2,5-Disubstituted 3,4-dihydro-2H-benzo[b][1,4]thiazepines as potent and selective V2 arginine vasopressin receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2003.08.051

A number of 2,5-disubstituted benzothiazepines were synthesized and screened for their ability to inhibit arginine vasopressin binding to the human V₂ and V_1a receptor subtypes. The more active compounds were subsequently analyzed for their antagonist activity in in vitro functional assays. The SAR showed a preference for an acidic unit appended from the benzothiazepine scaffold. This substitution pattern afforded the most potent and selective analogues in the series. The carboxymethyl analogue 4, showed a 140-fold greater selectivity for the V₂ over the V_1a receptor in the binding assay. In the cell-based functional assays this analogue was a potent and selective antagonist of the V₂ receptor. The in vitro SAR of the series and a description of the in vivo studies around compound 4 is described.

Full text of DOI:10.1016/j.bmcl.2003.08.051

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4031–4034
2,5-Disubstituted 3,4-dihydro-2H-benzo[b][1,4]thiazepines as
Potent and Selective V2 Arginine Vasopressin Receptor
Antagonists
Maud J. Urbanski,* Robert H. Chen,y Keith T. Demarest, Joseph Gunnet, Richard Look,
Eric Ericson, William V. Murray, Philip J. Rybczynski and Xiaoyan Zhang
Drug Discovery, Johnson & Johnson Pharmaceutical Research & Development L.L.C., 1000 Route 202,
Raritan, NJ 08869-0602, USA
Received 22 May 2003; accepted 25 August 2003
Abstract—A number of 2,5-disubstituted benzothiazepines were synthesized and screened for their ability to inhibit arginine vaso-
pressin binding to the human V2 and V1a receptor subtypes. The more active compounds were subsequently analyzed for their
antagonist activity in in vitro functional assays. The SAR showed a preference for an acidic unit appended from the benzothiaze-
pine scaffold. This substitution pattern afforded the most potent and selective analogues in the series. The carboxymethyl analogue
4, showed a 140-fold greater selectivity for the V2 over the V1a receptor in the binding assay. In the cell-based functional assays this
analogue was a potent and selective antagonist of the V2 receptor. The in vitro SAR of the series and a description of the in vivo
studies around compound 4 is described.
# 2003 Elsevier Ltd. All rights reserved.
Vasopressin is a nine amino acid pituitary peptide that
exerts multiple actions throughout the body by inter-
acting with three receptor subtypes, V1a, V1b and V2.
The V2 receptors, which are localized in the renal col-
lecting ducts, regulate water resorption.1 Stimulation of
these receptors by vasopressin, reduces the output of
urine exerting an antidiuretic or ‘water-sparing’ affect.
A V2 antagonist acts as an aquaretic, increasing water
output without promoting the loss of electrolytes.2
Diuretics are currently used to treat conditions such as
congestive heart failure, inappropriate antidiuretic hor-
mone secretion (SAIDH), brain edema, cirrhosis, and
hyponatremia. A V2 receptor antagonist could serve as
a potassium sparing diuretic thereby eliminating the
dangerous electrolyte imbalances associated with existing
diurectics.3
scaffold linked through N-1 to a substituted p-amido-
benzoyl side chain. This side chain contains key phar-
macophore elements and has therefore been preserved
in several of the reported compounds. Ancillary sub-
stitutions to the benzazepine scaffold ring consist pri-
marily of basic amines or nitrogen containing
heterocycles that were appended or fused to the benza-
zepine ring as in the first example of a non-peptide
vasopressin antagonist, OPC 312605 (1) and the clinical
candidate, Lixivaptan6 (2) (Fig. 1). These basic units
have been shown to improved compound potency and
aqueous solubility.
The biphenyl substitution pattern in the p-amido-
benzoyl side chain, exemplified in the mixed V1a/V2
antagonists Conivaptan7 (3), was preserved in several of
the benzothiazepines described and the SAR explored.
In contrast to many of the previously reported benza-
zepine compounds, the present studies show that a side
chain containing an acidic moiety extended from the 2-
position of the benzothiazepine scaffold provided a more
potent and selective V2 receptor antagonist 4 (Fig. 1). In
fact most of the derivatives containing an alkyl chain
substituted with a carboxylic acid or an acid bioisostere
show good potentcy and selectivity for the V2 receptor.
