Discovery of 4-(dimethylamino)quinazolines as potent and selective antagonists for the melanin-concentrating hormone receptor 1

Article abstract of DOI:10.1016/j.bmcl.2005.03.052

A series of 4-(dimethylamino)quinazoline based antagonists of the melanin-concentrating hormone receptor 1 (MCH-R1) is described. This series was derived from a lead compound, AR129330, identified by HTS of a GPCR-directed library using a functional assay with a constitutively activated (CART) form of the receptor. The preliminary optimization resulted in the identification of compounds 20, 21, and 23.

Full text of DOI:10.1016/j.bmcl.2005.03.052

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2565–2569
Discovery of 4-(dimethylamino)quinazolines as potent and
selective antagonists for the melanin-concentrating
hormone receptor 1
Kosuke Kanuma,a Katsunori Omodera,a Mariko Nishiguchi,a Takeo Funakoshi,a
Shigeyuki Chaki,a Graeme Semple,b Thuy-Anh Tran,b Bryan Kramer,b Debbie Hsu,b
Martin Casper,b Bill Thomsen,b Nigel Beeleyb and Yoshinori Sekiguchia,*
aMedicinal Research Laboratories, Taisho Pharmaceutical Co. Ltd, 1-403 Yoshino-cho, Kita-ku, Saitama, Saitama 331-9530, Japan
bArena Pharmaceuticals Inc., 6166 Nancy Ridge Drive, San Diego, CA 92121, USA
Received 28 January 2005; revised 9 March 2005; accepted 14 March 2005
Available online 11 April 2005
Abstract—A series of 4-(dimethylamino)quinazoline based antagonists of the melanin-concentrating hormone receptor 1 (MCH-R1)
is described. This series was derived from a lead compound, AR129330, identified by HTS of a GPCR-directed library using a func-
tional assay with a constitutively activated (CART) form of the receptor. The preliminary optimization resulted in the identification
of compounds 20, 21, and 23.
Ó 2005 Elsevier Ltd. All rights reserved.
Melanin-concentrating hormone (MCH) is a cyclic 19
amino acid peptide expressed in the lateral hypotha-
lamic area (LHA) and the rostromedial zona incerta.1,2
There has been significant interest in the pharmacology
of MCH since the discovery of the MCH-R1.3 Based on
the localization of MCH-containing neurons, a role for
MCH in feeding behavior and energy expenditure has
been suggested. Indeed, the intracerebroventricular
(icv) injection of MCH in rats results in increased food
intake.4,5 Moreover, MCH mRNA was reported to be
upregulated by fasting in both normal and ob/ob mice
and to be increased in ob/ob mice relative to wild-type
animals.4 In contrast, prepro-MCH-knockout mice were
found to be hypophagic and have reduced body weight
and increased leanness relative to their wild-type coun-
terparts.6 It is suggested that MCH-R1, the only form
of the receptor shown to be present in rodents, may
mediate the physiological effects of MCH in those spe-
cies. From these data it can be inferred that a centrally
acting MCH-R1 antagonist may reduce food intake
and bodyweight, and hence may be effective in treating
obesity. In addition, MCH was reported to induce an
anxiogenic effect when injected into the medical preoptic
area in rats,7 while icv injection of MCH increased the
number of entries into the open arms in the elevated
plus-maze test,8 suggesting that MCH may also be
implicated in the modulation of emotional states. This
hypothesis is supported by the recent finding that
SNAP-7941, a selective antagonist for the MCH-R1,
exhibited anxiolytic- and antidepressant-like effects in
rodent models.9
MCH-R1 antagonists reported in the literature to date
encompass a wide range of structural types (Fig. 1).
The first non-peptide MCH-R1 antagonist T-226296
has a tetralin substructure and has been found to have
low nanomolar affinity for the MCH-R1.10 One of the
most potent MCH-R1 antagonists hitherto described is
compound SNAP-7941, with a reported IC50 value of
0.04 nM in an MCH-R1 binding assay.11
Our initial hit identification however came not from
modification of such known structures, but from high-
throughput screening of our in-house GPCR-directed li-
brary. The library contains a series of collections based
on a number of privileged fragments, some of which
Keywords: Melanin-concentrating hormone receptor 1 antagonists;
MCH-R1 antagonists; AR129330.
