2,5-Diketopiperazines as potent, selective, and orally bioavailable oxytocin antagonists. 3. Synthesis, pharmacokinetics, and in vivo potency

Article abstract of DOI:10.1021/jm060073e

A short, efficient, and highly stereoselective synthesis of a series of (3R,6R,7R)-2,5-diketopiperazine oxytocin antagonists and their pharmacokinetics in rat and dog is described. Prediction of the estimated human oral absorption (EHOA) using measured li

Full text of DOI:10.1021/jm060073e

J. Med. Chem. 2006, 49, 4159-4170
4159
2,5-Diketopiperazines as Potent, Selective, and Orally Bioavailable Oxytocin Antagonists. 3.
Synthesis, Pharmacokinetics, and in Vivo Potency
Alan D. Borthwick,*,† Dave E. Davies,† Anne M. Exall,† Richard J. D. Hatley,† Jennifer A. Hughes,† Wendy R. Irving,†
David G. Livermore,† Steve L. Sollis,† Fabrizio Nerozzi,† Klara L. Valko,§ Michael J. Allen,| Marion Perren,|
Shalia S. Shabbir,‡ Patrick M. Woollard,‡ and Mark A. Price#
Departments of Medicinal Chemistry and DMPK, CardioVascular and Urogenital Centre of Excellence for Drug DiscoVery, GlaxoSmithKline
Research and DeVelopment, Medicines Research Centre, Gunnels Wood Road, SteVenage, Herts, SG1 2NY, U.K., Department of Assay
DeVelopment and Compound Profiling, GlaxoSmithKline, New Frontiers, Science Park, Third AVenue, Harlow, Essex, CM19 5AD, U.K., and
Safety Assessment, GlaxoSmithKline, Ware, Herts, SG12 0DP, U.K.
ReceiVed January 23, 2006
A short, efficient, and highly stereoselective synthesis of a series of (3R,6R,7R)-2,5-diketopiperazine oxytocin
antagonists and their pharmacokinetics in rat and dog is described. Prediction of the estimated human oral
absorption (EHOA) using measured lipophilicity (CHI log D) and calculated size (cMR) has allowed us to
rank various 2,5-diketopiperazine templates and enabled us to focus effort on those templates with the greatest
chance of high bioavailability in humans. This rapidly led to the 2′,4′-difluorophenyl-dimethylamide 25 and
the benzofuran 4 with high levels of potency (pKi) and good bioavailability in the rat and dog. Dimethylamide
25 is more potent (>20-fold) than 4 in vivo and has a high degree of selectivity toward the vasopressin
receptors, >10000 for hV1a/hV1b and ∼500 for hV2. It has a good Cyp450 profile with no time dependent
inhibition and was negative in the genotoxicity screens with a satisfactory oral safety profile in rats.
Introduction
was >1000-fold selective for the human oxytocin receptor
relative to all three vasopressin receptors hV1aR, hV2R, and
hV1bR. It had good rat pharmacokinetics with 46% bioavail-
ability. It was also active in vivo in the rat (DR10 ) 0.44 mg/
kg iv).
Preterm labor is a major clinical problem leading to death
and disability in newborns. It accounts for 10% of all births
and causes 70% of all infant mortality and morbidity.1 Oxytocin
is a potent stimulant of uterine contractions and is responsible
for the initiation of labor via the interaction with the oxytocin
receptors in mammalian female uterus. Oxytocin antagonists
have been shown to inhibit uterine contractions and delay
preterm delivery. Hence, there is increasing interest in oxytocin
antagonists as a consequence of their potential application in
the prevention of preterm labor and premature birth. Although
several tocolytics (uterine contraction inhibitors) have already
been approved in clinical practice, they have harmful maternal
or fetal side-effects.2 The first clinically tested oxytocin
antagonist, atosiban (Tractocile), has a much more tolerable side-
effect profile and has recently been approved for use in Europe.
However, atosiban is a peptide and a mixed oxytocin/vasopressin
V1a antagonist that has to be given by iv infusion and is unlikely
to be suitable for long-term maintenance treatment, because it
is not orally bioavailable.3 Our target is a potent, orally active
oxytocin antagonist with high levels of selectivity over the
vasopressin receptor, which would safely delay labor by greater
than 7 days and improve infant outcome.
However, it had poor aqueous solubility and was prepared
as the minor isomer in a short, nonstereospecific synthesis. Good
pharmacokinetics in a second species could only be obtained
by using a cyclodextrin formulation in the dog, which increased
the bioavailability from 13% to 31%.
In this report we address these issues by using structure-
property relationships to optimize the drug-like properties of
this template, in particular using models to select structures
having the potential for good human oral absorption. We
describe how we investigated the modification of the amide
function and developed a stereospecific, high yielding synthesis
to these derivatives, as well as further investigating the effect
of disubstitution in the aryl ring.
We recently reported4 the identification of a novel series of
(3R,6R,7R)-2,5-diketopiperazine (DKP) derivatives with an-
tagonist activity at the human oxytocin receptor (hOTR). The
most potent of these was the 2′,4′-difluorophenyl derivative 1
with a pKi ) 8.9 at the human oxytocin receptor that had only
a small shift in IC50 of 3.8-fold in the presence of HSA and
* To whom correspondence should be addressed: Phone: 07772112742.
Fax: +44(0)208 4555418. E-mail: alan.d.borthwick@drugmoldesign.com.
† Department of Medicinal Chemistry, Cardiovascular and Urogenital
Centre of Excellence for Drug Discovery, Stevenage, U.K.
§ Department Structural & Computational Science, Stevenage, U.K.
| Department Department of Assay Development and Compound Profil-
ing, Harlow Research 2, U.K.
Chemistry
In our previous paper4 we described the synthesis of 2,5-
diketopiperazine secondary amides in two steps using a four-
component Ugi reaction followed by ring closure. The com-
pounds 2-9 which are the 3R,6R,7R isomers were prepared in
e25% yield by this method (Figure 1, see Supporting Informa-
tion).
‡ Department of DMPK Cardiovascular and Urogenital Centre of
Excellence for Drug Discovery, Stevenage, U.K.
# Safety Assessment, Ware, U.K.
10.1021/jm060073e CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/21/2006

4160 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Borthwick et al.
Figure 1. 2,5-Diketopiperazine secondary amides.
Scheme 1. a Synthesis of 2,5-Diketopiperazine Tertiary Amides, Acids, and Primary Alcohol
a Reagents and conditions: (a) D-Leu methyl ester hydrochloride, methanol, aryl aldehyde, Et3N, room temp, 1-3 days; then Cbz-IndGly, 2-benzyloxyphenyl
isocynanide, room temp, 24 h; (b) H2, Pd/C, EtOAc, AcOH, 2-5 h; (c) CDI, DCM, 6 h; (d) R2NH, room temp, 16 h; (e) Me2CO, H+, H2O, 10 min; (f) Et3N,
DCM, MeCN then BOPCl, 10 min, followed by Me2NH, room temp, 16 h; (g) MeOH; (h) LiBHEt3 (3 equiv)/THF, -78 °C, 4 h.
A general synthetic procedure that enables the exocyclic
amide function to be readily modified is outlined in Scheme 1.
As a results of a sustained effort5 we have achieved a high level
of stereochemical control in the synthesis of the indane-2,5-
diketopiperazine tertiary amides which has enable us to obtain
>90% of our required 3R,6R,7R isomers. The four-component
Ugi reaction of Cbz indanyl R-glycine 10, R-aminoester (D-
leucine methyl ester) 11, convertible isonitrile6 12, and the 2′,4′-
difluorobenzaldehyde 13 gave the tripeptide intermediate 15,
which was followed by hydrogenation to remove the Cbz and
benzyl groups, enabling cyclization to occur to give the phenolic
2,5-diketopiperazine 17, in 70% yield from 13. Only 25% of
2,5-diketopiperazine 17 was the (3R,6R,7R) isomer.
The cyclic carbamate 19 obtained from 1,1′-carbonyldiimi-
dazole (CDI) and phenol 17 is an activated amide, which on
hydrolysis with water gave the carboxylic acid 21 in 86% yield.

