A palladium-catalysed urea arylation route to a CRF1 receptor antagonist

Article abstract of DOI:10.1055/s-2006-950276

A new synthetic approach to a potent CRF antagonist, GW808990 (NBI35583), is reported. The route hinges on the palladium-catalysed intramolecular arylation of a urea with 2-chloropyridine. Spontaneous piperazine ring closure means that a high-yielding bis

Full text of DOI:10.1055/s-2006-950276

2716
LETTER
A Palladium-Catalysed Urea Arylation Route to a CRF1 Receptor Antagonist
P
alladium-CatalysedU
a
r
ea Arylati
t
on
R
ou
t
te toa
h
CRF1Rece
p
e
tor
A
ntagon
w
ist E. Popkin,* Richard K. Bellingham, Jerome F. Hayes
Chemical Development, GlaxoSmithKline, Old Powder Mills, Tonbridge, Kent, TN11 9AN, UK
Fax +44(1732)372355; E-mail: matt.e.popkin@gsk.com
Received 20 March 2006
Abstract: A new synthetic approach to a potent CRF antagonist,
GW808990 (NBI35583), is reported. The route hinges on the palla-
dium-catalysed intramolecular arylation of a urea with 2-chloro-
OH
NH
NO2
pyridine. Spontaneous piperazine ring closure means that a high-
yielding bisannulation reaction is the final step.
a, b
NO2
O
N
O
Key words: CRF antagonist, palladium, catalysis, ligands, urea,
arylations
H
2
N
H
3
c, d
The first corticotrophin release factor (CRF), a 41-amino-
acid neurotransmitter was isolated from ovine hypothalmi
in 1981.1 Human CRF has been shown to produce pro-
found alterations in endocrine, nervous and immune sys-
tem functions. CRF is believed to initiate its biological
effects by binding to a plasma membrane receptor which
has been found to be distributed throughout the brain.2,3
More recently, numerous small molecule CRF receptor
antagonists have been disclosed which could be useful in
the treatment of endocrine, psychiatric and neurologic
conditions or illnesses, including stress-related disorders
in general.4
NH
N
NH
N
OMe
NH2
OMe
NO2
e
N
H
N
H
5
4
f
N
The 2,3-fused pyridyl urea, GW808990 (also identified as
NBI35583), 1 has been developed under a collaborative
agreement between GlaxoSmithKline and Neurocrine
Biosciences as a promising CRF1 antagonist.5 The origi-
nal synthesis is shown in Scheme 1. Other fused pyridyl
ureas have been reported as CRF receptor antagonists by
Pfizer6 and DuPont.7 To date, the reported syntheses of all
of these compounds follows the basic structure of the
route to 1 as outlined in Scheme 1. Sequential selective
substitutions at the 4- and then 2-positions of the readily
available nitropyridinone 2 are followed by nitro-reduc-
tion of 4 and installation of the urea by reaction of 5 with
triphosgene. The final step in the synthesis of 1 is the an-
nulation of the urea 6 with dibromoethane to install the
piperazine ring.
NH
N
H
N
N
N
O
g
O
N
N
GW808990
NBI35583
1
6
OMe
OMe
Scheme 1 Reagents and conditions: (a) PhSO2Cl, Et3N, THF, ref-
lux; (b) 4-heptylamine, Et3N, 50%, 2 steps; (c) POCl3, MeCN, 50 °C;
(d) 4-anisidine, MeCN, 50 °C, 75%, 2 steps; (e) Na2S2O4, NaOH,
MeCN, H2O, quant; (f) triphosgene, NaOH, CH2Cl2, 80%; (g) dibro-
moethane, NaOH, CH2Cl2, 70%.
2-Chloropyridine 8 was synthesised via an activation/sub-
stitution sequence analogous to that described in
Scheme 1. Following substitution at the 4-position with 7
the hydroxyl group was acylated and the chloropyridine 8
was then formed in 41% overall yield from 2 (Scheme 2).
Although this chemistry could indeed be elaborated to 1
along lines similar to the supply route in Scheme 1, we
discovered that the 4-anisidine substitution of the 2-chlo-
ropyridine 8 was very difficult. With this in mind we
elected to adopt a different approach and we report herein
a synthetically distinct route to 1 via a palladium-cataly-
sed bisannulation reaction of the key urea 11.
At the outset we discovered that the urea installation in tri-
aminopyridine 5 was hampered by poor regioselectivity
between the 2- and 4-amino groups. Our initial strategy
toward a better synthesis of 1 was designed simultaneous-
ly to eliminate this regiochemical issue and avoid the use
of genotoxic dibromoethane by exchanging 4-heptyl-
amine for 4-(aminoethyl)heptane (7),8 thus installing the
required ethylene functionality from the start.
SYNLETT 2006, No. 17, pp 2716–2718
1
7
.1
0
.2
0
0
6
Advanced online publication: 09.10.2006
DOI: 10.1055/s-2006-950276; Art ID: D08406ST
© Georg Thieme Verlag Stuttgart · New York
Chloropyridine 8 was cleanly reduced by hydrogen at at-
mospheric pressure with 1% Pt/C catalyst in ethyl acetate

