Design, synthesis and evaluation of constrained tetrahydroimidazopyrimidine derivatives as antagonists of corticotropin-releasing factor type 1 receptor (CRF1R)

Article abstract of DOI:10.1016/j.bmcl.2010.01.127

Several tetrahydroimidazopyrimidines were prepared using silver assisted cyclization as the key step. The binding affinities of compounds thus prepared were evaluated in vitro toward hCRF₁R. Initial lead compound 16 (K_i = 32 nM) demo

Full text of DOI:10.1016/j.bmcl.2010.01.127

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1905–1909
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and evaluation of constrained tetrahydroimidazopyrimidine
derivatives as antagonists of corticotropin-releasing factor type 1 receptor
(CRF1R)
*
Vivekananda M. Vrudhula , Bireshwar Dasgupta, Sokhom S. Pin, Kevin D. Burris, Lynn A. Balanda,
Lawrence K. Fung, Tracey Fiedler, Kaitlin E. Browman, Matthew T. Taber, Jie Zhang, John E. Macor,
Gene M. Dubowchik
Bristol-Myers Squibb Research & Development, 5 Research Parkway, Wallingford CT 06492, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Several tetrahydroimidazopyrimidines were prepared using silver assisted cyclization as the key step.
The binding affinities of compounds thus prepared were evaluated in vitro toward hCRF1R. Initial lead
compound 16 (Ki = 32 nM) demonstrated modest putative anxiolytic effects in the mouse canopy test.
Further optimization using parallel synthesis provided compounds with Ki’s <50 nM.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 25 November 2009
Revised 26 January 2010
Accepted 28 January 2010
Available online 4 February 2010
Keywords:
Tetrahydroimidazopyrimidine
CRF1R
CRF Anxiolytic
The hypothalamus, which is located at the base of the brain,
functions as a control center for the neuroendocrine system.1 In re-
sponse to stress the hypothalamus increases the production of a 41
amino acid neurohormone, Corticotropin Releasing Factor (CRF).2
CRF regulates the hypothalamic-pituitary-adrenal axis (HPA axis)
and is involved in coordinating the endocrine as well as autonomic
and behavioral responses of the organism to stress.3 CRF exerts its
effects by binding to and activating class 2 G protein-coupled
receptors. There are two subtypes of CRF receptors, namely, CRF1
Several research groups have successfully demonstrated the
efficacy of CRF1R antagonists in in vivo models of anxiety and
depression.5a–f Examination of a number of such non-peptide
chemotypes (Fig. 1) led to pharmacophoric models and identifica-
tion of certain structural requirements for antagonistic activity as
depicted in Figure 1.5c,10 Some key features of these mono or bicy-
clic systems containing an sp2 nitrogen include: (a) an orthogonal
ring system capable of interacting with a lipophilic pocket of the
receptor; (b) a lipophilic side-chain at the top; and (c) a short alkyl
group on the core as illustrated in Figure 1.
and CRF2, with three splice variants CRF2a, CRF2b and CRF2c for
the CRF2 receptors. These receptors exhibit differential expression
and pharmacology.4 The potential for CRF1R as a target for the
development of therapeutics for anxiety and depression has been
reviewed extensively in literature5a–f and stems from several
observations. For example, intracavernous administration of CRF
resulted in stimulation of ACTH, release of corticosterone, and an
increase in anxiety-like behavior in rats and primates.6 Further-
more, analysis of cerebrospinal fluid (CSF) of untreated patients
suffering from depression indicated elevated levels of CRF. Normal-
ization of CRF levels was achieved following the treatment with
antidepressants.7a,b Finally, mice over-expressing CRF display anx-
ious and depressive-like behavior8 while CRF1R gene knock out
mice demonstrate reduced anxiety.9
We sought to design and synthesize novel chemotypes incorpo-
rating such recognition elements and examine their activity as
antagonists of CRF1R. In this communication we describe the syn-
thesis and evaluation of constrained tetrahydroimidazopyrimi-
dines and dihydroimidazopyrimidinones as CRF1R antagonists.
Within the scope of these two chemotypes, the synthesis and eval-
uation of three classes of compounds viz., ‘bis-amides’, C3-amides
and C3-amines as antagonists of CRF1R is described.
The three general targets for synthesis are shown in Scheme 1.
