Synthesis and pharmacological evaluation of Tic-hydantoin derivatives as selective σ1 ligands. Part 2

Article abstract of DOI:10.1016/j.bmcl.2005.07.039

Herein is described a new class of selective σ₁ ligands consisting of tetrahydroisoquinoline-hydantoin (Tic-hydantoin) derivatives. Compound 1a has high affinity (IC₅₀ = 16 nM) for σ₁ receptor and is selective in a large panel of therapeutic targets. This study presents structural changes on the side chain of the Tic-hydantoin core. Analogs of higher affinity could be identified (IC₅₀ ≈ 2-3 nM).

Full text of DOI:10.1016/j.bmcl.2005.07.039

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4828–4832
Synthesis and pharmacological evaluation of
Tic-hydantoin derivatives as selective r1 ligands. Part 2
Amaury Cazenave Gassiot, Julie Charton,à Sophie Girault-Mizzi, Pauline Gilleron,§
Marie-Ange Debreu-Fontaine, Christian Sergheraert and Patricia Melnyk*
´
UMR CNRS 8525, Universite de Lille II, Institut de Biologie et Institut Pasteur de Lille 1 rue du Professeur Calmette,
B.P. 447, 59021 Lille cedex, France
Received 14 March 2005; revised 7 July 2005; accepted 14 July 2005
Available online 1 September 2005
Abstract—Herein is described a new class of selective r1 ligands consisting of tetrahydroisoquinoline-hydantoin (Tic-hydantoin)
derivatives. Compound 1a has high affinity (IC50 = 16 nM) for r1 receptor and is selective in a large panel of therapeutic targets.
This study presents structural changes on the side chain of the Tic-hydantoin core. Analogs of higher affinity could be identified
(IC50 ꢀ 2–3 nM).
Ó 2005 Elsevier Ltd. All rights reserved.
O
O
To date, multiple r binding sites have been identified
and r receptors are classified into two distinct subtypes
denoted as r1 and r2.1 Recent evidence has indicated
that r1 receptors may be involved in regulating a variety
of neurotransmitters in the central nervous system,
including cholinergic,2,3 dopaminergic,4 and glutamater-
gic systems.5,6 Thus, there is a sustained interest for
using selective r1 ligands, which stems from the possibil-
ity of developing new drug candidates particularly for
the treatment of depression,7 psychiatric disorders,
memory deficits, or drug addiction.8,9 r2 ligands may
be developed for attenuating motor side effects associat-
ed with typical antipsychotic agents or for diagnossis
and treatment of cancer.
N
N
N
N
N
N
1a
1b
O
O
Figure 1. Compounds 1a and 1b.
r1 receptor could be explained by the association of
an aromatic moiety and a nitrogen atom. That type of
structure has previously proved to be a pharmacophoric
element in the binding of a series of phenylalkylpiperi-
dines and phenylalkylpiperazines to r receptors.10
The tetrahydroisoquinoline-hydantoin (Tic-hydantoin)
core was of particular interest on account of the specific-
ity of the pharmacological profile obtained for com-
pound 1a.11
In a profile screening, compound 1a (Fig. 1) was identi-
fied as an interesting ligand of guinea pig r1 receptor
with an IC50 of 16 nM. Binding of this compound at
Two parallel studies were investigated to improve the
affinity of the lead compound 1a toward r1 receptor.
The first study, described in the previous article, dealt
with structural changes on the Tic-hydantoin core, while
preserving the alkylpiperidine moiety of compound 1a.
In this letter, parallel work consisting in the study of
the side chain of compound 1a while preserving the
Tic-hydantoin core is reported. Two series of com-
pounds were designed (Fig. 2).
Keywords: Hydantoin; Tic; r-Receptors; Selectivity.
*
Corresponding author. Tel.: +33 3 20 87 12 19; fax: +33 3 20 87 12
33; e-mail: patricia.melnyk@ibl.fr
Present address: School of Chemistry, University of Southampton—
Highfield, Southampton, SO17 1BJ, UK.
Present address: Laboratoire de Chimie Ge´ne´rale et Organique,
à
´
UMR8525, Faculte des Sciences Pharmaceutiques et Biologiques, 3
rue du Professeur Laguesse, B.P. 83, 59006 Lille, France.
´
Present address: ICPAL—EA 2692—Universite de Lille II, 3 rue du
Professeur Laguesse, B.P. 83, 59006 Lille, France.
§
First, the piperidino group was replaced by differently
substituted piperazino moieties, while conserving the
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.07.039

A. C. Gassiot et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4828–4832
4829
O
O
Table 1. r-binding assays on compounds 1a–b and 5a–j. Mean
IC50 SD values for two to three independent experiments are shown
NRR'
n N
R
n
N
m
N
N
N
O
O
O
NRR'
A series
B series
A series
n
N
N
Figure 2. Target compounds.
