 (LNCaP); 1.01 µmol l−1 (VCaP); 1.1 µmol l−1 (22Rv1)	Preclinical	PMID: 37139712
Z15		AR AF-1 and LBD inhibitor and AR variant inhibitor	220 nmol l−1 (LNCaP); 1.37 µmol l−1 (VCaP); 3.6 µM (22Rv1)	Preclinical	PMID: 36656639
EIQPN		AR AF-1 inhibitor and AR variant inhibitor	1.1 µmol l−1 (C4-2); 1.5 µmol l−1 (CWR22rv)	Preclinical	PMID: 33415022

AF-1: activation function-1; AR: androgen receptor; IC₅₀: half-maximal inhibitory concentration; LBD: ligand-binding domain; NTD: N-terminal domain

EPI-001

From a screening of marine sponge extracts, EPI-001 was discovered for its ability to bind to the NTD. Specifically, this screening was designed to identify small molecules to inhibit both ligand-dependent and ligand-independent AR activity. Since ligand binds in the LBD and the NTD are important for AR activation, the ability to inhibit both ligand-dependent and ligand-independent AR activity was evidence of inhibition of non-LBD AR regions. EPI-001 was shown to inhibit the activity of AR constructs containing amino acids 1–558 of the NTD of AR.83 This finding was further validated when high-resolution nuclear magnetic resonance spectroscopy showed EPI binding to the transactivation unit 5 in the NTD.84 Through this binding, EPI compounds induce the formation of partially collapsed helical states which give a structural mechanism for EPI inhibition of AR activity.85 Beyond showing binding and inhibition of AR activity in vitro, EPI-001 showed efficacy in slowing tumor growth in mice with LNCaP xenograft tumors.83 The promise of EPI-001 as a novel NTD-targeting agent led to the synthesis of hundreds of structural analogs for further development of this class of compounds. One of these analogs is ralaniten, also called EPI-002, which has an IC₅₀ at 7.4 µmol l⁻¹.86 EPI-506, which is ralaniten acetate, was tested in phase 1 clinical trial, but the trial was terminated early because of poor pharmacokinetic properties, specifically in oral bioavailability.87 Current work is trying to address these problems by developing additional EPI analogs. One of the analogs, EPI-7386 (masofaniten) exhibited an IC₅₀ at 421 nmol l⁻¹,88 which is about 20-fold more potent than EPI-002 (ralaniten), and is currently being tested in a phase 1/2 clinical trial.89

Sintokamides

Sintokamides were identified as small molecules that can inhibit the transactivation of the NTD of AR.90 Sintokamides are a class of small molecules that were isolated from marine sponge Dysidea sp. from Indonesia. The mechanism of NTD of AR inhibition of sintokamide A (SINT1) is through direct binding to the activation function-1 (AF-1) of the AR.91 The AF-1 is a critical region of the NTD of AR that contains two transcriptional activation units, Tau-1 and Tau-5.92 While the exact binding site of SINT1 has not been shown, its synergism with EPI compounds, which bind to Tau-5, suggests that SINT1 binds to a different site on the AF-1. SINT1 has been shown to selectively inhibit AR-positive cell growth, decrease the expression levels of genes regulated by AR and/or AR variants, and cause regression of CRPC tumors.91

IMTPPE

A high-throughput screening of molecules that could cause cytoplasmic localization of AR also identified a molecule known as 2-[(isoxazol-4-ylmethyl)thio]-1-(4-phenylpiperazin-1-yl)ethenone (IMTPPE).68 IMTPPE was shown to inhibit AR and AR splice variant 7 (ARv7) protein level with IMTPPE through dose-dependent degradation AR.93 Further development of structural analogs of IMTPPE led to the compounds JJ-450 and (+)-JJ-74-138. Consistent with the inhibition of non-LBD regions, JJ-450 and (+)-JJ-74-138 both were able to inhibit ARv7-positive enzalutamide-resistant prostate cancer cell growth.94 Further work demonstrated that this class of molecules directly binds to AR and does not compete for DHT binding. This class of molecules can inhibit AR transcriptional activity through several different mechanisms, consisting of noncompetitively binding to AR, preferentially causing AR and ARv7 degradation in the nucleus, and inhibiting liganded-AR binding to androgen response elements (AREs). The IMTPPE class of compounds also has shown efficacy in inhibiting the growth of prostate cancer xenografts that are resistant to second-generation AR antagonists.95 While it has been demonstrated that IMTPPE and analogs bind to a non-LBD region of AR, the exact binding site is not yet known. These compounds are being continually developed to determine the binding site and improve their efficacy.