LETTER
Palladium-Catalysed Urea Arylation Route to a CRF1 Receptor Antagonist
2717
Various conditions were investigated in an attempt to ef-
fect the key N-aryl bond-forming step and, indeed, a few
reports of urea substitution of 2-halopyridines are present
in the literature.10-12 Our successful final step to 1 how-
ever uses the urea analogue of a Buchwald–Hartwig-type
palladium-catalysed aryl-N coupling. This reaction has
recently been popularised by Beletskaya who has shown
that palladium-dibenzylideneacetone complexes and
Xantphos ligands will N-arylate a wide range of ureas
with aryl bromides and iodides.13,14 We explored the pos-
sibility of using the related commercial ligand 12, which
did indeed achieve the desired N-aryl bond formation and
the subsequent closure of the pyridylpiperazine ring sys-
tem in 50% yield. In our hands however, better results
were observed using Buchwald’s catalyst-ligand system
as applied to the coupling of amines with 2-chloro-
pyridines.15 To this end the use of methylbiphenyldicyclo-
hexyl phosphine 13 (9 mol%) and Pd2(dba)3 (3.6 mol%)
with potassium carbonate (2.2 equiv) in refluxing dioxane
gave 1 in 86% yield after 18 hours. Chromatographic pu-
rification gave 1 in 58% isolated yield. This bisannulation
reaction of 11 can be followed by HPLC–MS and it is
clear that under the reaction conditions the palladium-
catalysed N-aryl bond formation is the faster step
(Scheme 4). A related palladium-catalysed intramolecular
arylation of ureas has recently been reported.16
OH
OAc
NO2
N
N
NO2
O
a–d
N
H
Cl
2
8
e
OH
HN
OAc
N
7
OMe
NO2
N
N
H
9
steps
1
Scheme 2 Reagents and conditions: (a) PhSO2Cl, Et3N, THF, ref-
lux; (b) 7, Et3N, 54%, 2 steps; (c) Ac2O, pyridine, toluene, 85%; (d)
POCl3, EtCN, reflux, 90%; (e) 4-anisidine, HCl, MeCN, reflux, 48 h,
25%.
to give 10, in 74% yield (Scheme 3). After catalyst filtra-
tion, the aminopyridine 10, still in ethyl acetate, can be
treated with one equivalent of 4-methoxyphenyl
isocyanate9 to give the crystalline urea 11 in 73% isolated
yield.
OAc
N
H
N
a
11
1
O
N
N
Ar
14
OAc
N
Scheme 4 Reagents and conditions: (a) Pd2(dba)3 (3.6 mol%), 13
a
8
(9 mol%), K2CO3 (2.2 equiv), dioxane, reflux, 20 h, 86%.
NH2
10
N
Cl
In summary, a new six-step route to the CRF receptor
antagonist GW808990 1 has been developed. The un-
optimised route employs a palladium-catalysed bisannu-
lation–arylation of urea 1117 as the key step and allows for
the synthesis of 1 in 17% overall yield from 2.18
b
N
N
N
OAc
c or d
N
N
O
H
N
H
References and Notes
N
N
(1) Vale, W.; Spiess, J.; Rivier, J. Science 1981, 213, 1394.
(2) Gilligan, P. J.; Hartig, P. R.; Robertson, D. W.; Zaczek, R. In
Annual Reports in Medicinal Chemistry, Vol. 32; Bristol, J.
A., Ed.; Academic: San Diego, 1997, 1–50.
(3) Turnbull, A. V.; Rivier, C. Proc. Soc. Experim. Biol. Med.
1997, 215, 1.
O
OMe
Cl
1
11
OMe
PPh2
PPh2
O
(4) (a) Gutman, D. A.; Owens, M. J.; Nemeroff, C. B.
Pharmaceut. News 2001, 8, 18. (b) Heinrichs, S. C.; Tache,
Y. Exp. Opin. Invest. Drugs 2001, 10, 647.
13
12
PCy2
(5) Neurocrine Biosciences Inc. USA, PCT Int. Appl.,
WO0027846, 2000.
(6) Pfizer Inc. USA, PCT Int. Appl., WO 01/23389A2, 2001.
Scheme 3 Reagents and conditions: (a) 1% Pt/C, H2 (1 atm), AcOH,
EtOAc, 74%; (b) 4-methoxyphenyl isocyanate, EtOAc, 73%; (c)
Pd2(dba)3 (3.6 mol%), 12 (7.2 mol%), K2CO3 (2.0 equiv), dioxane, re-
flux, 22 h, 50%; or (d) Pd2(dba)3 (3.6 mol%), 13 (9 mol%), K2CO3
(2.2 equiv), dioxane, reflux, 20 h, 86%.
Synlett 2006, No. 17, 2716–2718 © Thieme Stuttgart · New York

