Discovery of (R)-2-(6-Methoxynaphthalen-2-yl)butanoic Acid as a Potent and Selective Aldo-keto Reductase 1C3 Inhibitor

Article abstract of DOI:10.1021/acs.jmedchem.6b00160

Type 5 17β-hydroxysteroid dehydrogenase, aldo-keto reductase 1C3 (AKR1C3) converts Δ⁴-androstene-3,17-dione and 5α-androstane-3,17-dione to testosterone (T) and 5α-dihydrotestosterone, respectively, in castration resistant prostate cancer (CRPC). In CRPC, AKR1C3 is implicated in drug resistance, and enzalutamide drug resistance can be surmounted by indomethacin a potent inhibitor of AKR1C3. We examined a series of naproxen analogues and find that (R)-2-(6-methoxynaphthalen-2-yl)butanoic acid (in which the methyl group of R-naproxen was replaced by an ethyl group) acts as a potent AKR1C3 inhibitor that displays selectivity for AKR1C3 over other AKR1C enzymes. This compound was devoid of inhibitory activity on COX isozymes and blocked AKR1C3 mediated production of T and induction of PSA in LNCaP-AKR1C3 cells as a model of a CRPC cell line. R-Profens are substrate selective COX-2 inhibitors and block the oxygenation of endocannabinoids and in the context of advanced prostate cancer R-profens could inhibit intratumoral androgen synthesis and act as analgesics for metastatic disease.

Full text of DOI:10.1021/acs.jmedchem.6b00160

Subscriber access provided by Northern Illinois University
Article
Discovery of (R)-2-(6-methoxynaphthalen-2-yl)butanoic acid
As a Potent and Selective Aldo-keto Reductase 1C3 Inhibitor
Adegoke O. Adeniji, Md. Jashim Uddin, Tianzhu Zang, Daniel Tamae,
Phumvadee Wangtrakuldee, Lawrence J. Marnett, and Trevor M. Penning
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00160 • Publication Date (Web): 03 Aug 2016
Downloaded from http://pubs.acs.org on August 3, 2016
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of (R)ꢀ2ꢀ(6ꢀmethoxynaphthalenꢀ
2
ꢀyl)butanoic acid As a Potent and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Selective Aldoꢀketo Reductase 1C3
Inhibitor.
1
2
1
1
Adegoke Adeniji; Md Jashim Uddin Tianzhu Zang Daniel Tamae;
1
2*
1*
Phumvadee Wangtrakuldee , Lawrence J Marnett; Trevor M Penning.
1
Department of Systems Pharmacology and Translational Therapeutics and the Center
for Excellence in Environmental Toxicology, Perelman School of Medicine, University of
Pennsylvania, 1315 BRBII/III, 421 Curie Blvd, Philadelphia, PA 19104ꢀ6160
2
Departments of Biochemistry, Chemistry, and Pharmacology, Vanderbilt Institute of
Chemical Biology, Center in Molecular Toxicology, VanderbiltꢀIngram Cancer Center,
Vanderbilt University School of Medicine, Nashville, Tennessee 37232ꢀ0146, United
States
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
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
(
150 word limit)
ABSTRACT: Type 5 17βꢀhydroxysteroid dehydrogenase, aldoꢀketo reductase 1C3
4
(AKR1C3) converts ꢀ ꢀandrosteneꢀ3,17ꢀdione and 5αꢀandrostaneꢀ3,17ꢀdione to
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
testosterone (T) and 5αꢀdihydrotestosterone, respectively in castration resistant prostate
cancer (CRPC). In CRPC, AKR1C3 is implicated in drug resistance, and enzalutamide
drug resistance can be surmounted by indomethacin a potent inhibitor of AKR1C3. We
examined a series of naproxen analogs and find that (R)ꢀ2ꢀ(6ꢀmethoxynaphthalenꢀ2ꢀ
yl)butanoic acid (in which the methyl group of Rꢀnaproxen was replaced by an ethyl
group) acts a potent AKR1C3 inhibitor that displays selectivity for AKR1C3 over other
AKR1C enzymes. This compound was devoid of inhibitory activity on COX isozymes
and blocked AKR1C3 mediated production of T and induction of PSA in LNCaPꢀ
AKR1C3 cells as a model of a CRPC cell line. RꢀProfens are substrate selective COXꢀ2
inhibitors and block the oxygenation of endocannabinoids, and in the context of advanced
prostate cancer Rꢀprofens could inhibit intratumoral androgen synthesis and act as
analgesics for metastatic disease.
2
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
The development of castrate resistant prostate cancer (CRPC) in patients that have
undergone androgen deprivation therapy (ADT) is driven by reactivation of androgen
receptor (AR) signaling within the tumor in the presence of castrate levels of circulating
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
, 2
3ꢀ5
androgens. AR reactivation results from adaptive intratumoral androgen biosynthesis
6
and from changes in the AR itself, including gene amplification , AR mutations that
7 8
9, 10
make the receptor ligand permissive , and the appearance of AR splice variants
which make the receptor constitutively active.
