Improvement in the selectivity and metabolic stability of the serotonin 5-HT1A ligand, S 15535: A series of cis- and trans-2-(arylcycloalkylamine) 1-indanols

Article abstract of DOI:10.1021/jm010975+

S 15535 (1) displays a distinctive profile of agonist and antagonist (weak partial agonist) activity at pre- and postsynaptic 5-HT_1A receptors, respectively. It has proven to be active in several models predictive of anxiolytic, antidepressant,

Full text of DOI:10.1021/jm010975+

J . Med. Chem. 2002, 45, 165-176
165
Im p r ovem en t in th e Selectivity a n d Meta bolic Sta bility of th e Ser oton in 5-HT1A
Liga n d , S 15535: A Ser ies of cis- a n d tr a n s-2-(Ar ylcycloa lk yla m in e) 1-In d a n ols
,
†
†
†
†
†
‡
J ean-Louis Peglion,* Bertrand Goument, Nicole Despaux, Val e´ rie Charlot, H e´ l e` ne Giraud, Christian Nisole,
§
§
⊥
#
§
Adrian Newman-Tancredi, Anne Dekeyne, Marc Bertrand, Patrick Genissel, and Mark J . Millan
Institut de Recherches Servier, 11 rue des Moulineaux, 92150 Suresnes et 125 Chemin de Ronde, 78290 Croissy-sur-Seine,
France, Technologie Servier, 25-27 rue Eug e` ne Vignat, 45007 Orl e´ ans, France, and Servier Research and Developement,
Fulmer Hall, Windmill Road, Fulmer, Slough SL3 6HH, Berks, England
Received J uly 4, 2001
S 15535 (1) displays a distinctive profile of agonist and antagonist (weak partial agonist) activity
at pre- and postsynaptic 5-HT1A receptors, respectively. It has proven to be active in several
models predictive of anxiolytic, antidepressant, and procognitive properties. In an attempt to
increase its selectivity and metabolic stability, and guided by the results of human metabolic
studies, we prepared a series of cis- and trans-2-(arylcycloalkylamine) 1-indanols. Irrespective
of the nature of the arylcycloalkylamine moiety or the presence of substituents on the indanol
ring, trans isomers invariably showed the highest affinity for human, recombinant h5-HT1A
receptors. Among them, compounds 39, 42, 45, 49, 52, 53, 54, 57, 61, 64, 67, and 70 displayed
similar or higher affinity than the parent compound 1 (pK
-adrenoceptors has been frequently encountered with 5-HT1A ligands. While S 15535 itself
presents reasonable selectivity (158-fold) in this respect, trans piperazine derivatives 4-tr a n s,
5, 39, 41, 47, 64, 68, 69, 70, 71 displayed even more pronounced selectivity vs R -adrenoceptors,
i
g 9.1). Lack of selectivity toward
R
1
3
1
with the nitro derivative 70 being highly selective (1259-fold). However, among the set of trans
piperidines prepared, only 64, which also bears a nitro on the indanol ring, displayed selectivity
greater than the parent compound 1. All trans derivatives behaved as partial agonists at h5-
3
5
HT1A receptors, as determined by their submaximal stimulation of [ S]GTPγS binding to a
level comparable to that observed with S 15535. In metabolic stability studies in vitro using
human microsomes and hepatocytes, only trans piperazines and, in particular, 35, 39, 41, 68,
6
3
9, and 70, showed an improvement relative to 1, whereas trans piperidines did not. Compounds
5, 39, 41, and 70, which combined both improved selectivity and metabolic stability, and which
retained the distinctive pharmacological characteristics of S 15535, were evaluated in animal
models of anxiety. Of these, 35, which showed the highest oral bioavailability in vivo in rats,
was resolved into its two isomers 36 and 37. The eutomer 37 displayed 47% oral bioavailability
in the rat and was potently active (0.1-0.5 mg/kg, s.c.) in the rat ultrasonic vocalization and
social interaction models, predictive of anxiolytic activity. In conclusion, 2-(arylcycloalkylamine)
1
-indanols represent a novel class of potent 5-HT1A ligands in which the presence of the hydroxyl
group in the benzylic position enhances selectivity, while substituents on the phenyl ring of
the indanol moiety improve both selectivity and metabolic stability.
In tr od u ction
AR) (158-fold), a characteristic seldom encountered in
1
0a,b
In a previous article,1
5
-HT1A receptors antagonists.
S 15535 (1), a benzodioxopiperazine derivative,1-3 is
we provided full details concerning the structure-
activity relationships which led to the selection of 1 for
further development and clinical evaluation.Unfortu-
nately, during the course of clinical trials, it was
established that S 15535 is extensively metabolized.
Several circulating hydroxylated metabolites were iden-
tified: the 5-hydroxyindane 3 was the most abundant,
whereas the 1-hydroxyindane 4 was present only in
small amounts. However, in a radioreceptor assay from
plasma of treated patients, only 4 appeared to have
marked affinity for 5-HT1A receptors.To study their
pharmacology and metabolic stability in detail, we
prepared both diastereoisomers of 4. The cis isomer of
active in diverse, experimental models predictive of
4
5,6
7
antidepressant, anxiolytic, and procognitive proper-
ties, and displays a distinctive profile of agonist and
antagonist (weak partial agonist) properties at pre- and
8
postsynaptic serotonin 5-HT1A receptors, respectively.
9
In analogy to the selective and pure 5-HT1A antagonist,
WAY 100635 (2), S 15535 shows high affinity for human
5
-HT1A receptors (pKi ) 9.1) and reasonable selectivity
over dopamine D2 (195-fold) and R1-adrenoceptors (R1-
*
To whom correspondence should be addressed: J ean-Louis
Peglion, Institut de Recherches Servier, 11 rue des Moulineaux, 92150
Suresnes. Tel: 33 1 55 72 22 13. Fax: 33 1 72 55 24 30. E-mail: jean-
louis.peglion@fr.netgrs.com.
4
displayed weak affinity for human serotonin h5-HT1A
†
Department of Chemistry B (Suresnes-France).
Department of Psychopharmacology (Croissy-sur-Seine-France).
Department of Analytical Chemistry (Suresnes-France).
Department of Biopharmaceutical Research (Orl e´ ans-France).
§
receptors (pKi ) 7.2), while 4-tr a n s retained almost the
same affinity as the parent compound 1. Moreover, in
vitro, metabolic stability studies using human hepatic
microsomes and hepatocytes revealed an improvement
‡
⊥
#
Department of Pharmacokinetics and Metabolism (Fulmer-
England).
1
0.1021/jm010975+ CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/05/2001