These are C-3 tertiary amides containing Me, Et and CF3 substit-
uents at the 2-position on the imidazole ring, C-3 tertiary amines
containing the electron-withdrawing CF3 substituent at C-2 of the
imidazole ring and finally bis-amides bearing a C-3 amide side-
chain and a 7-oxo group in the core. The amine targets in Scheme
1 were expected to be susceptible to elimination of the amine
side-chain and therefore were prepared with an electron-with-
* Corresponding author. Fax: +1 203 677 7702.
E-mail address: vivekananda.vrudhula@bms.com (V.M. Vrudhula).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.127

1906
V. M. Vrudhula et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1905–1909
NEtBu
N
large
N
lipophilic cavity
- HCl
N
N
N
N
Br
side chain
XQ046
CP-154526-1
Rat CRF Ki= 46 nM
hCRF1 Ki = 1 nM
(Pfizer)
Mono or
short
bicyclic
Core with
sp2 N
alkyl
group
NPr2
N
S
N
essential
small
N
N
orthogonal
lipophilic
pocket
arylgroup
N
Cl
N
Hydrophobic
pocket
Cl
NMe2
YY832 (Sanofi)
R121919
Rat CRF Ki = 8 nM
hCRF1 Ki = 2 nM
(Neurocrine - Janssen)
Figure 1. Key structural features of non-peptide CRF antagonists.
R2
N
R3
Tetrahydroimidazopyrimidine Amide Targets
R2 = R3 = Alkyl groups, Q = O, Y = H2, R = Me, Et, CF3
Tetrahydroimidazopyrimidine Amine Targets
R2 = R3 = Alkyl groups Q = Y = H2 R = CF3
5
Q
R
3
N
2
7
8
Y
N
N
1
Dihydroimidazopyrimidinone Bis-amide Targets
R2 = R3 = Alkyl groups, Q = O, Y = O, R = Me, Et
Scheme 1. Proposed targets for synthesis for evaluation as antagonists of CRF1R.
HN
HN
N
NH2
HN
Br
a, b
2
1
R
H
N
H
3
O
3
R
N
1
Br
HN
1
3
N
O
N
7
+
R
4
N
O
c
N
2
Br
5
1
5
N
O
4
H
O
Predominant
minor
O
3
a = R = CH3 =
b = R = C2H5 =
c = R = CF3
11
3
:
:
1
exclusive
O
O
R
N
Br
HN
N
d
R
O
N
N
N
O
4 a-c
6 a-c
Scheme 2. Reagents and conditions: (a) n-BuLi/dioxane–hexane/0 °C; (b) excess Br(CH2)3Br, 22% overall yield; (c) 2/DBU/acetone or Cs2CO3/DMF, 21–65% yield; (d) Ag2CO3 or
AgOTf/sulfolane/150 °C/20–60% yield.

V. M. Vrudhula et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1905–1909
1907
R1
O
O
N R2
O
R3
N
R3
a or
N
b, c, d
N
N
N
N
7-14
R1
6a-c
O
N
N R2
O
HO
N
N
N
N
F3C
F3C
f, g
F3C
e
N
N
N
N
6c
15
16 = R1 =
,
R2 = Pr
Scheme 3. Reagents and conditions: (a) R1NHR2/AlMe3/toluene/80 °C or; (b) LiOH/DME–H2O; (c) CF3COOC6F5/DMF, 60% yield of ester; (d) R1NHR2/Cs2CO3/DMF, 0–83% yield;
(e) LAH/THF, 97% yield; (f) SOCl2/DCE; (g) R1NHR2/MeCN.
drawing group at C-2. The dialkylamine reagents necessary for
the preparation of these targets are from a focused library previ-
ously reported from these laboratories.11 Our initial approach to
the synthesis of the 5,6-fused ring system was similar to the syn-
thesis of 5,5-fused ring system from cyclic guanidines reported
from our laboratories.12 Even though a cyclic guanidine such as
1 could be prepared from mesitylamine, subsequent elaboration
to final target was unsuccessful. It became necessary to devise
an alternative methodology for building such a fused system from
mesitylamine as shown in Scheme 2.