The importance of the conversion of adrenal androgens into testosterone (T) and 5αꢀ
dihydrotestosterone (5αꢀDHT) or de novo androgen biosynthesis within the tumor is
firmly established by the therapeutic efficacy of Abiraterone acetate (AA); an inhibitor of
11ꢀ15
P450c17 (17αꢀhydroxylase/17,20 lyase) in CRPC patients.
However, concerns over
the need for coꢀadministration of prednisone with AA to prevent adrenal insufficiency
and the rapid appearance of drug resistance indicates a pressing need for new therapeutic
1
6, 17
agents.
Aldoꢀketo reductase 1C3 (AKR1C3) also known as type 5, 17βꢀhydroxysteroid
dehydrogenase (HSD17B5), is a 17ꢀketoreductase that catalyzes the NADPH dependent
4
conversion of androgen precursors, 4ꢀandrosteneꢀ3,17ꢀdione (∆ –AD) and 5αꢀandrostanꢀ
3
,17ꢀdione (5ꢀAdione) to yield the potent androgens, T and 5αꢀDHT, respectively
1
8, 19
(
Scheme 1).
AKR1C3 also catalyzes the conversion of androsterone to 5αꢀ
2
0, 21
androstaneꢀ3α,17βꢀdiol the precursor of 5αꢀDHT in the backdoor pathway.
AKR1C3
is one of the most highly over expressed steroidogenic enzymes in CRPC compared to
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
22ꢀ26
normal prostate tissue and prostate cancer.
Moreover, it is dramatically upregulated
2
3, 27, 28
by ADT.
Upon ADT, AKR1C3 is induced by the TMPRSS2ꢀERG fusion protein,
whereby the ERG transcription factor can override the repressive effects of the AR
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
29
binding to the AKR1C3 promoter. AKR1C3 also plays a role in resistance to P450c17
1
6, 17
inhibition by AA observed in prostate cancer cell lines and xenografts
AKR1C3 may
also act as an AR selective coactivator that promotes tumor growth and this coactivator
3
0
function could be blocked with small molecule enzyme competitive inhibitors.
Scheme 1. Central role of AKR1C3 in androgen biosynthesis in prostate cancer.
4
4
DHEA = dehydroepiandrosterone; ꢁ ꢀAD = ꢁ ꢀandrosteneꢀ3,17ꢀdione; Adione= 5αꢀ
5
5
androstaneꢀ3,17ꢀdione; ꢁ ꢀAdiol = ꢁ ꢀandrosteneꢀ3β,17βꢀdiol; 5αꢀDHT = 5αꢀ
dihydrotestosterone; enzymes are referred to by their gene names and are italicized:
HSD3B1 = 3βꢀhydroxysteroid dehydrogenase type 1; SRD5A 5αꢀreductase type 1 or type
2; AKR1C2 = type 3 3αꢀhydroxysteroid dehydrogenase; and HSD17B6 = 17βꢀ
hydroxysteroid dehydrogenase type 6.
4
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Enzalutamide (ENZ) is an AR superꢀantagonist which is also used to treat CRPC
3
1ꢀ33
patients,
and drug resistance also occurs with this agent. ENZ drug resistance in cell
lines and in xenografts could be surmounted with indomethacin a nonꢀsteroidal
antiinflammatory drug (NSAID) and competitive inhibitor of AKR1C3 identified in this
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
4, 35
laboratory.
We are repurposing NSAIDs to be devoid of COXꢀ1 or COXꢀ2 inhibition
while maintaining their inhibitory potency for AKR1C3. We have previously reported on
indomethacin and fenamate analogs that have the requisite nanomolar inhibitory potency
36,37
and selectivity for AKR1C3.
These agents do not inhibit COXꢀisozymes, nor do they
inhibit AKR1C1 and AKR1C2, which are required for the inactivation of 5αꢀDHT within
3
8ꢀ40
the prostate.
Naproxen, (S)ꢀ2ꢀ(6ꢀmethoxynaphthalenꢀ2ꢀyl)propanoic acid is a NSAID that is used
clinically to block cyclooxygenase (COX) mediated inflammation. It is also a potent
AKR1C3 inhibitor that inhibits the AKR1C3 catalyzed reduction of the bioreductive drug
4
1
PRꢀ104 in multiple human cancer cells lines and a lung cancer xenograft model.
34
However, naproxen also inhibits AKR1C2
CRPC.
which limits its therapeutic potential in
We set out to develop naproxen based AKR1C3 inhibitors that would be useful in
CRPC but are devoid of inhibitory activity on COX and other AKR1C enzymes. We
report the discovery of (R)ꢀ2ꢀ(6ꢀmethoxynaphthalenꢀ2ꢀyl)butanoic acid as a competitive,
potent and selective AKR1C3 inhibitor which identifies Rꢀprofens as a new class of
repurposed NSAIDs for CRPC patients. Rꢀprofens are also substrate selective inhibitors
of COXꢀ2 oxygenation of endocannabinoids raising the prospect that in the context of
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
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
prostate cancer they may prevent intratumoral androgen biosynthesis and act as
4
2,43
analgesics in metastatic disease.
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
RESULTS
Chemistry
SꢀNaproxen 1 was the precursor of compounds 5, 6 and 9ꢁ11 (Schemes 2 and 3). A
common intermediate was Sꢀmethylꢀ2ꢀ(6ꢀtrifluoromethylsulfonyloxynaphthalene
propanoate, 4 This compound was synthesized from Sꢀnaproxen which was converted to
42
Oꢀdemethyl naproxen 2 under acidic conditions followed by esterification to yield, 3.
