Discovery and SAR of a series of agonists at orphan G protein-coupled receptor 139

Article abstract of DOI:10.1021/ml100293q

GPR139 is an orphan G-protein coupled receptor (GPCR) which is primarily expressed in the central nervous system (CNS). In order to explore the biological function of this receptor, selective tool compounds are required. A screening campaign identified compound 1a as a high potency GPR139 agonist with an EC₅₀ = 39 nM in a calcium mobilization assay in CHO-K1 cells stably expressing the GPR139 receptor. In the absence of a known endogenous ligand, the maximum effect was set as 100% for 1a. Screening against 90 diverse targets revealed no cross-reactivity issues. Assessment of the pharmacokinetic properties showed limited utility as in vivo tool compound in rat with a poor whole brain exposure of 61 ng/g and a brain/plasma (b/p) ratio of 0.03. Attempts to identify a more suitable analogue identified the des-nitrogen analogue 1s with a reduced polar surface area of 76.7 A² and an improved b/p ratio of 2.8. The whole brain exposure remained low at 95 ng/g due to a low plasma exposure.

Full text of DOI:10.1021/ml100293q

LETTER
pubs.acs.org/acsmedchemlett
Discovery and SAR of a Series of Agonists at Orphan G
Protein-Coupled Receptor 139
Feng Shi,† Jing Kang Shen,† Danqi Chen,† Karina Fog,§ Kenneth Thirstrup,§ Morten Hentzer,§
Jens-Jakob Karlsson,§ Veena Menon,‡ Kenneth A. Jones,‡ Kelli E. Smith,‡ and Garrick Smith*,§
†Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555 Zu-Chong-Zhi Road, Shanghai 201203, China
‡Lundbeck Research USA, 215 College Road, Paramus, New Jersey 07652-1431, United States
§Neuroscience Drug Discovery Denmark, H. Lundbeck A/S, 9 Ottiliavej, DK-2500 Copenhagen, Valby, Denmark
S Supporting Information
b
ABSTRACT: GPR139 is an orphan G-protein coupled receptor (GPCR) which is primarily
expressed in the central nervous system (CNS). In order to explore the biological function of
this receptor, selective tool compounds are required. A screening campaign identified
compound 1a as a high potency GPR139 agonist with an EC50 = 39 nM in a calcium
mobilization assay in CHO-K1 cells stably expressing the GPR139 receptor. In the absence of a
known endogenous ligand, the maximum effect was set as 100% for 1a. Screening against 90
diverse targets revealed no cross-reactivity issues. Assessment of the pharmacokinetic proper-
ties showed limited utility as in vivo tool compound in rat with a poor whole brain exposure of
61 ng/g and a brain/plasma (b/p) ratio of 0.03. Attempts to identify a more suitable analogue identified the des-nitrogen analogue
1s with a reduced polar surface area of 76.7 Å2 and an improved b/p ratio of 2.8. The whole brain exposure remained low at 95 ng/g
due to a low plasma exposure.
KEYWORDS: Orphan GPR-139, agonists, G-protein coupled receptor, CNS, hydrazinecarboxamide
Scheme 1. Synthesis of Compounds 1a-1oa
PR139 is an orphan G-protein coupled receptor, which has a
G
low sequence similarity (20-25%) to other members of
the GPCR rhodopsin1,2 family. Human and mouse GPR139
share 94% amino acid homology. Tissue distribution of mouse
and human GPR139 is primarily located in the CNS. Human
GPR139 is expressed in putamen, caudate nucleus, entopendund-
ular nucleus, olfactory bulbs, hapenular nucleus, parts of the
hypothalamus, substantia nigra, cerebellar nuclei, and vestibular
nuclei.1
a Reagents and conditions: (a) R2-C10H8-NCO, DCM, rt or R2-C10H8-
NC(O)OPh, DMSO, 80 °C.
In order to elucidate the function of GPR139, selective
ligands are required. Recently, the first surrogate agonists and
antagonists have been reported.3 In the present communica-
tion, we report the discovery of a novel GPR139 agonist series.
A screening campaign was executed at our laboratory to identify
GPR139 agonists. Compound 1a was identified as a GPR139
receptor agonist with an EC50 of 39 nM in a calcium mobiliza-
tion assay using a CHO-K1 cell line stably expressing the
human GPR139 receptor. Initial assessment of compound 1a
indicated no issues with selectivity, as no cross-reactivity was
observed against 90 diverse targets (data in Supporting In-
formation). A hit exploration program was undertaken to
examine whether the potency could be improved while reduc-
ing the polar surface area and hydrogen bond donor count of
1a, which could potentially lead to a brain penetrant tool
compound.4 In this paper we outline the synthesis of analogues
we undertook within the confines of the generic structure 2
(Figure 1), where we probed both the importance of peripheral
substitution R1 and R2 and the significance of the central linker
for GPR139 efficacy.
Variation of peripheral substituents R1 and R2 was achieved
by treating substituted benzhydrazides (3) with either substi-
tuted napthyl isocyanates or substituted napthyl carbamic-acid-
phenyl esters to prepare compounds 1a-1o, as shown in
Scheme 1.
Removal of the hydrazide oxygen was obtained by reaction of
aldehyde 4 with hydrazine in ethanol at reflux to give 1-(3,
5-dimethoxyphenyl)meth-(E)-ylidenehydrazine (5), which upon
treatment with 1-napthyl isocyanate and subsequent imine
reduction gave the des-oxygen analogue 1p (Scheme 2).
Replacement of the hydrazide nitrogens was undertaken as
shown in Scheme 3. N-BOC-glycine (6) was converted to
2-amino-1-(3,5-dimethoxyphenyl)ethanone hydrochloride
(7) and then reacted with 1-napthyl isocyanate to give 1q.
Received: December 8, 2010
Accepted: February 12, 2011
Published: February 28, 2011
r
2011 American Chemical Society
303
dx.doi.org/10.1021/ml100293q ACS Med. Chem. Lett. 2011, 2, 303–306
|