O
Compound
Haloperidol
n
NRR0
—
IC50 r1 (nM)
two and three methylene linkers, which were found to be
relevant in our parallel study (A series). Indeed, affinity
of compounds 1a and 1b (Fig. 1) for r1 receptor
(IC50 = 16 and 21.5 nM, respectively) and selectivity
were equivalent.
—
2.1 0.3
1a
5a
5b
2
2
2
N
16
3
N
N
N
N
>100
>1000
Second, the structure was varied according to a model
(Fig. 3) recently proposed to account for the binding
of ligands at r1 receptor (B series).10,12,13
N
N
N
OH
5c
5d
2
2
>100
>100
In this model, a nitrogen atom appears as a required
pharmacophoric element between two hydrophobic aro-
matic regions: the first corresponding, restrictively, to a
phenyl ring, whilst the second tolerates various structur-
al features because it likely spills over a region of bulk
tolerance (Fig. 3). The Tic-hydantoin core could play
the role of this second hydrophobic group.
O
N
H
N
N
N
N
1b
5e
3
3
21.5 4.0
>100
N
N
To obtain compounds 5a–j of A series (Table 1), the
starting material was commercial (S)-(ꢁ)-1,2,3,4-tetra-
hydroisoquinoline-3-carboxylic acid
2
(L-Tic-OH)
N
N
N
5f
5g
5h
5i
3
3
3
3
3
>100
>100
whose secondary amine function was protected using
Boc2O before its transformation into methyl ester 4
(Scheme 1). The release of the amine function was real-
ized by treatment with a TFA/CH2Cl2 1:1 mixture.
OH
N
The urea was formed and cyclization was achieved as
described in the previous article in a ‘‘one pot’’ reaction
by the action of appropriate 3-chloropropyl- or 2-chlo-
roethylisocyanate in a DIEA/CH2Cl2 mixture.11,14 Com-
pounds 1a–b and 5a–j were obtained by nucleophilic
substitution of the chlorine atom.
32
6
N
N
N
O
O
23.8 3.7
N
5j
82.3 10.9
N
HN
For compounds 10a–r of the B series (Table 2), the meth-
od described in Scheme 1 was first tried using appropriate
chloroalkylisocyanates with the additional step of substi-
tution of the chlorine atom by N-alkylmethylamine. This
procedure was feasible when using primary amines
(R = H), for instance N-benzylamine and 4-phenyl-1-
butylamine, led to compounds 10a and 10b, respectively,
COOH
NH
COOH
NBoc
COOCH3
a
b
NBoc
4
2
3
c
O
O
d
NRR'
N
n
Cl
n
N
Proton Donor Site
N
N
Region of
O
O
Bulk Tolerance
1a-b and 5a-j
6 - 10 Å
2,5 - 3,9 Å
(optimum binding : 8,3 Å)
Scheme 1. Reagents and conditions: (a) Boc2O 1.1 equiv, NaOH 1 M
1.1 equiv, dioxane, rt, 12 h, 98%; (b) (i) Cs2CO3 0.5 equiv, H2O,
MeOH, rt, 10 min, (ii) CH3I 1.1 equiv, DMF, rt, 12 h, 90%; (c) (i)
TFA/CH2Cl2 1:1, rt, 30 min, (ii) DIEA 15 equiv, CH2Cl2, rt, 15 min,
(iii) 3-chloropropyl- or 2-chloroethylisocyanate 2.5 equiv, CH2Cl2, rt,
12 h; (d) HNRR0 8 equiv, K2CO3 3 equiv, CH3CN, reflux, several
hours, 30–70%.
N
Region tolerates
small groups only
Secondary Binding Site .
