2-Aryl-3,6-dialkyl-5-dialkylaminopyrimidin-4-ones as novel CRF-1 receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(03)00483-9

The discovery, synthesis and structure-activity studies of a novel series of 2-arylpyrimidin-4-ones as CRF-1 receptor antagonists is described. These compounds are structurally simple and display appropriate physical properties for CNS agents

Full text of DOI:10.1016/S0960-894X(03)00483-9

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2497–2500
2-Aryl-3,6-dialkyl-5-dialkylaminopyrimidin-4-ones as Novel
CRF-1 Receptor Antagonists
Kevin J. Hodgetts, Taeyoung Yoon, Jianhua Huang, Michael Gulianello, Andrzej Kieltyka,
Renee Primus, Robbin Brodbeck, Stephane De Lombaert and Darıo Doller*
Neurogen Corporation, 35 NE Industrial Road, Branford, CT 06405, USA
Received 5 April 2003; revised 6 May 2003; accepted 7 May 2003
Dedicated to the memory of Prof. Eduardo G. Gros and Prof. Alicia M. Seldes
Abstract—The discovery, synthesis and structure–activity studies of a novel series of 2-arylpyrimidin-4-ones as CRF-1 receptor
antagonists is described. These compounds are structurally simple and display appropriate physical properties for CNS agents
# 2003 Elsevier Ltd. All rights reserved.
In spite of numerous marketed treatments, anxiety and
depression are psychiatric disorders which constitute a
major health concern. Agents with increased efficacy or
a reduced side-effects profile are actively sought by
pharmaceutical companies. Corticotropin-releasing fac-
tor (CRF) is a 41-aminoacid peptide characterized as
the principal mediator of the effects of stress via the
hypothalamic-pituitary-adrenocortical (HPA) axis.
CRF receptors are 7-transmembrane Gas-protein cou-
pled receptors. They are classified in two subtypes,
CRFR-1 and CRFR-2; these two share ca. 70%
homology. Extensive biological research led to the
hypothesis that CRF-1 receptor antagonists might be
useful in treating anxiety and depression.1 The conclu-
sions froman open-label clinical study involving
R121919, a small-molecule CRF-1 receptor antagonist,
were reported recently. The outcome was considered
encouraging as R121919 produced a decrease in symp-
toms of both depression and anxiety.2 Eventually, its
clinical development was discontinued, seemingly due to
elevation of liver enzymes. Other CRF-1 antagonists
(structures not disclosed) are reportedly undergoing
clinical trials at the present time.
particularly considering the compounds which reached
advanced preclinical development or clinical stage (e.g.,
topology 1: CP-154,526, R121919, DMP 904, NGD
98-1; topology 2: SSR125543A; Fig. 1). In our studies
we set out to explore the possibility of identifying novel
small-molecule CRF-1 receptor antagonists. Specifi-
cally, our strategy was to seek an appropriate linker to
replace the bicyclic or tricyclic cores found in most
reported antagonists. In addition, we envisioned that
such a linker would have the potential for decreasing
the molecular weight relative to existing antagonists.
Herein we report our work leading to the discovery of
2-aryl-3,6-dialkyl-5-dialkylaminopyrimidin-4-ones (e.g.,
1) as CRF-1 receptor antagonists.
Planning for an appropriate central scaffold we con-
sidered the following facts. Prior structure–activity
relationship studies for bicyclic and tricyclic CRF-1
receptor antagonists had demonstrated that certain
structural features are needed for small-molecule high-
affinity binding to the receptor.1 First, an out-of-plane
relationship between the bottomaromatic ring and the
heterocyclic core is key. Second, at least one nitrogen
atomis required on the central ring, presumably playing
the role of hydrogen bond acceptor. The position of this
nitrogen atomis critical for activity. Two topologies for
CRFR-1 antagonists have been proposed. These differ
in the number of atoms linking the bottom aromatic
group and this essential nitrogen atomon the central het-
erocycle.1 Findings fromour preliminary studies were in
agreement with this model (Fig. 2). We decided to keep
the top portion as a di-n-propylamino group for our
A number of CRF-1 receptor antagonists have been
reported. They can be classified into two topologies
based on distinct structural features.1 The diversity of
chemical matter among leading series is rather narrow,
*Corresponding author. Tel.: +1-203-488-8201x3114; fax: +1-203-
481-5290; e-mail: ddoller@nrgn.com
0960-894X/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00483-9

2498
K. J. Hodgetts et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2497–2500
stituted pyrimidones were typically inactive or much less
active than their N-alkyl analogues. N-ethyl substitution
on the pyrimidone ring provided a number of potent
analogues (e.g., 1, 6, 16, 25, 29). Further increasing the
size of the alkyl group or introducing solubilizing
groups (14, 26)9 resulted in significant loss of affinity.
