Potent and selective oxindole-based vasopressin 1b receptor antagonists with improved pharmacokinetic properties

Article abstract of DOI:10.1016/j.bmcl.2011.03.012

Herein we report the discovery of a novel series of vasopressin 1b (V1b) receptor antagonists, starting from potent but metabolically labile oxindole SSR149415. Masking the proline N,N-dimethyl amide moiety as an oxazole and attaching a benzylic amine moiety to the northern phenyl ring resulted in potent and selective V1b receptor antagonists with improved metabolic stability and improved pharmacokinetic properties in rat. Compound 18c was found to be efficacious in a rat model of anti-depressant activity.

Full text of DOI:10.1016/j.bmcl.2011.03.012

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3828–3831
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Potent and selective oxindole-based vasopressin 1b receptor antagonists
with improved pharmacokinetic properties
⇑
Thorsten Oost, Gisela Backfisch, Swati Bhowmik, Marcel M. van Gaalen, Hervé Geneste ,
Wilfried Hornberger, Wilfried Lubisch, Astrid Netz, Liliane Unger, Wolfgang Wernet
Abbott, Neuroscience Discovery, D-67008 Ludwigshafen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Herein we report the discovery of a novel series of vasopressin 1b (V1b) receptor antagonists, starting
from potent but metabolically labile oxindole SSR149415. Masking the proline N,N-dimethyl amide moi-
ety as an oxazole and attaching a benzylic amine moiety to the northern phenyl ring resulted in potent
and selective V1b receptor antagonists with improved metabolic stability and improved pharmacokinetic
properties in rat. Compound 18c was found to be efficacious in a rat model of anti-depressant activity.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 19 January 2011
Revised 1 March 2011
Accepted 2 March 2011
Available online 6 March 2011
Keywords:
Vasopressin 1b
V1b
SSR149415
Oxindole
Metabolic stability
The nine-amino acid neuropeptide arginine vasopressin plays a
critical role in the regulation of the hypothalamic–pituitary–adre-
nal axis during stress and is involved in various emotional
processes. Oxindole SSR149415, the first potent, selective, orally-
active vasopressin V1b receptor antagonist disclosed (Fig. 1),
showed anxiolytic and anti-depressant-like effects in several
rodent behavioral models.1 Blockade of V1b receptors may thus
represent a novel mechanism for the treatment of depression
and anxiety disorders. SSR149415 entered phase II trials for
depression and anxiety but was discontinued later on.2
We and other groups found that SSR149415 exhibits a sub-opti-
mal pharmacokinetic profile in rat.3 Specifically, it displayed short
half-life (<1 h) and high plasma clearance in rats, which is consis-
tent with the high intrinsic clearance observed in liver microsomes
(Table 1), indicating extensive first-pass metabolism.
In order to improve the sub-optimal pharmacokinetic profile,
we sought to increase the metabolic stability of analogs from the
oxindole series. One strategy to improve metabolic stability in a
given chemical series is to reduce the overall lipophilicity. This
sometimes works because the binding site of metabolizing CYP en-
zymes is generally lipophilic in nature, thus possessing an affinity
for lipophilic substrates.4 Given this working hypothesis, our initial
efforts were aimed at replacing the phenyl rings in SSR149415
with less lipophilic heteroaryl rings, such as pyridine, pyrazine,
pyrimidine, etc. Analogs in which the 3-phenyl ring is replaced
with a heteroaryl ring were prepared starting from 5-chloro isatin
by addition of lithiated heterocycles5 (Scheme 1). After chlorina-
tion in the 3-position with thionyl chloride, the proline moiety
was introduced via nucleophilic substitution with trans-
hydroxy-L-proline N,N-dimethyl amide or L-proline N,N-dimethyl
amide. Final compounds were obtained by sulfonylation with
2,4-dimethoxyphenylsulfonyl chloride. Diastereomers were either
separated after the amination or after the sulfonylation stage.
Replacement of the 2-methoxyphenyl in SSR149415 with the 2-
methoxypyridin-3-yl moiety led to compound 3b, which exhibited
equipotent nanomolar binding to the human vasopressin 1b
(hV1b) receptor, moderate selectivity for hV1b versus closely re-
lated vasopressin 1a (hV1a) and oxytocin (hOT) receptor and high
selectivity for hV1b versus vasopressin 2 (hV2) receptor (Table 1).
The proline analog 3c lacking the hydroxy group in 3b remained a
potent antagonist of the hV1b receptor. The 4-methoxypyridin-3-
yl analog 3d and the 2,4-dimethoxypyrimidin-5-yl analog 3f were
OH
O
Cl
N
N
O
O
N
O S O
O
⇑
Corresponding author. Tel.: +49 621 589 3423; fax: +49 621 589 63423.
