Subtlety of the Structure-Affinity and Structure-Efficacy Relationships around a Nonpeptide Oxytocin Receptor Agonist

Article abstract of DOI:10.1021/jm901084f

Very few nonpeptide oxytocin agonists have currently been reported, and none of them seem suitable for the in vivo investigation of the oxytocin mediated functions. In an attempt to rationalize the design of better tools, we have systematically studied the structural determinants of the affinity and efficacy of representative ligands of the V_1a, V₂, and OT receptor subtypes. Despite apparently obvious similarity between the ligand structures on one hand, and between the receptor subtypes on the other hand, the binding affinity and the functional activity profiles of truncated and hybrid ligands highlight the subtlety of ligand-receptor interactions for obtaining nonpeptide OT receptor agonists.

Full text of DOI:10.1021/jm901084f

1546 J. Med. Chem. 2010, 53, 1546–1562
DOI: 10.1021/jm901084f
Subtlety of the Structure-Affinity and Structure-Efficacy Relationships around a Nonpeptide
Oxytocin Receptor Agonist
Marie-Celine Frantz,†,§ Jordi Rodrigo,†, Laure Boudier,‡ Thierry Durroux,‡ Bernard Mouillac,‡ and Marcel Hibert*,†
ꢀ
†
ꢀ
ꢀ
Laboratoire d’Innovation Therapeutique, UMR 7200 CNRS/Universite de Strasbourg, Faculte de Pharmacie, 74 route du Rhin, BP60024, 67401
Illkirch, France and Institut de Genomique Fonctionnelle UMR CNRS 5203/INSERM U661/Universite Montpellier I & II, Dept Pharmacologie
ꢀ
‡
ꢀ
ꢀ
Moleculaire, 141 rue de la Cardonille, 34094 Montpellier, France. § Current address: University of Pittsburgh, Department of Chemistry,
ꢀ
Pittsburgh, Pennsylvania 15260. Current address: Universite Paris-Sud, CNRS, BioCIS-UMR 8076, Laboratoire de Chimie Therapeutique,
ꢀ
Faculte de Pharmacie, 5 rue J.-B. Clement, 92296 Chatenay-Malabry, France
ꢀ
ꢀ
ꢀ
^
Received July 22, 2009
Very few nonpeptide oxytocin agonists have currently been reported, and none of them seem suitable for
the in vivo investigation of the oxytocin mediated functions. In an attempt to rationalize the design of
better tools, we have systematically studied the structural determinants of the affinity and efficacy of
representative ligands of the V1a, V2, and OT receptor subtypes. Despite apparently obvious similarity
between the ligand structures on one hand, and between the receptor subtypes on the other hand, the
binding affinity and the functional activity profiles of truncated and hybrid ligands highlight the subtlety
of ligand-receptor interactions for obtaining nonpeptide OT receptor agonists.
Introduction
there was an urgent need for centrally active, potent, specific,
and bioavailable agonists and antagonists of the four vaso-
pressin and oxytocin receptor subtypes, namely the V1a, V1b,
V2, and OT receptors. Such antagonists became recently
available.21 However, very few nonpeptide agonists have been
reported.22 For OT receptor agonists, the most interesting to
date (compound 1, Figure 1) is the optimized analogue of a hit
discovered by systematic screening.23,24 However, its structure
does not seem optimal for CNS in vivo studies because its
molecular weight (MW = 600.8 Da) and its number of
H-bonds (14 H-bonds possible) remain rather high.25 In addi-
tion, some of its hydrophobicity parameters do not fit the
requirements for a significant blood-brain barrier penetration
Numerous animal studies have provided evidence support-
ing a role for oxytocin (OTa) and vasopressin (AVP) in
many central nervous system (CNS) functions including
mating, affiliation, adult pairing and faithfulness, anxiety,
and prosocial behaviors.1-6 In humans, OT has been reported
to improve trust7-9 and to decrease fear associated to social
phobia10-12 and symptoms related to autism.13-20 However,
the only molecular probe that is currently available to study
these functions, both in animals (usually after intracerebro-
ventricular administration) and humans (after intranasal
administration), is oxytocin itself or closely related peptide
analogues that have a poor bioavailability inherent with most
neuropeptides (e.g., poor intestinal and blood-brain barrier
penetration, short half-life, rapid clearance). To characterize
without ambiguity the different central mechanisms of action
and physiopathological functions of vasopressin and oxytocin,
2
˚
(ideal clogP = 2 but high PSA = 121 A ). Furthermore, its
efficacy and selectivity profile had to be further explored. We
have used this molecule and closely related analogues acting at
other vasopressin receptor subtypes to systematically explore
the structural determinants of affinity, efficacy, and selectivity
at V1a, V2, and OT receptor subtypes. Our strategy consisted in
two steps. First, we tried to dissect compound 1 to identify the
minimum molecular components required for affinity and to
evaluate their positive (agonism) or negative (antagonism)
contribution to efficacy on the three considered receptor sub-
types. To this end, truncated analogues were prepared and their
affinity and efficacy were measured in parallel on the target
receptors. Second, we systematically prepared hybrid mole-
cules combining the apparently homologous molecular frag-
ments of four representative vasopressin/oxytocin ligands
(Table 1): compound 1, a claimed OT receptor agonist;24
compound 2, a reported nonselective V1a/V2/OT receptor
antagonist;26 compound 3, WAY-VNA-932, a V2 receptor
agonist;27 and compound 4, L-371,257, the structurally closest
and selective OT receptor antagonist.28 The affinity and effi-
cacy of these chimeric molecules for the V1a, V2, and OT
receptor subtypes were studied to probe the contribution to
*To whom correspondence should be addressed. Phone:
þ33390244232. Fax: þ33390244310. E-mail: mhibert@pharma.
u-strasbg.fr.
a Abbreviations: AVP, arginine vasopressin; BSA, bovine serum
albumin; cAMP, cyclic adenosine 50-monophosphate; CDI, carbonyl-
diimidazole; CHO, Chinese hamster ovary; CNS, central nervous
system; CRE, cAMP response element; DCM, dichloromethane; DIEA,
diisopropylethylamine; DMAP, N,N-dimethyl-4-aminopyridine; D-
MEM, Dulbecco’s modified Eagle’s medium; DMF, dimethylforma-
mide; DMSO, dimethylsulfoxide; E, extracellular loop; EDCI, 1-(3-
dimethylaminopropyl)-3-ethyl-carbodiimide; EDTA, ethylenediamine-
tetraacetic acid; GPCR, G protein-coupled receptor; HOBt, N-hydro-
xybenzotriazole; IP1, myo-inositol 1-phosphate; μw, microwaves;
NFAT, nuclear factor of activated T-cells; NMP, N-methyl-2-pyrroli-
done; OT, oxytocin; OT-R, oxytocin receptor; Ro-20-1724, 4-(3-butoxy-
4-methoxyphenyl)methyl-2-imidazolidone; rt, room temperature; SAR,
structure-activity relationships; SEM, standard error of the mean;
TFA, trifluoroacetic acid; THF, tetrahydrofuran; TM, transmembrane
domain; Tris, 2-amino-2-hydroxymethyl-1,3-propanediol; V1a-R, vaso-
pressin V1a receptor subtype; V1b-R, vasopressin V1b receptor subtype;
V2-R, vasopressin V2 receptor subtype.
r
pubs.acs.org/jmc
Published on Web 01/27/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1547
affinity and efficacy of the different corresponding fragments.
The unexpected results shed an interesting and disturbing light
on the mode of action of very similar ligands on very similar
receptor subtypes.
isopropyl alcohol for 14 days in the presence of acetic acid was
difficult to reproduce. However, a 89% yield could be achieved
by heating the precursor 5 in a sealed tube at 160 °C for 48 h. As
described for compound 1, the urea truncated analogues 12a and
12b could be isolated in high yields after coupling with the
common amine intermediate 10 via preactivation with carbo-
nyldiimidazole (CDI). Alternatively, the acetylated truncated
analogue 12c was obtained by reacting 10 with acetic anhydride.
A similar synthetic route was followed for the preparation
of compounds 23a-d, according to the general Scheme 2. The
polycyclic amines 13, 14, and 15 were prepared according to
literature protocols.30-32 In addition, with regard to the
synthesis of the analogue 23c, the key acid intermediate 20
was prepared starting from the acyl chloride 16 via a Sand-
meyer reaction to introduce the nitrile group.
Chemistry
Reference compounds 2, 3, and 4 were prepared according to
literature protocols.26-28 Compound 1 and analogues 12a-c
were prepared according to Scheme 1. The general procedure is
similar to the synthesis previously described by Pitt et al.24,29 with
some optimization, specifically for the cyclization step leading to
the lactam 6. The original preparation consisting of reflux in
As described in Scheme 3, the hybrid compound 27 was
prepared in a similar manner as the reference compound 2.
Finally, both hybrids 31a and 31b were obtained following a
general route described by Kondo et al.33
Biology
The affinities of the different ligands for the human vaso-
pressin receptors subtypes and the oxytocin receptor were
determined on CHO cell membranes by competition experi-
ments against [3H]AVP, as described previously.34-36
Figure 1. Schematic representation of the main results of the
truncated analogues of compound 1.
Table 1. Structures and Biological Activities of the Reference Compoundsg
a The inhibition constants (Ki in nM) for human V1a, V2, and OT receptors were determined on CHO cell membranes by competition binding assays
(displacement of radioactive [3H]AVP). Results are expressed as mean ( SEM of at least three separate experiments performed in triplicate. When
Ki >1000 nM, these are results of at least two separate experiments performed in triplicate. b The activation constants (Kact in nM) were measured in
CHO cell lines expressing either vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and OT-R, IP1 accumulation) or cAMP dynamic 2 (V2-R,
cAMP accumulation) assays. These values are the mean ( SEM of at least three separate experiments performed in triplicate. The maximal stimulations
(Emax) are expressed as percentages of the endogenous agonist maximal stimulation (n = 3). In case of a weak stimulation, results are expressed as
percentages of the endogenous agonist maximal stimulation at [ligand] = 1 μM (n = 2). Results are not significant (ns) when there is <10% response at
[ligand] = 1 μM (n = 2). c The inactivation constants (Kinact in nM) were measured in CHO cell lines expressing either vasopressin V1a or V2 or oxytocin
receptor by IP-One (V1a-R and OT-R, inhibition of agonist-induced IP1 accumulation) or cAMP dynamic 2 (V2-R, inhibition of agonist-induced cAMP
accumulation) assays. These values are the mean ( SEM of at least three separate experiments performed in triplicate. In case of a weak inhibition,
results are expressed as percentages of endogenous agonist-response inhibition at [ligand] = 1 μM (n = 2). Results are not significant (ns) when there is
<10% response inhibition at [ligand] = 1 μM (n = 2). d Not determined. e n = 1. f n = 2. g Putatively homologous binding components are displayed
with similar colors. 1 is a claimed OT receptor agonist;23,24 2 is a nonselective V1a/V2/OT receptor antagonist;26 3, WAY-VNA-932, is a V2 receptor
agonist;27 4, L-371,257, is a selective OT receptor antagonist.28

1548 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
Scheme 1. Synthesis of Compound 1 and Truncated Analoguesa
a Reagents and conditions: (i) 1:9 AcOH/iPrOH, sealed flask, 160 °C, 48 h, 89%. (ii) LiAlH4, THF, reflux, 44 h, 86%. (iii) (a) 8, (COCl)2, cat. DMF,
DCM, 2 h; (b) 7, Et3N, DCM, 2 d, 75%. (iv) NaBH4, CoCl2 6H2O, MeOH, 1 h, quant. (v) (a) CDI, DIEA, DMF, 1 h; (b) 11 or pyrrolidine or
3
Me2NH HCl, DIEA, DMF, overnight, 58-82%. (vi) Ac2O, Et3N, DCM, 3 h, 38%.
3
Scheme 2. Synthesis of the Molecular Hybrids Conserving the Eastern Part of Compound 1a
a Reagents and conditions: (i) MeOH, 2 h, quant. (ii) SnCl2 2H2O, EtOH, reflux, 1 h, quant. (iii) (a) NaNO2, conc HCl/H2O, -15 °C, 15 min;
3
(b) CuCN, KCN, H2O, -10 °C-rt, 2 h, 61%. (iv) LiOH H2O, MeOH/THF, 1 h, 83%. (v) For 21a: (a) 8, (COCl)2, cat. DMF, DCM, 1 h; (b) 13, pyridine,
3
cat. DMAP, DCM, overnight, 76%; for 21b: (a) 8, (COCl)2, cat. DMF, DCM, 1 h; (b) 14, pyridine, DCM, 1 h, 65%; for 21c: (a) 20, (COCl)2, cat. DMF,
DCM, 0 °C, 1 h; (b) 14, pyridine, cat. DMAP, DCM, μw 80 °C, 6 min, 49%; for 21d: 8, 15, HOBt H2O, EDCI HCl, DIEA, DMF, 14 h, 65%.
3
3
(vi) NaBH4, CoCl2 6H2O, MeOH, 1-3 h, 12-89%. (vii) (a) CDI, DIEA, DMF, 1-3 h; (b) 11, DIEA, DMF, 16-40 h, 51-95%.
3
The functional agonist and competitive antagonist proper-
ties of each ligand were determined for the human vasopressin
and oxytocin receptors subtypes stably expressed in CHO
cells. The accumulation of myo-inositol 1-phosphate (V1a and
OT receptors) was determined by the IP-One assay and the
accumulation of cAMP (V2 receptor) was determined by the
cAMP dynamic 2 assay, both kindly provided by Cisbio
International. These immunoassays are based on the competi-
tion between free IP1 or cAMP and IP1-d2 or cAMP-d2
conjugate, respectively (for more information on the princi-
ple, see http://www.htrf.com/products/gpcr/ipone and http://
www.htrf.com/products/gpcr/camp).
model was assessed by docking first the endogenous peptide
ligand oxytocin (see Figure S1 in Supporting Information).
OT and compound 1 were docked to this preliminary model
using the Gold 3.0 program.38 The starting conformation of
OT was modeled from the X-ray structure of neurophysin-
bound OT.39 Compound 1 was built starting from the X-ray
structure of its truncated analogue 12c, and its geometry was
fully optimized with HF/6-31* level using Gaussian 03.40 The
resulting binding mode of compound 1 obtained was similar
to the oxytocin binding mode (Figure 3).
Results and Discussion
Full Characterization of Compound 1. Molecule 1 is the
most active of very few nonpeptide oxytocin agonists pub-
lished to date.22-24 It results from the optimization of a hit
discovered after high throughput screening. The activity of 1
on OT receptor (EC50 = 33 nM) was characterized using a
NFAT-luciferase gene reporter assay. Its agonist activity for
Molecular Modeling
A three-dimensional model [model coordinates are avail-
able upon request] of the oxytocin receptor was built and
refined based on the sequence alignment and the V1a model
previously published by our group.37 The quality of this new

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1549
Scheme 3. Synthesis of the Molecular Hybrids Conserving the
Western Part of Compound 1a
Table 2. Structures and Biological Activities of Truncated Analogues of
Compound 1
a Reagents and conditions: (i) (a) (COCl)2, cat. DMF, DCM, 30 min;
(b) 7, Et3N, DCM, overnight, 90%. (ii) SnCl2 2H2O, EtOH, reflux, 1 h,
3
a The inhibition constants (Ki in nM) for human V1a, V2, and OT
receptors were determined on CHO cell membranes by competition
binding assays (displacement of radioactive [3H]AVP). Results are
expressed as mean ( SEM of at least three separate experiments
performed in triplicate. When Ki >1000 nM, these are results of at least
two separate experiments performed in triplicate. b The activation
constants (Kact in nM) were measured in CHO cell lines expressing
either vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and
OT-R, IP1 accumulation) or cAMP dynamic 2 (V2-R, cAMP accumu-
lation) assays. These values are the mean ( SEM of at least three
separate experiments performed in triplicate. The maximal stimulations
(Emax) are expressed as percentages of the endogenous agonist maximal
stimulation (n = 3). In case of a weak stimulation, results are expressed
as percentages of the endogenous agonist maximal stimulation at
[ligand] = 1 μM (n = 2). Results are not significant (ns) when there is
<10% response at [ligand] = 1 μM (n = 2). c The inactivation constants
(Kinact in nM) were measured in CHO cell lines expressing either
vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and OT-
R, inhibition of agonist-induced IP1 accumulation) or cAMP dynamic 2
(V2-R, inhibition of agonist-induced cAMP accumulation) assays.
These values are the mean ( SEM of at least three separate experiments
performed in triplicate. In case of a weak inhibition, results are expressed
as percentages of endogenous agonist-response inhibition at [ligand] =
1 μM (n = 2). Results are not significant (ns) when there is <10%
response inhibition at [ligand] = 1 μM (n = 2). d [ligand] up to 10 μM.
82%. (iii) (a) 2-Phenylbenzoic acid, (COCl)2, cat. DMF, DCM, 5 h;
(b) 26, Et3N, DCM, 36 h, 87%. (iv) 3-Methylpyrazole, K2CO3, NMP,
120 °C, 6-25 h, 9-41%. (v) HCl/H2O/AcOH, reflux, 6 h, 96-99%.
(vi) (a) SOCl2, cat. NMP, DCM, overnight; (b) 7, pyridine, cat. DMAP,
DCM, 3 d, or μw 80 °C, 8 min, 53-81%.
the V2 receptor subtype (EC50 = 850 nM) was evaluated in a
similar way (CRE-luciferase). Furthermore, 1 was shown to
be active in vivo after iv bolus administration in a rat model
of uterine response. We have extended the characterization
in studying the binding profile of this lead compound at the
OT, V1a, and V2 oxytocin and vasopressin receptor subtypes.
Furthermore, its agonist and antagonist functional efficacy
was directly measured at the level of the second messengers.
In terms of binding affinity, 1 was less potent than
expected at the OT receptor with Ki = 147 nM (Table 1).
Its affinity for the V1a and V2 receptor subtypes was Ki =
330 nM andKi >1000 nM, respectively. Intermsof functional
efficacy, 1 showed a weak but substantial agonist activity at
theOT(Kact = 667nM) and V2 (Kact = 599 nM) receptors but
not at the V1a receptor. However, the activation did not reach
the level induced by OT or AVP, indicating a partial agonist
profile with an Emax of 58% and 84% for OT and V2
receptors, respectively. The capacity of 1 to antagonize the
endogenous ligand activation was also measured. In the
conditions of the study, compound 1 showed some weak
antagonist effect at the OT receptor, no significant effect at
determinants of affinity and efficacy in the vasopressin-
oxytocin system.
Determinants for the Agonism/Antagonism Switch. The
binding and the switch from agonist to antagonist among
G protein-coupled receptor (GPCR) ligands remain a key
issue for drug design. The vasopressin-oxytocin receptors
represent a very interesting system to investigate this ques-
tion. Indeed, the different receptor subtypes are very similar
(37% to 47% identity), they are activated by the same
endogenous agonist, arginine vasopressin (AVP), they have
been cloned in many species including humans,41 many
peptidic and nonpeptidic ligands are known, and three-
dimensional models validated by many experimental data
are available.34-37,42-59
Among the structurally diverse nonpeptide ligands of
vasopressin receptors disclosed over the years, the tetra-
hydrobenzazepine/benzodiazepine series is one of the most
important so far. Ligands of this class are active at three of
the four receptor subtypes, often exhibiting potent antago-
nist properties at the V1a and/or V2 receptors, and sometimes
a non-negligible OT-R antagonist activity (e.g., compound
2).21 To date, very few nonpeptide agonists acting at these
receptors have been reported.22 Interestingly, the only ones
the V2 receptor, and a marked V1a antagonist effect (Kinact
=
271 nM). Overall, compound 1 is a moderatelyactiveligand of
the OT receptor with a marginal selectivity versus the V1a
receptoranda largerselectivityversusthe V2 site(>7fold). At
the level of second messengers, it behaves as a partial agonist
of the OT and V2 receptors and as a slightly more potent
antagonist at the V1a receptor. The apparent discrepancy
between the reported EC50 and full efficacy at OT receptors
derived from a gene reporter assay24 and the affinity and
efficacy measured here from radioligand binding and second
messenger analysis can easily be explained by the fact that the
gene reporter assays reflect a downstream effect with several
levels of signal amplification. The extension of profiling to the
V1a receptor subtype unmasked the weak selectivity of com-
pound 1, hampering its usefulness as a pharmacological
investigation probe. However, this molecule remains a very
interesting tooltoinvestigate the ligand-receptor interactions
at the molecular level and more particularly the structural