2,5-Diketopiperazine Oxytocin Antagonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4161
Table 1. Bioavailability of Mono- and Diaryl Substituted (3R,6R,7R)-2,5-Diketopiperazines in the Rat and Dog
1
2
3
6
7
40a
R1
isoBu
8.9
36/46
11/13
60
isoBu
8.4
34/48
23/14
48
isoBu
8.4
39/27
14/9
46
isoBu
8.4
26/24
21/10
19
R-secBu
8.8
24/31
6/61
33
isoBu
8.5
48/2
15/50
46
pKib
rat Cl/F%c
dog C/F%c
EHOA %d
CHI log De
cMRf
3.4
13.51
3.3
14.54
3.4
14.85
2.4
16.61
3.0
15.69
3.4
14.78
a Prepared previously.4 b Displacement of [3H] oxytocin from hOTR by the test compound.7 pKi values are reported as mean values of three or more
determinations, and all results were obtained in binding assays with a standard deviation in pKi of less than 0.25. c Rat (n ) 4) and dog (n ) 3) PK: Cl in
mL min-1 kg-1 d EHOA ) estimated human oral absorption, eq 4. e An HPLC method-based measurement of lipophilicity.8 f Calculated molar refraction
.
(Daylight v 4.81).
Reaction of the acid 21 (activated with BOPCl) with dimethyl-
amine gave the 3R,6R,7R-dimethylamide 25 in 37% yield (12%
overall), together with 3R,6R,7S-dimethylamide 26 in 18% yield.
This is a 2:1 ratio of (3R,6R,7R): (3R,6R,7S) isomers, and we
sought to improve the ratio of isomers in favor of our required
(3R,6R,7R) isomer 25. We recognized that there were two
activated amide steps in the preparation of 25 via the acid 21
and therefore investigated the reaction of the first activated
amide 19 directly with dimethylamine. This reaction was much
more selective and gave a 10:1 ratio of the (3R,6R,7R) 25:
(3R,6R,7S) 26 dimethylamides in 68% yield. The overall yield
of 3R,6R,7R-dimethylamide 25 for the three stages from 13 is
40%. A similar 10:1 ratio of the (3R,6R,7R): (3R,6R,7S) isomers
was achieved directly from the activated amide 19 with other
secondary amines 27-30,5 whereas for methylamine and
ammonia which furnished 24 and 23 the (3R,6R,7R): (3R,6R,7S)
isomer ratio was 3:1. A similar sequence from 4-fluorobenzal-
dehyde 14 gave the monofluoro analogues 31-36 in similar
overall yields. Reaction of the cyclic carbamate 19 with
methanol gave a 2.4:1 ratio of the esters 37:38 in 93% yield.
Reduction of the (3R,6R,7R) isomer 37 with LiBHEt3 in THF
gave the alcohol 39 in 56% yield.
further exploit substitution at this position. Also further modi-
fications to the aryl ring were envisaged. However, we needed
to choose one of these substituted phenyl derivatives (Table 1)
to use to explore a range of amides. With the bioavailability of
1, 2, 3, 6, and 7 being good either in the rat or dog it was
considered that some other estimation of how these would
behave in humans would to be an advantage in choosing the
best template to progress.
Several approaches to predicting oral bioavailability in
humans have been investigated.9 A key part of oral bioavail-
ability is absorption across the gut wall by passive diffusion.
Human intestinal drug absorption is largely driven by lipophi-
licity, hydrogen bonding, and size.9 Although the most com-
monly used models, the Lipinski “rule of 5”,10 the molecular
polar surface area model,11 and the Abraham descriptors,12 are
very useful in general for selection of better compound librar-
ies, they are less accurate in ranking compounds within this
chemical series. We have found that the application of mea-
sured physicochemical data and models based on measured
in vitro and physical chemical properties of the molecules are
very helpful in selecting good candidate molecules within a
chemical series. Also for the current 2,5-diketopipera-
zine compounds there is a poor correlation between the
measured and calculated lipophilicity, possibly because the
calculation methods employed are 2D based and do not
compensate for potential conformational effects of these chirally
pure molecules, which contain three asymmetric centers. It is
possible that due to their conformation not all of the hydrophobic
regions of the molecule are exposed to the solvent; this could
result in measured values being significantly lower than
calculated, which is consistent with our observations (see
Supporting Information).
We have developed several types of high throughput HPLC-
based measurements of lipophilicity, such as the Chromato-
graphic Hydrophobicity Index (CHI) using C-18 stationary
phases and fast acetonitrile CHI (ACN) or methanol CHI
(MeOH) gradient.8 The CHI values can be expressed in the log
D scale and can be compared to the octanol/water log Ds using
the Abraham solvation equation approach.13 This rapid measure-
ment of lipophilicity has been used in the recently published
equation14 (1) that relates estimated human oral absorption
(EHOA) to measured lipophilicity CHI (MeOH) and calculated
size (cMR) using data for a range of drugs with known human
oral absorption. The simplified version of the model calculated
Results and Discussion
4
We recently had success in improving the bioavailabil-
ity of 4′-monosubstituted phenyl-2,5-diketopiperazines by
further substitution in the 4-fluorophenyl ring giving the 2′,4′-
difluorophenyl derivative 1 (F ) 46% in rat). To exploit
this, several further disubstituted aryl ring analogues were
prepared. The best of these together with the 4′-monosubsti-
tuted derivatives 6 and 40 (Table 1) had good bioavaila-
bility in animals and good potency in the oxytocin binding
assay (pKi ) 8.4-8.9). Those with the greatest bioavailabil-
ity in the rat were the 2′,4′-difluorophenyl derivative 1 and
the 3′,4′-methylenedioxyphenyl derivative 2, while the dihy-
drobenzofuran 3 and the 4-(hydroxypiperidinyl)phenyl deriva-
tive 6 had moderate bioavailability in this species. The 4′-
morpholino-2′-fluorophenyl derivative 7 and the 4′-dime-
thylaminophenyl derivative 40 had good bioavailability in the
dog.
Our objective was to modify the amide function in order to
further improve the bioavailability of these 2,5-diketopiperazine
oxytocin antagonists. Most amides prepared previously were
secondary amides so a range of tertiary amides were favored to

4162 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Borthwick et al.
Table 2. The Estimated Human Oral Absorption (EHOA) of Aryl Substituted (3R,6R,7R)-2,5-Diketopiperazines
5
1
8
3
7
6
9
R1
R2
CHMe2
isoBu
63
3.3
13.16
CHMe2
isoBu
60
3.4
13.51
CMe3
isoBu
49
3.4
14.58
CMe3
isoBu
46
3.4
14.85
CHMe2
R-secBu
33
3.0
15.69
CMe3
isoBu
19
2.4
16.61
CMe3
isoBu
4
3.4
18.62
EHOA%a
CHI log Db
cMR
a EHOA ) estimated human oral absorption. b An HPLC method-based measurement of lipophilicity.8 c Calculated molar refraction (Daylight v 4.81).
Table 3. The Estimated Human Oral Absorption (EHOA) of (3R,6R,7R)-4′-Fluorophenyl-2,5-diketopiperazine Amides
31
64
3.3
13.08
32
59
3.21
13.50
33
55
2.3
13.18
34
47
2.7
14.61
35
38
3.2
15.33
36
EHOA%a
CHI log Db
cMRc
32
2.0
15.54
a EHOA ) estimated human oral absorption. b An HPLC method-based measurement of lipophilicity.8 c Calculated molar refraction (Daylight v 4.81).
using the data of 52 known drug molecules is shown by eq 2.
For the 2,5-diketopiperazine series, a good correlation has been
found between CHI (MeOH) and CHI log D as well as the
number of positively chargeable groups (PosCh) values as is
described by eq 3 (see Supporting Information).
2 and 3. As bioavailability encompasses oral absorption less
those effects due to transporter efflux or to first-pass elimination,
the EHOA is the maximum possible bioavailability in the
absence of these effects. Therefore, templates with high EHOA
63-45% may have high or low bioavailability, depending on
their metabolic or efflux potential. Thus, the 2′,4′-difluorophenyl
isopropylamide 1 has high rat bioavailability 46% and a high
EHOA 60%; however, the 4′-dimethylaminophenyl isopropy-
lamide 40,4 prepared previously with a high EHOA (46%), has
low rat bioavailability (2%) possibly due to metabolism of the
dimethylamino function (Table 1). However templates with low
EHOA (<40%) would be expected to have low rat bioavail-
ability. This enabled us to rank these 2,5-diketopiperazine
oxytocin antagonists and work on those with the highest
potential for human oral absorption and avoid those templates
with low estimated human absorption. The data (Tables 2 and
3) showed that the highest EHOA would be achieved by keeping
down the size of the aryl ring or the terminal N-substituent on
the N4-glycinamide group.
EHOA% ) 0.0371 CHI (MeOH) - 18.65 cMR +
0.639 cMR2 + 103.8 (1)
n ) 46 r ) 0.73 s ) 18
EHOA% ) 1.31 CHI (MeOH) - 10.93 cMR + 88.6 (2)
n ) 52 r ) 0.81 s ) 19.7 F ) 15.9
CHI (MeOH) ) 5.51 CHI log D7.4 + 3.21PosCh + 72.0 (3)
N ) 70 r2 ) 0.91 s ) 1 F ) 346
where n is the number of compounds, r is the multiple
correlation coefficient, s is the standard error, and F is the Fisher
test values.
On the basis of eq 3 (which has been derived from the
measured and calculated data of 70 2,5-diketopiperazines), the
model described by eq 2 has been converted into eq 4, which
can be considered as a local model for this series:
Therefore, to obtain the highest estimated human oral
absorption, the priority was to focus on the smaller 2,5-
diketopiperazine templates. What was required was to avoid
large aryl rings in the exocyclic position and to choose a range
of amides which were small in size. Therefore, from the
disubstituted derivatives with good animal pharmacokinetics
(Table 1), the difluorophenyl 1 and methylenedioxy 2 templates
were chosen as starting points for a modification program and
not the larger aryl morpholine 7 even though it had 61%
bioavailability in dog. To this end first we made a range of
small amides of the 2′,4′-difluorophenyl diketopiperazine as
outlined in Table 4; all had good potency in the oxytocin binding
EHOA %) 182.9 + 7.22 CHI log D7.4 + 4.21 PosCh -
10.93 cMR (4)
We measured the CHI log D for a range of 98 2,5-
diketopiperazines and also calculated their cMRs and then
derived their estimated human oral absorption (EHOA) from
the above equation (see Supporting Information). This gave a
range of absorption values 63% to 4% as illustrated in Tables

2,5-Diketopiperazine Oxytocin Antagonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4163
Table 4. Pharmacokinetics of (3R,6R,7R) -2′,4′-Difluorophenyl-2,5-diketopiperazines in the Rat and Dog
a Displacement of [3H] oxytocin from hOTR by the test compound.7 pKi values are reported as mean values of three or more determinations and all results
were obtained in binding assays with a standard deviation in pKi of less than 0.25. b Rat PK (n ) 4): AUC (h ng mL-1) at 5 mg/kg, 5%DMSO/95%PEG400
formulation, Cl in mL min-1 kg-1 c Dog PK (n ) 3): AUC (h ng mL-1) at 0.6 mg/kg, 5% DMSO/95% PEG400 formulation, Cl in mL min-1 kg-1
. .
d Solubility (mg/mL) a precipitation/HPLC-based measurement.16 e An HPLC method-based measurement of lipophilicity.8
Table 5. Bioavailability of Heterocyclic
(3R,6R,7R)-2,5-Diketopiperazines in the Rat and Dog
assay (pKi ) 8.9-10.0). The smallest tertiary amide, the 2′,4′-
diflurophenyl N,N-dimethylamide 25 (EHOA 65%), was the best
with increased bioavailability in both rat (53%) and dog (51%)
and was the first time that good bioavailability in both species
had been obtained with this template.
The smaller secondary amide 24 had a similar pharmacoki-
netic profile to the previous lead 1 with good bioavailability in
the rat but not in the dog (Table 4). The azetidine tertiary amide
2
3
4
27 had moderate bioavailability in the dog, but only poor
bioavailability in the rat. Smaller, more polar, amides 29 and
30 showed an increase in solubility; however, while the polar
amides 28, 29 maintained good rat pharmacokinetics, they
exhibited poor pharmacokinetics in the dog, whereas the
3-hydroxy azetidine derivative 30 had poor bioavailability in
the rat. Similarly the parent acid 21R had moderate bioavail-
ability in the rat, but only poor bioavailability in the dog.
Additional SAR in this series4,15 has been developed by the
discovery of the importance of the exocyclic carbonyl for good
potency. Although the carboxylic acid 21R has the same potency
as the dimethylamide 25, the primary alcohol 39, prepared by
reduction of the methyl ester 37, resulted in a significant loss
in potency indicating that the carbonyl of these exocyclic
functions is important for good activity.
a
pKi
8.4
8.4
9.0
rat Cl/F%b
dog Cl/F%b
EHOA%c
34/48
23/14
48
39/27
14/9
46
9/68
6/95
49
a Displacement of [3H] oxytocin from hOTR by the test compound.7 pKi
values are reported as mean values of three or more determinations and all
results were obtained in binding assays with a standard deviation in pKi of
less than 0.25. b Rat (n ) 4) and dog (n ) 3) PK: Cl in mL min-1 kg-1
c EHOA ) estimated human oral absorption.
.
heteroaromatic benzofuran derivative 4, which avoided increas-
ing the overall size of the system (the estimated human oral
absorption remained the same).
The benzofuran 4 had reduced clearance in both the rat and
the dog relative to 2 and 3 and good bioavailability in both
species (Table 5). Thus, we had two leads, the 2′,4′-difluo-
rophenyl dimethylamide 25 and the benzofuran 4, with high
levels of potency (pKi), good oral exposure in the rat and the
dog, and good bioavailability in both species with low clearance.
Oxytocin Antagonist Potency and Selectivity vs Human
Vasopressin Receptors. A further measure of the potency of
our lead oxytocin antagonists 25 and 4 was established by their
Aryl modifications to the 3′,4′-methylenedioxyphenyl lead 2
were also explored in order to increase bioavailability in the
rat. Both saturated heterocyclic derivatives, the methylenedioxy
2 and the dihydrofuran 3, had high clearance in the dog and
rat, respectively (Table 5). In an attempt to minimize the
metabolic liability of these bicyclic systems, we made the

