Identification and optimization of novel 2-(4-oxo-2-aryl-quinazolin-3(4H)- yl)acetamide vasopressin V3 (V1b) receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.07.118

The discovery, synthesis, and preliminary structure-activity relationship (SAR) of a novel class of vasopressin V3 (V1b) receptor antagonists is described. Compound 1, identified by high throughput screening of a diverse, three million-member compound collection, prepared using ECLiPS technology, had good activity in a V3 binding assay (IC₅₀ = 0.20 μM), but less than desirable physicochemical properties. Optimization of compound 1 yielded potent analogs 19 (IC₅₀ = 0.31 μM) and 24 (IC₅₀ = 0.12 μM) with improved drug-like characteristics.

Full text of DOI:10.1016/j.bmcl.2010.07.118

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5394–5397
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and optimization of novel
2-(4-oxo-2-aryl-quinazolin-3(4H)-yl)acetamide vasopressin
V3 (V1b) receptor antagonists
a
a
a
a
Jeffrey J. Letourneau a, , Christopher M. Riviello , Hong Li , Andrew G. Cole , Koc-Kan Ho ,
Heather A. Zanetakos a, Hema Desai a, Jiuqiao Zhao a, Douglas S. Auld a, Susan E. Napier b, Fiona J. Thomson b,
Katharine A. Goan b, J. Richard Morphy b, Michael H. J. Ohlmeyer a, Maria L. Webb a
*
a Ligand Pharmaceuticals, Inc., 3000 Eastpark Boulevard, Cranbury, NJ 08512, USA
b MSD, Newhouse, Lanarkshire ML1 5SH, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery, synthesis, and preliminary structure–activity relationship (SAR) of a novel class of vaso-
pressin V3 (V1b) receptor antagonists is described. Compound 1, identified by high throughput screening
of a diverse, three million-member compound collection, prepared using ECLiPS™ technology, had good
activity in a V3 binding assay (IC50 = 0.20 lM), but less than desirable physicochemical properties. Opti-
mization of compound 1 yielded potent analogs 19 (IC50 = 0.31 lM) and 24 (IC50 = 0.12 lM) with
Received 23 June 2010
Revised 21 July 2010
Accepted 26 July 2010
Available online 1 August 2010
improved drug-like characteristics.
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Vasopressin (V1b) receptor
Antagonist
HPA axis
Quinazolinone
A majority of patients suffering from severe depressive disorder
exhibit a profound alteration in their ability to regulate the hypotha-
lamic–pituitary–adrenal (HPA) axis—the major route via which hu-
mans (and other mammals) cope with, and adapt to stress.1 Arginine
vasopressin (AVP) is thought to play a pivotal role in HPA axis regu-
lation, particularly in situations of chronic stress.2 AVP mediates its
effects via binding to the V3 (V1b) receptor expressed upon pituitary
corticotropes. Through this mechanism, AVP acts synergistically
with corticotropin releasing hormone (CRH) to elicit release of adre-
nocorticotropic hormone (ACTH), thus positively driving the HPA
axis. Much evidence indicates that while CRH may be relevant in
orchestrating HPA responsivity to acute stress,3 AVP-induced activ-
ity of V3 receptors may be the more important modulator during
chronic stress. This in turn may be more clinically relevant to the
development of affective disorders.2,4
activity,5,6 and this compound advanced into phase IIb clinical
trials.7
In a program aimed at the identification of a novel and structur-
ally distinct class of V3 antagonists, high-throughput screening
(HTS) of a diverse three million-member compound collection, pre-
pared using ECLiPS™ technology,8 was undertaken. The com-
pounds were screened utilizing a 384-well whole cell binding
displacement assay using tritium labeled arginine vasopressin
([3H]-AVP).9 Multiple active structures, or hits, were identified in
this HTS campaign. Among these hits was compound 1 (Fig. 1),
possessing an IC50 value of 0.20 lM. Compound 1 was chosen as
a starting point for a hit-to-lead (HtL) program, with the task of
evaluating this hit class in terms of both target independent (e.g.,
physicochemical properties) and target dependent (e.g., potency,
selectivity, etc.) criteria.10
Therefore, a selective antagonist of the V3 receptor represents a
therapeutically relevant approach for treatment of the dysfunc-
tional HPA axis in depressive illness. In fact, a previously reported
V3 selective antagonist, SSR149415, has been demonstrated to be
active in animal models predictive of antidepressant and anxiolytic
The main objective of this HtL program was to generate a lead
compound with appropriate potency and drug-like physicochemi-
cal properties11 to serve as the basis for further lead optimization
efforts. One immediate concern about compound 1, was that from
a perspective of ‘drug likeness’ it possessed less than optimal phys-
icochemical properties such as high molecular weight (MW), high
‘fast polar surface area’ (FPSA), a high degree of lipophilicity, and
* Corresponding author. Tel.: +1 609 452 3748; fax +1 609 452 3671.