1550 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
Table 3. Structures and Biological Activities of Hybrid Molecules Exchanging the Western Part of 1
a The inhibition constants (Ki in nM) for human V1a, V2, and OT receptors were determined on CHO cell membranes by competition binding assays
(displacement of radioactive [3H]AVP). Results are expressed as mean ( SEM of at least three separate experiments performed in triplicate. When
Ki >1000 nM, these are results of at least two separate experiments performed in triplicate. b The activation constants (Kact in nM) were measured in
CHO cell lines expressing either vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and OT-R, IP1 accumulation) or cAMP dynamic 2 (V2-R,
cAMP accumulation) assays. These values are the mean ( SEM of at least three separate experiments performed in triplicate. The maximal stimulations
(Emax) are expressed as percentages of the endogenous agonist maximal stimulation (n = 3). In case of a weak stimulation, results are expressed as
percentages of the endogenous agonist maximal stimulation at [ligand] = 1 μM (n = 2). Results are not significant (ns) when there is <10% response at
[ligand] = 1 μM (n = 2). c The inactivation constants (Kinact in nM) were measured in CHO cell lines expressing either vasopressin V1a or V2 or oxytocin
receptor by IP-One (V1a-R and OT-R, inhibition of agonist-induced IP1 accumulation) or cAMP dynamic 2 (V2-R, inhibition of agonist-induced cAMP
accumulation) assays. These values are the mean ( SEM of at least three separate experiments performed in triplicate. In case of a weak inhibition,
results are expressed as percentages of endogenous agonist-response inhibition at [ligand] = 1 μM (n = 2). Results are not significant (ns) when there is
<10% response inhibition at [ligand] = 1 μM (n = 2). d n = 2. e V2-R: 13 ( 1% of response at 10 μM (n = 3); OT-R: maximal stimulation could not be
reached at 100 μM. f V2-R: 43 ( 6% of inhibition at 10 μM (n = 3); OT-R: 11 ( 16% (n = 2) and 56% (n = 1) at 10 and 100 μM, respectively. g [ligand] up
to 100 μM (n = 3). h Inhibition at 10 and 100 μM, respectively (n = 2): V1a-R: 7.0 ( 0.8% and 92 ( 2%, V2-R: 25 ( 1% and 94 ( 1%, OT-R: 17 ( 12%
and 52 ( 14%. i [ligand] up to 10 μM (n = 3).
belong to the benzazepine series (e.g., compound 3).27,60-62
Their V2-R agonist properties were discovered as a result of
structural modifications of the antagonists.33,63 Intriguingly, the
first nonpeptide OT-R agonist (compound 1) is also related to
the said chemical class.24 Therefore, a systematic study of these
remarkably subtle relationships between these particular ligands
should be of outstanding interest to clarify the structural deter-
minants conferring them such agonist/antagonist properties.
Thus, in our attempts to understand the structural origin
of compound 1 OT receptor agonist activity, we decided to
cleave the structure progressively or to combine its fragments
with fragments from a reported nonselective V1a/V2/OT
receptor antagonist, compound 2, a V2 receptor agonist,
WAY-VNA-932, compound 3, and a benzoxazinone-type
selective OT receptor antagonist, L-371,257, compound 4.
The affinity and efficacy on second messenger activation of
the resulting products have been determined.
Truncated Analogues. As seen from Table 2, the removal of
the thioamide-homopiperazine from compound1 (compound
12a) led to a very significant decrease in affinity for OT-R
because a very low [3H]AVP displacement could be observed
at a 1 μM concentration. No agonist activity could be
detected, and a 37% inhibition of the response induced by
OT could be observed at 1 μM, suggesting a weak and
marginally reinforced antagonist effect compared to 1. Addi-
tional cleavage leading to the urea 12b and to the acetamide
12c resulted in a complete loss of affinity and efficacy for the
OT receptor. Again, some weak but significant antagonistic
activity could still be detected for 12b. One may then conclude
that the eastern fragment of molecule 1 contributes signifi-
cantly to its affinity for the OT receptor. The lack of agonist
efficacy and some residual antagonist activity in compound
12a suggested that the homopiperazine fragment is necessary
for the OT receptor activation. The same trend was observed
on the V2 receptor subtype. The initial affinity was very weak
and remained as such in compounds 12a and 12b. However, in
contrast to OT-R, the agonist character persisted in the series
of analogues although very weakened. No antagonist activity
could be detected. Finally, on the V1a receptor subtype, the
affinity was slightly reinforced upon truncating the molecule,
as indicated by both the affinity and the antagonist efficacy of
1, 12a, and 12b. However, the ultimate transformation of the
urea in 12b into an acetamide in 12c was completely detri-
mental to the activity.
This first part of the study highlighted the crucial role of
the entire chain for the OT-R affinity of 1 and the specific
contribution of the homopiperazine fragment to its agonist
efficacy. Unexpectedly, it also highlighted the divergent

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1551
Table 4. Structures and Biological Activities of Hybrid Mmolecules
Exchanging the Eastern part of 1
and 23b. Even more surprisingly, the chimeric ligand 23d
switched from OT agonist to OT antagonist upon this ring
replacement. In a general manner, the affinity and efficacy
on the V1a and V2 homologous receptors were also decreased
without change in the efficacy profile.
These results led to two conclusions. The highly similar
cyclic aromatic fragments of the different ligands do not
contribute in the same way to their activity, probably inter-
acting in different manners with a given receptor as well as
with the different homologous receptor subtypes. The aro-
matic cyclic fragments of molecules 1, 2, and 3 stabilize the
active conformation of the OT receptor, whereas the piper-
idino-benzoxazinone of 4 stabilizes an inactive conforma-
tion. Interestingly, both benzazepine and piperidino-
benzoxazinone moieties actually derive from structural
modifications of the quinolinone ligand OPC-21268, the first
nonpeptide vasopressin receptor ligand described.28,64,65
Structurally speaking, there is a reasonably good similarity
between the piperidino-benzoxazinone fragment of 4 and the
benzazepine moiety. They both contain a cyclic aromatic
moiety likely to interact with aromatic residues in the active
site. They are both connected to the rest of the molecules by
an amide link. The discriminant feature could be the distance
between the aromatic ring and the amide link that is slightly
superior in the piperidino-benzoxazinone moiety.
The contribution of the eastern part of the molecule was
then investigated by introducing the side chains of com-
pounds 2 and 3 on the central and western part of 1. Once
again, the results were quite unexpected (Table 4). The
replacement of the complex electronegative side chain in 1
by the hydrophobic biphenyl amide of 2 led to compound 27
with a conserved OT receptor affinity (Ki = 111 nM versus
147 nM). The replacement by the methyl pyrazole directly
branched on the central benzoyl core provided 31a, an even
more potent OT-R ligand with a Ki = 27.8 nM. This
significant increase was unexpected because compounds 2
and 3 have affinity constants similar to 1 (Ki = 109 and 199
nM, respectively). However, these replacements caused a
switch in functional efficacy because compounds 27 and
31a are potent OT receptor antagonists (Kinact = 557 and
14.8 nM, respectively) without any detectable agonist activ-
ity. These results highlight the critical contribution of the
large electron-rich side chain of 1 to its OT receptor activat-
ing properties. Replacement by smaller, more hydrophobic
fragments increases the affinity but locks the receptor in an
inactive conformation. A similar effect was observed on the
V2 receptor subtype. Compound 27 is also a potent and
selective V2 receptor antagonist (Ki = 6.4 nM, V2/V1a Kinact
ratio = 374, V2/OT Kinact ratio = 506). In addition, the
affinity for the V1a receptor subtype increased from 10- to
20-fold (compounds 27 and 31a, respectively) but the initial
antagonist activity was retained.
a The inhibition constants (Ki in nM) for human V1a, V2, and OT
receptors were determined on CHO cell membranes by competition
binding assays (displacement of radioactive [3H]AVP). Results are
expressed as mean ( SEM of at least three separate experiments
performed in triplicate. When Ki >1000 nM, these are results of at least
two separate experiments performed in triplicate. b The activation
constants (Kact in nM) were measured in CHO cell lines expressing
either vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and
OT-R, IP1 accumulation) or cAMP dynamic 2 (V2-R, cAMP accumu-
lation) assays. These values are the mean ( SEM of at least three
separate experiments performed in triplicate. The maximal stimulations
(Emax) are expressed as percentages of the endogenous agonist maximal
stimulation (n = 3). In case of a weak stimulation, results are expressed
as percentages of the endogenous agonist maximal stimulation at
[ligand] = 1 μM (n = 2). Results are not significant (ns) when there is
<10% response at [ligand] = 1 μM (n = 2). c The inactivation constants
(Kinact in nM) were measured in CHO cell lines expressing either
vasopressin V1a or V2 or oxytocin receptor by IP-One (V1a-R and
OT-R, inhibition of agonist-induced IP1 accumulation) or cAMP
dynamic 2 (V2-R, inhibition of agonist-induced cAMP accumulation)
assays. These values are the mean ( SEM of at least three separate
experiments performed in triplicate. In case of a weak inhibition, results
are expressed as percentages of endogenous agonist-response inhibition
at [ligand] = 1 μM (n = 2). Results are not significant (ns) when there is
<10% response inhibition at [ligand] = 1 μM (n = 2). d [ligand] up to
10 μM (n = 3).
structure-activity relationships among highly similar recep-
tor subtypes because structural variations led to a loss of
affinity and a switch from agonism to antagonism in one case
(OT-R), a retention of affinity and agonist efficacy in a
second case (V2-R) and an increase in affinity and antagonist
efficacy in the last case (V1a-R) (Figure 1).
Hybrid Molecules. To further explore this SAR diver-
gence, we prepared and analyzed the hybrid molecules in
Tables 3 and 4. A simple qualitative analysis of these ligands
strongly suggested a shared pharmacophore around a cen-
tral benzoyl fragment branched with a polycyclic aromatic
amide on the western side and an aromatic or electron-rich
side chain on the opposite para position (eastern). It was
tempting to assume that similar fragments bind to similar
regions and trigger similar effects in the highly homologous
receptor subtypes.
The contribution of the western tricyclic part was first
studied. Unexpectedly, despite the fact that compounds 2, 3,
and 4 had a good or very good affinity for the OT-R, better
or comparable to 1, the introduction of their aromatic
fragments in the structure of 1 led to a significant (23a,b,d)
to dramatic (23c) decrease in affinity for the OT receptor
(Table 3). The efficacy decreased in a parallel fashion for 23a
Of interest is also the differential action of the core benzoyl
nucleus common to all ligands. With a methyl group in
position 3, the compound 31a is a potent antagonist for the
V1a and OT receptors and a weak antagonist for the V2
receptor. Substitution of the methyl for a chlorine in position
2 led to an equipotent ligand (31b) for the three receptor
subtypes. However, 31b behaves as a full agonist at the V2
and as an antagonist at the V1a and OT receptor subtypes,
highlighting once more the divergent SAR among highly
homologous ligands acting at highly homologous receptors.
Furthermore, by applying the same modification when
having the large side chain of 1, the affinity for the three

1552 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
separation renders the design of small size, potent agonists
more difficult. However, morphine represents a very good
example showing that a small nonpeptide molecule is able to
mimic a peptide in combining affinity and functional activity
on a limited molecular entity. Furthermore, the switch from
agonist to antagonist on morphine skeleton follows essen-
tially the same rules as for biogenic amine receptor ligands
and can be achieved by substitution with hydrophobic frag-
ments.70 By contrast, other examples have shown that a
nonpeptide antagonist ligand of a peptide receptor could be
converted into an agonist by introduction of an additional
hydrophobic interaction or steric constraint.69,71,72 This
phenomenon has also been studied on several receptors via
mutagenesis and photoaffinity labeling of structurally simi-
lar agonist and antagonist ligands.73-87 As shown here, the
side-chain decrease strategy leads indeed to a conserved
agonist character or a switch from antagonist to agonist
for the V2 receptor subtype but did not operate for the OT
and V1a receptors, demonstrating that the structural deter-
minants of the antagonist/agonist switch are probably
specific to every receptor-ligand complex. We have at-
tempted to obtain an explanation for these observed subtle
affinity and efficacy switches in analyzing molecular models
of the OT receptor binding to OT, the endogenous ligand, or
to compound 1, a partial agonist.
Molecular Modeling. Models for the OT, V1a, V1b, and V2
receptors interacting with diverse ligands had previously
been proposed and partly validated experimentally.42-59
They were derived from reference GPCR experimental
structures with different levels of relevance and resolution.
However, they proved to be sufficiently qualitative to predict
binding pocket localization, residues responsible for affinity,
species, and receptor specificity. We attempted to go one step
further in understanding the agonism-antagonism switch
for compound 1 using the best current reference templates
for GPCR modeling. A three-dimensional model of the
oxytocin receptor was built and refined based on the
sequence alignment and the V1a model previously published
by our group.37 The quality of this new model was assessed
by docking the endogenous peptide ligand oxytocin (see
Figure S1 in Supporting Information). In agreement with
most studies, the hydrophobic part of oxytocin (Tyr2 and
Ile3) is accommodated by a hydrophobic pocket (M3.36,
W6.48, F6.51) lying in the deep 7-transmembrane (TM)
cavity between TMs III, V, VI, and VII. An important
H-bonds network is established between the conserved gluta-
mine residues (Q3.32, Q4.60, Q6.55) located at the upper rim
of the cavity and the polar part of the peptide hormone,
i.e., the main chain atoms as well as Gln4, Asn5, and also
the Tyr2 side chain. Similar interactions have also been
described for AVP into V1a-R and V1b-R in our most
recent models37 as well as in previously published binding
modes.43,48 Moreover, the N-terminal Gly9 of OT is inter-
acting with E1.35 via hydrogen bonding. Thus, a very tiny
network of hydrogen bonds is used to anchor OT to its
receptor. On the other hand, Leu8 of OT is pointed to the
F2.68, located close to the extracellular loop E1, as proposed
previously.44 Noteworthy, most residues proposed to parti-
cipate in OT binding have been shown to be of crucial
importance for AVP agonist peptide binding using site-
directed mutagenesis.43,48,49 Thus, although our OT binding
has not been experimentally validated, we believe that our
OT-R model is a good starting point for docking other
agonists.
Figure 2. Schematic representation of the main results of the
hybrid molecules of compound 1.
receptor subtypes is dramatically lost, as illustrated by
compounds 23b and 23c (Table 3). Therefore, one may
hypothesize that a methyl group on the central benzoyl core
seems to be required at position 3 to stabilize this flexible
chain in its active conformation.
The second part of this study highlighted the contribution
to affinity, efficacy, and specificity of each fragment of the
compound 1. With regard to the western tricyclic part, the
benzazepine polycyclic moiety contributes to the affinity and
the stabilization of the active conformation of the OT and V2
receptors, the pyrazolobenzodiazepine scaffold of 1 remain-
ing unequaled (Figure 2). By contrast, the piperidino-ben-
zoxazinone is responsible for the OT-R selectivity and
stabilizes an inactive conformation of this receptor. Clearly,
both scaffolds are not interchangeable. Regarding the east-
ern part of the molecule, the large electron-rich side chain of
1 is essential for the oxytocin receptor activation. Finally, the
core benzoyl nucleus is not just a common linker but can
lock/unlock a specific conformation of the ligand through its
different substitution patterns.
In summary, the design and synthesis of several truncated
analogues of 1 and of hybrid molecules enabled us to identify
the key structural features responsible for its affinity and
OT-R agonist efficacy. Although it was tempting to assume
that similar fragments bind to similar regions and trigger
similar effects in the highly homologous receptor subtypes,
the reality is definitely not so straightforward. Indeed, each
part of the molecule appears to play a specific role for its
activity on the oxytocin receptor, and quite surprising find-
ings arose from this systematic study regarding its behavior
toward the other vasopressin receptor subtypes. The main
conclusion remains that the OT-R agonist activity of com-
pound 1 is mainly conferred by this large, rather polar,
pseudopeptidic side chain present at its eastern part. The
cleavage of this polar side chain and its replacement by a
smaller group did not afford an OT-R agonist as it does with
the same backbone on the V2-R.33,63 For bioamine receptors
ligands, which have to reach a deep binding cleft between the
helices, agonists are typically smaller than antagonists.66,67
Indeed, the switch from an antagonist to an agonist for
nonpeptide receptors is generally achieved by the suppres-
sion of a bulky lipophilic side chain. Obviously, it may be
more subtle for peptide receptor ligands, whose binding site
also includes the extracellular region. The large size of
peptides often allows the discrimination of the binding and
the activation domains, both on the peptide ligand and the
target receptor. This address and message concept has been
widely explored and often validated.46,49,54,68,69 A priori, this

