Discovery of a novel non-steroidal GR antagonist with in vivo efficacy in the olanzapine-induced weight gain model in the rat

Article abstract of DOI:10.1016/j.bmcl.2012.10.074

We report the optimization of a series of non-steroidal GR antagonists that led to the identification of compound 7. This compound is efficacious when dosed orally in an olanzapine-induced weight gain model in rats.

Full text of DOI:10.1016/j.bmcl.2012.10.074

Bioorganic & Medicinal Chemistry Letters 22 (2012) 7376–7380
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a novel non-steroidal GR antagonist with in vivo efficacy in
the olanzapine-induced weight gain model in the rat
Hazel J. Hunt a, , Nicholas C. Ray b, George Hynd b, Jon Sutton b, Mohammed Sajad b, Elizabeth O’Connor b,
⇑
Shahadat Ahmed b, Peter Lockey b, Steve Daly b, Gerry Buckley b, Robin D. Clark a, Robert Roe a,
Christine Blasey a, Joe Belanoff a
a Corcept Therapeutics, 149 Commonwealth Drive, Menlo Park, CA 94025, USA
b Argenta, 8/9 Spire Green Centre, Flex Meadow, Harlow, Essex, CM19 5TR, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
We report the optimization of a series of non-steroidal GR antagonists that led to the identification of
compound 7. This compound is efficacious when dosed orally in an olanzapine-induced weight gain
model in rats.
Received 17 August 2012
Revised 10 October 2012
Accepted 15 October 2012
Available online 23 October 2012
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Glucocorticoid receptor antagonist
Olanzapine-induced weight gain
Non-steroidal
Pyrimidinedione
A growing body of evidence suggests that glucocorticoid recep-
tor (type II) antagonists, particularly mifepristone, can block the
weight gain caused by antipsychotic medication. The effect of
mifepristone has been demonstrated and replicated in both animal
and human studies. In the rat, mifepristone has been shown to
both prevent the onset of antipsychotic induced weight gain, and
reverse weight gain after it is induced.1 Subsequently, two ran-
domized clinical trials in humans demonstrated that mifepristone
significantly attenuated the weight gain caused by the antipsy-
chotic medications olanzapine2 and risperidone.3 In the 28-day
trial of weight gain induced by risperidone, mifepristone also re-
duced the adverse effects of risperidone on fasting insulin and
triglycerides.3
O
HN
O
N
H
N
Figure 1. Structure of compound 1.
We previously reported the discovery and optimization of a ser-
ies of non-steroidal GR antagonists based on a disubstituted pyrim-
idinedione template 6, exemplified by compound 1 in Figure 1.
Whilst 1 exhibited excellent affinity in a GR binding assay and
good selectivity over other nuclear receptors, activity in a reporter
gene assay was only modest. Since compound 1 represents a novel
scaffold for GR antagonism with reduced molecular weight com-
pared with many GR ligands, we considered that further optimiza-
tion of this series was justified. In addition, compound 1 exhibited
an acceptable in vitro ADME profile, with excellent stability in the
human S9 system and no significant inhibition of the five major
human CYP isoforms. Figure 2
Animal studies have shown that other, more specific glucocorti-
coid receptor antagonists, have effects that are similar to those ob-
served in the mifepristone studies. CORT 108297,
a potent
glucocorticoid receptor antagonist without the progesterone recep-
tor antagonist activity associated with mifepristone, was shown to
effectively mitigate olanzapine-induced weight gain in rats.4 More
recently, rats receiving olanzapine plus either of the structural ana-
logues CORT 112716 or CORT 113083 gained significantly less
weight than rats receiving only olanzapine.5 Differences in weight
gain were not attributable to decreased food intake.
Affinity for GR was determined as reported previously6 by mea-
suring displacement of [3H]dexamethasone from recombinant
⇑
Corresponding author.
E-mail address: hhunt@corcept.com (H.J. Hunt).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.10.074