*
Corresponding author. Tel.: +81 48 669 3064; fax: +81 48 652
7254; e-mail: yoshi.sekiguchi@po.rd.taisho.co.jp
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.052

2566
K. Kanuma et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2565–2569
O
N
N
N
N
N
H
Br
N
H
H
N
S
F
HCl
O O
OCF3
T-226296
AR129330
F
MCH IC50 = 160 nM
Y5 IC50 = 2.7 nM
α2A IC50 = 7.7 nM
F
O
MeO2C
MeO
Figure 2. Quinazoline lead structure from HTS.
N
N
H
N
H
N
N
H
O
The quinazoline cores 3a–c were synthesized in two
steps from commercially available 1H,3H-quinazoline-
2,4-dione 1. The starting material was reacted with
phosphorous oxychloride (POCl3) under reflux in the
presence of N,N-dimethylaniline to provide 2,4-di-
chloro-quinazoline 2. Selective substitution of the
chlorine at the 4-position with 50% aqueous dimethyl-
amine or 28% aqueous ammonia gave the
corresponding 4-amino-2-chloro-quinazoline 3a or 3b.
Alternatively, the chlorine at the 4-position could be
selectively reduced by activated zinc to yield 2-chloro-
quinazoline 3c.12 The mono-protected trans-1,4-cyclo-
hexylmethyldiamine 6 was prepared in six steps from
commercially available tranexamic acid 4. The carbox-
ylic acid was reduced to alcohol 5 by NaBH4 via the
mixed acid anhydride after the amine was protected as
its tert-butyl (BOC) carbamate. Tosylation of the alco-
hol followed by azidation gave an intermediate azide,
which was further converted into the amine 6 by reduc-
tion with lithium aluminum hydride. Coupling of 3a–c
with 6 was accomplished upon reflux in isopropanol to
afford coupling products 7a–c, respectively. Deprotec-
tion of the BOC group was achieved with hydrogen
chloride to provide amines as precursors for target quin-
azoline derivatives 8a–c. The amines were coupled to an
appropriate sulfonyl chloride to afford the desired qui-
nazoline sulfonamides 8a–c.
O
SNAP-7941
Figure 1. Representative non-peptide MCH-R1 antagonists.
have been previously well documented and others which
have been identified by proprietary pharmacophore
mapping algorithms. In addition, we have designed
and synthesized libraries based on specific core struc-
tures selected from a broad range of known ligands
for GPCRs, with the focus on class 1 GPCRs. One such
sub-library was built around the known NPY Y5 antag-
onist series of quinazolines, as exemplified by CGP-
71683A. Using a fluorescence-based assay to measure
inhibition of MCH-induced calcium flux in HEK-293
cells stably expressing a constitutively activated mutant
on the human MCH-R1 (hMCH-R1), we identified a
series of 4-(dimethylamino)quinazolines and 4-(methyl-
amino)quinazolines from this Y5-like sub-library as func-
tional MCH-R1 antagonists. Our most potent initial
screening hit was 4-bromo-N-{[trans-4-({[4-(dimethyla-
mino)quinazolin-2-yl]amino}methyl)cyclohexyl]methyl}-
2-(trifluoromethoxy)benzenesulfonamide hydrochloride
(AR129330), a resynthesized and purified sample dem-
onstrating an IC50 value of 160 nM in our assay. This
is by no means the first example of using families of
GPCR antagonist ligands to hop from activity at one
GPCR to another, and this initial success strengthens
the case for screening of such target class directed li-
braries as an approach to accelerate the hit detection
phase of drug discovery. Although AR129330 lacked
selectivity against a panel of other GPCRs, in particular
a2A adrenergic receptor and, unsurprisingly, Y5, we
used AR129330 as our initial chemical starting point
for our discovery program targeting MCH-R1 antago-
nists (Fig. 2). In addition to the poor selectivity, the in
vitro metabolic stability of most members of the initial
hit series was found to be moderate at best and these
two properties were identified as key areas in need of
improvement before we would consider this a lead ser-
ies. Our initial efforts to develop structure–activity rela-
tionships (SAR) at the MCH-R1 based on the 4-
(dimethylamino)quinazoline series and to improve selec-
tivity are described in this communication.