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1553
lead to the movement of TM VI and TM VII around the
highly conserved proline kinks, while the intracellular seg-
ments would move away from TM III to expose specific
epitopes, and the extracellular segments would get closer to
TM III to close the binding site. The receptor would be
maintained in its inactive conformation by an aromatic
cluster in TM V and VI, stabilizing the W6.48. The antago-
nist binding mode is located deep in the transmembrane
domain and based on sufficiently strong interactions with the
hydrophobic and aromatic residues of TM IV, V, VI, and
VII, in order to prevent W6.48 rotation, maintaining the
receptor in its inactive conformation. The agonist would
adopt a more superficial binding mode between TM III, VI,
and VII and close to TM III to allow rotation of the W6.48
located at the opposite part of the cavity. Established inter-
actions are thus rather polar, with some hydrophobic con-
tacts able to disturb the aromatic cluster. In stabilizing the
active receptor conformation, small agonists would act as a
molecular glue in the main cavity between the helices,
whereas peptides would play the role of Velcro at the upper
helical bundle and the extracellular loops, to maintain these
helices together for an extended time. Small V2-R agonists
would bind in a position close to TM III to not disturb the
W6.48 rotation. Because of subtle differences of topology of
the cavity in OT-R, interaction of these ligands with the
aromatic cluster of TM VI would be too strong. On the
contrary, the eastern chain of the OT-R agonist 1 would
bring additional interactions with the upper part of TM I and
II, allowing the molecule to move upward, thus acting as
Velcro in peptide mimicry.
Figure 3. Proposed binding mode of compound 1 to human OT
receptor. The seven TM helices of each receptor are displayed by
cylinders and labeled from I to VII. Some TM helices have been
omitted for clarity. Compound 1 is displayed by orange sticks. This
figure and Figure S1 in Supporting Information has been prepared
with Sybyl.99
Compound 1 was docked in this model in an attempt to
gain insight into the agonism-antagonism switch. The
resulting binding mode obtained was similar to the oxytocin
binding mode (Figure 3). More precisely, a well-defined
negatively charged subsite (E1.35, D2.65) perfectly accom-
modates the positively charged basic amine at the eastern
part of compound 1. Very recently, we demonstrated that
this subsite represents a key anchoring point for the Arg8
residue of AVP and related peptide agonists in V1a and V1b
receptors.37 The western part of compound 1 is anchored via
H-bonds formed between its nitrogen-enriched tricyclic
fused ring and the conserved Q4.60 as well as the carbonyl
belonging to the backbone of K3.29. A possible aromatic
stacking interaction is also likely between the central benzene
ring of the ligand and the conserved F6.51 of the receptor.
Consequently, this novel oxytocin nonpeptide agonist
series represented by compound 1 seems to bind to the
receptor in a similar way that the peptide ligand does. The
extended pseudopeptidic structure of 1 seems to mimic
the size and the shape of oxytocin itself and to establish the
interactions necessary to activate the oxytocin receptor.
Indeed, our results indicate that the end of the eastern part
of compound 1 is required for OT-R activation and one can
imagine that the basic amine of the homopiperazine frag-
ment would be the key anchoring point to the receptor in
order to confer agonism to the ligand. According to our
model, this nitrogen can be in close contact with three
possible acidic residues of the OT receptor, namely E1.35,
D2.65 and the conserved D of the extracellular loop E2.
Interestingly, only D2.65 is not conserved in the V2 receptor,
where it is replaced by a K. This might explain why com-
pound 1 has a very low affinity and efficacy for the V2
receptor which binds and responds to the small hydrophobic
agonist WAY-VNA-932. Nevertheless, this model is not
sufficient to explain the agonism of compound 1 at the OT
receptor and its antagonism at the V1a receptor because
oxytocin itself has no positive charge at the region considered
and the three acidic residues are conserved in V1a-R.
Conclusions
The binding mode and the switch from antagonist to
agonist among GPCR ligands remain a key issue for drug
design. The vasopressin-oxytocin receptors and their ligands
represent a very convenient model system to address such an
issue because highly homologous ligands binding to highly
homologous binding sites are known. We have explored the
affinity and efficacy shifts around compound 1, which is the
leadof thefirst series of oxytocinreceptor nonpeptideagonists
described to date. We have fully investigated its binding and
efficacy profiles on the OT, V1a, and V2 oxytocin-vasopressin
receptors. The first important result is that compound 1 is not
a specific, full OT-R agonist. In our hands, it displays only a
2-fold selectivity versus the V1a receptor in terms of binding
and behaves as a partial agonist (58%) at the OT-R, being
equipotent as a V2-R agonist and more potent as a V1a-R
antagonist. This, added to its poor bioavailability, preempts
its use as specific pharmacological probe for studying oxyto-
cin-related functional responses. It remains, however, a very
interesting and relevant tool to explore the oxytocin binding
and efficacy supramolecular determinants. Its study leads to
the second conclusion that structure-activity relationships
around vasopressin/oxytocin ligands are extremely subtle and
the determinants for the antagonism/agonism switch are very
difficult to identify. Thus, although qualitatively accurate,
three-dimensional models of GPCR still remain of weak
predictive value to address such subtle phenomenon that
might be under kinetic and/or polymolecular control. Indeed,
static models can not satisfactory take into account dynamic
processes, conformational changes, loops movements, solva-
tion/desolvation, allosteric interactions, dimerization/oligo-
merization,90 G proteinbinding, membrane environment, and
Thus, studying the GPCR activation mechanisms still
remains challenging. A lot of work has been done on
the rhodopsin and the β2-adrenergic class 1 receptors.88,89
According to the proposed toggle switch model,89 a 90°
rotation of the crucial conserved W6.48 side chain would

1554 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
the multitude of parameters intervening in a living cell.
Molecular modeling of GPCR, associated with extended
experimental validation, is definitely an additive value to drug
design but, in our hands and in this particular case study, it
reaches its limits. The recent crystallographic structure of the
adenosine A2A receptor further highlights the plasticity of
GPCRs and the difficulty to predict precisely a given ligand
binding mode.91 Thus, for the time being, the dual challengeof
discovering a full, bioavailable OT agonist on one hand
and rationally designing a nonpeptide agonist for a peptide
receptor on the other hand remains open. However, waiting
for anoxytocinreceptor crystallographic structureand having
in mind that peptide GPCRs represent targets of special
interest for novel therapeutics,92,93 we will continue to com-
bine molecular modeling, SAR studies, labeling,34-36,55 and
mutagenesis to take up the challenge.
(bs, 1H). 13C NMR (75 MHz, DMSO-d6) δ 35.0, 100.5, 120.3,
121.8, 123.4, 123.9, 129.4, 133.8, 139.9, 146.0, 163.8. IR (KBr) ν
(cm-1) 3310, 3221, 3098, 3056, 1636, 1615, 1584, 1504, 1446,
1421, 1398, 1329, 1268, 1206, 747.
1-Methyl-1,4,5,10-tetrahydropyrazolo[3,4-b][1,5]benzodiazepine
(7). To a stirred suspension of 6 (714 mg, 3.33 mmol) in dry THF
at 0 °C under argon was added portionwise LiAlH4 (1.01 g, 26.67
mmol) over 10-15 min. The resulting reaction mixture was stirred
at 0 °C for 15 min and was then refluxed. After 26 h, as TLC
showed incomplete reaction, an additional amount of LiAlH4
(755mg, 20.00mmol) wasadded. After 44 hrefluxing, the reaction
mixture was cooled to 0 °C and 35% NaOH (4 mL) was added
dropwise over 1 h. Stirring was continued at rt for 1 h, then the
mixture was filtered over celite and concentrated in vacuo. Flash
chromatography (DCM to 95:5 DCM/MeOH) afforded the title
compound as a brown foam (576 mg, 86%); mp 204-206 °C
1
(from MeOH). H NMR (200 MHz, CD3OD) δ 3.71 (s, 3H),
3.98 (s, 2H), 6.65-6.95 (m, 3H), 6.95-7.15 (m, 2H). 13C NMR
(75 MHz, CD3OD) δ 34.9, 45.6, 103.0, 120.5, 122.5, 123.6, 124.1,
135.4, 136.8, 139.9, 143.1. IR (KBr) ν (cm-1) 3364, 3262, 3153,
3049, 2980, 2791, 1607, 1574, 1506, 1490, 1432, 1395, 1318, 1292,
1247, 993, 821, 757. MS (ES) m/z 201.12 ([M þ H]þ).
Experimental Section
Chemistry: General Methods. Dry solvents were purchased
from Fluka (SureSeal). Microwaves irradiations were per-
formed on a Biotage Initiator 2.0 apparatus. Melting points
(mp) were determined with a Gallenkamp apparatus and are
uncorrected. Optical rotations ([R]D) were determined with a
2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b][1,5]benzo-
diazepin-5(1H)-yl) carbonyl]benzonitrile (9). To a stirred suspen-
sion of 8 (1.10 g, 6.81 mmol) in dry DCM (8 mL) at 0 °C under
argon were added successively dry DMF (5 drops) and (COCl)2
(1.17 mL, 13.61 mmol). The resulting reaction mixture was
stirred at 0 °C for 15 min and then was allowed to warm to rt
and stirring was continued for 2 h. The solvent was removed in
vacuo, and the resulting residue was dissolved in dry THF. The
solvent was removed in vacuo, and the resulting residue was
dried under high vacuum for 30 min. The resulting brown solid
was dissolved in dry DCM (5 mL), and this mixture was
added dropwise, under argon, to a stirred ice-cold solution of
7 (1.14 g, 5.69 mmol) and Et3N (1.60 mL, 11.38 mmol) in dry
DCM (10 mL). Stirring was continued at 0 °C for 30 min and
then at rt overnight. As TLC showed still presence of some
starting amine 7, an additional amount of acyl chloride was
prepared from the acid 8 (458 mg, 2.28 mmol), dissolved in dry
DCM (2 mL) with Et3N (0.80 mL, 5.69 mmol), and added
dropwise to the ice-cold reaction mixture. Stirring was contin-
ued at rt overnight and then the solvent was removed in vacuo.
The resulting residue was taken up in DCM (50 mL) and 0.3 M
KHSO4 (75 mL). The aqueous phase was separated and ex-
tracted with a 8:2 CHCl3/iPrOH mixture (100 then 50 mL). The
combined organic layers were washed with satd NaHCO3
(75 mL), dried (Na2SO4), and concentrated in vacuo to yield a
residue that was triturated in AcOEt. The precipitate was
filtered, washed with AcOEt, and dried under high vacuum to
yield the title compound as a pale-yellow powder (1.46 g, 75%);
mp 253-256 °C. 1H NMR (300 MHz, DMSO-d6) δ 2.34 (s, 3H),
3.77 (s, 3H), 3.95 (d, 1H, J = 14.3 Hz), 5.64 (d, 1H, J = 14.3 Hz),
6.64 (t, 1H, J = 7.5 Hz), 6.79 (d, 1H, J = 7.9 Hz), 7.01 (d, 1H,
J = 7.9 Hz), 7.12 (t, 1H, J = 7.5 Hz), 7.19 (s, 1H), 7.23-7.33 (m,
2H), 7.55 (d, 1H, J = 7.9 Hz), 8.67 (bs, 1H). 13C NMR (50 MHz,
DMSO-d6) δ 19.7, 35.4, 43.2, 100.0, 112.2, 117.3, 119.5, 121.2,
124.8, 128.7, 129.0, 130.4, 131.3, 131.9, 135.6, 139.1, 139.6,
140.8, 141.3, 166.9. IR (KBr) ν (cm-1) 3339, 2910, 2863, 2225,
1633, 1606, 1563, 1504, 1449, 1419, 1387, 1321, 1298, 1251, 1176,
1139, 991, 840, 818, 767.
1
Perkin-Elmer 241 MC or a Perkin-Elmer 341 polarimeter. H
and 13C NMR spectra were recorded at 298 K on Bruker
spectrometers at 200 or 300 and 50 or 75 MHz, respectively, in
CDCl3, DMSO-d6, or CD3OD, using residual solvent as an
internal reference. Chemical shifts (δ) are reported in parts per
million (ppm), coupling constants (J) are reported in hertz (Hz).
IR spectra (ν, cm-1) were obtained with a Perkin-Elmer FT-IR
1600 or a Perkin-Elmer FT-IR 2000 spectrometers, or with a
Thermo Nicolet 380 spectrometer using attenuated total reflec-
tion (ATR) technology. Mass spectra (MS) were recorded on a
Mariner 5155 from Applied Biosystems using electrospray (ES)
ionization mode and a time-of-flight (TOF) analyzer. High
resolution mass spectra (HRMS) were recorded on a MicroTof
mass spectrometer from Bruker using electrospray (ES) ioniza-
tion mode and a time-of-flight (TOF) analyzer. Thin-layer
chromatography (TLC) was performed on silica gel 60 F254
aluminum plates from Merck. Flash chromatography was
performed on silica gel 60 (230-400 mesh ASTM) from Merck.
Analytical RP-HPLC analyses were performed on a Symmetry
Shield column (4.6 mm ꢀ 150 mm, 5 μm, C18) from Waters
under the following conditions: buffer A, 0.1% aqueous TFA;
buffer B, 0.1% TFA in 8:2 CH3CN/H2O; buffer C, 0.1% TFA in
CH3CN; gradient 1: 0% buffer B in buffer A for 5 min then
0-100% buffer B in buffer A over 25 min with a flow rate of
0.8 mL/min; gradient 2, 0-100% buffer C in buffer A over
30 min with a flow rate of 1 mL/min; detection: λ = 220/254 nm.
Retention times (tR) from analytical RP-HPLC are reported in
minutes. Semipreparative RP-HPLC separations were per-
formed on a Symmetry Shield column (19 mm ꢀ 300 mm,
0.7 μm, C18) from Waters under the following conditions: flow
rate: 10 mL/min; buffer A, 0.1% aqueous TFA, buffer B, 0.1%
TFA in CH3CN; gradients (variable): 0-100% buffer B over
30 min; detection: λ = 220/254 nm. Purity of all target com-
pounds was determined by analytical RP-HPLC according to
the conditions described above, and was >95%.
1-Methyl-5,10-dihydropyrazolo[3,4-b][1,5]benzodiazepin-
1-{2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b][1,5]ben-
zodiazepin-5(1H)-yl) carbonyl]phenyl}methanamine (10). To a
stirred suspension of 9 (49.6 mg, 0.144 mmol) and CoCl2 6H2O
(70.1 mg, 0.289 mmol) in MeOH (1.5 mL) at 0 °C under argon
was added portionwise NaBH4 (55.7 mg, 1.444 mmol) over
5 min. The resulting black reaction mixture was stirred at
0 °C for 10 min and then was allowed to warm to rt, and stirring
was continued for 1 h. After neutralization (pH 7-8) with
1 M KHSO4 (2 mL), MeOH was removed in vacuo, and the
resulting aqueous residue was diluted with 1 M KHSO4 (15 mL).
4(1H)-one (6). A solution of 5 (1.42 g, 5.45 mmol) in a 9:1 iPrOH/
AcOH mixture (25 mL) was heated in a sealed tube at 160 °C
under argon. After 48 h, the resulting mixture was allowed to
stand at rt under argon for 1 or 2 days until complete crystal-
lization of the product. The creamy to pale-gray crystals were
then collected, washed with cold Et2O, and dried under high
vacuum to yield the title compound (1.04 g, 89%); mp 262-264 °C.
1H NMR (300 MHz, DMSO-d6) δ 3.66 (s, 3H), 6.75-6.95
(m, 3H), 7.03 (d, 1H, J = 7.5 Hz), 7.45 (s, 1H), 8.35 (bs, 1H), 8.96
3

