Substituted tetraazaacenaphthylenes as potent CRF1 receptor antagonists for the treatment of depression and anxiety

Article abstract of DOI:10.1016/j.bmcl.2005.05.040

Two isomers of the hexahydro-tetraazaacenaphthylene templates (1 and 2) are presented as novel, potent, and selective corticotropin releasing factor-1 (CRF₁) receptor antagonists. In this paper, we report the affinity and SAR of a series of compounds, as well as pharmacokinetic characterization of a chosen set. The anxiolitic activity of a selected example (2ba) in the rat pup vocalization model is also presented.

Full text of DOI:10.1016/j.bmcl.2005.05.040

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3713–3716
Substituted tetraazaacenaphthylenes as potent CRF1
receptor antagonists for the treatment of depression and anxiety
a,
a
a
a
a
Y. St-Denis,a,* R. Di Fabio, G. Bernasconi, E. Castiglioni, S. Contini, D. Donati,
*
E. Fazzolari,a G. Gentile,a D. Ghirlanda,a C. Marchionni,a,ꢀ F. Messina,a,ꢀ F. Micheli,a
F. Pavone,a A. Pasquarello,a F. M. Sabbatini,a M. G. Zampori,a,ꢁ R. Arbanb and G. Vitullic
aDepartment of Medicinal Chemistry, GlaxoSmithKline Medicines Research Center, Via A. Fleming 4, 37135 Verona, Italy
bDepartment of Pharmacology, GlaxoSmithKline Medicines Research Center, Via A. Fleming 4, 37135 Verona, Italy
cDepartment of Drug Metabolism and Pharmacokinetic, GlaxoSmithKline Medicines Research Center,
Via A. Fleming 4, 37135 Verona, Italy
Received 17 March 2005; revised 6 May 2005; accepted 11 May 2005
Available online 8 June 2005
Abstract—Two isomers of the hexahydro-tetraazaacenaphthylene templates (1 and 2) are presented as novel, potent, and selective
corticotropin releasing factor-1 (CRF1) receptor antagonists. In this paper, we report the affinity and SAR of a series of compounds,
as well as pharmacokinetic characterization of a chosen set. The anxiolitic activity of a selected example (2ba) in the rat pup vocal-
ization model is also presented.
Ó 2005 Elsevier Ltd. All rights reserved.
Depression and anxiety are psychiatric disorders that
constitute a major health concern worldwide, and new
pharmacological approaches for improved efficacy and
reduced side effects profiles relative to currently mar-
keted drugs are highly desired.1 Corticotropin-releasing
factor (CRF), a 41 amino acid peptide synthesized by
specific hypothalamic nuclei in the brain, was originally
isolated by Wale and colleagues2 in 1981 from ovine
hypothalamus. It plays an important role as a neuro-
transmitter in the mediation of anxiety and depression
related behaviors, and could represent a new opportu-
nity for the treatment of such diseases. The fundamental
role of CRF is to prepare the organism for a response to
various stressors, such as physical trauma, insults to the
immune system, and social interactions, through the con-
trol of the hypothalamic–pituitary–adrenal (HPA) axis.
CRF1 receptor antagonists.3 Compounds such as CP-
154,5264 and DMP-6965 were among the first to show
good binding affinities coupled with interesting in vivo
activities. A vast majority of the compounds published
to date have in common the structural features shared
by the two aforementioned antagonists: a bicyclic
heterocyclic core substituted with a highly lipophilic
alkylamine side chain and a 2,4-disubstituted or 2,4,
6-trisubstituted aromatic or heteroaromatic ring (Fig.
1). More recently, new high-affinity analogs containing
monocyclic and tricyclic cores have been disclosed,6 thus
broadening the structural diversity and known SAR
for the receptor. Notwithstanding these encouraging
progresses, very few molecules have reached the advanced
preclinical development or clinical phases to date.
During the last decade, several research groups have
published their work in the area of small molecule
*
Corresponding authors. Tel.: +39 045 8219806; fax: +39 045
9218196; e-mail: yves.a.stdenis@gsk.com
ꢀ Present address: Nerviano Medical Sciences, Viale Pasteur #10,
20014 Nerviano (MI), Italy.
ꢁ Present address: Pfizer Limited, Ramsgate Road, Sandwich, Kent
CT13 9NJ, UK.
Figure 1. Known bicyclic small molecule CRF1 receptor antagonists.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.040