Tolerates bulk
Primary
Hydrophobic site
in medium overall yields (50 and 35%, respectively). But
the first attempts using N-methylbenzylamine gave poor
overall yields, respectively, 16 and 9% for n = 2 and 3
Figure 3. Pictorial representation of proposed features for r1
binding.10

4830
A. C. Gassiot et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4828–4832
Table 2. r-binding assays on compounds 1a and 10a–o
O
N
n N
R
m
B series
N
O
IC50 r2 (nM)a
Ratio r2/r1
a
Compound
R
n
m
IC50 r1 (nM)
Haloperidol
1a
2.1 0.3
70 20
>1000
>1000
nd
33
—
H
—
3
3
2
2
2
2
2
3
3
4
4
5
5
6
6
—
1
4
1
2
3
4
5
1
2
1
2
1
2
1
2
16
9.6 1.4
3
>60
>100
nd
10a
10b
10c
H
10.5
2
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
O
4.5 0.1
12.6 1.6
13.6 1.9
3.2 0.2
2.9 0.1
3.9 0.1
2.1 0.1
4.2 0.1
9.9 0.1
5.0 0.9
11.7 0.3
9.4 1.8
29.5 3.6
972 41
nd
nd
216
nd
10d
10e
nd
10f
358 60
>1000
502 98
21 11
>100
nd
112
>340
129
10
10g
10h
10i
10j
>20
nd
10k
10l
81 16
nd
nd
16
nd
10m
10n
10o
nd
nd
nd
10p
3
1
>100
nd
nd
N
N
10q
10r
—
CH3
—
3
1
0
2.4 0.1
>100
24 11
nd
10
nd
Mean IC50 SD values for two to three independent experiments are shown.
a nd, not determined.
O
O
(compounds 10c and 10h). Under the same conditions,
c
a, b
S
n N
m
n
O
BocHN
n OH
BocHN
H2N
compound 10d was obtained from N-methylphenethyl-
amine in poor yield (15%). N-methylphenylalkylamines
showed indeed low reactivity, resulting in an increase in
time of reaction and then decomposition of the com-
pounds under the heating conditions. Thus, another strat-
egy was to introduce diversity directly on the amine,
before coupling with the Tic-hydantoin core. Protected
amines 8c–o were first synthesized in a three-step process
starting from an appropriate alkylaminoalcohol (Scheme
2). Then, the deprotected amines were coupled with Boc-
L-Tic-OH 3 using HOBt/EDCI activation. After depro-
tection of the secondary amino group, 1,10-carbonyldiim-
idazole (CDI) was used to yield hydantoins 10c–o. This
procedure allows easy access to diversify hydantoin based
products further.15 Experimental procedure was opti-
mized to allow rapid synthesis of a large set of hydantoins.
8c-o
6a-e
n = 2 to 6
7a-e
n = 2 to 6
d, e
O
O
f, g
n
N
m
N
N
n
m
N
H
N
NBoc
10c-o
9c-o
O
Scheme 2. Reagents and conditions: (a) Boc2O 1.1 equiv, CH2Cl2, rt,
12 h; (b) MsCl 1.3 equiv, TEA 2 equiv, CH2Cl2, 0 °C, 2 h; (c)
phenylalkylmethylamine 3 equiv, K2CO3 4 equiv, CH3CN, 40 °C, 24–
72 h; (d) (i) TFA/CH2Cl2 1:1, rt, 30 min, (ii) DIEA 15 equiv, CH2Cl2,
rt, 15 min; (e) compound 3 1 equiv, HOBt 1.1 equiv, EDCI 1.1 equiv,
CH2Cl2, rt, 12 h; (f) (i) TFA/CH2Cl2 1:1, rt, 30 min, (ii) DIEA 15
equiv, THF, rt, 15 min; (g) CDI 2 equiv, THF, reflux, 12 h.
methylaniline) to verify the importance of the binding
of this nitrogen with a proton donor site (Fig. 4).
This last method could not be used to obtain secondary
amine derivatives, such as compounds 10a and 10b. In-
deed, when trying to obtain compound 10a using the
CDI method, the secondary nitrogen atom of the amino
derivative 9a reacts with CDI during the cyclization step
to give the compound 10p (Scheme 3). The use of one
equivalent of CDI did not avoid this secondary reaction.
All the compounds were assayed in binding assays on
guinea pig cerebral cortex r1 receptor using haloperidol
as reference compound. For compounds showing high
r1 affinity, binding assays were also performed on rat
r2 receptor.16–19 The specific ligand binding to the
receptors is defined as the difference between the total
binding and the non-specific binding determined in the
presence of an excess of unlabeled ligand. The biochem-
ical results are presented as IC50 value, concentration
causing a half-maximal inhibition of control-specific
Finally, two more compounds were synthesized using
this last method: 10q with a more rigid scaffold (from
4-amino-1-benzylpiperidine) and 10r with a less basic
nitrogen atom (starting from N-(3-aminopropyl)-N-

A. C. Gassiot et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4828–4832
4831
O
O
a, b
N
n N
H
m
N
n N
m
H
N
NBoc
O
N
O
N
9a
10p
Scheme 3. Reagents and conditions: (a) (i) TFA/CH2Cl2 1:1, rt, 30 min, (ii) DIEA 15 equiv, THF, rt, 15 min; (b) CDI 2 equiv, THF, reflux, 12 h,
30%.