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1555
This aqueous phase was filtered over celite, extracted with Et2O
(2 ꢀ 20 mL), basified (pH > 10) with 2 N NaOH, and extracted
with CHCl3 (2 ꢀ 50 mL). The combined chloroformed layers
were dried (Na2SO4) and concentrated in vacuo to yield the
title compound as a pale-brown powder (50 mg, quant); mp
206-210 °C. 1H NMR (300 MHz, CDCl3) δ 2.11 (s, 3H), 3.68
(s, 2H), 3.69 (s, 3H), 3.95 (d, 1H, J = 14.3 Hz), 5.88 (d, 1H, J =
14.7 Hz), 6.65 (m, 2H), 6.76-7.08 (m, 6H), 7.08-7.43 (m, 3H).
13C NMR (50 MHz, CDCl3) δ 18.6, 34.8, 43.3, 43.5, 101.1, 119.1,
122.1, 125.7, 128.3, 123.0, 130.2, 130.6, 133.0, 134.1, 134.9, 136.2,
138.1, 139.6, 142.5, 169.5. IR (KBr) ν (cm-1) 3361, 3286, 3193,
3070, 2941, 2868, 1629, 1608, 1561, 1504, 1447, 1413, 1387, 1380,
1317, 1298, 1252, 1176, 1138, 993, 838, 817, 756. MS (ES) m/z
348.25 ([M þ H]þ), 695.47 ([2M þ H]þ).
dimethylamine hydrochloride (50 mg, 0.609 mmol) as des-
cribed for compound 1 and purified by flash chromatography
(98:1:1 to 96:3:1 CHCl3/MeOH/NH4OH); white foam (189 mg,
82%); mp 127-130 °C. 1H NMR (300 MHz, DMSO-d6) δ 2.11
(s, 3H), 2.79 (s, 6H), 3.77 (s, 3H), 3.88 (d, 1H, J = 14.7 Hz), 4.08
(app d, 2H, J = 5.0 Hz), 5.67 (d, 1H, J = 14.4 Hz), 6.60-6.77
(m, 3H), 6.90 (AB system, 2H, J = 8.4, 7.8 Hz), 6.99 (s, 1H),
7.11 (t, 1H, J = 7.5 Hz), 7.16 (s, 1H), 7.28 (d, 1H, J = 8.1 Hz),
8.59 (bs, 1H). 13C NMR (75 MHz, DMSO-d6) δ 18.4, 35.3,
35.9, 41.0, 43.2, 100.3, 119.5, 121.3, 124.7, 125.3, 128.2, 129.1,
130.3, 132.6, 134.1, 134.3, 135.6, 138.9, 139.8, 140.4, 158.0,
168.1. IR (KBr) ν (cm-1) 3389, 3125, 3031, 2941, 2863, 2823,
1624, 1558, 1538, 1506, 1488, 1449, 1417, 1379, 1320, 1259,
1230, 1064, 839, 759, 730, 663, 618. HRMS (ES): calcd for
(2S)-N-{2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b]-
[1,5]benzodiazepin-5(1H)-yl)carbonyl]benzyl}-2-[(4-methyl-1,4-
diazepan-1-yl)carbonothioyl] pyrrolidine-1-carboxamide, Bis-
C23H27N6O2, 419.2190; found, 419.2186. RP-HPLC (gradient
1): tR 17.5 min, purity >95%.
N-{2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b][1,5]benzo-
diazepin-5(1H)-yl) carbonyl]benzyl}acetamide (12c). To a stirred
solution of 10 (177 mg, 0.509 mmol) and Et3N (106 μL, 0.763
mmol) in dry DCM (4 mL) was added Ac2O (53 μL, 0.560 mmol).
Stirring was continued at rt for 3 h under argon. The reaction
mixture was diluted with DCM (10 mL), and the organic layer was
separated, washed with 10% citric acid (5 mL) and satd NaHCO3
(5 mL). The aqueous phases were basified with 10 N NaOH and
extracted with CHCl3 (2 ꢀ 20 mL). The combined organic layers
were dried (Na2SO4) and concentrated in vacuo. Flash chromato-
graphy (95:4:1 to 90:8:2 AcOEt/MeOH/Et3N) afforded the title
compound as a white powder (75 mg, 38%); mp 218-220 °C (from
(trifluoroacetate) (1 2TFA). To a stirred solution of 10 (152 mg,
3
0.439 mmol) and DIEA (76 μL, 0.439 mmol) in dry DMF (3 mL)
was added CDI (98 mg, 0.606 mmol). The resulting reaction
mixture was stirred at rt for 1 h under argon and then treated
with a solution of 11 (91 mg, 0.399 mmol) and DIEA (76 μL,
0.439 mmol) in dry DMF (1 mL). Stirring was continued at rt
overnight. As TLC showed still presence of some starting amine
11, an additional amount of activated 10 was prepared (42 mg,
0.120 mmol) and added to the reaction mixture. Stirring was
continued at rt for 9 h, and then the solvent was removed in
vacuo. Flash chromatography (98:1:1 to 94:5:1 CHCl3/MeOH/
NH4OH) afforded an oil that was further purified by semipre-
parative HPLC and lyophilized to yield the title compound as a
1
EtOH/AcOEt). H NMR (300 MHz, DMSO-d6) δ 1.84 (s, 3H),
2.11 (s, 3H), 3.77 (s, 3H), 3.89 (d, 1H, J = 14.7 Hz), 4.12 (app d, 2H,
J= 2.8 Hz), 5.67 (d, 1H, J=14.3Hz), 6.65(t,1H,J= 7.3 Hz), 6.72
(d, 1H, J = 7.2 Hz), 6.89 (AB system, 2H, J = 7.8, 7.8 Hz), 7.04 (s,
1H), 7.11 (t, 1H, J = 7.6 Hz), 7.17 (s, 1H), 7.29 (d, 1H, J = 8.1 Hz),
8.16 (app t, 1H), 8.61 (bs, 1H). 13C NMR (75 MHz, DMSO-d6) δ
18.4, 22.4, 35.3, 43.1, 100.3, 119.5, 121.3, 124.8, 126.1, 128.2, 129.3,
130.3, 132.5, 134.7, 134.9, 135.6, 138.6, 138.9, 139.8, 167.9, 169.1. IR
(KBr) ν(cm-1) 3311, 3294, 3081, 2964, 2919, 1622, 1610, 1564, 1507,
1449, 1420, 1385, 1299, 1262, 1250, 1175, 1025, 996, 836, 816, 768,
750, 730. HRMS (ES): calcd for C22H24N5O2, 390.1925; found,
390.1922. RP-HPLC (gradient 1): tR 17.0 min, purity >98%.
Methyl 2-Chloro-4-nitrobenzoate (17). A solution of 2-chloro-
4-nitrobenzoyle chloride 16 (1.020 g, 4.64 mmol) in dry MeOH
(5 mL) was stirred at 0 °C for 5 min and then allowed to warm to
rt and stirring was continued for 2 h. The solvent was removed in
vacuo and the residue dried under high vacuum to yield the title
compound as a white powder (1.086 g, quant); mp 62-64 °C. 1H
NMR (200 MHz, CDCl3) δ 3.99 (s, 3H), 7.97 (d, 1H, J =
8.3 Hz), 8.16 (dd, 1H, J = 8.6, 2.2 Hz), 8.31 (s, 1H). 13C NMR
(75 MHz, CDCl3) δ 53.1, 121.4, 126.0, 132.0, 134.8, 135.7, 149.4,
164.6. IR (KBr) ν (cm-1) 3098, 3031, 2964, 1721, 1589, 1527,
1430, 1387, 1358, 1299, 1245, 1189, 1129, 1043, 945, 890, 853,
825, 780, 758, 742.
22
white crystalline powder (193 mg, 58%); mp 160-164 °C. [R]D
þ72 (c = 1, MeOH). 1H NMR (300 MHz, DMSO-d6) δ
1.70-2.00 (m, 2H), 2.00-2.37 (m, including s at 2.09 ppm,
6H), 2.67-2.90 (m, 2H), 3.22 (m, 1H), 3.30-3.44 (m, 4H), 3.78
(s, 3H), 3.88 (d, 1H, J = 14.7 Hz), 4.08 (m, 2H), 4.38-5.10 (m,
8H), 5.67 (d, 1H, J = 14.7 Hz), 6.60-6.80 (m, 2H), 6.80-7.07
(m, 3H), 7.11 (t, 1H, J = 7.0 Hz), 7.20 (s, 1H), 7.29 (d, 1H, J =
7.9 Hz), 8.67 (bs, 1H). 13C NMR (50 MHz, CD3OD) δ 19.7,
26.2, 27.2, 33.2, 36.0, 43.5, 45.3, 46.0, 47.0, 51.7, 52.5, 53.8, 54.0,
54.5, 56.1, 56.7, 57.2, 65.2, 104.7, 122.2, 124.9, 127.4 (m), 130.9,
131.3, 132.4, 135.5, 136.1, 136.3, 137.1, 140.1, 142.3, 144.6,
144.8, 159.5, 172.5, 209.8, 210.0. IR (KBr) ν (cm-1) 3367,
3210, 3092, 2958, 2874, 2672, 1683, 1645, 1635, 1564, 1558,
1502, 1454, 1417, 1393, 1382, 1317, 1202, 1183, 1136, 834, 797,
761, 721, 705. HRMS (ES): calcd for C32H41N8O2S, 601.3068;
found, 601.3071. RP-HPLC (gradient 1): tR 20.3 min, purity
>98%. Anal. (C32H40N8O2S.2C2HF3O2 2H2O) C, H, N: calcd,
50.00, 5.36, 12.96; found 49.90, 5.09, 12.99.
3
N-{2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b][1,5]ben-
zodiazepin-5(1H)-yl) carbonyl]benzyl}pyrrolidine-1-carboxamide
(12a). Prepared from compound 10 (193 mg, 0.555 mmol) and
pyrrolidine (51 μL, 0.610 mmol) as described for compound 1 and
purified by flash chromatography (96:3:1 CHCl3/MeOH/Et3N);
white powder (189 mg, 77%); mp 253-255 °C (from EtOH). 1H
NMR (300 MHz, DMSO-d6) δ 1.78 (app t, 4H, J = 6.2 Hz), 2.11
(s, 3H), 3.21 (app t, 4H, J = 6.2 Hz), 3.77 (s, 3H), 3.88 (d, 1H, J =
14.3 Hz), 4.09 (app d, 2H, J = 5.0 Hz), 5.68 (d, 1H, J = 14.3 Hz),
6.48 (bt, 1H, J = 5.6 Hz), 6.65 (t, 1H, J = 7.3 Hz), 6.72 (d, 1H,
J = 7.2 Hz), 6.92 (AB system, 2H, J = 7.8, 7.8 Hz), 7.00 (s, 1H),
7.11 (t, 1H, J = 7.6 Hz), 7.16 (s, 1H), 7.28 (d, 1H, J = 7.8 Hz),
8.59 (bs, 1H). 13C NMR (75 MHz, DMSO-d6) δ 18.4, 25.0, 35.3,
40.7, 43.2, 45.3, 100.3, 119.5, 121.3, 124.7, 125.4, 128.2, 129.0,
130.3, 132.6, 134.1, 134.2, 135.6, 138.9, 139.8, 140.4, 156.4, 168.1.
IR (KBr) ν (cm-1) 3349, 3266, 3065, 2971, 2870, 1624, 1560, 1541,
1505, 1456, 1417, 1396, 1379, 1354, 1315, 1300, 1257, 1175, 990,
840, 815, 763, 730. HRMS (ES): calcd for C25H29N6O2, 445.2347;
found, 445.2344. RP-HPLC (gradient 1): tR 18.4 min, purity >98%.
N,N-Dimethyl-N0-{2-methyl-4-[(1-methyl-4,10-dihydropyrazolo-
[3,4-b][1,5]benzodiazepin-5(1H)-yl)carbonyl]benzyl}urea (12b).
Prepared from compound 10 (192 mg, 0.554 mmol) and
Methyl 4-Amino-2-chlorobenzoate (18). A mixture of 17 (974
mg, 4.52 mmol) and SnCl2 2H2O (5.10 g, 22.60 mmol) in EtOH
3
(50 mL) was refluxed under argon. After 1 h, the reaction
mixture was cooled to 0 °C and 1 N NaHCO3 was added slowly
until pH 7-8 (65 mL). The resulting milky mixture was stirred at
rt for 1 h, and then the solvent was removed in vacuo and the
resulting aqueous phase was extracted with AcOEt (200 mL).
The organic layer was washed with brine (50 mL), and the
aqueous phase was extracted with DCM (200 mL). The com-
bined organic layers were dried (Na2SO4) and concentrated in
vacuo to yield the title compound as an orange crystalline solid
(928 mg, quant); mp 100-102 °C. 1H NMR (200 MHz, CDCl3)
δ 3.85 (s, 3H), 3.96 (bs, 2H), 6.52 (dd, 1H, J = 8.5, 2.2 Hz), 6.69
(d, 1H, J = 2.2 Hz), 7.76 (d, 1H, J = 8.6 Hz). 13C NMR
(75 MHz, CDCl3) δ 51.8, 112.3, 116.3, 118.0, 133.8, 136.1, 150.5,
165.7. IR (KBr) ν (cm-1) 3427, 3333, 3003, 2952, 1703, 1633,
1598, 1550, 1499, 1429, 1329, 1264, 1248, 1186, 1122, 1040, 906,
862, 850, 831, 773, 685.

1556 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
Methyl 2-Chloro-4-cyanobenzoate (19). A suspension of 18
(510 mg, 2.75 mmol) in conc HCl (5.5 mL) was cooled to -15 °C
and ice (30 mL) was added. Then a solution of NaNO2 (235 mg,
3.30 mmol) in water (5 mL) was added dropwise under argon.
The resulting reaction mixture was stirred at -15 °C for 15 min.
The presence of HNO2 excess was checked with a starch-iodide
paper. The mixture was then carefully neutralized by adding
solid Na2CO3 with constant stirring (pH 5). At -10 °C, a
mixture of KCN (497 mg, 6.87 mmol) and CuCN (320 mg,
3.57 mmol) in cold water (8 mL) was added dropwise. Stirring
was continued for 2 h from -10 °C to rt. The reaction mixture
was then filtered and extracted with DCM (2 ꢀ 75 mL). The
combined organic layers were dried (Na2SO4) and concentrated
in vacuo. Flash chromatography (95:5 n-heptane/AcOEt) afforded
the title compound as a white woolly solid (328 mg, 61%); mp
76-80 °C. 1H NMR (200 MHz, CDCl3) δ 3.97 (s, 3H), 7.61 (dd,
1H, J = 8.1, 1.5 Hz), 7.75 (d, 1H, J = 1.5 Hz), 7.90 (d, 1H, J =
8.1 Hz). 13C NMR (50 MHz, CDCl3) δ 53.0, 116.2, 116.6, 130.0,
131.8, 132.2, 134.3, 134.6, 164.7. IR (KBr) ν (cm-1) 3070, 3048,
2969, 2235, 1722, 1435, 1387, 1306, 1295, 1259, 1194, 1136, 1054,
954, 895, 845, 805, 775, 607.
3-Chloro-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-10(11H)-yl-
carbonyl)benzonitrile (21c). The acyl chloride of the acid 20
(25 mg, 0.138 mmol) was prepared according to the procedure
described for compound 9. It was then dissolved in dry DCM
(0.3 mL), and this mixture was added dropwise, under argon, to
a stirred ice-cold solution of 14 (21 mg, 0.115 mmol), anhydrous
pyridine (28 μL, 0.345 mmol), and DMAP (0.4 mg, 0.003 mmol)
in dry DCM (0.3 mL). The reaction mixture was irradiated
under microwaves at 80 °C (1 bar) for 6 min and then diluted in
DCM (15 mL), washed with 1 N HCl (3 mL) and satd NaHCO3
(3 mL), dried (Na2SO4), and concentrated in vacuo. Flash
chromatography (8:2 to 7:3 n-heptane/AcOEt) afforded the title
compound as a white powder (20 mg, 49%); mp 104-106 °C. 1H
NMR (300 MHz, DMSO-d6) δ 4.90-5.50 (m, 4H), 5.90 (t, 1H,
J = 3.0Hz), 6.01 (m, 1H), 6.83(app t, 1H, J = 1.9Hz), 7.00-7.10
(m, 2H), 7.10-7.18 (m, 1H), 7.40 (d, 1H, J = 7.2 Hz), 7.45-7.65
(m, 1H), 7.69 (d, 1H, J = 7.5 Hz), 7.99 (s, 1H). 13C NMR
(75 MHz, DMSO-d6) δ 45.1, 49.2, 106.8, 108.7, 113.0, 116.9,
122.3, 124.6, 125.0, 128.1, 128.6, 128.8, 129.6, 130.8, 132.9, 135.7,
139.1, 140.6, 165.3. IR (KBr) ν (cm-1) 3098, 3064, 2952, 2919,
2852, 2230, 1653, 1600, 1580, 1493, 1390, 1371, 1351, 1303, 1273,
1194, 834, 758, 714, 686, 612.
2-Chloro-4-cyanobenzoic Acid (20). A solution of 19 (967 mg,
4.94 mmol) and LiOH H2O (318 mg, 7.41 mmol) in a 1:1
2-Methyl-4-{[4-(2-oxo-2H-3,1-benzoxazin-1(4H)-yl)piperidin-
1-yl]carbonyl} Benzonitrile (21d). To a solution of 15 (288 mg,
1.62 mmol) in dry DMF (6 mL) were added successively
EDCI HCl (381 mg, 1.95 mmol), HOBt H2O (269 mg, 1.95
3
MeOH/THF mixture (10 mL) was stirred at rt for 1 h. The
solvent was removed in vacuo, and then the resulting residue
was taken up with water (10 mL) and acidified with 6 N HCl
(pH 1). The precipitate was filtered and dried under vacuum
over P2O5 and then at 45 °C for 8 h to yield the title compound
as a pale-pink powder (744 mg, 83%); mp 160-162 °C. 1H
NMR (300 MHz, DMSO-d6) δ 7.90 (m, 2H), 8.16 (s, 1H), 14.01
(bs, 1H). 13C NMR (50 MHz, DMSO-d6) δ 114.6, 116.8, 130.9,
131.1, 131.8, 133.8, 136.3, 165.7. IR (KBr) ν (cm-1) 3300-2400,
3076, 3048, 2235, 1717, 1687, 1600, 1544, 1485, 1429, 1408, 1383,
1302, 1261, 1054, 903, 884, 848, 778, 683. MS (ES) m/z 179.97
([M - H]-).
3
3
mmol), 8 (377 mg, 1.78 mmol), and then DIEA (0.56 mL, 3.25
mmol) dropwise. Stirring was continued at rt for 14 h, and then
the solvent was removed in vacuo. The resulting residue was
taken up in AcOEt (50 mL). This organic layer was washed with
10% citric acid (20 mL), satd NaHCO3 (20 mL) and brine
(20 mL), dried (Na2SO4), and concentrated in vacuo. Flash
chromatography (5:5 to 3:7 n-heptane/AcOEt) afforded the title
1
compound as a white foam (394 mg, 65%); mp 64-68 °C. H
NMR (300 MHz, CDCl3) δ 1.75-2.10 (m, 2H), 2.58 (s, 3H),
2.60-3.00 (m, 3H), 3.13 (m, 1H), 3.80 (m, 1H), 4.00-4.20 (m,
1H), 4.92 (m, 1H), 5.10 (s, 2H), 7.00-7.24 (m, 3H), 7.30-7.50
(m, 3H), 7.65 (d, 1H, J = 7.9 Hz). 13C NMR (50 MHz, CDCl3) δ
20.5, 28.0, 28.8, 42.1, 47.3, 56.4, 67.1, 113.3, 113.9, 117.4, 122.4,
123.4, 124.5, 125.0, 128.6, 129.2, 132.7, 138.6, 140.1, 142.7,
152.5, 168.5. IR (KBr) ν (cm-1) 3048, 2975, 2934, 2870, 2226,
1716, 1633, 1607, 1498, 1468, 1454, 1447, 1389, 1293, 1273, 1258,
1242, 1204, 1042, 771, 755.
2-Methyl-4-(2,3,4,5-tetrahydro-1H-1-benzazepin-1-ylcarbonyl)-
benzonitrile (21a). The acyl chloride of the acid 8 (301 mg, 1.87
mmol) was prepared according to the procedure described for
compound 9. It was then dissolved in dry DCM (2 mL), and this
mixture was added dropwise, under argon, to a stirred ice-cold
solution of 13 (250 mg, 1.70 mmol), anhydrous pyridine (0.41 mL,
5.10 mmol), and DMAP (6.2 mg, 0.05 mmol) in dry DCM (4 mL).
Stirring was continued at 0 °C for 30 min then at rt overnight. The
reaction mixture was diluted in DCM (40 mL), washed with 1 N
HCl (5 mL) and satd NaHCO3 (5 mL), dried (Na2SO4), and
concentrated in vacuo. Flash chromatography (8:2 n-heptane/
AcOEt) afforded the title compound as a white powder (375 mg,
76%); mp 126-128 °C. 1H NMR (300 MHz, CDCl3) δ 1.40-1.70
(m, 1H), 1.90-2.20 (m, 3H), 2.41 (s, 3H), 2.70-3.10 (m, 3H), 4.96
(app bd, 1H, J = 13.1 Hz), 6.59 (d, 1H, J = 7.5 Hz), 6.87-6.97
(m, 2H), 7.10 (dt, 1H, J = 7.5, 0.9 Hz), 7.19-7.27 (m, 2H), 7.31 (d,
1H, J = 7.8 Hz). 13C NMR (75 MHz, CDCl3) δ 20.3, 26.3, 29.5,
34.9, 47.8, 113.4, 117.5, 125.4, 127.1, 127.7, 128.2, 129.8, 130.2,
131.7, 139.3, 140.5, 141.7, 143.1, 167.4. IR (KBr) ν (cm-1) 3059,
2930, 2851, 2222, 1644, 1603, 1578, 1558, 1492, 1438, 1413, 1384,
1359, 1317, 1274, 895, 851, 828, 775, 759, 741.
1-[2-Methyl-4-(2,3,4,5-tetrahydro-1H-1-benzazepin-1-ylcarbonyl)-
phenyl] Methanamine (22a). Prepared from compound 21a (367 mg,
1.26 mmol) as described for compound 10; cream-colored powder
1
(329 mg, 89%); mp 96-98 °C. H NMR (300 MHz, CDCl3) δ
1.30-2.20 (m, 4H), 2.16 (s, 3H), 2.50-3.10 (m, 3H), 3.73 (s, 2H),
5.00 (app bd, 1H, J = 12.8 Hz), 6.62 (d, 1H, J = 7.8 Hz), 6.75-6.92
(m, 2H), 6.95-7.08 (m, 2H), 7.11 (s, 1H), 7.20 (d, 1H, J = 7.2 Hz).
13C NMR (75 MHz, CDCl3) δ18.6, 26.3, 29.6, 34.9, 43.5, 47.6, 125.8,
126.9, 127.0, 128.2, 129.8, 130.2, 134.5, 135.0, 139.2, 142.1, 142.2,
144.1, 168.9. IR (KBr) ν (cm-1) 3379, 3316, 3064, 3045, 2942, 2924,
2879, 2840, 1633, 1578, 1564, 1492, 1456, 1435, 1411, 1377, 1356,
1312, 1276, 1259, 1209, 1166, 1052, 1033, 967, 889, 862, 819, 761, 755.
1-[2-Methyl-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-10(11H)-
ylcarbonyl)phenyl] Methanamine (22b). Prepared from com-
pound 21b (259 mg, 0.790 mmol) as described for compound
2-Methyl-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-10(11H)-yl-
carbonyl)benzonitrile (21b). Prepared from acid 8 (304 mg, 1.79
mmol) and amine 14 (300 mg, 1.63 mmol) as described for
compound 21a (without DMAP); pale-yellow powder (345 mg,
1
10; pale-brown foam (32 mg, 12%); mp 62-66 °C. H NMR
(300 MHz, CDCl3) δ 2.22 (s, 3H), 3.80 (s, 2H), 5.20 (app bs, 4H),
6.04 (s, 1H), 6.09 (t, 1H, J = 3.1 Hz), 6.69 (s, 1H), 6.87 (m, 1H),
6.99 (m, 1H), 7.04-7.21 (m, 3H), 7.24 (m, 1H), 7.36 (d, 1H, J =
7.2 Hz). 13C NMR (75 MHz, CDCl3) δ 18.7, 43.5, 46.5, 51.3,
107.5, 108.1, 121.5, 126.0, 126.5, 127.4, 128.7, 128.9, 129.3,
1
65%); mp 186-188 °C. H NMR (300 MHz, CDCl3) δ 2.46
(s, 3H), 5.19 (app bs, 4H), 6.06 (m, 1H), 6.10 (t, 1H, J = 3.0 Hz),
6.71 (m, 1H), 6.78 (m, 1H), 6.93-7.16 (m, 2H), 7.20 (t, 1H, J =
7.3 Hz), 7.32-7.50 (m, 3H). 13C NMR (75 MHz, CDCl3) δ 20.4,
46.4, 51.2, 107.8, 108.7, 114.1, 117.3, 121.9, 125.6, 125.8, 128.1,
129.0, 129.5, 130.1, 131.9, 134.1, 139.4, 141.7, 142.4, 168.5. IR
(ATR) ν (cm-1) 2921, 2854, 2229, 1644, 1584, 1492, 1462, 1447,
1407, 1368, 1301, 1270, 1246, 1214, 1199, 1076, 890, 853, 841,
827, 757, 728, 701, 682, 612. MS (ES) m/z 328.1 ([M þ H]þ),
350.1 ([M þ Na]þ).
130.4, 133.5, 134.0, 135.4, 142.9, 170.2. IR (ATR) ν (cm-1
)
3369, 3306, 2920, 2851, 1650, 1640, 1634, 1583, 1574, 1568, 1493,
1454, 1446, 1408, 1366, 1348, 1303, 1271, 1188, 1148, 1075, 825,
755, 745, 710, 611. MS (ES) m/z 332.2 ([M þ H]þ).
1-[3-Chloro-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-10(11H)-
ylcarbonyl)phenyl] Methanamine (22c). Prepared from compound