3714
Y. St-Denis et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3713–3716
protected with a t-butyl-dimethylsilyl group (P, in
Scheme 1).
The protected alcohol was then subjected to a nucleo-
philic aromatic substitution with deprotonated anilines
to give aminopyrimidines 6. The amino group of 6 was
protected (t-Boc) and the alcohol desilylated to give
intermediates 7. Mesylation of alcohols 7, followed by
cleavage of the t-Boc protecting group gave an amino-
mesylate intermediate, which cyclised to substituted
pyrrolidino-pyrimidines 8 upon treatment with Et3N in
THF. Ozonolysis of 8 to give aldehydes 9 was followed
by imine formation with the appropriate alkylamines,
which was then reduced in situ to give the desired com-
pound 1.
Figure 2. General structures 1 and 2.
We wish to report here our efforts toward the identifica-
tion of novel, potent, and selective CRF1 receptor
antagonists. In our search for heterocyclic templates
with high receptor affinities, hexahydro- and tetrahy-
dro-tetraazaacenaphthylenes of general structures 1
and 2 were identified (Fig. 2).7,8
Scheme 2 illustrates the preparation of tetraaza-
acenaphthylene template 2. The alcohol group of inter-
mediate 5 (from Scheme 1) was similarly protected
using a tert-butyldiphenylsilyl group (P, in Scheme 2)
to give better stability. Nucleophilic aromatic substitu-
tion with different anilines followed by protection of
the amino group were performed as reported in Scheme
1, to give intermediate 11. Ozonolysis of 11 followed by
reductive workup gave an alcohol, which was reacted
with methanesulfonyl chloride in the presence of Et3N
to give mesylates 12. Cleavage of the t-Boc protecting
group, followed by treatment with Et3N gave the sub-
stituted piperidinopyrimidines 13. Deprotection of the
alcohol group of 13 was followed by mesylate formation
and cyclization with the appropriate amines (neat) to
yield the desired compounds 2a.
The top amine substituents (R1 in 1 and 2) as well as the
bottom aromatic portion (substituents R2) were ex-
plored, leaving the central acenaphthylene core intact.
It was our belief that modification of these two regions
would allow for the optimization of the general physico-
chemical characteristics of the molecules, thus leading to
a better/suitable PK profile.
The synthetic procedure for the preparation of the tetra-
azaacenaphthylene template 1 is described in Scheme 1.
Intermediate 4 has already been described in the litera-
ture.9,10 Allylation of 4 followed by reduction of the
ethyl ester functionality yielded alcohol 5, which was
Both templates were initially prepared as racemic mix-
tures. In order to verify if the stereogenic center of the
different templates had an effect on the affinity for the
receptor, the two enantiomers of two examples of tem-
plate 2a were prepared, as described in Scheme 3. Inter-
mediate 14 (racemic, from Scheme 2) was acylated with
Scheme 1. Reagents and conditions: (a) (i) acetamidine hydrochloride,
Na, MeOH, rt, 20 min; (ii) 2-ethoxycarbonyl–succinic acid diethyl
ester, rt, 48 h; (iii) POCl3, reflux, 3.5 h, 73%; (b) 1 M LiHMDS/Hex,
allyl bromide, THF, 0 °C to rt, 68%; (c) 1 M DIBAl-H/Hex, CH2Cl2,
À78 to 0 °C, 3 h, 86%; (d) t-BDMSi-Cl, imidazole, DMF, 0 °C to rt,
3 h, 81%; (e) NaH 80%/oil, R2-aniline, THF, 0 °C, 15 min, then
compound 5, THF, reflux, 3 h, 72%; (f) (t-Boc)2O, DMAP, CH2Cl2, rt,
18 h, 83%; (g) Et3NÆ3HF, DMF, rt, 18 h, 74%; (h) CH3SO2Cl, Et3N,
CH2Cl2, rt, 18 h, 98%; (i) (i) TFA 20%/CH2Cl2, rt, 2 h, (ii) Et3N, THF,
rt, 1 h, quant.; (j) O3, CH2Cl2, À78 °C, 30 min, then Me2S, rt, 18 h,
27%; (k) R1-amine, THF, rt, 90 min, then 1 M NaBH3CN/THF, rt,
18 h, 7%. The yields reported are for compound 1b (Table 1) only. P,
protecting group.
Scheme 2. Reagents and conditions: (a) t-BDPSi-Cl, DMAP, imida-
zole, DMF, 0 °C to rt, 2 h, 90%; (b) NaH 80%/oil, R2-aniline, THF,
0 °C, 30 min, then compound 10, THF, reflux, 5 h, 46%; (c) (t-Boc)2O,
DMAP, CH2Cl2, rt, 48 h, 93%; (d) O3, CH2Cl2, À78 °C, 30 min, then
NaBH4, rt, 3 h, 54%; (e) CH3SO2Cl, Et3N, CH2Cl2, rt, 18 h, 95%; (f)
TFA 20%/CH2Cl2, rt, 18 h, 100%; (g) Et3N, THF, 0 °C to rt, 16 h,
48%; (h) Et3NÆ3HF, DMF, 40 °C, 4 h, 98%; (i) CH3SO2Cl, Et3N,
CH2Cl2, rt, 16 h, 93%; (j) neat R1-amine, 120 °C, 8 h, 47%. The yields
reported are for compound 2aa (Table 2) only. P, protecting group.