Niphatenones

Niphatenones are a class of compounds derived from the marine sponge Niphates digitalis in Dominica.96 Niphatenone B is a glycerol ether that was found to bind to the AF-1 of AR in the NTD. Consistent with this binding site, niphatenone B was able to inhibit both full-length AR and AR variants lacking the LBD. This binding is sufficient to inhibit the transactivation of full-length AR with IC₅₀ values of approximately 6 µmol l⁻¹. Niphatenone B mainly shows selectivity for the AR but has been shown to slightly inhibit and bind to the glucocorticoid receptor (GR). Despite the efficacy of niphatenones at inhibiting AR, the reactivity and the lack of specificity for only the AR has made further development of this class of molecules difficult.

UT-155 and UT-34

Following molecular modeling of the LBD of AR, molecules were designed and screened to find novel degraders of AR. This screening discovered the molecules UT-155 and UT-69, which were effectively able to antagonize AR and selectively promote AR proteasomal degradation. These compounds were able to inhibit the binding of radiolabeled DHT to AR, which demonstrated their binding to the LBD of AR. Despite selectively targeting the LBD of AR, these compounds showed some activity toward other steroid receptors at high doses and inhibited the transactivation of the PR at comparable concentrations as AR. Since these compounds were designed to inhibit and bind the LBD of AR, it was unexpected when these compounds were shown to degrade both full-length AR and AR variants that lacked the LBD. Based upon this finding, the binding site of these compounds was further evaluated, and they were shown to also bind to the AF-1 of AR in the NTD. Further experimentation showed that the R enantiomer of UT-155, (R)-UT-155, selectively binds to the AF-1 and does not bind to the LBD.97 To further improve upon the ability of this class of compounds to target the AF-1, a series of similar compounds were tested, and UT-34 was selected as a strong candidate for its low affinity for the LBD and nanomolar inhibition of full-length AR.98 Like its predecessor molecules, UT-34 maintained the degradation mechanism of both full-length AR and AR variants. Importantly, this targeting of the NTD coincided with the inhibition of in vivo tumor growth of enzalutamide-sensitive and enzalutamide-resistant xenografts.99 This class of compounds is currently being further developed and new structurally similar drugs have shown promise in inhibiting aggressive enzalutamide-resistant tumor growth in vivo.100

Mahanine

Mahanine is a molecule that was derived from the curry leaf plant Murraya koenigii. Mahanine was shown to inhibit both ligand-dependent and ligand-independent AR activity. Mahanine was shown to inhibit AR activity after a short treatment time and cause AR and AR variant polyubiquitination when treated for a longer duration.101 While the exact binding of mahanine was not stated, mahanine was shown to inhibit the phosphorylation of Ser81, an important site of post-translational modification that affects AR stability, transcriptional activity, and growth.102 This inhibition of Ser81 phosphorylation is mediated by the inhibition of CDK1, a well-established protein kinase for this site. This inhibition of CDK1 phosphorylation by mahanine was shown not to be through direct inhibition of CDK1.101 Despite this promising inhibition of AR activity, mahanine was also shown to inhibit the growth of AR-negative prostate cancer cells. This indicates that mahanine has mechanisms of action beyond AR inhibition.103

Cinobufagin-3-acetate

Cinobufagin-3-acetate is a natural molecule that was isolated from the skin of the toads Bufo gargarizans and Duttaphrynus melanostictus. It was discovered as a potential signal transducer and activator of transcription 3 (STAT3) signaling inhibitor also capable of inhibiting prostate cancer cell growth in a dose-dependent manner. At sub-micromolar concentrations, cinobufagin-3-acetate was shown to inhibit PSA protein levels in LNCaP cells. This reduction in PSA is likely driven by the ability of cinobufagin-3-acetate to affect AR protein levels without affecting AR mRNA levels. Cinobufagin-3-acetate inhibited both the protein levels of AR and AR variants, which indicated that its binding was not in the ligand-binding pocket. This reduction of AR protein levels occurred in the nucleus in the absence of the androgen R1881 but was not observed in the presence of R1881. Further analysis indicated that cinobufagin-3-acetate directly binds to the NTD of AR which was consistent with other results.104

JN compounds

A class of compounds known as the JN series of compounds has been developed and is currently being explored for inhibition of AR through the NTD. JN018 was the initial compound identified for its inhibition of an AR splice variant in yeast. Based on the structure of JN018, analogs were synthesized that show promise for inhibiting AR activity. The group that developed these compounds has reported that these compounds selectively inhibit the growth of AR-positive cancer cells, lower protein levels of AR and AR variants, and inhibit the growth of enzalutamide-resistant tumor xenografts (https://patents.google.com/patent/WO2018136792A1/en, last accessed on September 23, 2024). The JN series of compounds appears to be under further investigation.105