E-mail address: jletourneau@ligand.com (J.J. Letourneau).
In 2008, Pharmacopeia, Inc. was acquired by Ligand Pharmaceuticals, Inc.
a
high degree of conformational flexibility (MW = 613,
FPSA = 148.48 Å2, A log P = 4.73, # of rotatable bonds = 17).11,12
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.118

J. J. Letourneau et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5394–5397
5395
Table 2
HO
Exploration of chain length between amino functionality and scaffold
N
N
O
O
O
Cl
N
O
O
O
HN
O
N
HN
O
N
O
N
O
N
S
O
O
N
NH2
n
n
N
O
SSR149415
13-15
8-12
a
Compds
n
hV3 IC50 (lM)
HN
O
O
N
O
O
NH2
8
9
1
2
3
4
5
1
2
3
4.0 ( 1.6)
3.6 ( 0.2)
>30
N
N
H
10
11
12
13
14
15
>30
OH
1
>100
22b
Figure 1. V3 antagonists.
1.7 ( 0.6)
4.8 ( 1.0)
a
Values are means of at least two experiments unless otherwise indicated,
A calculation of the binding energy of compound 1 by Andrew’s
analysis13 gave
significantly higher binding energy
Gcalc = 18.3 kcal/mol) compared to the experimental value
Gobs = 9.2 kcal/mol). This large difference suggested that not all
standard deviation is given in parentheses.
a
b
Value was obtained from a single experiment.
(
D
(D
parts of the molecule were optimally involved in binding to the
receptor. This indicated the possibility for deletions and alterations
to the structure withouta major decrease in binding affinity. We rea-
soned that the molecule could be greatly simplified, by strategic
deletion of unnecessary structural elements, to deliver a potent
and tractable lead compound with improved physicochemical prop-
erties. In this report, we describe the synthesis and SAR of novel V3
antagonists resulting in the identification of potent lead compounds
with improved physicochemical properties versus compound 1.
The compounds detailed in Tables 1–3 were synthesized via the
outlined general syntheses (Schemes 1–3). The synthesis of the
Table 3
N-Methylpiperazine replacements
HN
O
N
O
N
R1
N
O
R2
16-24
NR1R2
hV3 IC50 (lM)
a
Compds
NH
16
25.8 ( 2.1)
N
*
Table 1
Simplification/truncation of right-hand side
17
18
NHMe
N(Et)2
7.6 ( 0.6)
3.8 ( 1.9)
H
N
H
19
20
21
0.31 ( 0.15)
1.8 ( 1.0)
N
*
O
N
N
O
N
O
N
R1
O
O
R2
OH
N
N
*
H
N
0.34 ( 0.11)
N
*
1, 3, 5-7
2
O
R1
R2
hV3 IC50 (lM)
a
22
0.71 ( 0.13)
1.1 ( 0.02)
Compds
N
*
O
OH
NH2
23
24
*
*
N
*
*
*
1
0.2 ( 0.07)
O
0.12 ( 0.046)
OH
N
*
O
a
NH2
NH2
5
6
5.1 ( 3.3)
Values are means of at least two experiments, standard deviation is given in
parentheses.
