Next-generation spirobenzazepines: Identification of RWJ-676070 as a balanced vasopressin V1a/V2 receptor antagonist for human clinical studies

Article abstract of DOI:10.1016/j.bmcl.2007.09.059

We have continued to explore spirobenzazepines as vasopressin receptor antagonists to follow up on RWJ-339489 (2), which had advanced into preclinical development. Further structural modifications were pursued to find a suitable backup compound for human clinical studies. Thus, we identified carboxylic acid derivative 3 (RWJ-676070; JNJ-17158063) as a potent, balanced vasopressin V_1a/V₂ receptor antagonist with favorable properties for clinical development. Compound 3 is currently undergoing human clinical investigation.

Full text of DOI:10.1016/j.bmcl.2007.09.059

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6623–6628
Next-generation spirobenzazepines: Identification of RWJ-676070
as a balanced vasopressin V1a/V2 receptor
antagonist for human clinical studies
Min Amy Xiang,a Philip J. Rybczynski,a Mona Patel,a Robert H. Chen,a
David F. McComsey,b Han-Cheng Zhang,b Joseph W. Gunnet,a Richard Look,a
Yuanping Wang,b Lisa K. Minor,b H. Marlon Zhong,b Frank J. Villani,b
Keith T. Demarest,b Bruce P. Damianob and Bruce E. Maryanoffb,*
aResearch and Early Development, Johnson & Johnson Pharmaceutical Research & Development,
8 Clarke Drive, Cranbury, NJ 08512, USA
bResearch and Early Development, Johnson & Johnson Pharmaceutical Research & Development,
Welsh & McKean Roads, Spring House, PA 19477-0776, USA
Received 27 June 2007; revised 14 September 2007; accepted 14 September 2007
Available online 19 September 2007
Abstract—We have continued to explore spirobenzazepines as vasopressin receptor antagonists to follow up on RWJ-339489 (2),
which had advanced into preclinical development. Further structural modifications were pursued to find a suitable backup com-
pound for human clinical studies. Thus, we identified carboxylic acid derivative 3 (RWJ-676070; JNJ-17158063) as a potent, bal-
anced vasopressin V1a/V2 receptor antagonist with favorable properties for clinical development. Compound 3 is currently
undergoing human clinical investigation.
Ó 2007 Elsevier Ltd. All rights reserved.
For several years, we have prospected for nonpeptide
vasopressin receptor antagonists to advance compounds
into human clinical trials.1 We managed to identify dif-
ferent series of potent V2-selective antagonists,2a–f from
which three compounds were selected for preclinical
development. One of these, RWJ-351647 (1), was inves-
tigated in humans and found to be well tolerated and
efficacious as an aquaretic agent.1,3 Potent V1a receptor
antagonism proved to be more elusive, until we discov-
ered a chemical series based on spirocyclic benzaze-
pines.4 This novel series also enabled us to surface
dual vasopressin V1a/V2 receptor antagonists, which
have potential for wider clinical utility, especially for
treating congestive heart failure and renal disorders.5
its excellent V1a and V2 affinity, good potency, and bal-
anced ratio in V1a and V2 functional assays. In rats, 2
had useful oral bioavailability (F = 22%; t1/2 = 6.5 h)
and produced dose-dependent aquaresis (at 10 mg/kg,
p.o., urine output was elevated by 1100% and urine
osmolality was reduced by 75%),1 the latter being a
gauge of V2 receptor antagonism in vivo. On the basis
of the many drugworthy properties of 2, we advanced
it into preclinical development as a dual vasopressin
V1a/V2 antagonist. However, because of unexpected tox-
icology findings at suprapharmacological doses, 2 was
abandoned as a clinical candidate. In this paper, we de-
scribe our effort to identify a suitable, next-generation
backup compound, which was realized in the form of
RWJ-676070 (3; JNJ-17158063).4c
Our initial reports on spirobenzazepines4a,b disclosed
RWJ-339489 (2; R-enantiomer),4b which is notable for
In seeking a backup compound for 2, we were interested
in reducing its relatively high molecular weight
(MW = 612.8 Da), which would increase the probability
of success.6 One approach was to decrease the size of the
substituent on the 5-membered ring. However, it was
also important to retain some polar character at that
Keywords: Vasopressin receptor; Antagonist; Spirobenzazepine.
