New Generation of Selective Androgen Receptor Degraders: Our Initial Design, Synthesis, and Biological Evaluation of New Compounds with Enzalutamide-Resistant Prostate Cancer Activity

Article abstract of DOI:10.1021/acs.jmedchem.8b00973

In our effort to find small-molecule treatments of advanced prostate cancers (PCs), a novel series of indolyl and indolinyl propanamides (series II and III) were discovered as selective androgen receptor degraders (SARDs). Initial studies of androgen receptor (AR) antagonist (1) and agonist (2) propanamides yielded a tertiary aniline (3) with novel SARD activity but poor metabolic stability. Cyclization to II and III produced submicromolar AR antagonism and protein degradation selective to AR and AR splice variant (AR SV). II and III maintained potency against enzalutamide-resistant (Enz-R) mutant ARs and PC cells and were efficacious in Enz-R xenografts, suggesting their potential to treat advanced PCs. Design, synthesis, and biological activity of novel SARDs that could potentially be used for the treatment of a wide spectrum of PCs including castration-resistant, Enz-R, and/or AR SV-dependent advanced PCs that are often untreatable with known hormone therapies are discussed.

Full text of DOI:10.1021/acs.jmedchem.8b00973

Subscriber access provided by YORK UNIV
Article
A New Generation of Selective Androgen Receptor Degraders:
Our Initial Design, Synthesis, and Biological Evaluation of New
Compounds with Enzalutamide-Resistant Prostate Cancer Activity
Dong-Jin Hwang, Yali He, Suriyan Ponnusamy, Michael L Mohler, Thirumagal
Thiyagarajan, Iain J. McEwan, Ramesh Narayanan, and Duane D Miller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00973 • Publication Date (Web): 07 Dec 2018
Downloaded from http://pubs.acs.org on December 8, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A New Generation of Selective Androgen Receptor Degraders: Our Initial Design,
Synthesis, and Biological Evaluation of New Compounds with Enzalutamide-Resistant
Prostate Cancer Activity
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
†
‡
†,#
Dong-Jin Hwang, Yali He, Suriyan Ponnusamy, Michael L. Mohler, Thirumagal
‡

‡
*,†
Thiyagarajan, Iain J. McEwan, Ramesh Narayanan, Duane D. Miller
†
. Department of Pharmaceutical Sciences, University of Tennessee Health Science Center,
Memphis, TN 38163.
‡. Department of Medicine, University of Tennessee Health Science Center, Memphis, TN
3
#
8103.
. GTx, Inc., Memphis, TN 38103.
&
. School of Medicine, Medical Sciences and Nutrition, Institute of Medical Sciences,
University of Aberdeen, Aberdeen AB25 2ZD, Scotland, UK
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT: In our effort to find small molecule treatments of advanced prostate cancers (PCs),
the novel series of indolyl and indolinyl propanamides (series II and III) were discovered as
selective androgen receptor degraders (SARDs). Initial studies of androgen receptor (AR)
antagonist (1) and agonist (2) propanamides yielded a tertiary aniline (3) with novel SARD activity
but poor metabolic stability. Cyclization to II and III produced sub-micromolar AR antagonism
and protein degradation selective to AR and AR splice variant (AR SV). II and III maintained
potency against enzalutamide-resistant (Enz-R) mutant ARs and PC cells, and were efficacious in
Enz-R xenografts, suggesting their potential to treat advanced PCs. Disclosed is the design,
synthesis, and biological activity of novel SARDs that could potentially be used for the treatment
of a wide spectrum of PCs including castration resistant, Enz-R, and/or AR SV dependent
advanced PCs that are often untreatable with known hormone therapies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KEYWORDS: Structure-activity relationship, Selective androgen receptor degrader, Androgen
receptor, Androgen receptor splice variant, AR-V7, AR escape mutants, Antagonist, Prostate
cancer, Antiandrogen resistance, Enzalutamide resistance, Castration-resistant prostate cancer,
N-terminal domain, Ligand binding domain, Prostate-specific antigen, AR activation function
domain-1.
1. INTRODUCTION
The androgen receptor (AR) has been the focus of prostate cancer (PC) therapies since
Huggins and Hodges’s discovery in the early 1940s that androgens promote PC growth.1-3 PC is
one of the most common malignancies diagnosed in men. According to the American Cancer
Society in 2017, 161,000 men were diagnosed with PC in the United States with 26,730 deaths
2
ACS Paragon Plus Environment

Page 3 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
estimated from the disease. More than 3 million men in the United States are currently living
5
with PC. The AR plays an important role not only for the development and function of the
normal prostate gland, but also for the growth and maintenance of PC cells.6-7 Androgen
deprivation therapy (ADT), via suppression of endogenous synthesis (e.g., goserelin or
abiraterone) and/or AR blockade (e.g., bicalutamide (4), enzalutamide (5), and apalutamide (6)
in Figure 1), is the standard of care for metastatic PC. Re-activation of the AR-axis despite
treatment can occur even despite castration levels of testosterone or 5-dihydrotestosterone
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
DHT), suggesting the need for new mechanisms for AR antagonism.8-10
An AR antagonist 5 improved survival of men with metastatic castration-resistant prostate
11
cancer (CRPC). Earlier this year, the AR antagonist 6 was approved for use in non-metastatic
CRPC based on its ability to extend the metastasis-free survival as compared to placebo.12-13
Unfortunately, primary and acquired resistance to 5 (and other antiandrogens) is common,14-15
e.g., the mutations AR F876L or F876L-T877A switch 5, at higher levels, from AR antagonist to
14
agonist making PCs enzalutamide-resistant (Enz-R). Further, cross-resistance between 5 and 6
1
6
based of F876L and cross-resistance in general between abiraterone and/or ligand binding
domain (LBD)–directed antiandrogens is well known.17-19 Moreover, AR splice variants (e.g.,
19
AR-V7 and D567es) lacking the LBD have been reported as pan-resistant CRPCs. The
development of resistance to antiandrogens is a growing concern and new strategies to block AR
function in CRPC are required.17-19
Our first generation of selective androgen receptor degraders (SARDs) were metabolically
labile secondary and tertiary amines (I) lacking in vivo activity when administered orally that
were designed by structural modification of the AR antagonist 1 [2-hydroxy-4-(4-
isothiocyanatophenyl)-2-methyl-N-(4-nitro-3-(trifluoromethyl)phenyl)butanamide] and the
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
tissue-selective AR agonist enobosarm (2, (S)-N-(4-cyano-3-(trifluoromethyl)phenyl)-3-(4-
cyanophenoxy)-2-hydroxy-2-methylpropanamide) (Figure 2). Class I was exemplified by UT-69
(3, (S)-N-(4-cyano-3-(trifluoromethyl)phenyl)-3-((6-cyano-[1,1'-biphenyl]-3-yl)(methyl)amino)-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
-hydroxy-2-methylpropanamide) whose tertiary amine was cyclized to form indoles (II) and
indolines (III), which are characterized herein as potent AR antagonists and SARDs with a broad
activity profile in models of prostate cancer, and in vivo AR antagonism when orally
administered. E.g., SARDs of II and III exhibited strong AR antagonistic activity in vitro in
transcriptional activation and cellular proliferative assays, including in models of enzalutamide-
sensitive and Enz-R PCs, and CRPCs. Additionally, II and III showed selective AR degradation
of full-length (FL; e.g., from LNCaP cells (T877A)) and splice variant (SV; e.g., from 22RV1
cells (AR-V7)) isoforms of AR, all at sub to low micromolar treatment levels, and in a variety of
PC cell contexts, including Enz-R PCs (e.g., MR49F cells).
The ability to degrade SV AR in this study suggested the potential of II and III to treat
various currently untreatable advanced and refractory PCs. E.g., those lacking the ligand binding
domain (LBD) of AR such as AR-V7 and D567es AR truncations, which are not susceptible to
ADT, abiraterone, or LBD-directed antiandrogens (e.g., 4-6 (Figure 1)), and are associated with
short survival.20-21 Further, in vivo investigations found that analogs within the II and III series
overcome a variety of escape mutants including F876L and F876L/T877A that are known to
emerge due to treatment with 5. These mutations convert 5 and 6 to agonists, conferring
2
2
resistance to prostate cancer cells and tumors via an agonist switch as seen with other LBD-
binding antiandrogens, e.g. W741L for 4 and T877A for flutamide (N-(4-nitro-3-
(
trifluoromethyl)phenyl)isobutyramide).23-24
4
ACS Paragon Plus Environment