Y. St-Denis et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3713–3716
3715
Table 2. Affinity results for template 2a
Compounds
R1
R2
pIC50
2aa
2ab
Pent-3-yl
2,4-Dichloro
2,4-dichloro
2,4-Dichloro
2,4-Bis-CF3
2,4-Dichloro
2,4-Dichloro
2,4-Bis-CF3
2,4-Bis-CF3
7.16
5.90
7.65
7.44
7.96
7.89
7.55
7.84
1,3-Di-MeO-prop-2-yl
Hept-4-yl
Hept-4-yl
2ac
2ad
2aea
2afa
2aga
2aha
Hept-4-yl
Hept-4-yl
Hept-4-yl
Hept-4-yl
Scheme 3. Reagents and conditions: (a) (S)-2-acetoxypropionyl chlo-
ride, DMAP, CH2Cl2, 0 °C, 30 min, then chiral HPLC separation,
46%; (b) LiOH, THF/H2O, 50 min, 90%. The yields reported are for
compounds 2ag and 2ah (Table 2) only.
a Chirally pure enantiomers, prepared as in Scheme 3.
Table 3. Affinity results for template 2b
enantiopure (S)-2-acetoxypropionyl chloride. Initial
attemps to separate diastereoisomers 15a and 15b using
conventional flash chromatography proved to be unsuc-
cessful, thus prompting us to use chiral HPLC.11 The
two separated enantiomerically pure compounds were
hydrolyzed to give alcohols 16a and 16b, which were sub-
jected to the conditions reported above in Scheme 2
(steps i and j) to complete the synthesis of the chirally
pure enantiomers.
Compound
R1
R2
pIC50
2ba
2bb
2bc
2bd
2be
Hept-4-yl
Hept-4-yl
Pent-3-yl
Pent-3-yl
Pent-3-yl
2,4-Bis-CF3
2-Me-4-CN
2,4-Bis-CF3
2-Me-4-CN
2-Cl-4-CN
7.85
7.35
7.58
7.57
7.21
Finally, the unsaturated tetraazaacenaphthylenes 2b
were easily obtained by the DDQ oxidation of saturated
templates 2a, as shown in Scheme 4.
CRF binding affinity has been determined in vitro by the
compoundsꢂ ability to displace 125I-oCRF from recom-
binant human CRF receptors expressed in Chinese
Hamster Ovary (CHO) cell membranes.7,8 Affinity re-
sults for templates 1, 2a, and 2b are presented in Tables
1–3, respectively. As a general rule, affinity for the recep-
tor seems to be controlled by the lipophilicity of the
compounds (see 1a vs 1c in Table 1, or 2ab vs 2aa vs
2ae in Table 2). The top region of the molecules (R1) re-
quires lipophilic chains in order to maintain high affini-
ties (see 1a vs 1b vs 1c in Table 1). Polar groups do not
seem to be well tolerated in that region (see 2aa vs 2ab in
Table 2). As far as the substituents on the bottom phenyl
ring are concerned, reduction of lipophilicity was toler-
ated, as can be seen in Table 3, where a cyano group has
replaced one of the lipophilic moieties (see 2bc vs 2bb or
2bd in Table 3).
Scheme 4. Reagents and conditions: (a) DDQ, CH2Cl2, rt, 3 h, 65%.
The yield reported is for compounds 2ba (Table 3) only.
Table 1. Affinity results for template 1 and standards
Compound
R1
R2
pIC50
As can be seen in Table 2, the configuration of the ste-
reogenic center in template 2b has a limited effect on
the affinity of this series for the receptor (see 2ae vs
2af, and 2ag vs 2ah in Table 2). The chirally pure enan-
tiomers have more or less the same affinity as the race-
mic mixtures. The same conclusion can be reached
looking at the compounds in Table 3, where the chiral
CP-154,526
DMP-696
7.76
7.42
5.35
6.74
7.52
7.39
1a
1b
1c
1d
Cyclopropylmethyl
3-Pentyl
4-Heptyl
2,4-Dichloro
2,4-Dichloro
2,4-Dichloro
2,4-Bis-CF3
4-Heptyl