H. J. Hunt et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7376–7380
7377
Figure 2. Change in Body weight (g) in Female Rats Receiving Olanzapine and CORT118335.
baculovirus derived human GR.7 Functional activity was assessed
in SW1353/MMTV-5 cells transfected with a plasmid encoding
firefly luciferase located behind a glucocorticoid response element
(GRE).7 GR antagonist activity was measured as inhibition of dexa-
methasone induced luciferase expression. Selected compounds
were tested for agonist activity by performing the assay in the ab-
acid and transesterification, followed by alkylation of the resultant
b-keto ethyl ester (vii) to provide (viii). Cyclization with thiourea
provided intermediates (ix), and subsequent hydrolysis provided
the desired pyrimidinediones (x).
Previous investigation in the piperidine series indicated that a
variety of small non-polar substituents were tolerated on the pen-
dant benzyl group, and similar results were obtained in this cyclo-
hexyl series. Compounds 4–7 and 8–13, which incorporate
halogens, alkyl, or haloalkyl substituents all have excellent affinity
in the binding assay and good activity in the reporter gene assay. In
contrast, compound 9, which incorporates a more polar sulfone
substituent, is significantly less active in the binding assay. Substi-
tution is acceptable at every position, and di-substitution is well
tolerated.
Next we investigated whether the benzyl substituent could be
replaced by a heteroarylmethyl substituent, with the aim of reduc-
ing lipophilicity and improving aqueous solubility. The incorpora-
tion of 2-pyridyl (compound 14), 2-pyrimidinyl (compound 15) or
2-thiazolyl (compound 18) resulted in a significant loss of GR bind-
ing affinity and activity in the reporter gene assay as shown in
Table 2. Judicious substitution was able to restore some affinity
for GR, see for example compounds 19, 21, 22 and 23. The pyrazole
compound 20 was inactive.
We undertook a limited exploration of substitution on the N-1
position (Table 3). We had shown previously1 in the piperidine ser-
ies that alkylation of N-1 was tolerated, whilst alkylation of N-3
was not. We prepared a small set of cylcohexyl compounds
incorporating substitution on N-1, with a particular emphasis on
analogues with solubilizing substituents. Unfortunately this
endeavour was not successful; although N-methylation was toler-
ated, other substituents resulted in a significant loss of activity in
the reporter gene assay.
sence of dexamethasone. Selectivity over the estrogen (ERa),
androgen (AR) and progesterone (PR) receptors was determined
using ligand binding assays8, whilst selectivity over mineralocorti-
coid receptors (MR) was determined using a reporter gene assay,
using T47D cells which express endogenous MR and PR.7
Our previous SAR investigation involved variation of the sub-
stituents on the pendant benzyl substituent and modification of
the piperidine substituent. All of the original compounds incorpo-
rated a heterocyclic group attached through the nitrogen to the 4-
position of the pyrimidinedione. We decided to replace the nitro-
gen by a carbon, that is, to use a cyclohexyl group in place of the
piperidine, since we considered that this might affect the orienta-
tion of the phenyl substituent.
The original synthetic route used to prepare the cyclohexyl
compounds is depicted in Scheme 1. Thus, conversion of 4-phenyl-
cyclohexanone (i) to the corresponding vinyl boronate (iii) was
achieved via a palladium-catalyzed cross-coupling reaction of
bis(pinacolato)diboron with the readily prepared vinyl triflate
(ii). Microwave assisted Suzuki cross-coupling of the vinyl boro-
nate (iii) with 6-chloropyrimidinedione (iv) using an air-stable pal-
ladium catalyst2 afforded the expected cyclohexenyl compound
(v). Catalytic hydrogenation of the cyclohexene double bond
yielded the target cyclohexyl compound as a 5:7 mixture of trans
and cis isomers, 2 and 3, respectively, which were separated by
preparative reverse phase HPLC.
We were gratified to discover that the trans isomer 2 had excel-
lent affinity for GR, with Ki = 4 nM. The cis isomer 3 was signifi-
cantly less active, with Ki = 32 nM. Of greater interest was the
improved potency of compound 2 in the reporter gene assay, with
Ki = 44 nM compared to 93 nM for compound 1 (Table 1). Having
determined that trans cyclohexyl was a good replacement for
piperidine, we required a synthetic route that would allow more
convenient access to trans compounds without the need for te-
dious separation from the corresponding cis isomer. The trans
selective route is depicted in Scheme 2, and involved the reaction
of trans 4-phenyl cyclohexane carboxylic acid (vi) with Meldrum’s
Finally, we confirmed the importance of the cyclohexyl ring by
replacing it with a phenyl or a 4-substituted piperidine. Both of
these compounds were inactive, data not shown.
Since we had identified a significant number of analogues with
comparable GR binding affinity and activity in the reporter gene
assay we carried out cassette dosing PK studies to prioritize com-
pounds for further evaluation. Each cassette was comprised of 3
new compounds and a standard compound (with an excellent PK
profile) from an unrelated series. The cassettes contained a total
dose of 5 mg/kg and were dosed orally to normal Sprague Dawley