Page 5 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The intractability of truncation mutants and the frequency of the agonist switch mutations
suggest that novel ways, potentially LBD-independent ways, of targeting the AR are needed.
The initial design, synthesis and biological evaluation of these SARDs as putative treatments of
Enz-R and other advanced PCs is discussed. Moreover, these SARDs are dual acting agents, i.e.,
potent inhibitors and degraders of AR, providing a higher evolutionary barrier to the
development of resistance to II and III. For all these reasons, we believe that SARDs may
provide a next generation of AR antagonists to treat a variety of refractory and/or advanced PCs,
including Enz-R and AR-V7 dependent PCs.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Herein we rationalize our profile of screening assays designed to overcome Enz-R PC
including our initial SARD structure activity relationship (SAR). Further, II-III were evaluated
in in vitro anti-proliferation and in vivo xenograft models of Enz-R PC. In overview, we report
three series of SARDs (I-III) possessing a novel mechanism of action in SV AR and escape
mutant models of Enz-R PC in vitro and in vivo assays. Despite a variety of competing
preclinical approaches to address Enz-R, there has been little success in the clinic, leaving this
and other refractory prostate cancers as unmet clinical needs. It is our belief that the SARDs
such as reported herein may resurrect the AR axis as a target for overcoming Enz-R, AR-V7
dependent, and/or refractory CRPCs.
2
. RESULTS
Design. Since even before our discovery of diaryl nonsteroidal androgens in 1998,25-27
we have designed and developed many variations of AR agonist and AR antagonist
propanamides with a recent focus on selective androgen receptor degraders (SARDs or AR
degraders; where selectivity refers to degrading only the AR protein) such as 3 and 7 (Figures 1
and 2).28 SAR studies of the propanamides demonstrated that the linkage to the B-ring plays a
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
key role in determining the agonistic vs. antagonistic activities (Figure 2), which can be fine-
tuned via B-ring substitution.25-27, 29-40 Many ethers (and thioethers in vitro) are agonists whereas
most other linkages are partial to full antagonist. E.g., enobosarm (ostarine, MK-2866, GTx-024,
S-22, 2 in Figure 1) showed strong in vivo36, 40-42 tissue-selective androgen receptor modulation
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(SARM) activity. In other words anabolic tissue-selectivity is revealed by increased lean body
mass reflective of muscle mass changes, increased bone mineral density and performance
enhancement in humans across at least 22 clinical trials.43,44
We also have reported high affinity p-NCS B-ring substituted propanamides with a
variety of linkages to the B-ring such as methylene (e.g., 1 which is a butanamide), secondary
2
7
and tertiary amine, ether, thioether, and sulfone propanamides. Though compounds like 1 were
designed as antagonists that irreversibly bind the AR via the NCS moiety, we were only able to
demonstrate moderate (M) anti-proliferation in prostate cancer cell lines (IC values of 20.6-
50
2
3.8 M in various PC cell lines). Seeking to improve the efficacy in models of PC, we explored
the B-ring substitution in a series of secondary and tertiary amines (i.e., class I SARDs). The
high inhibitory potency of the amines led to the discovery of 3 (Figure 2) and its characterization
as our initial SARD providing an entry into AR degradation and helping to define the structural
characteristics of propanamide SARDs.28 Fusion of the tertiary amine of 3 into the B-ring
retained both the AR inhibition and SARD activities at very high potency but with improved
2
8
ligand efficiency and in vitro stability, as exemplified by the indole 7 in Figure 1. Further, 3
and 7 did not degrade glucocorticoid receptor (GR), estrogen receptor (ER), or progesterone
receptor (PR) expressed in cells, demonstrating the AR selectivity of I-III.28 The ability to
degrade AR was generalizable to a series of N-linked indoles (II), indolines (III) (Figure 2), and
various other bicyclic heterocylces (not reported here).
6
ACS Paragon Plus Environment

Page 7 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Our initial [or seminal] SARD SAR exploration is detailed herein. Further, class II-III
SARDs in Figure 2 are evaluated for in vitro anti-proliferation and anti-xenograft activities in
models of Enz-R PC. Unlike previous antiandrogens lacking SARD activity such as 4 and 5, I-
III have the potential to overcome all AR-dependent PC models tested to date, including Enz-R
and/or SV AR expressing CRPCs.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Initial Biological Evaluation of Class I. Herein, we have sought to design new
molecules to potently inhibit and degrade AR with the specific aim to treat CRPC. Initially, we
tested a couple of high affinity but weakly antiproliferative butanamides (i.e., methylene-linked
propanamides) such as 1 and 8 for AR antagonistic activity and their ability to degrade AR at the
protein level as shown in Table 1. SARD activity reported as percent degradation captures the
AR protein levels relative to vehicle treated and are reported qualitatively using the following
abbreviations: -, no degradation (inactive); +, < 30% degradation (weak SARD activity); ++, 31
~ 60% degradation (moderate SARD activity); +++, 61 ~ 90% degradation (strong SARD
activity); ++++: > 90% degradation (complete SARD activity). [The terms inactive, weak,
moderate, strong and complete AR degradation and the symbols are used interchangeably.] High
2
7
affinity 1 (0.032 M) did not inhibit AR transactivation in vitro (> 10 M) despite weak (SV)
to moderate (FL) SARD activity; whereas non-binder 8 showed sub-micro molar level in vitro
inhibition (0.392 M) and moderate FL SARD activity.
Though interesting, these activity profiles did not provide potent inhibition and FL AR
and SV AR degradation in a single molecule, nor consistent antagonism. However, changing the
methylene linkage to amine provided a series of compounds with the in vitro profile we were
looking for, the ability to potently inhibit and degrade FL and SV AR (Table 1). Biaryl B-ring 3
emerged as the lead molecule from series I, demonstrating highly potent inhibition in vitro (48
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
nM) and strong SARD activity that almost completely degrades FL and SV AR at sub-M and
28
low M levels, respectively (Table 1). The shifted nitrogen atom illustrated in Figure 2
produced a unique biological action with the potential to overcome CRPC whether based on AR
point mutation or AR truncation. However, the biphenyl B-ring and methyl substituent of the
amine proved to be metabolic liabilities, limiting the bioavailability of 3, making observation of
AR antagonism in vivo inaccessible.28 These metabolic liabilities were removed by cyclization
of 3 to form classes II (indole) and III (indoline) SARDs possessing a nitrogenous heterobicyclic
B-ring (Figure 2) which did not require aryl substitution and provided new Lipinski-compliant45
chemotypes (Table 2) to co-optimize for antagonism, SARD activity (infra), and drug
metabolism and pharmacokinetic (DMPK) properties (infra), with emphasis on SARD efficacy.
We synthesized the indolyl II and indolinyl III classes of new derivatives and evaluated
the SAR of the SARD activities of these compounds. The A-ring was conserved except for two
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
variants at the 3’-position, namely 3’-CF and 3’-Cl anilides (Table 2). In overview, II-III
3
maintained the high efficacy degradation of FL AR at 1 M dose and SV AR at 10 M dose, and
high potency antagonism (two digit nano-molar range) of 3. Unlike the previous nonsteroidal
AR agonists or antagonists including 4-6 and abiraterone, the structurally optimized members of
II-III showed AR antagonism in in vivo models including models of CRPC.
2
7
Synthesis. Compounds 1 and 8 previously reported by our group , are representative
examples of a series of carbon-linked AR antagonists and were synthesized as previously
reported. 2, an ether linked agonist, was also synthesized as previously reported for other ether
linked propanamides.29, 46 1 and 2 each have a nitrogenous substituent on the aromatic B-rings.
In 1 the nitrogen atom in the isothiocyanate group is  to the B-ring whereas 2 has a  nitrogen
atom in the cyanide group (Figure 2).
8
ACS Paragon Plus Environment