Subsequent addition of trifluoromethane sulfonic anhydride in base gave 4. Compound 4
was converted to compound 5 via the 6ꢀvinylogous intermediate and deesterification. The
protected acid 3 was converted to the ethoxy intermediate using iodoethane and
desesterification gave compound 6 (Scheme 2). Coupling of 4 with sodium
triisopropylsilanethiolate followed by deprotection with tetrabutylammonium fluoride
gave (S)ꢀ2ꢀ(6ꢀ(methylthio)naphthalenꢀ2ꢀyl)propanoate 8, which upon base hydrolysis
yielded the racemic acid, 9. The racemic acid was further oxidized with mꢀ
chloroperoxybenzoic acid to yield racemic 2ꢀ(6ꢀmethylsulfinyl)naphthaleneꢀ2ꢀ
yl)propanoic acid 10, Scheme 3. The steps to compound 11 are identical except the final
oxidation of the methylꢀthio derivative to yield the methylsufonyl derivative used
potassium peroxymonosulfate (Oxone), Scheme 3. Compound 14 was synthesized in two
4
2
steps from 2ꢀbromoꢀ6ꢀmethoxyꢀnaphthalene 12. Compound 15 was synthesized from
racemic 14 via 1,1’ꢀcarbonyldiimidazole coupling of methanesulfonamide, Scheme 3.
6
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
a
Scheme 2. Synthesis of racemic naproxen analogs 2 and 3 Reagents and conditions: (i) 48%
o
o
HBr, AcOH, reflux, 3h; (ii) TMSCl, CH OH, 25 C, 2 h; (iii) (CF SO ) O, Et N, DCM, 25 C, 1 h; (iv)
3
3
2 2
3
o
o
CH =CHBF K, Cs CO , Pd(PPh ) , Et N, EtOH, 50 C, 16 h; (v) (OAc) Pd, tꢀBut P, HCO H, 25 C, 12 h;
2
3
2
3
3 4
3
2
3
2
o
(vi) 3M KOH/CH OH, reflux, 3h, (vii) KOH,C H I, 25 C, 30 min; (viii) 3M KOH/ CH OH, reflux, 2 h.
3
2
5
3
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
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
Scheme 3. Synthesis of racemic naproxen analogs . Reagents and conditions: (i) Pd(PPh ) ,
3
4
o
[
(CH ) CH] SiSH, C H , reflux, 4 h; (ii) TBAF, CH I, 25 C, 2 h; (iii) 3M KOH/ CH OH, reflux, 2h; (iv)
3 2 3 6 6 3 3
o
o
mCPBA, DCM 0 C, 1 h; (v) KHSO . 0.5 KHSO . 0.5 K SO , (CH ) C=O/H O 25 C, 2 h; (vi) Mg, I ,
5
4
2
4
3 2
2
2
THF, reflux, 1 h, (vii) CH CH CHBrCO CH , THF, reflux, 2 h; (viii) 3M KOH/CH OH, reflux, 2h, (ix)
3
2
2
3
3
8
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
o
CDI, CH SO NH , DBU, DCM, 25 C, 4 h.
3
2
2
AKR1C3 and AKR1C2 Inhibitor Screening
Naproxen analogs were synthesized to explore the effect of modifications of its
structure on AKR1C3 activity and selectivity. Based on the observation that other
NSAIDs require the carboxylic acid group for optimal inhibition of the AKR1C
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
44
enzymes, most of the synthesized compounds retained this functional group.
Modifications were made primarily on the chiral αꢀcarbon or the 6–position of the
naphthalene ring. Due to the presence of the chiral center in the molecule, most of
compounds were initially assayed as racemates and screened for AKR1C2 and AKR1C3
inhibition without separation of the enantiomers.
+
The ability of the compounds to inhibit the NADP dependent oxidation of Sꢀtetralol
catalyzed by AKR1C3 and AKR1C2 was determined and IC values obtained, Figure 1
5
0
&
Table 1. Since the Sꢀtetralol assays were performed at K the IC values for the two
m 50
enzymes were directly comparable. Naproxen inhibited AKR1C3 with an IC value of
50
1
80 nM and was not selective for AKR1C3 over AKR1C2, consistent with previous
45
studies. As naproxen, 1 is an Sꢀenantiomer, its Rꢀenantiomer, 1a was evaluated for
AKR1C3 inhibition and selectivity. Rꢀnaproxen inhibited AKR1C3 and AKR1C2 with
IC values of 50 nM and 2.75 ꢂM, respectively, which makes it more potent and more
5
0
selective for AKR1C3 than naproxen. Replacement of the 6ꢀOMe group of naproxen with
an ꢀEt to give 5, or an ꢀOEt to give 6, did not change the potency and selectivity for
AKR1C3. On the other hand, an isosteric replacement of the 6ꢀOMe with the thiomethyl
group, ꢀSMe to give compound 9, led to a 3ꢀfold increase in the inhibitory potency for
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
AKR1C3 (IC = 60 nM) over that seen with naproxen while the AKR1C2 inhibitory
5
0
potency remained unchanged. This translated to 25 fold selectivity for AKR1C3 over
AKR1C2 by compound 9. Because the ꢀSMe group can be metabolized to the S(=O)Me
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
and S(=O) Me groups, compounds 10 and 11 containing these functional groups at the 6ꢀ
2
position were synthesized and screened for AKR1C3 activity and selectivity. Compared
to 9, both compounds 10 and 11 displayed a greater than 15 fold loss of inhibitory
potency on AKR1C3.