Throughout our hit-to-lead and lead optimization pro-
gram, we employed a combination of two assays to as-
sess potency of new analogues at the hMCH-R1. In
each case, we used a constitutively activated (CART)
form of the receptor mutated at the third intracellular
loop of the WT-MCH-R1 and measured competitive
inhibition of binding of [125I](Phe13, Tyr19)MCH or
transient intracellular calcium-mobilization evoked by
the MCH agonist in HEK293 cells stably expressing
the CART-MCH-R1.13 We used the former to compare
binding to MCH-R1 with binding data at other recep-
tors to enable us to determine selectivity, whereas the
latter we used to measure functional antagonist potency.
We found that the CART form of the receptor provided
better signal-to-noise ratios in these assays than the
wild-type hMCH-R1 transfected into the same cell line.
However, to verify that the mutation would not signifi-
cantly impact the activity of compounds at the WT-
MCH-R1, we compared data from a number of our quin-
azoline analogues across two assay platforms with both
A representative scheme for the preparation of quinazo-
line derivatives described herein is shown in Scheme 1.

K. Kanuma et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2565–2569
2567
X
N
O
Cl
b: 3a
c: 3b
a
N
NH
N
d: 3c
Cl
N
H
1
O
N
Cl
3a X = NMe2
3b X = NH2
3c X = H
2
HO2C
g
f
e
HO
H2N
NH2
NHBOC
NHBOC
4
5
6
X
N
X
N
h
N
N
N
H
N
H
H
N
R
NHBOC
S
O O
8a X = NMe2
8b X = NH2
8c X = H
7a X = NMe2
7b X = NH2
7c X = H
Scheme 1. Representative synthetic route of quinazolines. Reagents and conditions: (a) POCl3, N,N-dimethylaniline, reflux, 7 h, 86%; (b) aq 50%
Me2NH, THF, room temp., 80 min, 94%; (c) aq 28% NH3, room temp., 28.5 h, 96%; (d) activated Zn, aq 9% NH3 saturated with NaCl, CH2Cl2,
reflux, 3 h, 70%; (e) (i) (BOC)2O, aq 1.3 M NaOH, tert-BuOH, room temp, 18 h, 67%, (ii) ClCO2Et, Et3N, CH2Cl2, 0 °C, 50 min, (iii) NaBH4, THF,
MeOH, 4 °C, 3 h, 84% in two steps; (f) (i) TsCl, pyridine, 4 °C to room temp, 15 h, (ii) NaN3, DMF, 50 °C, 4 h, 91% in two steps, (iii) LAH, THF,
room temp., 6 h, 91%; (g) isopropanol, reflux; (h) (i) 4 M HCl in EtOAc, EtOAc, room temp, 1.5 h, (ii) RSO2Cl, isoPr2NEt, CH2Cl2, 4 °C or room
temp.
the CART-MCH-R1 and WT-MCH-R1. The IC50 val-
ues of the lead compound AR129330 in the receptor
binding assay and intracellular calcium-mobilization
assay utilizing the CART-MCH-R1 were 160 9.0 nM
(n = 8) and 140 10 nM (n = 6), respectively, whereas
the IC50 in the receptor binding assay using the WT-
MCH-R1 was 200 38 nM (n = 3). From these data,
we were satisfied that measuring binding at the
CART-MCH-R1 would provide a meaningful SAR-
driving assay for our medicinal chemistry approaches
and downstream spot checks, where we never observed
statistically significant differences in IC50 values in the
CART- and WT-MCH-R1 assays, further validated this
assumption.