Page 9 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In our design, we shifted the nitrogen into the linkage to produce multiple series (I-III)
with a common unique biological profile which is optimally represented as 3 and 7. 3 was
synthesized as previously reported (see supplementary section).28 The present study illustrates a
series of indolyl (II) and indolinyl (III) small molecule SARDs which were designed and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
synthesized as diaryl (S)-propanamides with two variations of the anilide ring, 3’-CF and 3’-Cl
3
anilides. SAR studies of II-III focused on substitution of the B-ring. The synthesis of II and III
is outlined in Scheme 1. We prepared both the class II and III starting from a chiral hydroxyl
bromide 11 (Scheme 1) reacted with an aniline (9 or 10) to form an R-bromoaniline (12 or 13),
followed by oxirane (14 or 15) formation in the presence of K CO (or other basic conditions),
2
3
via a known procedure.27, 37, 47-48 In overview, compounds of II were prepared in very good yield
by Method A via activation of substituted-1H-indoles (16) with sodium hydride and, its
subsequent reaction with the oxiranes 14 or 15. Similarly, compounds of III were prepared by
Method B which uses LDA to activate substituted 1H-indolines (17) to react with the same
oxiranes (14 or 15), as depicted in Scheme 1. Indoline 21f was prepared by reduction of indole
1
9f as described in the Experimental Section. Although most compounds 16 and 17 were
commercially available, indoles 16a-c were de novo synthesized in our lab as described in the
Experimental Section.
The R-bromoanilides 12 and 13 were synthesized as enantiomerically pure products via
the conserved intermediate bromo-acid 11 (R-isomer) which was synthesized in large quantities
from D-proline as the chiral auxiliary, as previously reported.27, 49 Final products were diaryl (S)-
propanamides. To further probe the SAR, the opposite isomer of 7, namely 7r was prepared
from the S-isomer of compound 11 derived from L-proline via the same synthetic method to yield
an (R)-propanamide.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Overview of the Biological Evaluation of Class II and III SARDs. SAR-guided
substitution of the B-ring allowed co-optimization of AR antagonism and AR degradation
activities for the treatment of Enz-R or SV AR expressing CPRC. The screening regimen
included determination of AR LBD binding affinity, inhibition of R1881-activity in vitro
transcriptional activation (sometimes stated as in vitro antagonism herein), and degradation of
FL and SV ARs, as shown in Tables 1, 3 and 4. Once potent full antagonism and strong pan-
SARD activity were accomplished in a single molecule, DMPK criteria (e.g., see in vitro liver
microsome stabilities studies reported infra) were also considered in the selection of SARDs to
be tested for in vivo efficacy.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In vitro screening: AR LBD binding (K ), wild type antagonism (IC ), and FL and
i
50
SV SARD activity (% degradation). In separate batches of the stated in vitro assays, 7 (5-F
2
8
indole) was included as a positive control, so that variations in the assays could be recognized.
The desired screening profile consisted of: (1) strong (+++) or complete (++++) efficacy FL AR
degradation in LNCaP cells possessing T877A (recently reported as T878A with all other amino
acid numbers also shifted by +1, whereas traditionally numbered as T877A) mutant AR
conferring resistance to hydroxyflutamide, (2) strong (+++) or complete (++++) efficacy SV AR
degradation in 22RV1 cells possessing AR-V7 truncation mutant AR, which is theoretically
resistant to all LBD-targeted antiandrogens, and (3) at least mid nanomolar (nM) potency in
vitro antagonism to block any residual wildtype or mutant AR that might be present despite
strong to complete degradation. (Binding affinity to purified LBD AR was determined but was
not used to select leads, as discussed in the next section.)
Following screening, promising compounds were tested in DMPK assays, e.g., the mouse
liver microsome (MLM) assay reported infra, to determine their relative stability and/or to locate
10
ACS Paragon Plus Environment