*
Corresponding author. Tel.: +1 215 6285530; fax: +1 215 6284985;
e-mail: bmaryano@prdus.jnj.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.09.059

6624
M. A. Xiang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6623–6628
CO2H
O
NH(CH2)2NMe2
N
O
N
H
N
N
O
Cl
F
O
O
Ph
O
Ph
O
O
N
H
Cl
N
H
N
H
MeO
1
2
3
Table 1. Vasopressin binding and functional data for spirobenzazepine derivativesa
X
X
N
N
R3
R3
O
O
O
O
R1
R1
N
H
N
H
R2
R2
R4
R4
4a-j, l, m
2; 3; 4k, n-r
Compound
X
R1
R2
R3
R4
V1a bndgb
V2 bndgb
IC50, nM
V1a funct.c
IC50, nM
V2 funct.c
IC50, nM
IC50, nM
4a
CO2H
H
H
H
H
H
Cl
Cl
Cl
H
Ph
Ph
Ph
F
H
H
H
H
H
F
F
F
F
F
F
F
F
F
F
F
F
F
F
H
37
14
20
360
190
1800
ND
>10,000
ND
ND
ND
140
410
100
>10,000
120
52
4b
4cd
CH2OH
H
9
18
9
CH2NHMe
CO2H
H
18
44
630
ND
130
ND
ND
ND
11,600
780
380
190
89
4d
H
ꢀ100
15
15%
23%
67%
66%
17%
110
76
4e
CO2Me
H
F
4f
CO2H
H
Me
Me
Me
Me
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Ph
10%
ꢀ100
0%
15%
32
4g
4hd
CH2OH
H
CH2NHMe
CO2H
H
4i
MeO
H
H
4j
4ke
CO2H
H
CO2H
H
H
13
53
4l
CO2Me
H
H
40
>200
92
4m
3f
CO2H
H
MeO
MeO
EtO
MeO
MeO
MeO
MeO
H
19
CO2H
H
5
34
38
4ne
4og,h
4pg,i
4qg,j
4rg,k
2l
CO2H
H
8
71
50
150
40
Tetrazol-5-yl
C(O)NHSO2NH2
Imidazol-2-yl
Imidazolin-2-yl
C(O)NHRm
H
ND
ND
ND
ND
2
ND
ND
ND
ND
8
63
H
27
11
H
95
730
1050
36
H
320
45
H
VPA-985n
150
5
>5000
91
a Target compounds were purified by silica gel flash-column chromatography. All compounds were characterized by ESI-MS and 300-MHz 1H
NMR. Compounds are racemates unless noted otherwise. ND, not determined.
b Inhibition of [3H]-Arg-vasopressin binding to recombinant human vasopressin V1a or V2 receptors (N = 2–6). IC50 values are given unless noted
otherwise as % inhibition at 100 nM.
c Inhibition of Arg-vasopressin-induced effects on cells expressing either human V1a or V2 receptors. IC50 values represent the concentration of compound that
antagonizes50%ofthecellularcalciummobilizationforV1a receptorsorcellularcAMPaccumulationforV2receptorscausedby1 nMArg-vasopressin(N = 1–4).
d Isolated as a free base.
e Prepared from (R)-(+)-I (>99% ee).
25
f (R)-(+) enantiomer; ½aꢁD þ 19:5ꢂ (c 0.14, CHCl3); >99% enantiomeric purity by chiral HPLC (ChiralPak AD-H column).
g (R) enantiomer; prepared from 3.
25
h
½aꢁD þ 193ꢂ (c 0.84, CHCl3).
25
½aꢁD þ 103:9ꢂ (c 1.32, CHCl3).
i
25
½aꢁD þ 71:2ꢂ (c 1.23, CHCl3).
j
25
½aꢁD þ 13:1ꢂ (c 1.23, CHCl3).
k
l (R)-(+) enantiomer; >99% enantiomeric purity by chiral HPLC. The biodata reported here for 2 are taken from Ref. 4b.
m R = CH2CH2NMe2.
n Reference standard lixivaptan (Ref. 7a).