O
O
H
H
30.5 ( 0.7)
*
*
penultimate phenol intermediate
elsewhere.14
2
has been described
7
>100
As depicted in Scheme 1, a Mitsunobu reaction15 between phe-
nol 2 and either Boc-L-leucinol or Boc-2-aminoethanol followed by
OH
Boc deprotection gave amines 3 and 4, respectively. Amine 3 was
then coupled with Boc-protected tyrosine which, followed by Boc
deprotection, yielded compound 1. In a similar fashion both
amines 3 and 4 were coupled with Boc-protected glycine and
deprotected to yield compounds 5 and 6. Amine 4 was also coupled
with 3-(4-hydroxyphenyl)propanoic acid giving compound 7. As
2
3
>100
H
2.1 ( 0.1)
*
a
Values are means of at least two experiments, standard deviation is given in
parentheses.

5396
J. J. Letourneau et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5394–5397
H
H
H
N
O
N
N
O
N
N
O
N
R1
O
N
R1
O
O
N
R2
O
OH
O
d,c; e,c
or f
a,c
NH2
N
H
N
or b,c
3; R1 = i-Bu
1, 5-7
2
4; R1 = H
Scheme 1. Reagents and conditions: (a) N-Boc-2-aminoethanol, DIAD, PPh3, THF, 0–23 °C; (b) N-Boc-
L-leucinol, DIAD, PPh3, THF, 0–23 °C; (c) 50% TFA/DCM; (d) N-Boc-L-
tyrosine, EDCI, HOBt, DMF; (e) Boc-glycine, EDCI, HOBt, DMF; (f) 3-(4-hydroxyphenyl)propanoic acid, EDCI, HOBt, DMF.
inherent in the tyrosine, and glycine residues, might be required
for activity. This hypothesis was further strengthened by the fact
that the synthetic intermediate 3 (see Scheme 1), which also con-
tains a basic primary amine, was found to possess moderate activ-
HN
O
N
O
N
O
NH2
a,b
OH
n
ity (IC50 = 2.1 lM).
HN
O
N
O
Next, a series of aminoalkoxy analogs with an amino (NH2) or N-
methylpiperazine substituent at varying distances from the quinaz-
olinone scaffold were prepared to determine the preference in chain
length between the amine functionality and the quinazolinone scaf-
fold. As shown in Table 2, in both the primary amine and N-methyl-
8-12
N
c,d
2
piperazine series,
established (compounds 9 and 14). In this set of analogs, compound
14 had the best activity (IC50 = 1.7 M) and represented a significant
a preference of three carbon atoms was
HN
O
O
l
O
N
N
advancement in terms of improved physicochemical properties.
Compared with compound 1, compound 14 had a reduced molecular
weight (from 612 to 490), a reduced fast polar surface area (from
n
N
N
13-15
FPSA = 148.48 Å2 to FPSA = 77.72 Å2),
a reduced A log P (from
Scheme 2. Reagents and conditions: (a) HO(CH2)nNHBoc, DIAD, PPh3, THF, 0–23 °C;
(b) 50% TFA/DCM; (c) Br(CH2)nBr, K2CO3, CH3CN, reflux; d) N-methylpiperazine,
K2CO3, CH3CN, reflux.
A log P = 4.73 to A log P = 2.72), and a reduced number of rotatable
bonds (from 17 to 11). Consequently compound 14 was the basis
for further modifications to improve potency.