O
O
between the two aromatic groups (compound 10o:
n,m = 6,2) was detrimental to the r1 affinity.
N
N
N
N
N
N
O
O
The compounds of B series were designed according to
GlennonÕs model, which describes optimum length for
the chain linking the nitrogen atom and the phenyl ring.
Regardless of the lower r1 affinities of the compounds
of B series compared to the affinities of the phenylpiper-
azines or phenylpiperidines described in GlennonÕs
work,10 the spacer length on both sides of the central
nitrogen atom does not seem to significantly influence
the affinity. For instance compounds 10c (n,m = 2,1),
10f (n,m = 2,4), 10g (n,m = 2,5), and 10l (n,m = 5,1) dis-
played similar affinity. Assuming that the Tic core binds
at the left hydrophobic site would then lead to the con-
clusion that there is actually no direct correlation be-
tween the affinity and respective side chain length.
Nevertheless, present results could also be interpreted
in terms of different ways of binding. The decrease in
affinity from 10c (n,m = 2,1; IC50 = 4.5 nM) to 10d and
10e (n,m = 2,2 and 2,3; IC50 = 12.6 and 13.6, respective-
ly) could be explained by the hypothesis that these com-
pounds bind in one way which is disfavored by an
increase in chain length. The fact that compounds 10f
and 10g (n,m = 2,4 and 2,5; IC50 = 3.2 and 2.9 nM,
respectively) showed similar affinity as 10c, despite their
even longer side chain, could then be explained by the
existence of another (opposite) way of binding.
10q
10r
Figure 4. Compounds 10q and 10r.
binding20 (Tables 1 and 2). Some compounds were
tested for cytotoxicity upon a human diploid embryonic
lung cell line (MRC-5 cells) using the colorimetric MTT
assay (Table 3).21
With regard to A series (Table 1), introduction of ben-
zylpiperazine templates results in retention of r1 affinity
(compounds 5h and 5i). In contrast, the piperazines 5b–
g, which are not substituted by aromatic groups, lost r1
affinity. The 4-amino-1-benzylpiperidinyl compound 5j
showed moderate r1 affinity. In this series, a second
hydrophobic group seems to be required for r1 affinity,
which is consistent with the previously proposed model
(Fig. 3). As stated in Introduction, it was first considered
that the Tic core fitted in the left hydrophobic site of the
proposed model due to the fact that this site was likely
to tolerate steric bulk. However, it was recently suggest-
ed that the right hydrophobic, could tolerate steric bulk
as well.22 Taking that into account, it was suggested that
two modes of binding may be possible: binding of the
Tic core at the left hydrophobic site or at the right
one. Whether the Tic core binds at the left or right re-
gion might be depending on the presence or not of a
benzylic moiety.
However, conclusions can hardly be drawn without
further investigations.
The significant loss of affinity for compounds 10p and
10r confirmed the contribution of the proton-accepting
nitrogen atom for r1 receptor binding, as previously de-
scribed by Glennon. The secondary amino compound
10a showed a slightly lower r1 affinity than its tertiary
N-methylated analog 10h. More constrained compound
10q provided an increase in affinity but detriment to the
selectivity r2/r1.
Most active compounds of the A series were submitted
to r2 binding test, but very low inhibition percentages
were obtained at 10 lM, showing a good r2/r1 ratio.
With regard to B series, the r1 affinity was enhanced for
a great majority of compounds (except compounds 10d,
10e, 10o, 10p, and 10r) compared with the lead com-
pound 1a. This suggests that there is no minimal chain
length requirement. Conversely, increasing the distance
A preliminary study of the cytotoxicity upon MRC-5 cells
of some compounds from the B series was investigated to
evaluate the therapeutic potential of this series. As shown
in Table 3, tested compounds present a low cytotoxicity
with CC50 values superior to 100 lM for tertiary amino
compounds. All these compounds provided a selectivity
index (ratio CC50/IC50 on r1 receptor) superior to
10,000. Replacement of the tertiary nitrogen by a second-
ary nitrogen (compound 10b), in addition to be detrimen-
tal to the affinity, seems to increase the cytotoxicity.
Table 3. Cytotoxicity on MRC-5 cells and selectivity index
Compound
CC50 (lM)
IS
2952
10b
10c
10e
10h
10k
10l
31
107
136
1
4
7
23778
10000
50256
10707
20200
196 11
106
101 20
3
By holding the Tic-hydantoin template constant, it has
been possible to explore the nature of the side chain.
Mean CC50 SD values for two to three independent experiments are
shown.

4832
A. C. Gassiot et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4828–4832
9. Guitart, X.; Codony, X.; Monroy, X. Psychopharmacol-
ogy 2004, 174, 301.
