Targeting prostate cancer with compounds possessing dual activity as androgen receptor antagonists and HDAC6 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2016.09.058

While enzalutamide and abiraterone are approved for treatment of metastatic castration-resistant prostate cancer (mCRPC), approximately 20–40% of patients have no response to these agents. It has been stipulated that the lack of response and the development of secondary resistance to these drugs may be due to the presence of AR splice variants. HDAC6 has a role in regulating the androgen receptor (AR) by modulating heat shock protein 90 (Hsp90) acetylation, which controls the nuclear localization and activation of the AR in androgen-dependent and independent scenarios. With dual-acting AR–HDAC6 inhibitors it should be possible to target patients who don't respond to enzalutamide. Herein, we describe the design, synthesis and biological evaluation of dual-acting compounds which target AR and are also specific towards HDAC6. Our efforts led to compound 10 which was found to have potent dual activity (HDAC6 IC₅₀= 0.0356 μM and AR binding IC₅₀= <0.03 μM). Compound 10 was further evaluated for antagonist and other cell-based activities, in vitro stability and pharmacokinetics.

Full text of DOI:10.1016/j.bmcl.2016.09.058

Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Targeting prostate cancer with compounds possessing dual activity
as androgen receptor antagonists and HDAC6 inhibitors
Pradeep S. Jadhavar a, Sreekanth A. Ramachandran a, Eduardo Riquelme b, Ashu Gupta a, Kevin P. Quinn c,
Devleena Shivakumar d, Soumya Ray d, Dnyaneshwar Zende a, Anjan K. Nayak a, Sandeep K. Miglani a,
Balaji D. Sathe a, Mohd. Raja a, Olivia Farias b, Ivan Alfaro b, Sebastián Belmar b, Javier Guerrero b,
c,
Sebastián Bernales c, Sarvajit Chakravarty c, David T. Hung c, Jeffrey N. Lindquist c, , Roopa Rai
⇑
⇑
a Integral BioSciences Pvt. Ltd, C-64, Hosiery Complex Phase II Extension, Noida, Uttar Pradesh 201306, India
b Fundación Ciencia y Vida, Avenida Zañartu 1482, Ñuñoa, Santiago 7780272, Chile
c Medivation, 525 Market Street, 36th Floor, San Francisco, CA 94105, USA
d Schrodinger, Inc., Portland, OR, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
While enzalutamide and abiraterone are approved for treatment of metastatic castration-resistant pros-
tate cancer (mCRPC), approximately 20–40% of patients have no response to these agents. It has been
stipulated that the lack of response and the development of secondary resistance to these drugs may
be due to the presence of AR splice variants. HDAC6 has a role in regulating the androgen receptor
(AR) by modulating heat shock protein 90 (Hsp90) acetylation, which controls the nuclear localization
and activation of the AR in androgen-dependent and independent scenarios. With dual-acting AR–HDAC6
inhibitors it should be possible to target patients who don’t respond to enzalutamide. Herein, we describe
the design, synthesis and biological evaluation of dual-acting compounds which target AR and are also
specific towards HDAC6. Our efforts led to compound 10 which was found to have potent dual activity
Received 6 September 2016
Revised 22 September 2016
Accepted 23 September 2016
Available online xxxx
Keywords:
Prostate cancer
Enzalutamide
HDAC6 inhibitor
Androgen receptor (AR) antagonist
Induced fit docking (IFD)
(HDAC6 IC50 = 0.0356
lM and AR binding IC50 = <0.03 lM). Compound 10 was further evaluated for
antagonist and other cell-based activities, in vitro stability and pharmacokinetics.
Ó 2016 Elsevier Ltd. All rights reserved.
It is widely recognized that chemotherapy drugs are most effec-
tive when given in combination. The rationale for combination
therapy is to harness disparate mechanisms, thereby reducing
the likelihood of resistance. Deriving from the same principles, it
HDACs modulate histone acetylation, which controls gene
expression. HDAC inhibitors have been studied and tested in can-
cer treatment with numerous agents approved and others under-
going clinical trials.4,5 HDAC6 has been implicated in the
pathogenesis and treatment of cancer6 and its role in regulating
the androgen receptor (AR) by modulating heat shock protein 90
(Hsp90) acetylation has also been studied.7,8 Hsp90 acetylation
controls the stability, nuclear localization and activation of the
AR in androgen-dependent and independent scenarios. Inhibition
of HDAC6 therefore provides an opportunity to target castration
resistant prostate cancer.6–9
Enzalutamide has proven to be clinically beneficial in meta-
static castration-resistant prostate cancer (mCRPC).10–12 While
enzalutamide and abiraterone13 are approved for treatment of
mCRPC, approximately 20–40% of patients have no response to
these agents. Moreover, it has been stipulated that the lack of
response and the development of secondary resistance to these
drugs may be due to the presence of AR splice variants.14 Mean-
while, the clinical evaluation of HDAC inhibitors as monotherapy
for prostate cancer has not been promising. However using a
is possible to have
a single molecule with dual activity.