E-mail address: herve.geneste@abbott.com (H. Geneste).
O
Deceased.
Figure 1. SSR149415 (absolute configuration according to literature Ref. 1).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.03.012

T. Oost et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3828–3831
3829
Table 1
Receptor profilea and metabolic stabilityb of analogs from the proline series
O
O
Compounds hV1b Ki
hV1a Ki
(nM)
hV2 Ki
(nM)
hOT Ki
(nM)
Microsomal
stability mCl
H
O
O
O
(nM)
iii
i, ii
O
(l
L/min/mg)
Cl
OH
O
HO
Human Rat
O
N
H
5
I
I
SSR149415
1.7
94
290
25
343
432
4
(3a)
3b
3c
3d
3e
1.3
3.8
2.1
26
55
219
49
500
173
618
42
62
nd
271
45
323
359
158
302
203
446
446
413
1030
366
R1
O
O
O
>1000
>1000
>1000
636
O
O
vi, vii
iv, v
Cl
Cl
Cl
3f
5.6
N
N
a
N
H
6
N
H
7
O
hV1b Ki, hV1a Ki, hV2 and hOT Ki values were determined according to literature
O
O
N
N
procedures6 and are means of at least two experiments.
b
Microsomal intrinsic clearance values were determined according to literature
R1
procedures7 and are means of at least two experiments.
4
5
R1
N
O
O
viii
Cl
N
N
R1
N
O
N
O
O S O
R1
R2
O
O
Cl
i
Cl
O
N
O
R2
OH
O
O S O
R2
N
H
N
H
3a
3b
3c
3d
3e
OH
O
*
*
*
*
1a-e
O
O
N
N
R1
R2
OH
H
R1
R2
R1
R2
O
O
O
O
ii, iii
Cl
N
8a
pyrrolidin-1-yl
pyrrolidin-1-yl
OMe
H
9a
9b
9c
pyrrolidin-1-yl
pyrrolidin-1-yl
4-Me-piperazin-1-yl
OMe
H
8b
8c
N
O
O
N
H
4-Me-piperazin-1-yl
H
H
2a-e
N
Scheme 2. Synthesis of analogs 8a–8c from the benzyl amine–proline dimethyla-
mide series: (i) NaH, MeI, DMF; (ii) neopentylglycol, amberlyst-15, toluene reflux;
(iii) n-BuLi; sodium salt of 5-chloroisatin (prepared by addn of NaH in THF); (iv)
SOCl2, pyridine, CH2Cl2; (v) Pro-NMe2, DIPEA, CH2Cl2; (vi) 2 N aq HCl, CH2Cl2/
acetone; (vii) amine R1-H; MP-triacetoxyborohydride resin, THF, separation of
diasteromers; (viii) NaH, DMF, sulfonyl chloride. Analogs 9a–9e bearing the benzyl
amine moiety in the 5-position were prepared in analogous fashion starting from 3-
bromo-4-methoxybenzaldehyde (see Scheme 4).
OH
OH
R1
R2
iv
Cl
N
N
N
N
O S O
O
N
N
O
N
*
O
O
OH
3f
O
3a-e
O
*
deprotection via reductive amination. Separation of diastereomers
was accomplished at this stage and the final compounds were ob-
tained by sulfonylation.
Scheme 1. Synthesis of analogs from the proline series: (i) (a) 2-methoxy-
phenylmagnesium bromide; (b) and (c) bromomesitylene, tert-butyllithium, 2-
methoxypyridine; (d) bromo-mesitylene, tert-butyllithium, 4-methoxy-pyridine;
(e) LTMP, 2-methoxy-pyrazine; (f) 5-bromo-2,4-dimethoxy-pyrimidin, n-BuLi; (ii)
SOCl2, pyridine, CH2Cl2; (iii) Hyp-NMe2 or Pro-NMe2, DIPEA, CH2Cl2/THF, (iv) NaH,
DMF, 2,4-dimethoxyphenylsulfonyl chloride. Separation of diastereomers after step
(iii) (2a, 2b) or (iv) (3c–3f).