3716
Y. St-Denis et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3713–3716
Table 4. DMPK parameters for 1c, 2ac, 2ad, and 2baa
Acknowledgment
Parameters
1c
2ac
2ad
2ba
The authors would like to thank Dr. Carla Marchioro
for analytical support.
Plasma clearance (mL/min/kg)
Bioavailability (%)
Brain/Plasma
31
31
32
14
15
36
1.1
10
33
1.3
1.8
2.0
a Pharmacokinetic experiments were performed in Hans–Wistar rats
after oral (1 mg/kg) and intravenous (0.5 mg/kg) administrations;
brain/plasma ratios and brain levels were measured at 1 h after a
0.5 mg/kg iv dose.
References and notes
1. (a) Fawcett, J.; Barkin, R. L. J. Clin. Psychiat. 1997, 58,
32; (b) DeVane, C. L. J. Clin. Psychiat. 2000, 61, 4.
2. Wale, W.; Spiess, J.; Rivier, C.; Rivier, J. Science 1981,
213, 1394.
3. (a) For some recent reviews and opinions on the subject,
see Gilligan, P. J.; Li, Y.-W. Curr. Opin. Drug Discov. Dev.
2004, 7, 487; (b) Zorilla, E. P.; Koob, G. F. Expert Opin.
Investig. Drugs 2004, 13, 799.
4. Schulz, D. W.; Mansbach, R. S.; Sprouse, J.; Braselton, J.
P.; Collins, J.; Corman, M.; Dunaiskis, A.; Faraci, S.;
Schmidt, A. W.; Seeger, T.; Seymour, P.; Tingley, F. D.
III; Winston, E. N.; Chen, Y. L.; Heym, J. Proc. Natl.
Acad. Sci. USA 1996, 93, 10477.
5. He, L.; Gilligan, P. J.; Zaczek, R.; Fitzgerald, L. W.;
McElroy, J.; Shen, H. S.; Saye, J. A.; Kalin, N. H.;
Shelton, S.; Christ, D.; Trainor, G.; Hartig, P. J. Med.
Chem. 2000, 43, 449.
center has been eliminated through the introduction
of an insaturation. Analogs of template 2b are almost
equipotent to their equivalents of template 2a, whether
they are chirally pure or racemates (see 2ba vs 2ad vs
2ag, and 2ah).
Four compounds were selected for the in vivo PK char-
acterization (1c, 2ac, 2ad, and 2ba). Table 4 reports the
in vivo parameters measured for these compounds: low
to moderate plasma clearance, good oral bioavailability,
and brain penetration were observed.
6. Saunders, J.; Williams, J. Prog. Med. Chem. 2003, 41, 195.
7. Di Fabio, R.; Micheli, F.; Pasquarello, A.; St-Denis, Y.
WO 02/087573 A1, 2002.
8. Di Fabio, R.; Gentile, G.; Haddach, M.; St-Denis, Y.;
Williams, J.P. WO 02/094826, 2002.
9. Chen, Y.L. WO 95/33750 A1, 1995.
In particular, compound 2ba exhibited a high affinity for
the receptor and good PK characteristics. Based on this
positive profile, it was evaluated in the rat pup vocaliza-
tion model.12 2ba both at 30 and 60 mg/kg was able to
reduce significantly the rat pup vocalization time by
55% with respect to vehicle.
10. Chen, Y.L. US 5962479 A, 1999.
11. Daicel CHIRALPAK AD; mobile phase: n-hexane/iso-
propyl alcohol 9:1; flow rate: 6.5 mL/min; detector type:
diode array detector (DAD) (225 nm, 292 nm).
12. Rat pups between 9 and 11 days postnatal emit ultrasonic
vocalization in response to separation from their mother
and littermates. This behavior can reflect a state of
distress. This test has been originally proposed by Gardner
(J. Pharmacol. Methods, 1985, 14, 181) as a sensitive test
for anxiolytic drugs. Benzodiazepines, buspirone, CRF1
antagonists, fluovoxamine, and tianeptine have been
reported to be active in this test.
In summary, the synthesis of two novel tetra-
azaacenaphthylene templates was described. Com-
pounds belonging to these series exhibited high
affinities for the CRF1 receptor. Selected examples of
both templates showed good in vivo PK characteristics.
Furthermore, compound 2ba exhibited good activity in
the rat pup vocalization model, thus confirming that
CRF1 antagonists may play a role in the treatment of
anxiety and depression.