1
66 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
Ch a r t 1
illustrated in Scheme 1 by a classical set of reactions.
4
-Piperidinol 50 was prepared by an alternative route
depicted in Scheme 2. Because of their relative instabil-
ity, indanone intermediates V (Table 1) had to be
rapidly reduced to indanols by sodium borohydride,
affording a mixture of cis and trans isomers where the
cis isomer was always present in a much larger propor-
tion than the trans isomer (about 85/15), and the
workup procedure did not allow isolation of the trans
isomer in sufficient amounts to permit pharmacological
evaluation. Consequently, we sought a strategy where
the ratio of trans isomer could be significantly improved.
We developed, thus, a simple and quick method starting
from 2-bromo 1-indanones IV, which is characterized
by the ready generation of stable pyridinium ions VI
as key intermediates (Table 1). Reduction of VI by
sodium borohydride in methanol afforded a 50/50 mix-
ture of cis- and trans-1,2,5,6-tetrahydropyridine deriva-
tives I, which could be either separated by flash
chromatography to lead to the above-mentioned com-
pounds, or further reduced by catalytic hydrogenation
to give piperidine derivatives 52, 53, 55-59, 61-63
Ch a r t 2
(Scheme 3).
4
-Arylpyridines VIII A-E used in the first step of this
in the stability of 4-tr a n s relative to 1. Encouraged by
the observation that the indanol backbone could also
lead to potent and more stable 5-HT1A ligands, we
hypothesized that blockade of the probable metabolic
synthesis were obtained through a Suzuki reaction as
depicted in Scheme 4, starting from the corresponding
aryl triflate or aryl bromide VII.
However, this synthetic pathway applied only to
piperidines and tetrahydropyridines. Moreover, we pre-
sumed that, in comparison to 4-cis and 4-tr a n s, trans
derivatives would be more potent. Consequently, we
attempted to develop a diastereospecific and general
method (Scheme 5) for exclusively obtaining the trans
isomer of piperidines and piperazines. When epoxides
IX were treated by aryl cyclic amines IIA-IIL, a
regiospecific ring opening occurred. This led to trans
-indanols X (Table 1), which were isomerized into the
ester of trans 1-indanols XI by a Mitsunobu reaction at
25 °C in the presence of p-nitrobenzoic acid under
conditions described in the literature,
(hydroxylation) position of the phenyl ring of the indanol
moiety of 4 might generate novel ligands of superior
metabolic stability while retaining the original phar-
macological profile of S 15535. Surprisingly, few 1-in-
danols substituted by an arylpiperazine in position 2
are described in the literature.11 This prompted us to
investigate a series of cis and trans compounds with the
general structure I shown in Chart 2, where the
arylpiperazine group IIA could be extended to closed
analogues IIB-IIL such as shown in Chart 3.
In the present paper, we describe the synthetic
methodologies used for the preparation of compounds
of structure I and examine structure-activity relation-
ships based on the results of binding studies at human
2
-
1
2a,b
providing a
diastereomeric excess greater than 99%. Saponification
of the resulting p-nitrobenzoate XI led to compounds
2, 43, 51, 64-71 (Table 2). The mechanism of this
isomerization reaction could involve an aziridinium ion,
5
-HT1A receptors. We also assessed selectivity versus
4
native, rat R1-adrenoceptors and, for the most selective
ligands, efficacy at human 5-HT1A receptors, as well as
antagonist properties in the rat in vivo. In parallel, we
evaluated oral bioavailability prediction from in vitro
metabolic stability (human hepatic microsomes) and
intestinal permeability (Caco-2 cell line). Because of
possible primary metabolic pathways activated by phase
II conjugation enzymes, the evaluation was completed
for several key compounds, by assessing metabolic
stability using human hepatocytes and by determination
of oral bioavailability in the rat.
1
3
12b
as suggested by Freedman and Sol a` .
In regard to
the hindered nature of the secondary alcohols X, rather
unusual Mitsunobu conditions had to be employed.14
Assignment to either cis or trans configuration was
made by means of IR spectra as described by Rimek.15
In addition, X-ray crystallography was performed on
compound 4-cis and 4-tr a n s to verify attributions made
by the IR method.
Biology
Ch em istr y
The biological characterization of compounds pre-
pared in the present work was based partly upon
experimental strategies discussed previously.1 Affini-
ties at h5-HT1A receptors and R1-adrenoceptors were
By introducing an hydroxyl function on the benzylic
position of S 15535, we created geometrical and optical
isomers and were confronted with the problem of either
designing synthetic routes able to produce regiospecifi-
cally cis or trans isomers, or finding appropriate experi-
mental conditions generating an equilibrated mixture
of both isomers.
,2
3
assessed by inhibition of [ H]-8-OH-DPAT binding in
Chinese hamster ovary cells line stably expressing
1
6
3
recombinant h5-HT1A receptors and inhibition of [ H]-
prazosin binding to rat cortex, respectively. Efficacy of
compounds at h5-HT1A sites were determined by mea-
suring 5-HT1A receptor-linked G-protein activation by
Piperazines 4-cis, 4-tr a n s, 34-41, 46, 47, and pip-
eridines 44, 45 listed in Table 2 were prepared as

Selectivity of the Serotonin 5-HT1A Ligand, S 15535
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1 167
Ch a r t 3
Ta ble 1. Key Intermediates. Indanones V and VI, 2-Indanols
X (Preparation and Physical Properties)
8-OH-DPAT in rats was employed as a measure of
antagonist properties at postsynaptic 5-HT1A receptors.
Resu lts
Str u ctu r e-Activity Rela tion sh ip s. Concerning af-
finities at h5-HT1A receptors, the tendency observed
with compounds 4-cis and 4-tr a n s was confirmed with
the other couples of cis/trans isomers (Tables 2 and 3),
irrespective of the nature of the cyclic amine moiety or
the presence of substituents on the indanol ring. That
is, trans isomers invariably showed the highest affinity.
With respect to the influence of the heterocyclic
aromatic moiety upon affinity at 5-HT1A receptors (Table
4
), there were only minor variations in affinity when
we compared benzodioxan derivatives 41 and 45 to their
respective benzopyran analogues 47 and 67, and when
a chlorine atom was introduced into the benzene ring
of the benzodioxan part of 41 to give 71. However, a
marked decrease was seen when a fluorine atom was
introduced on the benzopyran ring of 67 leading to 59.
A reduction of affinity was also obtained when a
dihydrobenzofuran ring, as in 62, replaced the benzopy-
ran ring of 67. Finally, the most marked loss of affinity
was observed in the piperazine series upon transforma-
tion of the benzodioxan moiety of 41 to 66, as under-
1
8
taken by Hartog and colleagues with flesinoxan. As
for compound 1, the benzodioxan ring conferred maxi-
mal affinity at h5-HT1A sites.
On the other hand, the affinities of benzodioxan
derivatives at h5-HT1A receptors (Table 5) were only
slightly affected by variation of the cyclic amine com-
ponent inasmuch as piperazines 4-tr a n s and 41 dis-
played almost similar affinity to tetrahydropiperidines
4
9 and 54, and piperidine analogues 53 and 45 showed
only a slight increase in affinity. However, a piperidinol
was deleterious for affinity since compound 51 exhibited
a pKi < 6. Major differences were seen with respect to
the nature of the substituent on the indanol moiety
(
6
)
Table 6). In the piperazine series, a methoxy in position
, as for compound 39, conferred maximal activity (pKi
9.35) and a 5,6-difluoro lower activity (pKi ) 8.67 for
a
All melting points were determined on a Reichert Thermovar
compound 68). In the piperidine series, a 6-nitro (com-
pound 64) afforded the highest affinity among the
compounds prepared in this study, in contrast to the
5,6-dimethoxy analogue 63 (pKi ) 7.1), which showed
the lowest affinity of the series.
apparatus and are uncorrected.
[35
S]GTPγS binding.17a,b The ability of compounds to
inhibit hypothermia elicited by s.c. administration of

1
68 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
Ta ble 2. Preparation, Physical Properties, and Affinities at h5-HT1A Receptors of Compounds I
a
Resolution of 35 by HPLC on chiral column. b See corresponding footnote Table 1. c Compounds were purified by column
chromatography; C, H, and N analyses were within 0.4% of theoretical values for the formulas given, unless otherwise stated. All compounds
d
e
exhibited NMR consistent with assigned structures. C: % calculated ) 68.09, % found ) 67.39. N: % calculated ) 6.44, % found )
7
.06. f pKi ( SEM values are from two to five independent experiments. g N. T. ) not tested.
It is of importance to note that compounds 39, 42, 45,
9, 52, 53, 54, 57, 61, 64, 67, and 70 displayed similar
was responsible for decreased selectivity since affinity
at h5-HT1A receptors was not modified.
4
or higher affinity than the parent compound 1 (Table
).
Concerning selectivity toward R1-adrenoceptors, the
unsubstituted or monosubstituted benzodioxan seemed
to confer the highest selectivity, as illustrated by
compounds 41 and 71 (Table 4), but disubstitution led
to a decrease of selectivity mainly due to a decrease in
affinity at h5-HT1A receptors, since in 66 affinity at R1-
adrenoceptors was not affected. On the other hand, in
the benzopyran analogue 47, a rise in affinity at R1 sites
The high selectivity shown by 41 and 71 was also due
to the presence of the piperazine ring since the 1,2,5,6-
tetrahydropyridine ring, as in 54, or the piperidine ring,
as in 45, led to a decrease of selectivity due to an
increase in R1-AR affinity. Similar observations was
made upon comparing the piperazine derivative 4-tr a n s
with its 1,2,5,6-tetrahydropyridine analogue 49 and its
piperidine analogue 53, respectively (Table 5), or pip-
erazine 47, with its piperidine homologue 67 (Table 4).
Substitution on the phenyl ring of the indanol moiety
2

Selectivity of the Serotonin 5-HT1A Ligand, S 15535
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1 169
Sch em e 1. Synthesis of Piperazine and Piperidine Derivatives I (Method A)a
a
Reagents and conditions: (a) nBu4NBr3, CH2Cl2, MeOH, R.T.; (b) K2CO3, DMF, R.T.; (c) NaBH4, THF, R.T.
Sch em e 2. Synthesis of 50 (Method A′)a
a
Reagents and conditions: (a) K2CO3, DMF, R.T.; (b) NaBH4, THF, R.T.; (c) HCl concentrated, AcOH, 50 °C, 24 h; (d) THF, R.T., then
reflux 1 h.
Sch em e 3. Synthesis of 1,2,5,6-Tetrahydropyridine and Piperidine Derivatives I (Method B)a
a
Reagents and conditions: (a) methylethyl ketone or acetone, reflux; (b) 1. NaBH4, methanol, 2. AcOH; (c) H2/PtO2, MeOH, THF.
apparently played a limited role with regard to selectiv-
ity (Table 6), except in the case of a nitro substitution
in position 6 which appears crucial since piperazine 70
and piperidine 64 displayed the highest selectivity
among piperazine and piperidine series, respectively,
more than double those seen with their unsubstituted
analogues 4-tr a n s and 53, their 5-fluoro analogues 41
and 45, or their 6-fluoro analogues 69 and 52. Finally,
all piperazines listed in Table 6 and piperidine 64
showed a higher affinity than the parent compound 1.
Biologica l Resu lts. In analogy to the parent com-
pound 1, all ligands examined (Table 7) behaved as
partial agonists at h5-HT1A receptors expressed in CHO
cells, as determined by their submaximal stimulation
3
5
of [ S]GTPγS binding to a level comparable to that
observed with S 15535, and markedly inferior to 5-HT
itself (defined as a full agonist, 100% stimulation). No
systematic differences in this regard were noted be-
tween trans piperazines and trans piperidines. This
modest level of efficacy is sufficient to activate highly
sensitive 5-HT1A receptors but not low-sensitivity
1
9-21
postsynaptic 5-HT1A receptors.
Correspondingly, in
analogy to S 15535, both trans piperazine and trans
piperidine derivatives displayed pronounced and com-