Page 11 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
their metabolic liabilities such that they can be blocked. This allows us to advance only
relatively stable compounds possessing all three screening criteria to in vivo testing, including in
models of CPRC.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
LBD vs. Activation Function-1 (AF-1) Binding and Antagonism. Equilibrium-binding
affinity (K ) of series II and III was measured using binding to purified GST-tagged human AR
i
3
LBD in competition with a non-metabolizable and stable steroidal androgen, [ H]-mibolerone
(
[3H]MIB) (at 1 nM). The specific binding at each concentration of the test compounds was
3
obtained by subtracting the non-specific binding of [ H]MIB, and the values are expressed as the
percentage of the specific binding in the absence of the compound of interest.
All assay batches were normalized to the respective K value for DHT in each assay,
i
which was considered as 1 nM. The concentration of compounds II or III that reduced the
3
specific binding of a radiolabeled [ H]MIB by 50% (IC ) was determined by computer fitting
5
0
®
the data to the following equation using SigmaPlot and four parameter logistics.
We hypothesized that series I-III will need to make at least a transient interaction with AR
2
8
to enable its ubiquitination-dependent degradation, suggesting the importance of AR binding.
However, when we tested the degradation properties of II and III, these values correlated poorly
with the AR LBD affinity assay described herein (Tables 3 and 4). Better correlation was seen
between the level of degradation in LNCaP and 22RV1 cells and AR antagonism. Further, the
5
0
series II and III exhibited antiproliferative activity in Enz-R cells (see MR49F data infra)
suggestive of phenotypic effects resulting from the antagonism and SARD activities reported here.
Though generalization is difficult, SARDs of series II and III produced hit or miss nM level
binding (ranged from 0.047 to >10 M) and more consistently potent AR antagonism (ranged from
0
.026 M to 0.835 M). Regarding selectivity of antagonism, equipotent PR antagonism was
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
28
observed for 7 and selected members of II and III (data not shown), but not other steroid
hormone receptors. E.g., indoles 19f and 19b and indolines 21a and 21b lack significant GR
antagonism in in vitro transactivation studies, as shown in Supplementary Figure S1. Further, 19f
and 19b also lack the ability to antagonize the ER in vitro. Given the lack of antagonism of GR
or ER with the above and other SARDs, we do not routinely screen for GR and ER degradation.
The poor correlation between AR LBD binding and antagonist or SARD activity is clearly
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
8
seen for the optical isomers 7 and 7r. They demonstrated comparably high efficacy SARD
activities despite their disparity in binding affinities (0.267 M (7) vs. >10 M (7r)). We have
previously addressed possible rationalizations of this behavior, which include the N-terminal
domain (NTD) binding site for our SARDs reported for 7, 7r, and 3 which differentiates our
compounds from other SARDs or proteolysis-targeting chimaeras (PROTACs).28
Several lines of biophysical evidence substantiate that AF-1 binding is generalizable for I-
III, including fluorescence polarization for 19b (Supplemental Figure S2) and previously for 328
28
28
28
and 7 , and previous NMR studies for 7 and 7r , as well as these and other biophysical methods
applied to these and other structural classes of SARDs not reported here (unpublished results).
The fluorescence polarization study with 19b demonstrated an interaction of the SARD with AF-
1
domain. The interaction was revealed as a shift in the fluorescence emitted by the tyrosine and
tryptophan residues in the presence of 19b, which is consistent with the aforementioned previous
results. The AF-1 domain is present in the N-terminus of the AR and is common to all forms of
the AR expressed in cells. We believe AF-1 binding to be a general mechanism of action for our
SARDs but do not routinely screen new compounds biophysically.
12
ACS Paragon Plus Environment

Page 13 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Moreover, our SARDs may induce degradation by perturbing AR folding or modulating
the protein-protein interactions of AR. These would not necessarily be reflected in AR LBD
binding affinities.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The previous notwithstanding, the AR LBD affinity (for most compounds) and in vitro
antagonist properties of II and III ranged from comparable to favorable relative to the standard
known AR antagonists currently employed clinically for the treatment of PC. E.g., 4, 5, and 6
had binding affinities of 0.509 M, 3.641 M, and 1.452 M, and in vitro antagonism of 0.248
5
1

M, 0.216 M, and 0.016 M (the last value was previously reported ), respectively, compared
to 7 binding of 0.267 M and antagonism of 0.085 M. Despite the standard agents being potent
antagonists, only 6 demonstrated any SARD activity albeit with lower efficacy (<30%), as
shown in Table 3. As expected and consistent with steroidal AR agonists, the nonsteroidal
SARM 2 stimulated expression of FL and SV AR as seen by Western blot, affirming that AR
protein levels in the SARD assay behaved as expected. In contrast, the series II and III
degraded FL AR at 1 M dose and also SV AR at 10 M (22RV1) with efficacies ranging from
3
0-100%. In some cases, the reduction of AR protein levels was nearly 100% and occurred at
nM levels (infra). Our current view is that the SARD activity of these compounds is of primary
importance to their ability to overcome CRPC. Consequently the SAR discussion below will
focus on SARD activity.
Our SARDs display a novel mechanism of action, which, by its nature, provides a
rationale for an expanded disease scope of efficacy and a barrier to the development of
mutational resistance.28 Below, AR antagonism will only be discussed as a secondary
consideration in the biological evaluation.
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
SARD assays in LNCaP and 22RV1 cell lines. We screened all compounds of II and III
for their ability to degrade FL (LNCaP) and SV (22RV1) androgen receptors (Tables 1, 3 and 4).
Western blots for representative compounds from II and III are shown in Figure 3. SARD
activity was measured by treating cells with several doses, e.g., 0.1, 1.0 a n d / or 10 M, of
SARDs in the presence of agonist (0.1 nM R1881). LNCaP or 22RV1 cells were plated in full
serum containing medium. Medium was replaced with 1% charcoal-stripped fetal bovine serum
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(csFBS) containing medium and the cells were maintained in this medium for 2 days. Medium
was changed again after 2 days and the cells were treated as indicated in the figures in the
presence of 0.1 nM R1881. Cells were harvested 24 h after treatment and Western blot for AR
and actin was performed using specific antibodies.
The Western blots were quantified densitometrically and the AR/actin values are
represented as fold change or percent change from vehicle-treated cells. Panel A in Figure 3
showed the degradation of FL AR in LNCaP cells and Panel B showed degradation of SV AR in
2
2RV1 cells, while actin in each lane serves as an internal standard to correct for variations in
protein loading which complicate the visual interpretation of the immunoblots. The percent (%)
degradation values reported in Tables 1, 3, and 4 are normalized for variations in protein loading
and are relied upon for relative efficacy determinations. Dose-dependent degradation was seen
in LNCaP cells for 21f (3’-Cl, 5-F, 6-Ph indoline), 19f (3’-Cl, 5-F, 6-Ph indole), 19b (3’-Cl, 4-F
indole), 20b (3’-CF , 4-F indoline), 21b (3’-Cl, 5-F indoline) and 21d (3’-Cl, 6-F indoline).
3
From Figure 3A, it is apparent that >50% of FL AR is already degraded at 1 M of these
SARDs, i.e. nM range SARD activity. SV AR degradation (the lower molecular weight band in
Figure 3B; upper band is disregarded in % degradation values) for the selected SARDs 21f (3’-
Cl, 5-F, 6-Ph indoline), 19f (3’-Cl, 5-F, 6-Ph indole), 21a (3’-Cl, 4-F indoline), and 7 (3’-CF , 5-
3
14
ACS Paragon Plus Environment