Intramolecular cyclization of suitably substituted bromoimi-
dazoles such as 4a–c appeared to be a viable approach to prepare
the desired 5,6-fused system. Alkylation of unsymmetrically
substituted bromoimidazole 3a–c can lead to the formation of
two N-alkylated regioisomers. The ratio of the two regioisomers
formed during the alkylation of an unsymmetrically substituted
O
O
H
H
N
H
N
Br
O
N
O
a
b
N
N
R
R
3 a-b
a = R = Me
b = R = Et
17a-b
O
N
O
O
O
N
OTBDMS
c, d
R
R
e
O
NH
N
N
OH
NH
19 a-b
18 a-b
O
O
R2
N
O
N
N
f
R
R1
N
N
O
N
R
O
N
20 a-b
21-30
Scheme 4. Reagents and conditions: (a) Mesitylamine/AgOTf/diglyme/150 °C, yield = 81% (R = Et) and 89% (R = Me); (b) Br(CH2)3OTBDMS/Cs2CO3/DMF, yield = 52% (R = Me)
and 65% (R = Et); (c) TBAF/THF, yield = 63% (R = Me) and 86% (R = Et); (d) Jones oxidation, yield = 28% (R = Me) and 62% (R = Et); (e) EDCÁHCl/DIEA/DCM, yield = 22% (R = Me)
and 53% (R = Et); (f) R1NHR2/AlMe3/toluene/80 °C, 6–24% yield.

1908
V. M. Vrudhula et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1905–1909
Table 1
Table 2
Binding affinities of amides and bis-amides toward hCRF1R
Binding affinities of amines toward hCRF1R
R1
O
R1
N
N
R2
R2
N
N
R3
F3C
N
N
N
Q
N
H
N
H
Compd
R3
Q
Ki (nM)
N
Compd
Ki (nM)
R1
R2
R1
R2
Pr
N
H
7
8
Me
Me
H2
H2
3400
3300
N
H
16
31
32
8
Pr
N
H
N
H
CF3
CF3
H
N
H
N
32
14
9
Me
H2
1700
Pr
F
F
N
H
Pr
33
34
35
36
19
21
27
31
Ph
N
H
10
11
12
13
14
21
22
23
Me
Me
Me
Et
H2
H2
H2
H2
H2
O
>5000
>5000
>5000
2300
Pr
N
H
Pr
Pr
Pr
Pr
N
H
N
H
CF3
N
H
Et
N
H
N
H
H
N
37
38
67
38
Pr
N
H
Pr
CF3
Me
Me
Me
1200
H
N
Pr
CF3
N
H
>5000
>5000
>5000
N
H
O
Pr
N
H
39
40
48
63
Pr
N
H
O
H
N
F3C
Pr
F
NH
24
Me
O
>5000
41
42
43
44
45
245
364
N
NH
NH
NH
NH
O
25
26
Me
Et
O
O
>5000
>5000
NH
1510
441
Pr
Et
O
S
N
N
H
Ph
Pr
N
27
28
29
30
Et
Et
Et
Et
O
O
O
O
>5000
>5000
>5000
>5000
H
>5000
NH
Pr
N
H
N
H
derived cyclization product as the minor. When the alkyl group
on the imidazole was an electron-withdrawing CF3 group, alkyl-
ation led to exclusive formation of the N1-alkylated isomer. While
the cyclization of the minor N3-alkylated material occurred readily
under the conditions of alkylation itself, the cyclization of N1-alkyl-
ated material did not. After exploring several conditions, silver ion
assisted cyclization in sulfolane at 150 °C led to successful forma-
tion of the desired tetrahydroimidazopyrimidine ring system. Con-
version of the vinylogous urethane to the C-3 amide derivatives
was carried out either by Weinreb amidation14 or via coupling of
the amines with the activated pentafluorophenyl ester (Scheme
3). Sterically hindered amines such as di-sec-butylamine and
amines containing electron-withdrawing groups such as
PhCH2NHCH2CF3 failed to couple in this activated ester coupling.
n-Bu
Et
N
H
imidazole can be influenced by the conditions of alkylation,13a,b
and the nature of the alkyl group on the imidazole ring. For alkyl-
ation of 3, the N1-isomer was isolated from the reaction mixture,
while the N3-isomer under the conditions of alkylation underwent
cyclization to form 5. During the alkylation of 3a–c with 2, the
composition of the two regioisomers formed was found to be
influenced by the nature of the alkyl group on the imidazole ring.
Electron-donating groups such as Me and Et led to the formation
of N1-alkylated isomer as the major product and N3-alkylation

V. M. Vrudhula et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1905–1909
1909
amines such as 41 and 42 or cyclic tertiary amines containing a
second hetero atom such as 43 and 44 displayed diminished activ-
ity. Introduction of a second basic site in the amine residue as in 45
led to a significant loss of binding affinity.