The first analogs from the benzyl amine–proline dimethyl
amide series bearing a pyrrolidin-1-ylmethyl group in the 4- or
5-position of the northern phenyl ring (8a and 9a) exhibited sub-
nanomolar binding to the human V1b receptor (Table 2), indicating
that the northern benzyl amine appendages are well tolerated by
the hV1b receptor. Interestingly, switching to the 4-methoxyphe-
nylsulfonamide resulted in antagonists (8b to 8c and 9b to 9c) with
double-digit picomolar affinity and >100-fold selectivity for hV1b
versus hV1a, hV2 and hOT receptor. Remarkably, benzyl amine
also potent single-digit nanomolar antagonists of the hV1b recep-
tor, while the 2-methoxypyrazin-3-yl analog 3e showed only mod-
erate binding affinity. Despite the reduced lipohilicity of the
heteroaryl analogs 3b–3f, the metabolic stability, as measured by
incubation with human and rat liver microsomes, was as poor as
for the phenyl analog SSR149415 (Table 1).
Table 2
Receptor profilea and metabolic stabilityb of analogs from the benzyl amine–proline
dimethylamide series
Another approach to reducing the lipophilicity of a molecule is
to attach an extra amine group that is charged under physiological
conditions. Initially, introducing extra basic amines in the western,
southern and eastern portion of the molecule only resulted in ana-
logs with dramatically decreased V1b binding affinity (data not
shown). Eventually, we decided to examine the effect of benzylic
amino groups on the northern phenyl ring. Analogs from the ben-
zyl amine–proline dimethylamide series were synthesized starting
from 3-hydroxy-4-iodobenzaldehyde or 3-bromo-4-methoxybenz-
aldehyde (benzyl amine in the 4- or 5-position, respectively) as
illustrated in Scheme 2. After protecting the aldehyde as dioxolane,
the organometallic species was formed via lithium–halide ex-
change. Addition to the sodium salt of 5-chloroisatin yielded the
3-hydroxy oxindole 5, which was converted to the proline interme-
diate 6 as described before. The amino group was introduced after
Compounds hV1b Ki
hV1a Ki
(nM)
hV2 Ki
(nM)
hOT Ki
(nM)
Microsomal
stability mCl
(nM)
(
lL/min/mg)
Human Rat
8a
8b
8c
9a
9b
9c
0.81
0.08
0.04
0.26
0.03
0.03
>1000
101
15
74
10
>1000
>1000
>1000
>1000
>1000
>1000
45
28
7
29
17
20
155
127
378
394
506
414
117
106
56
148
101
78
13
a
hV1b Ki, hV1a Ki, hV2 and hOT Ki values were determined according to litera-
ture procedures6 and are means of at least two experiments.
b
Microsomal intrinsic clearance values were determined according to literature
procedures7 and are means of at least two experiments.

3830
T. Oost et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3828–3831
N
N
O
O
O
O
O
ii, iii
O
Cl
i
Cl
Cl
O
O
OH
O
N
ii
i
O
H
O
Cl
N
H
N
H
11a-f
N
H
12a-f
O
R1
OH
O
O
O
N
H
Br
Br
N
N
14
15
O
iv
Cl
O
O
N
R1
R2
O
O
O
13a
( )-phenyl
OMe
OMe
O
R1
R2
13b
13c
13d
13e
13f
( )-pyridin-2-yl
O S O
O
O
Cl
Cl
Cl
v
iii, iv
( )-thiophen-2-yl OMe
N
N
( )-thiazol-2-yl
( )-oxazol-2-yl
(S)-oxazol-2-yl
OMe
OMe
H
N
H
O
N
O
N
N
O
O S O
O
13a-f
16
17
O
R
Scheme 3. Synthesis of analogs 13a–13f bearing heterocyclic amide replacements:
(i) bromomesitylene, tert-butyllithium, 2-methoxypyridine; (ii) SOCl2, pyridine,
CH2Cl2; (iii) 2-substituted pyrroldine, DIPEA, CH2Cl2/THF, synthesis of 2-(S)-
pyrrolidin-2-yl-oxazol hydrobromide according to Ref. 8; (iv) NaH, DMF, sulfonyl
chloride. Separation of diastereomers after (iii) or (iv).
R
O
vi, vii
N
18a
18b
pyrrolidin-1-yl
NMe2
N
O
N
O
18c
NHMe
O S O
18d
4-Me-piperazin-1-yl
substituents of different sizes were tolerated equally well in the 4-
and 5-position of the northern phenyl ring. Analogs 8c and 9c
showed improved metabolic stability in rat liver microsomes, but
lower stability in human microsomes.