Cabozantinib is a small molecule tyrosine kinase inhibitor of cMet
and VEGFR2; this dual activity has culminated in US FDA approvals
for medullary thyroid cancer and advanced renal cell carcinoma
and it is currently being tested in the clinic for various other can-
cers, including prostate cancer (PC).1 At the research level, a recent
Letter describes the synthesis and evaluation of dual-acting estro-
gen receptor (ER) and histone deacetylase (HDAC) inhibitors
(HDACi).2,3 These ER–HDAC inhibitors combined ER
activity with potent HDAC inhibitor activity, resulting in better
anti-tumor efficacy in ER positive breast cancer cells in vitro
a antagonist
a
when compared to the approved drug Tamoxifen.
⇑
Corresponding authors.
E-mail addresses: jeffrey.lindquist@medivation.com (J.N. Lindquist), roopa.
rai@hotmail.com (R. Rai).
http://dx.doi.org/10.1016/j.bmcl.2016.09.058
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

2
P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
combination of an HDAC inhibitor with antiandrogens, a synergis-
tic increase in cytotoxicity has been demonstrated in a number of
hormone-sensitive and -resistant preclinical models.15 In such sce-
narios, dual-acting AR–HDAC6 inhibitors may be of value. The idea
of using AR binding as a means of directing HDACi to prostate
tumors has been explored, where AR binding is suggested as a
homing device.3 These authors reported using cyanonilutamide
as the AR binding element in their designs of dual AR-HDACi.3
Herein, we describe our efforts to make dual-acting compounds
which target AR and are also specific towards HDAC6, using enza-
lutamide as the AR binding element. During the preparation of our
manuscript, a similar hybrid molecule generated from enzalu-
tamide and HDAC inhibitor such as Vorinostat was reported to
inhibit viability of enzalutamide resistant PC cells by downregulat-
ing HSP90 and AR.16
In the design of our dual AR–HDAC6 inhibitors, the intent was
to maintain AR antagonist activity, while also inhibiting HDAC6.
We first examined the binding mode of enzalutamide, which when
compared to conventional agents like Bicalutamide, binds to the
AR with higher affinity and demonstrates pure antagonist activity
in preclinical models.10 After preparing the 3D structure of enzalu-
tamide using LigPrep,17 we performed IFD to the ligand binding
domain (LBD) of human AR (pdb code: 1T63)18 to understand its
binding mode. The computational methods used are described in
Supporting information (SI) section. Figure 1a illustrates the key
interactions in 2D, highlighting the hydrophobic nature of the
LBD and active site hydrogen bonds. The poses obtained from
induced fit docking (IFD) shows that the trifluoromethyl group
makes favorable van Der Waals (vDW) contacts with hydrophobic
residues Val746, Met749, Phe764 and Leu873. Also the A ring of
enzalutamide (Fig. 1b) forms a T-shaped pi-pi stacking interaction
with Phe764. The cyano group forms H-bonds with two key active
site residues—Arg752 and amide of Gln711. This group occupies
the same position as the keto group in the endogenous substrate
dihydrotestosterone (DHT). The amide oxygen is involved in an
H-bond with Asn705. We also observed that the methyl amide part
of enzalutamide points towards Helix12 (Fig. 1b) of the LBD, which
was crucial for our design strategy moving forward.
HDAC inhibitors typically have a zinc binding group (ZBG)
which is attached to a cap group through a hydrophobic linker.