ACS Medicinal Chemistry Letters
LETTER
Figure 1
Scheme 2. Synthesis of Compound 1pa
a Reagents and conditions: (a) hydrazine, EtOH, reflux, 90%; (b) 1-napthylisocyanate, THF, rt, 90%; (c) NaBH3(CN), HOAc, MeOH, 91%.
Scheme 3. Synthesis of Compounds 1q and 1ra
a Reagents and conditions: (a) N,O-dimethylhydroxylamine hydroxylamine, EDCI, DMAP, Et3N, rt, 83%; (b) 3,5-dimethoxyphenylmagnesium
bromide, THF -30-50 °C, 84%; (c) HCl, 1,4-dioxane, 100%; (d) 1-napthyl isocyanate, DIPEA, THF, rt, 59%; (e) 1-napthylamine, EDCI, HOBt,
DIPEA, DMF, rt, 84%; (f) HCl, 1,4-dioxane, rt, 100%; (g) 3,5-dimethoxybenzoic acid, EDCI, HOBt, DIPEA, DMF, rt, 77%.
Alternatively, 1-napthylamine was coupled with N-BOC-glycine
to give 8. Removal of the protecting group and coupling to 3,5-
dimethoxybenzoic acid yielded the glycine derivative, 1r.
The synthesis of the carbon analogue was obtained (Scheme 4)
by coupling 1-napthyl acetic acid (9) with 3,5-dimethoxybenz-
hydrazide to give 1s.
Introduction of methyl groups on the hydrazide nitrogens was
achieved by the use of N-methylhydrazine in treatment with 3,5-
dimethoxybenzoyl chloride or 1-napthylacetyl chloride to give 1t
and 1u, respectively, as shown in Schemes 4 and 5.
The compounds were tested in CHO-K1 cells stably ex-
pressing the human GPR139 receptor for a calcium mobiliza-
tion response measured using a Hamamatsu FDSS7000
imaging plate reader. In the absence of a known endogenous
GPR139 agonist, compound 1a was defined as a full agonist
(i.e., Emax = 100% stimulation). This assumption was sup-
ported by the observation that a saturating dose of 1a typically
produced a calcium mobilization response close to the re-
sponse to ionomycin, a potent and selective Ca2þ ionophore
causing re-equilibration of calcium gradients between mito-
chondria, cytosol, and the extracellular space (data not shown).
A detailed description of assay conditions is provided in the
Supporting Information.
Peripheral substitution of 2, first examining the effect of R1
substitution while maintaining the unsubstituted napthyl,
showed that the unsubstituted analogue 1b was significantly less
potent. Substitution in the 2 position with 2-ethoxy (1d)
markedly reduced activity while the smaller 2-methoxy was
tolerated (1c). Interestingly, the 3,5 diethyl analogue, 1e, was
inactive. The 4 position tolerated fluorine (1g) while the
introduction of an extra methoxy in the 4 position in the tris-
trimethoxy analogue (1h) was not detrimental to activity. The
bulky phenoxy substituent (1i) was also tolerated in the 4
position.
304
dx.doi.org/10.1021/ml100293q |ACS Med. Chem. Lett. 2011, 2, 303–306