QW07

A screening for molecules that can inhibit the transcriptional activity of the NTD of AR identified the molecule QW07. As expected, further analysis showed that QW07 was able to also inhibit full-length AR and AR variants. QW07 was found to bind to the AR-NTD with a dissociation constant (K_D) of 1.4 μmol l⁻¹. This was similar to the binding affinity of EPI-001 at 2.0 μmol l⁻¹. Further data showed that QW07 had a greater inhibition of AR transcriptional activity than EPI-001 and that QW07 and EPI-001 do not compete for binding, emphasizing that they bind to different sites. Different than some NTD of AR inhibitors that induce a reduction in AR protein levels, QW07 did not affect AR expression at the protein level. The growth inhibition effect of QW07 also appeared to be regulated through its AR inhibition, as the inhibition of growth was much more significant in AR-positive cell lines than in AR-negative cell lines. In vivo experiments further corroborated the efficacy of QW07, showing a dose-dependent effect in tumor xenografts for both 22Rv1 and VCaP models.106 Several analogs of QW07 were synthesized and one of them appeared to be more potent than QW07.107

VPC-220010

The molecule VPC-2055 was identified in a combination of virtual and experimental screenings of novel AR inhibitors108 and a structural comparison of this compound with EPI-001 showed that these two molecules contained a similar chlorohydrin moiety. Since EPI-001 is a known and established NTD of AR inhibitor, this led to testing of VPC-2055 for NTD inhibition. VPC-2055 showed efficacy in inhibiting ARv7 transcriptional activity, so structural analogs of VPC-2055 were synthesized to improve upon this inhibition. Among these newly synthesized compounds, the compound VPC-220010 was shown to have the highest efficacy at inhibiting ARv7 activity, with an IC₅₀ of 2.7 μmol l⁻¹. In comparison, the IC₅₀ in the same assay for EPI-001 was 17.1 μmol l⁻¹. Like EPI-001, VPC-220010 showed specificity for the AR and did not affect transcription of the GR, ER, or PR. While VPC-220010 shares similar structural components as EPI-001, an AF-1-binding molecule, the specific binding site of VPC-220010 on AR has not been studied. Despite the promising results in inhibiting AR activity, VPC-220010 needs to be further developed before testing in further preclinical studies due to issues with metabolic stability.109

SC428

SC428 is a novel structure that has been shown to inhibit full-length AR as well as different AR splice variants. SC428 exhibits selectivity for the AR and does not affect other nuclear steroid receptors. SC428 was shown to affect the thermostability of both AR and ARv7, suggesting its binding to a non-LBD region. However, the specific binding site of SC428 has not been reported. SC428 was able to inhibit the cell proliferation of both enzalutamide-resistant and enzalutamide-sensitive cells with IC₅₀ values of approximately 1 μmol l⁻¹. SC428 also inhibited the proliferation of AR-negative PC3 cells, indicating that some effects of SC428 are not mediated through the AR. Furthermore, SC428 was able to inhibit in vivo tumor growth of cells with high levels of ARv7 expression.110 Further studies should establish a specific binding site of SC428 to AR and further explore its mechanisms of action.

Z15

Using a rational drug design for AR antagonists, Z15 was recently identified as a bifunctional small molecule acting as both an effective AR antagonist and a selective AR degrader.111 Z15 inhibited DHT-induced transcriptional activities of both exogenous and endogenous AR in several different cell lines, including LNCaP, VCaP, and 22Rv1. Z15 appears to be highly potent, with an IC₅₀ at approximately 0.22 μmol l⁻¹ in LNCaP cells. Z15 also appears to be specific to AR, because its IC₅₀ is over 20 μmol l⁻¹ for GR and ER and 9.3 μmol l⁻¹ for PR. Z15 exhibited direct binding to the LBD of AR, with an IC₅₀ value of 63.3 nmol l⁻¹ but was also shown to directly bind to the AF-1 of AR, with K_D at 0.93 μmol l⁻¹. However, it is not clear if Z15 could bind to AF-1 and LBD of AR simultaneously. Consistent with its targeting of AF-1, Z15 was also shown to induce proteasome-dependent degradation of both full-length AR and ARv7 in prostate cancer cells. Z15 potently inhibited proliferation in AR-positive VCaP and 22Rv1 cells, but only weakly affected the proliferation in AR-negative PC3 and DU145 cells. Importantly, Z15 suppressed the growth of 22Rv1 xenografts in mice.