Several non-peptide V2 receptor antagonists have
appeared in the literature over the last decade.4 A num-
ber of the compounds described contain a benzazepine
*Corresponding author. Tel.: +1-908-704-4741; fax: +1-908-203-
8109; e-mail: murbansk@prdus.jnj.com
yCurrent address: Ace Chemicals and Pharmaceutical, M.L. King Jr.
Blvd., Newark, NJ 07102, USA.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.08.051

4032
M. J. Urbanski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4031–4034
The substituted benzothiazepines were synthesized as
illustrated in Scheme 1. The intermediate lactam 6 was
prepared in two steps by an addition of commercially
available 2-aminothiophenol to 5,6-dihydropyran-2-one
(5) followed by heating in refluxing xylene to promote
cyclization in 61% overall yield.8 Subsequent reduction
with borane-tetrahydrofuran complex provided the 2-
hydroxyethyl substituted benzothiazepine 7. The sub-
stitiuted p-amidobenzoyl side chain was introduced by
an in-situ silyl protection of the hydroxyl group using
N,O-bis(trimethylsilyl)-acetamide, followed by amine
acylation to form the 2,5-disubstituted benzothiazepine
8.
Alternatively, the intermediate hydroxyl group could be
protected as the TBDMS ether, and the substituted
p-amidobenzoylchloride introduced in either one step or
a stepwise fashion to allow SAR exploration of the aryl
rings in this portion of the molecule. The substituted
p-amidobenzoylchloride was prepared in three steps
from commercially available 4-aminomethylbenzoate
using standard procedures. The TBDMS protected
intermediates were treated with methane sulfonic acid to
remove the protecting group. By either route, a final
pyridinium dichromate oxidation provided the
carboxylic acid product 9.
As shown in Scheme 2, the hydroxyl intermediate 8
provided several opportunities to expand the SAR in
this series. Treatment of this intermediate 8 with di-
ethylaminosulfurtrifluoride (DAST) provided the alkyl
fluoride analogue 10. Oxidation of 8 with PCC to the
corresponding aldehyde followed by treatment with
DAST afforded the difluorinated analogue 11. PCC
oxidation also oxidized the sulfur to the sulfoxide;
however, subsequent treatment with DAST converted
the sulfoxide back to the sulfide.9 The hydroxyl group
of 8 was converted to the mesylate to provide the inter-
mediate 13. This versatile intermediate was readily dis-
placed with aliphatic amines to introduce basic groups
as in example 12, converted to the bromide then heated
in the presence of sodium sulfite to provide the sulfonic
acid analogue 14 or treated with sodium cyanide to
yield the nitrile 16. The nitrile was then further hydro-
lyzed to yield the homologated carboxylic acid analogue
Figure 1. Non-peptide V2 receptor antagonist.
Scheme 1. (i) CH2Cl2, Et3N (99%); (ii) xylene, reflux (61%); (iii) BH3
THF; (iv) THF, N,O-Bis(trimethylsilyl)acetamide; CH2Cl2, sub-
stituted p-amidobenzoyl chloride (85%); (v) (a) THF, N,O-bis
(trimethylsilyl)acetamide; CH2Cl2, p-nitrobenzoyl chloride; (b) H2, Pd/
C, EtOH; (c) substituted benzoylchloride (vi) PDC, DMF.
Scheme 2. (i) PCC, Celite, CH2Cl2 (55%); (ii) DAST, CH2Cl2 (61%);
(iii) MsCl, Et3N, CH2Cl2; (iv) NR2R4, 80 ꢀC; (v) (a) LiBr, THF;
(b) Na2SO3, EtOH/H2O; (vi) NaCN, DMF; (vii) NaN3, NH4Cl,
DMF; (viii) NaOH/ MeOH.

M. J. Urbanski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4031–4034
4033
15 or treated with sodium azide to provided the tetrazole
17, a carboxylic acid bioisostere.10
and 148-fold greater selectivity for the V2 receptor.
Compound 4 was selected for further studies.
The enantiomers of the intermediate 23 were resolved
by the addition of a chiral isocyanate, to form
diastereomeric carbamates 18 as illustrated in Scheme 3.