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1557
21c (207 mg, 0.594 mmol) as described for compound 10; pale-
yellow gum (64 mg, 30%). 1H NMR (300 MHz, CDCl3) δ
3.70-4.05 (m, 2H), 5.00-5.40 (m, 4H), 5.95-6.15 (m, 2H),
(c = 1, CHCl3). 1H NMR (200 MHz, CDCl3) δ 1.80-3.10 (m,
14H, including d at 2.44 ppm, J = 8.4 Hz), 3.30-3.53 (m, 1H),
3.58-4.20 (m, 4H), 4.22-4.53 (m, 2H), 4.53-4.87 (m, 2H),
4.87-5.40 (m, 4H), 6.06 (m, 2H), 6.66 (m, 1H), 6.94-7.7.58
(m, 7H). 13C NMR (50 MHz, CDCl3) δ 22.6, 22.9, 24.1, 24.2,
25.0, 26.4, 27.9, 29.6, 31.8, 43.4, 44.0, 45.6, 46.0, 46.5, 50.4,
50.7, 51.1, 51.3, 52.8, 53.9, 55.4, 55.7, 56.8, 57.5, 62.7, 62.9,
107.6, 108.8, 121.7, 125.2, 125.8, 126.1, 126.5, 126.7, 127.6,
128.1, 128.3, 128.8, 129.0, 131.1, 134.0, 134.5, 134.7, 140.5,
142.6, 156.1, 167.7, 205.1, 205.9. IR (KBr) ν (cm-1) 3400,
2930, 2863, 2796, 1641, 1531, 1494, 1458, 1382, 1200, 758, 716.
HRMS (ES): calcd for C32H38ClN6O2S, 605.2460; found,
605.2443. RP-HPLC (gradient 1): tR 21.2 min, purity >98%.
(2S)-N-{2-Methyl-4-[(4-(2-oxo-2H-3,1-benzoxazin-1(4H)-yl)-
piperidin-1-yl)carbonyl] benzyl}-2-[(4-methyl-1,4-diazepan-1-yl)-
carbonothioyl]pyrrolidine-1-carboxamide (23d). Prepared from
compound 22d (150 mg, 0.396 mmol) and compound 11 (75 mg,
0.330 mmol) as described for compound 1 and purified by flash
chromatography (98:1:1 to 97:2:1 CHCl3/MeOH/NH4OH); white
6.60-6.70 (m, 1H), 6.85-7.18 (m, 4H), 7.18-7.60 (m, 3H). 13
C
NMR (75 MHz, CDCl3) δ 45.6, 51.1, 51.3, 107.6, 108.9, 121.8,
125.7, 125.8, 127.7, 127.9, 128.0, 128.2, 128.4, 128.6, 128.9, 129.0,
131.1, 134.6, 134.8, 140.4, 148.8, 167.8. IR (CsI) ν (cm-1) 3383,
3042, 2919, 2852, 1652, 1606, 1495, 1398, 1306, 1274, 758, 718.
1-{2-Methyl-4-[(4-(2-oxo-2H-3,1-benzoxazin-1(4H)-yl)piperidin-
1-yl)carbonyl]phenyl}methanamine (22d).Prepared from compound
21d (338 mg, 0.90 mmol) as described for compound 10; beige
1
powder (188 mg, 55%); mp 128-130 °C. H NMR (200 MHz,
DMSO-d6) δ1.50-2.00 (m, 2H), 2.00-2.35 (m, 3H), 2.55-4.80 (m,
9H), 4.80-5.20 (m, 2H), 6.80-7.50 (m, 7H). 13C NMR (50 MHz,
DMSO-d6) δ 18.5, 28.4, 46.6, 55.3, 66.3, 114.0, 122.6, 123.0, 124.1,
124.8, 128.2, 128.4, 129.1, 135.5, 136.7, 138.6, 151.8, 168.8. IR (KBr)
ν (cm-1) 2963, 2934, 2863, 1719, 1709, 1607, 1564, 1499, 1467, 1446,
1390, 1303, 1292, 1259, 1206, 1111, 1041, 769, 755, 620. MS (ES) m/z
380.24 ([M þ H]þ).
22
(2S)-N-[2-Methyl-4-(2,3,4,5-tetrahydro-1H-1-benzazepin-1-
ylcarbonyl)benzyl]-2-[(4-methyl-1,4-diazepan-1-yl)carbonothioyl]-
pyrrolidine-1-carboxamide (23a). Prepared from compound 22a
(159 mg, 0.539 mmol) and compound 11 (88 mg, 0.385 mmol) as
described for compound 1 and purified by flash chromatography
(CHCl3/Et3N 0.7 to 2%). White foam (199 mg, 95%); mp
foam (153 mg, 74%); mp 136-140 °C; [R]D þ47 (c = 1, CHCl3).
1H NMR (200 MHz, CDCl3) δ 0.60-5.20 (m, including s at
2.32 ppm and s at 5.09 ppm, 37H), 6.80-7.70 (m, 7H). 13C NMR
(75 MHz, CDCl3) δ 19.1, 24.4, 28.9, 31.8, 42.2, 45.5, 45.8, 45.9,
46.3, 50.0, 51.5, 53.3, 55.0, 55.4, 56.6, 62.6, 67.1, 113.5, 121.4,
122.4, 123.3, 124.4, 124.9, 127.3, 128.7, 129.2, 134.6, 134.8, 136.4,
138.6, 138.9, 152.6, 156.2, 170.5. IR (KBr) ν (cm-1) 3337, 2947,
2937, 2870, 2801, 1717, 1625, 1608, 1541, 1498, 1467, 1457, 1388,
1348, 1291, 1258, 1204, 1133, 1041, 769, 753. HRMS (ES): calcd
for C34H45N6O4S, 633.3218; found, 633.3190. RP-HPLC
(gradient 1): tR 20.1 min, purity >98%.
22
1
98-100 °C; [R]D -67 (c = 1, CHCl3). H NMR (300 MHz,
DMSO-d6) δ 1.30-1.50 (m, 1H), 1.60-2.22 (m, 9H), 2.25 (s, 3H),
2.35-2.78 (m, 4H), 2.80-3.10 (m, 3H), 3.16-3.50 (m, 4H),
3.50-3.64 (m, 1H), 3.72-4.30 (m, 6H), 4.83 (app bd, 1H, J =
13.1 Hz), 4.96 (m, 1H), 6.48 (bs, 1H), 6.68 (d, 1H, J = 7.5 Hz),
6.80 (app bt, 1H, J = 9.0 Hz), 6.87-7.13 (m, 4H), 7.28 (d, 1H,
J = 7.5 Hz). 13C NMR (75 MHz, DMSO-d6) δ 18.4, 23.6, 24.6,
25.9, 26.6, 29.3, 31.3, 34.2, 45.3, 45.5, 45.6, 46.1, 46.2, 47.0, 49.9,
50.8, 53.0, 53.2, 54.9, 55.4, 55.8, 57.0, 61.9, 62.0, 124.6, 124.8,
125.2, 125.4, 126.9, 127.0, 128.1, 129.0, 129.8, 134.1, 134.2, 139.1,
140.2, 143.9, 155.5, 167.9, 205.0, 205.2. IR (KBr) ν (cm-1) 3328,
2939, 2863, 2801, 1634, 1540, 1505, 1492, 1456, 1438, 1415, 1393,
1379, 1354, 1314, 1202, 1138, 1032, 755, 744. HRMS (ES): calcd
for C31H42N5O2S, 548.3054; found, 548.3054. RP-HPLC
(gradient 1): tR 20.8 min, purity >98%.
1-Methyl-5-(3-methyl-4-nitrobenzoyl)-1,4,5,10-tetrahydro-
pyrazolo[3,4-b][1,5] Benzodiazepine (25). The acyl chloride of
3-methyl-4-nitrobenzoic acid 24 (271 mg, 1.50 mmol) was
prepared according to the procedure described for compound
9. It was then dissolved in dry DCM (2 mL), and this mixture
was added dropwise, under argon, to a stirred ice-cold solu-
tion of 7 (250 mg, 1.25 mmol) and Et3N (0.52 mL, 3.74 mmol)
in dry DCM (3 mL). Stirring was continued at 0 °C for 30 min
then at rt overnight. The reaction mixture was diluted in
DCM (25 mL), washed with 10% citric acid (10 mL) and satd
NaHCO3 (15 mL), dried (Na2SO4), and concentrated in
vacuo to yield a residue that was triturated in AcOEt. The
precipitate was filtered, washed with AcOEt, and dried
under high vacuum to yield the title compound as a yellow
powder (409 mg, 90%); mp 222-224 °C. 1H NMR (300 MHz,
DMSO-d6) δ 2.36 (s, 3H), 3.78 (s, 3H), 3.97 (d, 1H, J = 14.3 Hz),
5.65 (d, 1H, J = 14.7 Hz), 6.66 (td, 1H, J = 7.8, 1.2 Hz), 6.83 (dd,
1H, J = 7.8, 1.6 Hz), 7.07-7.17 (m, 2H), 7.20 (s, 1H), 7.29 (dd,
1H, J = 8.4, 1.3 Hz), 7.32 (s, 1H), 7.74 (d, 1H, J = 8.4 Hz), 8.67
(bs, 1H). 13C NMR (75MHz, DMSO-d6) δ 19.4, 35.5, 43.4, 100.2,
119.7, 121.5, 124.0, 125.8, 129.0, 130.6, 131.5, 131.6, 132.6, 135.9,
139.2, 139.8, 141.1, 148.9, 166.8. IR (KBr) ν (cm-1) 3343, 3098,
3076, 2986, 2930, 2858, 1632, 1605, 1565, 1524, 1503, 1414, 1382,
1356, 1318, 1293, 1252, 1180, 843, 769, 734.
(2S)-N-[2-Methyl-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-
10(11H)-ylcarbonyl) benzyl]-2-[(4-methyl-1,4-diazepan-1-yl)carbo-
nothioyl]pyrrolidine-1-carboxamide, Trifluoroacetate (23b TFA).
3
Prepared from compound 22b (32 mg, 0.096 mmol) and com-
pound 11 (24 mg, 0.106 mmol) as described for compound 1 and
purified by flash chromatography (98:1:1 to 94:5:1 CHCl3/
MeOH/NH4OH) followed by semipreparative HPLC; white
22
hygroscopic powder (29 mg, 51%); mp 140-142 °C; [R]D þ17
(c=0.1, MeOH). 1H NMR (300 MHz, CDCl3) δ 1.83-2.10 (m,
5H), 2.20 (s, 3H), 2.38 (d, 3H, J = 4.3 Hz), 2.47-2.92 (m, 4H),
3.34-3.48 (m, 1H), 3.58-3.72 (m, 1H), 3.75-4.22 (m, 4H),
4.22-4.55 (m, 4H), 5.03-5.30 (m, 4H), 6.03 (s, 1H), 6.09 (app t,
1H, J = 2.8 Hz), 6.69 (s, 1H), 6.88 (m, 1H), 7.01 (m, 1H),
7.06-7.23 (m, 4H), 7.36 (d, 1H, J = 7.2 Hz). 13C NMR
(75 MHz, CDCl3) δ 18.9, 24.0, 24.1, 27.1, 29.7, 31.7, 31.8, 42.2,
46.3, 46.4, 46.6, 50.6, 51.0, 51.3, 53.6, 54.0, 55.7, 55.9, 56.8, 57.8,
62.6, 62.9, 107.5, 108.0, 121.5, 121.8, 125.9, 126.5, 126.7, 127.6,
128.7, 128.9, 129.4, 130.4, 133.9, 135.9, 139.5, 142.8, 156.0, 170.1,
204.9, 205.6. IR (ATR) ν (cm-1) 3367, 2953, 2874, 2725, 2678,
1644, 1633, 1614, 1548, 1538, 1495, 1416, 1372, 1351, 1304, 1274,
1197, 1172, 1137, 796, 753, 717, 704, 593. HRMS (ES): calcd for
C33H41N6O2S, 585.3006; found, 585.3002. RP-HPLC (gradient 2):
2-Methyl-4-[(1-methyl-4,10-dihydropyrazolo[3,4-b][1,5]benzo-
diazepin-5(1H)-yl) carbonyl]aniline (26). A mixture of 25 (318 mg,
0.875 mmol) and SnCl2 2H2O (987 mg, 4.374 mmol) in EtOH
3
(11 mL) was refluxed under argon. After 1 h, the reaction mixture
was cooled to 0 °C and 1 N NaHCO3 was added slowly until pH
7-8 (7 mL). The resulting mixture was stirred at rt for 1.5 h and
then extracted with AcOEt (2 ꢀ 100 mL). The aqueous phase was
filtered over celite to eliminate the excess of salts and extracted
with AcOEt (100 mL). The combined organic layers were washed
with brine (30 mL), dried (Na2SO4), and concentrated in vacuo to
yield a residue that was triturated in AcOEt. The precipitate was
filtered, washed with AcOEt, and dried under high vacuum to
yield the title compound as a white powder (240 mg, 82%); mp
250-252 °C. 1H NMR (300 MHz, DMSO-d6) δ 1.88 (s, 3H), 3.76
(m, 3H), 3.75-3.90 (m, 1H), 5.17 (bs, 2H), 5.66 (app bd, 1H, J =
14.3 Hz), 6.25 (d, 1H, J = 8.1 Hz), 6.67 (m, 3H), 6.88 (s, 1H), 7.12
tR 18.8 min, purity >98%.
(2S)-N-[3-Chloro-4-(5H-pyrrolo[2,1-c][1,4]benzodiazepin-
10(11H)-ylcarbonyl) benzyl]-2-[(4-methyl-1,4-diazepan-1-yl)-
carbonothioyl]pyrrolidine-1-carboxamide (23c). Prepared
from compound 22c (55 mg, 0.157 mmol) and compound 11
(39 mg, 0.173 mmol) as described for compound 1 and
purified by flash chromatography (CHCl3/Et3N 1 to 2%);
22
cream powder (83 mg, 87%); mp 146-150 °C; [R]D -31

