Discovery of NBI-77860/GSK561679, a potent corticotropin-releasing factor (CRF1) receptor antagonist with improved pharmacokinetic properties

Article abstract of DOI:10.1016/j.bmcl.2010.10.095

Antagonists of the corticotropin-releasing factor (CRF) neuropeptide may prove effective in treating stress and anxiety related disorders. In an effort to identify antagonists with improved physico-chemical properties a new series of CRF₁ antagonists were designed to substitute the propyl groups at the C7 position of the pyrazolo[1,5-a]pyrimidine core of 1 with heterocycles. Compound (S)-8d was identified as a high affinity ligand with a pK_i value of 8.2 and a functional CRF₁ antagonist with pIC₅₀ value of 7.0 in the in vitro CRF ACTH production assay.

Full text of DOI:10.1016/j.bmcl.2010.10.095

Bioorganic & Medicinal Chemistry Letters 20 (2010) 7259–7264
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of NBI-77860/GSK561679, a potent corticotropin-releasing factor
(CRF1) receptor antagonist with improved pharmacokinetic properties
John E. Tellew a, Marion Lanier a, Manisha Moorjani a, Emily Lin a, Zhiyong Luo a, Deborah H. Slee a,
Xiaohu Zhang a, Sam R. J. Hoare b, Dimitri E. Grigoriadis b, Yves St. Denis c, Romano Di Fabio c,
Enza Di Modugno c, John Saunders a, John P. Williams a,
⇑
a Department of Medicinal Chemistry, Neurocrine Biosciences, 12780 El Camino Real, San Diego, CA 92130, USA
b Department of Pharmacology, Neurocrine Biosciences, 12780 El Camino Real, San Diego, CA 92130, USA
c Neurosciences Centre of Excellence for Drug Discovery and Molecular Discovery Research, GlaxoSmithKline Medicines Research Centre, Via A. Fleming 4, 37135 Verona, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Antagonists of the corticotropin-releasing factor (CRF) neuropeptide may prove effective in treating stress
and anxiety related disorders. In an effort to identify antagonists with improved physico-chemical prop-
erties a new series of CRF1 antagonists were designed to substitute the propyl groups at the C7 position of
the pyrazolo[1,5-a]pyrimidine core of 1 with heterocycles. Compound (S)-8d was identified as a high
affinity ligand with a pKi value of 8.2 and a functional CRF1 antagonist with pIC50 value of 7.0 in the
in vitro CRF ACTH production assay.
Received 9 September 2010
Revised 18 October 2010
Accepted 19 October 2010
Available online 25 October 2010
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Corticotropin-releasing factor
CRF1 antagonist
GSK561679
Corticotropin-releasing factor (CRF) is a 41 amino acid residue
C-amidated peptide that is secreted by cells of the paraventricular
nucleus of the hypothalamus in response to stressful stimuli and is
the key mediator of an organism’s response to stress. Secretion of
CRF causes release of adrenocorticotropin-releasing hormone
(ACTH) from corticotrophs in the anterior pituitary via binding to
the CRF1 receptor, a member of the class B family of G-protein cou-
pled receptors. The stress-induced release of ACTH, in turn, acts at
the adrenal gland and results in the secretion of glucocorticoids
which feed back at the level of the pituitary and the hypothalamus
to attenuate the further release of ACTH and CRF, respectively. This
hormone loop is known as the hypothalamic-pituitary adrenal
(HPA) axis and is tightly regulated in mediating the stress
response.1
N
N
N
N
N
N
N
N
N
Cl
N
Cl
1
2
The CRF1 receptor is localized to a number of tissues and brain
sites, notably the hypothalamus, cerebral cortex, pituitary,
adrenals, testes, placenta, and gastrointestinal tract.2 In the years
since the discovery of the CRF1 receptor, a large body of evidence
has accumulated that links hyperactivity of the CRF system to a
number of disease states, in particular depression, anxiety, irritable
bowel syndrome (IBS), premature labor, and addiction disorders.
Consequently, the search for antagonists to the CRF1 receptor has
been an active area of investigation in drug discovery for more
than two decades.3
NH
O
N
N
N
N
N
O
N
O
3
4
⇑
Corresponding author. Tel.: +1 858 617 7662; fax: +1 858 617 7925.
E-mail address: jwilliams@neurocrine.com (J.P. Williams).
Figure 1. Clinically evaluated CRF1 receptor antagonists.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.10.095