1
70 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
Sch em e 4. Synthesis of Arylpyridines VIIIa
deprived of functional phase II enzymes under the
incubation conditions used, the apparent improvement
in metabolic stability was confirmed using human
hepatocytes that present both functional microsomal
enzymes and phase II conjugation enzymes. Metabolic
bioavailability predictions using human hepatocytes
were confirmed to be very much in line with the results
of human hepatic microsomes for compounds 35, 39, and
4
1 (74, 62, and 58% vs 80, 62, and 65%). Such compari-
son also confirmed the low metabolic stability seen with
4
5, 52, 53, and 64 (15, 9, 10, and 22% vs 7, 8, 7 and
2
2% for human hepatocytes and human hepatic mi-
crosomes, respectively).
Discu ssion a n d Con clu sion
It is of interest to note that I-tr a n s compounds led
not only to the discovery of a novel family of antagonists/
partial agonists at postsynaptic 5-HT1A receptors but
also to the identification, with regard to piperidines 64
and 67, of compounds with affinities among the highest
ever obtained at these sites. In contrast, cis-I isomers
displayed much lower affinities than their trans ana-
logues. This result can lead to a further comprehension
of the structural requirements for optimal recognition
of 5-HT1A receptors. Molecular modeling studies are
currently underway to define which interactions are
decisive for high recognition and to explain the poor
affinity of I-cis isomers in comparison to I-tr a n s
analogues.
a
Reagents and conditions: (a) Na2CO3, Pd(PPh3)4, toluene,
ethanol, reflux; (b) nBu4NBr, KOH, Pd(PPh3)4, THF.
Sch em e 5. Synthesis of Trans Piperazine and
Piperidine Derivatives I (Method C)a
Among the numerous points that may be considered
regarding the structure-activity study, one particularly
intriguing aspect concerns the comparison between the
behavior of benzopyran derivatives 59 and 67, on one
hand, and benzodioxan derivatives 71 and 41, on the
other. In the former case, the presence of a fluorine led
to a decrease in affinity of 16-fold, but in the latter case,
a chlorine had no effect (Table 4). Interestingly, Kuipers22
made a similar observation: incorporation of a fluorine
atom into the 5-position of the benzofuran cycle of 1-(7-
benzofuranyl)piperazine led to a decrease in affinity of
1
0-fold, whereas a bromine atom at the same position
a
Reagents and conditions: (a) CH3CN, R.T.; (b) DEAD, PPh3,
caused only a 2.5-fold decrease. This observation led the
authors to speculate that steric effects play only a minor
role at this 5-position. In extension to their analysis,
we may hypothesize that an electronic effect on the
para-oxygen atom explains the influence of the fluorine
atom, in 59 and in Kuiper’s derivative. In general, this
oxygen atom is frequently found to confer high 5-HT1A
affinities in arylpiperazine derivatives (cf. compounds
1 and 2). The strong electronegative effect of a fluorine
atom may sufficiently modify the electronic environment
of a para-oxygen atom to induce a marked decrease in
affinity. In contrast, a chlorine (or a bromine) atom, with
its electronegative effect less pronounced than the
fluorine, does not reduce the electronic density on the
oxygen atom enough to cause a noticeable change in
affinity.
p-nitrobenzoic acid, toluene, THF, -25 °C; (c) KOH, CH3OH, H2O,
R.T.
parable activity in vivo in blocking the induction of
hypothermia by 8-OH-DPAT, with the exception of
compounds of 52, 64, and 68. This paradigm similarly
revealed antagonist properties of the trans 1,2,4,5-
tetrahydropyridines at postsynaptic 5-HT1A receptors.
Although trans piperazines and trans piperidines dis-
played, thus, very similar profiles as concerns their
affinities and functional activities at 5-HT1A receptors
and showed high Caco-2 permeability corresponding to
total oral absorption prediction (Table 7), they differed
markedly as concerns their metabolic stability. Thus,
the former ligands, with the exception of 4-tr a n s,
showed relatively substantial stability upon incubation
with human microsomes (60 to 80% of metabolic stabil-
ity prediction), whereas the latter revealed little im-
provement relative to S 15535.
Lack of selectivity toward R1-adrenoceptors is a major
consideration in the design of 5-HT1A ligands,23 and in
the resolution of this difficulty, diverse strategies have
2
3
As all these compounds include an hydroxyl group
that may be metabolized by phase II conjugation
enzymes (UDP-glucuronyl transferases and/or sulfo-
transferases), and inasmuch hepatic microsomes are
been adopted. For example, Hibert, on the basis of
graphical computer-generated maps of 5-HT1A receptors
and R1-adrenoceptors, increased the size of the aromatic
moiety of the 20-fold selective compound MDL 72832

Selectivity of the Serotonin 5-HT1A Ligand, S 15535
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1 171
Ta ble 3. Affinities at h5-HT1A Receptors: Comparison of Cis and Trans Compounds I
a
See corresponding footnote of Table 2.
Ta ble 4. Influence of the Nature of the Heterocyclic Moiety
upon Affinity at h5-HT1A Receptors and Selectivity vs
Ta ble 5. Influence of the Nature of the Central Amino Ring
upon Affinity at h5-HT1A Receptors and Selectivity vs
R1-Adrenoceptors for Trans Piperazines and Trans Piperidines I
R1-Adrenoceptors for Trans Isomers I
arylpiperazines bearing a 3-(diphenylmethylene or a
9
H-fluoren-9-ylidene) 2,5-pyrrolidinedione-1-yl group in
the N-4 position, demonstrated the preponderant role
for selectivity played by the length of the tether between
this group and the piperazine ring. In previous work,1
we showed that, when the substituent in the N-4
position of arylpiperazines is an indan-2-yl, reasonable
selectivity could be attained. In the present work, we
exploited this 4-(indan-2-yl) 1-arylpiperazine backbone
in structure I to further increase selectivity. Three
to obtain a 140-fold selective derivative, which poorly
recognized R1-adrenoceptors. Bolognesi,24 in introducing
a bulky tether into the mixed 5-HT1A/R1-AR ligand, WB
4
101, was able to prepare a selective (468-fold) 5-HT1A
2
5
ligand. Lopez-Rodriguez, working with a series of