EIQPN

In structure–activity relationship studies to develop AR antagonists,112,113 6-[6-ethoxy-5-ispropoxy-3,4-dihydroisoquinolin-2[1H)-yl]-N-[6-methylpyridin-2-yl]nicotinamide (EIQPN) was identified as a novel AR antagonist that did not bind to LBD. Further analysis showed that EIQPN can inhibit both AR activation induced by androgen or nonandrogen forskolin (FSK) and interleukin-6 (IL-6).114 EIQPN inhibited AR activation at multiple steps, including the DHT-induced nuclear import of AR, N/C interaction of AR, AR recruitment to AREs, and co-activator recruitment to AR. EIQPN was shown to interact with the AR AF1. EIQPN markedly decreased the protein levels of endogenous ARs in several AR-positive prostate cancer cell lines, including LNCaP and CWR22Rv, through the ubiquitin-proteasome pathway. As expected, EIQPN inhibited the proliferation of AR-positive prostate cancer cells such as LNCaP, C4-2, and CWR22Rv, with IC₅₀ around 1 µmol l⁻¹. It did not affect the proliferation of AR-negative DU145 cells, suggesting the specificity of EIQPN to AR. EIQPN also effectively inhibited 22Rv1 xenograft tumor growth in mice.

DBD-TARGETING AR ANTAGONISTS/INHIBITORS

In addition to NTD and LBD, DBD can also be targeted by small molecules. The availability of AR-DBD crystal structure allows for an in silico drug design approach. Similar to the NTD-targeting small molecules, DBD-targeting compounds should be able to inhibit both full-length AR and AR splice variants. Several different structures have been reported to target DBD and are summarized below and in Supplementary Table 3.

Supplementary Table 3.

DNA-binding domain-targeting androgen receptor antagonists/inhibitors

Compound	Structure	Description	IC50	Stage of development	Reference
Pyrvinium pamoate		Noncompetitive inhibitor and AR variant inhibitor	8–30 nmol l−1 (22Rv1, LNCaP, C4-2, and LAPC4)	Preclinical	PMID: 34117723 PMID: 24354286
VPC-14449		Selective inhibitor and AR variant inhibitor	Sub-micromolar (LNCaP, C4-2, and MR49F) low-micromolar (22Rv1)	Preclinical	PMID: 28775145
VPC-17005		AR dimerization inhibitor and AR variant inhibitor	1.5 µmol l−1 (LNCaP); 10 µmol l−1 (ARv7)	Preclinical	PMID: 30165195 PMID: 33801338
MF-15		Dual inhibitor and AR variant inhibitor	Not reported	Preclinical	PMID: 32731472

AR: androgen receptor, IC₅₀: half-maximal inhibitory concentration

Pyrvinium pamoate

Pyrvinium pamoate was reported as a noncompetitive AR inhibitor both in vitro and in vivo.115 Pyrvinium inhibited AR target gene expression both in cultured prostate cancer cells and in the mouse prostate. Pyrvinium pamoate was shown to directly bind to AR.116 Mutagenesis analysis revealed that pyrvinium pamoate binds to the DNA-binding domain of AR. Computational modeling suggests that pyrvinium binds at the interface of the DBD dimers and the AR response element. In fact, pyrvinium is the first reported small molecule inhibitor that targets the DBD of AR. It is also one of the first small molecules shown to directly inhibit the activity of AR splice variants (ARvs), including ARv7. As expected, pyrvinium inhibited the growth of ARv7-positive 22Rv1 CRPC tumors in mice. Although pyrvinium interacts with the AR/DNA complex, pyrvinium was not shown to influence the DNA-binding kinetics of AR.117 However, pyrvinium binding altered AR interactions with its cofactors such as DEAD-box helicase 17 (DDX17). This suggests that pyrvinium inhibits AR activity through interfering with AR interactions with its cofactors. However, pyrvinium was shown to also inhibit several other hormone nuclear receptors,116 suggesting that its non-AR targeting activities will need to be considered when pyrvinium is used to inhibit AR in prostate cancer.

VPC-14449

Using an in silico approach, VPC-14449 (4-(4-(4,5-bromo-1H-imidazol-1-yl)thiazol-2-yl)morpholine) was identified as a small molecule inhibitor targeting a surface exposed pocket on the DBD of AR.118 VPC-14449 was shown to inhibit both full-length AR and ARv7 splice variants at sub-micromolar concentrations. The inhibition was abrogated by mutations at the amino acid residues involved in interactions with VPC-14449, further supporting specific interactions of VPC-14449 with the DBD of AR. VPC-14449 also appears to be specific for AR, because it exhibited little or no cross-reactivity with other steroid receptors (ER, GR, and PR) in luciferase assays.118 VPC-14449 does not appear to affect androgen-induced AR nuclear translocation. Instead, VPC-14449 interferes with AR binding to chromatin, consistent with VPC-14449 targeting the AR-DNA interface, P-Box of the AR-DBD.119 In castrated mice, VPC-14449 inhibited the growth and PSA production in relapsed LNCaP xenograft tumors.