Compound 16 was evaluated in vivo as a prototype of the C-3
amine series. Intraperetoneal (ip) administration of 16 at a dose
50
40
30
20
10
0
50
40
30
20
10
0
of 32 mg/kg in
a vehicle containing cremophor/DMSO/water
(1:1:8) in mice showed that the compound achieved highest brain
penetration 30 min post administration (352 ng of 16/g of brain
with a brain/plasma ratio of 1.2). One hour post administration
the brain/plasma ratio for 16 was found to be 1.0 (217 ng of 16/g
of brain). The putative anxiolytic effects of 16 were evaluated in
a mouse canopy test as described previously.12,15 In this experi-
ment ip administration of 16 in a vehicle containing cremophor/
DMSO/water (1:1:8) demonstrated a dose-dependent reduction
of stretched attenuated postures, demonstrating anxiolytic effects
(Fig. 2).
*
*
Vehicle Buspar
16 @
16 @
16 @
16 mg/kg 32 mg/kg 64 mg/kg
Figure 2. Dose-dependent decrease in total stretched attend postures with ip
administration of 16 in mouse canopy test as a measure of putative anxiolytic
activity. Vehicle is DMSO/cremophor/water = 1:1:80. Asterisk indicates significant
difference from vehicle (p <0.05).
In summary, in this Letter we described the synthesis of novel
tetrahydroimidazopyrimidines and dihydroimidazopyrimidinones
as targets for evaluation as CRF1R antagonists. Silver assisted cycli-
zation was used as a methodology to prepare these constrained
imidazoles. Dihydromidazopyrimidinones (the bis-amides) were
inactive, and the tetrahydroimidazopyrimidine amides (C3-
amides) were only modestly active. Tetrahydroimidazopyrimidine
amines demonstrated significant CRF1R affinity. Numerous tetra-
hydroimidazopyrimidine amines demonstrated Ki’s <50 nM. In
vivo activity for 16 was demonstrated in the mouse canopy model,
demonstrating CRF1R antagonism in vivo. Further efforts to im-
prove the physiochemical properties of this series (i.e., reduce
the hydrophobic nature of these constrained imidazoles) while
maintaining their potency will be described in a future report.
The C-2 CF3, C-3-dialkylaminomethyl targets were prepared from
the vinylogous urethane by first reducing it to alcohol and convert-
ing the alcohol to the chloromethyl derivative followed by dis-
placement of chlorine with the appropriate dialkylamine as
shown in Scheme 3.
Synthesis of the bis-amide targets required a different route as
shown in Scheme 4. Silver assisted displacement of bromide from
3a–b by mesitylamine to afford 17a–b was performed as the first
step. Subsequently, alkylation of the imidazole with TBDMS pro-
tected bromoethanol, followed by TBAF-mediated removal of the
protecting TBDMS group and Jones oxidation gave the carboxylic
acids 19a–b which were converted to the dihydroimidazopyrimid-
inones 20a–b. The cyclic amides were then converted to N,N-dial-
kylated bis-amide targets 21–30 via Weinreb amidation.
The binding affinities of compounds thus prepared were deter-
mined by displacement of [125I]-Tyr-o-CRF from hCRF1R endoge-
nously expressed on IMR-32 human neuroblastoma cells
following the protocol described previously.12
Acknowledgment
The authors would like to thank Dr. Joanne Bronson for critical
reading of the Letter.
References and notes
1. Brain Facts. A Primer on Brain and Nervous System; Miller, M., Ed.; Chapter on
Stress; Publication of Society for Neuroscience, 2008; pp 31–33. Can be
accessed at www.sfn.org.
Initial results with several examples (21–30) of bis-amide
derivatives containing methyl or ethyl substituent on the imidaz-
ole ring demonstrated poor binding affinity with Ki values of
>10,000 nM (Table 1).
In the C-3 amide series which lacked one carbonyl function
compared to the bis-amide series (i.e., Q = H2, Table 1), N-cyclopro-
pylmethyl-N-propyl amides (compounds 12–14) showed a trend of
improved binding affinity when the C-2 substituent was varied
from electron-donating alkyl groups to electron- withdrawing
CF3 group (i.e., 14 vs 12).