Figure 1. Structure of Naproxen Analogs
Table 1: Structure and AKR1C3/AKR1C2 Inhibitory Potency of Naproxen Analogs
Compound
R1
R2
R3
AKR1C3
AKR1C2
Ratio IC50
IC (ꢂM) + IC (ꢂM) +
value
(AKR1C2:
AKR1C3)
7
5
0
50
S.E. (n= 4)
S.E. (n =4)
0
.18 + 0.04
.05 + 0.004
.12 + 0.0075
0.10 + 0.0076
.060 + 0.0054
.05 + 0.0015
1.26 + 0.15
2.75 + 0.35
1.9 + 0.11
2.4 + 0.24
1.5 + 0.14
4.35 + 0.61
1.35 + 0.09
6.3 + 0.35
3.0 + 0.14
7.6 + 0.95
48.1 + 5.0
1.72 + 0.37
34 + 1.8
SꢀNaproxen (1)
RꢀNaproxen (1a)
(±) 5
±) 6
±) 9
R) 9a
S) 9b
±) 10
±) 11
±) 14
R) 14a
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀOH
ꢀMe
ꢀMe
ꢀMe
ꢀMe
ꢀMe
ꢀMe
ꢀMe
ꢀMe
ꢀOMe
ꢀOMe
ꢀEt
ꢀOEt
ꢀSMe
0
56
15
24
25
87
19
6
0
(
(
(
(
0
0
ꢀSMe
ꢀSMe
ꢀS(=O)Me
0
.07 + 0.004
1.05 + 0.13
(
0
.82 + 0.1
4
(
(
(
ꢀMe ꢀS(=O) Me
2
0
.12 + 0.014
58
437
14
6
ꢀEt
ꢀEt
ꢀEt
ꢀEt
ꢀH
ꢀOMe
ꢀOMe
ꢀOMe
ꢀOMe
ꢀOMe
0
.11 + 0.01
.12 + 0.01
6.0 + 1.0
0
(S) 14b
±1) 15
(
ꢀNHSO Me
2
0
.65 + 0.11
19.04 + 3.7
29
1
6
ꢀOH
1
0
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
Windsor et al., ACS Med Chem Lett. 2012, 13 759ꢀ63
The Nꢀ(methylsulfonyl)acetamide analogue, 15 displayed a significant loss of
inhibitory activity for AKR1C3 and AKR1C2 with IC values of 6.0 ꢂM and 34 ꢂM,
5
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
respectively underscoring the need for a free carboxylic acid group for optimal inhibition
of the AKR1C enzymes. Next, the replacement of the αꢀMe group of naproxen with an ꢀ
Et gave 14, which was the most AKR1C3 selective racemate evaluated. With an IC50
value of 120 nM against AKR1C3, compound 14 was more potent as an AKR1C3
inhibitor than naproxen and was 58 fold selective for AKR1C3 over AKR1C2. Due to its
favorable properties the racemic mixture of 14 was separated to give the Rꢀ (14a) and the
Sꢀ (14b) enantiomers. The configuration of the enantiomers was inferred by comparing
the elution order of the enantiomers with that of naproxen and Rꢀnaproxen when the same
chiral column and mobile phase was used (see supporting information). Surprisingly,
while the AKR1C3 inhibitory potency of the enantiomers was similar and not
significantly different from the inhibitory potency of the racemate, the AKR1C2
inhibitory potency of the enantiomers was markedly different. The Sꢀenantiomer
inhibited AKR1C2 with an IC of 1.72 ꢂM while the Rꢀenantiomer displayed an IC
5
0
50
value of 46.4 ꢂM against AKR1C2. This translated to 14ꢀ and 437ꢀfold selectivity for
AKR1C3 inhibition over AKR1C2, respectively. Eliminating the chiral center of
naproxen by the removal of the αꢀMe group to give 16, led to a loss of inhibitory
potency and selectivity for AKR1C3 indicating the importance in retaining the Rꢀ
configuration. The Rꢀ and Sꢀ enantiomers of compound 9 were also examined based on
the selectivity achieved with the Rꢀenantiomer 14a. Compound 9a showed 87ꢀfold
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
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
selectivity for AKR1C3 which was 5ꢀfold less than the selectivity observed with
compound 14a, demonstrating that the –OMe was preferred over the more bulky –SMe.
Compound 14a was the most selective AKR1C3 inhibitor identified from the primary
screen. When tested for inhibition of AKR1C1, it displayed selectivity for AKR1C3 over
AKR1C1 inhibiting the latter with an IC value of 50 ꢂM, which translates to a 500 fold
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
0
selectivity for AKR1C3 (Figure 2.) Similar, selectivity was observed when the
compound was screened against AKR1C4.