VPC-17005

In addition to the identification of small molecules such as VPC-14449 that target the P-Box of AR-DBD, in silico screen was also performed to identify small molecules targeting a D-Box pocket between amino acid residues L594-S613 at the dimerization interface of the DBD domain in AR. This screen has led to the identification of VPC-17005, which is capable of inhibiting AR at sub-micromolar level.119 Further analysis showed that VPC-17005 also inhibited ARv7 activity. VPC-17005 inhibition appears to be specific to AR because VPC-17005 did not inhibit the transcriptional activity of other steroid hormone receptors, ER, GR, and PR. Specificity for AR was shown when VPC-17005 inhibited the growth of AR-positive LNCaP, 22Rv1, and MR49F cells, but not AR-negative PC3 cells.119 Despite promising in vitro efficacy, VPC-17005 exhibited poor metabolic stability and is not suitable for in vivo studies. Thus, Radaeva et al.120 tested more chemotypes using a combination of docking, pharmacophore modeling, and biological assays. This led to the discovery of two lead compounds, VPC-17160 and VPC-17281, which showed stronger inhibition of ARv7 transcriptional activity and superior activity in mammalian two-hybrid assays compared to VPC-17005. VPC-17281 also appeared to have significantly improved microsomal stability, suggesting that it is an excellent candidate for further hit-to-lead optimization.

AR TARGETING PROTEOLYSIS TARGETING CHIMERAS (PROTACS)

PROTACs are heterobivalent molecules, with one part targeting a protein specifically and the other part binding to an E3 ubiquitin ligase, that can degrade the target protein specifically within cells.121 PROTACs function by bringing the target protein and an E3 ubiquitin ligase into close proximity and facilitating the ubiquitination of the target protein, which then activates the ubiquitin-proteasome system (UPS). Once ubiquitinated, the target protein undergoes degradation by the proteasome. The PROTACs system offers a groundbreaking strategy for drug discovery and development. A salient feature of the PROTACs approach is its ability to utilize the endogenous cellular quality control systems to degrade target proteins. This results in high target selectivity and low toxicity as well as decreased systemic drug exposure.122 PROTACs pave the way for targeting molecules previously considered undruggable.

PROTACs consist of a targeting ligand, a linker region, and an E3 ligase ligand, which together facilitate the selective degradation of target proteins. Despite their clever design, the inclusion of three components typically results in high molecular weights, which makes the molecules large and complex. This complexity can lead to difficult synthesis and component optimization, as each component must be designed and optimized for effective interaction with the target protein.123 Furthermore, this large size and structure of PROTACs violates Lipinski’s rule of 5, a general rule for orally bioavailable drugs.124,125 This potentially limits the route of administration of PROTACs to alternative delivery mechanisms such as intravenous injection. Despite these challenges, advances in drug delivery systems are actively addressing this bioavailability problem.126,127 Beyond oral availability, PROTACs can sometimes suffer from poor tissue adsorption and short metabolic stability.128 These problems are also being actively investigated. Fortunately, PROTACs can take advantage of targeting ligands with high affinity to the target protein. This allows PROTACs to have efficacy at low concentrations and evade some of the bioavailability and metabolism problems.

One significant advantage of PROTACs is their flexibility in binding sites. Unlike traditional inhibitors, PROTACs do not need to target the active site of a protein; they can, in theory, bind to any region of the protein that can be bound with high affinity, provided the site is accessible to the recruited E3 ligase. This flexibility allows for targeting proteins that are otherwise difficult to inhibit. However, PROTACs may be less effective if the binding sites are not well exposed.129 In addition, the efficacy of PROTACs depends on the availability and expression of the E3 ligase in the targeted tissue, though the choice of E3 ligase ligand can be optimized for specific applications.

Since the first PROTAC was reported in 2001,121 various PROTACs targeting a range of proteins have been developed. PROTACs targeting AR were first reported in 2003. Currently, there are over 200 PROTAC candidate drugs targeting AR.130,131 This section is aimed at providing a brief overview of some PROTACs that target AR. These PROTACs of AR are summarized in Supplementary Table 4. Since the effects of PROTACs currently in clinical trials on CRPC patients have only been partially published, the major weaknesses of individual PROTACs are not yet fully understood.

Supplementary Table 4.

Proteolysis targeting chimeras targeting androgen receptor

Compound	Structure	DC50	E3 ligase	Stage of development	Reference
ARV-110		<1 nmol l−1	CRBN	Phase 2	https://doi.org/10.1158/1538-7445.AM2018-5236 PMID: 36770875
ARV-766		<1 nmol l−1	CRBN	Phase 2	PMID: 36508999 https://doi.org/10.1158/1538-7445.AM2023-ND03 https://www.medchemexpress.com/arv-766.html
CC-94676	Undisclosed	Undisclosed	CRBN	Phase 1	https://www.mskcc.org/cancer-care/clinical-trials/20-479 https://aacrjournals.org/cancerres/article/84/7_ Supplement/ND02/742763/Abstract-ND02- Discovery-of-BMS-986365-a-ligand
HP518	Undisclosed	Undisclosed	CRBN	Phase 1	PMID: 36508999 https://www.clinicaltrialsarena.com/news/ hinova-doses-first-patient-in-phase-i- castration-resistant-prostate-cancer-trial/
AC0176	Undisclosed	Undisclosed	CRBN	Phase 1	PMID: 36508999 https://www.accutarbio.com/accutar-biotechnology- announces-first-patient-dosed-with-ac0176-in-phase-1-study-in- patients-with-metastatic-castration-resistant-prostate-cancer-mcrpc/
MTX-23		AR-FL: 2 µmol l−1 ARv7: 370 nmol l−1	VHL	Preclinical	PMID: 33277442
Au-AR pep-PROTAC	See reference	AR: 48.8 nmol l−1 ARv7: 79.2 nmol l−1	MDM2	Preclinical	PMID: 35971165

