Identification and optimisation of novel sulfonamide, selective vasopressin V1B receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2011.04.104

The synthesis and preliminary structure-activity relationships (SAR) of a novel class of vasopressin V_1B receptor antagonists are described. Hit compound 5, identified via high throughput screening of the corporate collection, showed good activity in a V_1B binding assay (K _i 63 nM) but did not possess the lead-like physicochemical properties typically required in a hit compound. A 'deletion approach' on the HTS hit 5 was performed, with the focus on improvement of physicochemical properties, yielding the selective V_1B antagonist 9f (K_i 190 nM), with improved druglike characteristics.

Full text of DOI:10.1016/j.bmcl.2011.04.104

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3603–3607
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and optimisation of novel sulfonamide, selective vasopressin
V1B receptor antagonists
James Baker c, Matilda Bingham a, , Ruth Blackburn-Munro a, Jiaqiang Cai a, Mark Craighead b,
⇑
Robert Gilfillan a, Kate Goan b, David Jaap a, Rachel Milne b, J. Richard Morphy a,
b
a
b
Susan Napier a, , Jeremy Presland , Gayle Spinks , Fiona Thomson
⇑
a Department of Medicinal Chemistry, MSD, Newhouse, Motherwell ML1 5SH, United Kingdom
b Department of Molecular Pharmacology, MSD, Newhouse, Motherwell ML1 5SH, United Kingdom
c Department of Pharmacology, MSD, Newhouse, Motherwell ML1 5SH, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and preliminary structure–activity relationships (SAR) of a novel class of vasopressin V1B
receptor antagonists are described. Hit compound 5, identified via high throughput screening of the cor-
porate collection, showed good activity in a V1B binding assay (Ki 63 nM) but did not possess the lead-like
physicochemical properties typically required in a hit compound. A ‘deletion approach’ on the HTS hit 5
was performed, with the focus on improvement of physicochemical properties, yielding the selective V1B
antagonist 9f (Ki 190 nM), with improved druglike characteristics.
Received 11 March 2011
Revised 20 April 2011
Accepted 22 April 2011
Available online 1 May 2011
Keywords:
Vasopressin antagonist
GPCR antagonist
Ó 2011 Elsevier Ltd. All rights reserved.
The hypothalamic-pituitary-adrenal (HPA) axis is one of the
main routes by which humans and other mammals cope with
and adapt to stress.1 A majority of patients suffering from severe
depressive disorder exhibit a profound alteration in their ability
to regulate the HPA, indeed, hyperactivity of the HPA axis is a
neuroendocrine abnormality that has been reported to occur in
a number of psychiatric conditions including major depressive
disorder.1 There are strong data to indicate that hyperactivity
of the HPA axis during chronic stress and in depression is caused
by a shift towards a predominant vasopressin (AVP) regulation
of this system via binding and activation of the pituitary V1B
receptor.2 Thus, antagonists of the pituitary V1B receptor are pro-
posed to normalise HPA overactivity and, as such, could provide
therapeutic benefit in the treatment of diseases such as major
depression and stress-related disorders.3 Preclinical studies with
the literature V1B antagonist SSR 149415 in animal models pre-
dictive of antidepressant and anxiolytic activity further support
this hypothesis.4
and the oxytocin (OT) receptor. This emphasises the challenges
inherent in the identification of small molecule non-peptide antag-
onists for this neuropeptide GPCR family of receptors and is re-
flected in the scant reports of selective non-peptide V1B ligands
since the discovery of the receptor in 1994.5 Our programme aim
was thus to identify a reasonably potent (Ki <500 nM) and selective
lead compound which was structurally distinct from known V1B
antagonists with suitable druglike and physicochemical properties
to serve as a basis for further lead optimisation efforts.