4164 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Borthwick et al.
Table 6. Inhibition of Whole Cell Oxytocin Activitya and Inhibition of
OT binding at the Human OT (hOT) and Vasopressin Binding at the
Human (V1a, V1b, and V2) Receptorsb
PK/PD comparisons of 25 with other oxytocin antagonists:
atosiban, our previous oxytocin antagonist template benzoxazine
4c (N-[(5-{[4-(2-oxo-2H-3,1-benzoxazin-1(4H)-yl)-1-piperidi-
nyl]carbonyl}-1-benzofuran-2-yl)methyl]-2-[2-oxo-5-(trifluoro-
methyl)-1(2H)-pyridinyl]acetamide),18 and Merck L-371257, in
the rat uterine contractility model (Figure 3) using the same
oxytocin challenge of 0.3 µg/mL throughout show the higher
efficacy of 25 in this species.
The ability of 25 to inhibit OT-induced calcium mobilization
in hOT-CHO cells has been demonstrated (measured using
FLIPR,20 Figure 4). The dimethylamide 25 was devoid of any
agonist activity in this sensitive system and inhibited oxytocin
(1 nM)-induced calcium mobilization with an IC50 of 3.7 nM
(estimated pKB 9.1).
Additional predevelopment studies revealed that 4 inhibited
several Cyp450 isozymes, which was not the case with 25 (Table
8). Also no time-dependent inhibition22 was observed for 25
against these five Cyp450 isozymes. The 2′,4′-diflurophenyl
dimethylamide 25 was therefore chosen as the compound for
further progression.
oxytocin
vasopressin
hOTR,
hVIaR, hVIbR, hV2R,
compd
R
R1
pKib/fpKia
pKib
pKib
pKib
25
4
2,4-diF-phenyl NMe2
5-benzofuran NHiPr
9.2/8.8
9.0/8.4
<5.2
<5.2
<5.2
<5.2
6.5
5.3
a Functional OT antagonist activity measured by FLIPR assay; see
Experimental Section.17 b Displacement of [3H] oxytocin from hOTR or
vasopressin from hV1aR, hV1bR, and hV2R by the test compound. pKi
values are reported as mean values of three or more determinations, and
all results were obtained in binding assays with a standard deviation in pKi
of less than 0.25; see ref. 4.
functional oxytocin antagonist activity (fpKi) in whole cells (by
the FLIPR assay17) and is shown together with the binding
inhibition data (pKi) against the isolated recombinant oxytocin
receptor (Table 6). In both assays, 25 is slightly more potent
than 4. Selectivity relative to human vasopressin receptors for
dimethylamide 25 and isopropylamide 4 has been established
by measuring their pKi against the hV1aR, hV2R, and hV1bR
(Table 6). Both lead oxytocin antagonists, 25 and 4, have low
affinity for all three human vasopressin receptors and OT
receptor selectivity for 25 versus these receptors is >10000 for
hV1a/hV1b and ∼500 for hV2, while for 4 it is >6300 for hV1a/
hV1b and ∼7900 for hV2 (Table 6).
Further Predevelopment Studies. The 2′,4′-diflurophenyl
dimethylamide 25 was shown not to be mutagenic in vitro in
both the bacterial (HTFT/mini-Ames assay)23 and mammalian
cell (mouse lymphoma assay)24 genotoxicity screens. A four
day oral Safety Assessment study was carried out with 25 in
female Sprague-Dawley rats at doses of 0, 30, 100, 300 (mg/
kg/day) suspended in 0.5% (w/w) hydroxypropylmethyl-cel-
lulose (HPMC) with 0.1% (w/w) Polysorbate 80 (Tween) in
sterile water. There were no adverse clinical signs, no significant
effects on body weight, food consumption, hematology, or
clinical chemistry parameters at any dose level. Histopatho-
logical examination did not reveal any treatment-related effects
at a dose up to 300 mg/kg/day.
In Vivo Potency. The in vivo efficacy of the most promising
compounds 25 and 4 was estimated using two anaesthetized
rat models, where uterine contractions were elicited by iv bolus
administration of oxytocin. In both methods, the reduction in
intensity of uterine contractions in response to oxytocin was
measured after iv bolus doses of 25 and 4. In the first method,
oxytocin was administered repeatedly at a constant dose, and
the contractions together with the associated plasma concentra-
tions of 25 and 4 were measured after increasing doses of the
antagonists, enabling the determination of a plasma IC5018 (the
concentration that caused 50% inhibition of response). In the
second method, dose-response curves to iv administration of
oxytocin were constructed after each increasing dose of 25 and
4, and a DR10 value was determined (the antagonist dose that
shifted the response to oxytocin 10-fold).4
The in vivo comparison of the two leads 25, 4 and atosiban
showed (Table 7) that the difluoro derivative 25 is of comparable
efficacy to atosiban in both models, while the benzofuran 4 is
25-38-fold less active than 25 in the DR10 assay. In contrast,
in vitro assessment using the rat receptor indicated that 25 and
4 have similar potency. Experiments carried out using the human
receptor, where the potency of the compounds was determined
with and without addition of HSA, suggested that the loss of
activity in vivo is probably due to increased plasma protein
binding. The presence of HSA reduced the activity of 4 by a
factor of >11, while only reducing the activity of 25 by
approximately 4.
Conclusion
We have developed a short, efficient, and highly stereose-
lective synthesis of a series of (3R,6R,7R)-2,5-diketopiperazine
amides. Disubstitution in the 7-aryl ring of the (3R,6R,7R)-2,5-
diketopiperazines has produced a range of potent oxytocin
antagonists with good oral bioavailablity in either the rat or dog.
Property-based design has been used to enable us to focus effort
on those templates with the greatest chance of high bioavail-
ability in humans. Estimation of human oral absorption (EHOA),
using measured lipophilicity (CHI log D) and calculated size
(cMR), allowed us to rank a range of 2,5-diketopiperazine
templates with the greatest chance of high bioavailability and
suggested that 2,5-diketopiperazine templates which were small
in size would be preferred. This led to the 2′,4′-diflurophenyl
dimethylamide 25 and the benzofuran 4 with high levels of
potency in the oxytocin binding assay (pKi), good oral exposure
in the rat and dog, and good bioavailability in both species with
low clearance (Table 9). Evaluation in vivo has shown that 25
is (>20-fold) more potent than 4 and has comparable potency
to atosiban (a marketed iv peptide oxytocin antagonist) in the
rat, while being >60-fold more potent than the latter in vitro at
the human receptor. Also 25 shows a high degree of selectivity
toward the vasopressin receptors >10000 for hV1a/hV1b and
∼500 for hV2 and has a good Cyp450 profile with no significant
inhibition (IC50 > 49µM). It was negative in the genotoxicity
screens and had a satisfactory safety profile warranting further
consideration.
Using a dose range of 0.01, 0.1, 0.5 mg/kg, the PK/PD
relationship for 25 was determined (n ) 6) by plotting the
contractile response to oxytocin (0.3 µg/mL) against the plasma
concentration fitted using an inhibitory sigmoidal Cmax model,
which gave a slope 1.36 and an IC50 of 110 ng/mL (227 nM)
(Figure 2).
Experimental Procedures
General Procedures. Melting points were obtained using an
Electrothermal digital melting point apparatus and are uncorrected.