1
72 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
Ta ble 6. Influence of the Nature of (R)x Substituents upon
Affinity at h5-HT1A Receptors and Selectivity vs
R1-Adrenoceptors for Trans Piperazines and Trans Piperidines I
Third, substitution of the phenyl component of thein-
danol ring may play a major role, in certain cases, in
influencing both affinity and selectivity. Thus, for both
piperazines and piperidines, a 6-nitro substitution
simultaneously conferred high affinity at 5-HT1A sites
and low affinity at R1-AR sites, a property not shared
by any other substituent and leading, in the case of 70,
to a pronounced improvement of selectivity in compari-
son to 1. Compounds 4-tr a n s, 35, 39, 41, 64, 68, 69 also
showed high selectivity, although less marked than 70,
underlining the key role of the hydroxyl function as the
main structural feature underlying this increased se-
lectivity in comparison to 1 (Table 6). The choice of the
4
-(indan-2-yl) 1-arylpiperazine backbone seemed judi-
cious since appropriate substitution on the indan ring
generated selective 5-HT1A ligands.
In addition, the present series of derivatives retained
the distinctive pharmacological characteristics of 1
inasmuch as they behaved as partial agonists at cloned
h5-HT1A receptors (Table 7), thereby preserving agonist
properties at 5-HT1A autoreceptors (unpublished obser-
vations) and antagonist actions at postsynaptic sites
(Table 7).
Concerning metabolic stability in human microsomes,
the hydroxyl function in the benzylic position appeared
insufficient to afford in itself a superior pharmacokinetic
profile than 1, since 4-tr a n s presented only a modest
improvement in metabolic stability. Further, replace-
ment of the piperazine ring of 4-tr a n s by a 1,2,5,6-
tetrahydropyridine ring as in 49 or by piperidine ring
as in 53 did not result in an improvement in stability
(Table 7). Even heteroaryl groups different to benzo-
dioxan were not able to confer improved metabolic
stability (data not shown). In the tr a n s-I piperidine
series, mono or disubstitutions on the aromatic part of
the indanol ring did not stabilize these compounds
toward metabolic attack, in contrast to the clear im-
provement shown by substituted piperazines 35, 39, 41,
and 70 and, to a lesser extent by 68 and 69 (Table 7).
These observations were confirmed with hepatocytes in
which possible hydroxyl conjugation could be taken into
account. Together with the satisfactory absorption
predicted by high permeability in the in vitro Caco-2
model, the data suggest that these compounds should
display good oral bioavailability in man.
regions of molecules I were of significance concerning
selectivity for 5-HT1A receptors versus R1-adrenoceptors.
First, the heteroaryl region had a greater influence on
selectivity than affinity, even if affinity at 5-HT1A
receptors was somewhat dependent on the presence of
substituents on the ring, since the fluorine atom of 59
and the chlorine atom plus the hydroxymethyl group of
To summarize, the presence of an hydroxyl group in
the benzylic position of S 15535 increased its selectivity,
and substituents on the phenyl ring of the indanol
moiety enhanced both selectivity and metabolic stabil-
ity. Compounds 35, 39, 41, and 70, which combined both
improvements, were evaluated in animal models of
6
6 were responsible for the decreased 5-HT1A affinity
of their respective analogues 67 and 41. However, as
far as selectivity is concerned, marked differences were
noted even for minor structural variations, for example,
in comparing the benzodioxan 41 (selectivity 562) to its
benzopyran analogue 47 (selectivity 166), or 71 versus
6
anxiety and depression. Of these, 35, which showed a
marked increase in oral bioavailability relative to 1 in
in vivo metabolic studies in the rat, was resolved into
its two isomers 36 and 37. 37 showed oral bioavailability
in the rat of 47% and was potently active (0.1-0.5 mg/
kg, s.c.) in rat the ultrasonic vocalization and social
6
6 (selectivity of 645 vs 64). Second, the cyclic amino
region also played a major role in determining selectiv-
ity, mainly reflecting its influence upon R1-AR affinity.
Piperazines 4-tr a n s and 41 were far more selective than
their tetrahydropyridine analogues 49 and 54 and their
piperidine analogues 53 and 45 (Table 5). The presence
of a nitrogen atom linked to the phenyl ring of the
heteroaromatic part was of great importance for selec-
tivity, since it markedly suppressed R1-AR affinity, but
only lowered 5-HT1A receptor affinity to a lesser extent.
6
interaction models predictive of anxiolytic activity. 41
and 70, which also demonstrated robust activity in the
ultrasonic vocalisation test, are currently being resolved.
In conclusion, this study demonstrates that structural
modification of S 15535 results in compounds retaining
its distinctive profile of interaction with 5-HT1A recep-
tors and pronounced anxiolytic properties, yet present-

Selectivity of the Serotonin 5-HT1A Ligand, S 15535
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1 173
Ta ble 7. Receptor Binding Affinities, Efficacies, Antagonist Properties at Postsynaptic 5-HT1A Receptors and in Vitro Oral
Bioavailability Predictions of Trans Piperazines, Trans Piperidines, and Trans 1,2,5,6-Tetrahydropyridines I
8
-OH-DPATb
pKi
[35S] GTPγSa
(n ))
hypothermia ID50
mg/kg s.c. (95% C. L.)
MF % human
e
Abs %f
Caco-2 cells
compd
h5-HT1A
microsomes
trans piperazines
4
8.79 ( 0.09
8.90 ( 0.26
9.35 ( 0.13
8.90 ( 0.01
8.67 ( 0.13
8.88 ( 0.25
9.13 ( 0.29
52 ( 2% (3)
21 ( 2% (3)
31 ( 7% (2)
60 ( 5% (3)
54 ( 1% (2)
41% (1)
0.96 (0.26-3.52)
0.72 (0.21-2.51)
1.25 (0.37-4.27)
1.42 (0.45-4.51)
21
80
62
65
51
46
71
100
100
100
100
100
100
100
3
3
4
6
6
7
5
9
1
8
9
0
c
∼2.50 (N.C.)
1.00 (0.40-2.70)