An in-house HTS campaign identified hit 5 as a potent non-selec-
tive V1B antagonist (Fig. 2, Table 2). This compound would not rou-
tinely be a selected as a suitable start point for a Hit-to-Lead
programme due to its poor physicochemical properties however,
with favourable precedent in the vasopressin field, we elected to
employ a strategy successfully utilized in our discovery of a novel
series of 2-(4-oxo-2-aryl-quinazolin-3(4H)-yl)acetamide V1B antag-
onists,8 where a similarly high molecular weight hit was optimised
via careful monitoring of ligand efficiency and physicochemical
properties.
A number of V1B receptor antagonists have been reported in the
literature (Fig. 1, Table 1). However, these ligands are often charac-
terised by high molecular weight, non lead-like properties, and
with the exception of the recently reported and structurally related
3 and 4, show off-target activity at the related GPCR receptors in
the vasopressin receptor family the V1A, V2 vasopressin receptors
Scheme 1 illustrates a typical synthetic route which was used to
prepare analogues of the hit compound 5. Final compounds were
prepared using parallel synthesis technology, either via resin
chemistry as illustrated in Scheme 1, or via solution phase chemis-
try as in Scheme 2. Final products were isolated and purified by
preparative LCMS.
The compounds thus prepared were evaluated for their ability
to displace the binding of tritium labelled arginine vasopressin
([3H]-AVP) to the human V1B receptor in a whole cell binding assay
⇑
Corresponding authors. Tel.: +44 1698736000; fax: +44 1698736187 (M.B.).
E-mail addresses: matilda.bingham@virginmedia.com
(M. Bingham),
susannapier67@googlemail.com (S. Napier).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.104

3604
J. Baker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3603–3607
H
N
O
N
O
N
O
OH
O
N
Me
Me
Cl
N
O
H
N
N
3
Me
O
O
N
N
N
S
S
O
H
N
O2
O
S
O
N
O
O
O
2
Me
Me
O
O
O
O
Me
Me
SSR 149415
O
N
N
1
N
4
Cl
Figure 1.
Table 1
Table 2
Physicochemical properties of hit compound 5
Physicochemical properties of literature V1B antagonists
MWt
Rotatable
bonds
c Log P PSA
Donor/
acceptors
Reference
MWt
Rotatable
bonds
c Log P PSA
Ligand
efficiency
Donor/
acceptors
1
2
3
4
630.1
619.8 11
476.6 10
8
3.9
5.1
3.7
4.1
125.9 9/1
105.9 7/1
83.5 6/1
79.7 5/1
6
7
8
9
5
630.1
8
3.9
125.9 0.21
9/1
486.0
9
using CHO cells recombinantly expressing the human V1B
receptor.10
H2N
Our initial studies were focused on reducing the molecular
weight of compound 5 hence a series of deletion analogues were
prepared to determine the minimum pharmacophore required
for activity at V1B (Table 3).
These deletion analogues in Table 3 suggested that the (S),
(S)-enantiomer was responsible for the potency at the V1B receptor,
however, it was notable that the (S)-CH2(napthalen-2-yl) analogue
10c, which more closely resembled the steric requirements of hit
compound 5, was not active in contrast to the (S)-CH2Ph analogue
10b. This led us to hypothesise that the basic group in the
(1-aminoisoquinolin-6-yl) side chain of hit 5 was responsible for
a key interaction with the V1B receptor. To test this hypothesis,
and further reduce the molecular weight of the compounds in line
N
Me
Me
O
Me
S
O
Me
H
N
N
N
H
O
O
O
5
V1b pKi 63nM
Figure 2.
Me
N
Me
O
Me
S
O
O
CO2H
Me
H
N
H
N
a) b)
c)
CO2H
O
BocHN
Me
S
O
O
Me
O
O
Me
OH
Me
N
d)
H
N
R1
R2
Me
O
Me
O
R2
R1
Me
S
Me
S
N
H
N
Cl
O
6
O
Me
Me
O
O
O
Me
Me
7a-g
Scheme 1. (a) TFA (b) 6, DIPEA, DMF:H2O (2:1) 61% (c) DIC, HOBt, DMAP, DMF (d) R1R2NH, DCM, rt 18 h.