2,5-Diketopiperazine Oxytocin Antagonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4165
Table 7. Potency of 2,5-Diketopiperazines 4, 25 and Atosiban at the Rat and Human Oxytocin Receptor Compared with the Efficacy Obtained in the
Rat Uterine Contractility Model
rat iv
compd
R
R1
DR10,a mg/kg
IC50,b nM
rOTR,c pKi
hOTR,d pKi
hOTR IC50e shift + HSA
25
4
2,4-diF-phenyl
5-benzofuran
atosiban
NMe2
NHiPr
0.22
5.4-8.4
0.63
227
-
∼186
8.0
7.8
7.2
9.2
9.0
7.4
3.6
11.2
-
a DR10 determination.4 b Plasma IC50.18 c Displacement of [3H] oxytocin from rOT by the test compound.7d Displacement of [3H] oxytocin from hOTR
by the test compound.7 pKi values are reported as mean values of three or more determinations, and all results were obtained in binding assays with a
standard deviation in pKi of less than 0.25. e HSA shift ) ratio of displacement of [3H] oxytocin from hOTR by the test compound in the presence and
absence of 50 mg/mL human serum albumin.19
Figure 2. Inhibition of uterine contractions in the rat PK/PD model
by 25.
Figure 4. Inhibition of oxytocin (1 nM)-induced calcium signal on
hOT-CHO cells.
Gradient 0%B 2 min, 0%-100% B 40 min,100% B 10 min; flow
rate ) 1 mL/min,, λ ) 215 nm). Retention times (tr) are given in
minutes. LCMS were run on a Hewlett-Packard 1050 coupled with
a Micromass Platform II equipped with a Supelco ABZplus column.
Standard conditions were eluent system A (H2O, 0.1% formic acid,
10 mmol ammonium acetate) B (MeCN, 0.05% formic acid):
gradient 1 100%A 0.7 min, 100%A to 100% B 3.5 min,100% B
3.5 min, 100% to 0%B 0.3 min; flow rate ) 1 mL/min). Gradient
2 100%A 0.7 min, 100%A to 100% B 4.2 min, 100% B 1.1 min,
100% to 0%B 0.2 min; flow rate ) 1 mL/min). Gradient 3 100%A
3 min, 100%A to 100% B 20 min, 100% B 5 min, 100% to 0%B
2 min; flow rate ) 1 mL/min). All NMR spectra were run on a
Bruker 400 MHz instrument generally as solutions in CDCl3 unless
otherwise stated. IR spectra were recorded on a Bio-rad FTS7
spectrometer from thin films on NaCl plates, a KBr mix, or solutions
in the solvent specified. Mass spectra were run on an electrospray
Hewlett-Packard 5989B instrument. CD spectra were recorded in
acetonitrile on a Jasco J-720A spectropolarimeter. Optical rotations
Figure 3. Comparison of in vivo efficacies for various oxytocin
antagonists in the rat uterine contractility model.
were taken with a Perkin-Elmer model 241 polarimeter. Enantio-
meric excess (% ee) was determined by chiral HPLC anaylsis using
a Chiracel OJ or Chiral Pak AD464 column with a UV detector λ
) 215 nm and with eluents and flow rates as indicated in each
All purifications by flash chromatography were performed using
Kieselgel 60, Merck 9385 silica gel or using a Biotage Quad 3
system with Biotage silica cartridges. Hydrophobic frits refer to
filtration tubes sold by Whatman. Preparative plate chromatography
was performed using Whatman PK6F silica gel 60A plates eluting
with ethyl acetate-cyclohexane or 2-propanol-dichloromethane
mixtures. Monitoring of reactions by TLC used Merck 60 F254
silica gel glass backed plates (5 × 10 cm), eluted with mixtures of
ethyl acetate and cyclohexane and visualized by UV light, followed
by heating with aqueous phosphomolybdic acid. Analytical HPLC
were run on a Hewlett-Packard 1090 HPLC instrument, equipped
with a Intersil M column ODS2. Standard conditions were eluent
system A (H2O, 0.1% H3PO4) B (95% MeCN/H2O, 0.1% H3PO4):
case. Final organic solutions were dried over MgSO4 before
filtration and evaporation using a Buchi Rotavapor. Ambient
temperature was 20 °C. All solvents used were Fisons analytical
reagents except for pentane (Aldrich Chemical Co.) and anhydrous
THF (Fluka sureseal). All other reagents were usually obtained from
Aldrich, Fluka, or Lancaster. Elemental microanalyses were
determined by Butterworths.
(2R)-2,3-Dihydro-1H-inden-2-yl({[(phenylmethyl)oxy]car-
bonyl}amino)ethanoic Acid (10). R-Indanylglycine4 (1.91 g, 10
mmol) was suspended in dioxane (10 mL) and water (10 mL). To
this were added triethylamine (1.7 mL) and N-(benzyloxycarbo-
nyloxy)succinimde (2.54 g), and the reaction mixture was stirred

4166 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Table 8. Cyp45021 Inhibition Profile
Borthwick et al.
IC50, µM
compd
R
R1
1A2
2C9
2C19
2D6
3A4-DEFa
3A4-PPRa
25
4
2, 4-diF-phenyl
5-benzofuran
NMe2
NHiPr
64
73
49
3.5
>100
>100
>100
89
2.5
62
4.3
62
a Diethoxyfluorescein (DEF) and 7-{3-(4-phenylpiperazin-1-ylmethyl)benzyl}resorufin (PPR) are fluorogenic substrates for Cyp3A4.
Table 9. Profiles of 2,5-Diketopiperazines Oxytocin Antagonist 25 and 4
ratb
t1/2
dogc
t1/2
in vitro
rat in vivo efficacy
Cyp450g
compd
25
pKia
F%
Cl
Vdss
F%
Cl
Vdss
IC50 shift + HSAd
DR10, emg/kg, iv
IC50, nM,f iv
IC50, µM, 2C9/3A4
9.2
9.0
53
68
15
9
1.1
2
1.2
1.4
51
95
7
6
2.1
2.4
1.1
0.9
3.6
11.2
0.22
227
>49
3.5/4.3
4
5.4-8.4
0.63
atosiban
∼186
a Displacement of [3H] oxytocin from hOTR by the test compound.7 pKi values are reported as mean values of three or more determinations, and all
results were obtained in binding assays with a standard deviation in pKi of less than 0.25. b Rat PK (n ) 4): AUC (h ng mL-1) at 5 mg/kg, 5% DMSO/95%
PEG400 formulation, Cl in mL min-1 kg-1; Vdss in L kg-1; t1/2 in h. c Dog PK (n ) 3): AUC (h ng mL-1) at 0.6 mg/kg, 5% DMSO/95% PEG400 formulation,
Cl in mL min-1 kg-1; Vdss in L kg-1; t1/2 in h. d HSA shift ) ratio of displacement of [3H] oxytocin from hOTR by the test compound in the presence and
absence of 50 mg/mL human serum albumin.19 e DR10 determination.4 f Plasma IC50.18 g Cyp45021 inhibition profile.
rapidly at room temperature for 2 days. The reaction mixture was
poured into water (50 mL) and extracted with chloroform (3 × 30
mL). The combined organic phase was washed with 1 M hydro-
chloric acid (50 mL) and water (50 mL). This was dried over
magnesium sulfate and the solvent removed in vacuo to give 10
(3.06 g, 94%): 1H NMR (CDCl3) δ 7.40-7.29 (m, 5H), 7.21-
7.11 (m, 4H), 5.28 (d, 1H, J ) 8.6 Hz), 5.11 (s, 2H), 4.57 (m, 1H),
3.14-2.79 (m, 5H); LCMS m/z 326 (MH+) single component,
Gradient 2 (tR 3.35 min); Chiral HPLC: 100.00%, (tR 11.77 min),
(Chiral HPLC (Chiralpak AD, 20% EtOH/heptane (0.1% TFA) @
215 nm) ee 100.0%, [Chiral HPLC of racemate shows clear
separation of S and R isomers: 49.9%, (tR 10.11 min) and 50.1%
(tR 11.70 min)]; HRMS calcd for for C19H19NO4 (MH+) 326.1392,
1H), 9.55 (m, 1H), 8.65, 8.56 (d, 1H, J ) 3.8 Hz), 7.90, 7.87 (d,
1H, J ) 8 Hz), 7.54-7.38 (m, 1H), 7.21 (m, 3H), 7.12 (m, 2H),
6.99-6.74 (m, 3H), 6.48, 6.16 (s, 1H), 3.88, 3.47 (m, 1H), 3.83,
3.74 (dd, 1H, J ) 4, 10 Hz), 3.07-2.94 (m, 2H), 2.91-2.71 (m,
3H), 1.77-1.54 (m, 2H), 1.46, 0.77 (m, 1H), 0.72, 0.65 (d, 3H, J
) 7 Hz), 0.64 (d, 3H, J ) 6 Hz); LCMS m/z 548 (MH+) single
component, Gradient 2 (tR 3.59 min); HRMS calcd for for
C31H31F2N3O4 (MNa+) 570.2180, found 570.2184; HPLC: 33%
(tR 14.81 min) and 66% (tR 14.93 min).
The following were similarly prepared.
(2RS)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-(2-hy-
droxyphenyl)ethanamide (18). Similarly prepared as for 17, using
4-fluorobenzaldehyde, the phenol 18 was isolated as a white solid
found 326.1387; HPLC: 99.5% (tR 13.9 min). [R]23 ) -5.5° (c
d
1
) 0.018 in MeOH, 10 cm cell).
(73%): H NMR (CD3OD) δ 7.82 (m, 1H), 7.55 (m, 2H), 7.27-
(2RS)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-
inden-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]-N-(2-hy-
droxyphenyl)ethanamide (17). To a solution of D-leucine methyl
ester hydrochloride 11 (1.45 g, 8 mmol) in methanol (10 mL) were
added triethylamine (1.12 mL) and 2,4-difluorobenzaldehyde 13
(0.875 mL, 8 mmol). The mixture was stirred for 3 days before
the acid 10 (2.6 g, 8 mmol) and 2-benzyloxyphenylisocynanide6
12 (1.76 g, 8.4 mmol) were sequentially added. The reaction mixture
was left to stand for 24 h. The solvent was removed in vacuo, and
the residue 15 was separated between ethyl acetate (200 mL) and
water (200 mL). The organic phase was washed with brine. To
this solution were added palladium on carbon (2.0 g) and acetic
acid (10 mL), and the reaction mixture was stirred under an
atmosphere of hydrogen for 2 h. The mixture was filtered through
Celite, washed with water (3 × 100 mL), saturated sodium
bicarbonate solution, and brine, and dried over magnesium sulfate.
The solvent was evaporated, and the crude product was purified
by flash column chromatography eluting with ethyl acetate-
cyclohexane (1:1 to 2:1) to give as a mixture of diastereomers the
phenol 17 (2.0 g, 46%): 1H NMR (DMSO-d6) δ 9.81, 9.76 (m,
7.15 (m, 4H), 7.10 (m, 2H), 6.98 (m, 1H), 6.81 (m, 2H), 5.89 (s,
1H), 3.95 (d, 1H, J ) 9.3 Hz), 3.74 (dd, 1H, J ) 4.5, 10.5 Hz),
3.13-2.96 (m, 3H), 2.94-2.81 (m, 2H), 1.87-1.55 (m, 3H), 0.76
(d, 3H, J ) 6.5 Hz), 0.59 (d, 3H, J ) 6.5 Hz); LCMS m/z 530
(MH+) single component, Gradient 2 (tR 3.49 min); HRMS calcd
for C31H32FN3O4 (MNa+) 552.2275, found 552.2274; HPLC: 95%
(tR 14.82 min).
(2RS)-(2,4-Difluorophenyl)[(3R,6R)-3-(2,3-dihydro-1H-inden-
2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]ethanoic Acid
(21). Carbonyldiimidazole (1.42 g, 1.6 equiv) was suspended in
anhydrous dichloromethane (10 mL), and the suspension was left
at room temperature for 15 min. The phenol 17 (3.00 g, 5.5 mmol)
was then added, and the resultant solution was stirred at room
temperature for 16 h. The resulting yellow solution was then
evaporated under reduced pressure, and the residue was treated with
a mixture of water-acetone (1:1, 10 mL). The mixture was stirred
for 2 h, then the acetone was removed under reduced pressure and
the residue was partitioned between dichloromethane (100 mL) and
0.1 M hydrochloric acid (100 mL). The organic phase was separated
using a hydrophobic frit and then evaporated to low volume and

