New benzylureas as a novel series of potent, nonpeptidic vasopressin V2 receptor agonists

Article abstract of DOI:10.1021/jm8008162

Vasopressin (AVP) is a hormone that stimulates an increase in water permeability through activation of V2 receptors in the kidney. The analogue of AVP, desmopressin, has proven an effective drug for diseases where a reduction of urine output is desired. However, its peptidic nature limits its bioavailability. We report herein the discovery of potent, nonpeptidic, benzylurea derived agonists of the vasopressin V2 receptor. We describe substitutions on the benzyl group to give improvements in potency and subsequent modifications to the urea end group to provide improvements in solubility and increased oral efficacy in a rat model of diuresis. The lead compound 20e (VA106483) is reported for the first time and has been selected for clinical development.

Full text of DOI:10.1021/jm8008162

8124
J. Med. Chem. 2008, 51, 8124–8134
New Benzylureas as a Novel Series of Potent, Nonpeptidic Vasopressin V2 Receptor Agonists
Christopher M. Yea,*,† Christine E. Allan,† Doreen M. Ashworth,‡ James Barnett,‡ Andy J. Baxter,‡ Janice D. Broadbridge,‡
Richard J. Franklin,‡ Sally L. Hampton,† Peter Hudson,‡ John A. Horton,‡ Paul D. Jenkins,‡ Andy M. Penson,‡ Gary R.
W. Pitt,‡ Pierre Rivie`re,§ Peter A. Robson,† David P. Rooker,† Graeme Semple,‡ Andy Sheppard,† Robert M. Haigh,† and
Michael B. Roe†
Vantia Ltd, Southampton Science Park, Southampton SO16 7NP, United Kingdom, Ferring Research Ltd, Southampton Science Park,
Southampton SO16 7NP, United Kingdom, and Ferring Research Institute Inc., 3550 General Atomics Court, San Diego, California 92121
ReceiVed October 3, 2008
Vasopressin (AVP) is a hormone that stimulates an increase in water permeability through activation of V2
receptors in the kidney. The analogue of AVP, desmopressin, has proven an effective drug for diseases
where a reduction of urine output is desired. However, its peptidic nature limits its bioavailability. We
report herein the discovery of potent, nonpeptidic, benzylurea derived agonists of the vasopressin V2 receptor.
We describe substitutions on the benzyl group to give improvements in potency and subsequent modifications
to the urea end group to provide improvements in solubility and increased oral efficacy in a rat model of
diuresis. The lead compound 20e (VA106483) is reported for the first time and has been selected for clinical
development.
Introduction
The cyclic nonapeptide hormone arginine vasopressin (AVPa)
plays a crucial role in water balance. It is released from the posterior
pituitary in response to decreased blood pressure or volume or by
an increase in plasma osmolality. It stimulates an increase in water
permeability in the collecting ducts of the kidney by activation of
V2 receptors and subsequent regulation of aquaporin-2 water
channels, ultimately leading to a decrease in urine volume.
Additionally, AVP interacts with receptor subtypes V1a and V1b,
resulting in a number of effects including contraction of vascular
Figure 1. Nonpeptide V2 agonists.
smooth muscle, platelet aggregation, glycogenolysis, ACTH, and
ꢀ-endorphin release.1-7 The only commercially available V2
Table 1. Potencies of Compounds 3a-g
agonist, desmopressin (DDAVP), is effective as a treatment for
diseases where a reduction of urine output is desired including
diabetes insipidus, enuresis, and nocturia.8,9 While it is successfully
administered orally, its peptidic nature limits its bioavailability.
Increased utility of this therapeutic class has been enhanced by
identification of orally available nonpeptides such as 1 and 2 (Figure
1), but these agonists have yet to progress to market.10,11 There is
therefore a need to develop new orally active agonists that may be
used as treatments for indications where decreased urine production
would have clinical benefit. We have previously reported the
syntheses of libraries targeted to the vasopressin V2 and the related
oxytocin receptors.12,13 Screening of these libraries identified 3a
as a hit compound (Table 1). We report here, for the first time, the
functional activity at the human V2 receptor of the resulting series
of novel agonists and their SAR, leading to the selection of 20e
(VA106483) as a clinical candidate.
Compound
R
EC50 (nM)
n
3a
3b
3c
3d
3e
3f
H
2-F
47 ( 15
180 ( 40
40 ( 25
110 ( 30
2.1 ( 1.7
10 ( 4
3
3
5
3
6
3
6
2-Cl
2-OMe
3-Me
3-Et
3,5-DiMe
3g
82 ( 73
For example, compound 3a was prepared starting with 4-cy-
anobenzoic acid 4a, which was coupled with 2,3,4,5-tetrahydro-
1H-1-benzazepine 5 to give 7a. The nitrile was reduced by
hydrogenation to give 8a. Finally the urea was prepared from
2,6-difluorophenylisocyanate 17 to afford the test compound 3a.
The aromatic ureas in Table 2 (18a-c) and the aliphatic ureas
in Table 3 (19a-l) were prepared from the common methyl
substituted benzylamine intermediate 8e (Scheme 2). Reaction
with the appropriate isocyanate analogously to the preparation
of 3e provided 18a, 18b, and 19c. Coupling of 8e to the
appropriate amine with CDI gave 19d, e, and j. For the
preparation of 18c, diphenylphosphoryl azide was used to
prepare the isocyanate in situ via a Curtius rearrangement.
Carboxylic acid derivative 19a was prepared by hydrolysis of
glycine ester 19c. 19a was subsequently coupled to alcohols or
Chemical Synthesis
The chemistry to prepare the difluorophenyl ureas in Table
1 is shown in Scheme 1, details of which have been disclosed.14,15
* To whom correspondence should be addressed. Phone: +44 23 8076
3440. Fax: +44 23 8076 6755. E-mail: chris.yea@vantiatherapeutics.com.
†
Vantia Ltd.
Ferring Research Ltd.
Ferring Research Institute Inc.
‡
§
a Abbreviations: AVP, arginine vasopressin; CDI, 1,1′-carbonyldiimi-
dazole; DAST, (diethylamino)sulfur trifluoride; DDAVP, 1-desamino-8-D-
arginine-vasopressin; WSCDI, 1-ethyl-3-(3′-dimethylaminopropyl)carbo-
diimide·HCl.
10.1021/jm8008162 CCC: $40.75
2008 American Chemical Society
Published on Web 11/24/2008

Benzylureas as Vasopressin V2 Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24 8125
Scheme 1a
a Reagents and conditions: (a) nBuLi, THF, CO2, -78°C (56-73%); (b) SOCl2, toluene, reflux, 2.5 h then MeOH, CH2Cl2 (77%); (c) N-bromosuccinimide,
CCl4 then NH3, EtOH, (BOC)2O, NaOH, dioxane, H2O (30%); (d) HCl, NaNO2, H2O, 0-5 °C then Na2CO3, CuCN, KCN, H2O, 70 °C (50-69%); (e)
LiOH, dioxane, H2O (95-97%); (f) 2,3,4,5-tetrahydro-1H-1-benzazepine 5, DMAP, WSCDI, Et3N, CH2Cl2, reflux (69-77%); (g) 4N HCl/dioxan (60%);
(h) for the preparation of 8a, 8e, and 8f, 10% Pd/C, HCl(aq), MeOH (74-98%); (i) for the preparation of 8c, 8d, and 8g, NaBH4, CoCl2 ·6H2O MeOH
(57-69%); (j) 2,6-difluorophenylisocyanate 17, Et3N, CH2Cl2 (31-88%).
Table 2. Potencies of Compounds 18a-c Compared to 3e
proline was coupled with dimethylamine, deprotected, and
coupled with 8e to provide 20g. The thioamide 20h was
prepared by treating BOC-proline-N,N-dimethylamide 21 with
Lawesson’s reagent to give 22, followed by hydrogen chloride
solution in 1,4-dioxane to give 23. The usual coupling with 8e
gave 20h.