14a (ꢀM)
Figure 2. Inhibitory effect of compound 14a on AKR1C1ꢀ4
1
2
ACS Paragon Plus Environment

Page 13 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Effect of 14a on AKR1C2 Catalyzed Reduction of 5αꢁDHT
Compounds 14a and 14b were next evaluated for the ability to inhibit AKR1C2
catalyzed NADPH dependent reduction of 5αꢀDHT. This was necessary to establish the
lack of inhibitory activity on AKR1C2 using its physiologically relevant ketosteroid
substrate. In this reaction, compound 14b inhibited AKR1C2 in a dose dependent manner
over the concentration range of 3 – 100 ꢂM achieving 70% inhibition of enzyme activity
at 100 ꢂM (Figure 3). In contrast, 14a did not display any significant inhibition of
AKR1C2 at <100 ꢂM concentrations. Flufenamic acid (FLU), a nonꢀselective inhibitor of
the AKR1C enzymes displayed about 70% inhibition of AKR1C2 at 10 ꢂM
concentration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
00
8
0
60
4
0
0
0
2
0
1µM
3µM 10µM 30µM 100µM FLU
10µM)
(
Inhibitor Conc
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
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
Figure 3: Effect of compounds 14a (green) and 14b (blue) on AKR1C2 catalyzed
reduction of 5αꢀDHT.
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
Mode of AKR1C3 Inhibition by 14a
+
The pattern of AKR1C3 inhibition by 14a was evaluated in by measuring the NADP
dependent oxidation of Sꢀtetralol and by measuring the NADPH dependent reduction of
4
∆
ꢀandrostenꢀ3,17ꢀdione. Compound 14a competitively inhibited the AKR1C3 catalyzed
oxidation of Sꢀtetralol with a K value of 31 nM (Figure 4a). The same mode of
i
4
inhibition was also observed when the reduction of ∆ –AD was monitored, albeit with a
much higher K value of 750 nM (Figure 4b).
i
1
4
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 4a. Competitive Inhibition of AKR1C3 catalyzed oxidation of Sꢀtetralol by
1
4a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
Figure 4b. Competitive Inhibition of AKR1C3 catalyzed Reduction of ∆ –AD by 14a
Inhibition of COXꢁ1
Naproxen inhibited COXꢀ1 with an IC value of 61 nM. Relative to naproxen, 14b
50
displayed a 30 fold loss of inhibitory potency on COXꢀ1 with an IC value of 1.93 ꢂM
5
0
(
Figure 5). There was a profound loss of inhibitory activity on COXꢀ1 activity by the
respective Rꢀenantiomers, compounds 1a and 14a. Both compounds displayed less than
0% inhibition of COXꢀ1 activity at the highest inhibitor concentration (100 ꢂM) tested.
2
This is consistent with reported structure activity relationship studies on the COXꢀ1
inhibitory effects of naproxen and its analogs.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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
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
1
4b (S-)
4a (R-)
Figure 5. Inhibition of COXꢀ1 by naproxen analogs
Table 2. Inhibitory potency of compounds on AKR1C3 and COXꢀ1
Compound
AKR1C3
IC50
COXꢁ1
IC50
(ꢂM)
0.061
>100
>100
COXꢁ2
IC50
(ꢀM)
0.90
>25
No
inhibition*
No
Ratio
(COXꢁ1 IC50 :
AKR1C3 IC )*
Ratio
(COXꢁ2 IC50 :
AKR1C3 IC50)
15
(
ꢂM)
5
0
Naproxen 1
Rꢀnaproxen 1a
14a
0.18
0.05
0.11
0.34
>2000
>910
>500
∞
14b
0.12
1.93
16
∞
inhibition*
*
Taken from Duggan et al. J. Biol. Chem. 285 (2010) 34950 (ref. 42).
1
6
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A comparison of inhibitory potency of the compounds against AKR1C3 and COXꢀ1
(
Table 2) shows that 14a was almost a thousand fold more selective for AKR1C3 over
COXꢀ1. Compound 14 which is a racemate of 14a and 14b had no inhibitory activity on
COXꢀ2 at the maximum concentration of arachidonic acid tested as substrate (10 ꢀM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Effect on AR Reporter Gene Assay:
Compound 14a was next evaluated for a direct effect on AR signaling. Compound
1
4a did not effect the transꢀactivation of the AR mediated by 5αꢀDHT There was a
concentrationꢀdependent increase in luciferase activity when HeLa13 cells containing a
stably transfected AR and an androgen response element driven luciferase gene were
treated with increasing concentrations of 5αꢀDHT. The luciferase activity peaked at 0.3
nM 5αꢀDHT, however, there were no significant change in the 5αꢀDHT induced
luciferase expression when cells were treated with increasing concentrations of 5αꢀDHT
in the presence of 10 µM compound 14a, Figure 6.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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
Figure 6: Effect of Compound 14a on DHT Induced AR Gene Expression. DHT
alone (purple) and DHT plus 10 ꢀM compound 14a
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
Inhibition of AKR1C3ꢁmediated Production of Testosterone
4
LNCaPꢀAKR1C3 cells were used to probe the metabolism of ∆ –AD in the presence
3
4
and absence of compound 14a. Radioꢀlabeled [ H]ꢀꢁ ꢀAD was incubated with carrier
4
∆
–AD (100 nM) with LNCaPꢀAKR1C3 cells placed in medium supplemented with
charcoal dextran stripped fetal bovine serum (CDꢀFBS) which is devoid of androgens.