2. Vale, W.; Spiess, J.; Rivier, C.; Rivier, J. Science 1981, 213, 1394.
3. Owens, M. J.; Nemeroff, C. B. Pharmacol. Rev. 1991, 43, 425.
4. McCarthy, J. R.; Heinrichs, S. C.; Grigoriadis, D. E. Curr. Pharm. Des. 1999, 5, 289.
5. (a) Dzierba, C. D.; Hartz, R. A.; Bronson, J. J. Ann. Rep. Med. Chem. 2008, 43, 3; (b)
Müller, M. B.; Wurst, W. Trends Mol. Med. 2004, 8, 409; (c) Heinrichs, S. C.;
Taché, Y. Exp. Opin. Invest. Drugs 2001, 10, 647; (d) Gilligan, P. J.; Robertson, D.
W.; Zaczek, R. J. Med. Chem. 2000, 43, 1641; (e) Keller, P. A.; Elfick, L.; Garner, J.;
Morgan, J.; McCluskey, A. Bioorg. Med. Chem. 2000, 8, 1213; (f) De Souza, E. B.;
Grigoriadis, D. E. Psychopharmacology. In The Fourth Generation of Progress;
Bloom, F. E., Kupfer, D. J., Eds.; Raven Press: New York, 1995; pp 505–517.
6. Dunn, A. J.; Berridge, C. W. Brain Res. Rev. 1990, 15, 71.
Removing the C3 amide carbonyl afforded the dialkylaminom-
ethyl analogs lacking a carbonyl group. In this series (Table 2)
the corresponding derivative 16 containing the N-cyclopropylm-
ethyl-N-propyl aminomethyl moiety had a Ki value of 32 nM indi-
cating a significant improvement in affinity. Hence our attention
was turned to these tetrahydroimidazo pyrimidine derivatives.
In an effort to improve upon the activity in this series, we uti-
lized a focused library of sixty amines. Many of the dialkylamines
necessary for this purpose were prepared as described in literature
from our laboratories.11 The results obtained with these com-
pounds are shown in Table 2. Hydrophobic secondary amine resi-
dues containing small rings such as cyclopropyl and cyclobutyl
as alkyl groups (e.g., 16, 34 and 36) demonstrated low nanomolar
range binding affinity. Amines containing fluorinated alkyl groups
as one of the alkyl groups also demonstrated pronounced binding
affinity as illustrated by examples such as 31 and 32. Cyclic tertiary
7. (a) Nemeroff, C. B.; Widerlov, E.; Bisette, G.; Walleus, H.; Karlsson, I.; Eklund, K.;
Kilts, C. D.; Loosen, P. T.; Vale, W. Science 1984, 226, 1342; (b) Nemeroff, C. B.;
Bisette, G.; Akil, H.; Fink, M. J. Psychiatry 1991, 158, 59.
8. Stenzel-Poore, M. P.; Heinrichs, S. C.; Rivest, R.; Koob, G. F.; Vale, W. W. J.
Neurosci. 1994, 14, 2579.
9. Smith, G. W.; Aubry, J. M.; Dellu, R.; Contarino, A.; Bilezekijian, L. M.; Gold, L. H.;
Chen, R.; Vale, W.; Lee, K. F. Neuron 1998, 20, 1093.
10. Keller, P. A.; Bowman, M.; Dang, K. H.; Garner, J.; Leach, S. P.; Smith, R.;
McCluskey, A. J. Med. Chem. 1999, 42, 2351.
11. Dubowchik, G. M.; Michne, J. A.; Zuev, D. Bioorg. Med. Chem. Lett. 2004, 14,
3147.
12. Han, X.; Michne, J. A.; Pin, S. S.; Burris, K. D.; Balanda, L. A.; Fung, L. K.; Fiedler,
T.; Browman, K. E.; Taber, M. T.; Zhang, J.; Dubowchik, G. M. Bioorg. Med. Chem.
Lett. 2005, 15, 3870. Ki values presented for all compounds are obtained as
averages from determinations which were run in duplicate.
13. (a) Leone-Bay, A. U. S. Patent 4,711,962, 1987.; (b) Benjes, P.; Grimmett, R.
Heterocycles 1994, 37, 735.
14. Levin, J. I.; Turos, E.; Weinreb, S. M. Synth. Commun. 1982, 12, 989.
15. Grewal, S. S.; Shepherd, J. K.; Bill, D. J.; Fletcher, A.; Dourish, C. T.
Psychopharmocology 1997, 133, 29.