0.90 (0.10-6.20)
49 ( 2% (2)
trans piperidines
4
5
5
5
6
6
6
5
2
3
8
1
4
5
9.36 ( 0.31
9.11 ( 0.01
9.15 ( 0.16
8.66 ( 0.11
9.28 ( 0.05
9.54 ( 0.17
8.96 ( 0.31
50% (1)
45% (1)
47% (1)
45% (1)
24% (1)
54% (1)
N. T.
N. T.d
7
8
7
5
100
100
100
100
100
100
2.10 (0.80-5.80)
0.50 (0.20-1.60)
5.00 (1.50-16.5)
1.00 (0.60-1.60)
8.10 (2.20-30.5)
0.60 (0.20-1.80)
8
22
N. T.d
N. T.
d
trans 1,2,5,6-tetrahydropyridines
1.30 (0.40-4.00)
4
5
9
4
9.08 ( 0.19
9.07 ( 0.05
N. T.
N. T.
4
8
100
100
1.70 (0.50-5.30)
1
1
00
00
1
2
9.10 ( 0.09
9.30 ( 0.21
39 ( 2% (2)
0% (3)
1.44 (0.72-2.87)
0.01 (0.003-0.03)
4
1
100
100
a
Agonist efficacies (compared to 5-HT at 5-HT1A receptors) ( SEM values are from two to three independent experiments, except
when n ) 1. ID50 ) inhibitory dose50; n g 5 per dose tested. c N. C. ) not computable. N. T. ) not tested. e MF % ) metabolic
b
d
f
bioavailability prediction; experiment were performed once (100% ) no metabolism). Abs % ) oral absorption prediction; experiments
were performed in duplicate and led to the same prediction (100% ) total intestinal absorption)
ing a marked improvement in selectivity and metabolic
stability. Of these, the optical isomers of 35 are under
more intensive pharmacological evaluation.
washed with water, dried (MgSO
reduced pressure. The residue obtained (14 g) was chromato-
graphed (CH Cl /MeOH, 97/3). The first compound eluted (9
g) was recrystallized from CH CN: yield 7.4 g (35.5%) mp
96-198 °C and identified to be 4-cis by its IR spectrum. IR
4
), and concentrated under
2
2
3
1
(
Exp er im en ta l Section
) γmax (cm ): 3380 (br, OH). 1H NMR (300 MHz,
-
1
CHCl
DMSO, δ): 2.65 (m, 2H, CHH pip), 2.8 (m, 3H, CHH pip +
CHN indan), 2.93 (m, 2H, CH indan), 3.04 (m, 4H, CH pip),
4.23 (m, 4H, OCH CH O), 4.52 (d, 1H, OH) 4.87 (t, 1H, CHOH),
3
Ch em istr y. Reactions performed in nonaqueous solvents
were carried out under an atmosphere of nitrogen. Column
chromatography was carried out with Merck Kieselgel 60
2
2
2
2
(
230-400 mesh) under a nitrogen pressure of 0.5 atm. Mi-
6.48 (d, 1H, Bzd H-6), 6.51 (d, 1H, Bzd H-8), 6.73 (t, 1H, Bzd
H-7), 7.21 (m, 1H, arom), 7.26 (d, 2H, arom), 7.39 (d, 1H, arom).
Confirmation of the cis configuration was done by X-ray
crystallography.
croanalyses were performed on solid samples only, with a
Carlo-Erba autoanalyzer. IR spectra were recorded on a
Bruker IFS 28 infrared spectrometer. H NMR spectra were
1
recorded on either a Bruker AC 200 or Bruker AM 300
spectrometer at 200 and 300 MHz, respectively. Chemical
shifts are reported as δ values in parts per million (ppm)
relative to tetramethylsilane (δ, 0.00) used as internal stan-
dard. Assignment to either cis or trans configuration was made
The second compound eluted (1.3 g) was washed with hot
acetonitrile: yield 1.1 g (5.4%); mp 211-214 °C and identified
-
1
1
to be 4-tr a n s. IR (CHCl ) γmax (cm ): 3580 (w, OH). H NMR
3
(300 MHz, DMSO, δ): 2.8-3.25 (cluster of 11H, 8H pip + 1H,
CHN indan, + 2H, CH indan), 4.25-4.35 (m, 4H, OCH CH O),
2
2
2
1
5
by means of IR spectra as described by Rimek. Cis com-
pounds, which form intramolecular hydrogen bonds, showed
5.22 (d, 1H, CHOH), 6.55 (d, 1H, Bzd H-8), 6.6 (d, 1H, Bzd
H-6), 6.8 (t, 1H, Bzd H-7), 7.22 (m, 3H, arom), 7.4 (m, 1H,
arom). Confirmation of the trans configuration was done by
X-ray crystallography.
-1
in solution, a broad hydroxyl band around 3400 cm , whereas
trans compounds where the hydroxyl function is free, showed
-1
a weak band around 3600 cm . Optical activity was measured
at 20 °C with a Perkin-Elmer 241 polarimeter.
cis-2-[4-(2,3-Dih yd r oben zo [1,4] d ioxin -5-yl) p ip er a zin -
1-yl] 5-m eth oxyin d a n -1-ol (34) a n d tr a n s-2-[4-(2,3-Dih y-
d r oben zo [1,4] d ioxin -5-yl) p ip er a zin -1-yl] 5-m eth oxyin -
d a n -1-ol (35). Prepared as described for 4-cis and 4-tr a n s
starting from compound 6.
(+)-tr a n s-2-[4-(2,3-Dih yd r oben zo [1,4] d ioxin -5-yl) p ip -
er a zin -1-yl] 5-m eth oxyin d a n -1-ol (36) a n d (-)-tr a n s 2-[4-
(2,3-Dih ydr oben zo [1,4]dioxin -5-yl)piper azin -1-yl] 5-m eth -
oxyin d a n -1-ol (37). Compound 35 (3 g) was resolved in its
enantiomers on a Chiralpak AD column by HPLC (eluting
system: EtOH/diethylamine, 1000/1). The first enantiomer
Gen er al Meth od A. 2-[4-(2,3-Dih ydr oben zo [1,4] dioxin -
5
-yl) p ip er a zin -1-yl] 1-in d a n on e (5). A suspension of 2-bro-
mo 1-indanone (15.9 g, 75 mmol), compound II-A (19.3 g, 75
mmol), and potassium carbonate (10.4 g, 75 mmol) in dimeth-
ylformamide (100 mL) was stirred at room temperature for
4
8 h. The mixture was poured in water (750 mL) and extracted
by AcOEt (2 × 200 mL). After the samples were washed, the
combined extracts were dried (MgSO ) and concentrated under
4
reduce pressure to yield 26.1 g of a violet oil (100%).
cis-2-[4-(2,3-Dih yd r oben zo [1,4] d ioxin -5-yl) p ip er a zin -
-yl] in d a n -1-ol (4-cis) a n d tr a n s 2-[4-(2,3-Dih yd r oben zo
1,4] d ioxin -5-yl) p ip er a zin -1-yl] in d a n -1-ol (4-t r a n s).
Sodium borohydride (2.2 g, 4.3 mmol) was added by portion,
at room temperature, to a solution of compound 5 (20 g, 57
mmol) in tetrahydrofuran (170 mL). At the end of the addition,
the mixture was stirred overnight and evaporated to dryness,
and the residue was taken up in methylene chloride (200 mL),
eluted (880 mg) was recrystallized from CH
(20%) mp 149-152 °C; [R ] +12.5° (c 0.5, MeOH) and identified
to be (+)-36. IR (CHCl ) γmax (cm ): 3500 (w, OH). H NMR
(300 MHz, CDCl , δ): 2 (br m, 1H, OH), 2.75-3.1 (cluster of
7H, 4H pip + 1H, CHN indan, + 2H, CH indan), 3.15 (m,
4H, 4H pip), 3.8 (s, 3H, OMe), 4.3 (m, 4H, OCH CH O), 5.2 (d,
3
CN: yield 0.61 g
1
[
D
-
1
1
3
3
2
2
2
1H, CHOH), 6.55 (m, 2H, Bzd H-6 and H-8), 6.75 (m, 3H, 1H
Bzd H-7 + 2H arom), 7.25 (d, 1H, arom).