7260
J. E. Tellew et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7259–7264
In the first published clinical study of a CRF1 receptor antago-
the pyrazolo[1,5-a]pyrimidine core structure. Based on calculated
log P values10 it was anticipated that replacement one methylene
unit of the starting propyl side chain with an oxadiazolyl heterocy-
cle would lower the overall calculated log P by half of a log unit. By
decreasing the log P (<4) it was hoped that the compounds would
retain potent CRF1 receptor functional inhibition while decreasing
the volume of distribution (Vdss <10 L/kg) and clearance (<33
mL/min kg) in rat pharmacokinetic studies.
The synthesis of analogs where an 1,2,4-oxadiazolyl heterocycle
is introduced into the top region side chain is summarized in
Scheme 1. Key intermediate 5 has been previously described11
and addition of the requisite amino esters provided the intermedi-
ates 6a–f and 7. Condensation with methylamidoxime and the es-
ter intermediates, in the presence of sodium hydride, followed by
thermal cyclization gave the compounds 8a–f and 9.
The 1,2,4-oxadiazolyl analogs 8a–f and 9 were assayed for bind-
ing to the recombinant human CRF1 receptor and selected high
affinity compounds were subsequently tested for in vitro meta-
bolic stability in human liver microsomes expressed as maximum
estimated bioavailability (max F% values) (Table 1).12 Compound 1
was utilized as a positive control and it has been demonstrated to
be a high affinity CRF1 receptor antagonist with a pKi value of 8.4
and a calculated log P value of 4.9. This structure typifies CRF1
receptor antagonists where most have a branched alkyl ‘top re-
gion’, a core heterocycle and an aromatic bottom region with sub-
stitution at the 2 and 4 positions on the aryl ring. It appears that
the core heterocycle may play the role of a hydrogen bond acceptor
when interacting with the receptor13 and the substituent at the 2
position of the bottom aromatic ring acts as a steering group14
for that moiety.
nist, NBI-30775/R121919 (compound 1, Fig. 1) demonstrated sig-
nificant improvement in depression and anxiety scores in patients
with severe depression without demonstrating any untoward ef-
fects in this population. While this trial was not considered proof
of concept due to the design of an open label trial, the apparent
improvement in both depression and anxiety scores over the
30-day treatment period provided a pharmacological rationale
for blocking this system in the primary disease indication. Further
clinical studies demonstrated that this particular compound
caused a reversible increase in liver enzymes thus precluding
any further development.4 More recent reports of clinical trial re-
sults have been mixed, for example, it was reported that NBI-
34041/SB-723620 (2) inhibited the stress-induced increase of
ACTH and cortisol levels in healthy volunteers during a Trier So-
cial Stress Test.5 On the other hand, CP-316,311 (3) and pexacer-
font (4) failed to show any evidence of efficacy in patients
suffering from major depression or generalized anxiety disorder,
respectively.6,7 In spite of these seemingly disparate results, there
is still a need for a CRF1 receptor antagonist that can be used in
later stage clinical trials using an extended dosing regimen to ad-
dress the apparent ambiguities raised by the earlier reported clin-
ical studies.
Typically CRF1 receptor antagonists are thought to possess
physico-chemical properties that are sub-optimal for drugs that
target the central nervous system. Previously reported compounds
have high hepatic clearance, large volumes of distribution and high
plasma protein binding.8 The main objective of this current work
was to lower the lipophilicity of 1 (log P = 4.9)9 by the introduction
of a heterocycle into one of the N-alkyl chains at the C7 position of
R 1
O
O
R 2
Cl
N
O
NH
O
N
N
N
a
N
N
N
+
N
N
N
O
O
O
5
6a (R1 = H, R2 =H)
7
6b (R1= H , R2 = Et)
6c (R1 = Me, R2 = H)
6d (R1 = Et, R2 = H)
6e (R1= (CH2)2, R2 = H)
b
6f (R1 = (CH2)3, R2 = H)
R 1
O
N
R 2
N
N
N
NH
O
N
N
N
N
N
+
N
N
O
O
9
8a (R1 = H, R2 =H)
8b (R1= H , R2 = Et)
8c (R1 = Me, R2 = H)
8d (R1 = Et, R2 = H)
8e (R1= (CH2)2, R2 = H)
8f (R1 = (CH2)3, R2 = H)
Scheme 1. (a) Amino esters, Et3N, acetonitrile, 80 °C; (b) MeCNOHNH, NaH, THF, reflux.