2,5-Diketopiperazine Oxytocin Antagonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4167
purified by flash column chromatography eluting with ethyl
acetate-cyclohexane (1:1), then ethyl acetate, followed by ethyl
acetate-methanol (1:1), to give acid 21 as a mixture of diastere-
omers (2.16 g 86%): 1H NMR (CD3OD): δ 7.94 (s, 0.4H), 7.44-
7.33 (m, 1H), 7.08-6.80 (m, 7.3H), 5.92 (s, 0.7H), 5.53 (s, 0.3H),
3.93-3.88 (m, 0.6H), 3.75-3.67 (m, 1H), 3.53-3.48 (m, 0.4H),
2.97-2.56 (m, 5H), 1.81-1.71 (m, 0.4H), 1.62-1.38 (m, 2H),
0.80-0.70 (m, 0.6H), 0.62 (d, J ) 6.5 Hz, 1.1H), 0.54 (d, J ) 6.5
Hz, 1.9H), 0.49 (d, J ) 6.5 Hz, 1.1H), 0.44 (d, J ) 6.5 Hz, 1.9H);
LCMS m/z 457 (MH+) two component, Gradient 2 (tR 3.47 and
3.68 min).
These were separated on a chiral reverse-phase column (Chiralcel
OD, eluted with 15% propan-2-ol/heptane containing 0.1% TFA)
to give the 2R,3R,6R acid 21R diastereomer as a white solid (1.60
g, 64%): 1H NMR (CDCl3) δ 7.90 (m, 1H), 7.72 (m, 1H), 7.20-
7.11 (m, 4H), 6.92 (m, 1H), 6.86 (m, 1H), 5.39 (s, 1H), 4.01 (m,
1H), 3.94 (dd, 1H, J ) 3.8, 9.4 Hz), 3.15 (dd, 1H, J ) 7, 15 Hz),
3.03 (m, 2H), 2.88 (m, 1H), 2.77 (dd, 1H, J ) 9, 15 Hz), 1.92 (m,
1H), 1.85 (m, 1H), 1.55 (m, 1H), 0.91 (d, 3H, J ) 6.5 Hz), 0.88
(d, 3H, J ) 6.5 Hz); LCMS m/z 457 (MH+) single component,
Gradient 2 (tR 3.4 min); HRMS calcd for for C25H26F2N2O4 (MH+)
457.1939, found 457.1942; HPLC: 99% (tR 16.65 min).
The dimethylamide diastereomer 25 was isolated as a colorless
solid (0.285 g, 36%) indentical in all respects as that prepared from
phenol 17.
(2R)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-in-
den-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]-N,N-di-
methylethanamide (25). Carbonyldiimidazole (4.80 g, 1.54 equiv)
was suspended in anhydrous dichloromethane (40 mL), and the
suspension was left at room temperature for 15 min. The phenol
17 (10.50 g, 19.17 mmol), predried in vacuo over P4O10 for 24 h,
was then added with stirring, and the resultant solution was stirred
at room temperature for 6 h. The resulting yellow solution was
then treated with a 2.0 M solution of dimethylamine in tetrahy-
drofuran (54 mL, 5.6 equiv), and the resulting mixture was stirred
at room temperature for 16 h. The solvents plus residual dimethyl-
amine were removed under reduced pressure, and the reaction
mixture was taken up in dichloromethane (200 mL) and washed
with 1 M hydrochloric acid (200 mL). The organic phase was
separated using a hydrophobic frit and evaporated under reduced
pressure to a gum (NMR indicated that the ratio of 25 to 26 was
10:1). The crude product was purified by flash column chroma-
tography eluting with ethyl acetate-cyclohexane (2:1), then ethyl
acetate, followed by ethyl acetate-methanol (9:1) to give the
dimethylamide 25 as a colorless solid. (5.753 g, 62%). m.p 176-
The following were similarly prepared.
1
177 °C; [R]23 ) -167° (c1.0, EtOH); H NMR (CDCl3) δ 7.44
D
(2RS)-[(3R,6R)-3-(2,3-dihydro-1H-inden-2-yl)-6-(2-methylpro-
pyl)-2,5-dioxo-1-piperazinyl](4-fluorophenyl)ethanoic Acid (22).
Similarly prepared as for 21, the acid 22 as a mixture of
diastereomers was isolated as a white solid (76%): 1H NMR
(CDCl3): δ 7.79-7.72 (m, 1H), 7.46-7.05 (m, 1H), 7.08-6.80
(m, 8H), 5.49 (s, 0.3H), 5.38 (s, 0.7H), 4.07-3.90 (m, 1.7H), 3.75-
3.67 (m, 0.3H), 3.17-2.73 (m, 5H), 1.83-1.62 (m, 2H), 1.29-
1.19 (m, 1H), 0.82-0.62 (m, 6H); LCMS m/z 439 (MH+) two
component, Gradient 2 (tR 3.36 and 3.49 min). A sample (50 mg)
of this mixture was separated by preparative TLC to give the
2R,3R,6R 22R diastereomer as a white solid (27 mg): 1H NMR
(CDCl3): δ 7.89 (m, 1H), 7.42 (m, 2H), 7.20-7.05 (m, 6H), 5.44
(s, 1H), 4.02 (d, 1H, J ) 4, 9.5 Hz), 3.93 (dd, 1H, J ) 3.5, 11 Hz),
3.15-2.98 (m, 5H), 2.90-2.73 (m, 2H), 1.73 (m, 1H), 1.64 (m,
1H), 1.18 (m, 1H), 0.78 (d, 3H, J ) 6.5 Hz), 0.63 (d, 3H, J ) 6.5
Hz); LCMS m/z 439 (MH+) single component, Gradient 2 (tR 3.47
min); HRMS calcd for C25H27FN2O4 (MH+) 439.2033, found
439.2030; HPLC: 98% (tR 15.55 min).
(2S)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-in-
den-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]-N,N-di-
methylethanamide (26) and (2R)-2-(2,4-Difluorophenyl)-2-[(3R,
6R)-3-(2,3-dihydro-1H-inden-2-yl)-6-(2-methylpropyl)-2,5-dioxo-
1-piperazinyl]-N,N-dimethylethanamide (25). The acid 21 (0.724
g, 1.59 mmol), predried over P4O10 in vacuo for 5 h, was dissolved
in anhydrous dichloromethane-acetonitrile (1:1, 6 mL) and treated
with triethylamine (0.223 mL) and BOP-Cl (bis(2-oxo-3-oxazo-
lidinyl)phosphinic chloride (0.450 g, 1.77 mmol)), and the mixture
was sonicated for 1 min to give a gelatinous mass. After 10 min at
room temperature, a solution of dimethylamine in tetrahydrofuran
(10 mL of 2 M solution) was added to give a clear solution, and
this was left for 16 h at room temperature. The solvents were
removed under reduced pressure, and the mixture was partitioned
between dichloromethane and 0.1 M hydrochloric acid. The organic
phase was separated using a hydrophobic frit and evaporated under
reduced pressure. The crude product was purified by flash column
chromatography eluting with ethyl acetate-cyclohexane (1:1) then
ethyl acetate followed by ethyl acetate-methanol (9:1) to give the
dimethylamide diastereomer 26 as a colorless solid (143 mg,
18%): 1H NMR (CDCl3) δ 7.31 (m, 1H), 7.20 (m, 2H), 7.14 (m,
2H), 6.99-6.91 (m, 2H), 6.64 (d, 1H, J ) 4 Hz), 6.32 (s, 1H),
3.95 (dd, 1H, J ) 4, 10 Hz), 3.42 (m, 1H), 3.15 (m, 1H), 3.08 (m,
2H), 3.06 (s, 3H), 2.97 (m, 1H), 2.91 (s, 3H), 2.76 (m, 1H), 2.04
(m, 1H), 1.77 (m, 1H), 1.73 (m, 1H), 0.84 (d, 3H, J ) 6.6 Hz),
0.52 (d, 3H, J ) 6.6 Hz); Circular dichroism (CH3CN) λmax 226.2
nm, dE -3.90; E3457; LCMS m/z 484 (MH+) single component,
Gradient 2 (tR 3.30 min); HRMS calcd for C27H31F2N3O3 (MH+)
484.2412, found 484.2405; HPLC: 99% (tR 13.0 min).
(dt, J ) 8.5 Hz, 6.3 Hz, 1H, difluorophenyl-6H), 7.26-7.14 (m,
4H, indanyl-arylH), 7.02-6.91 (m, 2H, difluorophenyl-5H,-3H),
6.62 (broad s, 2H, NCHdifluorophenyl, CONH), 4.09 (dd, J ) 11.3
Hz, 3.0 Hz, 1H, NCHisobutyl), 3.98 (dd, J ) 10.3 Hz, 4.5 Hz, 1H,
NCHindanyl), 3.19-3.02 (m, 3H, indanyl-3H,-1H), 2.99 (s, 3H,
CONMeMe), 2.93-2.81(m, 4H, indanyl-2H, CONMeMe), 2.80-
2.72 (m, 1H, indanyl-1H), 1.61-1.51 (m, 2H, CHHCHMe2),
CH2CHMe2), 0.75-0.64 (m, 4H, CHHCHMe2, CH2CHMeMe), 0.42
(d, J ) 6.5 Hz, 3H, CH2CHMeMe); Circular dichroism (CH3CN)
λmax 205.8 nm, dE 12.04; E31054; λmax 229.2 nm, dE -14.95;
E3708; LCMS m/z 484 (MH+) single component, Gradient 2 (tR
3.43 min); HRMS calcd for C27H31F2N3O3 (MNa+) 506.2231, found
506.2225; HPLC 99% (tR 13.76 min). Anal. (C27H31F2N3O3) C, H,
N.
The dimethylamide 26 was isolated as a white solid (0.578 g,
6%) indentical in all respects as that prepared from acid 21.
The following compounds were similarly prepared.
(2R)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-in-
den-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]ethana-
mide (23). Similarly prepared as for 25, using ammonia and phenol
17, the amide 23 was isolated as a white solid (36%): 1H NMR
(CD3OD) δ 7.58 (m, 1H), 7.19 (m, 2H), 7.11 (m, 4H), 6.19 (s,
1H), 4.00 (dd, 1H, J ) 3.5, 11 Hz), 3.87 (d, 1H, J ) 10 Hz), 3.13-
2.93 (m, 3H), 2.87-2.74 (m, 2H), 1.67-1.52 (m, 2H), 0.90 (m,
1H), 0.71 (d, 3H, J ) 6.5 Hz), 0.53 (d, 3H, J ) 6.5 Hz); LCMS
m/z 456 (MH+) single component, Gradient 2 (tR 3.30 min); HRMS
calcd for C25H27F2N3O3 (MNa+) 478.1918, found 478.1928; HPLC
99% (tR 12.86 min). Anal. (C25H27F2N3O3) C, H, N.
(2R)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-in-
den-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]-N-meth-
ylethanamide (24). Similarly prepared as for 25, using methylamine
and phenol 17, the amide 24 was isolated as a white solid (41%):
1H NMR (CDCl3) δ 7.71 (dt, 1H, J ) 6.5, 8.5 Hz), 7.60 (d, 1H, J
) 4 Hz), 7.18 (m, 2H), 7.14 (m, 2H), 6.95 (m, 1H), 6.89 (m, 1H,
J ) 2.5, 8.5, 11 Hz), 6.12 (q, 1H, J ) 5 Hz), 5.45 (s, 1H), 3.99
(dd, 1H, J ) 3, 10 Hz), 3.93 (dd, 1H, J ) 4, 9 Hz), 3.14 (dd, 1H,
J ) 7, 15 Hz), 3.03 (d, 2H, J ) 8 Hz), 2.89 (m, 1H), 2.85 (d, 3H,
J ) 5 Hz), 2.80 (dd, 1H, J ) 8.5, 15 Hz), 1.84 (m, 1H), 1.79 (m,
1H), 1.46 (m, 1H), 0.87 (d, 3H, J ) 6.5 Hz), 0.80 (d, 3H, J ) 6.5
Hz); LCMS m/z 470 (MH+) single component, Gradient 2 (tR 3.36
min); HRMS calcd for C26H29F2N3O3 (MNa+) 492.2075, found
492.2062; HPLC 99% (tR 13.13 min).
(3R,6R)-1-[(1R)-2-(1-Azetidinyl)-1-(2,4-difluorophenyl)-2-oxo-
ethyl]-3-(2,3-dihydro-1H-inden-2-yl)-6-(2-methylpropyl)-2,5-pip-
erazinedione (27). Similarly prepared as for 25, using azetidine
and phenol 17, the amide 27 was isolated as a white solid (21%):
1H NMR (CDCl3) δ 7.65 (m, 1H), 7.24-7.11 (m, 4H), 6.99 (m,