1
74 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
The second compound eluted (870 mg) was recrystallized
mmol) in acetone (35 mL) was heated under reflux for 24 h.
On cooling (ice), a precipitate appeared that was filtered off,
washed with cold acetone, and dried on vacuum under P O :
2 5
from CH
3
CN: yield 0.62 g (21%) mp 151-153 °C; [R
D
] -12.1°
3
(
(
1
+
(
c 0.5, MeOH) and identified to be (-)-37. IR (CHCl
) γmax
-
1
1
cm ): 3500 (w, OH). H NMR (300 MHz, CDCl
H, OH), 2.75-3.1 (cluster of 7H, 4H pip + 1H, CHN indan,
2H, CH indan), 3.15 (m, 4H, 4H pip), 3.8 (s, 3H, OMe), 4.3
m, 4H, OCH CH O), 5.2 (d, 1H, CHOH), 6.55 (m, 2H, Bzd H-6
3
, δ): 2 (br m,
yield 3.1 g (64%); mp 251-254 °C.
cis-2-[4-(2,3-Dih yd r oben zo[1,4]d ioxin -5-yl) 3,6-d ih yd r o-
H -p yr id in -1-yl] in d a n -1-ol (48) a n d tr a n s-2-[4-(2,3-
2
2
2
2
Dih yd r ob en zo[1,4]d ioxin -5-yl) 3,6-d ih yd r o-2H -p yr id in -
-yl] in d a n -1-ol (49). Sodium borohydride (1.18 g, 31 mmol)
and H-8), 6.75 (m, 3H, 1H Bzd H-7 + 2H arom), 7.25 (d, 1H,
arom).
cis-2-[4-(2,3-Dih yd r oben zo [1,4] d ioxin -5-yl) 4-h yd r oxy
p ip er id -1-yl] 5-flu or oin d a n -1-ol (50). (Meth od A′). cis-2-
1,4-Dioxo 8-azaspiro [4.5] dec-8-yl) 5-fluoroindan-1-ol 74-cis
7.04 g, 24 mmol), obtained as described for compound 4-cis
1
was added by portion, at room temperature, to a suspension
of compound 11 (2.64 g, 6.2 mmol) in methanol (50 mL). At
the end of the addition, the mixture was stirred 1 h at room
temperature, and acetic acid (2.8 mL, 49.6 mmol) was added.
After evaporation to dryness, the residue was taken up in
NaOH 1 N (50 mL) and extracted by CH Cl (2 × 50 mL). The
(
(
in reacting 2-bromo 5-fluoro indan-1-one IV-D with 1,4-dioxo
-azaspiro [4.5] decane 72 and then reducing the ketone
obtained (73) by NaBH , was deprotected in 1-(5-fluoro 1-hy-
2
2
8
combined organic layers were concentrated under reduce
pressure, and the residue was chromatographed (CH Cl /
4
2
2
droxy indan-2-yl) piperid-4-one 75 (mp 178-181 °C) by HCl
concentrated (24 mL) and AcOH (48 mL). Compound 75 (0.85
g, 3.41 mmol) in THF (34 mL) was introduced at 0 °C in 22
mL of a solution prepared by reacting magnesium (485 mg)
on 5-bromo 2,3-dihydrobenzo[1,4]dioxine (4.3 g, 20 mmol) in
THF (50 mL). At the end of the addition, the reaction mixture
was stirred 1 h at room temperature then 1 h at reflux. After
AcOEt, 90/10). The first compound eluted was recrystallized
from C H OH: yield 0.57 g (26%); mp 109-112 °C and
2
5
identified to be of cis configuration by IR spectrum. IR (CHCl )
3
-1
1
γmax (cm ): 3400 (br, OH). H NMR (300 MHz, CDCl , δ): 2.5-
3
3.25 (cluster of 6H, 4H NCH CH pip + 2H, CH indan), 3.40
2
2
2
(m, 2H, NCH CHdC), 4.30 (m, 5H, 4H, OCH CH O + 1H,
2
2
2
CHN indan), 5.0 (d, 1H, CHOH), 5.90 (m, 1H, NCH CHdC),
2
hydrolysis by a saturated solution of NH
4
Cl (50 mL) and
6.8 (m, 3H, Bzd), 7.25-7.5 (m, 4H, arom).
extraction by ether (3 × 50 mL), evaporation to dryness led to
The second compound eluted was recrystallized from CH
3
-
-1
5
(
0: yield 1.52 g, mp 189-191 °C. IR (CHCl
3
) γmax (cm ): 3366
br, OH). H NMR (300 MHz, DMSO, δ): 2.03-2.25 (m, 4H,
CH pip), 2.7-3.18 (cluster, 7H, 2H, CH NCH + 3H, CH CH
indan), 3.68 (s, 1H, OH), 4.25 (m, 1H, OH), 4.32 (m, 4H, OCH
CH O), 4.9 (d, 1H, CHOH), 6.8-7 (cluster, 5H, 3H, Bzd + 2H,
arom), 7.42 (dd, 1H, arom).
CN: yield 0.67 g (28%); mp 182-184 °C and identified to be
1
) γmax (cm ): 3600 (w,OH). 1H NMR (200
CH ), 2.8-3.2 (cluster of
indan), 3.95
-1
trans. IR (CHCl
MHz, CDCl , δ): 2.6 (m, 2H, NCH
H, 2H NCH CH + 1H, CHN indan + 2H, CH
m, 2H, NCH CHdC), 4.25 (s, 4H, OCH CH O), 5.25 (d, 1H,
CHdC), 6.8 (m, 3H, Bzd), 7.1-7.5
3
2
2
2
2
3
2
2
2
-
5
(
2
2
2
2
2
2
2
CHOH), 5.85 (m, 1H, NCH
2
Gen er a l Meth od B. 4-(2,3-Dih yd r oben zo[1,4]d ioxin -5-
yl) p yr id in e (VIII-A). 4-Pyridineboronic acid (9.85 g, 80
mmol) in EtOH (120 mL) and sodium carbonate (59.4 g, 560
mmol) in water (240 mL) were successively added to a solution
of 5-bromo 2,3-dihydrobenzo[1,4]dioxine (22.4 g, 104 mmol),
tetrakis(triphenylphosphine) palladium (3.7 g, 3.2 mmol), and
toluene (380 mL), and the resulting mixture was heated at
reflux for 28 h and then poured in water (1 L). After
decantation, the aqueous phase was extracted by AcOEt (2 ×
(m, 4H, arom).
tr a n s-2-[4-(2,3-Dih yd r oben zo[1,4]d ioxin -5-yl) p ip er id -
-yl] in d a n -1-ol (53). Compound 49 (0.4 g, 1.15 mmol) in
1
2
methanol (20 mL) was hydrogenated over PtO during 19 h,
at room temperature and under atmospheric pressure. After
filtration and concentration of the solvent, the solid residue
was recrystallized from CH CN: yield 0.21 g (52%); mp 173-
3
-1
1
1
74 °C. IR (CHCl
MHz, CDCl , δ): 1.84 (m, 4H, CH
CH ), 2.85-3.42 (cluster of6H, 2H, NCH
1H, CHN indan + 2H, CH indan), 4.28 (s, 4H, OCH
.25 (d, 1H, CHOH), 6.75 (m, 3H, Bzd), 7.1-7.45 (m, 4H,
arom).
Gen er a l Meth od C. tr a n s-1-[4-(2,3-Dih yd r oben zo[1,4]-
3
) γmax (cm ): 3600 (w,OH). H NMR (300
3
2
CHCH
2
), 2.34 (m, 2H, NCH
+1H, CH CHCH
CH
2
-
2
2
1
50 mL), and the joint organic phases were extracted by HCl
N (2 × 200 mL). Acidic phases were basified by concentrated
2
2
CH
2
2
+
5
2
2
2
O),
2 2
NaOH and extracted by CH Cl . After the usual work up, 13.4
1
g of VIII-A were isolated: yield 78%; mp < 60 °C. H NMR
200 MHz, CDCl , δ): 4.3 (s, 4H, OCH CH ), 6.95 (s, 3H, Bzd),
.5 (d, 2H, CHCCH pyr), 8.65 (d, 2H, CHNCH).
Compounds VIII-C, VIII-D, VIII-E were prepared accord-
ing to the same procedure:
(
7
3
2
2
d ioxin -5-yl) p ip er id -1-yl] 6-n itr oin d a n -2-ol (27). A solution
of 6-nitroindene oxide IX-H (6.2 g, 35 mmol) and compound
II-A in CH
0 h and at reflux for 4 h. After concentration under vacuum,
the residue was taken up in CH Cl , washed with water, dried,
and concentrated. The residue was purified by flash-chroma-
tography (CH Cl
05 °C. IR (CHCl
3
CN (30 mL) was stirred at room temperature for
1
4
-Ben zofu r a n -7-yl p yr id in e (VIII-C): yield 71%; H NMR
2
(
7
2
200 MHz, CDCl , δ): 6.82 (d, 1H, OCHCH), 7.34 (t, 1H, Bzd),
3
2
2
.54 (d, 1H, Bzd), 7.65 (m, 2H, 1H, OCH + 1H, Bzd), 7.78 (dd,
H, CHCCH pyr), 8.68 (dd, 2H, CHNCH).
2
2
/EtOH, 98/2): yield 5.9 g (52%); mp 202-
1
4
-(2,3-Dih yd r oben zofu r a n -7-yl) p yr id in e (VIII-D): yield
3%.
-Th ioch r om a n -8-yl p yr id in e (VIII-E): using the triflate
derivative instead of the bromo: yield 36%. 3.24 (t, 2H,
OCH CH ), 4.61 (t, 2H, OCH ), 6.92 (t, 1H, Bzd), 7.23 and 7.32
d, 2H, Bzd), 7.65 (d, 2H, CHCCH pyr), 8.61 (d, 2H, CHNCH).
-(6-F lu or och r om a n -8-yl) p yr id in e (VIII-B). A mixture
of diethyl 4-pyridinyl borane (0.45 g, 3 mmol), 8-bromo
-fluorochroman (1 g, 4.5 mmol), KOH (0.5 g, 9 mmol),
tetrabutylammonium bromide (0.48 g, 1.5 mmol), and tetrakis-
triphenylphosphine) palladium in THF (15 mL) was heated
-1
2
3
) γmax (cm ): 3600 (w,OH). H NMR (200
7
MHz, CDCl
3
, δ): 1.9 (br m, 1H, OH), 2.85 (m, 4H, CH
2
NCH
2
4
pip), 2.9 (dd, 1H, CHH indan), 3.15 (cluster, 5H, 4H, CH NCH
2
2
pip + 1H, CHN indan), 3.35 (dd, 1H, CHH indan), 4.3 (m, 4H,
OCH CH O), 4.85 (d, 1H, CHOH), 6.6 (m, 2H, Bzd-6 + Bzd-
2
2
2
2
2
(
8), 6.8 (t, 1H, Bzd-7), 7.35 (d, 1H, arom-4), 8.15 (dd, 1H, arom-
5), 8.25 (d, 1H, arom-7).
4
tr a n s-2-[4-(2,3-Dih yd r oben zo[1,4]d ioxin -5-yl) p ip er id -
1-yl] 6-n itr oin d a n -1-ol (70). Diethylazodicarboxylate (9.5
mL, 60.3 mmol) was added dropwise, in 15 min, to a suspen-
sion of compound 27 (4.9 g, 12.3 mmol), p-nitrobenzoic acid
(9.0 g, 54.1 mmol), triphenylphosphine (15.8 g, 60.3 mmol), in
THF (125 mL) and toluene (125 mL) cooled at -20 °C. The
mixture was kept at this temperature for 1 h and overnight
at room temperature. After the sample was evaporated to
dryness, the residue was filtered off on silica (CH Cl /AcOEt,
6
(
at reflux for 3 days. At the end of the reaction, AcOEt (75 mL)
was added. The resulting mixture was washed with brine,
dried over MgSO
chromatography (CH
VIII-B was isolated; yield 59%; mp 76-78 °C. H NMR (200
MHz, CDCl , δ): 2 (m, 2H, OCH CH CH ), 2.82 (m, 2H, OCH
CH CH ), 4.14 (m, 2H, OCH ), 6.7-6.9 (m, 2H, chroman), 7.44
d, 2H, CHCCH pyr), 8.58 (d, 2H, CHNCH).
-(2,3-Dih yd r oben zo[1,4]d ioxin -5-yl) 1-(1-oxoin d a n -2-
4
, and concentrated. After purification by
2
Cl then CH Cl /AcOEt, 90/10), 0.6 g of
2
2
2
1
2
2
3
2
2
2
2
-
98/2) to yield 16 g of a residue which was taken up in ether
and HCl 1 N to lead, after filtration, to 3.7 g (48%) of the
dihydrochloride salt of p-nitrobenzoic acid 2-[4-(2,3-dihydro-
benzo[1,4]dioxin-5-yl) piperazin-1-yl] 6-nitro indan-1-yl ester.
The dihydrochloride salt was suspended in a solution of KOH
(1.63 g, 29 mmol), water (16 mL), and methanol (160 mL) and
2
2
2
(
4
yl) p yr id in iu m , br om id e (11). A solution of 2-bromo 1-in-
danone (2.45 g, 11.6 mmol) and compound VIII-A (2.41 g, 11.3