The substituted proline derivatives 21a-h shown in Table 5
were prepared according to Schemes 3, 4, and 5. Scheme 3
describes the routes that began with (4R)-BOC-4-hydroxyproline
24. This allowed access to the hydroxyl analogue 21a, by
coupling of 24 to dimethylamine, followed by deprotection, then
CDI coupling to the intermediate 8e as described earlier.
Subsequent derivatization to the ethyl ether 21c was achieved
with iodoethane. Oxidation of 21a with Dess-Martin periodi-
nane afforded the carbonyl derivative 21f. Application of the
Dess-Martin periodinane, this time to the hydroxyproline
intermediate 25, provided a substrate 27 suitable for fluorination
with DAST. Subsequent deprotection and CDI coupling with
8e provided the difluoro derivative 21g. Scheme 4 describes
the syntheses of the methyl, tert-butyl, and benzyl ether
derivatives, 21b, 21d, and 21e. The protected proline methyl
ether 30b is known and the protected benzyl and tert-butyl ethers
30e and 33d were commercially available.16 These were coupled
with dimethylamine, deprotected, and coupled to intermediate
8e with CDI. Scheme 5 describes the synthesis of the dimethyl
acetal derivative beginning with the known compound 4,4-
dimethoxy-L-proline methyl ester, 35h.17 In this instance, the
order of reactions was modified such that coupling with
intermediate 8e was undertaken first, followed by ester hydroly-
sis and coupling to dimethylamine.
amines using established coupling procedures to yield the
remaining compounds in Table 3.
The proline analogues 20a-g and the thioamide proline
analogue 20h in Table 4 were also prepared from the common
intermediate 8e using CDI (Scheme 2). Compounds 20b-h were
prepared directly from the appropriate proline derivative,
whereas the carboxylic acid 20a was obtained by hydrolysis of
the methyl ester 20b. The starting prolines were available
commercially for compounds 20b-e. The proline derivatives
required for the syntheses of compounds 20f and 20g were
obtained by conventional methods (not shown in scheme).
Specifically, BOC-L-proline was coupled with diethylamine and
deprotected before coupling with 8e to provide 20f. BOC-D-
The syntheses of compounds 22a-g exploring the SAR about
the azepine ring system (Table 6) are shown in Schemes 6, 7,
and 8. Scheme 6 begins with the nitrile intermediate 7e from
Scheme 1. Oxidation of the 5-position of the benzazepine end

8126 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24
Yea et al.
Scheme 2a
a Reagents and conditions: (a) R-NCO, Et3N, CH2Cl2 (65-81%); (b) 2-hydroxynicotinic acid, diphenylphosphoryl azide, Et3N, dioxane, reflux then 8e,
i
i
reflux, 18 h (43%); (c) LiOH, THF, H2O (46-99%); (d) coupling to alcohol or amine; (e) CDI, Pr2NEt, DMF, 90 min then RR′NH, Pr2NEt, DMF, 18 h
(30-65%); (f) Lawesson’s reagent, toluene, reflux, 2 h (70%); (g) 4N HCl/dioxane (100%).
group provided the carbonyl derivative 37a. This was amenable
to reduction, concomitantly with the nitrile reduction, to provide
the hydroxyl substituted intermediate 38a. Subsequent CDI
coupling to the commercially available L-proline-N,N-dimethy-
lamide provided the target hydroxylated compound 22a. How-
ever, the diastereoisomeric products of this route were insepa-
rable. The difluoro analogue 22c was prepared in a similar
sequence but with the inclusion of a DAST fluorination of the
carbonyl intermediate 37a. The 5-methylbenzazepine and the
benzoxazepine end groups 41b and 41d were prepared from
the corresponding tetralone 39b and chromanone 39d, respec-
tively, as shown in Scheme 7. These were converted to their
respective oximes and reduced with regiospecific rearrangement
with DIBAL-H.18 The remaining, tricyclic, top groups shown
in Table 6 were prepared according to known procedures.19-21
Scheme 8 shows the sequence for coupling top end groups 41b
and 41d-g with 4-cyano-3-methylbenzoic acid 4e, subsequent
reduction, and finally the CDI coupling with L-proline-N,N-
dimethylamide to yield the target compounds 22b and 22d-g.
The diastereoisomers of 22b were inseparable.
cAMP responsive gene elements (“reporter gene”) and subse-
quently measured as luminescence. EC50 values were determined
and data are reported as geometric means ( standard deviation.
In vivo data were obtained using Brattleboro rats, a strain of
Long-Evans hooded rats that have no circulating AVP, thus
constituting a model of diabetes insipidus.22 These animals
produce large volumes of dilute urine (approximately 10 mL/
h), which were collected and measured over 1 h time periods.
Following oral compound administration in carboxymethyl
cellulose, voided urine volumes were measured and compared
to vehicle-treated animals. Each treatment group contained at
least six animals and data are expressed as means.
Results and Discussion
The first SAR to be investigated was around the central linker.
The hit compound 3a, with an unsubstituted aromatic linker,
was potent (EC50 ) 47 nM) and fully efficacious in a human in
vitro assay. However, in the equivalent rat assay, the maximum
response was only 30% of the maximum AVP response (data
not shown), rendering it unsuitable as a pharmacological tool.
In an effort to improve the efficacy at the rat receptor and
provide a pharmacological tool, a series of compounds with
aromatic substituents was prepared. Table 1 shows the potencies
of these compounds at the human V2 receptor. Addition of small
substituents to either the 2- or the 3-position was generally well
tolerated. In particular, small alkyl substituents such as methyl
or ethyl improved potency significantly. The 3-methyl substitu-
Biological Assays
In vitro potency and relative efficacy data were obtained by
expressing the V2 receptor in human embryonic kidney
(HEK293) cells. On agonist stimulation the V2 receptor signals
an increase in intracellular cAMP levels. These are amplified
through the production of luciferase protein by cotransfected

Benzylureas as Vasopressin V2 Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24 8127
Table 5. Potencies of Compounds 21a-h Compared to 20e
Table 3. Potencies of Compounds 19a-l
Table 4. Potencies of Compounds 20a-h
Compound
R
EC50 (nM)
n
20a
20b
20c
20d
20e
20f
(S)-COOH
680 ( 110
100 ( 75
950 ( 350
390 ( 80
24 ( 40
350 ( 140
380 ( 140
3.2 ( 0.5
3
6
3
3
6
3
3
3
(S)-COOMe
(S)-COOiPr
(S)-COOtBu
(S)-CONMe2
(S)-CONEt2
(R)-CONMe2
(S)-CSNMe2
Alternative derivatizations were therefore designed with the aim
of incorporating more hydrophilic groups (Table 3). The glycine
derivative, 19a (EC50 ) 180 nM), was targeted. Although
incorporation of the acidic functionality contributed to a
significant loss in potency compared to the difluorophenyl
analogue 3e, the compound retained maximum efficacy and was
a useful lead for a new series of amino acid derivatives. Thus
a series of esters (19b-g) and amides (19h-l) were prepared
of increasing size. While small ester groups such as methyl,
ethyl, iso-propyl, and tert-butyl (19b-e) displayed the highest
potency, it was judged that these would likely be hydrolyzed
rapidly to the acid in vivo. Incorporation of bulkier esters phenyl
and benzyl (19f-g) resulted in a reduction in potency. The
amides that were prepared retained potency compared to the
glycine lead. In particular, the synthesis of 19j (EC50 ) 58 nM)
resulted in the desired improvement in solubility from <4 µg/
mL to 177 µg/mL.
20g
20h
ent was highly potent (3e, EC50 ) 2.1 nM) and showed 20-fold
improvement over 3a. In addition, the substituted compounds
showed full and potent agonism of the rat receptor in the in
vitro assay. Using the highly potent 3e as the lead for the next
SAR study, the urea substituent was investigated (Tables 2, 3,
and 4).