4168 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Borthwick et al.
1H), 6.91 (m, 1H), 6.80 (d, 1H, J ) 4 Hz), 6.05 (s, 1H), 4.32 (m,
1H), 4.14 (m, 1H), 4.06 (m, 1H), 4.02-3.93 (m, 2H), 3.75 (m,
1H), 3.19-3.03(m, 3H), 2.86 (m, 1H), 2.77 (m, 1H), 2.34-2.14
(m, 2H), 1.63 (m, 1H), 1.55 (m, 1H), 0.87 (m, 1H), 0.70 (d, 3H, J
) 6.5 Hz), 0.45 (d, 3H, J ) 6.5 Hz); LCMS m/z 496 (MH+) single
component, Gradient 2 (tR 3.53 min); HRMS calcd for C28H31F2N3O3
(MNa+) 518.2231, found 518.2241; HPLC 99% (tR 13.81 min);
Anal. (C28H31F2N3O3_0.5H2O) C, H, N.
droxyethyl)ethanamide (33). Similarly prepared as for 25, using
2-aminoethanol and phenol 18, the amide 33 was isolated as a white
solid (54%): 1H NMR (CDCl3) δ 7.45 (m, 2H), 7.23-7.09 (m,
6H), 7.05 (d, 1H, J ) 3.5 Hz), 6.05 (t, 1H, J ) 5.5 Hz), 4.96 (s,
1H), 3.97 (dd, 1H, J ) 4, 10 Hz), 3.87-3.81 (m, 2H), 3.65-3.53
(m, 2H), 3.23-3.13 (m, 2H), 3.03 (m, 2H), 2.90 (m, 1H), 2.79
(dd, 1H, J ) 9, 15 Hz), 1.89 (m, 1H), 1.83 (m, 1H), 1.64 (m, 1H),
0.89 (d, 3H, J ) 6.5 Hz), 0.83 (d, 3H, J ) 6.5 Hz); LCMS m/z
482 (MH+) single component, Gradient 2 (tR 3.02 min); HRMS
calcd for C27H32FN3O4 (MH+) 482.2455, found 482.2459; HPLC
>98% (tR 12.22 min).
(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-1-[(1R)-1-(4-fluorophe-
nyl)-2-(4-methyl-1-piperazinyl)-2-oxoethyl]-6-(2-methylpropyl)-
2,5-piperazinedione (34). Similarly prepared as for 25, using
N-methylpiperazine and phenol 18, the amide 34 was isolated as a
white solid (34%): 1H NMR (CDCl3) δ 7.41 (m, 2H), 7.25-7.12
(m, 6H), 6.54 (s, 1H), 6.31 (d, 1H, J ) 4 Hz), 4.12 (m, 1H), 3.98
(dd, 1H, J ) 4.5, 10.5 Hz), 3.68 (m, 2H), 3.36 (m, 1H), 3.21-3.01
(m, 4H), 2.87 (m, 1H), 2.75 (dd, 1H, J ) 9, 15 Hz), 2.41 (m, 1H),
2.30 (m, 2H), 2.23 (s, 3H), 1.95 (m, 1H), 1.53 (m, 1H), 1.43 (m,
1H), 0.62 (d, 3H, J ) 6.5 Hz), 0.57 (m, 1H), 0.41 (d, 3H, J ) 6.3
Hz); LCMS m/z 521 (MH+) single component, Gradient 2 (tR 2.62
min); HRMS calcd for C30H37FN4O3 (MH+) 521.2928, found
521.2919; HPLC 98% (tR 10.13 min).
(2R)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-methyl-
N-(2-pyridinylmethyl)ethanamide (35). Similarly prepared as for
25, using methyl(2-pyridinylmethyl)amine and phenol 18, the amide
35 was obtained as a white solid (19%): 1H NMR (CDCl3) δ 8.50
(m, 1H), 7.67 (m, 1H), 7.16, 7.48 (m, 2H), 7.25-7.06 (m, 8H),
6.50, 6.48 (s, 1H), 6.31 (m, 1H), 4.74, 4.58, 4.34 (m, 2H), 4.15
(m, 1H), 4.02 (m, 1H), 3.22-3.04 (m, 3H), 3.02, 2.88 (s, 3H),
2.94-2.73 (m, 2H), 1.55 (m, 1H), 1.47 (m, 1H), 0.67 (m, 1H),
0.63 (m, 3H), 0.39 (m, 3H); LCMS m/z 543 (MH+) single
component, Gradient 2 (tR 3.28 min); HRMS calcd for C32H35FN4O3
(MH+) 543.2771, found 543.2769; HPLC 99% (tR 13.63 min).
(2R)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-[2-(1-pi-
peridinyl)ethyl]ethanamide (36). Similarly prepared as for 25,
using 2-(1-piperidinyl)ethanamine and phenol 18, the amide 36 was
isolated as a white solid (30%): 1H NMR (CD3OD) δ 7.46 (m,
2H), 7.14-7.07 (m, 4H), 7.02 (m, 2H), 5.64 (s, 1H), 3.91 (dd, 1H,
J ) 3.5, 11 Hz), 3.81 (d, 1H, J ) 9.5 Hz), 3.33-3.17 (m, 2H),
3.02-2.88 (m, 3H), 2.74 (m, 2H), 2.39-2.28 (m, 6H), 1.54-1.41
(m, 6H), 1.35 (m, 2H), 0.97 (m, 1H), 0.62 (d, 3H, J ) 6.5 Hz),
0.43 (d, 3H, J ) 6.3 Hz); LCMS m/z 549 (MH+) single component,
Gradient 2 (tR 2.63 min); HRMS calcd for C32H41FN4O3 (MH+)
549.3241, found 549.3237; HPLC 98% (tR 10.09 min).
Methyl (2R)-(2,4-Difluorophenyl)[(3R,6R)-3-(2,3-dihydro-1H-
inden-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]etha-
noate (37) and Methyl (2S)-(2,4-Difluorophenyl)[(3R,6R)-3-(2,3-
dihydro-1H-inden-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piper-
azinyl]ethanoate (38). Carbonyldiimidazole (0.324 g, 1.6 equiv)
was suspended in anhydrous dichloromethane (4 mL), and the
suspension was left at room temperature for 15 min. The phenol
17 (0.800 g, 1.46 mmol) was then added with stirring, and the
resultant solution was left at room temperature for 16 h. The mixture
was then treated with methanol (10 mL) and left at room-
temperature overnight. The solvents were removed under reduced
pressure, and the residue was purified by preparative plate chro-
matography eluting with ethyl acetate-cyclohexane (1:3) to give
the 2R,3R,6R methyl ester 37 (0.453 g, 66%): 1H NMR (CDCl3)
δ 7.69-7.61 (m, 1H, difluorophenyl-6H), 7.26-7.15 (m, 4H,
indanyl-arylH), 6.95-6.82 (m, 2H, difluorophenyl-5H,-3H), 6.70
(broad d, J ) 3.5 Hz, 1H, CONH), 5.32 (s, 1H, NCHdifluorophe-
nyl), 3.98-3.91 (m, 2H, NCHisobutyl, NCHindanyl), 3.81 (s, 3H,
CO2Me), 3.20-2.74 (m, 5H, indanyl-3H, -1H, -2H), 1.99-1.80 (m,
2H, CHHCHMe2, CH2CHMe2), 1.59-1.51 (m, 1H, CHHCHMe2),
0.93 and 0.89 (2d, J ) 6.5 Hz, 6H, CH2CHMe2); LCMS m/z 471
(MH+) single component, Gradient 2 (tR 3.42 min); HRMS calcd
(3R,6R)-1-[(1R)-1-(2,4-Difluorophenyl)-2-(4-morpholinyl)-2-
oxoethyl]-3-(2,3-dihydro-1H-inden-2-yl)-6-(2-methylpropyl)-2,5-
piperazinedione (28). Similarly prepared as for 25, using mor-
pholine and phenol 17, the amide 28 was isolated as a white solid
(61%): 1H NMR (CDCl3) δ 7.43 (dt, 1H, J ) 6, 8.5 Hz), 7.25-
7.13 (m, 4H), 7.01 (m, 1H), 6.96 (m, 1H), 6.89 (d, 1H, J ) 4 Hz),
6.63 (s, 1H), 4.06 (m, 1H), 3.99 (dd, 1H, J ) 4.5, 10 Hz), 3.76-
3.56 (m, 5H), 3.39 (m, 1H), 3.30 (m, 1H), 3.19-3.01(m, 4H), 2.87
(m, 1H), 2.77 (m, 1H, J ) 9, 15 Hz), 1.56 (m, 2H), 0.72-0.63 (m,
4H), 0.43 (d, 3H, J ) 6.5 Hz); LCMS m/z 526 (MH+) single
component, Gradient 2 (tR 3.35 min); HRMS calcd for C29H33F2N3O
(MNa+) 548.2337, found 548.2357; HPLC 95% (tR 13.54 min).
(2R)-2-(2,4-Difluorophenyl)-2-[(3R,6R)-3-(2,3-dihydro-1H-in-
den-2-yl)-6-(2-methylpropyl)-2,5-dioxo-1-piperazinyl]-N-(2-hy-
droxyethyl)-N-methylethanamide (29). Similarly prepared as for
25, using 2-(methylamino)ethanol and phenol 17, the amide 29 was
isolated as a white solid (54%): 1H NMR (CDCl3) δ Rotamers
7.63-7.48 (m, 1H), 7.26-7.12 (m, 4H), 7.04-6.84 (m, 3H), 6.61,
6.57 (s, 1H), 4.07-3.53 (m, 6H), 3.19-2.99 (m, 3H), 3.00, 2.92
(s, 3H), 2.90-2.71 (m, 2H), 1.67-1.51 (m, 2H), 0.96-0.76 (m,
1H), 0.69 (m, 3H), 0.40, 0.34 (d, 3H, J ) 6.3 Hz); LCMS m/z 514
(MH+) single component, Gradient 2 (tR 3.27 min); HRMS calcd
for for C28H33F2N3O4 (MH+) 514.2517, found 514.2523; HPLC:
98% (tR 13.08 min).
(3R,6R)-1-[(1R)-1-(2,4-Difluorophenyl)-2-(3-hydroxy-1-aze-
tidinyl)-2-oxoethyl]-3-(2,3-dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-piperazinedione (30). Similarly prepared as for 25,
using azetidin-3-ol25 and phenol 17, the amide 30 was isolated as
a colorless solid (45%): 1H NMR (CDCl3) δ Rotamers 7.58 (m,
1H), 7.25-7.10 (m, 4H), 7.00 (m, 1H), 6.92 (m, 1H), 6.11, 6.05
(s, 1H), 4.70-4.15 (m, 4H), 4.05-3.92 (m, 3H), 3.84, 3.59 (m,
1H), 3.23-2.99 (m, mH), 2.90-2.74 (m, 2H), 1.65-1.50 (m, 2H),
0.90, 0.74 (m, 1H), 0.69 (m, 3H), 0.47, 0.42 (d, 3H, J ) 6.3 Hz);
LCMS m/z 512 (MH+) single component, Gradient 2 (tR 3.2 min);
HRMS calcd for for C28H31F2N3O4 (MH+) 512.2361, found
512.2370; HPLC: 98% (tR 12.85 min).
(2R)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-(2,2,2-tri-
fluoroethyl)ethanamide (31). Similarly prepared as for 25, using
(2,2,2-trifluoroethyl)amine and phenol 18, the amide 31 was isolated
as a white solid (27%): 1H NMR (CDCl3) δ 7.41 (m, 2H), 7.23-
7.10 (m, 6H), 6.74 (d, 1H, J ) 3.5 Hz), 6.45 (t, 1H, J ) 6.5 Hz),
5.14 (s, 1H), 4.06-3.82 (m, 4H), 3.16 (dd, 1H, J ) 7, 14.5 Hz),
3.04 (m, 2H), 2.89 (m, 1H), 2.78 (dd, 1H, J ) 9, 14.5 Hz), 1.85
(m, 1H), 1.77 (m, 1H), 1.47 (m, 1H), 0.87 (d, 3H, J ) 6.5 Hz),
0.82 (d, 3H, J ) 6.5 Hz); LCMS m/z 520 (MH+) single component,
Gradient 2 (tR 3.48 min); HRMS calcd for C27H29F4N3O3 (MH+)
520.2223, found 520.2220; HPLC: 99% (tR 14.2 min).
(2R)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-(1-meth-
ylethyl)ethanamide (32). Similarly prepared as for 25, using
isopropylamine and phenol 18, the amide 32 was isolated as a white
solid (18%): 1H NMR (CDCl3) δ 7.46-7.41 (m, 2H), 7.25-7.08
(m, 6H), 6.51 (d, 1H, J ) 3.5 Hz), 6.60 (d, 1H, J ) 7.5 Hz), 5.10
(s, 1H), 4.09 (m, 1H), 3.98-3.91 (m, 2H), 3.16 (dd, 1H, J ) 7, 15
Hz), 3.07 (m, 2H), 2.90 (m, 1H), 2.87 (dd, 1H, J ) 8.5, 15 Hz),
1.83 (m, 1H), 1.73 (m, 1H), 1.43 (m, 1H), 1.12 (d, 6H, J ) 6.5
Hz), 0.84 (d, 3H, J ) 6.5 Hz), 0.79 (d, 3H, J ) 6 Hz); LCMS m/z
480 (MH+) single component, Gradient 2 (tR 3.31 min); HRMS
calcd for C28H34FN3O3 (MH+) 480.2662, found 480.2657; HPLC
99% (tR 13.85 min).
(2R)-2-[(3R,6R)-3-(2,3-Dihydro-1H-inden-2-yl)-6-(2-methyl-
propyl)-2,5-dioxo-1-piperazinyl]-2-(4-fluorophenyl)-N-(2-hy-