1558 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
(m,2H),7.29(d,1H,J=8.1Hz),8.53(bs,1H).13CNMR(75MHz,
DMSO-d6) δ 17.2, 35.3, 43.2, 100.6, 111.9, 119.3, 119.5, 121.3, 122.4,
127.2, 127.6, 130.2, 130.8, 133.6, 135.5, 138.6, 140.0, 148.3, 168.3. IR
(KBr) ν (cm-1) 3428, 3346, 3221, 3064, 2969, 2935, 1620, 1605, 1562,
1506, 1446, 1421, 1381, 1317, 1302, 1261, 1177, 1157, 824, 755, 732,
629. MS (ES) m/z 334.23 ([M þ H]þ).
1.9 Hz), 7.62 (s, 1H), 7.67 (d, 1H, J = 2.2 Hz), 7.98 (d, 1H, J =
8.4 Hz). 13C NMR (75 MHz, CDCl3) δ 12.8, 52.5, 108.4, 121.7,
126.3, 128.1, 132.2, 134.7, 139.0, 141.0, 143.0, 165.3. IR (KBr)
ν (cm-1) 3111, 3053, 2989, 2950, 2846, 1733, 1604, 1403, 1261,
1119, 1054, 504.
3-Methyl-4-(3-methyl-1H-pyrazol-1-yl)benzoic Acid (30a). A
solution of 29a (137 mg, 0.597 mmol) in a 1:1 6 N HCl/AcOH
mixture (2 mL) was refluxed for 6 h and then poured in an ice/
water mixture (30 mL) and extracted with CHCl3 (2 ꢀ 50 mL).
The combined organic layers were dried (Na2SO4) and concen-
trated in vacuo to yield the title compound as a white powder
(128 mg, 99%); mp 142-144 °C. 1H NMR (300 MHz, CDCl3)
δ 2.39 (s, 3H), 2.40 (s, 3H), 6.27 (d, 1H, J = 2.2 Hz), 7.46 (d, 1H,
J = 8.1 Hz), 7.57 (d, 1H, J = 2.2 Hz), 8.00 (dd, 1H, J = 8.3,
1.7 Hz), 8.06 (s, 1H), 10.53 (bs, 1H). 13C NMR (75 MHz, CDCl3)
δ 13.5, 18.7, 106.9, 125.8, 128.5, 128.6, 131.3, 132.9, 133.5, 143.8,
150.4, 170.8. IR (KBr) ν (cm-1) 3200-2400, 2986, 2929, 1678,
1610, 1579, 1540, 1508, 1433, 1402, 1367, 1317, 1304, 1284, 1221,
1068, 1041, 952, 926, 911, 834, 771, 754, 577.
5-[4-[(1,10-Biphenyl-2-ylcarbonyl)amino]-3-methylbenzoyl]-1-
methyl-1,4,5,10-tetrahydropyrazolo[3,4-b][1,5]benzodiazepine (27).
Prepared from 2-phenylbenzoic acid (445 mg, 2.25 mmol) and
amine 26 (150 mg, 0.45 mmol) as described for compound 25 and
purified by flash chromatography (95:5 AcOEt/MeOH) followed
by recrystallization in EtOH. Yellow crystalline powder (202 mg,
87%); mp 240-242 °C. 1H NMR (300 MHz, DMSO-d6) δ 1.79
(s, 3H), 3.77 (s, 3H), 3.89 (d, 1H, J = 14.7 Hz), 5.67 (d, 1H, J =
14.7 Hz), 6.60-6.80 (m, 2H), 6.87 (d, 1H, J = 7.5 Hz), 7.03 (s,
1H), 7.08-7.23 (m, 3H), 7.23-7.65 (m, 10H), 8.60 (s, 1H), 9.45 (s,
1H). 13C NMR (50 MHz, DMSO-d6) δ 17.2, 35.3, 43.2, 100.3,
119.6, 121.4, 123.4, 125.1, 127.2, 127.4, 127.9, 128.3, 128.5, 129.7,
129.9, 130.3, 132.5, 132.8, 135.6, 136.9, 137.1, 138.9, 139.1, 139.8,
140.1, 167.7, 167.9. IR (KBr) ν (cm-1) 3303, 3203, 3154, 3014,
2968, 2913, 2863, 1650, 1641, 1611, 1591, 1563, 1539, 1505, 1449,
1386, 1326, 1298, 1268, 1254, 1177, 1136, 1000, 893, 830, 761, 742,
699. HRMS (ES): calcd for C32H28N5O2, 514.2238; found,
514.2255. RP-HPLC (gradient 1): tR 22.3 min, purity >98%.
Anal. (C32H27N5O2) C, H, N: calcd, 74.84, 5.30, 13.64; found
74.60, 5.49, 13.55.
2-Chloro-4-(3-methyl-1H-pyrazol-1-yl)benzoic Acid (30b). A
solution of 29b (545 mg, 2.17 mmol) in a 1:1 6 N HCl/AcOH
mixture (7 mL) was refluxed for 6 h and then poured in an ice/
water mixture (50 mL). The precipitate was filtered, washed with
water, and dried in vacuo under P2O5 to yield the title compound
as a white powder (493 mg, 99%); mp 166-168 °C (sublimation).
1H NMR (300 MHz, DMSO-d6) δ 2.26 (s, 3H), 6.40 (d, 1H, J =
2.5 Hz), 7.80-8.00 (m, 3H), 8.53 (d, 1H, J = 2.2 Hz), 13.32 (bs,
1H). 13C NMR (75 MHz, DMSO-d6) δ 13.4, 109.0, 115.8, 119.0,
Methyl 3-Methyl-4-(3-methyl-1H-pyrazol-1-yl)benzoate (29a).
A solution of methyl 4-fluoro-3-methylbenzoate 28a (417 mg,
2.48 mmol), 3-methylpyrazole (0.30 mL, 3.72 mmol), and K2CO3
(686 mg, 4.96 mmol) in dry NMP (3 mL) was heated at 120 °C
under argon. As TLC showed incomplete reaction after 5 h, an
additional amount of 3-methylpyrazole (0.30 mL, 3.72 mmol)
was added and heating was continued for 20 h. The reaction
mixture was then poured into water (30 mL) and extracted with
AcOEt (2 ꢀ 150 mL). The combined organic layers were dried
(Na2SO4) and concentrated in vacuo. Flash chromatography
(n-heptane to 75:25 n-heptane/Et2O) afforded the title compound
as a colorless oil (Rf = 0.2, 49.2 mg, 9%). 1H NMR (200 MHz,
CDCl3) δ 2.37 (s, 3H), 2.38 (s, 3H), 3.93 (s, 3H), 6.25 (d, 1H, J =
2.4 Hz), 7.42 (d, 1H, J = 8.3 Hz), 7.55 (d, 1H, J = 2.2 Hz), 7.93
(m, 1H), 7.99 (m, 1H). 13C NMR (75 MHz, CDCl3) δ 13.5, 18.6,
52.2, 106.7, 125.7, 128.0, 129.2, 131.2, 132.8, 132.9, 143.4, 150.2,
166.4. IR (CsI) ν (cm-1) 2953, 2928, 2852, 1721, 1610, 1585, 1535,
1505, 1437, 1394, 1364, 1288, 1267, 1248, 1218, 1198, 1130, 1112,
1040, 1012, 980, 949, 843, 788, 770, 716. The isomer methyl
3-methyl-4-(5-methyl-1H-pyrazol-1-yl)benzoate was also iso-
lated as a yellow solid (Rf = 0.1, 3.7 mg, 1%); mp 58-60 °C.
1H NMR (200 MHz, CDCl3) δ 2.10 (s, 3H), 2.12 (s, 3H), 3.94
(s, 3H), 6.20 (s, 1H), 7.30 (d, 1H, J = 8.0 Hz), 7.60 (d, 1H, J =
1.6 Hz), 7.96 (m, 1H), 8.02 (m, 1H). 13C NMR (75 MHz, CDCl3)
δ 11.3, 17.3, 52.3, 105.7, 127.8, 130.7, 132.3, 136.6, 139.46, 139.9,
142.3, 142.4, 166.4. IR (CsI) ν (cm-1) 2954, 2924, 2852, 1726,
1613, 1586, 1544, 1509, 1438, 1411, 1391, 1297, 1257, 1201, 1132,
1106, 1018, 976, 923, 786, 773, 724.
Methyl 2-Chloro-4-(3-methyl-1H-pyrazol-1-yl)benzoate (29b).
Prepared from methyl 2-chloro-4-fluorobenzoate 28b (1.00 g,
5.30 mmol) and 3-methylpyrazole (0.51 mL, 6.36 mmol) as
described for compound 29a (6 h reflux); white solid (545 mg,
41%); Rf = 0.5 (7:3 n-heptane/AcOEt); mp 64-65 °C. 1H NMR
(200 MHz, CDCl3) δ 2.38 (s, 3H), 3.94 (s, 3H), 6.31 (d, 1H, J =
2.4 Hz), 7.60 (dd, 1H, J = 8.6, 2.2 Hz), 7.82 (d, 1H, J = 2.2 Hz),
7.86 (d, 1H, J = 2.4 Hz), 7.97 (d, 1H, J = 8.6 Hz). 13C NMR
(50 MHz, CDCl3) δ 13.7, 52.4, 109.0, 115.6, 120.4, 126.1, 127.5,
133.0, 135.6, 142.8, 152.0, 165.3. IR (KBr) ν (cm-1) 3153, 2987,
2947, 2929, 2848, 1730, 1604, 1572, 1541, 1505, 1439, 1395, 1378,
1359, 1310, 1264, 1251, 1226, 1194, 1124, 1088, 1044, 965, 951,
834, 769, 750. Isomer methyl 2-chloro-4-(5-methyl-1H-pyrazol-1-
yl)benzoate; cream-colored solid (93 mg, 7%); Rf = 0.4 (7:3
n-heptane/AcOEt); mp 80-81 °C. 1H NMR (300 MHz, CDCl3)
δ 2.44 (s, 3H), 3.97 (s, 3H), 6.24 (s, 1H), 7.48 (dd, 1H, J = 8.4,
126.9, 129.1, 132.8, 133.6, 142.1, 151.1, 165.8. IR (KBr) ν (cm-1
)
3146, 3081, 2952, 2800-2300, 1731, 1602, 1538, 1433, 1369, 1298,
1252, 1123, 1047, 965, 831.
1-Methyl-5-[3-methyl-4-(3-methyl-1H-pyrazol-1-yl)benzoyl]-
1,4,5,10-tetrahydropyrazolo[3,4-b][1,5]benzodiazepine (31a). To
a stirred solution of 30a (114 mg, 0.530 mmol) in dry DCM
(3 mL) at 0 °C under argon were added successively dry NMP
(3 drops) and SOCl2 (0.10 mL, 1.325 mmol). The resulting
reaction mixture was stirred at rt overnight. The solvent was
removed in vacuo, and the resulting residue was dissolved in
dry THF. The solvent was removed in vacuo, and the resulting
residue was dried under high vacuum for 1 h. Amide coupling
was then performed starting from the acyl chloride and amine 7
(116.7 mg, 0.583 mmol) as described for compound 21a
(stirring at rt for 3 days, 10% citric acid wash instead of 1 N
HCl, flash chromatography in 98:2 DCM/MeOH); cream-
colored foam (113 mg, 53%); mp 140-142 °C. 1H NMR (300
MHz, CDCl3) δ 2.15 (s, 3H), 2.32 (s, 3H), 3.84 (s, 3H), 3.98 (d,
1H, J = 14.6 Hz), 5.92 (d, 1H, J = 14.6 Hz), 6.17 (s, 1H), 6.72
(s, 2H), 6.91 (bs, 1H), 6.95-7.15 (m, 4H), 7.30 (s, 2H), 7.39 (s,
1H). 13C NMR (50 MHz, CDCl3) δ 13.5, 18.3, 35.0, 43.3, 101.7,
106.4, 119.4, 122.9, 124.9, 126.1, 128.7, 130.6, 131.2, 131.4,
132.6, 132.7, 135.1, 135.5, 137.7, 140.2, 140.8, 149.8, 168.6. IR
(KBr) ν (cm-1) 3289, 3137, 3064, 2958, 2927, 2872, 1634, 1606,
1561, 1532, 1504, 1447, 1397, 1380, 1322, 1295, 1251, 1178,
1136, 1040, 992, 951, 837, 815, 758, 733. HRMS (ES): calcd for
C23H23N6O, 399.1928; found, 399.1922. RP-HPLC (gradient 1,
flow rate: 1 mL/min): tR 19.6 min, purity >98%.
5-[2-Chloro-4-(3-methyl-1H-pyrazol-1-yl)benzoyl]-1-methyl-
1,4,5,10-tetrahydro pyrazolo[3,4-b][1,5]benzodiazepine (31b).
Prepared from acid 30b (280 mg, 1.18 mmol) and amine 7
(182 mg, 0.91 mmol) as described for compound 31a, but the
reaction mixture was irradiated under microwaves at 80 °C
(2 bar) for 8 min; beige powder (309 mg, 81%); two isomers
present in 3:1 proportions; mp 176-178 °C (from EtOH). 1H
NMR (300 MHz, CDCl3) δ 2.32 (s, 2.27H), 2.39 (s, 0.73H), 3.76
(s, 3H), 4.05 (d, 1H, J = 14.3 Hz), 5.87 (d, 1H, J = 14.4 Hz),
6.22 (d, 0.9H, J = 2.2 Hz), 6.32 (s, 1H), 6.68 (t, 1H, J = 7.5 Hz),
6.86 (d, 1H, J = 7.8 Hz), 6.90-7.20 (m, 3H), 7.20-7.45 (m,
2H), 7.58 (d, 0.9H, J = 1.9 Hz), 7.70 (d, 0.9H, J = 2.2 Hz), 7.83
(d, 0.2H, J = 2.2 Hz), 7.91 (app bs, 0.2H). 13C NMR (75 MHz,
CDCl3) δ 13.65, 34.80, 42.98, 100.96, 108.4, 115.7, 119.1, 119.2,