Page 15 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
F indole) was observed to be dose-dependent and generally about 10-fold less potent (Figure 3B)
than FL AR degradation, which is consistent with our other SARDs. Some compounds degrade
FL AR better than SV AR (e.g., 21d) or vice versa (e.g., 18f) (Tables 3 and 4), whereas the
optimal SARD will potently and completely (i.e., ++++) degrade both and has a high potency
antagonism. 21f comes closest to displaying the perfect profile with complete/strong
degradation of FL/SV and antagonism comparable to 5, 0.244 M (21f) vs. 0.216 M (5), but
weaker than other SARDs reported here.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Selectivity and mechanistic aspects of SARD activity: The AR selectivity of degradation
among steroidal receptors has been shown previously, e.g. 3 and 7 were shown to not degrade
28
the expression of PR, ER, and GR . Supplemental Figure S3 further supports this view as it
demonstrates that 19b and 19f do not degrade ER in MCF-7 cells. Moreover, the absence of
significant antagonism of GR by 19b, 19f, 21a and 21b nor ER antagonism by 19b and 19f
(Figure S1 discussed supra) further supports the lack of degradation of these receptors.
Although LBD binding may not correlate to degradation, steroidal receptor antagonism would be
expected if that steroid receptor is degraded.
To validate that SARDs function by reducing the stability of protein (as opposed to
decreasing rate of protein synthesis, e.g., lowering mRNA levels, etc.), we conducted an
experiment with 19f in the presence of the general protein synthesis inhibitor, cycloheximide
(Supplemental Figure S4). Treatment of LNCaP cells with 50 µM cycloheximide or 10 µM 19f
did not significantly reduce the protein levels at the evaluated time-points. However, when the
cells were treated with the two molecules together, significant down-regulation of the AR was
observed. These results suggest that the SARDs function by destabilizing the AR protein that has
already been synthesized as opposing inhibiting protein synthesis. This agrees well with an
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
earlier validation of 3 and 7 using very similar methodology.28 Further, 7 was previously
characterized as increasing rate of AR degradation by targeting AR to the proteasome in LNCaP
2
8
cells using bortezomib. That result was confirmed for indoles reported first herein using the
same methodology but in an EnzR LNCaP cell line (see infra), suggesting that proteosomal
degradation of AR can overcome EnzR.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SARD SAR. As described above, two classes of SARDs were prepared, II (indoles) and
III (indolines), as either 3’-CF (18 and 20 series) or 3’-Cl (19 and 21 series) anilines as shown
3
in Table 2. The B-rings of II and III were substituted with a variety of electron withdrawing
(e.g., halogens, nitro, trifluoromethyl, etc.) and/or phenyl groups. FL and SV SARD analyses
were routine screening of II and III, along with the binding and antagonism assays described
above. SARD activities were determined via Western blots employing NTD directed antibody to
determine AR levels. The FL AR and SV AR experiments were performed in LNCaP and
22RV1 cells, respectively, treated at 1 M and 10 M, respectively. We have found that, in
general, FL AR is more sensitive than SV AR to degradation by II and III, however, the ability
to degrade both is preferable for treating CRPC.
We used Western blot to determine degradation of proteins. As it is difficult to obtain
IC values due to low throughput nature of the system, we quantified the AR bands and
5
0
normalized to loading controls and provided as percent degradation. The percent degradation
values were segregated into five qualitative ranges of values: 1. no degradation is symbolized as
(-) and termed as inactive; 2. < 30% degradation (weak SARD activity) is symbolized as (+); 3.
3
1% to 60% degradation (moderate SARD activity) is symbolized as (++); 4. 61% to 90%
degradation (strong SARD activity) is symbolized as (+++); or 5. > 90% degradation (complete
AR degradation) symbolized as (++++). [The terms inactive, weak, moderate, strong and
16
ACS Paragon Plus Environment

Page 17 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
complete AR degradation and the symbols are used interchangeably.] The SARD SAR
discussion below describes the SARD data in Tables 3 and 4. Based on the current data set, it is
clear that SARD values for FL (LNCaP) and SV (22RV1) AR are not univocal, which may result
in differential efficacy across different PC models. Since the goal is to degrade both FL and SV,
these data are discussed in aggregate as SARD activity (see below).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
’-CF (18 series) vs. 3’-Cl (19 series) A-ring: In general, it is unclear whether series 18
3
(
3’-CF ) or series 19 (3’-Cl) is superior. E.g., comparison in Table 3 of 7 vs. 19c (both 5-F
3
indoles) and comparison of 18b vs. 19b (both 4-F indoles) seem to slightly favor 3’-CF3.
However, it is less clear in the case of 6-F substitution, 18c (-/+++ for FL/SV) vs. 19d (+++/- for
FL/SV). In general, the substitution with 3’-CF vs. 3’-Cl produced approximately equivocal
3
SARD results.
Indole (18 series and 19 series) vs. Indoline (20 series and 21 series): In many cases
only a small advantage is seen for indolines, e.g., 20a vs. 18a (both unsubstituted), 20e vs. 18c
(both 6-F), and 21d vs. 19d (both 6-F). However, comparing indoline 21f vs. indole 19f (both 5-
F, 6-Ph substituted) suggests that the double bond reduction contributes to 21f SARD activity.
Further, both compounds that induced complete (++++) FL degradation were indolines, 21d and
2
1f.
Electron withdrawing groups (EWG) at positions 3-7 of B-ring: In general, electron
withdrawal on the B-ring seems to favor the SARD activities of II and III. For indoles (II),
compare 19a (3-F; +++/+++), 7 (5-F; +++/+++), 7r (R-isomer of 7; 5-F; +++/+++), or 18b-18f
(monofluorination or mononitration at positions 4-6) to the unsubstituted 18a (+/++). For
indolines (III), compare 20b (4-F; +++/+) and 20e (6-F; +++/++) to the unsubstituted 20a
(+++/++). As outlined above, fluorination generally augmented for indoles or retained for
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
indolines SARD activity relative to the unsubstituted analog. Indoles seemed to benefit more
from EWG substitution and fluorination at the 3- (19a) or 5- (7, 7r) position is generally optimal.
Nitration of the indole ring, e.g., 18e (5-NO ) and 18f (6-NO ), showed strong SV SARD
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
activity, and very potent in vitro antagonism with IC values of 0.103 M and 0.058 M,
5
0
respectively, but weaker FL SARD activity. Fluorination at the 4 (18b, 19b, 20b, 21a), 5 (19c,
21b), or 6 (18c, 19d, 20e, 21d) positions of II or III produced SARD portfolios with lower %
degradation values for FL or SV than 7 (5-F). Larger 5-halogenations on the indole ring (II)
abolished activity (18j (5-Br), 18k (5-I)), whereas on the indoline ring (III) it was better
tolerated, e.g., see 21c (5-Br; ++/-) even though ~1 M antagonism was seen for all. 5,6-
Dihalogenation of 3’CF indolines (III) such as 20f (5,6-diF; ++/++) and 20g (5-Cl, 6-F; -/-;
3
agonism) did not improve upon the SARD activity over the monohalo analog 20e (6-F; +++/++),
and further did not conform to the desired screening profile due to partial agonism in vitro which
represents a liability in the treatment of PC. Contrary to this, 21e (5,6-diF in the context of 3’-
Cl; +++/++) had an improved SARD profile vs. 21b (5-F; ++/+) and 21d (6-F; ++++/+) and was
a very potent and pure antagonist (0.032 M). Although 21d (6-F indoline) produced complete
FL AR degradation (and very potent antagonism (0.037 M)), the SV AR degradation was weak
making it a less than ideal compound for CRPC. [Moreover, insertion of a nitrogen atom into
the B-ring of II (18n; 5-CN) produced only weak degradation (+/+) compared to 7 (5-F).]
B-ring 3-position: As suggested above, 3-F (on B-ring) seemed to contribute to SARD
activity based on 19a vs. 18a, however, the 3’-position (A-ring) is also a variable in this
comparison. Also, 3-methylation was tolerated (18g vs. 7), but did not increase activity. Larger
substituents, polar and nonpolar, abolished SARD activity completely such as 18l (3-CO H),
2
1
8m (3-CO Et), and 19g (3-Ph, 5-F), possibly suggesting limited steric tolerance. However,
2
18
ACS Paragon Plus Environment