Similarly, ethyl substitution on position 6 of the pyr-
imidone ring led to analogues with high affinity (1 and
13). While 2,4,6-trisubstitution on the aryl ring provided
some of the more potent analogues, it was not essential
for potency, as appropriate 2,4-disubstitution also
resulted in compounds of comparably high affinity (e.g.,
29). The presence of a methoxy group in one of the
ortho positions resulted in an increase in affinity with
respect to a methyl substituent (e.g., 4 vs 15, 6 vs 16).
The poor affinity of analogues bearing a hydrogen on
the para position underscores the importance of such
substitution (19, 20 and 22). These substituent effects on
affinity parallel those observed for a number of CRF-1
receptor antagonists.1
By appropriate tuning of the bottomphenyl ring and the
substitution on the pyrimidone core, potent CRF-1
antagonists were obtained. As expected, the required
introduction of alkyl groups on the pyrimidone core
resulted in an overall increase in lipophilic character and a
decrease in aqueous solubility at pH 7.4. Calculated Mlog
P and polar surface area (PSA) values were generally
within the range which predicts reasonably good mem-
brane permeability and brain penetration.10 Disappoint-
ingly, human microsomal half-lives for these compounds
were too short to predict appropriate plasma concentra-
tions over a useful period of time. To explore the potential
for oral plasma and brain exposure of this series a rat phar-
macokinetic study was carried out on a representative.
Compound 6 showed negligible plasma exposure following
oral administration in rats (F=2.6Æ0.5% at a 20 mg/kg
dose, n=3; Cmax=234Æ52 ng/mL, Tmax=0.6Æ0.4 h, T1/
2=3.3Æ0.3 h, Cl=16.6Æ9.5 mL/min/kg, Vd=11.5Æ0.2
L/Kg). The brain to plasma ratio was B/P=0.9Æ0.4 mL/g.
Figure 1. Structures of selected CRF-1 receptor antagonists.
Figure 2. Early analogues defining the need for a nitrogen atomon a
specific position of the central ring.
exploratory work, as it produced analogues with acceptable
affinity levels. While no significant binding wasobserved for
aniline 2 or pyrimidine3, pyrim idone4 did show affinity for
CRFR-1. This suggested that a nitrogen atomon a specific
location of the central ring is also needed for CRFR-1
binding in this new pyrimidone-based template, and that
the optimal position was that in 4 and not that in 3.
Noticing that some N-unsubstituted pyrimidin-4-ones
(5, 12, 17, 18, 21, 23) had more-acceptable microsomal
stabilities than their N-Me or N-Et counterparts, we
speculated that CYP-450-mediated N-dealkylation
might be the major metabolic event leading to short
half-lives. Unfortunatelly, attempts at increasing micro-
somal half-lives by blocking this dealkylation process
through fluorination or introduction of heteroatoms on
the N-alkyl substituent group did not prove successful.
These compounds were found to be much weaker (9, 14,
26) or reasonably potent but still metabolically unstable
(10). This can be explained by the existence of several
metabolic routes for these compounds.
After defining a novel CRF-1 receptor binding tem-
plate, the next step became trying to improve its binding
affinity towards the receptor. Results fromour struc-
ture–activity relationship studies are summarized in
Table 1. Affinities were determined by examining the
displacement of [125I]-sauvagine fromCRF-1 receptors
endogenously expressed in IMR-32 human neuro-
blastoma cells.6,7 Functional antagonismof this series
was established in a secondary assay, measuring inhibi-
tion of sauvagine-stimulated cAMP accumulation in
AtT20 cells which endogenously express CRFR-1.6,8
Compounds screened in this functional assay had Ki
values between 10 and 100 nM (Table 1). These values
tracked reasonably well with the binding affinity.