Next, analogs of compound 14 were prepared in which the N-
methylpiperazine group was replaced with other amines (Table
3). Demethylation of the piperazine, to give 16, led to a substantial
loss of potency. Analog 17 (IC50 = 7.6 lM), containing a secondary,
N-methyl amine, was twofold less active than both the primary
HN
O
N
O
R1
N
HN
O
N
O
N
O
a,b
R2
OH
amine 9 (i.e., –NH2, see Table 2) and the tertiary, N,N-diethylamine
N
18. The piperidine 19 (IC50 = 0.31
crease in potency compared to N-methyl piperazine 14
(IC50 = 1.7 M), demonstrating that the presence of the distal, basic
lM) resulted in a fivefold in-
2
16-24
l
Scheme 3. Reagents and conditions: (a) 1,3-dibromopropane, K2CO3, CH3CN,
piperazine nitrogen was not optimal for activity. The piperidine 19
was also about 10-fold more active than the N,N-diethylamine 18,
indicating a preference for cyclic amines. Pyrrolidine 20 was six-
fold less potent than piperidine 19 whereas homopiperidine 21
was better tolerated, but did not improve potency compared to
19. Introduction of a methoxy (compound 22) or hydroxyl (com-
pound 23) group at the 4-position of the piperidine ring in 19 re-
sulted in a 2- to 4-fold decrease in potency. Finally, replacement
of piperidine with morpholine, yielding compound 24, resulted in
a further threefold increase in potency to give the most potent ana-
log (IC50 = 120 nM) discussed herein.
reflux; (b) HNR1R2, K2CO3, CH3CN, reflux.
depicted in Scheme 2, Mitsunobu reaction of the phenol 2 with the
appropriate Boc-protected hydroxyalkylamines followed by Boc
deprotection yielded compounds 8–12. Alternatively, alkylation
with the appropriate dibromoalkanes followed by reaction with
N-methylpiperazine yielded compounds 13–15. Finally, as shown
in Scheme 3, phenol 2 was alkylated with 1,3-dibromopropane fol-
lowed by reaction with the appropriate amines to give compounds
16–24.
The rat V3, and human V1a, V2, and oxytocin affinities were also
determinedfor compound24(Table 4). Affinityat theratV3 receptor
was comparable (within twofold) versus the human V3 receptor.
This indicated that further optimization of this hit class was
likely to deliver tool compounds for evaluation in rat in vivo mod-
The compounds in Tables 1–3 were evaluated for their ability to
displace the binding of [3H]-AVP to human recombinant V3 recep-
tor expressed in CHO cells.9 In order to gain insight into the struc-
tural requirements for activity in the peptidic, right-hand portion
of the molecule, a series of strategic deletions within this domain
was performed (Table 1). Removal of the tyrosine sidechain, as in
compound 5, gave an active structure (IC50 = 5.1 lM), but with a
Table 4
26-fold decrease in potency versus compound 1. Further removal
of the isobutyl sidechain led to an additional sixfold loss of potency
Evaluation of compound 24 for affinity at rat (r) V3 and human (h) V1a, V2 and
oxytocin receptors versus affinity at human V3 receptor
Compd
IC50a
(lV)
(compound 6; IC50 = 30.5 lM). Interestingly, incorporation of the
tyrosine sidechain back into the molecule with simultaneous dele-
tion of the amine functionality gave compound 7 which was inac-
tive in the assay. Finally, removal of the complete dipeptidic
sidechain, resulting in phenol 2, led to an inactive compound. This
preliminary SAR indicated that the basic amine functionality,
hV3 (V1b)
rV3
0.21 ( 0.016)
hV1a
hV2
hOxt
>100
24
0.12 ( 0.046)
6% inh. at 5
l
M
>100
a
Values are means of at least two experiments, standard deviation is given in
parentheses.