Removal of one of the fluorine substituents from the
difluorophenylurea resulted in a 10-fold reduction in potency
(compound 18a, EC50 ) 30 nM). Removal of both fluorine
substituents reduced potency a further 4-fold (compound 18b,
EC50 ) 130 nM). It was attempted to mimic the interaction, if
any, of the fluorine with the receptor with an oxygen substituent.
In vivo studies with 3e and 19j confirmed the improvement
in oral activity hypothesized by increasing solubility. Treatment
of Brattleboro rats with 3e at 1 mg/kg po in methylcellulose
had no effect on urine output over a 5 h period. However,
treatment with 19j at the same dose resulted in marked
antidiuresis for 1-2 h.12 Following that period, no reduction in
This was achieved through the pyridine analogue, 18c (EC50
)
140 nM), but it was equipotent with the unsubstituted phenyl
group. There was extensive further work on the SAR of aromatic
ureas, and while potent and efficacious agonists were designed,
the compounds all suffered from limited aqueous solubility.

8128 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24
Yea et al.
Scheme 3a
a Reagents and conditions: (a) Me2NH·HCl, DMAP, iPr2NEt, CH2Cl2 (100%); (b) 4N HCl/dioxane (17-100%); (c) CDI, iPr2NEt, DMF, 90 min then 8e,
iPr2NEt, DMF, 18 h (66-80%); (d) NaH, EtI, DMF (39%); (e) Dess-Martin periodinane, CH2Cl2 (32-50%); (f) DAST, CH2Cl2 (35%).
Scheme 4a
i
a Reagents and conditions: (a) Me2NH.HCl, WSCDI, HOBT, Et3N, CH2Cl2 (57-100%); (b) 4N HCl/dioxane (75-100%); (c) CDI, Pr2NEt, DMF, 90
i
min then 8e, Pr2NEt, DMF, 18 h (24-75%); (d) piperidine, DMF (100%).
Scheme 5a
a Reagents and conditions: (a) 8e, CDI, DMF (57%); (b), LiOH ·H2O, THF, H2O (98%); (c) Me2NH·HCl, WSCDI, HOBT, Et3N, CH2Cl2 (50%).
urine output was detected. It was further hypothesized that the
urea functionality may be limiting intestinal permeability. We
believed that if the number of hydrogen bond donors could be
reduced then absorption may be increased, thus a series of cyclic
proline derivatives was prepared (Table 4). The L-proline
analogue (20a, EC50 ) 680 nM) showed a reduction in potency
compared to the glycine analogue 19a (EC50 ) 180 nM).
Nonetheless, a series of small ester and amide derivatives was
synthesized (20b-20f). These showed equivalent or improved
potency compared to the carboxylic acid 20a, with the most
potent being the dimethylamide derivative 20e (EC50 ) 24 nM).
Having introduced a chiral center, the opposite, unnatural
stereoisomer of this molecule was prepared (20g, EC50 ) 380
nM) but it showed more than 10-fold reduction in potency
compared to the natural stereoisomer. Interestingly, the thioa-
mide analogue 20h (EC50 ) 3.2 nM) showed a large increase
in potency and was the first nonaromatic compound to match
the potency of 3e. However, it was deemed unsuitable for
progression because of the well-known toxicity risks associated
with thioamides.
Further studies around 20e were conducted in an effort to
explore potency and drug-like properties. The first study
explored the effect of substitution on the proline as shown in
Table 5. Some of these analogues, substituted in the 4-position,

Benzylureas as Vasopressin V2 Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24 8129
Table 6. Potencies of Compounds 22a-g Compared to 20e
data in Table 5 suggested that in order to gain potency, the
additional chiral center would be required. A selection of
compounds were tested in an equivalent in vitro functional
reporter gene assay transfected with the rat V2 receptor. These
data, shown in Table 7, showed modest improvements in
potency in vitro at the rat receptor with compounds 21a-c
compared to 20e. The Brattleboro rat model data shown in Table
7, measuring percentage inhibition of urine output, suggested
approximately equivalent responses between these compounds.
Thus, given no demonstrable superiority in vivo of any one of
substituted proline analogues and given the risk of increased
cost of goods of such compounds, it was decided to continue
to pursue SAR studies with the unsubstituted dimethylprolin-
eamide analogue, 20e, as the lead.
Thus far, all SAR studies had retained a benzazepine end
group. Indeed, benzazepines and benzodiazepines have been the
predominant scaffolds utilized by many groups for V2 ligands.23
For example, a 1,4-benzodiazepine based end group features
in the V2 agonist 2.11 Thus they have been termed privileged
structures for vasopressin and oxytocin receptors.13 To explore
the SAR, we sought to utilize similar ring systems in new
analogues of 20e. In particular, we were aware of certain
metabolic liabilities in benzazepine. For example, the azepine
ring was known to be prone to metabolic hydroxylation although
the precise position of hydroxylation had yet to be identified. It
was speculated that it was most likely the benzylic 5-position,
and we sought to block this with the analogues shown in Table
6. Compound 22a is the diastereoisomeric mixture of the
5-hydroxylated derivative of the lead compound 20e. The
hydroxyl group was speculated to both reduce the metabolic
liability of the end group and to mimic a putative metabolite.
However, it demonstrated a 10-fold loss in potency (EC50
)
230 nM), so it was decided not to pursue the separation of
isomers. A similar conclusion was reached with the methylated
derivative 22b (EC50 ) 220 nM). The 5,5-difluorobenzazepine
derivative 22c (EC50 ) 1700 nM) and the 1,5-benzoxazepine
derivative 22d (EC50 ) 1000 nM) both retained good relative
efficacy but were only weakly potent. Inclusion of pyrrole fused
to the 3,4-position of the seven-membered ring provided 22e
(EC50 ) 120 nM), the exact analogue of the known V2 agonist
2. Interestingly, 2 is reported to be a potent agonist with an
EC50 value of <1 nM.11 However, the same end group, when
applied to our benzylurea series, gave a reduced potency
compared to either 2 or to our own lead compound 20e. The
reasons for this unexpected drop in potency may only be
speculated. However, the known agonists 1 and 2 both possess
relatively small aniline derived substituents in place of the urea
substituents in our series. Also, it is reported that the steric
requirements about this end of the molecule are rigorously
limited and it may be supposed that urea derivatives of the
published compounds would be too large to be tolerated.10a
Perhaps, therefore, this difference is simply due to there being
a different binding mode for our benzylureas compared to other,
previously reported, small molecule V2 agonists. Also previ-
ously disclosed as having affinity for the vasopressin V2
receptor, albeit as end groups in compounds with antagonist
activities, were the tricyclic ring systems exemplified within
compounds 22f (EC50 ) 84 nM) and 22g (EC50 ) 140
nM).20,21,24 All three tricyclic analogues prepared, 22e, 22f, and
22g, were of similar potency. To summarize Table 6, it was
found to be possible to incorporate a number of analogues of
the benzazepine end group that retained potency. However, none
showed any improvement in potency over benzazepine itself.
were highly potent compounds. For example, incorporation of
a hydroxyl group to give 21a (EC50 ) 2.3 nM) improved
potency 10-fold over the unsubstituted compound 20e. This
provided a compound equipotent with the initial lead of 3e but
which had perceived advantages in terms of drug-likeness. Also,
additions of small alkyl groups to provide the methyl ether 21b
(EC50 ) 4.7 nM) and the ethyl ether 21c (EC50 ) 4.3 nM) were
well tolerated, giving compounds equipotent with 21a and 3e.