AR: androgen receptor; AR-FL: full-length AR; ARv7: AR splice variant 7; CRBN: cereblon; DC₅₀: half-maximal degradation concentration; MDM2: mouse double minute 2 homolog; PROTAC: proteolysis targeting chimeras; VHL: von Hippel-Lindau

Clinical trials are currently in progress on five PROTACs targeting the AR: ARV-110 (NCT03888612), ARV-766 (NCT05067140), CC-94676 (NCT04428788), HP518 (NCT05252364), and AC0176 (NCT05241613). Both ARV-110 and ARV-766 have progressed to phase 2 clinical trials, and their drug characteristics in CRPC patients are becoming increasingly clear.

ARV-110

ARV-110 was a first-in-class targeted protein degrader that moved into clinical trials. Preclinical studies revealed that ARV-110 has potent degradation against AR as shown by a half-maximal degradation concentration (DC₅₀) of less than 1 nmol l⁻¹, but it does not suppress the growth of prostate cancer cells with specific AR mutations/alterations, such as ARv7 and AR L702H,132 indicating that ARV-110 is dependent on binding to LBD of AR. However, results from the phase 1 clinical trial showed that among CRPC patients treated with ARV-110, those with T878A/S and H875Y mutations still experienced PSA decline and tumor regression. T878A/S LBD mutations are commonly associated with resistance to AR antagonists. Despite its resistance to some LBD mutations of AR, the L702H mutation considerably reduced the efficacy of ARV-110.133 The ongoing phase 2 clinical trial is assessing the effect of ARV-110 by subgrouping patients based on AR mutations/alterations such as T878, H875, L702H, and ARv7.132 Regarding the safety profile, no adverse events beyond grade 4 treatment-related events were reported in the phase 2 dose. As this clinical trial is still ongoing, further detailed reports on the characteristics of ARV-110 are awaited.

ARV-766

The clinical trial for ARV-766 began in September 2021 (NCT05067140). Even though ARV-766 bears structural similarities to ARV-110, it has shown effectiveness against not only AR H878 and H875 but also the L702 mutation, which was not responsive to ARV-110.134 A progress report from the phase 1/2 dose escalation and expansion trial revealed that 42% of patients with LBD mutations of AR achieved a PSA reduction of over 50% (PSA₅₀). Moreover, all three patients who had co-occurring mutations of T878, H875, or L702 reached PSA₅₀.135 These results suggested that ARV-766 may be also effective for patients resistant to ARV-110.

CC-94676, HP518, and AC0176

CC-94676, HP518, and AC0176 are PROTACs currently in phase 1 clinical trials (summarized in Supplementary Table 4). Details about their progress have not yet been extensively reported. However, the scientific community keenly anticipates forthcoming updates on these compounds. Since AR remains the primary target for CRPC, a growing number of PROTAC-candidate drugs are expected to emerge in the future.

Since ARv7 lacks the LBD region entirely, AR signaling is activated without ligand interaction.136 While ARV-110 has been shown to be incapable of degrading ARv7 in preclinical studies,133 the effect of PROTACs in other clinical trials on ARv7 has not yet been published (NCT04428788).133,137,138 This section will highlight some preclinical studies on candidate PROTACs of both full-length AR and ARv7, which may become clinically significant in the future.

MTX-23

MTX-23 is a recently reported PROTAC molecule.139 A preclinical study indicates its capability to degrade both the full-length AR and ARv7. Unlike the molecular structure of PROTACs currently in clinical trials, MTX-23 targets the DBD of AR and also binds to the Von Hippel-Lindau (VHL) E3 ligase. Lee et al.139 reported that MTX-23 inhibited the growth of 10 human prostate cancer cell lines resistant to second-generation oral AR inhibitors and had no effect on two AR-negative cell lines. In addition, MTX-23 was found to degrade ARv7 more efficiently than full-length AR. Most PROTACs currently in clinical trials utilize E3 ligase CRBN, with few employing the E3 ligase VHL having advanced to clinical trials. One of the primary challenges with VHL-PROTACs entering clinical trials is their low oral bioavailability.124 Despite these challenges, MTX-23 has exhibited the first reported anti-tumor effect against ARv7-positive tumor growth in preclinical studies of PROTACs, and its further development may provide medical benefits to CRPC patients in the future.