Androgens were extracted after 48 hrs and the aqueous fraction subject to βꢀ
glucuronidase treatment to liberate androgen conjugates that we have previously shown
to exist due to the robust uridineꢀ5’ꢀdiphosphoꢀglucuronosyltransferse (UGT) activity in
these cells. Metabolites were separated by radiochromatography using TLC. LNCaPꢀ
3
4
3
AKR1C3 cells were able to metabolize [ H]ꢀꢁ ꢀAD to [ H]ꢀandrosterone glucuronide and
3
[ H]ꢀtestosterone glucuronide. However, in cells that were treated with 30 µM of
3
compound 14a, the production of [ H]ꢀtestosterone glucuronide was significantly
inhibited (Figure 7).
1
8
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Distance
Figure 7: Inhibition of testosterone formation in LNCaPꢀ1C3 cells with Compound 14a.
Panel A shows conversion of 100 nM ꢁ ꢀAD to testosterone in LNCaPꢀAKR1C3 cells
following digestion with βꢀglucurondiase. Panel B, shows the same experiment
performed in the presence of 30 ꢀM compound 14a; Panel C shows statistical analysis of
n=3 versus indomethacin as a positive control * p value < 0.001; p value = 0.001
4
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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
Inhibition of AKR1C3 Mediated AR Gene Expression:
4
Compound 14a was also tested for its ability to block the ꢁ ꢀAD mediated expression
of PSA in LNCaPꢀAKR1C3 cells by western blot analysis Figure 8. Treatment of these
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
4
cells with 100 nM ∆ –AD led to a robust increase in PSA expression. This increase in
4
PSA was reduced when the cells were treated with 100 nM ∆ –AD in the presence of 30
ꢂM of compound 14a.
4
Figure 8: Inhibition of ∆ –AD induced PSA expression in LNCaPꢀAKR1C3 cells by
compound 14a. Panel A, immunoblot; Panel B, densitometric traces of immunoblots with
normalization of PSA to βꢀtubulin for biological replicates (n=3).
2
0
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DISCUSSION
CRPC is currently treated with either AA or ENZ, however, patients rapidly develop
drug resistance leading to an increase in median survival time of only 3ꢀ 4 months. One
mechanism of drug resistance is overexpression of AKR1C3. AKR1C3 represents a
rational target due to its vital role as “gatekeeper” for the production of potent androgens
regardless of the pathway used and its ability to function as a coactivator for the AR. This
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
30, 36,37,
underscores the intense attention the search for AKR1C3 inhibitors has generated.
4
6ꢀ53
NSAIDs are known to be pan inhibitors of the AKR1C enzymes. This is thought to be
as a result of the interaction of the carboxylate groups of NSAIDs with the catalytic site
53
of the AKR1C enzymes. The inhibition of AKR1C3 by NSAIDs is attained at
therapeutic concentrations that are required for COX inhibition. This makes using these
compounds as leads a promising approach since the resulting analogs may have similar
pharmacokinetic profiles and therefore may be well tolerated. We have conducted
structure activity relationship studies on the NSAID, naproxen and identified a Rꢀ
enantiomer that differs from Rꢀnaproxen by the simple substitution of an ethyl group for a
methyl group with therapeutic potential for CRPC.
The presence of small lipophilic groups at the 6ꢀposition of naproxen is optimal for
5
4
COX inhibitory activity.
To determine the contribution of the 6ꢀOMe group to the
inhibitory potency of naproxen for AKR1C3, analogs with small substituents were
synthesized and evaluated. With the exception of the 6ꢀSMe, all the other substituents at
the 6ꢀposition did not significantly improve AKR1C3 potency or selectivity.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
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
All the compounds evaluated were superior inhibitors of AKR1C3 than AKR1C2.
Since the AKR1C enzymes differ primarily in the enzyme subpockets, it is likely that the
larger and more flexible subpockets of AKR1C3 allow for better interaction with the
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
44.56
enzyme.
Naproxen, with a chiral center at the αꢀcarbon is used as the Sꢀenantiomer for antiꢀ
inflammatory activity. This is due to the Rꢀenantiomer being significantly less active as a
4
2,54
COX inhibitor, Figure 5.
Naproxen is likewise a potent AKR1C3 inhibitor.