7378
H. J. Hunt et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7376–7380
O
O
O
B
B
OTf
O
O
B
O
n-BuLi, HN(iPr)2, THF, -20 °C
O
PdCl2(dppf).DCM, KOAc,
Dioxane, 80 °C
(83%)
to -78 °C then PhN(OTf)2
(83%)
(i)
(iii)
(ii)
O
PdCl2(tBu2P(Ph-NMe2))2,
CsF, 10% aq. Dioxane,
MW, 140 °C, 40 mins
(32%)
HN
N
H
Cl
O
(iv)
O
O
O
10% Pd(OH)2/C, IMS
DCM, 45 Psi, 50 °C
HN
HN
HN
+
N
H
N
H
O
N
H
O
O
3
2
(v)
Scheme 1. Synthesis of cyclohexyl pyrimidinediones.
Table 1
significant inhibition of the five major human CYPS (3A4, 2D6, 1A2,
2C9, 2C19) and was stable in human and rat microsomes. In a sin-
gle compound iv/po PK study in rats an acceptable profile was
obtained.9 High plasma protein binding (>99.9% in rats and hu-
mans) did not appear to interfere with in vivo activity, vide infra.
CORT118335 was tested in the previously described olanza-
pine-induced weight gain model in female rats. After a three week
acclimatisation period, female Sprague Dawley rats were randomly
assigned to receive twenty-one days dosing with one of the follow-
ing six treatments: vehicle only, olanzapine (2.4 mg/kg/day) plus
vehicle, or olanzapine (2.4 mg/kg/day) plus CORT118335 at 0.2,
0.6, 2, 6 or 10 mg/kg/day. All compounds were administered by
oral gavage in divided doses twice daily. The dose of olanzapine
was selected based on previous experience.4,5
GR binding and reporter gene assay data for substituted benzyl compounds
R
O
HN
O
N
H
Compound
R
GR binding Ki (nM)
Reporter gene Ki (nM)
1
2
8
4
93
44
H
Figure 1 shows the mean weights for each treatment group.
Rats receiving no treatment (the vehicle group) gained significant
weight throughout the study (mean change from baseline to day
28: 18.6 11 g, F = 47, df = 10, p 0.0001). Rats receiving olanzapine
plus vehicle gained approximately 2 times more weight than rats
in the vehicle only group (mean change: +36.5 10 grams), and
this difference was statistically significant (p = 0.0006). Rats trea-
ted with olanzapine and a concomitant treatment of 0 .2 mg/kg
of CORT118335 did not differ in weight change when compared
to rats receiving only olanzapine (mean change: +25.6 13 grams,
p versus olanzapine = 0.78). Statistically significant differences in
body weight were observed when comparing rats receiving higher
doses of CORT118335 to rats receiving olanzapine. Weight change
was significantly attenuated in the groups of rats receiving 0.6 mg/
kg of CORT118335 (mean change: +9.8 12 grams, p versus olan-
zapine = 0.034); 2 mg/kg (mean change: ꢀ10.2 12 grams, p ver-
sus olanzapine <0.00001); 6 mg/kg (mean change: ꢀ23.6 11
grams, p versus olanzapine <0.00001); and 10 mg/kg (mean change
ꢀ23.7 11 grams, p versus olanzapine <0.00001). Rats receiving
olanzapine consumed more food than rats receiving vehicle, and
this increased food consumption was prevented by CORT118335
(0.6 mg/kg and above). Although the mechanism by which GR
antagonists prevent olanzapine-induced weight gain has not been
4
5
6
7
8
9
10
11
12
13
3-CH3
2-CH2CH3
2-Cl
3
19
7
8
557
9
8
5
11
11
21
78
37
24
3-CF3
2-SO2CH3
2-OCH3
2,5 Di-Cl
2,3 Di-Cl
2,4 Di-Cl
2,6 Di-Cl
Not tested
56
23
25
44
67
rats. Based on the results of the cassette PK studies we selected
compound 7 (CORT118335) for further progression. Whereas the
majority of the other compounds evaluated in these cassette PK
studies provided Cmax well below 20 ng/ml, compound 7 achieved
a Cmax of 43 ng/ml, which was not much lower than the level
achieved with the standard compound. CORT118335 exhibited
excellent selectivity against PR, ER and AR, with no significant
affinity in receptor binding assays. Whilst selectivity over MR
was moderate (8 fold), as measured in reporter gene assays, we felt
that it was appropriate for the therapeutic indications of interest to
us, for example, Cushing’s Syndrome. CORT118335 did not exhibit