The diastereomers were easily separated by column
chromatography.11 Once isolated, each pure diaste-
reomer was hydrolyzed back to the corresponding alco-
hols using sodium methoxide. The final PDC oxidation
step provided the enantiomerically pure acids 4 and 21
in greater that 98% ee.12
Strong species differences have been reported for vaso-
pressin antagonists.13 To explore that possibility, com-
pound 4 was tested in functional assays using cells
expressing rat V1a or V2 receptors. Compound 4 had
functional IC50 values of 0.017 and 0.133 mM against
the rat V2 and V1a, respectively. This compound was
therefore similar in potency at both the human and rat
V2 receptor but showed a greater selectivity with the
human receptors.
In the in vitro binding assay, substituted benzothiaze-
pines analogues were evaluated for their ability to dis-
place H-arginine vasopressin from cloned human V2
The acute diuretic effect of 4 was measured in vivo in
Sprague–Dawley rats (Table 2). The animals were dosed
orally with either the vehicle (0.5% hydroxypropyl–
methylcellulose), or the test compound. The urine was
then collected over a 4-h period. As shown in Table 2,
compound 4 demonstrated good aquaretic activity,
producing a dose-related increase in urine volume and an
associated decrease in urine osmolality. A significant
increase in urine output was observed at doses of 10 mg/kg
3
and V1a receptors. The functional activity was then
subsequently determined by measuring the activation or
inhibition of vasopressin induced cAMP accumulation
and calcium immobilization, respectively, in V2 and V1a
receptor expressing cells. Table 1 summarizes the results
of these studies. A biphenyl substitution on the p-ami-
dobenzoyl side chain was preferred to achieve the high-
est level of potency. However, selectivity for the V2
receptor was clearly enhanced with the addition of the
appended carboxylic acid side chain to the 2-position of
the benzothiazepine scaffold as in example 4, 20 and 21.
Diminished activities were observed in both V1a and V2
binding assays with the addition of amine appendages
as in example 22. The analogues containing alkyl chains
terminating with either a hydroxyl group (23) or with
fluorides (24, 25) maintained good binding affinity yet
displayed a mixed selectivity profile for the two receptor
subtypes. Good potency and selectivity were preserved
with bioisosteric replacements (28, 29) and with carboxy
side chain homologation (26, 27). In the functional
assay, both enantiomers 4 and 21 of the most potent
and selective benzothiazepine analogue displayed a 120-
Table 1. In-vitro binding and functional data
Compd R1
Config.
Binding
Functional
(IC50, mMa)
(IC50,mMa)
V1a
V2
V1a
V2
4
20
21
CO2H
CO2H
CO2H
N(CH3)2
CH2OH
CH2F
S
R/S
R
R/S
R/S
R/S
R/S
R
1.14
0.729
0.44
1.2
0.008
0.007
0.007
0.43
0.008
0.085
0.1
0.015
0.016
0.012
1.76
0.74
2.13
NDc
0.15
NDc
NDc
2.01
4.63
NDc
NDc
0.03
0.04
0.036
NDc
0.1
NDc
NDc
0.03
0.04
NDc
NDc
22
23
24
25
26
27
28
29
0.097
0.18
0.16
0.14
0.314
1.15
CHF2
CH2CO2H
CH2CO2H
CH2SO3H
CH2tetrazole
S
R/S
R/S
65%b 0.02
aIC50 values unless otherwise specified.
bPercent inhibition at 0.2 mM.
cND, not determined.
Table 2. Effect of Compound 4 on urine volume and osmolality in
conscious hydrated male rats
Treatment Number of animalsb Urine volume Urine osmolality
dose
(mL)a
(mOsmol/Kg)a
Vehicle
10 (9)
10
10
10 (9)
10 (9)
1.3 (ꢁ0.2)
2.2 (ꢁ0.3)
843 (ꢁ123)
529 (ꢁ63)*
299 (ꢁ31)*
161 (ꢁ15)*
177 (ꢁ11)*
3 mg/kg
10 mg/kg
30 mg/kg
100 mg/kg
7.3 (ꢁ1.0)*
15.7 (ꢁ2.1)*
13.8 (ꢁ0.9)*
*p<0.05 versus vehicle values.
aEach value represents the meanꢁSE.
bValues in parentheses indicate the number of samples available for
osmolality determination.
Scheme 3. (i) S-(ꢂ)-a-Methylbenzylisocyanate, (CH3)2NCH2CH2OH,
toluene (97%); (ii) NaOCH3/CH3OH (97%); (iii) PDC, DMF, (60%).

4034
M. J. Urbanski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4031–4034
and greater while osmolality decreased at all doses tes-
ted (3 mg/kg and above). Maximal aquaretic efficacy
was seen at 10 mg/kg with no greater effects apparent at
100 mg/kg.