M. A. Xiang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6623–6628
6625
site, such as a carboxylic acid or amine, to confer
reasonable aqueous solubility to the target molecules.
Thus, analogues of general structure 4 were considered
(Table 1). Our prior, intensive studies on vasopressin
receptor antagonists,2,4 and the published results of oth-
ers,7 have provided a good foundation for appreciating
the structure–activity relationships (SARs) associated
with aromatic ring substitution. Therefore, we did not
revisit this particular facet. Rather, we decided to begin
with certain most-favored arrangements for the 4-(benz-
amido)benzoyl subunit, especially those that would
enhance V1a receptor antagonist action.2,4,7 In general,
these subunits are: 4-(2-phenylbenzamido)benzoyl,
4-(2-methyl-5-fluorobenzamido)-2-chlorobenzoyl, and
4-(2-chloro-5-fluorobenzamido)benzoyl.
established the absolute stereochemistry (vide infra),
which was conveyed into the relevant target com-
pounds (Table 1). The chemistry starting with I is
shown in Scheme 1. For the second step, the 4-(benz-
amido)benzoyl subunits were generated in a standard
manner, except for the 3-alkoxy versions. Illustrating
for the 3-methoxy case, we synthesized (2-chloro-5-flu-
orobenzamido)-3-methoxybenzoyl chloride (IX) to
assemble 3, 4m, and 4o–r (Scheme 2). The alcohol
(4b, 4g) and amine (4c, 4h) target compounds were pre-
pared starting with known spirobenzazepine X,4b
according to the chemistry outlined in Scheme 3. The
tetrazole analogue 4o of carboxylic acid 3 was prepared
by forming the primary carboxamide, dehydrating it to
the nitrile with cyanuric chloride (DMF, 3 h), and
reacting the nitrile with sodium azide (DMF, 125 °C,
15 h). Acylsulfamide 4p was obtained by reacting the
acid chloride of 3 with sulfamide (100 °C, 1 h). Imidazo-
line 4r was prepared by converting the primary carbox-
The synthesis of 3 and its analogues departs from the
synthesis described earlier for 2 and its analogues,4b
in that we now employed spirobenzazepine I as a key
intermediate.8 Compound 3 was separated into its indi-
vidual enantiomers, and a single-crystal X-ray determi-
nation on the (ꢃ)-10-camphorsulfonate salt of (R)-I
amide to a methyl imidate ðMe3OþBF ꢃ; CH2Cl2Þ, fol-
4
lowed by reaction with ethylene diamine (CH2Cl2,
24 h). Imidazole 4q was obtained by oxidizing 4r with
MnO2 (CH2Cl2, 72 h).
2
1
3
R
R
R
CO2H
CO2Et
Intermediate I formed a crystalline salt with (ꢃ)-10-cam-
phorsulfonic acid, as one diastereomer. A single crystal of
the salt (from MeOH) was subjected to X-ray diffraction
to furnish the structure and absolute stereochemistry of
the spirobenzazepine unit, which turned out to be R.9
The structure of the salt is depicted in Figure 1.
CONH
COCl
4
H2SO4
R
EtOH
94%
Et3N
CH2Cl2
N
H
N
H
II
I
Since our goal was a potent, next-generation, dual
V1a/V2 antagonist, we directed our sights on obtaining
candidate compounds expeditiously by capitalizing on
already-published structure–activity correlations.2,4,7
Use of the 4-(benzamido)benzoyl subunits mentioned
above would increase the chances of obtaining the more
elusive V1a receptor antagonist activity. At first, the ana-
logues were evaluated for affinity to V1a and V2 recep-
tors in a binding assay;10 then, compounds of interest
were further evaluated for cellular V1a and V2 functional
activity (Table 1).11 It was not sufficient just to have high
potency in receptor binding, since high functional antag-
onism is crucial. Suitably potent dual antagonists in the
CO2H
CO2Et
LiOH, H2O
EtOH, THF
N
N
O
O
R3
R3
O
O
R1
R1
N
H
N
H
R2
R2
R4
R4
III
IV
Scheme 1.