O
Scheme 4. Synthesis of analogs 18a–18d from the benzyl amine–pyrrolidin-2-yl-
oxazole series: (i) neopentylglycol, amberlyst-15, toluene reflux; (ii) Mg, THF;
sodium salt of 5-chloroisatin; (iii) SOCl2, pyridine, CH2Cl2; (iv) 2-(S)-pyrrolidin-2-yl-
oxazol hydrobromide,8 DIPEA, CH2Cl2; separation of diastereomers (v) NaH, DMF, 4-
methoxyphenylsulfonyl chloride; (vi) 2 N aq HCl, acetone; (vii) amine R1-H; MP-
triacetoxyborohydride resin, THF.
In order to obtain a better understanding of the structural ele-
ments that are susceptible to oxidative CYP metabolism, we
decided to identify the metabolites in the proline series. Incubation
of 3b with human and rat liver microsomes revealed the proline
N,N-dimethylamide moiety as the major metabolic liability. In par-
ticular, it was found that one of the N-methyl groups is being oxi-
dized and cleaved (data not shown). Therefore, we attempted to
remove the metabolically susceptible site by replacing the N,N-
dimethylamide with a variety of potential bioisosteres, such as
small heterocycles. The synthesis of heterocyclic derivatives 13a–
13f was performed in analogous fashion to the corresponding pro-
line derivatives, employing 2-substitued pyrrolidines (Scheme 3).
While the phenyl and thiophene analogs 13a and 13c lacked affin-
ity to the hV1b receptor, the pyridine, thiazole and oxazole analogs
13b, 13d and 13e retained sub-micromolar activity, with the oxa-
zole analog being the most potent one (Table 3). Similar to the ben-
zyl amine–proline series, affinity could be further improved by
employing the 4-methoxyphenyl instead of the 2,4-dimethoxy-
phenyl sulfonamide, and the single stereoisomer 13f exhibited
20 nM binding affinity to the hV1b receptor. Encouragingly, the
pyrrolidin-2-yl-oxazole 13f was found to be moderately stable in
human (but not in rat) microsomes.
proline series. We further speculated that combination of the more
oxazole bioisostere (leading to improved metabolic stability in hu-
man microsomes) with the benzyl amine moiety (leading to im-
proved metabolic stability in rat microsomes) may result in
compounds with improved metabolic stability in both species.
The synthesis of the analogs from the benzyl amine–pyrrolidin-
2-yl-oxazole series was performed in analogous fashion to the cor-
responding proline derivatives, replacing L-proline N,N-dimethyl
amide with 2-(S)-pyrrolidin-2-yl-oxazol8 (Scheme 4). To maximize
synthetic efficiency, the reductive amination diversity step was
moved to the end.
Analogs 18a–18d bearing the benzyl amine appendage in the 5-
position exhibited potent single-digit nanomolar binding to the
hV1b receptor with overall good selectivity versus hV1a, hV2 and
hOT receptors. Gratifyingly, methylamine analog 18c and pipera-
zine analog 18d exhibited moderate metabolic stability in human
as well as in rat liver microsomes, prompting further in vivo phar-
macokinetic evaluation (Table 4).
In order to increase the moderate hV1b binding in the pyrroli-
din-2-yl-oxazole series, we wanted to take advantage of the dra-
matic affinity boost that was observed in the benzyl amine–
Table 4
Table 3
Receptor profilea and metabolic stabilityb of analogs from the benzyl amine–
pyrrolidin-2-yl-oxazole series
V1b affinitya and metabolic stabilityb of analogs bearing heterocyclic amide
replacements
Compds hV1b Ki
hV1a Ki
(nM)
hV2 Ki
(nM)
hOT Ki
(nM)
Microsomal
stability mCl
Compounds
hV1b Ki (nM)
Microsomal stability (mCl, lL/min/mg)
(nM)
(l
L/min/mg)
Human
Rat
Human Rat
13a
13b
13c
13d
13e
13f
>10,000
786
>10,000
276
101
20
nd
183
nd
373
nd
75
nd
702
nd
476
nd
356
18a
18b
18c
18d
1.2
2.0
3.9
1.3
193
289
243
231
>1000
>1000
>1000
>1000
170
115
135
89
80
204
64
164
217
62
76
56
a
hV1b Ki, hV1a Ki, hV2 and hOT Ki values were determined according to litera-
hV1b Ki values were determined according to literature procedures6 and are
means of at least two experiments.
a
ture procedures6 and are means of at least two experiments.
b
Microsomal intrinsic clearance values were determined according to literature
b
Microsomal intrinsic clearance values were determined according to literature
procedures7 and are means of at least two experiments.
procedures7 and are means of at least two experiments.

T. Oost et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3828–3831
3831
Table 5
tral V1b receptors. Overall, the PK characteristics of benzyl amine–
pyrrolidin-2-yl-oxazole analog 18c represent considerable
Pharmacokinetic profile in rat of 18c and SSR149415
a
improvement over SSR149415, in particular with respect to oral
bioavailabity, half-life and brain penetration (Table 5).