J. J. Letourneau et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5394–5397
5397
CO2 in DMEM/F12 (Invitrogen) with 10% FBS (HyClone Labs, FetalClone II,
#206007), 0.4 mg/mL G418, and 1% penicillin and streptomycin. Cells are
els. In addition, compound 24 demonstrated excellent selectivity
(>1000-fold) for the human V3 receptor versus the related human
V1a, V2, and oxytocin receptors.16,17
seeded at 30,000 cells into either 384-well (50
well) clear bottom plate and cultured at 37 °C overnight. Cells are washed
twice with 100 L/well (for 384-well plate) or 200 L/well (for 96-well plate)
of PBS. For initial HTS, 384-well plates containing dried down compounds are
resuspended with 30
L/well of 5 nM [3H]-AVP in assay buffer (PBS with
10 mM MgCl2 and 0.1% BSA) with 1% DMSO. After solubilization, 20 L/well is
lL/well) or 96-well (100 lL/
l
l
In conclusion, hit-to-lead efforts starting from the initial library
hit, compound 1, led to the identification of highly active lead com-
pounds 19 and 24. Importantly, a significant improvement in phys-
icochemical properties versus hit compound 1 was achieved (e.g.,
compound 24: MW = 477, A log P = 2.45, FPSA = 83.30 Å2 versus
compound 1: MW = 612, A log P = 4.73, FPSA = 148.48 Å2). In addi-
tion, compound 24 was demonstrated to have comparable affinity
for the rat V3 receptor versus the human V3 receptor, and an excel-
lent selectivity profile versus the related human V1a, V2, and oxy-
tocin receptors. Further optimization within this series to provide
highly selective V3 antagonists with improvements in affinity
and with in vivo activity will be reported in due course.
l
l
transferred from compound plate to cell plate. For compound evaluation, test
compounds are initially serially diluted in DMSO and then intermediately
diluted at 1:50 with assay buffer (PBS with 10 mM MgCl2 and 0.1% BSA). To
each well, 25 lL of serially diluted compounds are added followed by 25 lL of
10 nM [3H]-AVP in assay buffer. The final reaction contains 5 nM [3H]-AVP and
1% DMSO. The mixtures are incubated at room temperature for 30 min. The
plates are washed twice with 100
lL/well (for 384-well plate) or 200 lL/well
(for 96-well plate) of PBS and air dried. Detection is performed by adding 50
lL
of scintillation fluid and shaking for 30 min followed by counting at 1 min/well
in the microbeta.
10. (a) Davies, A. M.; Keeling, D. J.; Steele, J.; Tomkinson, N. P.; Tinker, A. C. Curr.
Top. Med. Chem. 2005, 5, 421; (b) Alanine, A.; Nettekoven, M.; Roberts, E.;
Thomas, A. W. Comb. Chem. High Throughput Screening 2003, 6, 51; (c) Michne,
W. F. Pharm. News 1996, 3, 19.
11. (a) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Delivery
Rev. 1997, 23, 3; (b) Egan, W. J.; Merz, K. M.; Baldwin, J. J. J. Med. Chem. 2000, 43,
3867; (c) Veber, D. F.; Johnson, S. R.; Cheng, H. Y.; Smith, B. R.; Ward, K. W., et al
J. Med. Chem. 2002, 45, 2615.
Acknowledgments
Cell lines for this research were gratefully received from Orga-
non Laboratories Ltd (part of MSD).
12. Physicochemical properties calculated using ADME Profiler: Program for
discreet compound analysis and calculation of physicochemical properties,
version 1.6.0; Pharmacopeia Inc., 2004.
References and notes
13. Andrews, P. R.; Craik, D. J.; Martin, J. L. J. Med. Chem. 1984, 27, 1648.
14. Letourneau, J.; Riviello, C.; Ho, K.-K.; Chan, J.-H.; Ohlmeyer, M.; Jokiel, P.;
Neagu, I.; Morphy, J. R.; Napier, S. E. 2-(4-oxo-4H-quinazolin-3-yl)Acetamides
and their use as vasopressin V3 antagonists. PCT Int. Appl. WO 06095014,
2006.
15. First report: (a) Mitsunobu, O.; Yamada, M.; Mukaiyama, T. Bull. Chem. Soc. Jpn.
1967, 40, 935; For reviews, see: (b) Mitsunobu, O. Synthesis 1981, 1; (c) Hughes,
D. L. Org. React. 1992, 42, 335; (d) Hughes, D. L. Org. Prep. Proced. Int. 1996, 28,
127.
1. (a) Holsboer, F.; Barden, N. Endocr. Rev. 1996, 17, 187; (b) Barden, N.; Reul, J. M.