CO2H
CO2Me
CO2Me
Me2SO4
H2/Pd-C
ArC(O)Cl
Et3N
CH2Cl2
K2CO3
acetone
EtOAc
95%
OH
OMe
OMe
NO2
NO2
NH2
98%
95%
V
VI
VII
Cl
Ar = 2-Cl,5-FC6H3
MeO2C
1) LiOH, THF
16 h (93%)
O
O
N
O
2) (COCl)2
CH2Cl2 (99%)
Ar
H
N
H
MeO
VIII
Ar
MeO
IX
Scheme 2.

6626
M. A. Xiang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6623–6628
CHO
ArCOCl
Et3N
O3, TsOH, CH2Cl2
then Me2S
50%
N
N
H
N
O
Ar
75-80%
O
Ar
XII
X
XI
NHMe
OH
NaCNBH3
MeNH3Cl
MeOH
HOAc
NaBH4
MeOH/HOAc
N
N
O
Ar
O
Ar
4b; 4g
4c; 4h
Me
F
O
Ph
O
Ar =
Cl
N
N
H
H
4b; 4c
4g; 4h
Scheme 3.
ers, although 4c was less potent in the functional assays.
Replacement of the phenyl group of the acid with a
fluoro (4d) or a methyl (4f–i) was not auspicious.
2-Chloro-5-fluoro substitution (4j) led to positive results
in binding affinity, but not in functional activity. How-
ever, in exploring some new SAR territory, we found
that addition of a 3-methoxy (4m, 3) or 3-ethoxy (4n)
group to this chloro/fluoro motif can yield a desirable
profile (cf. 2). The methoxy derivative as the R-enantio-
mer, 3, was potent in the functional V1a and V2 assays
with a favorable V1a/V2 ratio of nearly 1. Compound 3
is a white crystalline solid with an acceptable molecular
weight (Mw = 549.0 Da) for a nonpeptide vasopressin
antagonist2,4,7 and good aqueous solubility at pH 7.4
(>1 mg/mL). Given the favorable attributes of 3, we
sought some close-in derivatives with acidic or basic
replacements for the CO2H group, represented by
4o–r. The functional V1a and V2 potencies, and V1a/V2
ratios, for 4o and 4p (IC50 values of 63/40 and 27/11,
respectively; cf. 2 and 3) are particularly noteworthy.
Clearly, it is possible to obtain potent V1a and V2 recep-
tor antagonist activity with either a basic or acidic moi-
ety attached to the spirocyclopentene ring.
Figure 1. An ORTEP plot of the molecular structure of (R)-(+)-I,
(4R)-1,2,3,5-tetrahydrospiro[4H-1-benzazepine-4,10-[2]cyclopentene]-30-
carboxylic acid, as a salt with (1R)-(ꢃ)10-camphorsulfonic acid, showing
the atom-numbering scheme for the heteroatoms.
Additional studies were conducted with 3. With rat V1a
and V2 vasopressin receptors, its functional IC50 values
were 640 and 30 nM.12 The pharmacokinetics of 3 were
assessed in rats, dogs, and cynomolgus monkeys. Thus,
3 was found to have excellent oral bioavailability
(F = 68%, 45%, and 23%, respectively) with good oral
plasma half-lives (t1/2 = 8.7, 5.1, and 10.9 h).
functional assays were assessed for oral bioavailability
in rats, metabolic stability in vitro, and inhibition of
cytochrome P450 isozymes in vitro.
Oral administration of 3 to hydrated conscious rats pro-
duced a dose-dependent aquaretic effect (Table 2). For
example, at an oral dose of 30 mg/kg, it caused a
The carboxylic acid parent of 2, 4a, exhibited activity of
interest, as did the carbinol (4b) and amine (4c) congen-

M. A. Xiang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6623–6628
Table 2. Effect of 3 on urine volume and osmolality in ratsa
6627
2. (a) Matthews, J. M.; Greco, M. N.; Hecker, L. R.;
Hoekstra, W. J.; Andrade-Gordon, P.; de Garavilla, L.;
Demarest, K. T.; Ericson, E.; Gunnet, J. W.; Hageman,
W.; Look, R.; Moore, J. B.; Maryanoff, B. E. Bioorg. Med.