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1559
122.3, 127.4, 128.5, 129.0, 129.9, 132.2, 133.4, 136.4, 138.1, 139.1,
140.6, 151.4, 166.6. IR (ATR) ν (cm-1) 3304, 2929, 1627, 1603,
1563, 1558, 1537, 1504, 1448, 1410, 1398, 1360, 1323, 1301, 1254,
1246, 1229, 1151, 1049, 956, 947, 838, 827, 788, 767, 756, 731, 718,
702, 522, 440. HRMS (ES): calcd for C22H20ClN6O, 419.1382;
found, 419.1376. RP-HPLC (gradient 2): tR 18.7 min, purity
97.2% (254 nm).
Biology: Material and Methods. AVP and OT were purchased
from Bachem or Novabiochem, [3H]AVP from Amersham or
PerkinElmer Life and Analytical Sciences, BSA from Sigma and
Ro-20-1724 (4-(3-butoxy-4-methoxyphenyl)methyl-2-imidazo-
lidone) from Calbiochem. Synthesized compounds were initially
dissolved in DMSO at 10 or 1 μM and then diluted to the desired
concentration with the assay buffer.
Cell Culture. The CHO cell lines, which stably express the
human vasopressin V1a, V2, or oxytocin receptors, were maintained
in culture in Dulbecco’s modified Eagle’s medium (D-MEM)
supplemented with 10% fetal calf serum, one dose of nonessential
aminoacids, 100 units/mL penicillin, and 100 μg/mL streptomycin
in an atmosphere of 95% air and 5% CO2 at 37 °C.
Membrane Preparations. To increase the level of expression of
some receptors (OT-R, V1a-R), the cells were treated overnight
with 5 mM sodium butyrate.35,55 As already published, this
treatment does not modify the pharmacological properties of
the receptors. The membranes were prepared as already
described.34-36,50,55 Briefly, culture dishes of CHO cells expres-
sing the human vasopressin V1a, V2, or oxytocin receptors were
washed twice in phosphate-buffered saline without Ca2þ and
Mg2þ, and cold lysis buffer (15 mM Tris-HCl, 2 mM MgCl2,
0.3 mM EDTA, pH 7.4) was added. Cells were scraped with a
rubber, Polytron-homogenized, and centrifuged at 800g for
7 min at 4 °C. Supernatants were recovered and centrifuged at
44000g for 25 min at 4 °C. Pellets were resuspended in a
suspension medium (50 mM Tris-HCl, 5 mM MgCl2, pH 7.4)
and centrifuged at 44000g for 25 min at 4 °C. Pellets were
resuspended in an appropriate volume of the same buffer, and
the protein content was determined by the Bradford method
(Bio-Rad) using BSA as the standard. Aliquots of membranes
were used immediately for binding assays or stored at -80 °C.
Radioligand Binding Assays. Binding assays were performed
at 30 °C using [3H]AVP as the radioligand and CHO cells-
membrane proteins (10-15 μg for V1a and OT receptors, 5 μg
for V2 receptor), as previously described.34-36,50,55 Briefly,
membranes of CHO cells stably expressing human vasopressin
V1a, V2, or oxytocin receptors were incubated for 30 min at 30 °C
in the binding buffer (50 mM Tris-HCl, 5 mM MgCl2, 1 mg/mL
BSA, pH 7.4) with the radiolabeled and displacing ligands. The
affinities (Kd) of [3H]AVP for the human vasopressin V1a, V2,
and oxytocin receptors have already been described earlier in
saturation experiments, respectively Kd = (0.70 ( 0.17) nM,
(1.36 ( 0.45) nM, and (1.36 ( 1.00) nM.50,55 The affinities (Ki) of
the unlabeled ligands were determined by competition experi-
ments using [3H]AVP (1-2 nM) and varying the concentrations
of the displacing ligands from 0.1 pM to 100 μM (without
exceeding a 4% final concentration in DMSO). Nonspecific
binding was determined by adding an excess of AVP (10 μM).
Bound and free radioactivity were separated by filtration on a
Brandel apparatus over Whatman GF/C filters presoaked in a
10 mg/mL BSA solution for 3-4 h. Radioactivity on the filters
was counted on a β-counter Tricarb 2100TR (Packard). The
ligand binding data were analyzed by nonlinear least-squares
regression using the computer program Kell Radlig. The inhibi-
tion constants (Ki) for unlabeled ligands were calculated accord-
ing to the Cheng and Prusoff equation: Ki = IC50/(1 þ [L*]/Kd),
where IC50 is the concentration of unlabeled ligand leading to
half-maximal inhibition of specific binding, [L*] is the concentra-
tion of the radioligand present in the assay, and Kd is its affinity
for the receptor studied. The given Ki values are expressed as the
mean ( SEM of at least three independent experiments, each
performed in triplicate, unless otherwise specified.
Functional Assays. The functional agonist and competitive
antagonist properties of each ligand were determined for the
human vasopressin and oxytocin receptors subtypes stably
expressed in CHO cells. The accumulation of myo-inositol
1-phosphate (V1a and OT receptors) was determined by the
IP-One assay, and the accumulation of cAMP (V2 receptor) was
determined by the cAMP dynamic 2 assay, both kindly provided
by Cisbio International. Briefly, CHO cells stably expressing the
human vasopressin V1a, V2, or oxytocin receptors were seeded
(at 80-90% confluence) at a density of 80000 cells/well for V1a
and OT receptors, and 1000 cells/well for V2 receptor, in 96-well
culture plates (Greiner) in their culture medium in an atmo-
sphere of 95% air and 5% CO2 at 37 °C overnight. After
removal of the culture medium, ligands in the stimulation buffer
(already furnished in the IP-One kit; D-MEM supplemented
with 0.1 mM of the phosphodiesterases inhibitor Ro-20-1724
and 1 mg/mL BSA for the cAMP assay) were added in increas-
ing final concentrations ranging from 10 fM to 100 μM (without
exceeding a 1% final concentration in DMSO). To test the
agonist effect, cells were incubated with the ligands for 30 min
(37 °C þ 5% CO2), the positive control used being 1 μM AVP for
V1a-R, 0.1 μM AVP for V2-R, and 1 μM OT for OT-R. To test
the competitive antagonist effect, cells were preincubated with
the ligands for 10 min (37 °C þ 5% CO2) and then stimulated by
the corresponding endogenous agonist (2 nM AVP for V1a-R,
1 nM AVP for V2-R, and 50 nM OT for OT-R) and incubated
for 30 min (37 °C þ 5% CO2). The cells were lysed and incubated
at rt with the IP1-d2 and anti-IP1-cryptate conjugates (IP-One),
or cAMP-d2 and anti-cAMP-cryptate conjugates (cAMP
dynamic 2). After 1-2 h incubation, fluorescence emissions were
measured both at 620 nm and at 665 nm on a RubyStar
fluorometer (BMG Labtechnologies) equipped with a nitrogen
laser as excitation source (337 nm). A 400 μs reading was
measured after a 50 μs delay to eliminate the short-lived fluore-
scence background from the specific signal. The fluorescence
intensities measured at 620 nm and at 665 nm correspond to the
total europium cryptate emission and to the FRET signal,
respectively. Thespecificsignal was calculatedusing the following
equation: ΔF = (R - Rneg)/Rneg. R is the ratio (fluorescence
665 nm/fluorescence 620 nm) ꢀ 100 calculated in wells incubated
with both donor- and acceptor-labeled antibodies, whereas Rneg
is the same ratio for the negative control. The functional data
were analyzed by nonlinear regression using GraphPad Prism.
For the agonist properties, concentrations of ligands leading to
half-maximal IP1 or cAMP accumulation (EC50 or activation
constant Kact) were calculated from the sigmoidal dose-response
curves, and the maximal responses induced by ligands (Emax) are
expressed as percentages of maximal stimulation of IP1 or cAMP
accumulation by the corresponding endogenous agonist. For the
antagonist properties, concentrations of ligands leading to half-
maximal inhibition (IC50) of IP1 or cAMP accumulation were
calculated from the inversed sigmoidal dose-response curve. The
inactivation constants (Kinact) were calculated as: Kinact = IC50/
(1 þ [ago]/Kact), where [ago] is the concentration of the endogen-
ous agonist present in the assay ([AVP] = 2 nM for V1a-R,
[AVP] = 1 nM for V2-R, and [OT] = 50 nM for OT-R), and Kact
is the concentration of this agonist inducing half-maximal stimu-
lation of IP1 or cAMP accumulation (Kact (AVP) = (1.20 ( 0.33)
nM for V1a-R, Kact (AVP) = (0.25 ( 0.07) nM for V2-R, and Kact
(OT) = (10.6 ( 0.3) nM for OT-R). Results are expressed as the
mean ( SEM of at least three independent experiments, each
performed in triplicate, unless otherwise specified.
Molecular Modeling. The protocol used for molecular mod-
eling studies was as published recently by our group.37 Very
briefly, the residue numbering proposed by Ballesteros and
Weinstein94 was used throughout this manuscript. It allows an
unambiguous comparison of TM residues for any class A
GPCR by assigning position 50 to a fully conserved amino acid
at each TM and numbering other amino acids according to this
reference residue. Residue x.y is thus the amino acid describing

1560 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
position y of TMx. For purposes of clarification, amino acids
from the peptide ligands will be labeled using a three-letter code,
whereas receptor residues will be labeled using a single-letter
code.
(6) Porges, S. W. Love: an emergent property of the mammalian
autonomic nervous system. Psychoneuroendocrinology 1998, 23,
837–861.
(7) Zak, P. J.; Kurzban, R.; Matzner, W. T. The neurobiology of trust.
Ann. N. Y. Acad. Sci. 2004, 1032, 224–227.
The amino acid sequences of human OT receptor was retrieved
from the Swiss-Prot database and aligned to the sequence of
bovine rhodopsin using the in-house developed GPCRmod
program95 focusing on transmembrane (TM) domains only.
The alignment of the amino and carboxy-terminal domains as
well as of the intra- and extracellular loops was realized using
ClustalW.96 Modeling the AVP-bound conformation of human
(8) Zak, P. J.; Kurzban, R.; Matzner, W. T. Oxytocin is associated with
human trustworthiness. Horm. Behav. 2005, 48, 522–527.
(9) Kosfeld, M.; Heinrichs, M.; Zak, P. J.; Fischbacher, U.; Fehr, E.
Oxytocin increases trust in humans. Nature 2005, 435, 673–676.
(10) Heinrichs, M.; Baumgartner, T.; Kirschbaum, C.; Ehlert, U. Social
support and oxytocin interact to suppress cortisol and subjective
responses to psychosocial stress. Biol. Psychiatry 2003, 54, 1389–
1398.
V
V
1a and V1b receptors3-D ground-statemodels ofhuman V1a and
€
(11) Uvnas-Moberg, K. Oxytocin may mediate the benefits of positive
1b receptors have recently been reported by our group.37,55,97,98
social interaction and emotions. Psychoneuroendocrinology 1998,
23, 819–835.
To achieve an agonist-bound model from an antagonist-bound
model, we followed a five-step protocol as proposed by Bissantz
etal.97 Itshouldbeobserved that the starting conformation ofOT
was modeled from the X-ray structure of neurophysin-bound
OT.39 Compound 1 was built starting from the X-ray structure of
its truncated analogue 12c, and then it was fully optimized with
HF/6-31* level using geometry optimization with Gaussian 03.40
OT and compound 1 were docked to this preliminary model using
the Gold 3.0 program.38 For each of the 10 independent genetic
algorithm(GA) runs, a maximumnumberof1000GA operations
was performed on a single population of 50 individuals. Operator
weights for crossover, mutation, and migration were set to 100,
100, and 0, respectively. To allow poor nonbonded contacts at the
start of each GA run, the maximum distance between hydrogen
(12) DeVries, A. C.; Glasper, E. R.; Detillion, C. E. Social modulation
of stress responses. Physiol. Behav. 2003, 79, 339–407.
(13) Insel, T. R.; O’Brien, D. J.; Leckman, J. F. Oxytocin, vasopressin,
and autism: is there a connection? Biol. Psychiatry 1999, 45, 145–
157.
(14) Kim, S. J.; Young, L. J.; Gonen, D.; Veenstra-VanderWeele, J.;
Courchesne, R.; Courchesne, E.; Lord, C.; Leventhal, B. L.; Cook,
E. H., Jr.; Insel, T. R. Transmission disequilibrium testing of
arginine vasopressin receptor 1A (AVPR1A) polymorphisms in
autism. Mol. Psychiatry 2002, 7, 503–507.
(15) Wu, S.; Jia, M.; Ruan, Y.; Liu, J.; Guo, Y.; Shuang, M.; Gong, X.;
Zhang, Y.; Yang, X.; Zhang, D. Positive association of the
oxytocin receptor gene (OXTR) with autism in the Chinese Han
population. Biol. Psychiatry 2005, 58, 74–77.
(16) Jacob, S.; Brune, C. W.; Carter, C. S.; Leventhal, B. L.; Lord, C.;
Cook, E. H., Jr. Association of the oxytocin receptor gene (OXTR)
in Caucasian children and adolescents with autism. Neurosci. Lett.
2007, 417, 6–9.
˚
donors and fitting points was set to 5 A, and nonbonded van der
Waals energies were cut off at a value equal to kij (well depth of
the van der Waals energy for the atom pair i,j). To further speed
up the calculation, the GA docking was stopped when the top
(17) Modahl, C.; Green, L.; Fein, D.; Morris, M.; Waterhouse, L.;
Feinstein, C.; Levin, H. Plasma oxytocin levels in autistic children.
Biol. Psychiatry 1998, 43, 270–277.
(18) Green, L.; Fein, D.; Modahl, C.; Feinstein, C.; Waterhouse, L.;
Morris, M. Oxytocin and autistic disorder: alterations in peptide
forms. Biol. Psychiatry 2001, 50, 609–613.
˚
three solutions were within 1.5 A rmsd. If this criterion is met, we
can assume that these top solutions represent a reproducible pose
for the ligand.
(19) Hollander, E.; Novotny, S.; Hanratty, M.; Yaffe, R.; DeCaria,
C. M.; Aronowitz, B. R.; Mosovich, S. Oxytocin infusion reduces
repetitive behaviors in adults with autistic and Asperger’s dis-
orders. Neuropsychopharmacology 2003, 28, 193–198.
(20) Hollander, E.; Bartz, J.; Chaplin, W.; Phillips, A.; Sumner, J.;
Soorya, L.; Anagnostou, E.; Wasserman, S. Oxytocin increases
retention of social cognition in autism. Biol. Psychiatry 2007, 61,
498–503.
Acknowledgment. Warm thanks are due to Dr. Hadjila
Chabane for the preparation of compound 15 and for her
constant connection to the project, and to Dr. Dominique
Bonnet and Didier Rognan for their scientific support. This
ꢁ
work was supported by the Ministere de la Recherche for
fundings to MCF and by the European Commission for the
Marie-Curie fellowship to J.R. (HPMF-CT-2002-02141).
L.B. was supported by grants from the French Agence
Nationale de la Recherche (ANR-06-PCVI-0021-03). This
work was possible due to the Plateforme Pharmacologie-
(21) Vassart, G. Non-peptide arginine-vasopressin antagonists: the
vaptans. Lancet 2008, 371, 1624–1632.
(22) Manning, M.; Stoev, S.; Chini, B.; Durroux, T.; Mouillac, B.;
Guillon, G. Peptide and non-peptide agonists and antagonists for
the vasopressin and oxytocin V1a, V1b, V2 and OT receptors:
research tools and potential therapeutic agents. Prog. Brain Res.
2008, 170, 473–512.
(23) Ashworth, D. M.; Batt, A. R.; Baxter, A. J.; Broqua, P.; Haigh,
R. M.; Hudson, P.; Heeney, C. M. S.; Laporte, R.; Penson, A.; Pitt,
G. R. W.; Robson, P. A.; Rooker, D. P.; Tartar, A. L.; Yea, C. M.;
Roe, M. B. Nonpeptide oxytocin agonists. Drugs Future 2006, 31,
345–353.
(24) 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.
ꢀ ꢀ
Criblage-Interactome of the Institut Federatif de Recherche
no. 3 in Montpellier. This work was supported by grants from
ꢀ
ꢀ
CNRS, INSERM, Universite de Strasbourg, and Universite
Montpellier 1 and 2. We thank Pascale Buisine and Patrick
Werhung (IFR85) for MS analyses and Cyril Antheaume for
NMR analyses.
Supporting Information Available: Proposed binding mode of
oxytocin to human oxytocin receptor and HPLC tracings for
target compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
(25) Pardridge, W. M. Blood-brain barrier delivery. Drug Discovery
Today 2007, 12, 54–61.
(26) Matsuhisa, A.; Tanaka, A.; Kikuchi, K.; Shimada, Y.; Yatsu, T.;
Yanagisawa, I. Nonpeptide arginine vasopressin antagonists for
both V1a and V2 receptors: synthesis and pharmacological proper-
ties of 2-phenyl-40-[(2,3,4,5-tetrahydro-1H-1-benzazepin-1-yl)-
carbonyl]benzanilide derivatives. Chem. Pharm. Bull. 1997, 45,
1870–1874.
References
(1) Carter, C. S. Neuroendocrine perspectives on social attachment
and love. Psychoneuroendocrinology 1998, 23, 779–818.
(2) Insel, T. R.; Young, L. J. The neurobiology of attachment. Nat.
Rev. Neurosci. 2001, 2, 129–136.
(3) Komisaruk, B. R.; Whipple, B. Love as sensory stimulation:
physiological consequences of its deprivation and expression.
Psychoneuroendocrinology 1998, 23, 927–944.
(4) Zeki, S. The neurobiology of love. FEBS Lett. 2007, 581, 2575–
2579.
(5) Debiec, J. From affiliative behaviours to romantic feelings: a role of
(27) Caggiano, T. J. WAY-VNA-932. Drugs Future 2002, 27, 248–
253.
(28) Williams, P. D.; Clineschmidt, B. V.; Erb, J. M.; Freidinger, R. M.;
Guidotti, M. T.; Lis, E. V.; Pawluczyk, J. M.; Pettibone, D. J.;
Reiss, D. R.; Veber, D. F.; Woyden, C. J. 1-{1-[4-[(N-Acetyl-
4-piperidinyl)oxy]-2-methoxybenzoyl]piperidin-4-yl}-4H-3,1-
benzoxazin-2(1H)-one (L-371,257): a new, orally bioavailable, non-
peptide oxytocin antagonist. J. Med. Chem. 1995, 38, 4634–4636.
(29) Pitt, G. R. W.; Roe, M. B.; Rooker, D. P. Oxytocin agonists. Patent
WO 03/000692 A3, 2003.
nanopeptides. FEBS Lett. 2007, 581, 2580–2586.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4 1561
(30) Sasatani, S.; Miyazaki, T.; Maruoka, K.; Yamamoto, H. Diiso-
butylaluminium hydride: a novel reagent for the reduction of
oximes. Tetrahedron Lett. 1983, 24, 4711–4712.
(31) Artico, M.; De Martino, G.; Filacchioni, G.; Giuliano, R. Ricerche
su sostanze ad attivita’ antiblastica. Il Farmaco;Ed. Sci. 1969, 24,
276–284.
(32) Bell, I. M.; Erb, J. M.; Freidinger, R. M.; Gallicchio, S. N.; Guare,
J. P.; Guidotti, M. T.; Halpin, R. A.; Hobbs, D. W.; Homnick,
C. F.; Kuo, M. S.; Lis, E. V.; Mathre, D. J.; Michelson, S. R.;
Pawluczyk, J. M.; Pettibone, D. J.; Reiss, D. R.; Vickers, S.;
Williams, P. D.; Woyden, C. J. Development of orally active
oxytocin antagonists: studies on 1-(1-{4-[1-(2-methyl-1-oxidopyridin-
3-ylmethyl)piperidin-4-yloxy]-2-methoxybenzoyl}piperidin-4-yl)-1,4-
dihydrobenz[d][1,3]oxazin-2-one (L-372,662) and related pyridines.
J. Med. Chem. 1998, 41, 2146–2163.
(33) 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 template. J. Med. Chem. 2000, 43, 4388–4397.
Barberis, C.; Mouillac, B. Direct identification of human oxytocin
receptor-binding domains using a photoactivatable cyclic peptide
antagonist: comparison with the human V1a vasopressin receptor.
J. Biol. Chem. 2001, 276, 26931–26941.
(51) Thibonnier, M.; Coles, P.; Thibonnier, A.; Shoham, M. Molecular
pharmacology and modelling of vasopressin receptors. Prog. Brain
Res. 2002, 139, 179–196.
(52) Hawtin, S. R.; Wesley, V. J.; Parslow, R. A.; Simms, J.; Miles, A.;
McEwan, K.; Wheatley, M. A single residue (Arg46) located within
the N-terminus of the V1a vasopressin receptor is critical for
binding vasopressin but not peptide or nonpeptide antagonists.
Mol. Endocrinol. 2002, 16, 600–609.
(53) Wesley, V. J.; Hawtin, S. R.; Howard, H. C.; Wheatley, M.
Agonist-specific, high-affinity binding epitopes are contributed
by an arginine in the N-terminus of the human oxytocin receptor.
Biochemistry 2002, 41, 5086–5092.
(54) Hawtin, S. R.; Wesley, V. J.; Simms, J.; Parslow, R. A.; Miles, A.;
McEwan, K.; Keen, M.; Wheatley, M. An arginyl in the
N-terminus of the V1a vasopressin receptor is part of the conforma-
tional switch controlling activation by agonist. Eur. J. Biochem.
2003, 270, 4681–4688.
(34) Carnazzi, E.; Aumelas, A.; Phalipou, S.; Mouillac, B.; Guillon, G.;
Barberis, C.; Seyer, R. Efficient photoaffinity labeling of the rat V1a
vasopressin receptor using a linear azidopeptidic antagonist. Eur.
J. Biochem. 1997, 247, 906–913.
(55) Tahtaoui, C.; Balestre, M. N.; Klotz, P.; Rognan, D.; Barberis, C.;
Mouillac, B.; Hibert, M. Identification of the binding sites of the
SR49059 nonpeptide antagonist into the V1a vasopressin receptor
using sulfydryl-reactive ligands and cysteine mutants as chemical
sensors. J. Biol. Chem. 2003, 278, 40010–40019.
ꢀ
(35) Phalipou, S.; Cotte, N.; Carnazzi, E.; Seyer, R.; Mahe, E.; Jard, S.;
Barberis, C.; Mouillac, B. Mapping peptide-binding domains of
the human V1a vasopressin receptor with a photoactivatable linear
peptide antagonist. J. Biol. Chem. 1997, 272, 26536–26544.
(36) Phalipou, S.; Seyer, R.; Cotte, N.; Breton, C.; Barberis, C.; Hibert,
M.; Mouillac, B. Docking of linear peptide antagonists into the
human V1a vasopressin receptor: identification of binding domains
by photoaffinity labelling. J. Biol. Chem. 1999, 274, 23316–23327.
(37) Rodrigo, J.; Pena, A.; Murat, B.; Trueba, M.; Durroux, T.;
Guillon, G.; Rognan, D. Mapping the binding site of arginine
vasopressin to V1a and V1b vasopressin receptors. Mol. Endocrinol.
2007, 21, 512–523.
(38) Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.; Murray, C. W.;
Taylor, R. D. Improved protein-ligand docking using GOLD.
Proteins 2003, 52, 609–623.
(39) Rose, J. P.; Wu, C. K.; Hsiao, C. D.; Breslow, E.; Wang, B. C.
Crystal structure of the neurophysin-oxytocin complex. Nat.
Struct. Biol. 1996, 3, 163–169.
(56) Gimpl, G.; Postina, R.; Fahrenholz, F.; Reinheimer, T. Binding
domains of the oxytocin receptor for the selective oxytocin receptor
antagonist barusiban in comparison to the agonists oxytocin and
carbetocin. Eur. J. Pharmacol. 2005, 510, 9–16.
(57) Hawtin, S. R.; Simms, J.; Conner, M.; Lawson, Z.; Parslow, R. A.;
Trim, J.; Sheppard, A.; Wheatley, M. Charged extracellular resi-
dues, conserved throughout a G-protein-coupled receptor family,
are required for ligand binding, receptor activation, and cell-
surface expression. J. Biol. Chem. 2006, 281, 38478–38488.
(58) Macion-Dazard, R.; Callahan, N.; Xu, Z.; Wu, N.; Thibonnier,
M.; Shoham, M. Mapping the binding site of six nonpeptide
antagonists to the human V2-renal vasopressin receptor. J. Phar-
macol. Exp. Ther. 2006, 316, 564–571.
(59) Conner, M.; Hawtin, S. R.; Simms, J.; Wootten, D.; Lawson, Z.;
Conner, A. C.; Parslow, R. A.; Wheatley, M. Systematic analysis of
the entire second extracellular loop of the V1a vasopressin receptor:
key residues, conserved throughout a G-protein-coupled receptor
family, identified. J. Biol. Chem. 2007, 282, 17405–17412.
(60) Nakamura, S.; Yamamura, Y.; Itoh, S.; Hirano, T.; Tsujimae, K.;
Aoyama, M.; Kondo, K.; Ogawa, H.; Shinohara, T.; Kan, K.;
Tanada, Y.; Teramoto, S.; Sumida, T.; Nakayama, S.; Sekiguchi,
K.; Kambe, T.; Tsujimoto, G.; Mori, T.; Tominaga, M. Characteri-
zation of a novel nonpeptide vasopressin V2-agonist, OPC-51803,
in cells transfected human vasopressin receptor subtypes. Br. J.
Pharmacol. 2000, 129, 1700–1706.
(40) Gaussian, Inc., Wallingford CT, 2004.
(41) Barberis, C.; Mouillac, B.; Durroux, T. Structural bases of vaso-
pressin/oxytocin function. J. Endocrinol. 1998, 156, 223–229.
(42) Hojro, E.; Eich, P.; Gimpl, G.; Fahrenholz, F. Direct identification
of an extracellular agonist binding site in the renal V2 vasopressin
receptor. Biochemistry 1993, 32, 13537–13544.
(43) Mouillac, B.; Chini, B.; Balestre, M. N.; Elands, J.; Trumpp-
Kallmeyer, S.; Hoflack, J.; Hibert, M.; Jard, S.; Barberis, C. The
binding site of neuropeptide vasopressin V1a receptor: evidence for
a major localization within transmembrane regions. J. Biol. Chem.
1995, 270, 25771–25777.
(44) Chini, B.; Mouillac, B.; Ala, Y.; Balestre, M. N.; Trumpp-Kallmeyer,
S.; Hoflack, J.; Elands, J.; Hibert, M.; Manning, M.; Jard, S.;
Barberis, C Tyr 115 is the key residue for determining agonist
selectivity in the V1a vasopressin receptor. EMBO J. 1995, 14,
2176–2182.
(45) Ufer, E.; Postina, R.; Gorbulev, V.; Fahrenholz, F. An extra-
cellular residue determines the agonist specificity of V2 vasopressin
receptors. FEBS Lett. 1995, 362, 19–23.
(46) Chini, B.; Mouillac, B.; Balestre, M. N.; Trumpp-Kallmeyer, S.;
Hoflack, J.; Hibert, M.; Andriolo, M.; Pupier, S.; Jard, S.; Barberis,
C. Two aromatic residues regulate the response of the human
oxytocin receptor to the partial agonist arginine vasopressin. FEBS
Lett. 1996, 397, 201–206.
(47) Postina, R.; Kojro, E.; Fahrenholz, F. Separate agonist and
peptide antagonist binding sites of the oxytocin receptor defined
by their transfer into the V2 vasopressin receptor. J. Biol. Chem.
1996, 271, 31593–31601.
(61) WAY-141608, Drug Data Rep. 2000, 22, 520.
(62) Kondo, K.; Kan, K.; Tanada, 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 agonists: 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-isopropylacetamide. J. Med. Chem. 2002, 45,
3805–3808.
(63) Failli, A. A.; Shumsky, J. S.; Steffan, R. J.; Caggiano, T. J.;
Williams, D. K.; Trybulski, E. J.; Ning, X.; Lock, Y.; Tanikella,
T.; Hartmann, D.; Chan, P. S.; Park, C. H. Pyridobenzodiazepines:
a novel class of orally active, vasopressin V2 receptor selective
agonists. Bioorg. Med. Chem. Lett. 2006, 16, 954–959.
(64) Yamamura, Y.; Ogawa, H.; Chihara, T.; Kondo, K.; Onogawa, T.;
Nakamura, S.; Mori, T.; Tominaga, M.; Yabuuchi, Y. OPC-21268,
an orally effective, nonpeptide vasopressin V1 receptor antagonist.
Science 1991, 252, 572–574.
(65) Ogawa, H.; Yamashita, H.; Kondo, K.; Yamamura, Y.; Miyamoto,
H.; Kan, K.; Kitano, K.; Tanaka, M.; Nakaya, K.; Nakamura, S.;
Mori, T.; Tominaga, M.; Yabuuchi, Y. Orally active, nonpeptide
vasopressin V2 receptor antagonists: a novel series of 1-[4-(benzo-
ylamino)benzoyl]-2,3,4,5-tetrahydro-1H-benzazepines and related
compounds. J. Med. Chem. 1996, 39, 3547–3555.
(48) Hibert, M.; Hoflack, J.; Trumpp-Kallmeyer, S.; Mouillac, B.;
ꢀ
Chini, B.; Mahe, E.; Cotte, N.; Jard, S.; Manning, M.; Barberis,
C. Functional architecture of vasopressin/oxytocin receptors.
J. Recept. Signal Transduction Res. 1999, 19, 589–596.
ꢀ
€
(49) Cotte, N.; Balestre, M. N.; Aumelas, A.; Mahe, E.; Phalipou, S.;
(66) Ariens, E. J. A molecular basis for the action of drugs. I. Receptor-
Morin, D.; Hibert, M.; Manning, M.; Durroux, T.; Barberis, C.;
Mouillac, B. Conserved aromatic residues in the transmembrane
region VI of the V1a vasopressin receptor differenciate agonist vs
antagonist ligand binding. Eur. J. Biochem. 2000, 267, 4253–4263.
(50) Breton, C.; Chellil, H.; Kabbaj-Benmansour, M.; Carnazzi, E.;
Seyer, R.; Phalipou, S.; Morin, D.; Durroux, T.; Zingg, H.;
theory and structure-effect relations. Arzneim. Forsch. 1966, 16,
1376–1395.
(67) Goddard, W. A., III; Abrol, R. 3-Dimensional structures of
G-protein-coupled receptors and binding sites of agonists and
antagonists. J. Nutr. 2007, 137, 1528S-1538S, and references cited
therein.