ACS Medicinal Chemistry Letters
LETTER
A limited examination (see Table 1) of R2 while maintaining
R1 as the preferred 3,5-bis-methoxy showed toleration of only the
2-methyl substitutent (1l) whereas Br, CN, and Cl substitution in
the 4 position (1j, 1k, 1m), Br in the 5 position (1n), or MeO in
the 7 position (1o) all reduced activity.
Examination of the central linker (see Table 2) showed that
the hydrazide oxygen plays an important role, as its removal in 1p
completely removed activity whereas replacement of the first
hydrazide nitrogen with carbon to give the ketone analogue (1q)
maintained micromolar potency. In contrast, replacement of the
other nitrogens in 1r and 1s did not markedly affect activity,
suggesting that only the amide NH was vital for activity.
However, methylation of both of these nitrogens also markedly
affected activity (1t and 1u).
and improved the apparent permeability in MDCK cells. The
brain/plasma ratio improved markedly, but low plasma levels
meant that the whole brain exposure did not improve. The brain
exposure of 1s was 95 ng/g, corresponding to an unbound brain
concentration of 10.4 nM (see Table 3). The aqueous solubility
of both 1a and 1s at pH 7.4 in 25 mM phosphate buffer was low,
at 1.5 μg/mL and 1.6 μg/mL, respectively.
We have identified novel agonists for the GPR139 orphan
receptor. ADME assessment indicates poor brain exposure in
rat brain after oral dosing, limiting the usefulness of these com-
pounds in vivo.
Table 1. Effect of R1 and R2 on Efficacy at Human GPR139
Receptors
In vitro assessment of the ADME properties of the original hit
1a for suitability as a tool compound for in vivo studies indicated
good stability in rat and human liver microsomes with values well
below the liver blood flow threshold values of 20 mL/min5 and
1.4 L/min, respectively.6 A high polar surface area7 and a
potential efflux issue, as indicated by the MDCK ratio for 1a,
were indeed predictive for the resulting low brain/plasma (b/p)
ratio in rat of 0.03 and a whole brain exposure of 65 ng/g after a p.
o. dose of 50 mg/kg dosed in PEG400. The brain concentration
of 1a corresponded to an unbound brain concentration of 6.23
nM, which is significantly lower than the EC50 of 39 nM (see
Table 3). The most promising compound 1s resulting from the
lead optimization work reduced the PSA from 88.7 Å2 to 76.7 Å2
compd
R1
R2
EC50 (nM)*
Emax (%)
1a
1b
1c
1d
1e
1f
3,5-diMeO
H
H
H
H
H
H
H
H
H
H
39
2400
620
nd
100
85
79
53
6
2-MeO
2-EtO
3,5 diEt
nd
3-MeO
1000
440
63
100
96
76
61
73
20
86
64
54
50
1g
1h
1i
4-F
3,4,5-triMeO
4-PhO
2100
nd
1j
3,5-diMeO
3,5-diMeO
3,5-diMeO
3,5-diMeO
3,5-diMeO
3,5-diMeO
4-Br
1k
1l
4-CN
2-Me
4-Cl
nd
Scheme 4. Synthesis of Compounds of 1s and 1ta
86
1m
1n
1o
nd
5-Br
nd
7-MeO
nd
* If an EC50 could not be determined, the activity is reported as percent
stimulation at 50 μM. nd, not determined. Data are the mean of at least
three experiments. The efficacy of saturating concentrations of com-
pound 1a was set to 100%.
Table 2. Effect of the Central Linker on Efficacy at Human
GPR139 Receptorsa
compd
EC50 (nM)
Emax (%)
1p
1q
1r
1s
1t
>10000
3300
180
100
98
95
74
84
88
2500
a Reagents and conditions: (a) SOCl2, DCM, rt; (b) 3,5-dimethoxy-
benzhydrazide, DIPEA, CH2Cl2, 91%; (c) SOCl2, MeOH, rt, 99%; (d)
CH3(NH)NH2, reflux, 71%; (e) 3,5-dimethoxybenzoyl chloride, DI-
PEA, THF, rt, 69%.
1u
5100
a Data are the mean of at least three experiments. The efficacy of
saturating concentrations of compound 1a was set to 100%.
Scheme 5. Synthesis of Compound 1ua
a Reagents and conditions: (a) SOCl2, MeOH, rt, 98%; (b) CH3(NH)NH2, reflux, 76%; (c) 1-napthylacetyl chloride, DIPEA, THF, rt, 72%.
305
dx.doi.org/10.1021/ml100293q |ACS Med. Chem. Lett. 2011, 2, 303–306