Chem. Lett. 2003, 13, 753; (b) Dyatkin, A. B.; Hoekstra,
W. J.; Hlasta, D. J.; Andrade-Gordon, P.; de Garavilla,
L.; Demarest, K. T.; Gunnet, J. W.; Hageman, W.; Look,
R.; Maryanoff, B. E. Bioorg. Med. Chem. Lett. 2002, 12,
3081; (c) Matthews, J. M.; Hoekstra, W. J.; Dyatkin, A.
B.; Hecker, L. R.; Hlasta, D. J.; Poulter, B. L.; Andrade-
Gordon, P.; de Garavilla, L.; Demarest, K. T.; Ericson,
E.; Gunnet, J. W.; Hageman, W.; Look, R.; Moore, J. B.;
Reynolds, C. H.; Maryanoff, B. E. Bioorg. Med. Chem.
Lett. 2004, 14, 2747; (d) Matthews, J. M.; Hlasta, D. J.;
Andrade-Gordon, P.; Demarest, K. T.; Ericson, E.;
Gunnet, J. W.; Hageman, W.; Look, R.; Moore, J. B.;
Maryanoff, B. E. Lett. Drug Des. Discov. 2005, 2, 601; (e)
Hoekstra, W. J.; Dyatkin, A. B.; Maryanoff, B. E.;
Matthews, J. M. U.S. Patent 6,713,475, 2004; (f) Urban-
ski, M. J.; Chen, R. H.; Demarest, K. T.; Gunnet, J.;
Look, R.; Ericson, E.; Murray, W. V.; Rybczynski, P. J.;
Zhang, X. Bioorg. Med. Chem. Lett. 2003, 13, 4031.
3. (a) Gunnet, J. W.; Matthews, J. M.; Maryanoff, B. E.; de
Garavilla, L.; Andrade-Gordon, P.; Damiano, B.; Hag-
eman, W.; Look, R.; Stahle, P.; Streeter, A.; Wines, P. G.;
Demarest, K. T. Clin. Exp. Pharmacol. Physiol. 2006, 33,
320; (b) Ros, J.; Fernandez-Varo, G.; Munoz-Luque, J.;
Arroyo, V.; Rodes, J.; Gunnet, J. W.; Demarest, K. T.;
Jimenez, W. Br. J. Pharmacol. 2005, 146, 654.
4. (a) Xiang, M. A.; Chen, R. H.; Demarest, K. T.; Gunnet,
J.; Look, R.; Hageman, W.; Murray, W. V.; Combs, D.
W.; Patel, M. Bioorg. Med. Chem. Lett. 2004, 14, 2987; (b)
Xiang, M. A.; Chen, R. H.; Demarest, K. T.; Gunnet, J.;
Look, R.; Hageman, W.; Murray, W. V.; Combs, D. W.;
Rybczynski, P. J.; Patel, M. Bioorg. Med. Chem. Lett.
2004, 14, 3143; (c) Patel, M.; Rybczynski, P. J.; Xiang, M.
A. U.S. Pat. Appl., US2004266752, 2004; (d) Chen, R. H.;
Xiang, M. A. PCT Int. Appl., WO2002002531, 2002.
5. (a) Francis, G. S.; Tang, W. H. W. JAMA 2004, 291, 2017;
(b) Goldsmith, S. R.; Gheorghiade, M. J. Am. Coll.
Cardiol. 2005, 46, 1785.
Dose, p.o.
(mg/kg)
Urine
volume
(mL)
Urine
Urine
Urine
volume
% vehicle
osmolality
(mOsm/kg)
osmolality
% vehicle
Vehicle
2.3 0.4
3.0 0.2
3.8 0.5b
7.2 1.5b
15.9 2.5b
1070 150
920 89
510 110b
560 110b
180 26b
1
3
130
165
310
690
87
48
52
17
10
30
a Compound 3 was administered orally to conscious hydrated male rats
at the specified dose level. Each value represents the mean SE
(N = 6–10).
b P < 0.05 versus the vehicle values (Dunnett’s multiple comparison
test).