J. E. Tellew et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7259–7264
7261
Table 1
In vitro CRF1 receptor binding, metabolism and calculated log P
R1
N
N
N
Ar
Compound
R1
Ar
CRF1 pKi SDa
HLM max F%b
Calcd log Pc
N
1
8.4 0.2
17
4.9
3.3
4.3
3.8
4.3
4.3
4.3
3.7
4.1
3.9
N
N
N
NH
N
O
O
N
8a
7.2 0.1
8.3 0.1
7.6 0.1
8.1 0.1
8.0 0.1
8.2 0.2
7.5 0.1
7.3 0.1
7.5 0.1
7.0 0.1
ND
17
38
42
63
42
43
37
32
64
O
N
N
8b
O
O
O
O
O
O
O
O
O
N
NH
8c
O
N
N
NH
NH
8d
O
O
N
N
(R)-8d
(S)-8d
8e
N
N
NH
NH
O
O
N
N
N
N
NH
8f
O
O
N
N
N
NH
9
O
NH
14
2.5
N
N
(continued on next page)

7262
J. E. Tellew et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7259–7264
Table 1 (continued)
Compound
R1
Ar
CRF1 pKi SDa
HLM max F%b
Calcd log Pc
N
N
N
O
NH
NH
15
19
20
21
6.6 0.1
ND
2.1
O
O
N
7.1 0.1
7.3 0.1
7.3 0.2
26
3.9
3.7
3.6
O
N
NH
ND
ND
N
O
O
N
N
O
N
NH
N
O
a
See Ref. 12.
b
Maximum estimated bioavailability values were determined in human liver microsomes: ND = not determined.
See Ref. 10.
c
Introduction of a single 1,2,4-oxadiazolyl methyl amino side
in calculated log P did not result in a more stable compound by
chain at the C7 position of the pyrazolo[1,5-a]pyrimidine (8a) pro-
vided a baseline binding affinity (pKi = 7.2) for the series. It was
clear at this point that additional substitution would be required
to achieve the objective of increasing the affinity by 10-fold and
the ethyl derivative 8b accomplished this task (pKi = 8.3). Addi-
tional characterization of compound 8b showed that the decrease
comparing human liver microsomal predicted clearance values.
Branching at the
a position of 1,2,4-oxadiazolyl methyl substituent
(8c), which re-introduced a secondary amine NH, did result in a
compound with improved metabolic stability (estimated max
F% = 38) with comparable CRF1 receptor binding affinity
(pKi = 7.6). Extending the
a
position substituent by one methylene
O
O
H
N
O
H2N
NH
O
+
NH
N
NH
H2N
NH
O
H
N
N
N
O
N
N
a
N
N
N
+
N
N
N
N
OMe
OMe
OMe
OMe
6d
7
10
11
b
N
O
O
N
H
O
H
N
O
N
NH
N
N
NH
NH
N
NH
H
N
O
N
H
N
N
N
O
N
N
N
+
c
N
+
N
N
N
N
OMe
OMe
OMe
OMe
14
15
12
13
Scheme 2. (a) H2NNH2, EtOH, 80 °C, 16 h; (b) acetic anhydride, NEt3, CH2Cl2; (c) p-TsCl, DBU, THF, 150 °C, 10 min, microwave.