2,5-Diketopiperazine Oxytocin Antagonists
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14 4169
for C26H29F2N2O4 (MH+) 471.20899, found 471.20956; HPLC 98%
(tR 14.46 min).
(4) Borthwick, A. D.; Davies, D. E.; Exall, A. M.; Livermore, D. G.;
Sollis, S. L.; Nerozzi, F.; Allen, M. J.; Perren, M.; Shabbir, S. S.;
Woollard, P. M.; Wyatt, P. G. 2,5-Diketopiperazines as potent,
selective and orally bioavailable oxytocin antagonists 2: Synthesis,
chirality and pharmacokinetics. J. Med. Chem. 2005, 48, 6956-
6969.
(5) Davies, D. E. Stereospecific synthesis of 2,5-diketopiperazines
Tetrahedron Lett. 2006, 47, in press.
(6) 2-Benzyloxyphenylisocynanide isonitrile was prepared by the method
of Obrecht, R.; Herrmann, R.; Ugi, I. Isocyanide synthesis with
phosphoryl chloride and diisopropylamine. Synthesis 1985, 4, 400-
402.
(7) For human OT binding assay method, see the following: Wyatt, P.
G.; Hickin, G.; Miller, N. D.; Allen, M. J.; Chilcott, J.; Woollard, P.
M. Structure-Activity relationship investigations of a potent and
selective benzodiazepine oxytocin antagonist. Bioorg. Med. Chem.
Lett. 2001, 11, 1301-1305. Human V1a, V1b, and V2 binding assay
method is identical to the hOT method except 10 µg of V1a-CHO
membrane was used and 0.5 nM [3H] vasopressin was used as ligand
(Kd ) 0.79 nM), 10 µg of V1b-CHO membrane was used and 1 nM
[3H] vasopressin was used as ligand (Kd ) 3.9 nM), and 10 µg of
V2-CHO membrane was used and 0.5 nM [3H] vasopressin was used
as ligand (Kd ) 3.8 nM).
(8) Chromatographic hydrophobicity index (CHI) utilizes the gradient
retention time from rapid gradient reverse-phase elution to measure
lipophilicity. See the following: Valko, K.; Du, C. M.; Bevan, C.;
Reynolds, D. P.; Abrahams, M. H. Rapid method for the estimation
of octanol/water partition coefficient (log Poct) from gradient RP-
HPLC retention and a hydrogen bond acidity term (ΣR2H). Curr.
Med. Chem. 2001, 8, 1137-1146.
(9) (a) Van De Waterbeemd, H.; Jones, B. C. Predicting oral absorption
and bioavailability. Prog. Med. Chem. 2003, 41, 1-59. (b) Lajiness,
M. S.; Vieth, M.; Erickson, J. Molecular properties that influence
oral drug-like behaviour Curr. Opin. Drug Discuss. DeV. 2004, 7,
470-477.
(10) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeny, P. Experi-
mental and computational approaches to estimate solubility and
permeability in drug discovery and development settings. AdV. Drug
DeliVery ReV. 1997, 23, 3-25.
The 2S,3R,6R methyl ester 38 was isolated as a colorless solid
(0.180 g, 26%): 1H NMR (CDCl3) δ 7.37-7.29 (m, 1H, difluo-
rophenyl-6H), 7.26-7.15 (m, 4H, indanyl-arylH), 7.97-6.87 (m,
2H, difluorophenyl-5H,-3H), 6.46 (broad d, J ) 3.3 Hz, 1H,
CONH), 5.78 (s, 1H, NCHdifluorophenyl), 4.02 (dd, J ) 9.5 Hz,
4.3 Hz, 1H, NCHindanyl), 3.81 (s, 3H, CO2Me), 3.67 (dd, J )
10.3 Hz, 3.8 Hz, 1H, NCHisobutyl), 3.20-2.75 (m, 5H, indanyl-
3H,-1H, -2H), 1.80-1.61 (m, 3H, CH2CHMe2, CH2CHMe2), 0.81
and 0.67 (2d, J ) 6.3 Hz, 6H, CH2CHMe2); LCMS m/z 471 (MH+)
single component, Gradient 2 (tR 3.42 min); HRMS calcd for
C26H29F2N2O4 (MH+) 471.20899, found 471.20953; HPLC 95%
(tR 14.27 min).
(3R,6R)-1-[(1R)-1-(2,4-Difluorophenyl)-2-hydroxyethyl]-3-
(2,3-dihydro-1H-inden-2-yl)-6-(2-methylpropyl)-2,5-piperazinedi-
one (39). The methyl ester 37 (360 mg, 0.765 mmol) was dissolved
in anhydrous tetrahydrofuran ((10 mL), and the solution was stiirred
under nitrogen at -78 °C. Lithium triethylborohydride (2.30 mL,
3 equiv) of a 1.0 M solution in tetrahydrofuran was added dropwise
via syringe over 1 min to the solution, which was stirred at -78
°C for 4 h then allowed to warm to room temperature. After 30
min at room temperature, the mixture was again cooled to -78 °C
and quenched by addition of saturated aqueous ammonium chloride
(5 mL) added via syringe over 1 min to give a white suspension.
This was allowed to warm to room temperature with stirring and
was left at room-temperature overnight. The mixture was diluted
with dichloromethane (100 mL), and the organic phase was
separated using a hydrophobic frit and evaporated under reduced
pressure to give the crude product which was purified by flash
column chromatography eluting with dichloromethane, then ethyl
acetate-cyclohexane (1:2), followed by ethyl acetate-cyclohexane
(1:1) and then ethyl acetate. The ethyl acetate fractions were
combined and evaporated under reduced pressure to give the alcohol
39 as a white solid (190 mg, 56% yield): 1H NMR (CDCl3) δ
7.56-7.49 (m, 1H, difluorophenyl-6H), 7.26-7.15 (m, 4H, indanyl-
arylH), 6.93-6.77 (m, 3H, difluorophenyl-5H,-3H, CONH), 4.76
(dd, J ) 7.3 Hz, 4.5 Hz, 1H, NCHdifluorophenyl), 4.41-4.32 and
4.15-4.06 (2m, 2H, difluorophenylCHCH2), 3.96 (dd, J ) 9.5 Hz,
4.0 Hz, 1H, NCHindanyl), 3.76 (dd, J ) 10.0 Hz, 6.0 Hz, 1H,
NCHisobutyl), 3.50-3.44 (m, 1H, OH), 3.20-2.99 and 2.92-2.75
(2m, 5H, indanyl-3H,-1H, -2H), 1.98-1.80 (m, 2H, CHHCHMe2,
CH2CHMe2), 1.63-1.55 (m, 1H, CHHCHMe2), 0.95 and 0.94 (2d,
J ) 6.5 Hz, 6H, CH2CHMe2); LCMS m/z 443 (MH+) single
component, Gradient 2 (tR 3.21 min); HRMS calcd for C25H29F2N2O3
(MH+) 443.21408, found 443.21427; HPLC 98% (tR 13.37 min).
(11) Clark, D. E. Rapid calculation of polar molecular surface area and
its application to the prediction of transport phenomena. 1. Prediction
of intestinal absorption. J. Pharm. Sci. 1999, 88, 807-814.
(12) (a) Abraham, M. H.; Zhao, Y. H.; Le, J.; Hersey, A.; Eddershaw, P.
J.; Luscombe, C. N.; Reynolds, D. P.; Beck, G.; Sherborne, B.;
Cooper, I. On the mechanism of human intestinal absorption. Eur.
J. Med. Chem. 2002, 37, 595-605. (b) Zhao, Y. H.; Le, J.; Abraham,
M. H.; Hersey, A.; Eddershaw, P. J.; Luscombe, C. N.; Butina, D.;
Beck, G.; Sherborne, B.; Cooper, I.; Platts, J. A. Evaluation of human
intestinal absorption data and subsequent derivation of a quantitative
structure-activity relationship (QSAR) with the Abraham descriptors.
J. Pharm. Sci. 2001, 90, 749-784. (c) Zhao, Y. H.; Abraham, M.
H.; Le, J.; Hersey, A.; Luscombe, C. N.; Beck, G.; Sherborne, B.;
Cooper, I. Rate-limited steps of human oral absorption and QSAR
studies. Pharm. Res. 2002, 19, 1446-1457.
(13) Du, C. M.; Valko, K.; Bevan, C.; Reynolds, D.; Abraham, M. H.
Characterizing the selectivity of stationary phases and organic
modifiers in reversed-phase high-performance liquid chromatographic
systems by a general solvation equation using gradient elution. J.
Chromatog. Sci. 2000, 38, 503-511.
(14) Valko, K. Measurements and predictions of physicochemical proper-
ties. In High-Throughput ADMETox Estimation; Darvas, F.; Dormain,
G., Eds.; Eaton Publishing: Westbrough, MA, 2002; Chapter 1, pp
1-24.
(15) Wyatt,P. G.; Allen, M. J.; Borthwick, A. D.; Davies, D. E.; Exall,
A. M.; Hatley, R. J. D.; Irving, W. R.; Livermore, D. G.; Miller, N.
D.; Nerozzi, F.; Sollis, S. L.; Szardenings, A. K. 2,5-Diketopipera-
zines as potent and selective oxytocin antagonists 1: Identification,
stereochemistry and initial SAR. Bioorg. Med. Chem. Lett. 2005, 15,
2579-2582.
Acknowledgment. We thank Alan Hill, Structural and
Computational Science, Discovery Research, GlaxoSmithKline
Research and Development, Stevenage, for Log D measurements
and analysis.
Supporting Information Available: C, H, N analysis for
compounds 4, 23, 24, 25, and 27 and the CD spectra of 25 and 26.
Experimental and spectroscopic data for compounds 2-9. Lipo-
philicity calculations and measurements, absorption models, and
estimated human oral absorption (EHOA) tables. This material is
available free of charge via the Internet at http://pubs.acs.org.
(16) Valko, K. Separation Methods in Drug Synthesis and Purification.
In Handbook of Analytical Separations; Valko, K., Ed.; Elsevier:
Amsterdam, 2000; Vol. 1, Chapter 12, pp 535-583.
References
(1) Goldenberg, R. L.; Rouse, D. Prevention of premature Birth. N. Engl.
J. Med. 1998, 339, 313.
(17) Determination of antagonist affinity at human oxytocin-1 receptors
using FLIPR. Cell Culture. Adherent Chinese Hamster Ovary (CHO)
cells, stably expressing the recombinant human oxytocin-1 (hOT)
receptor, were maintained in culture in DMEM:F12 medium (Sigma,
cat. no. D6421), supplemented with 10% heat inactivated fetal calf
serum (Gibco/Invitrogen, cat. no. 01000-147), 2 mM L-glutamine
(Gibco/Invitrogen, cat. no. 25030-024) and 0.2 mg/mL G418 (Gibco/
Invitrogen, cat. no. 10131-027). Cells were grown as monolayers
under 95%:5% air:CO2 at 37 °C and passaged every 3-4 days using
TrypLE Express (Gibco/Invitrogen, cat no. 12604-013). Measure-
(2) Enkin, M.; Kierse, M.; Neilson, J.; et al. Preterm Labour: A Guide
to EffectiVe Care in Pregnancy and Childbirth, 3rd ed.; Oxford
University Press: Oxford, U.K., 2000; pp 211-225.
(3) (a) Bossmar, T. Treatment of preterm labor with the oxytocin and
vasopressin antagonist atosiban. J. Perinat. Med. 1998, 26, 458-
465. See also (b) Coomarasamy, A.; Knox, E. M.; Gee, H.; Khan,
K. S. Oxytocin antagonists for tocolysis in preterm labour - a
systematic review. Med. Sci. Monit. 2002, 8, RA268-
RA273.