Analogue Syntheses
Pyrimidin-4-ones reported in Table 1 were synthesized
starting fromcompound 30 as exemplified in Scheme 1
for analogue 1. Nitration of commercially available
2,4-dihydroxy-6-methylpyrimidine 30 yielded nitropyr-
imidine 31, which was stepwise converted to chloropyr-
Conclusions fromour SAR studies are as follows. In
terms of affinity for the CRF-1 receptor, N-unsub-

K. J. Hodgetts et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2497–2500
2499
Table 1. CRF-1 receptor affinity, microsomal stability and physical properties for compounds 1 and 4–29
Compd R1
R2
Ara CRF-1 Ki,
Microsomal Solubility PSA, MlogPe AtT20
T1/2, m in
Definitions
c
nMb
(mg/mL)d
A2e
Ki, nMf
4
5
6
7
8
9
10
11
12
13
1
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Me
Me
Me
Me
Me
Me
Me
Me
Et
Me
H
Et
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
C
C
C
D
D
E
E
E
E
E
E
F
90
930
21
190
47
1630
39
IA
560
7
4
IA
135
72
IA
IA
530
310
IA
620
IA
470
38
4000
720
260
15
2.5
15.5
3.6
3.0
2.1
3.4
3.0
2.2
8.1
3.3
2.3
4.2
4.6
3.6
7.5
13.3
1.8
1.0
21.6
1.8
9.4
6.0
3.0
1.3
2.3
1.5
1.9
3.3
10.3
1.3
<0.1
<0.1
<0.1
<0.1
0.4
3.8
2.5
<0.1
25.2
2.2
0.9
0.4
23.6
6.5
20.5
ND
<0.1
7.9
41.5
17.5
n/a
41.2
53.2
40.4
36.3
28.9
35.5
36.3
34.9
52.4
39.9
35.3
31.8
29.6
26.2
40.6
65.0
52.6
48.2
64.5
47.8
60.2
53.7
52.3
64.5
52.3
50.2
29.4
3.30
3.08
3.52
3.74
3.74
3.84
3.63
4.33
3.30
3.52
3.74
4.09
3.82
4.04
3.82
2.35
2.57
2.80
2.69
3.12
2.35
2.57
2.80
2.26
3.02
3.23
4.36
26
n-Pr
iPr
CH2CF3
CH2CH2F
Bn
90
7
H
Me
Et
Et
Et
Me (CH2)2N(CH2)4
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
H
H
Me
Et
H
Et
H
Me
Et
(CH2)2OMe
nPr
CH2(CH2)3
Et
68
36
4.9
20.1
0.9
aFor Ar group definitions see Table 1.
bAccording to ref 7. Ki values are means of at least two determinations; standard errors lower than 25%. IA=inactive. ND=not determined.
cHuman microsomes.
dAt pH 7.4.
eValues calculated using AIDD software.11
fFunctional assay. See refs 6 and 8.
imidine 33. Suzuki coupling of 33 under standard condi-
tions afforded nitropyrimidine 34. Reduction to aminopyr-
imidine 35 by catalytic hydrogenation at atmospheric
pressure (balloon), followed by reductive amination yielded
quantitatively dialkylaminopyrimidine 36. Alkylation of
the acidic m ethyl group on the pyrim idine ring was accom -
plished by treatment with a slight excess of lithium diiso-
propylamide followed by quenching with methyl iodide.
Hydrolysis of the methoxypyrimidine to pyrimidone 37
under acidic conditions, followed by base-promoted
alkylation produced N-ethyl pyrimidone 1. Small
amounts of the isomeric 4-methoxypyrimidines were
reaction byproducts.
Synthesis of probe compounds 2 and 3 was carried out
as outlined in Schemes 2 and 3. For the case of 2,
starting out with commercially available 4-iodo-2,6-
dimethylaniline 38, reductive amination to dialkyla-
mine 39, followed by Suzuki coupling yielded bipheny-
lamine 2. For pyrimidine 3 (Scheme 3) the starting
material was compound 40. Reductive amination fol-
lowed by bromination produced bromide 41.
Scheme 1. Reagents and conditions: (a) KNO3, H2SO4 (75%); (b)
OPCl3 PhNEt2, reflux (60%); (c) MeONa, MeOH (35%); (d)
Pd[(PPh)3]4, 2,4-Me2-6-(MeO)C6H2B(OH)2, NaHCO3, toluene, 70ꢀC
(65%); (e) H2, Pd/C (100%); (f) Propionaldehyde, AcOH, NaBH(OAc)3,
1,2-dichloroethane, rt (100%); (g) (1) LDA, THF, À78ꢀC. (2) MeI,
À78ꢀC to rt (75%); (h) HCl 1M (100%); (i) EtI, NaH, THF, rt (50%).
Scheme 2. Reagents and conditions: (a) Propionaldehyde, AcOH,
NaBH(OAc)3, 1,2-dichloroethane (100%); (b) Pd[(PPh)3]4, 2,4-Me2-6-
(MeO)C6H2B(OH)2, NaHCO3, toluene, 70 ꢀC (35%).