Many of the reported HDAC inhibitors have an aromatic cap group
such as indole (Panobinostat19) or phenyl (suberoylanilide hydrox-
amic acid, SAHA/Vorinostat19) groups. Until recently, no structure
for HDAC6 or any other class IIb HDACs was available to guide inhi-
bitor design. Two recent Letters detailing the structure of human
and zebrafish HDAC620,21 have added valuable insights to our
design approach. Crystal structures of Vorinostat and Panobinostat
bound to HDAC621 show that the hydroxamic acid in these pan–
HDAC inhibitors interacted with the active-site Zn2+ in a bidentate
mode that was distinctly different from the HDAC6-selective
inhibitor
N-hydroxy-4-(2-((2-hydroxyethyl)(phenyl)amino)-2-
oxoethyl)benzamide (HPOB), which had a monodentate interac-
tion with Zn2+. Moreover, the cap group of these pan–HDAC
inhibitors appear to make interactions with the L1 loop (D460–
P484) that may favor binding to HDAC1–3. Despite having almost
the same first zinc coordination shell, different hydroxamate-zinc
coordination modes are observed in HDACs. For example, the zinc
binding mode is bidentate in HDAC8 (PDB code 1T69) and mon-
odentate in HDAC7 (PDB code C0Z20). The nature of this coordina-
tion appears to be strongly driven by the local environment around
zinc and water in this binding channel.22 Without further detailed
investigation and comparing binding energies of the ligands being
investigated here (such as using high level quantum mechanical
methods), it is not possible to determine if the zinc coordination
mode is indeed a key player for selectivity.
With the knowledge that the methyl amide attached to the dis-
tal phenyl group of enzalutamide points towards Helix12, we pro-
posed to use this phenyl group as the point of attachment of the
HDAC inhibitor, through a suitable linker, as illustrated in Figure 2.
The cyanotrifluoromethyl phenyl group (A-ring) and the thiohy-
dantoin (B-ring) would be maintained to participate in key interac-
tions with the AR ligand binding domain. In this design, the HDAC
inhibitor is positioned such that it putatively points towards
Helix12, thus maintaining AR antagonist activity. Enzalutamide
would replace the aromatic cap of typical HDAC inhibitors. We
planned to explore a variety of linkers as well as zinc-binding
groups to accomplish our objective of dual AR–HDAC6 inhibition.
Our compound designs included a number of published Zn
chelating groups (Fig. 3). These incorporated a thiophene trifluo-
romethyl ketone (1), 2-aminobenzamides (2, 11 and 12), hydroxy-
pyridones (4, 5), hydroxylpyridine-thione (3), sulfamides (6, 7) and
hydroxamic acids (8, 9, 10, 13 and 14).
In order to synthesize the proposed compounds, we prepared
two key intermediates 18 and 19. The synthesis of compound 18
has been reported;23 a similar process was used to synthesize the
bromophenyl compound 19. The routes for the synthesis of the tar-
get molecules are shown in Scheme 1. The details of the syntheses
can be found in SI.
Our evaluation of the synthesized compounds began with bind-
ing assays for AR and HDAC family proteins (Table 1). To identify
activity against HDAC family proteins we screened for inhibition of
several Class I HDACs (HDAC1, 2, and 3) and our target, the Class
IIb HDAC6 using a fluorogenic readout of de-acetylation of target
peptides to quantitate de-acetylation activity.24 While pan–HDAC
inhibitors have shown therapeutic promise, we focused on the
described synergy of HDAC6 specific inhibition with an AR
Figure 1. IFD pose of enzalutamide bound to AR. (a) Interaction of enzalutamide
with AR represented in 2D. (b) The methyl amide points towards Helix12 shown in
green. Enzalutamide is shown in blue ball and stick model. The key residues that
interact with enzalutamide are labeled. The residues Arg752 and amide of Gln711
form H-bonds with the cyano group whereas Asn705 engages the methyl amide tail
of enzalutamide. Phe764 forms a pi-pi stacking interaction with the cyanotriflu-
oromethyl phenyl-ring.
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
3
methyl amide points
towards Helix12
B ring
A ring
C ring
O
F
F
F
F
H
N
N
N
S
F
F
N
O
O
N
F
Enzalutamide
N
F
Zn chelating group
Zn chelating group
N
H
N
S
Linker
O
CAP
Zn Binding Group
Linker
F
F
O
F
O
N
H
N
N
F
N
OH
N
H
S
O
Linker
Vorinostat
O
OH
N
H
H
N
HN
Panobinostat
Figure 2. Compound designs incorporating enzalutamide for AR binding and a Zn-chelating group to bind HDAC.