J. E. Tellew et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7259–7264
7263
further increases the CRF1 receptor activity (8d, pKi = 8.1) while
decreasing the overall lipophilicity as measured by calculated log P
values as compared to compound 1. In an effort to maintain this
new level of reduced lipophilicity, the branching carbon atoms
were also incorporated into cyclopropyl 8e (pKi = 7.5) and cyclobu-
tyl 8f (pKi = 7.5) ring structures but it was clear that straight alkyl
chains provided higher affinity analogs.
In an effort to explore the nature of the interaction of the het-
erocycles with the CRF1 receptor, the orientation of the oxadiazole
heterocycle was also varied. Scheme 2 outlines the synthesis 1,3,4-
oxadiazole analogs 14 and 15. Conversion of ester intermediates 6d
and 7 to the hydrazides 10 and 11 was followed by acetylation to
provide the intermediates 12 and 13. Cyclization of the acylhydraz-
ides in the presence of p-TsCl and DBU under microwave condi-
tions provided the final compounds.
Additional analogs of compound 8d were prepared using the
procedure described in Scheme 1 with the objective of maintaining
the CRF1 receptor affinity but with lower lipophilicity. The bottom
region aromatic pyridine compound (20) was prepared as a direct
analog of compound 1 as well as the 2,4-dimethoxyphenyl deriva-
tive (21) to further drive down the lipophilicity but unfortunately
both analogs did not maintain the desired level of CRF1 receptor
affinity (pKi = 7.3).
Cl
HO
NH
a
N
N
O
N
N
N
N
Scheme 3 outlines the synthesis of a 1,2,4-oxaxidiazole analog
where the orientation of the ring has been reversed. Due to the re-
quired stability of the precursor nitrile (17) only the homologated
analog 19 was prepared. Analogous to Scheme 1, 2-aminopropanol
was used to displace the 7-chloro group of 5 which resulted in 16.
Activation of the hydroxyl group with methanesulfonyl chloride
followed by displacement of the mesylate group with sodium cya-
nide provided 17. Condensation with hydroxylamine followed by
1,2,4-oxadiazole formation by heating in the presence of dimethyl-
formamide dimethyl diacetal (DMFDMA) provided compound 19.
Biological evaluation of these additional analogs demonstrated
that exchange of the nitrogen and oxygen atoms within the hetero-
cycle, surprisingly, resulted in a 10-fold loss in affinity for the CRF1
receptor as seen in the 1,3,4-oxadiazolyl analog 14 (pKi = 7.0) com-
pared to 8d. A final avenue of top region investigation was to fur-
ther probe the area of chemical space occupied by the newly
identified heterocyclic side chain. The homologated analog 9
(pKi = 7.5) was less active but it was appreciated that this new ana-
log may significantly change how the heterocycle interacts with
the receptor. As a result, the orientation of the 1,2,4-oxadiazole
ring was reversed (compound 19) and the homologated 1,3,4-oxa-
diazole analog 15 was also prepared but both proved less active
with pKi values of 7.1 and 6.6, respectively.
O
O
5
(S)-22
b
NH2
N
O
N
NH
NH
O
O
c
N
N
N
N
N
N
N
O
O
(S)-23
(S)-8d
Scheme 4. (a) (S)-2-Aminobutyric acid, NaHCO3, dioxane/H2O, 100 °C, 14 h;
(b) methyl amidoxime, DIC, HOBt, CH2Cl2/DMF, À15 °C to room temperature;
(c) pyridine, 100 °C.
N
Cl
N
HO
NH
NH
N
N
N
a
b,c
N
N
N
N
N
OMe
OMe
OMe
5
16
17
d
HO
O
N
N
N
NH
H2N
NH
e
N
N
N
N
N
N
OMe
OMe
19
18
Scheme 3. (a) 2-Aminopropanol, NEt3, acetonitrile, 80 °C; (b) MsCl, NEt3, CH2Cl2; (c) NaCN, K2CO3, DMF; (d) NH2OHÁHCl, KOH, EtOH, reflux; (e) DMFDMA, neat, 110 °C.