Larger alkyl ethers such as tert-butoxy (21d, EC50 ) 46 nM)
and benzyloxy (21e, EC50 ) 10 nM) were slightly less well
tolerated but were still equipotent with the unsubstituted proline
20e. Finally, a number of derivatives were prepared that did
not generate an additional chiral center. These included the
ketone 21f (EC50 ) 27 nM), the difluoro derivative 21g (EC50
) 50 nM), and the dimethylacetal 21h (EC50 ) 20 nM). The

8130 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24
Yea et al.
Scheme 6a
a Reagents and conditions: (a) KMnO4, Mg(NO3)2 ·6H2O, tBuOH, H2O (25%); (b) CoCl2 ·6H2O, NaBH4, MeOH (31-60%); (c) CDI, DMF then L-proline-
N,N-dimethylamide (20-86%); (d) DAST, CH2Cl2, reflux (25%).
Scheme 7a
compounds reported here is driven by the linker substituent and
its effect on binding affinity because SAR was driven by a
functional assay. In fact, in contrast to the reported binding SAR
of V2 antagonists, these compounds actually show enhanced
agonist potency when the linker substituent was at the 3-position
compared to the 2-position (Table 1). Interestingly, when 3a,
3e, or 20e were tested for binding to the V2 receptor using a
radioligand binding assay they showed only weak affinities with
Ki values ranging from 0.2 µM for 3e (n ) 2) to 2 µM for 20e
(n ) 3). All of these observations are consistent with the lack
of correlation between agonist potency and binding affinity
previously reported for dopamine receptors.26 Similarly to this
example with dopamine receptors, rather than risk the use of
a Reagents and conditions: (a) NH2OH·HCl, NaOH, EtOH (90%); (b)
potentially misleading binding data, we found that using SAR
DIBAL-H, CH2Cl2 (63-86%).
generated from a functional assay was the correct approach for
optimization of these molecules.27
These structures offer the possibility of alternative routes of
In terms of intestinal absorption, it had been earlier hypoth-
metabolism and studies are underway to investigate this.
esized that the urea functionality would be limiting. It was
Despite extensive SAR work, there were no analogues
prepared that showed any obvious advantages over 20e. Its dose
response curve in the reporter gene assay confirmed its activity
at the human V2 receptor, showing 100% efficacy compared
to the maximum response of AVP (Figure 2). The specificity
of 20e was confirmed by testing its activity in a sham
transfection where the usual assay method was performed except
observed that in fact 20e showed no improvement in Caco-2
cell permeability despite the reduced number of hydrogen bond
donors compared to 19j (data not shown). However, studies
did show that 20e was significantly more soluble than other
analogues previously studied in animal models (7000 µg/mL).
Thus it was submitted for in vivo study in Brattleboro rats
(Figure 3). A reduction in urine volume after oral administration
no V2 receptor was transfected into the cells. No biological
(3 mg/kg) was observed with almost full inhibition of urine
activity of 20e in this assay was observed (data not shown).
Also, no agonist activities were observed by 20e or the earlier
lead compounds 3a and 3e in reporter gene assays utilizing
human V1a or V1b receptors at test concentrations up to 10
output for 2 h. Toward the end of the study, 5 h after dosing,
the cumulative volume output was observed to be returning to
normal levels.
µM, thus demonstrating excellent functional selectivity of this
series of compounds for the V2 receptor. Selectivity against
Summary and Conclusion
the closely related oxytocin receptor was obtained through
binding studies, which suggested no appreciable binding (Ki >10
µM) of 3a, 3e, or 20e to this receptor. SAR of V2 antagonists
of a similar structural template has been reported to suggest
that selectivity could be modified through judicious selection
linker substituent. It was suggested a 2-chloro or 2-methyl
substituent weakened V1a affinity and enhanced V2 affinity.23,25
This is consistent with the previously reported V2 agonists 1
and 2, which both bear a 2-chloro substituent. It is not possible
to say whether the source of the V2 selectivity of the series of
The hit compound 3a was identified as a novel and potent
agonist at the human vasopressin V2 receptor. However, owing
to selectivity across species, it was not suitable for study in rat
models of antidiuresis. Addition of substituents to the aromatic
linker provided compounds such as 3e with improved efficacy
and potency at both human and rat receptors. These suffered
from poor aqueous solubility. Replacement of the aromatic urea
with aliphatic ureas, such as the glycine amide in 19j, gave
compounds with improved solubility and improved oral efficacy,
albeit with reduced potency in vitro. The proline derivative 20e

Benzylureas as Vasopressin V2 Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24 8131
Scheme 8a
a Reagents and conditions: (a) SOCl2, toluene, reflux then amine 41b, d, e, f, or g, Et3N, CH2Cl2 (45-89%); (b) CoCl2 ·6H2O, NaBH4, MeOH (52-76%);
(c) CDI, DMF then L-proline-N,N-dimethylamide (29-70%).
Table 7. Potencies of Selected Compounds at the Rat V2 Receptor in
Vitro and Reduction in Urine Production in Vivo (Brattleboro Rat
Model)
rat EC50 (nM)
% inhibition of urine output;
Compound
n ) 1
n ) 2
1 mg/kg po; 1 h post dose
3e
50
100
13
50
0
82
81
68
84
60
19j
20e
21a
21b
21c
150a
65b
41
15
61
a Mean (n ) 24). b Mean (n ) 4).
showed yet further improvement in aqueous solubility and was
used as a lead compound for further SAR studies. In vitro
potency was increased further by addition of substituents onto
the proline, but there was no parallel increase in in vivo potency.
Substituted and tricyclic analogues of benzazepine were incor-
porated. These did not improve potency but may have potential
if they usefully modify the metabolism profile. Thus, compound
20e remained the lead compound, it is orally active in rat
models, and it has been selected for further progression into
clinical studies. Compound 20e essentially mimics the V2
activity of AVP, thus it may be useful in indications where
increased V2 activity would be beneficial, for example in
diabetes insipidus. Generally, it may have clinical benefit as a
treatment for indications where decreased urine output is desired.
Figure 2. Dose response curve of 20e at the human V2 receptor in
the reporter gene assay. The control used is AVP and data are reported
as the mean of data from six independent experiments, ( standard
deviation. EC50 ) 24 nM.
Experimental Section
1H NMR spectra were recorded on a Jeol EX 270 (270 MHz)
spectrometer with reference to deuterium solvent and at room
temperature. Samples were dissolved in CDCl3 unless stated
otherwise. HRMS data were obtained on a Waters LCT Premier
time-of-flight mass spectrometer. Samples were dissolved in 0.05%
HCO2H in H2O/MeCN (1:1), which were each combined with a
solution of nifedipine and reserpine (reference compounds) in 0.05%
HCO2H in H2O/MeCN (1:1) and infused by MilliGAT pump into
the ion source. Data is quoted for MH+ unless otherwise stated.
Elemental analyses were carried out by Medac Ltd. (Egham, UK).
Compounds were routinely checked by TLC and/or by LCMS. TLC
was carried out on Merck silica gel 60 F254 precoated plates. LCMS
was carried out using a Chromolith Speedrod RP-18e column, 50
mm × 4.6 mm, with a linear gradient 10% to 90% 0.1% HCO2H/
MeCN into 0.1% HCO2H/H2O over 11 min, flow rate 1.5 mL/min.
Data was collected using a Thermofinnigan Surveyor MSQ mass
spectrometer with electospray ionization in conjunction with a
Thermofinnigan Surveyor LC system. All target compounds were
analyzed by HPLC and were found to be >95% pure. HPLC was
carried out with a Varian Pursuit XRS-Diphenyl column, 50 mm
× 4.6 mm, 5 µm particle size, with a linear gradient 10% to 90%
0.1% CF3CO2H/MeOH into 0.1% CF3CO2H/H2O over 25 min, flow
rate 1.5 mL/min. UV detection was carried out at wavelength 254
nm. Full HPLC data and chromatograms are provided in Supporting
Information. Chromatography refers to flash chromatography, which
was carried out on Merck silica gel 60 (0.040-0.063 mm). All
materials were commercially available unless otherwise noted. 3a,
Figure 3. Antidiuretic action of compound 20e dosed orally to
Brattleboro rats.