F
F
R
F
F
S
N
N
4
N
O
R =
O
O
NH2
O
S
O
H
N
CF3
O
OH
S
N
H
N
H
N
N
H
4
4
O
N
1
2
3
O
O
O
O
O
O
NH2
OH
OH
S
N
H
N
N
N
N
H
H
H
4
4
6
4
5
O
O
O
H
H
N
O
O
NH2
N
O
S
O
OH
N
OH
N
N
H
H
H
3
O
O
7
8
9
O
O
O
H
N
N
H
OH
OH
N
N
H
H
3
O
O
NH2
NH2
10
11
12
O
H
N
N
OH
N
H
3
13
14
Figure 3. Structures of compounds to target AR and HDAC6.
antagonist to minimize toxicity issues and maximize the therapeu-
tic benefit. The results are listed in Table 1 with Trichostatin A used
as a positive control as a pan–HDAC inhibitor. The hybrid conjugates
with hydroxamic acids as zinc binding groups (compounds 8–10)
showed HDAC6 specific inhibition while retaining AR antagonistic
activity. Replacing the linker amide or ester in 8 and 9 with methy-
lene groups as in 13, tends to reduce HDAC6 binding activity. Among
the other zinc chelating moieties tested, the sulfamide (6) may also
provide an interesting starting point for optimization of a dual inhi-
bitor of AR and HDAC6. Focusing on the more potent hydroxamic
acid leads, we recognized that compound 8 (an ester) may not be
metabolically stable in vivo. While it provided an interesting SAR
point, we decided to follow-up on the amide analog 10 to gain fur-
ther insights about the potential of such dual-acting compounds.
We tested for AR binding using a radio-ligand binding assay,
where compounds were examined for the ability to inhibit
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

4
P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
O
H2N
B
r
K2CO3, DMF
i)
i)
HATU, DIPEA
O
4
5
1
8
Br
CF3
O
ii) NH2OH.HCl, Na/MeOH
S
CuI
PdCl2(PPh3)2
iii)
H2, PtO2
ii)
O
O
i) H2N
P
yBOP, D
IPEA
O
O
H2
N
i)
HATU, DIPEA
O
9
6
ii) NH2OH.HCl, Na/MeOH
2
NH2
F
ii) aq.NaOH/THF iii)
HATU, DIPEA
O
O
S
NH2
NC
X
OH
H2
N
HATU, DIPEA
O
N
N
O
5
H2N
N
10
6
F3C
HATU, DIPEA
3, (X=S)
, (X=O)
ii) NH2OH.HCl, Na/MeOH
O
4
18
X= S, O
Br HATU, DIPEA
iii)
NH2
HATU, DIPEA
i) NH2
NH2
5
11
14
O
ii)
r
BB 3
O
HN
i) MeI, K2CO3 ; ii) NaBH4, CaCl2 ; iii) MeI, K2CO3
i)
v)
O
H2N
NHBoc
4
iv) BBr3
,
1
HATU, DIPEA
HN
Cs2CO3, TBAI
6, 7
O
ii) TFA, DCM, RT
vi) LiOH.H2O; vi) NH2OH.HCl, HATU, DIPEA
iii) sulfuric diamide, triethylamine
Scheme 1a. Synthesis of compounds 1–11 and 14.
O
i)
Pd(OAc)2, P(tBu)3
O
TFA, DCM
NH2
ii)
12
13
HATU, DIPEA
iii)
F
NH2
Br
CN
F
S
Triethylamine,
Ethanol, rt
CF3
Br
NC
F3C
N
+
N
O
HO
N
H
O
O
i)
NCS
O
17
15
19
Pd(OAc)2, P(tBu)3
PtO , H
ii)
iii) NH2OH.HCl, Na/MeOH
2
2
Scheme 1b. Synthesis of compounds 12 and 13.
Table 1
AR and HDAC panel binding affinities
Compound ID
HDAC1 IC50
(l
M)
HDAC2 IC50
(
lM)
HDAC3 IC50
(
lM)
HDAC6 IC50
(
lM)
AR binding IC50 (lM)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
>10
>10
>10
>10
>10
0.28
>10
>10
>10
>10
>100
14
1.12
>100
0.002
0.495
0.036
47.3
60
0.17
2.03
0.005
—
0.068
<0.03
<0.03
0.25
0.56
0.11
3.37
3.33
>100
>100
>100
>100
1.1
86.5
2.25
14.1
>100
15
>100
>100
>100
>100
3.15
58.1
2.89
6.19
>100
20.6
>100
0.023
—
>100
>100
>100
>100
2
49.2
1.83
5.4
20.2
10.7
>100
0.014
—
0.35
0.043
0.078
<0.03
0.33
0.23
<0.03
<0.03
—
>100
0.024
—
Trichostatin A
Enzalutamide
0.02110
[3H]-Mibolerone binding to AR purified from membrane frac-
tions.25 All the compounds have sub-micromolar AR binding affin-
ity, as shown in Table 1. This was not surprising, considering our
design had maintained most of the enzalutamide structure intact.