1562 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 4
Frantz et al.
(68) Portoghese, P. S.; Lipkowski, A. W.; Takemori, A. E. Bimorphans
as highly selective, potent kappa opioid receptor antagonist.
J. Med. Chem. 1987, 30, 238–239.
(69) Perlman, S.; Costa-Neto, C. M.; Miyakawa, A. A.; Schambye,
H. T.; Hjorth, S. A.; Paiva, A. C. M.; Rivero, R. A.; Greenlee, W. J.;
Schwartz, T. W. Dual agonistic and antagonistic property of
nonpeptide angiotensin AT1 ligands: susceptibility to receptor
mutations. Mol. Pharmacol. 1997, 51, 301–311.
(83) Fleck, B. A.; Chen, C.; Yang, W.; Huntley, R.; Markison, S.;
Nickolls, S. A.; Foster, A. C.; Hoare, S. R. J. Molecular interactions
of nonpeptide agonists and antagonists with the melanocortin-4
receptor. Biochemistry 2005, 44, 14494–14508.
(84) Fleck, B. A.; Ling, N.; Chen, C. Substituted NDP-MSH peptides
paired with mutant melanocortin-4 receptors demonstrate the role
of transmembrane 6 in receptor activation. Biochemistry 2007, 46,
10473–10483.
€
(70) Furst, S.; Hosztafi, S.; Friedmann, T. Structure-activity relation-
(85) Foucaud, M.; Tikhonova, I. G.; Langer, I.; Escrieut, C.; Dufresne,
M.; Seva, C.; Maigret, B.; Fourmy, D. Partial agonism, neutral
antagonism, and inverse agonism at the human wild-type and
constitutively active cholecystokinin-2 receptors. Mol. Pharmacol.
2006, 69, 680–690.
(86) Kim, S. K.; Gao, Z. G.; Van Rompaey, P.; Gross, A. S.; Chen, A.;
Van Calenbergh, S.; Jacobson, K. A. Modeling the adenosine
receptors: comparison of the binding domains of A2A agonists
and antagonists. J. Med. Chem. 2003, 46, 4847–4859.
(87) Ghanouni, P.; Gryczynski, Z.; Steenhuis, J. J.; Lee, T. W.; Farrens,
D. L.; Lakowicz, J. R.; Kobilka, B. K. Functionally different
agonists induce distinct conformations in the G protein coupling
domain of the β2 adrenergic receptor. J. Biol. Chem. 2001, 276,
24433–24436.
(88) Palczewski, K.; Kumasaka, T.; Hori, T.; Behnke, C. A.; Motoshima,
H.; Fox, B. A.; Le Trong, I.; Teller, D. C.; Okada, T.; Stenkamp,
R. E.; Yamamoto, M.; Miyano, M. Crystal structure of rhodopsin: a
G protein-coupled receptor. Science 2000, 289, 739–745.
(89) Schwartz, T. W.; Frimurer, T. M.; Holst, B.; Rosenkilde, M. M.;
Elling, C. E. Molecular mechanism of 7TM receptor activation;a
global toggle switch model. Annu. Rev. Pharmacol. Toxicol. 2006,
46, 481-519, and references cited therein.
(90) Durroux, T. Principles: a model for the allosteric interactions
between ligand binding sites within a dimeric GPCR. Trends
Pharmacol. Sci. 2005, 26, 376–384.
(91) Jaakola, V.-P.; Griffith, M.; Hanson, M. A.; Cherezov, V.; Chien,
E. Y.; Lane, J; IJzerman, J. R.; Stevens, A. P. R.C. The 2.6
Angstrom Crystal Structure of a Human A2A Adenosine Receptor
Bound to an Antagonist. Science 2008, 322, 1211–1217.
ships of synthetic and semisynthetic opioid agonists and antago-
nists. Curr. Med. Chem. 1995, 1, 423–440.
(71) Aquino, C. J.; Armour, D. R.; Berman, J. M.; Birkemo, L. S.; Carr,
R. A. E.; Croom, D. K.; Dezube, M.; Dougherty, R. W.; Ervin,
G. N.; Grizzle, M. K.; Head, J. E.; Hirst, G. C.; James, M. K.;
Johnson, M. F.; Miller, L. J.; Queen, K. L.; Rimele, T. J.; Smith,
D. N.; Sugg, E. E. Discovery of 1,5-benzodiazepines with peri-
pheral cholecystokinin (CCK-A) receptor agonist activity. 1.
Optimization of the agonist “trigger”. J. Med. Chem. 1996, 39,
562–569.
(72) Asano, M.; Hatori, C.; Sawai, H.; Johki, S.; Inamura, N.; Kayakiri,
H.; Satoh, S.; Abe, Y.; Inoue, T.; Sawada, Y.; Mizutani, T.; Oku,
T.; Nakahara, K. Pharmacological characterization of a nonpep-
tide bradykinin B2 receptor antagonist, FR165649, and agonist,
FR190997. Br. J. Pharmacol. 1998, 124, 441–446.
(73) Gouldson, P.; Legoux, P.; Carillon, C.; Delpech, B.; Le Fur, G.;
Ferrara, P.; Shire, D. The agonist SR146131 and the antagonist
SR27897 occupy different sites on the human CCK1 receptor. Eur.
J. Pharmacol. 2000, 400, 185–194.
(74) Escrieut, C.; Gigoux, V.; Archer, E.; Verrier, S.; Maigret, B.;
Behrendt, R.; Moroder, L.; Bignon, E.; Silvente-Poirot, S.; Pradayrol,
L.; Fourmy, D. The biologically crucial C terminus of cholecysto-
kinin and the non-peptide agonist SR-146,131 share a common
binding site in the human CCK1 receptor: evidence for a crucial
role of Met-121 in the activation process. J. Biol. Chem. 2002, 277,
7546–7555.
(75) Holst, B.; Mokrosinski, J.; Lang, M.; Brandt, E.; Nygaard, R.;
Frimurer, T. M.; Beck-Sickinger, A. G.; Schwartz, T. W. Identifi-
cation of an efficacy switch region in the ghrelin receptor respon-
sible for interchange between agonism and inverse agonism. J. Biol.
Chem. 2007, 282, 15799–15811.
(92) Gurrath, M. Peptide-binding,
G protein-coupled receptors:
new opportunities for drug design. Curr. Med. Chem. 2001, 8,
1605–1648.
(76) Buck, E.; Wells, J. A. Disulfide trapping to localize small-molecule
agonists and antagonists for a G protein-coupled receptor. Proc.
Natl. Acad. Sci. U.S.A. 2005, 102, 2719–2724.
(93) Blakeney, J. S.; Reid, R. C.; Le, G. T.; Fairlie, D. P. Nonpeptidic
ligands for peptide-activated G protein-coupled receptors. Chem.
Rev. 2007, 107, 2960–3041.
(94) Ballesteros, J. A.; Weinstein, H. Integrated methods for the con-
struction of three dimensional models and computational probing
of structure-function relations in G-protein coupled receptors.
Methods Neurosci. 1995, 25, 366–428.
(95) Bissantz, C.; Logean, A.; Rognan, D. High-throughput modeling
of human G-protein coupled receptors: amino acid sequence
alignment, three-dimensional model building, and receptor library
screening. J. Chem. Inf. Comput. Sci. 2004, 44, 1162–1176.
(96) Thompson, J. D.; Higgins, D. G.; Gibson, T. J. CLUSTAL W:
improving the sensitivity of progressive multiple sequence align-
ment through sequence weighting, position-specific gap penalties
and weight matrix choice. Nucleic Acids Res. 1994, 22, 4673–
4680.
(97) Bissantz, C.; Bernard, P.; Hibert, M.; Rognan, D. Protein-based
virtual screening of chemical databases. II. Are homology models
of G-protein coupled receptors suitable targets? Proteins 2003, 50,
5–25.
(77) Wise, E. L.; Duchesnes, C.; da Fonseca, P. C. A.; Allen, R. A.;
Williams, T. J.; Pease, J. E. Small molecule receptor agonists and
antagonists of CCR3 provide insight into mechanisms of chemo-
kine receptor activation. J. Biol. Chem. 2007, 282, 27935–27943.
(78) Shen, C. P.; Xiao, J. C.; Armstrong, H.; Hagmann, W.; Fong, T. M.
F200A substitution in the third transmembrane helix of human
cannabinoid CB1 receptor converts AM2233 from receptor agonist
to inverse agonist. Eur. J. Pharmacol. 2006, 531, 41–46.
(79) Jensen, P. C.; Nygaard, R.; Thiele, S.; Elder, A.; Zhu, G.; Kolbeck,
R.; Ghosh, S.; Schwartz, T. W.; Rosenkilde, M. M. Molecular
interaction of a potent nonpeptide agonist with the chemokine
receptor CCR8. Mol. Pharmacol. 2007, 72, 327–340.
(80) Hadac, E. M.; Dawson, E. S.; Darrow, J. W.; Sugg, E. E.; Lybrand,
T. P.; Miller, L. J. Novel benzodiazepine photoaffinity probe
stereoselectively labels a site deep within the membrane-spanning
domain of the cholecystokinin receptor. J. Med. Chem. 2006, 49,
850–863.
(81) Scarselli, M.; Li, B.; Kim, S. K.; Wess, J. Multiple residues in the
second extracellular loop are critical for M3 muscarinic acetylcho-
line receptor activation. J. Biol. Chem. 2007, 282, 7385–7396.
(98) Derick, S.; Pena, A.; Durroux, T.; Wagnon, J.; Serradeil-, C.;
Hibert, M.; Rognan, D.; Guillon, G. Key amino acids located
within the transmembrane domains 5 and 7 account for the
pharmacological specificity of the human V1b vasopressin receptor.
Mol. Endocrinol. 2004, 18, 2777–2789.
(99) SYBYL 7.3; Tripos International, 1699 South Hanley Road, St. Louis,
MO, 63144, 2006.
ꢁ
(82) Baneres, J. L.; Mesnier, D.; Martin, A.; Joubert, L.; Dumuis, A.;
Bockaert, J. Molecular characterization of a purified 5-HT4
receptor: a structural basis for drug efficacy. J. Biol. Chem. 2005,
280, 20253–20260.