10. Ablordeppey, S. Y.; Fischer, J. B.; El-Ashmawy, M.;
Glennon, R. A. Eur. J. Med. Chem. 1998, 33, 625.
11. See previous article.
12. Ablordeppey, S. Y.; Fischer, J. B.; Glennon, R. A. Bioorg.
Med. Chem. 2000, 8, 2105.
13. Ablordeppey, S. Y.; Fischer, J. B.; Law, H.; Glennon, R.
A. Bioorg. Med. Chem. 2002, 10, 2759.
It led us to optimize the affinity of the lead compound 1a
by highlighting new selective r1 ligands with high affin-
ity, among them being compounds 10f, 10g, and 10h.
This study was performed in parallel to the evaluation
and optimization of the Tic-hydantoin core described
in the previous article. Further investigations combining
the results of these two studies should allow us to reach
r1 ligands with higher affinity and better selectivity.
Nevertheless, with such a high selectivity index, the de-
scribed compounds could be considered as potential
leads or candidates for therapy.
14. Charton, J.; Delarue, S.; Vendeville, S.; Debreu-Fontaine,
M.-A.; Girault-Mizzi, S.; Sergheraert, C. Tetrahedron
Lett. 2001, 42, 7559.
15. Charton, J.; Cazenave Gassiot, A.; Melnyk, P.; Girault-
Mizzi, S.; Sergheraert, C. Tetrahedron Lett. 2004, 45,
7081.
Acknowledgments
16. r1 receptors were extracted from guinea-pig cerebral
cortex according to Bowen et al.17 Membranes were
incubated with 2 nM [3H](+)pentazocine in 5 mM
K2HPO4/KH2PO4 buffer (pH 7.5) for 150 min at 22 °C.
Non-specific binding was determined under similar con-
ditions but in the presence of 10 lM unlabeled
haloperidol.18
17. Bowen, W. D.; de Costa, B. R.; Hellewell, S. B.; Walker,
M.; Rice, K. C. Mol. Neuropharmacol. 1993, 3, 117.
18. Ganapathy, M. E.; Prasad, P. D.; Huang, W.; Seth, P.;
Leibach, F. H.; Ganapathy, V. J. Pharmacol. Exp. Ther.
1999, 289, 251.
Thanks are due to Pr. Philippe Grellier (Museum dÕHis-
toire Naturelle de Paris, France) for CC50 evaluation,
´
´
Gerard Montagne for NMR spectra, and Herve Dro-
becq for MALDI-TOF experiments. Julie Charton was
`
a recipient of fellowships from the Ministere de lÕEduca-
tion Nationale et de la Recherche.
References and notes
1. Bowen, W. D. Pharm. Acta. Helv. 2000, 74, 211.
2. Senda, K.; Matsuno, K.; Okamoto, K.; Kobajashi, T.;
Nakata, K.; Mita, S. Eur. J. Pharmacol. 1996, 315, 1.
3. Senda, K.; Matsuno, K.; Kobajashi, T.; Nakazawa, M.;
Nakata, K.; Mita, S. Pharmacol. Biochem. Behav. 1998,
59, 129.
4. Kamei, H.; Noda, Y.; Kameyama, T.; Nabeshima, T. Eur.
J. Pharmacol. 1997, 319, 165.
5. Bergeron, R.; de Montigny, C.; Debonnel, G. J. Neurosci.
1996, 16, 1193.
19. r2 receptors were extracted from rat cerebral cortex
according to Bowen et al.17 Membranes were incubated
with 5 nM [3H]DTG (+300 nM (+)pentazocine) in 5 mM
K2HPO4/KH2PO4 buffer (pH 7.5) for 120 min at 22 °C.
Non-specific binding was determined under similar con-
ditions but in the presence of 10 lM unlabeled
haloperidol.
20. IC50 values from competitive inhibition experiments were
determined using the Marquardt–Levenberg nonlinear
curve fitting procedure of Xfit macro (Microsoft Excel).
21. Mossman, T. J. Immunol. Methods 1983, 65, 55.
22. Glennon, R. A.; Ismaiel, A. M.; Ablordeppey, S. A.; El-
Ashmawy, M.; Fisher, J. B. Bioorg. Med. Chem. Lett.
2004, 14, 2217.
6. Maurice, T.; Privat, A. Eur. J. Pharmacol. 1997, 328, 9.
7. Skuza, G. Pol. J. Pharmacol. 2003, 55, 923.
8. Maurice, T.; Martin-Fardon, R.; Romieu, P.; Matsumoto,
R. R. Neurosci. Biobehav. Rev. 2002, 26, 499.