Our lead HDAC6 selective compounds (8–10) also show potent
AR binding affinity (Table 1). Compound 8 has a binding affinity
to AR that is comparable to that published for enzalutamide at
21 nM.10
In order to understand the binding and HDAC6 selectivity of our
lead compounds, the recently determined crystal structure of
HDAC6 (pdb code: 5G0J)20 was used as a starting point for docking
calculations. The raw X-ray coordinates for the structure of HDAC6
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
5
was subjected to additional refinement steps using the protein
100.0
80.0
60.0
40.0
20.0
0.0
preparation wizard26 as discussed in SI. The ligands were prepared
using the LigPrep workflow with a metal binding state for the
ligands. Default IFD27 workflow was used for docking. Docking of
compounds 8, 9, 10, 13 and 14 produced reasonable binding poses
where the ligand binds in a way such that the hydroxamate group
is positioned correctly to ‘chelate’ Zn, a binding mode observed in
other HDAC inhibitors (see Fig. 4). The alkyl chain linker attached
to the hydroxamate group is sandwiched between a number of
hydrophobic aromatic residues such as F583 and F643, which is
consistent with binding similar inhibitors to other HDACs. In order
to explain the SAR across this series, we estimated the ligand strain
originating from varying the linker. Ligand strain calculations were
carried out using Prime MM-GBSA method.28,29 Calculations were
done on the free ligand and on the ligand bound in the geometry
it adopts in the HDAC6 complex, both with implicit solvent pre-
sent. The strain energy is the difference between the two energies.
Qualitatively, we observe a good correlation between ligand strain
Compound 10
Compound 8
Enzalutamide
10
100
1000 10000
Concentraꢀon (nM)
Figure 5. Cellular antagonist activity of compounds 8 and 10, compared with
enzalutamide. Compounds were tested at 10, 100, 1,000 and 10,000 nM for
inhibition of DHT-stimulated AR activity in charcoal-stripped-serum (CSS) using
the MDA-kb2 reporter cell line. MBD-kb2 cells were seeded at a density of 10,000
cells per well and kept in complete media for 24 h. Then, the media was replaced
with media containing 5% CSS and cells were incubated for another 48 h. Cells were
then treated with 10 nM DHT in presence of different concentrations of compounds
8 and 10 or the positive control enzalutamide, as noted. Relative luciferase levels
were determined and normalized to vehicle-treated cells.
and measured IC50 values. For example, compound 8 (IC50
0.002 M) has a ligand strain of 3.4 kcal/mol as compared to a
weaker binder such as compound 13 (IC50 = 0.17 M), which has
a ligand strain of 9.6 kcal/mol. The increased ligand strain of com-
pound 13 can be explained by the flexibility in the linker in com-
pound 13, compared to the amide linker of compound 8. Based
=
l
agents for the effects of a 24 h treatment on the LNCaP prostate
cancer cell line stimulated with DHT (Fig. 6a). Treatment with
enzalutamide alone did not alter AR or acetylation levels of tubulin,
as expected. The benzamide HDAC-1 and -3 inhibitor MS-27531 did
not affect AR levels, and mildly increased tubulin acetylation. The
broad spectrum HDACi SAHA resulted in a lowering of AR levels,
possibly due to inhibition of general transcriptional activity, and
a dramatic increase in alpha-tubulin acetylation. The HDAC6 speci-
fic inhibitor, Tubastatin A,32 demonstrated a decrease in AR protein
levels concomitantly with an increase in tubulin acetylation, sup-
porting the reported AR–HDAC6 link.7,8
We tested the effects of the dual AR–HDAC6 inhibitor molecules
(8 and 10) as well and observed a reduction in the steady state AR
protein level compared to the controls (Fig. 6b). Likewise, we
observed hyperacetylation of tubulin compared to control, indicat-
ing that compounds 8 and 10 also have HDAC inhibitory activity in
a cell-based assay. In these cell based assays the levels of AR was
normalized using the ubiquitously expressed cytoskeletal protein
l
on the predicted binding mode, compound 9 (IC50 = 0.5 lM) has a
hydrophilic O which may not be ideal to fit in the hydrophobic
channel lined with F583 and F643, which may result in lower
potency.