H. M.; Holsboer, F. TINS 1995, 18, 6.
2. (a) Scaccianoce, S.; Muscolo, L. A. A.; Cigliana, G.; Navarra, D.; Nicolai, R.;
Angelucci, L. Endocrinology 1991, 128, 3138; (b) Ma, X.-M.; Levy, A.; Lightman,
S. L. Endocrinology 1997, 138, 4351.
3. (a) Pitts, A. F.; Samuelson, S. D.; Meller, W. H.; Bissette, G.; Nemeroff, C. B.;
Kathol, R. G. Biol. Psych. 1995, 38, 330; (b) Nemeroff, C. B. Pharmacopsychiatry
1988, 21, 76.
4. Purba, J. S.; Hoogendijk, W. J. G.; Hofman, M. A.; Swaab, D. F. Arch. Gen.
Psychiatry 1996, 53, 137.
5. (a) Griebel, G.; Simiand, J.; Serradeil-Le Gal, C.; Wagnon, J.; Pascal, M.; Scatton,
B.; Maffrand, J.-P.; Soubrie, P. PNAS 2002, 99, 6370; (b) Serradeil-Le Gal, C.;
Wagnon, J.; Tonnerre, B.; Roux, R.; Garcia, G.; Griebel, G.; Aulombard, A. CNS
Drug Rev. 2005, 11, 53.
16. Human V2, oxytocin (OT), and rat V3 (V1b) receptors whole cell binding assays
were performed similarly to human V3 (V1b) whole cell assay in 96-well
format except that CHO cells stably expressing human V2, oxytocin (OT), and
rat V3 (V1b) receptors were used, respectively. The final reactions contain
5 nM [3H]-AVP for human V2 and rat V3 (V1b) receptor whole cell binding
assays, and 5 nM [3H]-oxytocin for human oxytocin receptor whole cell
binding assay.
6. For
a recent review describing various vasopressin receptor agonists and
antagonists, including SSR149415, see: Ryckmans, T. Annu. Rep. Med. Chem.
2009, 44, 129.
17. Human V1a membrane binding assay performed in 96-well microtitre plates
containing (final concentrations): 50
CHO cells stably expressing human V1a receptor, 5 nM [3H]-AVP and test
compounds in a total volume of 100 L of assay buffer (50 mM Tris–HCl pH 7.4,
lg/well membrane protein prepared from
7. ClinicalTrials.gov
web-based
search.
http://clinicaltrials.gov/ct2/search
(accessed Jan 2008), 3 phase II trials listed using search term ‘SSR149415’.
8. (a) Baldwin, J. J.; Burbaum, J. J.; Henderson, I.; Ohlmeyer, M. H. J. J. Am. Chem.
Soc. 1995, 117, 5588; (b) Ohlmeyer, M. H. J.; Swanson, R. N.; Dillard, L. W.;
Reader, J. C.; Asouline, G.; Kobayashi, R.; Wigler, M.; Still, W. C. Proc. Natl. Acad.
Sci. U.S.A. 1993, 90, 10922; (c) Nestler, H. P.; Bartlett, P. A.; Still, W. C. J. Org.
Chem. 1994, 59, 4723.
9. V3 (V1b) whole cell binding assay: The initial HTS is performed in 384-well
format whereas the compound evaluation is done in 96-well format. The CHO
cells stably expressing human V3 (V1b) receptor are cultured at 37 °C with 5%
l
5 mM MgCl2, 1 mg/mL BSA, and 0.5% DMSO). The assay mixture was incubated
at room temperature for 60 min. The reaction was terminated by rapid
filtration of the mixture over 96-well GF/B filters (pre-soaked in 0.3%
polyethylenimine) using a Tomtec Harvester. The filters were washed four
times with approximately 200
dried at 55 °C for 30 min. 50
filters counted using a Packard Topcount.
l
L of ice-cold 50 mM Tris–HCl (pH 7.4), then
l
L scintillation fluids were then added and the