690% increase of urine output over untreated controls,
with an 83% reduction of urine osmolality. By way of
comparison, a 10 mg/kg oral dose of lixivaptan (VPA-
985) is reported to alter urine output/osmolality by
+450%/ꢃ70%.7a Compound 3 was also an efficacious
aquaretic agent in female beagle dogs (at 30 mg/kg,
p.o., urine output and osmolality were altered by
+1130% and ꢃ78%) and cynomolgus monkeys.13
To assess the V1a activity of 3 in vivo, we evaluated its
ability to reverse hypertension in rats induced by Arg-
vasopressin. We infused a saline solution of Arg-vaso-
pressin intravenously (30 ng/kg/min) to male Long–
Evans rats to produce a stable increase in mean arterial
pressure of 40–60 mm Hg. The test compound was
administered i.v. in cumulative doses and the reduction
in blood pressure was monitored. The ED50 for 3 (from
a linear section of the dose–response curve) was
644 123 lg/kg (N = 6).14
In conclusion, we manipulated the structure of dual
vasopressin V1a/V2 receptor antagonist 2 to devise a
suitable backup compound for clinical development.
The amide group was changed to a carboxylic acid;
the pendant phenyl group was changed to a chloro;
and fluoro and methoxy substituents were added. The
resultant analogue, 3, is a potent, dual V1a/V2 antagonist
with functional activity in vitro and in vivo. In addition,
3 possesses favorable attributes for advancement as a
clinical candidate.15 Consequently, 3 (RWJ-676070;
JNJ-17158063)1a,4c is now undergoing clinical studies
in humans.16
6. (a) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney,
P. J. Adv. Drug Delivery Rev. 1997, 23, 3; (b) Veber, D. F.;
Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.; Ward, K. W.;
Kopple, K. D. J. Med. Chem. 2002, 45, 2615.
7. (a) Albright, J. D.; Reich, M. F.; Santos, E. G. D.; Dusza,
J. P.; Sum, F. W.; Venkatesan, A. M.; Coupet, J.; Chan, P.
S.; Ru, X.; Mazandarani, H.; Bailey, T. J. Med. Chem.
1998, 41, 2442; (b) Matsuhisa, A.; Taniguchi, N.; Koshio,
H.; Yatsu, T.; Tanaka, A. Chem. Pharm. Bull. 2000, 48, 21.
8. (a) Zhong, H. M. et al., Unpublished results; presented at
the 228th ACS National Meeting, Philadelphia, PA, Aug
22–26, 2004, ORGN-524; (b) See: Deng, X.; Kenny, B.;
Liang, J. T.; Mani, N.; Villani, F. J.; Zhang-Plasket, F.;
Zhong, H. U.S. Pat. Appl., US2004259857, 2004.
Acknowledgments
9. Single crystals of [(R)-C15H18NO2][(1R)-(ꢃ)-C10H15SO4]
(from MeOH, mp 289–291 °C) are orthorhombic [space
group P212121–D2 (No. 19)] with a = 7.095(3) A,
We thank Michael J. Costanzo for helpful discussions
and advice. We thank Victor Day of the Crystalytics
Company for the single-crystal X-ray diffraction study
on intermediate I.
4
˚
3
b = 15.592(5) A, c = 21.478(7) A, V = 2376(1) A , and
˚
˚
˚
Z = 4 cation/anion formula units [dcalcd = 1.330 g cmꢃ3
;
la(MoKa) = 0.178 mmꢃ1]. A total of 3285 reflections
having 2h (Mo Ka) < 50.7° were collected on a Bruker
P4 diffractometer using 1.20°-wide x scans and graphite-
˚
monochromated Mo Ka radiation (k = 0.71073 A); 3081
References and notes
of these reflections were unique. Lattice constants were
determined with the Bruker XSCANS software package
using 35 centered reflections. ‘Direct methods’ were used
to solve the structure and the resulting structural param-
eters were refined with F2 data to convergence by using
1. (a) Maryanoff, B. E. Acc. Chem. Res. 2006, 39, 831; (b)
Maryanoff, B. E. In Drug Discovery and Development;
Chorghade, M. S., Ed.; Wiley: Hoboken, NJ, 2006; Vol. 1,
pp. 313–337.