P.; de Garabvilla, L.; Demarest, K. T.; Ericson, E.; Greco,
M. N.; Gunnet, J. W.; Hageman, W.; Hecker, L. R.; Look, R.;
Moore, J. B.; Maryanoff, B. E. Bioorg. Med. Chem. Lett. 2003,
13, 753. (c) Trybulski, E. In Ann. Rep. Med. Chem., Doherty,
A. H., Hagmann, W. K., Eds.; Academic, 2001; pp 159–168.
5. (a) Yamamura, Y.; Ogawa, H.; Yamashita, H.; Chihara,
T.; Miyamoto, H.; Nakamura, S.; Onogawa, T.; Yamashita,
T.; Hosokawa, T.; Mori, T. Br. J. Pharmacol. 1992, 105, 787.
(b) Martinez-Castelao, A. Curr. Opin. Cardiovasc. Pulmonary
Renal Investig. Drugs 1999, 1, 423.
A new series of substituted benzothiazepine analogues
were synthesized and shown to inhibit binding of vaso-
pressin to the V1a and V2 receptors. Several of these
compounds also demonstrated good antagonist activity
in an in vitro functional assay. Analogues containing
carboxylic acid or acid isosteres appended from the
benzothiazepine ring, showed higher selectivity for the
V2 over the V1a receptor subtype in vitro. One example
of this, compound 4, was orally active in vivo, showing
aquaretic activity at doses as low as 3 mg/kg. The pro-
duction of hypo-osmotic urine differentiates this and
other V2 antagonists from the diuretics, such as fur-
osemide and hydrochlorothiazide, now used in the clinic.
6. Martinez-Castelao, A. Curr. Opin. Investig. Drugs 2001, 2,
525.
7. (a) Taniguchi N., Tanaka, A., Matsuhisa, A., Sakamoto,
K., Kishio, H., Yatsu, T. WO 9506035, 1995. (b) Tahara, A.;
Tomura, Y.; Wada, K. I.; Kusayama, T.; Tsukada, J.; Taka-
nashi, M.; Yatsu, T.; Uchida, W.; Tanaka, A. J. Pharmacol.
Exp. Ther. 1997, 282, 301.
8. Urbanski, M. J.; Chen, R. H. WO 0132639, 2001. Chem.
Abstr. 2001, 134, 353326. Preparation of this intermediate is
described herein.
9. Robins, M. J.; Wnuk, S. F. Tetrahedron Lett. 1988, 29,
5729.
References and Notes
10. Dhanoa, D. S.; Bagley, S. W.; Chang, R. S.; Lotti, V. J.;
Chen, T. B.; Kivlighn, S. D.; Zingaro, G. J.; Siegl, P. K.;
Patchett, A. A.; Greenlee, W. J. J. Med. Chem. 1993, 36, 4230.
11. Chiralpak AD 5 cmꢃ50 cm; Chiral Technologies, Inc.:
730 Springdale Dr., Exton, PA, USA.
1. Reeves, W. B.; Andreoli, T. E. In Williams Textbook of
Endocrinology, 8th ed, Wilson, J. D., Foster, D. W., Eds.;
W. B. Saunders: Philadelphia, 1992; pp 331–356.
2. Mayinger, B.; Hensen, J. Exp. Clin. Endocrinol. Diabetes
1999, 107, 157.
3. (a) Palm, C.; Reimann, D.; Gross, P. Cardiovasc. Res. 2001,
51, 403. (b) Paranjape, S. B.; Thibonnier, M. Expert Opin.
Investig. Drugs 2001, 10, 825. (c) Laszlo, F. A.; Laszlo, F., Jr.;
De Wied, D. Pharmacol. Rev. 1991, 43, 73.
12. The absolute stereochemistry of compound 4 was deter-
mined by crystal structure analysis of an intermediate pre-
pared by Chemical Development at JJPRD. Hydroquinine
was used as the counterion for this analysis. We would like to
acknowledge Fuqiang Liu and Ann Gelormini for scale-up
efforts and contributions to this project.
4. (a) Albright, J. D.; Chan, P. S. Curr. Pharm. Des. 1997, 3,
615. (b) Matthews, J. M.; Hoekstra, W. J.; Andrade-Gordon,
13. Ruffolo, R. R.; Brooks, D. P.; Huffman, W. F.; Poste, G.
Drug News Perspect. 1991, 4, 217.