2718
M. E. Popkin et al.
LETTER
(7) Arvanitis, A. G.; Beck, J. P.; Curry, M. A.; Fitzgerald, L. W.;
Gilligan, P. J.; Trainor, G. L.; Zaczek, R. Bioorg. Med.
Chem. Lett. 1999, 9, 967.
(8) Cope, A. C.; Hancock, E. J. Am. Chem. Soc. 1942, 64, 1503.
(9) 4-Methoxyphenyl isocyanate was purchased from Sigma-
Aldrich.
(100 MHz, CDCl3): d = 171.1, 157.1, 156.7, 156.4, 154.2,
152.7, 130.8, 123.2, 118.3, 114.2, 112.4, 91.4, 61.9, 61.7,
41.1, 35.7, 24.3, 20.6, 20.2, 14.2. IR (ATR mode): 1735,
1584, 1510, 1244, 1027, 986, 934, 836 cm–1. HRMS (ES): m/
z [MH+] calcd for C25H36ClN4O4: 491.2425; found:
491.2406.
(10) Meth-Cohn, O.; Zegui, Y. J. Chem. Soc., Perkin Trans. 1
1998, 423.
(11) Bianchi, M.; Butti, A.; Rossi, S.; Barzaghi, F.; Marcaria, V.
Eur. J. Med. Chem. Chim. Ther. 1983, 18, 495.
(12) Maruyama, T.; Kozai, S.; Uchida, M. Nucleosides
Nucleotides 1999, 18, 661.
(13) (a) Beletskaya, I. P.; Artamkina, G. A.; Sergeev, A. G.
Tetrahedron Lett. 2001, 42, 4381. (b) Beletskaya, I. P.;
Artamkina, G. A.; Sergeev, A. G. J. Org. Chem. USSR
(Engl. Transl.) 2003, 39, 1741. (c) Beletskaya, I. P.;
Artamkina, G. A.; Sergeev, A. G. Tetrahedron Lett. 2003,
44, 4719.
(14) Buchwald, S. L.; Yin, J. J. Am. Chem. Soc. 2002, 124, 6043.
(15) Buchwald, S. L.; Harris, M. C. J. Org. Chem. 2000, 65,
5327.
(16) Ferraccioli, R.; Carenzi, D. Synthesis 2003, 1383.
(17) Data for 10: mp 128–129 °C (EtOAc). 1H NMR (400 MHz,
CDCl3): d = 7.22 (d, J = 9.1 Hz, 2 H), 6.80 (d, J = 9.1 Hz, 2
H), 6.70 (br s, 1 H), 6.67 (s, 1 H), 6.36 (br s, 1 H), 4.04 (t,
J = 6.1 Hz, 2 H), 3.75 (s, 3 H), 3.74–3.78 (m, 1 H), 3.46 (t,
J = 6.1 Hz, 2 H), 2.44 (s, 3 H), 1.89 (s, 3 H), 1.44–1.50 (m,
4 H), 1.21–1.28 (m, 4 H), 0.83 (t, J = 7.3 Hz, 6 H). 13C NMR
(18) The urea 10 (300 mg, 0.61 mmol) and K2CO3 (190 mg, 1.38
mmol) were suspended in dioxane (6.0 mL). Pd2(dba)3 (20
mg, 0.022 mmol, 0.036 equiv) and 2-(dicyclohexylphos-
phino)-2¢-methylbiphenyl (20 mg, 0.055 mmol, 0.09 equiv)
were added and the reaction vessel was flushed with Ar and
the reaction mixture was heated to reflux for 20 h. Solution
yield of the product 1 at this point was 86%. The reaction
mixture was then evaporated and partitioned between
EtOAc (10 mL) and 5% aq NaHCO3 (5 mL). The EtOAc
layer was then separated and dried over MgSO4 before being
filtered and evaporated to give the crude product 1 which
was purified by chromatography (hexane–EtOAc, 1:1) to
give pure 1 in 58% yield. Data for 1: 1H NMR (400 MHz,
CDCl3): d = 7.63 (dd, J = 2.2, 6.8 Hz, 2 H), 7.00 (dd, J = 2.2,
6.8 Hz, 2 H), 6.25 (s, 1 H), 3.98 (t, J = 5.0 Hz, 2 H), 3.83 (s,
3 H), 3.74 (m, 1 H), 3.38 (t, J = 5.0 Hz, 2 H), 2.44 (s, 3 H),
1.55 (m, 4 H), 1.32 (m, 4 H), 0.92 (t, J = 7.3 Hz, 6 H). 13
C
NMR (100 MHz, CDCl3): d = 158.3, 152.0, 151.2, 141.4,
138.3, 127.2, 126.8, 114.4, 106.8, 99.2, 55.7, 55.5, 39.4,
34.7, 34.7, 25.2, 20.0, 14.0. IR (ATR mode): 1712, 1654,
1515, 1246, 1036, 854, 793, 735, 691 cm-1. HRMS (ES): m/
z [MH+] calcd for C23H31N4O2: 395.2447; found: 395.2441.
Synlett 2006, No. 17, 2716–2718 © Thieme Stuttgart · New York