We proceeded to evaluate the cell-based activity of selected
compounds in two key assays. First, we tested the potential AR
antagonist activity by using a cell line, MDA-kb2, that stably
expresses an androgen-responsive firefly luciferase reporter.30 To
this end, we co-incubated cells with DHT and different concentra-
tion of our hybrid molecules and we calculated how much inhibi-
tion they have when compared to a DHT only control. We also
included enzalutamide as a positive control for the inhibition of
this DHT-mediated AR activation. We observed a concentration
dependent inhibition that was comparable to enzalutamide with
our compounds (Fig. 5), suggesting that the inherent antagonist
activity associated with the AR-targeted portion of the molecule
remains intact. To further determine the effects of AR antagonists
and differential inhibition of HDAC activity, we then tested several
a)
b)
Compd.
8
Compd.
10
Veh.
Veh.
AR
Tub-Ac
Acꢀn
Figure 6. (a) Examination of AR and acetyl tubulin status with AR antagonist
(enzalutamide, 10 M), Broad spectrum HDACi (MS-275, SAHA 10 M) or an
l
l
HDAC6 specific inhibitor (Tubastatin A, doses as noted) in LNCaP cells stimulated
with 10 nM DHT and indicated agents for 24 h. (b) Tubulin acetylation and AR
protein levels in LNCaP cells stimulated with 10 nM DHT and treated with
Figure 4. Glide docking of compound 10 to HDAC6 X-ray structure. The hydrox-
amate forms bi-dentate interaction with Zn2+
.
The linker occupies
a
rather
compounds 8 and 10 at 10 lM for 24 h. Post treatment whole cell extracts made
hydrophobic channel lined by residues-F583 and F643. An additional H-bond with
the backbone amide of G582 is seen as well. The cap region forms a hydrogen bond
with H462 from the loop L1.
and resolved on an acrylamide gel. Proteins detected with antibodies against AR,
acetylated tubulin and actin as a loading control. (Proteins were detected with
antibodies against AR, acetylated tubulin and actin as a loading control.)
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

6
P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
Table 2
Microsomal stability and glucuronidation potential of compound 10
Compound #
Average human LMa (% rem)
29.1
Average mouse LMa (% rem)
21.9
Average hUGT (% rem)a
97.6
Average mUGT (% rem)a
46.1
10
a
Methods are described in Supporting information.
Systemic clearance greater than hepatic blood flow was observed
for compound 10 in mice which was approximately consistent
with its high in vitro metabolic turnover in the mouse microsomal
stability assay. Similar systemic exposure was achieved by the IP
and SC routes of administration, suggesting that either route of
administration would be acceptable for preclinical testing.
A.
Dose Level
2 mg/kg
IV
10 mg/kg
Route
IP
SC
0.269
0.5
0.706
0.083
0.206
0.427
16.7
0.419
1.0
Cmax (µM)
We have described efforts that have led to the discovery and
in vitro and in vivo evaluation of compound 10, a dual inhibitor
of AR and HDAC6. The hydroxamic acid, which is key to HDAC inhi-
bition is a known liability; such compounds are associated with
poor pharmacokinetics and lack of selectivity among the HDAC
isozymes.34 Compound 10 may be a useful tool compound in phar-
macology studies to evaluate the value of dual inhibition of AR and
HDAC6. It would have to be dosed via SC or IP route to enable suf-
ficient exposure in vivo. Recognizing the value of non-hydroxam-
ate containing HDAC inhibitors, we will focus future efforts to
develop the hit compound 6. Compound 6 embodies a sulfamide
as a Zn binding moiety, replacing the hydroxamate group. Docking
of compound 6 into HDAC6 shows the sulfamide group engaging in
a mono-dentate chelation of Zn, in addition to other interactions
with the protein (Fig. S1 in SI). Compounds containing sulfamide
have been previously shown to be active against Zn containing
enzymes, such as human carbonic anhydrase II (for reference bind-
ing mode, see PDB code: 4FU5). Compound 6 is active as a dual
Tmax (h)
0.846
11.0
NA
1.03
3.42
NA
AUClast (µM *h)
Terminal t1/2 (h)
CL (L/h/kg)
Vz (L/kg)
10.3
NA
NA
Bioavailability (%)
NA
82.1
100
B.
inhibitor (HDAC6 IC50 = 1 lM, AR IC50 = 0.1 lM) and is selective
for HDAC6 (>100 fold over the other isozymes tested), providing
a good starting point for optimization to a lead compound.