J. Baker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3603–3607
3605
O
Cl
S
O
Me
Me
Me
Me
a)
8
O
O
Me
Me
R1
R1
R2
R2
OH
N
N
( )n
R1
( )n
b)
( )n
c)
R1
O
R1
O
O
Me
Me
Me
N
N
N
H
H
O
S
O
9b-i
Me
O
Me
Me
Scheme 2. (a) PCl5, DCM, 0 °C, ClSO3H, 96% (b) EDCI, DIEA, HOBt, DCM (c) KOtBu, THF, 8, 20–96% after LCMS (2 steps).
Table 3
Table 4
V1B binding assay results for deletion analogues 10a-d
V1B binding assay results for compounds 7a–7g
Me
Me
O
Me
S
O
Me
S
R1
O
O
Me
Me
H
N
H
N
R1
N
N
H
N
R2
O
O
O
Me
O
O
Me
10a-d
7a-g
R1
V1B Ki (lM)
MWt
R1
H
R2
V1B Ki (
l
M)
MWt
a
a
10a
10b
10c
10d
(R)-CH2Ph
(S)-CH2Ph
(S)-CH2(naphthalen-2-yl)
>10
0.63
>10
>10
591.8
591.8
641.8
501.6
N
7a
7b
7c
0.50
503.7
516.7
515.7
N
H
N
N
N
H
H
0.25
2.5
a
Values are mean of at least 2 independent experiments carried out in duplicate.
N
with the stated strategy, a diverse array of compounds where the
terminal piperidinyl carboxyamide amino acid was replaced with
an amine group were synthesised. In addition to reducing the
molecular weight this would also allow removal of one of the
amide bonds, improving the predicted oral absorption properties.
Selected active compounds from this array are highlighted in Table
4.
Table 4 details a number of amines which successfully replaced
the terminal piperidinyl carboxyamide amino acid of hit 5, whilst
maintaining activity at V1B. The distance of the basic centre from
the amide bond appears to be key, with the propyl spacer as opti-
mal (7a vs 7b). The distal amino group is most important for activ-
ity since both the 4-methylpiperidine and morpholino equivalents
of 7b were less active. The secondary amide is also preferred,
although tertiary amides can be tolerated, as both 7f and the
homopiperazine 7g were less active. Importantly compound 7b
represents a breakthrough since it possesses significantly better
properties compared with the hit 5 in terms of molecular weight
(516.7 vs 630.1), a reduced number of amide bonds (1 vs 2) and
an improved ligand efficiency (0.26 vs 0.21).
7d
7e
H
10
503.7
461.6
O
N
N
N
H
2.0
7f
Me
1.0
3.2
530.7
473.6
7g
N-Methyl homopiperazine
a
Values are mean of at least 2 independent experiments carried out in duplicate.
(not shown), however only small changes to the electron rich
aromatic ring were tolerated with the original sulfonamide present
in compounds 7a–g remaining the best substituent. Removal of the
aryl 6-methyl group was tolerated and this sulfonamide was used
in subsequent compounds 9a–i due to the modest improvement it
afforded in reduction of MW and c Log P.
The next focus for optimisation was the number of rotatable
bonds. A number of cyclic constraints were investigated in this
regard, the two most successful both involved formation of a
6,5-bicycle via cyclisation of the central phenylalanine aryl ring
in compounds 7a–g (Table 4) onto the sulfonamide NH, to afford
An exploration of the SAR of the aryl-sulfonamide region of the
molecule proved less fruitful. A diverse array of sulfonamides were
synthesized including alkyl, heterocyclic and aryl sulfonamides

3606
J. Baker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3603–3607
Table 5
V1B binding assay results for compounds 9a–i
R3
O
Me
O
Me
Me
O
NR1R2
Me
NR1R2
N
S
O
N
S
O
n
Me
O
Me
O
O
Me
9b-i
9a
R2
R1
R3
n
V1B Ki (
l
M)
Rotatable bonds
8
a
N
N
9a
9b
H
H
—
H
—
1.2
N
N
0
0.56
8
9c
9d
9e
9f
N-Methyl homopiperazine
Homopiperazine
Homopiperazine
Homopiperazine
Homopiperazine
H
Me
F
OMe
OMe
0
0
0
0
1
0.50
0.79
0.31
0.19
0.18
4
4
4
5
6
9g
N
9h
9i
H
F
F
1
1
0.43
0.26
9
8
N
Me
N
a
Values are mean of at least 2 independent experiments carried out in duplicate.