Au-AR pep-PROTAC

Au-AR pep-PROTAC, first reported in 2022, is an advanced peptide PROTAC targeting both full-length AR and ARv7.140 It is well-documented that peptide-based drugs grapple with challenges such as a short half-life and low cell permeability.141 However, Au-AR pep-PROTAC, designed through artificial intelligence (AI) methodology, exhibits a pronounced affinity for the DBD of AR. In addition, its delivery efficacy and stability have been notably augmented through the synthesis of ultrasmall Au nanoparticles integrated into the AR pep-PROTAC structure. Ma et al.140 reported that Au-AR pep-PROTAC was particularly effective against the prostate cancer cell line, CWR22Rv1, which is an androgen-independent human prostate cancer cell line derived from the castration-resistant/recurrent CWR22R tumors.142 Au-AR pep-PROTAC’s half-maximal degradation concentration for ARv7 is five times lower than that of MTX-23, making it potentially more effective in curbing ARv7-driven prostate cancer progression.

PROTACs represent a potentially revolutionary therapeutic approach, aiming to degrade target proteins through ubiquitination. They hold promise in surmounting challenges tied to conventional AR inhibitors. Several PROTACs targeting the AR are now in clinical trials, with several novel candidates expressly targeting ARv7 under development. It is expected that CRPC patients will increasingly benefit from the entry of PROTACs in CRPC treatment.

OTHER AR-TARGETING AGENTS

Direct targeting of the AR is not the only way to inhibit AR activity. Mechanisms regulating AR expression and/or activity can be potentially targeted as well. For example, a productive AR transcriptional complex requires cofactors that give functionality and stability to the AR protein. As such, a variety of different novel approaches have shown efficacy in inhibiting AR activity. Although direct targeting of AR appears to be preferable because factors regulating AR could also play important roles in other pathways, some innovative ways to target mechanisms regulating AR have shown preclinical efficacy and should be further explored. These other AR targeting agents are summarized in Supplementary Table 5.

Supplementary Table 5.

Other androgen receptor targeting agents

Compound	Structure	Description	IC50	Stage of development	Reference
A4B17		Interrupts AR and BAG1L interaction and shows efficacy against ARv7	2.8 µmol l−1 (LNCaP)	Preclinical	PMID: 35479411 PMID: 34506104
Niclosamide		Induces ARv7 proteasomal degradation	800 nmol l−1 (LNCaP); 1.6 µmol l−1 (22Rv1)	Phase 1b	PMID: 24740322 PMID: 35772635 PMID: 33737681
ASC-J9		Induces AR and AR variant degradation through ARA55 and ARA70	9.5 µmol l−1 (LNCaP); 10.5 µmol l−1 (C4-2); 9.0 µmol l−1 (22Rv1)	Preclinical	PMID: 23219429 PMID: 29203251
TAS3681		AR and AR variant inhibitor	60.9 nmol l−1 (DHT induced VCaP); 52.7 nmol l−1 (COS-7/wt-AR)	Phase 1	https://ascopubs.org/doi/abs/10.1200/JCO.2021.39.15_suppl.5031 PMID: 38600681
Pyrrole-imidazole polyamides		DNA binder	>30 µmol l−1 (VCaP); 7±3 µmol l−1 (LNCaP)	Preclinical	PMID: 26571387

AR: androgen receptor, ARA55: AR-associated protein 55, ARA70: AR-associated protein 70, ARv7: AR splice variant 7, BAG1L: BCL-2-associated athanogene-1L, DHT: dihydrotestosterone, IC₅₀: half-maximal inhibitory concentration

A4B17

One important co-chaperone that has been implicated as a key factor required for the proper function of AF-1 of AR and, therefore, AR transactivation is BCL-2-associated athanogene-1L (BAG1L) binds to two regions on the AR; the AF-1 in the NTD of AR and the binding function-3 (BF-3) in the LBD of AR. Targeting the NTD-binding site of BAG1L has emerged as a potential therapeutic approach for inhibiting AR.143,144 A4B17 is a molecule that was discovered for its ability to inhibit AR activity by disrupting the AR-BAG1L interaction. Importantly, A4B17 was not able to affect BAG1L binding to other proteins, suggesting that the effect of A4B17 is selective for the AR.145 Additional experiments including a prostate cancer LAPC4 xenograft study demonstrated that A4B17 was effective at slowing tumor growth in vivo.146 Taken together, these results indicate that targeting the binding sites of AR co-chaperones could be a potential way to indirectly target regions of the intrinsically disordered NTD of AR.