3b, 3e, 3g, 4e, 4g, 6b, 7a, 7e, 7 g, 8a, 8b, 8e, 8g, 9b, 11b, 12b,
14g, 18c, 19a, 19b, 19c, and 19h were prepared as described
previously and 3c, 3d, 3f, 7c, 7d, 7f, 8c, 8d, 8f, 18a, and 18b were
prepared using analogous methods.14 4c, 4d, 4f, 14f, 16c, and 16d
were prepared as described previously.15
1-(4-[N-(2-Phenoxy-2-oxoethylcarbamoyl]aminomethyl]-3-
methylbenzoyl)-2,3,4,5-tetrahydro-1H-1-benzazepine (19f). To
a solution of 19a (0.134 g, 0.33 mmol) in CH2Cl2 (25 mL) were
added phenol (0.155 g, 1.6 mmol), 4-(dimethylamino)pyridine (0.04
g, 0.33 mmol), and WSCDI (0.135 mg, 0.66 mmol). The mixture
was stirred at reflux for 48 h and cooled. The mixture was washed
with 1 M KHSO4, saturated NaHCO3 solution and brine, dried over
MgSO4, and concentrated in vacuo. The crude material was purified
by chromatography (eluant EtOAc:pet. ether 80:20) to give a white
solid; yield 0.038 g (33%). 1H NMR δ 1.40-1.60 (1H, m),
1.62-2.08 (6H, m), 2.66-3.08 (3H, m), 4.12 (2H, d, J ) 5.0 Hz),
4.19 (2H, d, J ) 5.6 Hz), 4.95 (1H, d, J ) 13.2 Hz), 5.62 (1H, br

8132 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24
Yea et al.
s), 5.82 (1H, t, J ) 5.0 Hz), 6.57 (1H, d, J ) 7.6 Hz), 6.64-6.96
(4H, m), 6.98-7.14 (3H, m), 7.16-7.26 (2H, m), 7.28-7.42 (2H,
m). HRMS: 472.2229 (calculated) and 472.2119 (observed).
19g, 19i, 19k, and 19l were prepared using the same procedure
as 19f.
BOC-proline-N,N-dimethylthioamide (22). A solution of BOC-
proline-N,N-dimethylamide (21) (0.85 g, 3.51 mmol) and Lawes-
son’s reagent (0.78 g, 1.93 mmol) in toluene (30 mL) was stirred
at reflux for 2 h. The mixture was cooled, diluted with EtOAc,
washed with 0.5 M KHSO4, saturated NaHCO3 solution and brine,
dried over Na2SO4, and reduced in vacuo. The crude material was
purified by chromatography (eluant EtOAc:pet. ether 50:50) to give
a colorless gum; yield 0.63 g (70%).
Proline-N,N-dimethylthioamide (23). A solution of 22 (0.63
g, 2.44 mmol) in 4N HCl/dioxan was stirred for 90 min. The
mixture was reduced in vacuo to give a white solid that was used
without further purification; yield 0.52 g.
1-(2-Methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-1-ylcarbon-
yl)benzylcarbamoyl)-L-proline (20a). To a solution of 20b (0.176
g, 0.39 mmol) in THF (20 mL) and water (5 mL) was added
LiOH·H2O (0.038 g, 0.91 mmol). The mixture was stirred at room
temperature for 4 h, concentrated in vacuo, and the residue diluted
with water and washed with Et2O. The aqueous layer was acidified
to pH 1 by addition of 1 M HCl and extracted with EtOAc (3 times).
The combined organic extracts were washed with brine, dried over
Na2SO4, and concentrated in vacuo to give a white solid; yield 0.119
g (70%). 1H NMR δ 1.35-1.52 (1H, m), 1.75-2.10 (6H, m), 2.05
(3H, s), 2.30-2.40 (1H, m), 2.60-3.00 (3H, m), 3.10-3.35 (2H,
m), 4.20 (2H, d, J ) 5.2 Hz), 4.30-4.36 (1H, m), 4.82-4.90 (1H,
m), 5.24-5.30 (1H, m), 6.54 (1H, d, J ) 7.6 Hz), 6.70-7.02 (6H,
m), 7.13 (1H, d, J ) 6.9 Hz). HRMS: 436.2235 (calculated) and
436.1905 (observed).
27 was prepared using the same procedure as 21f.
BOC-4,4-difluoroproline-N,N-dimethylamide (28). To a solu-
tion of 27 (0.30 g, 1.17 mmol) in CH2Cl2 at 0 °C was added DAST
(155 µL, 0.189 g, 1.17 mmol). The mixture was stirred for 4 h. A
further amount of DAST (78 µL, 0.095 g, 0.59 mmol) was added
and allowed to warm to room temperature. The mixture was stirred
for 18 h, and a trace of water was added. The mixture was reduced
in vacuo and adsorbed onto silica gel. Chromatography (eluant
EtOAc:pet. ether 50:50) afforded a yellow gum; yield 0.11 g (35%).
4-Difluoroproline-N,N-dimethylamide hydrochloride (29). 28
(0.11 g, 0.40 mmol) was reacted with 4N HCI/dioxan (20 mL)
following the method of 26 and used immediately in the next step
without purification.
4,4-Dimethoxy-1-(2-methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-
1-ylcarbonyl)benzyl-carbamoyl)-L-proline methyl ester (36h). To
a solution of 8e (0.281 g, 0.90 mmol) in DMF (15 mL) at 0 °C
was added CDI (0.176 g, 0.90 mmol). The mixture was allowed to
warm to room temperature and stirred for 50 min. 4,4-Dimethoxy-
L-proline methyl ester (35h)17 (0.180 g, 0.90 mmol) was added,
and the mixture was stirred at room temperature for a further 16 h.
The solvent was removed in vacuo, and the crude material was
purified by chromatography (eluant MeOH:CH2Cl2 2:98) to give a
white solid; yield 0.260 g (57%).
4,4-Dimethoxy-1-(2-methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-
1-ylcarbonyl)benzyl-carbamoyl)-L-proline-N,N-dimethyla-
mide (21h). To a solution of the methyl ester 36h (0.230 g, 0.45
mmol) in water (5 mL) and THF (15 mL) was added LiOH·H2O
(56 mg, 1.4 mmol) and stirred for 18 h. Water (20 mL) was added
and the mixture washed with Et2O. The aqueous layer was acidified
with 2 M citric acid solution and extracted with EtOAc. The organic
phase was washed with water, brine, dried over MgSO4, and reduced
to give the carboxylic acid as a white solid; yield 219 mg (98%).
To a solution of the carboxylic acid (0.209 g, 0.42 mmol) in CH2Cl2
(10 mL) were added WSCDI (0.101 g, 0.50 mmol) and hydroxy-
benzotriazole (0.057 g, 0.46 mmol). The mixture was stirred and
to it was added Et3N (150 µL, 0.105mmol). After 10 min,
Me2N·HCl (0.086 g, 0.105 mmol) was added. The mixture was
stirred for 18 h and adsorbed onto silica gel. Chromatography
(eluant MeOH:CH2Cl2 3:97) gave a glassy solid; yield 0.11 g (50%).
1H NMR δ 1.38-1.58 (1H, m), 1.82-2.18 (3H, m), 2.21 (3H, s),
2.36 (1H, dd, J ) 7.9, 12.1 Hz), 2.60-3.26 (4H, m), 2.88 (3H, s),
3.07 (3H, s), 3.17 (3H, s), 3.21 (3H, s), 3.45 (1H, d, J ) 9.6 Hz),
3.54 (1H, d, J ) 9.6 Hz), 4.02-4.18 (1H, m), 4.28-4.44 (1H, m),
4.46-4.64 (1H, m), 4.76 (1H, t, J ) 8.2 Hz), 4.97 (1H, d, J )
13.6 Hz), 6.60 (1H, d, J ) 7.7 Hz), 6.76-6.98 (3H, m), 7.00-7.10
(2H, m), 7.18 (1H, d, J ) 6.7 Hz). HRMS: 523.2923 (calculated)
and 523.2510 (observed).