compounds 9g–i was also tolerated, as was the use of a less basic
side chain such as pyridine in 9i, however these modifications,
although interesting from an SAR standpoint did little to improve
the physicochemical properties.
Table 6
Selectivity profile for compound 9f
Me
O
Compound 9f was chosen for further selectivity profiling
against the family of related receptors V1A, V2 and OT (Table 6)
and was shown to be a selective V1B ligand which retained selectiv-
ity against 5HT6. Furthermore functional antagonism was demon-
strated in an hV1B luciferase reporter assay linked to AVP-mediated
intracellular calcium mobilisation (9f IC50 371 nM). Compound 9f
has good solubility (Solkin at pH7.4 72 mg/L), moderate rat and
mouse microsomal stability (RLM Clint 55, MLM Clint 58, HLM
217) and measured log D7.4 2.0. The addition of the methylene
spacer as in 9g improves the microsomal stability (RLM Clint 25,
MLM Clint 16, HLM 16), demonstrating there is scope for optimisa-
tion within the series however this is at the expense of the hOT
selectivity (9g hOT Ki 320 nM).
O
Me
NH
Me
Me
N
N
S
O
O
O
Me
9f
M)
hV1BaKi (
l
M)
hV1AaKi (
l
hV2aKi (
l
M)
hOTaKi (
lM)
9f
0.19
>10
>10
>10
a
Values are mean of at least 2 independent experiments carried out in duplicate.
In conclusion, hit optimisation efforts focussing on the improve-
ment of physicochemical properties have lead to the identification
of selective V1B antagonist 9f representing a novel chemotype in
the vasopressin field. Significant improvements have been made
to the physicochemical properties in comparison with the original
hit compound 5 (e.g., compound 5 MW = 630, c Log P 3.9, PSA
125.9, Rotatable bonds 11, L.E. 0.21 vs 9f MW = 485, c Log P 4.1,
Log D7.4 2.0, PSA 71.8, Rotatable bonds 5 L.E. 0.28). Compound 9f
thus represents a novel start point, with physicochemical proper-
ties suitable for further optimisation.
an indoline-3-carboxamide or 2-methyl-indole-3-carboxamide.
Examples of active compounds from these series are indoline-
3-carboxamide 9a and the 2-methyl-1H-indole-3-carboxamides
and (2-methyl-1H-indol-3-yl)acetamides 9b–i (Table 5). Interest-
ingly, carboxamide substitution at the 3-position of the indole
and indoline rather than 2-substitution (not shown) afforded
the active isosteric replacement, highlighting the importance of
testing all the possible regioisomers. The indoline chemotype
represented by 9a was found to have significant off target affin-
ity for 5HT6 and was therefore deprioritised in favour of 9b–i
(Table 5).11
References and notes
It was possible to further reduce the rotatable bond count by
replacing the (4-methylpiperazin-1-yl)propyl side chain in 9b with
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; (c) Dinan, T. G.; O’Brien, S.; Lavelle, E.; Scott,
L. V. Psychol. Med. 2004, 34, 169.
a
homopiperizine side chain. The modest potency of these
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) Scott, L. V.; Dinan, T. G. Life Sci. 1998, 62, 1985; (b) Aquilera, G.; Rabdan-
Diehl, C. Regul. Pept. 2000, 96, 23.