Niclosamide

Niclosamide is an FDA-approved drug that is currently used to treat human tapeworm infection. Niclosamide has also been shown to decrease ARv7 protein level, but not mRNA level. This effect on ARv7 occurs due to increased proteasomal degradation. Although the full-length AR mostly contains the amino acid sequence of ARv7, the degradation was specific to ARv7.147 Since the effect of niclosamide was mediated through the AR variants, the most pronounced efficacy of niclosamide was in combination with the second-generation AR antagonist enzalutamide. A phase 1 trial was conducted to further explore this synergism with enzalutamide but was determined to not be viable due to poor oral bioavailability.148 A new formulation was developed such that niclosamide can reach a therapeutic concentration,149 which will allow future clinical trials to test the efficacy of niclosamide in the treatment of CRPC.

ASC-J9

ASC-J9 is a curcumin analog that has been developed and shown to increase AR degradation.150 ASC-J9 has been shown to induce AR degradation by interrupting the interaction of AR with the cofactors AR-associated protein 55 (ARA55) and AR-associated protein 70 (ARA70), which then enhanced the association of AR with the E3 ligase mouse double minute 2 (MDM2).151 ARA55 has been specifically implicated as a cofactor that may play an important role in regulating AR transactivation in prostate cancer tumor cells, as its expression level is different in normal prostate compared to prostate cancer.152 Further testing of ASC-J9 has suggested preferential inhibition of AR-positive prostate cancer cells compared to AR-negative cells.153

TAS3681

While the binding site and mechanisms of TAS3681 are not fully understood, TAS3681 was shown to decrease the levels of AR protein levels and AR transactivation. TAS3681 was able to reduce levels of AR and ARv7, indicating potential efficacy against androgen-independent AR activation.154 Moreover, TAS3681 blocked AR transcriptional activity in mutated AR that confers resistance to both enzalutamide and darolutamide. Building upon this data, TAS3681 was able to suppress tumor growth in enzalutamide-resistant xenografts. TAS3681 also shows remarkably low IC₅₀ values that are comparable to enzalutamide, suggesting a strong affinity for AR.155 An in-human trial of TAS3681 showed promise, with patients seeing a dose-dependent decrease in PSA. These findings have prompted further investigation of TAS3681, particularly in patients with progressive disease. TAS3681 is currently being tested in a phase 1 clinical trial for use in patients with mCRPC (NCT02566772).

Pyrrole-imidazole polyamides (Py-Im polyamides)

Py-Im polyamides are modular sequence-specific DNA-binding molecules that can recognize DNA with the same affinity as DNA-binding proteins.156 Polyamides have been shown to target AREs in CRPC. In VCaP cells, Py-Im polyamides can target AREs and significantly downregulate AR-driven gene expression, thereby reducing the proliferation of VCaP cells and inhibiting the growth of the VCaP xenograft tumors in mice.157 Py-Im polyamides caused a 6% reduction in body weight, which was recovered within 5 days without visible signs of distress, indicating the low toxicity of Py-Im polyamides.158 Chromatin immunoprecipitation (ChIP) analysis confirmed sequence-specific suppression of AR binding to DNA in vivo by a Py-Im polyamide.159 These findings suggest the potential to further develop Py-Im polyamides as therapeutics to suppress AR function in prostate cancer.

FUTURE PERSPECTIVES

As prostate cancer progresses, it often develops resistance to existing therapies. Despite this resistance, the cancer remains dependent on the AR in most cases. Due to this resistance, significant efforts have been made to target the AR in innovative ways, different than the second-generation AR antagonists. These efforts have led to the development of several novel strategies, such as targeting specific regions of the AR, inhibiting its cofactors and post-translational modifications, and promoting selective degradation of the AR. Exploring different approaches opens up the potential for synergistic effects between different classes of molecules. The transition from first-generation to second-generation AR antagonists illustrates how iterative advancements can significantly benefit patients and has inspired researchers today to continue taking that same approach to improve their AR-targeting molecules. In this era where predictive biology is rapidly improving, we can anticipate the discovery of novel AR inhibitors and enhancements to existing treatments at an accelerated pace. Future research should continue to develop AR antagonists that overcome the shortcomings of the currently approved therapies. It is hopeful that the ongoing research and development of AR inhibitors will ultimately be effective, either alone or in combination with other therapeutic agents, for the treatment of patients with CRPC that are resistant to the current AR-targeting agents.

AUTHOR CONTRIBUTIONS

ZW organized the structure of the manuscript. RNC, QF, KM, and ZW wrote the manuscript. RNC and KM created the tables and QF made the figure. All authors read, edited, and approved the final manuscript.

COMPETING INTERESTS

ZW is listed as a co-inventor on a US patent held by the University of Pittsburgh (US 10004730, “Small molecules targeting androgen receptor nuclear localization and/or level in prostate cancer”). RNC, QF, and KM declare no competing interests.