7264
J. E. Tellew et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7259–7264
Compound 8d proved to be the best combination of CRF1
Acknowledgments
affinity and metabolic stability; consequently, the individual
enantiomers were prepared following the procedure outlined in
Scheme 4. Although the procedure as illustrated in Scheme 1 is
convenient, the protocol resulted in complete racemization of the
final compound when optically pure amino esters were utilized.
To avoid the troublesome ester hydrolysis step (S)- and (R)-2-ami-
nobutyric acids were substituted for the 2-aminobutyric acid
methyl ester in the addition to intermediate 5. Coupling of the
intermediates (R)-22 and (S)-22 with methyl amidoxime, followed
by cyclization under mildly basic15 conditions, provided final prod-
ucts (R)-8d and (S)-8d in high enantiomeric purity (>99% ee).16
Both enantiomers of 8d proved to be high affinity CRF1 receptor
antagonists with pKi values of 8.2 for (S)-8d and 8.0 for (R)-8d. Both
compounds proved to be metabolically stable (estimated max
F% = 42% and 63%, respectively) and an experimentally (shake
flask) determined log D7.4 of 3.8. The compounds were further
assayed for in vitro suppression of sauvagine-induced ACTH re-
lease in rat anterior pituitary cells11 and unexpectedly (S)-8d was
10-fold more active than its enantiomer with pIC50 values of 7.0
and 6.0, respectively. As a result of the functional activity, com-
pound (S)-8d was evaluated in a 24 h rat pharmacokinetic study
at an oral dose of 10 mg/kg. Compound (S)-8d demonstrated good
oral bioavailability (66%) and good exposure in both plasma
(AUC0–24 = 3130 ng h/mL) and brain (1 h brain/plasma ratio = 1.6).
A 72 h iv (5 mg/kg) study established that (S)-8d had moderate
clearance (14 mL/min kg) and a volume of distribution at steady
state (Vdss = 7.5 L/kg) within our target range. In addition, the hu-
man plasma protein binding for the compound was 94% as deter-
mined by equilibrium dialysis.
We thank Neurocrine and GSK Analytical departments as well a
Neurocrine Preclinical department for analytical measurements
and pharmacokinetic studies.
References and notes
1. Arborelius, L.; Owens, M. J.; Plotsky, P. M.; Nemeroff, C. B. J. Endocrinol. 1999,
160, 1.
2. Grigoriadis, D. E. Expert Opin. Ther. Targets 2005, 9, 651.
3. (a) Gilligan, P. J.; Li, Y.-W. Curr. Opin. Drug Discov. Devel. 2004, 7, 487; (b) Tellew,
J. E.; Luo, Z. Curr. Top. Med. Chem. 2008, 7, 506; (c) Dzierba, C. D.; Hartz, R. A.;
Bronson, J. J. Annu. Rep. Med. Chem. 2008, 43, 3.
4. (a) Holsboer, F.; Ising, M. Eur. J. Pharmacol. 2008, 583, 350; (b) Chen, C.;
Grigoriadis, D. E. Drug Dev. Res. 2005, 65, 216.
5. Ising, M.; Zimmermann, U. S.; Künzel, H. E.; Uhr, M.; Foster, A. C.; Learned-
Coughlin, S. M.; Holsboer, F.; Grigoriadis, D. E. Neuropsychopharmacology 2007,
32, 1941.
6. Binneman, B.; Feltner, D.; Kolluri, S.; Shi, Y.; Qiu, R.; Stiger, T. Am. J. Psychiatry
2008, 165, 617.
7. Coric, V.; Feldman, H. H.; Oren, D. A.; Shekhar, A.; Pultz, J.; Dockens, R. C.; Wu, X.;
Gentile, K. A.; Huang, S.-P.; Emison, E.; Delmonte, T.; D’Souza, B. B.;Zimbroff, D. L.;
Grebb, J. A.; Goddard, A. W.; Stock, E. G. Depress. Anxiety 2010, 27, 417.
8. Chen, C. Curr. Med. Chem. 2006, 13, 1261.
9. Chen, C.; Wilcoxen, K. M.; Huang, C. Q.; Xie, Y.-F.; McCarthy, J. R.; Webb, T. R.;
Zhu, Y. F.; Saunders, J.; Liu, X.-J.; Chen, T.-K.; Bozigian, H.; Grigoriadis, D. E. J.
Med. Chem. 2004, 47, 4787.
10. Calculator Plugins were used for Structure Property Prediction and Calculation,
Marvin 5.1.2 2008ChemAxon (http://www.chemaxon.com).
11. Gilligan, P. J.; Baldauf, C.; Cocuzza, A.; Chidester, D.; Zaczek, R.; Fitzgerald, L.
W.; McElroy, J.; Smith, M. A.; Shen, H.-S. L.; Saye, J.; Christ, D.; Trainor, G.;
Robertson, D. W.; Hartig, P. Bioorg. Med. Chem. 2000, 8, 181.
12. Gross, R. S.; Guo, Z.; Dyck, B.; Coon, T.; Huang, C. Q.; Lowe, R. F.; Marinkovic, D.;
Moorjani, M.; Nelson, J.; Zamani-Kord, S.; Grigoriadis, D. E.; Hoare, S. R. J.;
Crowe, P. D.; Bu, J. H.; Haddach, M.; McCarthy, J.; Saunders, J.; Sullivan, R.;
Chen, T.; Williams, J. P. J. Med. Chem. 2005, 48, 5780.
In summary, a series of novel and potent CRF1 receptor antago-
nists have been synthesized where analogs of 1 have heterocycles
replacing alkyl chains in the top region of the molecule which re-
sulted in analogs with reduced overall lipophilicity. The most
promising compound, (S)-8d (NBI-77860/GSK561679), possesses
a good pharmacokinetic profile and was selected as a candidate
for further preclinical investigation. Results of further preclinical
and clinical studies with this compound will be disclosed in future
publications.
13. Liaw, C. W.; Grigoriadis, D. E.; Lorang, M. T.; De Souza, E. B.; Maki, R. A. Mol.
Endocrinol. 1997, 11, 2048.
14. Hodge, C. N.; Aldrich, P. E.; Wasserman, Z. R.; Fernandez, C. H.; Nemeth, G. A.;
Arvanitis, A.; Cheeseman, R. S.; Chorvat, R. J.; Ciganek, E.; Christos, T. E.; Gilligan,
P. J.; Krenitsky, P.; Scholfield, E.; Strucely, P. J. Med. Chem. 1999, 42, 819.
15. Hamzé, A.; Hernandez, J.-F.; Fulcrand, P.; Martinez, J. J. Org. Chem. 2003, 68,
7316.
16. Enantiomeric excess was determined by normal-phase chiral HPLC using
Chiralcel OD columns with hexane/ethanol or hexane/isopropanol as eluant (in
the presence of diethylamine as an additive in some cases).