Selectivity of the Serotonin 5-HT1A Ligand, S 15535
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1 175
heated at reflux for 1 h. The solution was concentrated, and
the residue was taken up in water (100 mL) and extracted
The in vitro intrinsic clearances (vitroClint) were the slope
(after LN linearization) of the curve of the unchanged drug
remaining concentration versus incubation time. In vitro Clint
were then scaled up to in vivo whole body (vivoClint) using
0.045 mg of protein/kg of liver and liver weight of 11 g for the
rat and 1.2 kg for man. In vivo Clint were then transformed
into hepatic clearances (HepCl) using the well-stirred model
three times by CH
2
Cl
2
(70 mL) and after usual workup led to
2
.4 g of a light dark solid which was recrystallized from CH -
3
-1
CN: yield 1.7 g (74%); mp 194-196 °C. IR (CHCl
3
) γmax (cm ):
1
3
600 (w, OH). H NMR (300 MHz, CDCl
3
, δ): 2.2 (br m, 1H,
pip + 1H, CHN indan
1H, CHH indan), 3.9 (cluster, 5H, 4H, CH NCH pip + 1H,
O), 5.25 (m, 1H, CHOH),
2 2
OH), 3.20 (cluster, 6H, 4H, CH NCH
+
2
2
(
HepCl ) vivoClint*HBF/(vivoClint + HBF) where HBF
CHH indan), 4.3 (m, 4H, OCH
2
CH
2
(hepatic blood flow) were taken as 22 mL/min for the rat (250
6
.6 (m, 2H, Bzd-6 + Bzd-8), 6.8 (t, 1H, Bzd-7) 7.33 (d, 1H,
arom-4), 8.13 (dd, 1H, arom-5), 8.23 (s, 1H, arom-7).
Biology. 1. Bin d in g Stu d ies a t r
Recep tor s. (a ) Deter m in a tion of Affin ity for Ra t Cor tex
-Recep tor s. Binding affinity at rat R -adrenoceptors was
determined by competition with [ H]-prazosin (Amersham, Les
g) and 1500 mL/min for man. The MF% were then deducted
from the extraction ratio with the following equation (MF% )
1 - HepCl/HBF). This method was shown internally to give a
good in vitro/in vivo correlation on about 40 chemically
nonrelated compounds. When human hepatocytes were re-
quired to confirm the metabolic stability measurements the
same method was employed using in that case an hepatocytes
1
a n d a t Hu m a n 5-HT1A
r
1
1
3
2
6
Ulis, France) as described by Millan. Briefly, membranes
were prepared from rat frontal cortex and incubated in
6
suspension in a classical culture media (0.2 × 10 cells/mL),
3
triplicate with 0.2 nM [ H]-prazosin and competing ligand in
and kinetic time points were 10 min, 30 min, 1 h, 2 h, and 4
h.
a final volume of 0.5 mL, for 1 h at 22 °C. The incubation buffer
2
contained: 50 mM TRIS-HCl, pH 7.4, 4 mM CaCl , 0.1% w/v
5
. Deter m in a tion of In testin a l Absor p tion .29 Assuming
no possible limitation due to solubility, absorbed fractions
Abs%) were estimated using permeability measurements
ascorbic acid. Nonspecific binding was defined with 10 µM
phentolamine.
(
(
b) Deter m in a tion of Affin ity for Recom bin a n t Hu m a n
through Caco-2 cell monolayers. Briefly, unchanged drugs were
incubated with Caco-2 monolayers (apical compartment) at 20
µM and concentration versus incubation time appearance in
the basal compartment were quantified by LC-MS-MS after
0, 15, 30, 60, and 120 min of incubation. Apparent permeabili-
ties (Papp) were calculated from the slope of the appearance
line (permeability rate) with the following equation (Papp )
permeability rate/initial apical concentration/surface area of
the monolayer). The absorbed fraction prediction (Abs%) were
then obtained from the plot against an internal calibration
curve performed from Papp of six reference compounds for
which oral absorption in man was described (mannitol with
5
-HT1A Recep tor s. Binding affinity at human 5-hydrox-
ytryptamine 5-HT1A receptors was determined by competition
with [ H]-8-OH-DPAT (Amersham, Les Ulis, France) as de-
scribed by Newman-Tancredi. Briefly, membranes were
incubated with [ H]-8-OH-DPAT at 22 °C for 2.5 h. Nonspecific
3
2
7
3
binding was defined with 5-HT (10 µM).
(
c) Da ta An a lysis. At the end of the incubation period,
membranes were filtered through Whatman GF/B filters
pretreated with 0.1% de polyethylenimine. Radioactivity re-
tained on the filters was determined by scintillation counting.
Binding isotherms were analyzed by nonlinear regression
using the program Prism (GraphPad Software Inc., San Diego,
USA) to determine IC50 values. These were converted to
inhibition constants (K ) by the use of the Cheng-Prusoff
equation K ) IC50/{(L/K ) -1} where L is the concentration
of [ H]-ligand and K is its dissociation constant. The K values
were 0.1 nM for [ H]-prazosin, 0.6 nM for [ H]-8-OH-DPAT.
. Deter m in a tion of Agon ist Effica cy a t Recom bin a n t
Hu m a n 5-HT1A Recep tor s. Efficacy was determined by
1
9
6%, atenolol with 50%, cimetidine with 70%, propranolol with
0%, and antipyrine and testosterone with 100%). This method
i
was shown internally and by others to give a good in vitro/in
vivo correlation for human intestinal absorption.
i
D
3
D
D
3
3
2
Ack n ow led gm en t. We thank Rozenn De Sousa for
excellent technical assistance during the preparation of
this tapuscript.
3
5
measuring agonist stimulation of [ S]GTPγS binding, as
described previously by Newman-Tancredi. Briefly, CHO-h5-
2
7
HT1A membranes (50 µg of protein) were incubated (20 min,
Refer en ces
2
7
[
2 °C) in triplicate in a buffer containing 20 mM HEPES (pH
.4), 3 µM GDP, 3 mM MgSO
4
, NaCl 100 mM, and 0.1 nM
S]GTPγS (1300 Ci/mmol, NEN). Nonspecific binding was
(
(
(
1) Peglion, J .-L.; Canton, H.; Bervoets, K.; Audinot, V.; Brocco, M.;
Gobert, A.; Le Marouille-Girardon, S.; Millan, M. J . Character-
ization of potent and selective antagonists at postsynaptic 5-HT1A
receptors in a series of N4-substitutes arylpiperazines. J . Med.
Chem. 1995, 38, 4044-4055.
2) Millan, M. J .; Canton, J .-M.; Lejeune, F.; Gobert, A.; Widdowson,
P.; Bervoets, K.; Brocco, M.; Peglion, J .-L. S 15535: a highly
selective benzodioxopiperazine 5-HT1A receptor ligand which acts
as an agonist and an antagonist at presynaptic and postsynaptic
sites, respectively. Eur. J . Pharmacol. 1993, 230, 92-102.
3) (a) Newman-Tancredi, A.; Chaput, C.; Verri e` le, L.; Millan, M.
3
5
defined with 10 µM GTPγS. Agonist efficacy is expressed
relative to that of 5-HT () 100%), which was tested at a
maximally effective concentration (10 µM) in each experiment.
3
. In h ibition of Hyp oth er m ia In d u ced by 8-OH-DP AT.
Male Wistar rats of 200-220 g, housed singly, were removed
from home cages, and core (rectal) temperature was deter-
mined by use of a digital thermistoprobe. Then the rats were
treated with either vehicle or the putative antagonist and
returned to their home cages. Thirty minutes later, they were
reinjected with either vehicle or 8-OH-DPAT (0.16 mg/kg) and
returned to their cages for a further 30 min, and then the core
temperature was redetermined. The difference between basal
and posttreatment values was calculated. The mean decrease
in core temperature elicited by 8-OH-DPAT in the presence
of vehicle was, typically, -2.2 ( 0.2 °C (n ) 10). The ID50 (95%
confidence limits) was calculated according to the method of
Finney, 1964. All drugs were dissolved in distilled water and
given s.c. in a volume of 1.0 mL/kg.
3
5
J . S 15535 and WAY 100, 635 antagonise 5-HT-stimulated [ S]-
GTPγS binding at cloned human 5-HT1A receptors. Eur. J .
Pharmacol. 1996, 307, 107-111. (b) Newman-Tancredi, A.;
Verri e` le, L.; Chaput, C.; Millan, M. J . Labelling of recombinant
human and native rat 5-HT1A receptors by a novel, selective
3
radioligand [ H]-S 15535: definition of its binding profile using
agonists, antagonists and inverse agonists. Naunyn Schmiede-
berg’s Arch. Pharmacol. 1998, 357, 205-217.
4) Newman-Tancredi, A.; Rivet, J .-M.; Brocco, M.; Lacroix, P.;
Audinot, V.; Cistarelli, L.; Gobert, A. S15535, a novel benzo-
dioxopiperazine ligand of serotonin (5-HT)1A receptors: I. In-
(
2 3
teraction with cloned human (h)5-HT1A, dopamine hD /hD and
4
. Deter m in a tion of Meta bolic Bioa va ila bility.28 Meta-
hR2A-adrenergic receptors in relation to modulation of cortical
monoamine release and activity in models of potential antide-
pressant activity. J . Pharmacol. Exp. Ther. 1997, 232, 132-147.
5) Millan, M. J .; Hjorth, S.; Samanin, R.; Schreiber, R.; J affard,
R.; De Ladonchamps, B.; Veiga, S.; Goument, B.; Peglion, J .-L.;
Spedding, M.; Brocco, M. S15535, a novel benzodioxopiperazine
ligand of serotonin (5-HT)1A receptors: II. Modulation of hip-
pocampal serotonin release in relation to potential anxiolytic
properties. J . Pharmacol. Exp. Ther. 1997, 282, 148-161.
bolic bioavailability predictions (MF%) were based on in vitro
metabolic stability measurements with hepatic microsomes
assuming total absorption. Briefly, unchanged drugs were
(
-
7
quantified by LC-MS-MS following incubation (10 M) with
rat and human hepatic microsomes (0.33 mg of protein/mL)
after 0, 5, 15, 30, and 60 min of incubation. Hepatic mi-
crosomes were obtained from a pool of five different subjects.