6628
M. A. Xiang et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6623–6628
counter-weighted, full-matrix, least-squares techniques
10. Receptor binding was performed with recombinant
human V1a or V2 receptor preparations derived from the
membranes of transfected HEK-293 cells. Compounds
were evaluated for their ability to displace [3H]-Arg-
vasopressin.
and a structural model that incorporated anisotropic
thermal parameters for all nonhydrogen atoms and
isotropic thermal parameters for all hydrogen atoms.
The final agreement factors are: R1 (unweighted, based on
F) = 0.056 for 2194 independent reflections having 2h(Mo
Ka) < 50.7° and I > 2r(I); R1 (unweighted, based on
F) = 0.091 and wR2 (weighted, based on F2) = 0.131 for
all 3081 independent reflections having 2h(Mo
Ka) < 50.7°. The absolute configuration assigned to the
chiral cation was confirmed experimentally by the refined
value of 0.10(7) for the ‘Flack’ absolute structure param-
11. Inhibition of receptor activation induced by Arg-vaso-
pressin was quantitated by measuring intracellular Ca2+
concentration in HEK-293 cells expressing human V1a
receptors and by measuring cAMP levels in HEK-293 cells
expressing human V2 receptors.
12. Inhibition of AVP-induced activation of rat V1a (Ca2+
and V2 (cAMP) receptors in transfected HEK-293 cells.
)
eter. The carboxyl (H1O) and amine (H1N1 and H1N2
)
13. Details on the aquaresis studies with 3 in different animal
species will be published separately.
14. Details on the V1a antagonist activity of 3 in in vivo assays
will be published separately.
hydrogen atoms were located from a difference Fourier
map and refined as independent isotropic atoms. The
positional parameters of the amine hydrogens were
allowed to vary in least-squares cycles, but the carboxyl
hydrogen was placed at an idealized tetrahedral position,
with free rotation about the C–O bond in least-squares
cycles. The remaining hydrogen atoms were included in
the structure-factor calculations as idealized atoms
(assuming sp2- or sp3-hybridized carbon atoms and C–H
15. Compound 3 has good pharmacokinetics in multiple
species, does not display hERG binding, does not inhibit
cytochrome P450 enzymes, does not develop tolerance on
chronic administration, and shows no adverse CNS,
cardiovascular, metabolic, or endocrine effects in vivo at
elevated doses. At suprapharmacological doses, 3 did not
display the unexpected toxicology that was observed with 2.
16. (a) Conivaptan16b,c is a dual V1a/V2 receptor antagonist
that has received marketing approval in the USA for
treating hyponatremia, albeit by intravenous administra-
tion only (http://www.fda.gov/Cder/drug/InfoSheets/
patient/conivaptanPIS.htm; accessed Aug 2007). For
background on conivaptan (YM-087), see: (b) Matsuhisa,
A.; Taniguchi, N.; Koshio, H.; Yatsu, T.; Tanaka, A.
Chem. Pharm. Bull. 2000, 48, 21; (c) Tahara, A. et al. J.
Pharmacol. Exp. Ther. 1997, 282, 301.
˚
bond lengths of 0.94–0.99 A) ‘riding’ on their respective
carbon atoms. The isotropic thermal parameters for H1O
,
1N1, and H1N2 refined to final Uiso values of 0.18(5),
H
2
˚
0.04(1), and 0.07(2) A , respectively; those for the remain-
ing hydrogens were fixed at values of 1.2 (nonmethyl) or
1.5 (methyl) times the equivalent isotropic thermal param-
eters of the carbon atoms to which they are bonded.
Atomic coordinates have been deposited with the Cam-
bridge Crystallographic Data Centre (Accession No.:
CCDC 660933).