4170 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 14
Borthwick et al.
ment of [Ca2+]i using the FLIPR. CHO-hOT cells were seeded into
black-walled clear-base 384-well plates (Nunc) at a density of 10000
cells per well in culture medium as described above and maintained
overnight (95%:5% air:CO2 at 37 °C). After removal of culture
medium, cells were incubated for 1 h at 37 °C in Tyrode’s medium
(NaCl, 145 mM; KCl, 2.5 mM; HEPES, 10 mM; glucose, 10 mM.;
MgCl2, 1.2 mM; CaCl2, 1.5 mM) containing probenacid (0.7 mg/
mL), the cytoplasmic calcium indicator, Fluo-4 (4 µM; Teflabs,
Austin, TX) and the quenching agent Brilliant Black (250 µM;
Molecular Devices, UK). Cells were then incubated for an additional
30 min at 37 °C with either buffer alone or buffer containing OT
antagonist, before being placed into a FLIPR (Molecular Devices,
UK) to monitor cell fluorescence (λex ) 488 nm, λEM ) 540 nm)
before and after the addition of a submaximal concentration of
oxytocin (EC80). Data Analysis. Functional responses using FLIPR
were analysed using Activity Base Version 5.0.10.
(20) Agonist responses were measured according to ref 17 except that
test compound was added to cells in FLIPR reader in place of
oxytocin EC80.
(21) Microsomes, from human lymphoblastoid cell lines transfected with
cDNA for specific cytochrome P450 enzymes, were obtained from
Gentest Corporation (Woburn, MA). Reaction progress was moni-
tored using fluorescent substrates. Each test compound was dissolved
in methanol at 5 mM and serially diluted to determine IC50 values.
A positive control was included for each isozyme.
(22) Time-dependent inhibition assays were preformed as described in
ref 21, except that the fluorescence in each well was measured for
30 min at 1 min intervals. Each assay plate also contained a CYP3A4
time-dependent positive control.
(23) Reifferscheid, G.; Heil, J.; Oda, Y.; Zahn, R. K. A microplate version
of the SOS/umu-test for rapid detection of genotoxins and genotoxic
potentials of environmental samples. Mutat. Res. 1991, 253, 215-
222.
(24) Clive, D.; Caspary, W.; Kirby, P.; Krehl. R.; Moore, M.; Mayo, J.;
Oberly, T. J. Guide for performing the mouse lymphoma assay for
mammalian cell mutagenicity. Mutat. Res. 1987, 189, 143-
156.
(25) Prepared by the method of Chatterjee, S. S. and Triggle, D. J.
Synthesis of azetidin-3-ol. J. Chem. Soc., Chem. Commun. 1968, 2,
93.
(18) Wyatt, P. G.; Allen, M. J.; Chilcott, J.; Gardner, C. J.; Livermore,
D. G.; Mordaunt, J. E.; Nerozzi, F.; Patel, M.; Perren, M. J.;
Weingarten, G. G.; Shabbir, S.; Woollard, P. M.; Zhou, P. Identifica-
tion of potent and selective oxytocin antagonists. Part 2: further
investigation of benzofuran derivatives. Bioorg. Med. Chem. Lett.
2002, 12, 1405-1411.
(19) Ratio of the IC50 hOT in the presence of 50 mg/mL HSA to IC50
hOT in the absence of 50 mg/mL HSA. Method as in ref 7. The
ligand [3H]oxytocin was made up in buffer containing 100 mg/mL
HSA. HSA was obtained from Sigma (Poole, Dorset, U.K.).
JM060073E