However, Rꢀnaproxen, compound 1a has never been evaluated for AKR1C3 inhibitory
activity. Relative to naproxen, compound 1a was found to be more potent as an AKR1C3
inhibitor and less potent as an AKR1C2 inhibitor which translated into an increased
selectivity for AKR1C3 relative to AKR1C2. The higher affinity of 1a over naproxen for
4
8
AKR1C3 is supported by crystallographic studies conducted by Flanagan et al. These
authors reported that while both Sꢀ and Rꢀnaproxen bind to AKR1C3 when the respective
+
enantiomers were soaked separately into preformed crystals of the AKR1C3.NADP
complex, only the Rꢀ enantiomer was bound in the these crystals when coꢀsoaked with
both enantiomers. This observation was attributed to the number of intermolecular
interactions made by the enantiomers with the enzyme despite having relatively similar
binding poses with AKR1C3. Relative to naproxen, Rꢀnaproxen made more contacts with
48
side chains in the active site and cofactor leading to a greater complementarity. This
would suggest the existence of stereochemical specificity in the interaction of Rꢀprofens
with AKR1C3 and loss of specificity with AKR1C2. To further explore this observation,
compound 14 with an αꢀEt group and its respective enantiomers 14a(Rꢀ) and 14b(Sꢀ)
2
2
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were subsequently evaluated. Surprisingly, the expected stereochemical specificity was
not apparent in the interaction of the enantiomers, 14a and 14b with AKR1C3 but was
observed with AKR1C2 as the two enantiomers displayed markedly different inhibitory
potency towards AKR1C2. The different inhibitory potency of the enantiomers on
AKR1C2 was subsequently confirmed using a steroidal substrate for AKR1C2. Modeling
studies of enantiomers, 14a and 14b indicate that they adopt remarkably distinct binding
poses with AKR1C2, (see Figure 9 & 10).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
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
Figure 9. Alignment of 14b and 14a in the AKR1C3 active site. AKR1C3 residues (green);
4b (yellow), 14a (purple), Dotted line: possible hydrogen bond; OX: oxyanion site (residues
1
highlighted in pink), Ligand alignments were performed using LigAlign plugin in Pymol. The
+
template crystal structures of the AKR1C3•NADP complexes were chosen from the RCSB
protein data bank (PDB code: 3UFY and 3R58) [also see Supporting Information].
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
Ligand alignments of enantiomers 14a and 14b to the existing crystal structures
(3UFY and 3R58) indicate that they adopt similar binding poses with AKR1C3 (Figure
9
). Both Analogs 14a and 14b show a similar binding mode to that of Rꢀnaproxen with
the ethyl group occupying the sub pocket 3 (SP3) formed by residues Tyrꢀ24, Gluꢀ192,
47
Serꢀ217, Serꢀ221, and Glnꢀ222. The remainder of the molecule extends into the Sub
pocket 1 (SP1) pocket composed of Serꢀ118, Asnꢀ167, Pheꢀ306, Pheꢀ311, and Tyrꢀ319.
The ligand alignments revealed closer proximity of carboxylic group of 14a compared to
14b at the oxyanion site (Tyrꢀ55, Hisꢀ117, and the NADP+ cofactor). A slight increase in
inhibitory potency and selectivity of 14a may be due to the favorable Hꢀbond interaction
with the oxyanion site.
Compounds 14a and 14b display remarkably different binding poses to each other
with AKR1C2 (Figure 10). While analog 14a binds in the active site of AKR1C2 similar
to AKR1C3, compound 14b further extends into the SP3 pocket with no interaction with
the SP1 pocket. When the structures of AKR1C2 and AKR1C3 with either 14a or 14b
bound were overlaid additional reasons for the AKR1C3 selectivity of compound 14a
were found. The Leu 308 side chain at the AKR1C2 binding site sterically clashes with
1
4a, which may significantly decrease its preference towards binding in the SP1 pocket
of AKR1C2, resulting in a loss of inhibitory potency for AKR1C2.
2
4
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Comparison of 14b and 14a binding to AKR1C2 and AKR1C3 The 14b
yellow) and 14a (purple) binding modes in AKR1C2 (cyan) and AKR1C3 (green). The template
crystal structure of the AKR1C2.NADP+ complex was taken from the RCSB protein data bank
PDB code: 4JQ1) [also see Supporting Information].
(
(
Compound 14a competes with Sꢀtetralol and the physiologically relevant steroidal
4
substrate, ∆ –AD for binding to AKR1C3, albeit with different binding constants. The
difference in binding constant is related to the formation of two different inhibitor
34
+
complexes. During the competitive inhibition of Sꢀtetralol oxidation the E.NADP .I
complex forms (where I = inhibitor) to yield a K = 31 nM for 14a. By contrast during the
i
4
competitive inhibition of ∆ –AD reduction the E.NADPH.I complex forms to yield a K
i
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
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
+
=
750 nM for 14a. Thus compound 14a displays a 20ꢀfold preference for the E.NADP
complex, Figure 11. Which complex is inhibited in cells is unclear.
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
A) Oxidation
NADPH
NADP+
S
P
EꢀNADP+
EꢀNADP ꢀS
+
E
EꢀNADPHꢀP
EꢀNADPH
E
8
a
+
EꢀNADP ꢀ8a
B) Reduction
NADPH
NADP+
S
P
+
EꢀNADP+
E
EꢀNADPH
EꢀNADPHꢀS
EꢀNADP ꢀP
E
8
a
EꢀNADPHꢀ8a
Figure 11. Kinetic mechanism of the reactions catalyzed by AKR1C3. (A) oxidation
reaction where the substrate S is Sꢀtetralol and the product P is tetralone; and (B)
4
reduction reaction where the substrate is ꢁ ꢀandrosteneꢀ3,17ꢀdione and the product is
testosterone. S= substrate, P = product and E= enzyme.