H. J. Hunt et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7376–7380
7379
O
O
O
O
O
O
O
O
O
NaH, THF
Br
O
O
HO
R
R
(viii)
(vi)
(vii)
Na, EtOH,
thiourea, reflux
R
R
O
O
10% aq. chloroacetic acid,
1,4-dioxane, 100°C
HN
HN
S
N
H
O
N
H
(ix)
(x)
Scheme 2. Selective synthesis of trans-cyclohexylpyrimidinediones.
Table 2
Data for heteroarylmethyl compounds
Ar
O
HN
N
H
O
Compound
Ar
GR binding Ki (nM)
Reporter gene Ki (nM)
14
15
16
17
18
19
20
21
22
23
2-Pyridyl
2-Pyrimidyl
188
400
49
>50
139
28
>1500
22
18
337
>3000
116
>3000
>300
119
>3000
57
3-Chloro-2-pyridyl
2-Methyl-4-pyridyl
2-Thiazolyl
4-Methyl-2-thiazole
2-Methyl-3-pyrazolyl
6-Methyl-2-pyridyl
6-Trifluoromethyl-2-pyridyl
4-Methyl-2-pyridyl
47
44
21
Table 3
Data for N-1alkylated compounds
R
O
R'
N
O
N
H
Compound
R0
R
GR binding Ki (nM)
Reporter gene Ki (nM)
24
25
26
27
28
29
CH3
CH3
H
24
19
9
16
98
>50
57
39
3-CH3
3-CH3
3-CF3
3-CF3
3-CF3
CH2CH2N(CH3)2
CH3
CH2CH2OH
CH2CH2OCH3
>300
162
>3000
Not tested

7380
H. J. Hunt et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7376–7380
conclusively determined, we believe that olanzapine causes insulin
insensitivity both peripherally and centrally, and that corticoste-
rone blockade increases insulin sensitivity, peripherally and cen-
trally. The observed results indicate a return to normal satiety
and enhanced metabolism, both mediated by a return to normal
insulin function. CORT118335 achieves impressive efficacy in this
olanzapine-induced weight gain at low doses, with a minimal
effective dose of 0.6 mg/kg/day.
In summary, we have identified a series of novel non-steroidal
GR antagonists with excellent affinity for GR and good activity in
a reporter gene assay. A practical synthetic route has been devel-
oped, that allows selective synthesis of the more active trans iso-
mer. Compounds in this series exhibit excellent selectivity over
PR, AR and ER and modest selectivity over MR. A representative
from this series, CORT118335, has an acceptable in vitro and
in vivo ADME profile and demonstrates impressive efficacy at a
low dose in an olanzapine-induced weight gain model in rats.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2012.
10.074. These data include MOL files and InChiKeys of the most
important compounds described in this article.
References and notes
1. Beebe, K. L.; Block, T.; DeBattista, C.; Blasey, C.; Belanoff, J. K. Behav. Brain Res.
2006, 171, 225.
2. Gross, C.; Blasey, C. M.; Roe, R. L.; Allen, K.; Block, T. S.; Belanoff, J. K. Adv. Ther.
2009, 26, 959.
3. Gross, C.; Blasey, C. M.; Roe, R. L.; Belanoff, J. K. Obesity 2010, 18, 2295.
4. Belanoff, J. K.,; Blasey, C. M.; Clark, R. D.; Roe, R. L.; 2009. Diabetes. Obes. Metab.
2009, 12, 545.
5. Belanoff, J. K.; Blasey, C. M.; Clark, R. D.; Roe, R. L. Eur. J. Pharmacol. 2010, 655,
117.
6. Ray, N. C.; Clark, R. D.; Clark, D. E.; Williams, K.; Hickin, H. G.; Crackett, P. H.;
Dyke, H. J.; Lockey, P. M.; Wong, M.; Devos, R.; White, A.; Belanoff, J. K. Bioorg.
Med. Chem. Lett. 2007, 17, 4901.
Acknowledgments
7. Morgan, B. P.; Swick, A. G.; Hargrove, D. M.; LaFlamme, J. A.; Moynihan, M. S.;
Carroll, R. S.; Martin, K. A.; Lee, E.; Decosta, D.; Bordner, J. J. Med. Chem. 2002, 45,
2417.
The authors gratefully acknowledge Sharon Wickes, Roshini
Markandu, Anne White and Colin Bright for running and analyzing
the in vitro ADME studies; Graham Hagger for carrying out the
in vivo PK studies; Mike Podmore and Russel Scammell for analyt-
ical chemistry support.
8. Estrogen receptor: [3H]estradiol, Pan Vera 26467A Era; androgen receptor:
[3H]dihydrotestosterone, Pan Vera 24938 AR; progesterone receptor:
[3H]progesterone, Pan Vera 24900 PR.
9. CORT118335 was dosed at 1mg/kg iv and 5mg/kg po. The iv clearance was
22ml/min/kg with a half-life of 1.8 h. Bioavailability was 20%.