Page 19 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
8m and 19g did retained moderate antagonism of 0.972 M and 1.032 M, respectively. This
position was not substituted for indolines III, but would explore novel structural space due to the
tetrahedral carbon.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Addition of phenyl to B-ring: The phenylation of indoles (II) was not tolerated at 3, 4,
and 5 positions. E.g., see 19g (3-Ph, 5-F), 18o (4-Ph), and 18p (4-F, 5-Ph) in which activity was
abolished. This is consistent with steric intolerance at the 3 and 5 positions as discussed above.
Steric tolerance at the 6-position of the indole ring is suggested by 18i (5-F, 6-Ph; ++/+) and 19f
(5-F, 6-Ph; ++/++) where mostly moderate SARD activity was maintained upon addition of 6-Ph
group, but lower than 7 (5-F; +++/+++). In contrast, 18q (4-F, 6-(4-fluorophenyl)) was inactive
despite weak antagonism (0.898 M) and strong binding (0.062 M). Interestingly, 6-
phenylation on the indoline ring (III) improved SARD activity, i.e., 21f (5-F, 6-Ph indoline;
++++/+++) was much improved over 21b (5-F indoline; ++/+) [or 19c (5-F indole; +++/+)].
Unlike other indolines, 21f produced complete FL AR and strong SV AR degradation efficacy,
and demonstrated potent antagonism (0.244 M). The structure of 21f is reminiscent of the
biphenyl B-ring of 3 of class I, and may suggest further exploration of the indoline 6-position. 7
(
5-F; +++/+++/ 0.085 M) and 19a (3-F +++/+++/ 0.045 M) also produced impressive CRPC
in vitro screening portfolios.
In vitro metabolic stability in mouse liver microsomes (MLM). Several compounds of
each class were selected to be further evaluated for in vitro metabolic stability in MLM with co-
factors for enzymes of both phase I and phase II metabolism. This allows the calculation of half-
life (T ) and intrinsic clearance (CL ) values as a predictor of the DMPK of series II and III as
1
/2
int
shown in Table 5.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In overview, the CL of II and III is rapid (T ranges from 9.13 min to 36.32 min),
int
1/2
5
1
especially compared to 5 (10.04 h in humans; as previously reported) and 2 (360 min in MLM)
4
3
which are orally active and dosed daily in humans. The unsubstituted indole 18a has a T of
1
/2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
3.66 min. 5-Fluorinated 7 has an equivocal half-life but improved SARD activity (Table 3)
whereas 4-fluorination (18b; 36.32 min) increased the half-life of 18a by approximately 3-fold,
but reduced SARD activity. In general, 4-substituted indoles, and to a lesser extent, 3- or 5-
substituted indoles have improved stability compared to unsubstituted 18a. This may be due to
the steric blocking, or stabilization, of metabolically labile aryl protons on the indole. The most
stable SARD reported is 18b (4-F indole; 36.32 min) is ~3-fold more stable than 18a
(unsubstituted) or 7 (5-F), but this is still far from optimal metabolic stability. Although SARD
activity is preserved in some of these longer T compounds, the relative instability of II and III
1
/2
compared to the standard agents suggested that the pharmacokinetics of these SARDs may need
to be vastly improved in order to reveal the in vivo pharmacodynamic profile of SARDs II and
III. Nonetheless, II and III were tested in in vitro (e.g., antagonism, SARD, and anti-
proliferative assays) and in vivo (e.g., xenografts) models of CPRC (e.g., F876L, MR49F, and
2
2RV1) and demonstrated significant efficacies as will be discussed infra. Efforts to improve
the SARD efficacy and DMPK characteristics of II and III are ongoing.
Models of refractory and CRPC
Overview of refractory PC models employed: LNCaP is a well characterized PC cell line
which expresses a mutant full-length AR (T877A). The T877A mutation confers resistance to
5
2
hydroxyflutamide, but in the absence of other mutants, is sensitive to 5. 22RV1 cells express
53
both FL AR (H874Y) and SV AR (AR-V7). 22RV1 cells demonstrate a more promiscuous
ligand binding than wild-type AR (LBD AR) and respond to low levels of androgens and a wide
20
ACS Paragon Plus Environment

Page 21 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
spectrum of other natural and synthetic steroid hormones, mechanisms proposed to contribute to
5
3
tumor progression following androgen ablation . Further, the AR-V7 in 22RV1 supports
constitutive AR activity and lacks a LBD such that ADT, abiraterone, and traditional
antiandrogens including 5 and 6 cannot block growth of these cells. Our SARD screening
employs the combination of LNCaP and 22RV1 cells, respectively, to predict whether a
particular SARD can degrade [wildtype] and various escape mutant FL (e.g., T877A, H874Y,
W741L, etc.) and SV (e.g., AR-V7 and D567es, etc.) ARs, respectively, which emerge due to
treatment with clinically approved antiandrogens. As new agents are approved, new mutant FL
and SV AR will be discovered as resistance conferring escape mutants.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Recently, a basis for 5 and 6 resistance was discovered to be either the F876L point
mutation or F876F/T877A double mutant. Enz-R MR49F LNCaP cells harbor the double mutant
and serve as a model of Enz-R CRPC, which can be tested, e.g., in in vitro assays of antagonism
of transcriptional activation, SARD activity, or cell anti-proliferation, or as in vivo xenografts.
Xenografts are particularly informative as they reveal the adequacy of the combination of the
pharmacodynamic and DMPK profiles [in the animal species tested] for any given molecule and
provide a proof of concept for whether a particular class of agents is ready to be translated
toward human testing (e.g., IND-enabling studies such rat and dog toxicity, etc.). We produced
3
CRPC xenografts by growing implanted MR49F cells to 100-300 mm in intact animals (i.e.,
endogenous androgens are present), removing androgens via castration, and re-growing in the
absence of androgen. These castration and Enz-R PC xenografts better reflect the CPRC
phenotype seen clinically where patients typically are treated with ADT and 5 or 6 and despite
this, reactivation of the AR axis occurs. Various members of II and III demonstrated activity in
each of the models described above (see infra), including in vivo models, despite poor metabolic
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
stability (Table 5) in the same species, suggesting that metabolically stable SARDs with
comparable or improved pharmacodynamic profiles would be promising candidates for treatment
of many refractory PC’s, including Enz-R CPRC.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In vitro models of CRPC: Antagonism, SARD, and antiproliferation in Enz-R cell lines
SARDs inhibit transcriptional activation of F876L. To validate that II and III can
antagonize the R1881-driven transcriptional activation of mutant AR F876L, we transfected COS
cells with F876L AR with a glucocorticoid response element (GRE)-driven luciferase reporter
construct, and a Renilla reporter construct as a control for transfection efficiency. The GRE-
LUC construct consists of synthetic consensus response elements that are detected by AR, PR,
GR, and mineralocorticoid receptor. This construct is widely used in the nuclear receptor field to
determine the activity of these receptors. Cells were treated 24 h after transfection with 0.1 nM
R1881 (AR agonist) and a dose response of SARDs II and III. Luciferase (and Renilla) assays
were performed 48 h after transfection and reported as relative light units (RLU). COS is not a
prostate cancer cell line, so transfection with F876L does not convert them to Enz-R PC. Figure
4
A (top middle panel) demonstrated potent (low nM) but not full efficacy antagonism by 5 of
R1881-driven F876L transactivation, whereas wt AR inhibition was less potent (low M) and
full efficacy. Importantly, at high concentrations (>1 M), 5 acts as an agonist of F876L
transactivation (top right panel of Figure 4A), which is not seen in wt AR. This is indicative that
F876L acts like an agonist switch escape mutant of 5 therapy.
Given that II and III were structurally novel high potency AR antagonists with a unique
biological activity profile, representative compounds [i.e., 7 (5-F indole), 18b (4-F indole), 18c
(6-F indole), 19c (5-F indole), 19b (4-F indole), and 20b (4-F indoline)] were tested for their
22
ACS Paragon Plus Environment