In the rat forced swimming test, a model with predictive valid-
ity for measuring anti-depressant-like effects,9 18c showed effi-
cacy after intraperitoneal administration of 3 mg/kg (Fig. 2,
p <0.05).
In conclusion, starting from potent but metabolically labile
oxindole SSR149415, optimization of metabolic stability resulted
in a new series of potent and selective V1b receptor antagonists
with an improved pharmacokinetic profile in rat and efficacy in a
rat model of anti-depressant activity.
SSR149415
18c
t1/2 (h)a
0.5
2.8
5.2
12
3.3
13
3.0
68
VSS (L/kg)a
ClP (L/h/kg)a
Foral (%)b
Brain/plasma ratioc
0.03
0.31
a
b
c
After intravenous administration (2 mg/kg).
After oral administration (10 mg/kg).
After intraperitoneal administration (10 mg/kg), calculated based on the AUC in
plasma and brain over an 8 h period.
Acknowledgments
We would like to thank Jayne Froggett and Martina Schmidt
and the Analytical Chemistry Group in Ludwigshafen.
References and notes
1. (a) Serradeil-Le Gal, C.; Wagnon, J.; Simiand, J.; Griebel, G.; Lacour, C.; Guillon,
G.; Barberis, C.; Brossard, G.; Soubrie, P.; Nisato, D.; Pascal, M.; Pruss, R.; Scatton,
B.; Maffrand, J.-P.; Le Fur, G. J. Pharmacol. Exp. Ther. 2002, 300, 1122; (b) Griebel,
G.; Simiand, J.; Serradeil-Le Gal, C.; Wagnon, J.; Pascal, M.; Scatton, B.; Maffrand,
J.-P.; Soubrie, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 6370.
2. According to information on nelivaptan from thomson-pharma.com (search
performed on December 3, 2010)
3. (a) Hodgson, R. A.; Higgins, G. A.; Guthrie, D. H.; Lu, S. X.; Pond, A. J.; Mullins, D.
E.; Guzzi, M. F.; Parker, E. M.; Varty, G. B. Pharmacol., Biochem. Behav. 2007, 3,
431; (b) Kumagai, T.; Kuwada, T.; Shibata, T.; Hayashi, M.; Fujisawa, Y.;
Sekiguchi, Y. EP 1659121, 2006.; (c) Schönberger, M.; Leggett, C.; Kim, S. W.;
Hooker, J. M. Bioorg. Med. Chem. Lett. 2010, 20, 3103.
Figure 2. Effect of 18c in forced swim test in rats. Compound 18c showed efficacy
reaching significance at 3 mg/kg. Data are average + SEM.
4. Nassar, A.-E. F.; Kamel, A. M.; Clarimont, C. Drug Discovery Today 2004, 9, 1020.
5. (a) Mongin, F.; Queguiner, G. Tetrahedron 2001, 57, 4059; (b) Turck, A.; Ple, N.;
Mongin, F.; Queguiner, G. Tetrahedron 2001, 57, 4489.
6. (a) Tahara, A.; Saito, M.; Sugimoto, T.; Tomura, Y.; Wada, K.; Kusayama, T.;
Tsukada, J.; Ishii, N.; Yatsu, T.; Uchida, W.; Tanaka, A. Br. J. Pharmacol. 1998, 125,
1463; (b) Strakova, Z.; Kumar, A.; Watson, A. J.; Soloff, M. S. Mol. Pharmacol.
1997, 51, 217; (c) Jasper, J. R.; Harrell, C. M.; O’Brien, J. A.; Pettibone, D. J. Life Sci.
1995, 57, 2253.
The pharmacokinetic profile of methylamine 18c in rat after
intravenous administration of 2 mg/kg 18c was characterized by
high plasma clearance (3.0 L/h/kg) in combination with a high vol-
ume of distribution (13 L/kg) and a long terminal half-life (3.3 h).
The bioavailability after oral administration of 10 mg/kg 18c was
good (68%). The brain:plasma ratio was 0.31, indicating that this
compound can penetrate the blood–brain barrier and bind to cen-
7. Obach, S. Drug Metab. Dispos. 1999, 27, 1350.
8. Jacobs, J.; Jain, R. K.; Lewis, J. G.; Patel, D. V.; Yuan, Z. US 2003/0069223, 2003.
9. Porsolt, R. D.; Le Pichon, M.; Jalfre, M. Nature 1977, 266, 730.