10 lM. Simultaneously, we attempted to replace the ter-
minal phenyl ring with linear and branched alkyl
groups. Replacement with either methyl substituent 11
or isopropyl 12 was found to be completely lacking for
MCH-R1 activity whilst retaining significant affinity
for the Y5 and the a2A receptors. Likewise, butyl substi-
tuent 13 and octyl substituent 14 markedly decreased
MCH-R1 activity compared to AR129330. These early
results indicated that both 4-amino-substitution of the
quinazoline and the terminal phenyl group are critical
for MCH-R1 activity and form two key parts of the
MCH-R1 pharmacophore with the trans-1,4-cyclohex-
ane–dimethyleneamine serving to arrange the two phar-
macophoric groups into three-dimensional space. We
expected that chemical modification of the length or
conformation of the spacer between the quinazoline ring
and the sulfonamide linker could be used to optimize
activity at the receptor by more closely defining the
required spatial arrangement. This hypothesis then, led
us to focus our next series of investigations around this
area of our hit series.
We initially examined the effect of changes in the dime-
thylamino group at the 4-position of the quinazoline as
shown in Table 1. With some SAR information already
in hand from screening of our directed library, we knew
that substituents larger than dimethylamino in the qui-
nazoline 4-position were very poorly tolerated so we
were able to focus our efforts around very conservative
changes in this region of the molecule. Compound 9,
in which the dimethylamino group was replaced by an
amino group, was approximately 12-fold less potent at
the MCH-R1 than AR129330, and still showed very
high affinity for the Y5 (IC50 for Y5 = 1.6 nM). Com-
pound 10 with no substituent at the 4-position was de-
void of activity for the MCH-R1 at a concentration of
Next we examined the effects of altering the spacer por-
tion upon MCH-R1 activity and the selectivity over the
Y5 and the a2A receptors. Shortening the length of the
trans-1,4-cyclohexane spacer by removing one or both
methylene groups significantly reduced MCH-R1 affinity
(e.g., 15: m = 1, n = 0; 16: m = 0, n = 1; 17: m = 0, n = 0).
Changing the relative orientation of the substituents

2568
K. Kanuma et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2565–2569
Table 1. In vitro data of quinazoline derivatives
X
X
N
N
N
m
n
N
N
H
N
H
H
nN
R
N
R
S
S
O
O
O O
I
II
a
IC50 (nM)
Compound
Structure
Conformation
m
n
X
R
MCH-R1
Y5
a2A
CGP71683A
AR129330
I
trans
trans
trans
trans
trans
trans
trans
trans
trans
trans
trans
cis
1
1
1
1
1
1
1
1
1
0
1
0
1
0
1
0
1
0
NH2
NMe2
NH2
1-Naphthyl
3500
160
1.3
2.7
97
I
1
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
CH3
7.7
9
I
1
1900
> 10000
> 10000
> 10000
2900
6700
810
1.6
100
140
86
10
11
12
13
14
15
16
17
18
19
20
21
22
23
I
1
H
78
41
NEb
I
1
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
NMe2
I
1
isoC3H7
C4H9
C8H17
NE
58
I
1
4.9
NE
2.5
I
1
NE
21
I
1
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
4-Br-2-OCF3-Ph
I
0
1500
2200
2900
2700
140
150
31
110
47
I
0
I
1
NE
NE
120
880
3000
8900
NE
NE
130
57
I
cis
1
I
cis
0
I
cis
—
0
83
II
II
—
—
6000
250
5.6
170
—
a Compounds were evaluated for their ability to compete with [125I](Phe13, Tyr19)MCH, [125I]PYY, or [3H]MK912 at a human CA-MCH-R1 stably
expressed in HEK293 cells or a human Y5 or a2A transiently expressed in COS-1 cells. Data represent the mean of 1–3 separate experiments
performed from 5 concentrations and in duplicate.
b Not evaluated.
across the cyclohexyl ring of AR129330 from trans to cis
provided compound 18, which resulted in an 18-fold de-
crease in activity. Similarly 4-aminomethyl-cis-cyclohex-
ylamine derivative 19 showed significantly reduced
affinity for the MCH-R1. Compound 20, in which the
spacer used in 19 was reversed, exhibited comparable
affinity to AR129330 but still lacked selectivity over the
other receptors. Use of the shorter cis-1,4-cyclohexyldi-
amine spacer, as in 21 however, resulted in a 2-fold in-
crease in activity relative to AR129330 and more
importantly provided our first hint of selectivity over
the Y5 receptor. Further shortening of the linker by
replacement with piperidine rings led to divergent selec-
tivities. Compound 22 showed very potent affinity for
the a2A receptor (IC50 for a2A = 5.6 nM) and was signif-
icantly less active at both the MCH-R1 and the Y5 recep-
tor. In contrast, compound 23 maintained MCH-R1
activity and showed remarkably decreased Y5 activity
compared to AR129330. These data emphasize the
importance of the nature, ring conformation, and length
of the spacer to the MCH-R1 activity.