ACS Medicinal Chemistry Letters
LETTER
Table 3. In Vitro and In Vivo ADME Assessment of Compounds 1a and 1sa
ClintR
ClintH
MDCK Papp
MDCK
b/p rat
whole brain
exposure
compd PSA (Å2) c log P (mL/min) (L/min) (1 ꢀ 10-6 cm/s) ratio (BA/AB) 50 mg/kg p.o. @ 2 hb UBP (%) UBBr (%)
1a
88.7
3.34
8.2
1
28.8
4.1
0.03
1.5
3.5
61 ng/g
95 ng/g
1s
76.7
3.48
34
3.1
54
2.6
2.8
0.4
4.0
a Abbreviations: Papp, apparent permeability; PSA, polar surface area; ClintR, intrinsic clearance in rat liver microsomes; ClintH, intrinsic clearance in
human liver microsomes; MDC, Madin-Darby canine kidney cells; b/p, brain plasma ratio; UBP, % unbound in plasma; UBBr, % unbound in brain.
The basolateral-to-apical (B f A) and apical-to-basolateral (A f B) apparent permeability coefficient ratio in the same transport assay was used to
determine the extent of efflux or P-gp substrate specificity of 1a and 1s. b Dosed in PEG400, n = 2.
’ ASSOCIATED CONTENT
S
Supporting Information. Description of the GPR139
b
assay, cross reactivity data, and experimental details for the
synthesis of all compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: gsm@lundbeck.com. Phone: þ45 36 43 32 26.
’ ACKNOWLEDGMENT
We thank Gitte Mikkelsen, Benny Bang-Andersen, Tine Bryan
Stensbøl, and John Paul Kilburn for helpful input in preparation
of the manuscript.
’ REFERENCES
(1) Matsuo, A.; Matsumoto, S.; Nagano, M.; Masumoto, K.; Takasaki,
J.; Matsumoto, M.; Kobori, M.; Masao, K.; Shigeyoshi, Y. Molecular cloning
and characterization of a novel Gq-coupled orphan receptor GPRg1
exclusively expressed in the central nervous system. Biochem. Biophys. Res.
2005, 331, 363–369.
(2) S€usens, U.; Hermans-Bormeyer, I.; Urny, J.; Schaller, H. C.
Characterisation and differential expression of two very closely related
G-protein-coupled receptors, GPR139 and GPR142, in mouse tissue
and during mouse development. Neuropharmacology 2006, 50 (4),
512–520.
(3) Hu, L. A.; Tang, P. M.; Eslahi, N. K.; Zhou, T.; Barbosa, J.; Liu, Q.
Identification of Surrogate Agonists and Antagonists for Orphan G-Pro-
tein-Coupled GPR139. J. Biomol. Screen. 2009, 14, 789–799.
(4) Doan, K. M. M.; Humphreys, J. E.; Webster, L. O.; Wring, S. A.;
Shampine, L. J.; Serabjit-Singh, C. J.; Adkison, K. K.; Polli, J. W. Passive
permeability and P-Glycoprotein-mediated efflux differentiate central
nervous system (CNS) and non-CNS marketed drugs. J. Pharmacol. Exp.
Ther. 2002, 303 (39), 1029–1037.
(5) Houston, B. J. Utility of in vitro drug metabolism data in
predicting in vivo metabolic clearance. Biochem. Pharmacol. 1994,
9, 1469–1479.
(6) Kemme, M. J. B.; Burggraaf, J.; Schoemaker, R. C.; Paulson, S.;
Karim, A.; Lentjes, E. G. W. M.; Childs, A.; Braeckman, R. A.; Cohen,
A. F. The influence of reduced liver blood flow on the pharmacokinetics
and pharmacodynamics of recombinant tissue factor pathway inhibitor
2000, 67 (8), 504–511.
(7) Pajouhesh, H.; Lenz, G. R. Medicinal chemical properties of
successful central nervous drugs. NeuroRx 2005, 2, 541–552.
306
dx.doi.org/10.1021/ml100293q |ACS Med. Chem. Lett. 2011, 2, 303–306