analogues could be improved by substitution in the indole 5-posi-
tion (compounds 9d–f) to afford compounds such as 9f with the
target potency and greatly reduced rotatable bond count (4 vs 11
for compound 9b). The addition of a methylene spacer as in

J. Baker et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3603–3607
3607
4. (a) Griebel, G.; Simiand, J.; Serradeil-Le Gal, C.; Wagnon, J.; Pascal, M.; Scatton,
B.; Maffrand, J.-P.; Soubrie, P. PNAS 2002, I99, 6370–6375; (b) Serradeil-Le Gal,
C.; Wagnon, J.; Tonnerre, B.; Roux, R.; Garcia, G.; Griebel, G.; Aulombard, A. C. N.
S. Drug Rev. 2005, 11, 53.
5. Keyzer, Yde.; Auzan, C.; Lenne, F.; Beljord, C.; Thibonnier, M.; Bertagna, X.;
Clauser, E. FEBS Lett. 1994, 356, 215.
6. (a) Serradeil-Le Gal, C.; Wagnon, J.; Simiand, J.; Griebel, G.; Lacour, C.; Guillon,
G.; Barberis, C.; Brossard, G.; Sßoubrié, S.; Nisato, D.; Pascal, M.; Pruss, R.;
Scatton, B.; Maffrand, J.-P.; Le Fur, G. J. Pharmacol. Exp. Ther. 2002, 300, 1122; (b)
Griffante, C.; Green, A.; Curcuruto, O.; Haslam, C. P.; Dickinson, B. A.; Arban, R.
Br. J. Pharmacol. 2005, 146, 744.
7. Scott, J. D.; Miller, M. W.; Li, S. W.; Lin, S.-I.; Vaccaro, H. A.; Hong, L.; Mullins, D.
E.; Guzzi, M.; Weinstein, J.; Hodgson, R. A.; Varty, G. B.; Stamford, A. W.; Chan,
T.-Y.; McKittrick, B. A.; Greenlee, W. J.; Priestley, T.; Parker, E. M. Bioog. Med.
Chem. Lett. 2009, 19, 6018.
Morphy, J. R.; Ohlmeyer, M. H. J.; Webb, M. L. Bioorg. Med. Chem. Lett. 2010,
20, 5394.
9. Arban, R.; Bianchi, F.; Buson, A.; Cremonesi, S.; Fabio, R. D.; Gentile, G.; Micheli,
F.; Pasquarello, A.; Pozzan, A.; Tarsi, L.; Terreni, S.; Tonelli, F. Bioorg. Med. Chem.
Lett. 2010, 20, 5044.
10. CHO cells (V1B and OT) or membranes prepared from CHO cells (V1A and V2)
expressing the human versions of the receptors were incubated with tritiated
ligand ([3H] AVP for V1A, V1B and V2; [3H]oxytocin for OT) and increasing
concentrations of the compound. Nonspecific binding was determined in the
presence of 10 lM lysine vasopressin for V1A, V1B and V2; 10 lM oxytocin for
OT. The concentration of compound was plotted against specific binding and
data were analysed by non-linear regression to calculate affinities of the
compound for each receptor.
11. Reid, M.; Carlyle, I.; Caulfield, W.; Clarkson, T. R.; Cusick, F.; Epemolu, O.;
Gilfillan, R.; Goodwin, R.; Goodwin, R.; Jaap, D.; O’Donnell, E. C.; Presland, J.;
Rankovic, Z.; Spinks, D.; Spinks, G.; Thomson, A. M.; Thomson, F.; Strain, J.;
Wishart, G. Bioorg. Med. Chem. Lett. 2010, 20, 3713.
8. Letourneau, J. L.; Rivello, C.; Li, H.; Cole, A. G.; Ho, K.-K.; Zanetakos, H. A.;
Desai, H.; Zhao, J.; Auld, D. S.; Napier, S. E.; Thomson, F. J.; Goan, K. A.;