tion of the spacer portion between the quinazoline ring
and the sulfonamide linker resulted in the identification
of compounds 20, 21, and 23, which in next investiga-
tion led us to the finding of other series involving oral
active ATC0175.
References and notes
1. Vaughan, J. M.; Fischer, W. H.; Hoeger, C.; Rivier, J.;
Vale, W. Endocrinology 1989, 125, 1660–1665.
2. Kolakowski, L. F., Jr.; Jung, B. P.; Nguyen, T.; Johnson,
M. P.; Lynch, K. R.; Cheng, R.; Heng, H. H. Q.; George,
S. R.; OÕDowd, B. F. FEBS Lett. 1996, 398, 253–258.
3. Chambers, J.; Ames, R. S.; Bergsma, D.; Muir, A.;
Fitzgerald, L. R.; Hervieu, G.; Dytko, G. M.; Foley, J. J.;
Martin, J.; Liu, W. S.; Park, J.; Ellis, C.; Ganguly, S.;
Konchar, S.; Cluderay, J.; Leslie, R.; Wilson, S.; Sarau, H.
M. Nature 1999, 400, 261–265.
4. Qu, D.; Ludwig, D. S.; Grammeltoft, S.; Piper, M.;
Pelleymounter, M. A.; Cullen, M. J.; Mathes, W. F.;
Przypek, J.; Kanarek, R.; Maratos-Flier, E. Nature 1996,
380, 243–247.
5. Rossi, M.; Choi, S. J.; OÕShea, D.; Miyoshi, T.; Ghatei, M.
A.; Bloom, S. R. Endocrinology 1997, 138, 351–355.
6. Shimada, M.; Tritos, N. A.; Lowell, B. B.; Flier, J. S.;
Maratos-Flier, E. Nature 1998, 396, 670–674.
7. Gonzalez, M. I.; Vaziri, S.; Wilson, C. A. Peptides 1996,
17, 171–177.
In summary, high-throughput screening of our in-house
GPCR-directed library detected AR129330 as a poten-
tial chemical starting point for the discovery of MCH-R1
antagonists. Through the preliminary SAR investigation
we demonstrated that both 4-amino-substitution of the
quinazoline and the terminal phenyl group were critical
for MCH-R1 activity. Subsequently, further optimiza-
8. Monzon, M. E.; De Barioglio, S. R. Physiol. Behav. 1999,
67, 813–817.

K. Kanuma et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2565–2569
2569
9. Borowsky, B.; Durkin, M. M.; Ogozalek, K.; Marzabadi,
M. R.; Deleon, J.; Heurich, R.; Lichtblau, H.; Shaposh-
nik, Z.; Daniewska, I.; Blackburn, T. P.; Branchek, T. A.;
Gerald, C.; Vaysse, P. J.; Forray, C. Nat. Medicine 2002,
8, 825–830.
11. Lagu, B.; Wetzel, J.; Marzabadi, M.; Deleon, J. E.;
Gluchowski, C.; Noble, S.; Nagarathnam, D.; Chiu, G.
International Patent 6245, 2002.
12. William, G.; Ashit, G.; Jan, W. F. W. U.S. Patent
6040448, 2000.
10. Kato, K.; Terauchi, J.; Mori, M.; Suzuki, N.; Shimomura,
Y.; Takekawa, S; Ishihara, Y. International Patent WO
21577, 2001.
13. Sekiguchi, Y.; Kanuma, K.; Omodera, K.; Tran, T. A.;
Kramer, B. A.; Beeley, N. R. A. International Patent WO
28641, 2003.