1
76 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 1
Peglion et al.
(
6) Dekeyne, A.; Brocco, M.; Adhumeau, A.; Gobert, A.; Millan, M.
J . The selective serotonin (5-HT)1A receptor ligand, S 15535,
displays anxiolytic-like effects in the social interaction and Vogel
models and suppresses dialysate levels of 5-HT in the dorsal
hippocampus of freely moving rats. A comparison with other
anxiolytic agents. Psychopharmacology 2000, 152, 55-66.
7) Carli, M.; Balducci, C.; Millan, M. J .; Bonalumi, P.; Samanin,
R. S15535, a benzodioxopiperazine acting as presynaptic agonist
and postsynaptic 5-HT1A receptor antagonist, prevents the
impairment of spatial learning caused by intrahippocampal
scopolamine. Br. J . Pharmacol. 1999, 128, 1207-1214.
8) Millan, M. J .; Canton, H.; Gobert, A.; Lejeune, F.; Rivet J .-M.;
Bervoets, K.; Brocco, M.; Widdowson, P.; Mennini, T.; Audinot,
V.; Honor e´ , P.; Renouard, A.; Le Marouille-Girardon, S.; Verri e` le,
S.; Gressier, H.; Peglion, J .-L. Novel benzodioxopiperazines
acting as antagonists at postsynaptic 5-HT1A receptors and as
agonists at 5-HT1A autoreceptors: a comparative pharmacologi-
cal characterization with proposed 5-HT1A antagonists. J . Phar-
macol. Exp. Ther. 1994, 268, 337-352.
(18) Hartog, J .; Mos, J . EP 372657.
(19) Millan, M. J .; Brocco, M. Serotonin and anxiety: mixed 5-HT1A
agonists-5-HT1C/2 antagonists as potential anxiolytic agents. In
Hamon, M.; Ollat, H.; Thi e´ bot, M.-H. (eds) Anxiety, Neurobiol-
ogy, clinical and therapeutic perspectives. Libbey Eurotext,
London, 1993, pp 153-165.
(20) Meller, E.; Goldstein, M.; Bohmaker, K. Receptor reserve for
5-hydroxytryptamine1A-mediated inhibition of serotonin synthe-
sis: possible relationship to anxiolytic properties of 5-hydroxy-
tryptamine1A agonists. Mol. Pharmacol. 1990, 37, 231-237.
(
(
(
(
(
21) Newman-Tancredi, A.; Rivet, J .-M.; Chaput, C.; Touzard, M.;
Verri e` le, L.; Millan, M. J . The 5-HT1A receptor ligand, S 15535,
35
3
antagonises G-protein activation: a [ S]GTPγS and [ H]S 15535
autoradiography study. Eur. J . Pharmacol. 1999, 384, 109-119.
22) Kuipers, W.; van Wijngaarden, I.; Kruse, C. G.; ter Horst-van
4
Amstel, M.; Tulp, M. Th. M.; Ijzerman, A. P. N -Unsubstituted
1
N -Arylpiperazines as High-Affinity 5-HT1A Receptor Ligands.
J . Med. Chem. 1995, 38, 1942-1954.
23) Hibert, M. F.; Gittos, M. W.; Middlemiss, D. N.; Mir, A. K.;
Fozard, J . R. Graphics computer-aided receptor mapping as a
predictive tool for drug design: development of potent, selective
and stereospecific ligands for the 5-HT1A receptor. J . Med Chem.
(
9) Fletcher, A.; Forster, E. A.; Bill, D. J .; Brown, G.; Cliffe, I. A.;
Hartley, J . E.; J ones, D. E.; McLenachan, A.; Stanhope, K. J .;
Critchley, D. J . P.; Childs, K. J .; Middlefell, V. C.; Lanfumey,
L.; Caorradetti, R.; Laporte, A. M.; Gozlan, H.; Hamon, H.;
Dourish, C. T. Electrophysiological, biochemical, neurohormonal
and behavioural studies with WAY-100635, a potent, selective
and silent 5-HT1A receptor antagonist. Behav. Brain Res. 1996,
1
988, 31, 1087-1093.
(
24) Bolognesi, M.; Budriesi, R.; Cavalli, A.; Chiarini, A.; Gotti, R.,
Leonardi, A.; Minarini, A.; Poggesi, E.; Recanatini, M.; Rosini,
M.; Tumiatti, V.; Melchiorre, C. WB 4101-related compounds.
7
3, 337-353.
2
. Role of the ethylene chain separating amine and phenoxy
-adrenoreceptor subtypes and 5-HT1A
(
10) (a) Van Wijngaarden, I.; Tulp, M. Th. M.; Soudijn, W. The
concept of selectivity in 5-HT receptor research. Eur. J . Phar-
macol. 1990, 188, 301-312. (b) Glennon, R. A.; Dukat, M.
Serotonin receptors and their ligands; a lack of selective agents.
Pharmacol. Biochem. Behav. 1991, 40, 1009-1017.
11) Picciola, G.; Riva, M.; Ravenna, F.; Gentili, P. USP 4826975.
12) (a) Sol a` , L.; Vidal-Ferran, A.; Moyano, A.; Peric a` s, M. A.; Riera,
A. New Indane derived amino alcohols as chiral ligands for the
catalytic enantioselective addition of diethylzinc to aldehydes.
Tetrahedron: Asymmetry 1997, 8 (10), 1559-1568. (b) Sol a` , L.;
Vidal-Ferran, A.; Moyano, A.; Peric a` s, M. A.; Riera, A. Corri-
gendum. Tetrahedron: Asymmetry 1997, 8 (24), 3985.
units on the affinity for R
receptors. J . Med. Chem. 1999, 42, 4214-4224.
1
(
25) Lopez-Rodriguez, M. L.; Morcillo, M. J .; Rovat, T. K.; Fernandez,
E.; Vicente, B.; Sanz, A. M.; Hernandez, M.; Orensanz, L.
Synthesis and structure-activity relationships of a new model
of arylpiperazines. 4. 1-[ω-(arylpiperazin-1-yl)alkyl]-3-(diphe-
nylmethylene)-2,5-pyrrolidinediones and -3-(9H-fluoren-9-yliden)-
(
(
2
,5-pyrrolidinediones: study of the steric requirements of the
terminal amide fragment on 5-HT1A affinity/selectivity. J . Med.
Chem. 1999, 42, 36-49.
26) Millan, M. J .; Peglion, J .-L.; Vian, J .; Rivet J .-M.; Brocco, M.;
Gobert, A.; Newman-Tancredi, A.; Dacquet, C.; Bervoets, K.; Le
Marouille-Girardon, S.; J acques, V.; Chaput, C.; Audinot, V.
(
(
(
(
(
13) Freedman, J .; Vaal, M. J .; Huber, E. W. The Mitsunobu reaction
of some indan amino alcohols. J . Org. Chem. 1991, 56, 670-
6
72.
Functional correlates of dopamine D
rat in vivo and their modulation by the selective antagonist, (+)-
S14297: 1. Activation of postsynaptic D receptors mediates
hypothermia while blockade of D receptors elicits prolactin
secretion. J . Pharmacol. Exp. Ther. 1995, 275, 885-898.
3
receptor activation in the
14) Coleman, R. S.; Grant, E. B. A low-temperature Mitsunubu
reaction for the inversion of sterically hindered secondary
alcohols. Tetrahedron Lett. 1994, 35 (45), 8341-8344.
15) Rimek, H. J .; Yupraphat, T.; Zymalkowski, F. Konfigurations-
bestimmung an 2-amino-indanolen. Liebigs Ann. Chem. 1969,
3
2
(27) Newman-Tancredi, A.; Gavaudan, S.; Conte, C.; Chaput, C.;
Touzard, M.; Verri e` le, L.; Audinot, V.; Pasteau, V.; Millan, M.
J . Agonist and antagonist actions of antipsychotic agents at
serotonin 5-HT1A receptors: a [ S]GTPγS binding study. Eur.
J . Pharmacol. 1998, 355, 245-256.
(28) Bertrand, M.; J ackson, P.; Walther, B. Rapid assessment of drug
metabolism in the drug discovery process. Eur. J . Pharm. Sci.
2000, 11 Suppl. 2, S61-S72.
7
26, 25-29.
16) Peglion, J .-L.; Newman-Tancredi, A.; Verri e` le, L.; Millan, M. J .
3
35
[
H]-S 15535, a novel selective radioligand at serotonin 5-HT1A
receptors: characterisation of binding to cloned human and rat
hippocampal 5-HT1A receptors. Am. Soc. Neurosci. 1995, 21,
1
855.
(
17) (a) Newman-Tancredi, A. Clozapine is a partial agonist at cloned,
human serotonin 5-HT1A receptors. Neuropharmacology 1996,
(29) Yee, S. In vitro permeability across Caco-2 cells (colonic) can
predict in vivo (small intestinal) absorption in Man-Fact or Myth.
Pharmacol. Res. 1997, 14, 763-766.
3
5 (1), 119-121. (b) Newman-Tancredi, A.; Verri e` le, L.; Millan,
M. J . Differential modulation by GTPγS of agonist and inverse
agonist binding to h5-HT1A receptors revealed by [ H]-WAY 100,
3
6
35. Br. J . Pharmacol. 2001, 132, 518-524.
J M010975+