1-(4-Cyano-3-methylbenzoyl)-5-oxo-2,3,4,5-tetrahydro-1H-1-
benzazepine (37a). To a stirred suspension of 7e (4.00 g, 13.78
mmol) in tert-butanol (120 mL) and water (300 mL) were added
KMnO4 (15.25 g, 96.50 mmol) and Mg(NO3)2 ·6H2O (24.68 g,
96.25 mmol). The mixture was stirred for 72 h and diluted with 3
N HCl(aq) (100 mL). Sodium bisulfite was added portionwise until
a clear colorless solution was obtained. The mixture was extracted
with CH2Cl2. The organic extract was washed with saturated brine,
dried over MgSO4 and reduced. The residue was purified by
chromatography (eluant EtOAc:pet. ether 40:60) to give a pale-
orange gum; yield 1.06 g (25%).
1-(2-Methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-1-ylcarbon-
yl)benzylcarbamoyl)-L-proline-N,N-dimethylamide (20e). 14 To
a solution of 8e (0.10 g, 0.302 mmol) in DMF (10 mL), under a
nitrogen atmosphere, were added N,N-diisopropylethylamine (0.043
g, 0.332 mmol) and CDI (0.074 g, 0.453 mmol). The mixture was
stirred at room temperature for 40 min. A solution of proline-N,N-
dimethylamide (0.107 g, 0.76 mmol) in DMF (1 mL) was added.
The mixture was stirred at room temperature for a further 16 h.
The solvent was removed in vacuo, and the crude material was
purified by chromatography (eluant MeOH:CH2Cl2 5:95) to give a
1
white solid; yield 0.115 g (82%). H NMR δ 1.34-1.55 (1H, m),
1.70-2.20 (7H, m), 2.10 (3H, s), 2.60-3.10 (3H, m), 2.89 (3H,
s), 3.04 (3H, s), 3.20-3.34 (1H, m), 3.40-3.58 (1H, m), 4.00-4.18
(1H, m), 4.28-4.45 (1H, m), 4.58-4.84 (2H, m), 4.95 (1H, d, J )
13.1 Hz), 6.56 (1H, d, J ) 7.4 Hz), 6.73-6.89 (2H, m), 6.90-7.08
(3H, m), 7.15 (1H, d, J ) 7.4 Hz). HRMS: 463.2708 (calculated)
and 463.2450 (observed). Anal. (C27H34N4O3 + 0.25 H2O) C, H,
N.
19d, 19e, 19j, 20b, 20c, 20d, 20f, 20g, 20h, 21a, 21b, 21d, 21e,
and 21g were prepared using the same procedure as 20e.
(4R)-4-Ethoxy-1-(2-methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-
1-ylcarbonyl)benzyl-carbamoyl)-L-proline-N,N-dimethyla-
mide (21c). To a solution of 21a (0.080 g, 0.17 mmol) in DMF
(10 mL) at -5 °C was added NaH (60%, 8 mg, 2.0 mmol). The
mixture was stirred for 20 min, iodoethane (0.155 g, 10 mmol)
was added and was allowed to warm to room temperature and
stirred for 3 days. The mixture was reduced in vacuo. The residue
was purified by chromatography (eluant CHCl3:MeOH:acetic acid
98:1:1) and freeze-dried to give a white solid; yield 0.034 g (39%).
1H NMR δ 1.16 (3H, t, J ) 7.2 Hz), 1.35-1.55 (1H, m), 1.85-2.13
(5H, m), 2.14 (3H, s), 2.66-3.16 (3H, m), 2.91 (3H, s), 3.11 (3H,
s), 3.20-3.30 (1H, m), 3.44 (2H, q, J ) 7.4 Hz), 3.65-3.68 (1H,
t, J ) 6.7 Hz), 4.05-4.20 (1H, m), 4.25-4.39 (3H, m), 4.81-5.10
(2H, m), 6.61 (1H, d, J ) 7.7 Hz), 6.80-7.00 (3H, m), 7.07-7.17
(2H, m), 7.20-7.24 (1H, m). HRMS: 507.2973 (calculated) and
507.2576 (observed).
1-(2-Methyl-4-(2,3,4,5-tetrahydro-1-benzazepin-1-ylcarbon-
yl)benzyl-carbamoyl)-4-oxo-L-proline-N,N-dimethylamide (21f).
To a solution of 21a (0.080 g, 0.20 mmol) in CH2Cl2 (20 mL) at
0 °C was added Dess-Martin periodinane (0.176 g, 0.50 mmol).
The mixture was allowed to warm to room temperature and stirred
for 18 h. The mixture was adsorbed onto silica gel. Chromatography
(eluant CHCl3:MeOH:acetic acid 97:2:1) afforded a white solid;
1
yield 0.040 g (50%). H NMR δ 1.38-1.70 (1H, m), 1.86-2.18
(3H, m), 2.15 (3H, s), 2.45 (1H, d, J ) 17.6 Hz), 2.66-3.10 (4H,
m), 2.93 (3H, s), 3.20 (3H, s), 3.65 (1H, d, J ) 15.6 Hz), 3.95
(1H, d, J ) 15.6 Hz), 4.14-4.30 (1H, m), 4.32-4.48 (2H, m),
4.98 (1H, d, J ) 13.2 Hz), 5.33 (1H, d, J ) 7.7 Hz), 6.61 (1H, d,
J ) 7.9 Hz), 6.78-6.96 (3H, m), 7.00-7.28 (3H, m). HRMS:
477.2499 (calculated) and 477.2122 (observed).
(()-1-(4-Aminomethyl-3-methylbenzoyl)-5-hydroxy-2,3,4,5-

Benzylureas as Vasopressin V2 Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24 8133
tetrahydro-1H-1-benzazepine (38a). To a solution of 37a (0.450
g, 1.48 mmol) and CoCl2 ·6H2O (706 mg, 2.97 mmol) in MeOH
(40 mL) was added NaBH4 (0.573 g, 15.15 mmol) portionwise.
The mixture was stirred for 90 min, acidified with 1 N HCl(aq),
and reduced in vacuo. The residue was taken up in water, washed
with Et2O, basified with saturated NaHCO3 solution, and extracted
with CHCl3. The organic extract was washed with saturated brine,
dried over MgSO4, and reduced to give a white solid; yield 0.142
g (31%).
temperature, stirred for 18 h, and reduced in vacuo. The residue
was taken up in CHCl3 and washed with 0.3 M KHSO4(aq),
NaHCO3(aq), and brine. The organic phase was dried over MgSO4
and reduced to afford a brown solid; yield 3.50 g (89%).
37b, 37e, 37f, and 37g were prepared using the same procedure
as 37d.
CRE-Luciferase Reporter Gene Assay. Cell Transfection. The
human V2 receptor was cloned from kidney total RNA using the
RT-PCR. The human receptor was subcloned into pcDNA5/FRT
(InVitrogen). The cAMP response element and the synthetic Firefly
Luciferase gene Luc2 were provided in the pGL4.29-vector
(Promega). Plasmids were amplified in Novablue Escherichia coli
(Merck Biosciences Ltd.) and plasmid DNA was purified using
UltraMobius Endotoxin free plasmid purification kits (Merck
Biosciences Ltd.).