Page 23 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ability to overcome the agonist switch behavior. Approximately equipotent nM range, full
efficacy antagonism of R1881-driven transcriptional activation was observed in both F876L and
wt. This suggested that our SARDs would also exhibit activity in models of Enz-R (e.g., MR49F
cells) and primary PC possessing wt AR.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SARD activity and cellular anti-proliferation in a model of Enz-R PC (MR49F): To
ensure that SARD activity was also maintained in a Enz-R cell line, SARD assays were
performed in MR49F LNCaP cells containing the F876L/T877A double mutant. As seen in
Figure 4B, 19f (5-F, 6-Ph indole) and 19b (4-F indole) degraded this mutant FL AR in MR49F
cells in the low M and high nM range, respectively, consistent with the relative activities seen
in Tables 3 and 4 Immunoblots suggest that 19b and/or 19f demonstrated comparable potency of
SARD activity in the Enz-R LNCaP (Figure 4B) when compared to the parental 5 sensitive
LNCaP shown supra (Figure 3A). This conservation of SARD activity of these compounds
suggests that their SARD activity is not highly sensitive to small changes in AR amino acid
sequence or the transformed cellular context of the Enz-R cells. 5 was inactive in SARD activity
assays in LNCaP (FL) and 22RV1 (SV) cells and was not expected to be a SARD in the MR49F
context (or any cellular context), as demonstrated in the lower panel of Figure 4B. The
preservation of SARD activity even in the Enz-R context suggested that II and III may exhibit
anti-proliferative and/or anti-tumor activities.
Anti-proliferative assays in MR49F cells showed that 21a (4-F indoline), 19b (4-F
indole), and 19f (5-F, 6-Ph indole) completely and dose-dependently inhibited cell growth with
estimated IC values of less than 3 M for 21a and 19b, and less than 1 M for 19f (Figure 5A).
5
0
For 19b at least, this correlates well with in vitro antagonism and SARD activity in MR49F cells
(Figure 4), suggesting that II and III retained their unique biological profile in Enz-R PC. By
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
comparison, 5 demonstrated weak and incomplete efficacy as revealed by poor dose-dependence
and only partial inhibition of growth. E.g., growth inhibitions at 3 M, 10 M and 30 M were
approximately 30%, 15%, and 45%. This result demonstrated the Enz-R nature of these MR49F
cells, and further affirmed our ability to overcome the Enz-R phenotype with representative
examples of II and III, supporting testing in MR49F xenografts. Further, the AR-dependence of
the anti-proliferative action of our SARDs is demonstrated in Figure 5B by the lack of activity of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
9b and 19f in the AR-negative PC cell line PC-3 at doses as high as 10 M, removing the
possibility of non-specific cytotoxicity underlying anti-proliferative and AR degradation
activities.
Reversal of SARD activity in MR49F cells by inhibition of the proteasome: To
provide a mechanistic basis for the AR degradation observed with II and III, we tested the
effect of a proteasome inhibitor on AR degradation by 19f and 19b in the same model of Enz-R
PC as the xenografts (Figure 6). This experiment builds upon the observed AR destabilization
by 19f as shown in Figure S4, discussed supra. MR49F cells were treated with 19f and 19b in
the presence and absence of bortezomib. Cycloheximide, an inhibitor of protein synthesis, was
included in all treatments to prevent de novo AR synthesis. AR/GAPDH ratios based on
densitometric quantitation demonstrate that bortezomib treatment alone (lane 3; ratio of 1.6)
increased AR levels compared to vehicle treatment (lane 1; ratio of 1). In contrast, 19f and 19b
produced comparable high efficacy AR degradation (lanes 2 and5; ratios of 0.3 and 0.3) that
was completely reversed by equimolar bortezomib (lanes 4 and 6; ratios of 1.2 and 0.9)
producing AR levels comparable to vehicle treatment. This reversal of 19b- and 19f-dependent
degradation by bortezomib supports the hypothesis that ubiquitination-dependent proteasomal
degradation is enhanced by 19b and 19f. The ubiqutin-proteasome system is a tightly
24
ACS Paragon Plus Environment

Page 25 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
regulated, highly specific pathway responsible for the vast majority of protein turnover within
the cell. The ability to target the AR to the ubiquitin-proteasomal pathway even in this model
of EnzR PC helps to rationalize the anti-tumor activity (discussed infra) observed despite poor
PK properties.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In vivo antagonism: Hershberger assays and enzalutamide resistant LNCaP xenografts
Hershberger assays. Uncertain of whether our combination of pharmacodynamic (pan-
antagonism and pan-SARD activities) and pharmacokinetic (e.g., poor stability in MLM)
properties would be sufficiently robust to observe AR antagonism in vivo, we performed
Hershberger assays on several selected II and III orally administered in intact mice and rats.
Surprisingly, despite poor MLM stabilities, the tested SARDs (21a, 21b, 18c, 19f, 7, 19b, 19c,
1
9a and 21f) caused atrophy of AR-dependent seminal vesicles tissue in intact mice (Figure 7A,
left panels) whereas vehicle did not have any effect (0% change). Similar efficacy atrophy was
also observed for 21a and 21b in rats (Figure 7A, right panel) and was demonstrated to be dose-
dependent in prostate and seminal vesicles, with up to ~40% change in organ weights relative to
castrated control (100%). Though only partial efficacy at 40 mg per kg, this confirms that
orally administered compounds are being absorbed and distributed to the site of action in these
organs and suggests that these compounds should also distribute to tumors in xenograft models
to exert anti-tumor effects in sensitive models.
MR49F Xenografts in mice: Following the demonstration of in vitro activity in MR49F
and in vivo antagonism in Hershberger assays, there was confidence in our ability to demonstrate
activity in Enz-R xenografts. MR49F xenograft experiments were completed with 19f and 19b.
1
9b and 19f satisfied all the criteria for a next generation AR antagonist for EnzR PC that
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
includes stronger AR-LBD binding, lower AR antagonistic IC , better degradation of AR and
5
0
AR-SV degradation than enzalutamide. 19b was the best 3’-Cl indole (19 series) in vitro (Table
3), however 19f produced superior in vivo antagonism in intact animals (Figure 7A), despite non-
optimal in vitro values. Hence, we chose 19b and 19f for further characterization to see whether
in vitro or in vivo data was more predictive of anti-cancer activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MR49F xenografts were established by implanting these Enz-R LNCaP cells (a kind gift
from Dr. Martin Gleeve, University of British Columbia) mixed with Matrigel (BD Biosciences,
San Jose, CA) at 1:1 ratio and injecting subcutaneously in NOD SCID gamma (NSG) mice.
3
Once tumor sizes reached 100-200 mm , the mice were castrated and the tumors were allowed to
regrow as CRPC. The animals were randomized once the tumors started to regrow and treated
with vehicle (polyethylene glycol-300:DMSO 85:15 ratio) or 100 mg per kg of SARDs 19f or
1
9b for 14 d. In Figure 7B, 19f and 19b significantly reduced the tumor volume with a 40-60%
tumor growth inhibition (TGI), whereas 5 did not significantly reduce the growth of MR49F
xenografts. Though almost equivocal, 19f appeared to perform slightly better than 19b in this
proof-of-concept experiment, possibly suggesting in vivo data was a better predictor of anti-
cancer activity.
Further, the significant levels of TGI activity indicated that the oral bioavailability
demonstrated in Hershberger assays translated to adequate levels of 19f and 19b in tumor to
reveal [at least to some extent] the pharmacodynamic behavior of our SARDs. Though complete
tumor regression was not accomplished as may be possible with improved pharmacokinetic
properties, the proof-of-concept that our SARDs can overcome Enz-R CRPC in vivo was
established through the susceptibility of these Enz-R xenografts to 19f and 19b. This promising
result is surprising given the poor metabolic stability of II and III as a whole in the same species
26
ACS Paragon Plus Environment