2500
K. J. Hodgetts et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2497–2500
References and Notes
1. For recent reviews, see: (a) Holsboer, F. Curr. Opinion
Invest. Drugs 2003, 4, 46. (b) Kehne, J.; De Lombaert, S. Curr.
Drug Targets 2002, 1, 467. (c) Lanier, M.; Williams, J. P.
Expert Opin. Ther. Patents 2002, 12, 1619. (d) Grigoriadis,
D. E.; Haddach, M.; Ling, N.; Saunders, J. Curr. Med. Chem.
2001, 1, 63. (e) Gilligan, P. J.; Robertson, D. W.; Zaczek, R.
J. Med. Chem. 2000, 43, 1641.
2. Zobel, A. W.; Nickel, T.; Kunzel, H. F.; Ackl, N.;
Sonntag, A.; Ising, M.; Holsboer, F. J. Psych. Res. 2000,
34, 171.
Scheme 3. Reagents and conditions: (a) Propionaldehyde, AcOH,
NaBH(OAc)3, 1,2-dichloroethane (70%). (b) Br2, CHCl3, rt, 95%. (c)
Pd[(PPh)3]4, 2,4-(MeO)2C6H3B(OH)2, NaHCO3, toluene, 70 ꢀC (15%).
3. Webster, E. L.; Lewis, D. B.; Torpy, D. J.; Zachman, E. K.;
Rice, K. C.; Chrousos, G. P. Endocrinology 1996, 137, 5747.
4. He, L.; Gilligan, P. J.; Zaczek, R.; Fitzgerald, L. W.; Mc
Elroy, J.; Shen, H. S. L.; Saye, J. A.; Kalin, N. H.; Shelton, S.;
Christ, D.; Traimor, G.; Hartig, P. J. Med. Chem. 2000, 43,
449.
5. Gully, D.; Geslin, M.; Serva, L.; Fontaine, E.; Roger,
P.; Lair, C.; Dane, V.; Marcy, C.; Rouby, P. E.;
Simiand, J.; Guitard, J.; Gout, G.; Steinberg, R.; Rodier,
D.; Griebel, G.; Soubrie, P.; Pascal, M.; Pruss, R.; Scat-
ton, B.; Maffrand, J. P.; Le Fur, G. J. Pharmacol. Exp.
Ther. 2002, 301, 322. Griebel, G.; Simiand, J.; Steinberg, R.;
Jung, M.; Gully, D.; Roger, P.; Geslin, M.; Scatton, B.; Maf-
frand, J. P.; Soubrie, P. J. Pharmacol. Exp. Ther. 2002, 301,
333.
Suzuki coupling yielded final pyrimidine 3, albeit in
low overall yield.
In summary, we have described the discovery, synthesis
and structure–affinity relationship study of a new series of
CRF-1 receptor antagonists. These are structurally simple
and have physical properties consistent with CNS-active
drugs. At this stage, their unacceptable pharmacokinetic
profile, presumably due to low microsomal stability or
poor absorption, rendered this series of compounds
unsuitable for further preclinical development. However,
it can be expected that related series could be found with
similar pharmacophoric elements to those reported herein
and possessing improved permeability characteristics and/
or metabolic stability. The results of our continuing efforts
to solve these issues will be reported in due course.
6. Yuan, J.; Gulianello, M.; De Lombaert, S.; Brodbeck, R.;
Kieltyka, A.; Hodgetts, K. J. Bioorg. Med. Chem. Lett. 2002,
12, 2133.
7. Grigoriadis, D. E.; De Souza, E. B. Methods Neurosci.
1991, 5, 510.
8. Chaki, S.; Okuyama, S.; Nakazato, A.; Kumagai, T.;
Okubo, T.; Ikeda, Y.; Oshida, Y.; Hamajima, Y.; Tomisawa,
K. Eur. J. Pharmacol. 1999, 371, 205.
Acknowledgements
9. Wermuth E. G. In The Practice of Medicinal Chemistry.
Wermuth, E. G., Ed. Academic Press: New York, 1996;
chapter 35, pp 755–776.
10. Atkinson, F.; Cole, S.; Green, C.; van de Waterbeemd, H.
Curr. Med. Chem-CNS Agents 2002, 2, 229.
11. http://www.neurogen.com/technology.htm
We are grateful to Dr. Jan Wasley and Dr. Ray-
mond Horvath for their support during this work.
We also thank Mrs. Marta Day for solubility
determinations and our DMPK department for PK
determination of compound 6.