Cell Culture. Human embryonic kidney (HEK293) cells (EC-
CAC, UK) were maintained in Dulbecco’s modified Eagles medium
without phenol red, supplemented with 50000 U penicillin, 50 mg
streptomycin, and 10% fetal calf serum (InVitrogen). Cells were
grown to 70-80% confluency and transiently transfected with the
V2 receptor and pGL4 CRE-Luc reporter constructs using Genejuice
(Merck Biosciences Ltd.). These transfected cells were incubated
at 37 °C with 5% CO2 for 24 h before plating out at 10000 cells/
well into 96 well white Optiplates (Perkin-Elmer). The plates were
incubated for a further 24 h at 37 °C with 5% CO2 before testing
for functional agonism.
Functional Assays. Compounds were prepared in DMSO to a
concentration of 10 mM and 0.5 log serially diluted in duplicate to
give the appropriate concentrations when added to the cells. Final
DMSO concentrations were 0.1%. Levels of luciferase synthesis
were measured following 5 h incubation at 37 °C using SteadyGlo
luciferase reagent (Promega). The data were expressed as percentage
of the 100 nM Arginine-Vasopressin (AVP) (Sigma no. V-9879)
controls and then plotted against the log of the concentration. The
data were fitted to a 4-parameter logistic equation and EC50 values
determined. Data are reported as geometric means ( standard
deviation.
38b, 38c, 38d, 38e, 38f, 38g were prepared using the same
procedure as 38a.
1-(4-(5-Hydroxy-2,3,4,5-tetrahydro-1-benzazepin-1-ylcarbo-
nyl)-2-methylbenzylcarbamoyl)-L-proline-N,N-dimethylamide
(22a). To a solution of 38a (0.142 g, 0.46 mmol) in DMF (10 mL),
under a nitrogen atmosphere, was added CDI (0.083 g, 0.52 mmol).
The mixture was stirred at room temperature for 1 h. A solution of
L-proline-N,N-dimethylamide (0.107 g, 0.76 mmol) in DMF (1 mL)
was added. The mixture was stirred at room temperature for a
further 4 h, and the solvent was removed in vacuo. The residue
was taken up in EtOAc, washed with 0.3 M KHSO4, saturated brine,
dried over MgSO4, and reduced in vacuo. Chromatography (eluant
CHCl3:MeOH 92:8) gave a white solid; yield 0.045 g (20%). HPLC
data showed evidence of diastereoisomers, isomer A 15.86 min
(35.2%) and isomer B (61.3%). NMR data showed evidence of
diastereoisomers. 1H NMR δ 1.50-2.25 (11H, m), 2.64-2.90 (1H,
m), 2.83 and 2.85 (total 3H, each s), 3.01 (3H, s), 3.27-3.50 (1H,
m), 3.54-3.66 (1H, m), 3.98-4.30 (3H, m), 4.60-5.05 (3H, m),
6.48-6.60 and 6.70-7.20 (total 7H, each m), 7.60 (1H, d, J ) 7.4
Hz). HRMS: 479.2658 (calculated) and 479.2288 (observed).
22b, 22c, 22d, 22e, 22f, and 22g were prepared using the same
procedure as 22a.
1-(4-Cyano-3-methylbenzoyl)-5,5-difluoro-2,3,4,5-tetrahydro-
1H-1-benzazepine (37c). To a solution of 37a (0.60 g, 1.97 mmol)
in CH2Cl2 (20 mL) at 0 °C was added DAST (522 µL, 3.95 mmol).
The mixture was allowed to warm to room temperature and stirred
for 18 h. Additional amounts of DAST were successively added to
the mixture and heated at reflux until TLC indicated a significant
amount of product had formed. The mixture was cooled to room
temperature and poured into water (5 mL). Saturated NaHCO3(aq)
was added slowly until effervescence ceased. The mixture was
extracted with CH2Cl2. The organic phase was washed with water,
dried over MgSO4, and reduced in vacuo. Chromatography (eluant
EtOAc:pet. ether 30:70) afforded a white solid; yield 164 mg,
(25%).
Acknowledgment. We thank Andre´ Tartar (Faculte´ des
Sciences Pharmaceutiques et Biologiques de Lille) for helpful
and insightful discussions and Stephen Pethen (Vantia Ltd) for
provision of HRMS data.
Supporting Information Available: Routine experimental and
spectroscopic data, elemental analyses on selected compounds, and
HPLC traces on target compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
(()-4-Methyl-1-tetralone oxime (40b). To a solution of (()-
1-methyl-4-tetralone (39b) (2.0 g, 0.012 mol) and NH2OH·HCl
(1.388 g, 0.020 mol) in EtOH was added a solution of NaOH (2.4
g, 0.060 mol) in water (2.0 mL). The mixture was stirred at reflux
for 2 h, cooled, poured into hydrochloric acid, and stirred for 30
min. The precipitate was filtered and dried to give a yellow solid;
yield 1.90 g (90%).
References
(1) Altura, B. M.; Altura, B. T. Vascular smooth muscle and neurohy-
40d was prepared using the same procedure as 40b.
pophyseal hormones. Fed. Proc. 1977, 36, 1853–1860.
(()-5-Methyl-2,3,4,5-tetrahydro-1H-1-benzazepine (41b). To
a solution of DIBAL-H (4.7 mL, 0.026 mol) in CH2Cl2 (20 mL)
under N2 was added slowly a solution of 40b (1.9 g, 0.0108 mol)
in CH2Cl2 (10 mL) so as to achieve a gentle reflux. The mixture
was stirred at room temperature for 18 h. KF (11.2 g, 0.176 mol)
was added portionwise (CAUTION) and the mixture stirred for 10
min. Water was added dropwise (CAUTION), and the mixture
stirred for 20 min. The mixture was poured into CH2Cl2 (200 mL),
stirred for 30 min, filtered, and reduced in vacuo. The residue was
purified by chromatography (eluant EtOAc:pet. ether. 10:90) to
afford a pale-yellow oil; yield 1.20 g (86%).
(2) Howl, J.; Ismail, T.; Strain, A. J.; Kirk, C. J.; Anderson, D.; Wheatley,
M. Characterization of the human liver vasopressin receptor. Profound
differences between human and rat vasopressin-receptor-mediated
responses suggest only a minor role for vasopressin in regulating
human hepatic function. Biochem. J. 1991, 276 (1), 189–195.
(3) Jard, S.; Lombard, C.; Marie, J.; Devilliers, G. Vasopressin receptors
from cultured mesangial cells resemble V1a type. Am. J. Physiol. 1987,
253 (1 Pt 2), F41-F49.
(4) Penit, J.; Faure, M.; Jard, S. Vasopressin and angiotensin II receptors
in rat aorta smooth muscle cells in culture. Am. J. Physiol. 1983, 244
(1), E72-E82.
41d was prepared using the same procedure as 41b.
(5) Thibonnier, M.; Roberts, J. M. Characterization of human platelet
5-(4-Cyano-3-methylbenzoyl)-2,3,4,5-tetrahydro-1,5-benzox-
azepine (37d). To a solution of 4e (2.16 g, 13.4mmol) in toluene
(20 mL) was added SOCl2 (2.93 mL, 40.2 mmol). The mixture
was heated at reflux for 90 min and reduced in vacuo. The residue
was taken up in CH2Cl2 (5.0 mL) and added to a solution of 41d
(2.0 g, 13.4 mmol) and Et3N (3.74 mL, 26.8 mmol) in CH2Cl2 (2.0
mL) at 0 °C. The mixture was allowed to warm to room
vasopressin receptors. J. Clin. InVest. 1985, 76, 1857–1864.
(6) Laski, M. E.; Pressley, T. A. Aquaporin mediated water flux as a target
for diuretic development. Semin. Nephrol. 1999, 19 (6), 533–550.
(7) Dorsa, D. M.; Majumdar, L. A.; Petracca, F. M.; Baskin, D. G.;
Cornett, L. E. Characterization and localization of 3H-arginine8-
vasopressin binding to rat kidney and brain tissue. Peptides 1983, 4,
699–706.