Page 27 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
mice) as seen in MLM (Table 5; T for 19f and 19b were 9.13 min and 11.77 min). These
1
/2
experiments provide hope that our SARDs with their unique biological profile could be used to
overcome 5 and by extension 6 and abiraterone resistances in CRPC patients.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
. CONCLUSIONS AND DISCUSSION
In our effort to find bioactive small molecules for the treatment of advanced prostate
cancers (PCs), a novel series of indolyl and indolinyl propanamides (series II and III,
respectively) were discovered to be selective androgen receptor degraders (SARDs). The first
generation of SARDs were metabolically labile secondary and tertiary amines (I) lacking in vivo
activity that were designed by structural modification of the androgen receptor (AR) antagonist 1
[2-hydroxy-4-(4-isothiocyanatophenyl)-2-methyl-N-(4-nitro-3-
(trifluoromethyl)phenyl)butanamide] and tissue-selective AR agonist enobosarm (2, (S)-N-(4-
cyano-3-(trifluoromethyl)phenyl)-3-(4-cyanophenoxy)-2-hydroxy-2-methylpropanamide). Class
I was exemplified by 3 [(S)-N-(4-cyano-3-(trifluoromethyl)phenyl)-3-((6-cyano-[1,1'-biphenyl]-
3-yl)(methyl)amino)-2-hydroxy-2-methylpropanamide] whose tertiary amine was cyclized to
form indoles (II) and indolines (III). II and III are characterized herein as potent AR
antagonists and SARDs with a broad activity profile in models of PC, and in vivo AR
antagonism when orally administered. E.g., SARDs of II and III exhibited strong AR
antagonistic activity in vitro in transcriptional activation and cellular proliferative assays
including in models of 5 sensitive and resistant PCs, and castration resistant PCs (CRPCs).
Additionally, II and III showed selective AR [protein] degradation of FL (e.g., from
LNCaP cells (T877A)) and SV (e.g., from 22RV1 cells (AR-V7)) isoforms of AR, all at sub to
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 77
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
low micromolar treatment levels, and in a variety of prostate cancer cell contexts including Enz-
R PCs (e.g., MR49F). The ability to degrade SV AR in this study suggested the potential of II
and III to treat various currently untreatable advanced and refractory PCs. E.g., those lacking
the LBD of AR such as AR-V7 and D567es AR truncations, which are not susceptible to ADT,
abiraterone, or LBD-directed antiandrogens (e.g., 4-6), and are associated with short survival.21,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
4
Further, in vivo investigations found that analogs within the II and III series overcome a
variety of escape mutants including F876L and F876L/T877A (MR49F) that are known to
emerge due to enzalutamide (5) treatment. These mutations convert 5 and 6 to agonists,
2
2
conferring resistance to PC cells and tumors via an agonist switch mechanism as seen with
other LBD-binding antiandrogens, e.g. W741L for 4 and T877A for flutamide (N-(4-nitro-3-
(trifluoromethyl)phenyl)isobutyramide). The intractability of truncation mutants and the
frequency of the agonist switch mutations suggest that novel ways, potentially LBD-independent
ways, of targeting the AR are needed. The design, synthesis and biological evaluation of these
SARDs as putative treatments of Enz-R and other advanced PCs is discussed. Moreover, these
orally bioavailable SARDs are dual acting agents, i.e., potent inhibitors and degraders of AR,
providing a higher evolutionary barrier to the development of resistance to II and III. For all
2
8
these reasons, we believe that N-terminally directed SARDs may provide a next generation of
AR antagonists to treat a variety of refractory and/or advanced prostate cancers, including Enz-
R, castration resistant, and/or AR-V7 dependent PCs which are not amendable with current
hormone therapies. As such, SARDs may delay the need to rely solely on chemotherapy.
The lack of satisfactory hormonal pharmacotherapy for metastatic patients that have
failed to respond to abiraterone and/or 5 and 6 [approved in February 2018] has piqued interest
in the development of therapies to overcome Enz-R AR mutations. Some of the promising
28
ACS Paragon Plus Environment

Page 29 of 77
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
55
preclinical approaches include: 1.) combination therapies, e.g., sorafenib or CDK-4/6
5
6
inhibitors with 5 to revert the Enz-R phenotype; or 2.) AR-directed monotherapies such as the
emerging PROTACs technology that exploits cellular quality control machinery to selectively
degrade specific target proteins57-58 by creating in this case AR-PROTAC-E3 ubiquitin ligase
ternary complexes; or 3.) AR-independent approaches to treat CRPC, e.g., by activating Natural
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
9
Killer (NK) cells to attack the cancer ; and each have shown promise in Enz-R or AR-V7
CRPC preclinical models. These approaches are still early in their development cycles, and
many technical and regulatory hurdles remain for these approaches before any might reach the
clinic for Enz-R patients. Other drugs in the clinical pipeline such as second generation
6
0
antiandrogens, e.g., darolutamide (OEM-201) , are just LBD-directed antiandrogens like 5 that
will be susceptible to single amino acid and truncation escape mutations as observed for 5 and 6
and all first generation antiandrogens. Further, in the final analysis, none of the above may be
approved which would leave no viable alternatives to taxanes in antiandrogen resistant CPRC.
Importantly, to date none of the approaches mentioned above are dual targeted
antagonists capable of inhibiting and destroying various FL and SV ARs including all escape
mutants tested to date. Herein we report for the first time, an initial SAR series for our SARD
program for the indole (II) and indoline (III) SARDs described above that are capable of
2
8
destroying FL and SV ARs with high efficacy. Herein and recently, we characterized several
members of groups I-III as having a unique biological activity profile optimized to address Enz-
R CPRC. E.g., these compounds:
1
2
) Generally bind to LBD of wt AR;
) Inhibit transcriptional activation of wt AR (Tables 1, 3, and 4), escape mutant ARs
(F876L in Figure 4A, and T877A, Q711A, L704A, and N705A), and SV AR;28
29
ACS Paragon Plus Environment