Indomethacin, another NSAID has been shown to have greater affinity for the
+
E.NADP complex presumably due to the interaction between the negative charge on the
+
. 44
compound and the positive charge of NADP
Since most AKR1C3 inhibitors that have
2
6
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a carboxylic acid group are anchored to the enzyme’s catalytic site via this functional
3
7, 47, 48, 56, 57
group,
compound 14a with a negatively charged carboxylic acid group at pH
7
will likewise be expected to have similar interaction. This would bring the negative
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
+
charge of 14a close to the cofactor where it will form a stronger interaction with NADP
+
relative to NADPH. The stronger interaction between 14a and NADP would account for
the lower K value obtained for 14a during the oxidation of Sꢀtetralol by AKR1C3 when
i
4
compared to the value obtained during the reduction of ∆ –AD by AKR1C3.
Compound 14a, had no effect on the transactivation of the AR mediated by 5αꢀDHT
in HeLa cells. These cells were also shown to express AKR1C3 endogenously (Adeniji
and Penning, unpublished). Thus if AKR1C3 acts as an AR coactivator in these cells then
3
0
compound 14a differs from GTx560 in that it does not block coꢀactivator function . This
would suggest that not all competitive inhibitors of AKR1C3 are able to inhibit the coꢀ
activator function of this protein. However, coꢀactivator function and its inhibition can be
cell context dependent.
Consistent with the stereoselective inhibition of COX observed with naproxen,
compound 14a was likewise devoid of inhibitory activity on COX which is essential
since chronic COX inhibition is not desirable in the context of CRPC management. We
anticipate that due to the close structural similarity of naproxen with compound 14a
where the methyl group of Rꢀnaproxen was simply replaced by an ethyl group in
compound 14a, that compound 14b would have favorable absorption, distribution,
metabolism, elimination and toxicological (ADMET) properties that would enable its
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
rapid transit into man. However, it is recognized that the addition of the extra methylene
group may effect the compounds lipohilicity and hence affect these properties.
We also note with interest that (R)ꢀprofens are substrateꢀselective inhibitors 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
endocannabinoid oxygenation by COXꢀ2 and suggest that they will have analgesic
42,43
benefit by this mechanism.
As a consequence (R)ꢀprofens in CRPC may not only
inhibit intratumoral androgen synthesis but may alleviate pain associated with the
.
metastatic disease
Conclusion
We have identified (R)ꢀ2ꢀ(6ꢀmethoxynaphthalenꢀ2ꢀyl)butanoic acid as a potent
AKR1C3 inhibitor. This compound competitively inhibits AKR1C3 and displays
selectivity for AKR1C3 over other AKR1C enzymes and COX. It was also efficacious at
blocking AKR1C3 in a LNCaPꢀAKR1C3 cells which is a model for CRPC cells. This
compound represents a potential therapeutic agent that can be used in the management of
CRPC either alone or in combination with AA or ENZ where it may improve efficacy
and reduce the incidence of resistance to these drugs.
EXPERIMENTAL SECTION
Chemistry
GeneralꢀAll commercially available reagents and anhydrous solvents were ACS grade
or better and were used as received.
Analytical thinꢀlayer chromatography was carried out using glassbacked plates coated
2
8
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with fluorescent silica gel 60 F254 from Whatman (Partisil LK6D). Spots were visualized
under natural light and UV illumination at λ = 254 and 365 nm. Flash chromatography
was conducted on a Biotage SP1 automated flash chromatography system equipped with
a fixed wavelength UV detector (λ = 254 nm). Samples were preabsorbed onto readyꢀ
made silica gel samplets and then applied on to normalꢀphase flash chromatography
cartridges (Biotage KPꢀSIL, size according to requirements) and elution with a 0−100%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
13
EtOAc/hexane (0.5% acetic acid) gradient. H and C NMR spectra were recorded at
400 and 100 MHz, respectively, using a Bruker AVꢀ400
with sample changer (BACS 60). A purity of ≥95% (unless otherwise indicated) for
the final compounds in this study was confirmed by analytical HPLC on a Waters HPLC
system with PDA detector (set at λ = 254 nm) equipped with a Supelco Supelcosil LCꢀ18
reverseꢀphase column (15 cm Å~ 3 mm, 5 ꢂm). Compounds screened in the current study
also were characterized with respect to their m.pt. and HRMS.
Preparation of Compounds
(S)ꢁMethyl 2ꢁ(6ꢁ(trifluoromethylsulfonyloxy)naphthalenꢁ2ꢁyl)propanoate, 4. To (S)ꢀ
naproxen (1, 3 g, 13 mmol) in acetic acid (20 mL) was added 48% HBr (11.2 g, 7.5 mL,
o
57
1
38 mmol) at 0 C. Following 3 h of reflux, water (40 mL) was added to precipitate out
the product 2 which was isolated by filtration. To the crude product 2 was added
methanol (35 mL) and TMSCl (1.7 g, 2.0 mL, 15.7 mmol), and the mixture was stirred at
5
8, 59
room temperature for 2 hours.
The solvent was removed in vacuo resulting in a tan
solid 3, which was dissolved in CH Cl (20 mL). Following the addition of triethylamine
2
2
2
9
ACS Paragon Plus Environment