Funding: This research did not receive any specific grant from
funding agencies in the public, commercial or not-for-profit sec-
tors. It was funded solely by Medivation.
Acknowledgements
Figure 7. Pharmacokinetic evaluation of compound 10. (A) Noncompartmental PK
parameters (WinNonlin Phoenix v6.4) determined from mice (N = 3/time-point) in
the sparse sampling mode, (B) concentration–time profiles in mice following iv
(2 mg/kg), ip (10 mg/kg) or sc (10 mg/kg) administration. Methods are described in
Supporting information.
The authors thank Dr. Andy Protter, Dr. Michael J. Green and Dr.
Son M. Pham for careful reading of the manuscript and the Analyt-
ical and Purification group at Integral BioSciences for purification
of compounds and for acquiring NMR data.
Supplementary data
actin. The ability of these dual inhibitor molecules to inhibit AR sig-
naling and to reduce the AR protein levels indicates a new
approach to targeting prostate cancer that may be applicable to
patients who are resistant to currently approved therapies.
With compounds that demonstrated effects on AR and HDAC6
in LNCaP cells, we began to focus on PK properties for this chemical
scaffold. Lead compound 10 was evaluated for in vitro microsomal
stability in both human and mouse liver microsomes (Table 2). We
observed low microsomal stability in human and mouse NADPH-
supplemented microsomal incubations. Furthermore, since
hydroxamic acid based HDAC inhibitors are known to undergo
extensive glucuronidation,33 we tested the stability of compound
10 in the in vitro human and mouse UGT assays (Table 2). Interest-
ingly, it showed a lower potential for glucuronidation in a human
assay than in a mouse assay.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.09.
058.
References and notes
1. lvaro, P. Cancer Chemother. Pharmacol. 2014, 73, 219.
2. Tang, C.; Li, C.; Zhang, S.; Hu, Z.; Wu, J.; Dong, C.; Huang, J.; Zhou, H.-B. J. Med.
Chem. 2015, 58, 4550.
3. Gryder, B. E.; Akbashev, M. J.; Rood, M. K.; Raftery, E. D.; Meyers, W. M.; Dillard,
P.; Khan, S.; Oyelere, A. K. ACS Chem. Biol. 2013, 8, 2550.
4. West, A. C.; Johnstone, R. W. J. Clin. Invest. 2014, 124, 30.
5. Falkenberg, K. J.; Johnstone, R. W. Nat. Rev. Drug Disc. 2014, 13, 673.
6. Aldana-Masangkay, G. I.; Sakamoto, K. M. J. Biomed. Biotechnol. 2011, 2011,
875824.
7. Ai, J.; Wang, Y.; Dar, J. A.; Liu, J.; Liu, L.; Nelson, J. B.; Wang, Z. Mol. Endocrinol.
2009, 23, 1963.
8. Seidel, C.; Schnekenburger, M.; Mazumder, A.; Teiten, M.-H.; Kirsch, G.; Dicato,
M.; Diederich, M. Biochem. Pharmacol. 2016, 99, 31.
The lead compound 10 was also profiled for its pharmacokinet-
ics in mouse (Fig. 7). Given its poor metabolic stability in mouse
liver microsomes, we dosed 10 by the intraperitoneal (IP) and sub-
cutaneous (SC) route of administration rather than an oral route.
9. Abbas, A.; Gupta, S. Epigenetics 2008, 3, 300.
10. Tran, C.; Ouk, S.; Clegg, N. J.; Chen, Y.; Watson, P. A.; Arora, V.; Wongvipat, J.;
Smith-Jones, P. M.; Yoo, D.; Kwon, A.; Wasielewska, T.; Welsbie, D.; Chen, C. D.;
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058

P. S. Jadhavar et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
7
Higano, C. S.; Beer, T. M.; Hung, D. T.; Scher, H. I.; Jung, M. E.; Sawyers, C. L.
21. Hai, Y.; Christianson, D. W. Nat. Chem. Biol. 2016.
Science 2009, 324, 787.