8134 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 24
Yea et al.
(8) Andersson, K. E.; Bengtsson, B.; Paulsen, O. Desamino-8-D-arginine
vasopressin (DDAVP): pharmacology and clinical use. Drugs Today
(Barc) 1988, 24 (7), 509–528.
(9) Skoog, S. J.; Stokes, A.; Turner, K. L. Oral desmopressin: a
randomized double-blind placebo controlled study of effectiveness in
children with primary nocturnal enuresis. J. Urol. 1997, 158 (3 Pt 2),
1035–1040.
(10) (a) Kondo, K.; Ogawa, H.; Shinohara, T.; Kurimura, M.; Tanada, Y.;
Kan, K.; Yamashita, H.; Nakamura, S.; Hirano, T.; Yamamura, Y.;
Mori, T.; Tominaga, M.; Itai, A. Novel Design of Nonpeptide AVP
V2 Receptor Agonists: Structural Requirements for an Agonist Having
1-(4-Aminobenzoyl)-2,3,4,5-tetrahydro-1H-1-benzazepine as a Tem-
plate. J. Med. Chem. 2000, 43, 4388–4397. (b) Kondo, K.; Kan, K.;
Tananda, Y.; Bando, M.; Shinohara, T.; Kurimura, M.; Ogawa, H.;
Nakamura, S.; Hirano, T.; Yamamura, Y.; Kido, M.; Mori, T.;
Tominaga, M. Characterization of orally active nonpeptide vasopressin
V2 receptor agonist. Synthesis and biological evaluation of both the
(5R)- and (5S)-enantioisomers of 2-[1-(2-chloro-4-pyrrolidin-1-yl-
benzoyl)-2,3,4,5-tetrahydro-1H-1-benzazepin-5-yl]-N-isopropylaceta-
mide. J. Med. Chem. 2002, 45, 3805–3808.
(11) Caggiano, T. J. WAY-VNA-932. Drugs Future 2002, 27, 248–253.
(12) Pitt, G. R. Non-Peptide Vasopressin V2 Receptor Agonists. Abstracts
of Papers, RSC-SCI Symposium on Medicinal Chemistry, September
9-12, 2001.
(13) Pitt, G. R. W.; Batt, A. R.; Haigh, R. M.; Penson, A. M.; Robson,
P. A.; Rooker, D. P.; Tartar, A. L.; Trim, J. E.; Yea, C. M.; Roe,
M. B. Non-peptide oxytocin agonists. Bioorg. Med. Chem. Lett. 2004,
14, 4585–4589.
(18) Cho, H.; Murakami, K.; Nakanishi, H.; Isoshima, H.; Hayakawa, K.;
Uchida, I. Regioselective synthesis of several heterocyclic fused
azepines using diisobutylaluminum hydride. Heterocycles 1998, 48,
919–927.
(19) Albright, J. D.; Reich, M. F.; Delos Santos, E. G.; Dusza, J. P.; Sum,
F.-W.; Venkatesan, A. M.; Coupet, J.; Chan, P. S.; Ru, X.; Mazan-
darani, H.; Bailey, T. 5-Fluoro-2-methyl-N-[4-(5H-pyrrolo[2,1-c]-
[1,4]benzodiazepin-10(11H)-ylcarbonyl)-3-chlorophenyl]benzamide
(VPA-985): an orally active arginine vasopressin antagonist with
selectivity for V2 receptors. J. Med. Chem. 1998, 41, 2442–2444.
(20) Ohta, T.; Miyake, S.; Shudo, K. A trifluoromethanesulfonic acid-
catalyzed reaction of arylhydrazines with benzene. Tetrahedron Lett.
1985, 26, 5811–5814.
(21) Chimenti, F.; Vomero, S.; Giuliano, R.; Artico, M. Research on
substances having antiblastic activity. Note LIX. Anthramycin and
related compounds. Part VIII: Synthesis of 4H-pyrrolo[1,2-
a][1,5]benzodiazepine derivatives. Farmaco, Ed. Sci. 1977, 32, 339–
347.
(22) (a) Schmale, H.; Richter, D. Single base deletion in the vasopressin
gene is the cause of diabetes insipidus in Brattleboro rats. Nature 1984,
308, 705–709. (b) Valtin, H.; Schroeder, H. A. Familial hypothalamic
diabetes insipidus in rats (Brattleboro strain). 1964. J. Am. Soc.
Nephrol. 1997, 8, 1333–1341.
(23) Trybulski, E. J. Vasopressin Receptor Modulators: From Non-peptide
Antagonists to Agonists. In Annual Reports in Medicinal Chemistry,
Volume 36; Doherty, A. M., Ed.; Academic Press: San Diego, 2001;
pp 159-168.
(24) Sum, F.-W.; Dusza, J.; Santos, E. D.; Grosu, G.; Reich, M.; Du, X.;
Albright, J. D.; Chan, P.; Coupet, J.; Ru, X.; Mazandarani, H.;
Saunders, T. Structure-Activity Study of Novel Tricyclic Benzazepine
Arginine Vasopressin Antagonists. Bioorg. Med. Chem. Lett. 2003,
13, 2195–2198.
(25) Kondo, K.; Ogawa, H.; Yamashita, H.; Miyamoto, H.; Tanaka, M.;
Nakaya, K.; Kitano, K.; Yamamura, Y.; Nakamura, S.; Onogawa, T.;
Mori, T.; Tominaga, M. 7-Chloro-5-hydroxy-1-[2-methyl-4-(2-meth-
ylbenzoyl-amino)benzoyl]-2,3,4,5 -tetrahydro-1H-1-benzazepine (OPC-
41061): A potent, orally active nonpeptide arginine vasopressin V2
receptor antagonist. Bioorg. Med. Chem. 1999, 7, 1743–1757.
(26) Chio, C. L.; Lajiness, M. E.; Huff, R. M. Activation of heterologously
expressed D3 dopamine receptors: comparison with D2 dopamine
receptors. Mol. Pharmacol. 1994, 45, 51–60.
(27) Kenakin, T. A Pharmacology Primer: Theory, Application and
Methods; Elsevier Academic Press: San Diego, 2004; pp 73-92.
(14) Ashworth, D. M.; Pitt, G. R. W.; Hudson, P.; Yea, C. M.; Franklin,
R. J.; Semple, G. Preparation of fused azepine derivatives and their
use as antidiuretic agents. PCT Int. Appl. WO 000626 A1, 2002.
(15) Hudson, P.; Batt, A. R.; Laurent, C. M. S.; Baxter, A. J.; Roe, M. B.;
Robson, P. A. Preparation of substituted pyrazolo-fused benzodiazepine
analogs as oxytocin agonists and vasopressin antagonists. Eur. Pat.
Appl. EP 1449844 A1, 2004.
(16) Chiba, J.; Takayama, G.; Takashi, T.; Yokoyama, M.; Nakayama, A.;
Baldwin, J. J.; McDonald, E.; Moriarty, K. J.; Sarko, C. R.; Saionz,
K. W.; Swanson, R.; Hussain, Z.; Wong, A.; Machinaga, N. Synthesis,
biological evaluation, and pharmacokinetic study of prolyl-1-piper-
azinylacetic acid and prolyl-4-piperidinylacetic acid derivatives as
VLA-4 antagonists. Bioorg. Med. Chem. 2006, 14, 2725–2746.
(17) Smith, E. M.; Swiss, G. F.; Neustadt, B. R.; Gold, E. H.; Sommer,
J. A.; Brown, A. D.; Chiu, P. J. S.; Moran, R.; Sybertz, E. J.; Baum,
T. Synthesis and pharmacological activity of angiotensin-converting
enzyme inhibitors: N-(mercaptoacyl)-4-substituted-(S)-prolines. J. Med.
Chem. 1988, 31, 875–885.
JM8008162