22. Wu, R.; Lu, Z.; Cao, Z.; Zhang, Y. J. Am. Chem. Soc. 2011, 133, 6110.
11. Rathkopf, D.; Scher, H. I. Cancer J. 2013, 19, 43.
23. Jain, R. P.; Angelaud, R.; Thompson, A.; Lamberson, C.; Greenfield, S. PCT Int.
Appl. 2011.
24. Howitz, K. T. Drug Discovery Today Technol. 2015, 18, 38.
25. Traish, A. M.; Muller, R. E.; Wotiz, H. H. Endocrinology 1986, 118, 1327.
26. Sastry, G. M.; Adzhigirey, M.; Day, T.; Annabhimoju, R.; Sherman, W. J. Comput.
Aided Mol. Des. 2013, 27, 221.
12. Chen, Y.; Scher, H. I. Nat. Rev. Clin. Oncol. 2012, 9, 70.
13. Rescigno, P.; Lorente, D.; Bianchini, D.; Ferraldeschi, R.; Kolinsky, M. P.; Sideris,
S.; Zafeiriou, Z.; Sumanasuriya, S.; Smith, A. D.; Mehra, N.; Jayaram, A.; Perez-
Lopez, R.; Mateo, J.; Parker, C.; Dearnaley, D. P.; Tunariu, N.; Reid, A.; Attard, G.;
de Bono, J. S. Eur. Urol. 2016.
14. Antonarakis, E. S.; Lu, C.; Luber, B.; Wang, H.; Chen, Y.; Nakazawa, M.; Nadal, R.;
Paller, C. J.; Denmeade, S. R.; Carducci, M. A.; Eisenberger, M. A.; Luo, J. JAMA
Oncol. 2015, 1, 582.
27. Sherman, W.; Beard, H. S.; Farid, R. Chem. Biol. Drug Des. 2006, 67, 83.
28. Miller, E. B.; Murrett, C. S.; Zhu, K.; Zhao, S.; Goldfeld, D. A.; Bylund, J. H.;
Friesner, R. A. J. Chem. Theory Comput. 2013, 9, 1846.
15. Marrocco, D. L.; Tilley, W. D.; Bianco-Miotto, T.; Evdokiou, A.; Scher, H. I.;
Rifkind, R. A.; Marks, P. A.; Richon, V. M.; Butler, L. M. Mol. Cancer Ther. 2007, 6,
51.
29. Guimaraes, C. R. W.; Cardozo, M. J. Chem. Inf. Model. 2008, 48, 958.
30. Wilson, V. S.; Bobseine, K.; Lambright, C. R.; Gray, L. E. J. Toxicol. Sci. 2002, 66,
69.
16. Rosati, R.; Chen, B.; Patki, M.; McFall, T.; Ou, S.; Heath, E.; Ratnam, M.; Qin, Z.
Mol. Pharmacol. 2016, 90, 225.
17. Jain, R. P.; Vederas, J. C. Bioorg. Med. Chem. Lett. 2004, 14, 3655.
18. Estebanez-Perpina, E.; Moore, J. M. R.; Mar, E.; Delgado-Rodrigues, E.; Nguyen,
P.; Baxter, J. D.; Buehrer, B. M.; Webb, P.; Fletterick, R. J.; Guy, R. K. J. Biol. Chem.
2005, 280, 8060.
31. Saito, A.; Yamashita, T.; Mariko, Y.; Nosaka, Y.; Tsuchiya, K.; Ando, T.; Suzuki,
T.; Tsuruo, T.; Nakanishi, O. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 4592.
32. Butler, K. V.; Kalin, J.; Brochier, C.; Vistoli, G.; Langley, B.; Kozikowski, A. P. J.
Am. Chem. Soc. 2010, 132, 10842.
33. Kang, S. P.; Ramirez, J.; House, L.; Zhang, W.; Mirkov, S.; Liu, W.; Haverfield, E.;
Ratain, M. J. Pharmacogenet. Genomics 2010, 20, 638.
19. Wagner, J. M.; Hackanson, B.; Lubbert, M.; Jung, M. Clin. Epigenet. 2010, 1, 117.
20. Miyake, Y.; Keusch, J. J.; Wang, L.; Saito, M.; Hess, D.; Wang, X.; Melancon, B. J.;
Helquist, P.; Gut, H.; Matthias, P. Nat. Chem. Biol. 2016.
34. Suzuki, T.; Miyata, N. Curr. Med. Chem. 2005, 12, 2867.
Please cite this article in press as: Jadhavar, P. S.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.09.058