Synthesis and in vitro evaluation of novel derivatives of diphenylsulfide as serotonin transporter ligands

Article abstract of DOI:10.1016/j.bmcl.2005.11.066

As the serotonin transporter (SERT) is involved in several neurodegenerative and psychiatric disorders, radiopharmaceuticals to image the SERT by PET or SPECT would be very valuable in studying these diseases. For the development of imaging agents, we hav

Full text of DOI:10.1016/j.bmcl.2005.11.066

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1297–1300
Synthesis and in vitro evaluation of novel derivatives
of diphenylsulfide as serotonin transporter ligands
Johnny Vercouillie,a Sylvie Mavel,a Laurent Galineau,a Tiziana Ragusa,a Robert Innis,b
a
e
a
Michael Kassiou,c,d Sylvie Chalon, Frederic Dolle, Jean-Claude Besnard,
´ ´
´
Denis Guilloteaua and Patrick Emonda,*
aUniversite´ Franc¸ois Rabelais de Tours, INSERM U619, Laboratoire de Biophysique Me´dicale et Pharmaceutique,
31 avenue Monge, 37200 Tours, France
bMolecular Imaging Branch, National Institute of Mental Health, Bethesda, MD 20892, USA
cDepartment of PET and Nuclear Medicine, Royal Prince Alfred Hospital, Missenden Road,
Camperdown, NSW, Sydney 2050, Australia
dDepartment of Pharmacology, University of Sydney, NSW 2006, Australia
eCEA-SHFJ, 4 place du Ge´ne´ral Leclerc, 91401 Orsay Cdx, France
Received 6 October 2005; revised 18 November 2005; accepted 18 November 2005
Available online 7 December 2005
Abstract—As the serotonin transporter (SERT) is involved in several neurodegenerative and psychiatric disorders, radiopharmaceu-
ticals to image the SERT by PET or SPECT would be very valuable in studying these diseases. For the development of imaging
agents, we have synthesized novel derivatives of recently reported diphenylsulfide SERT ligands, in which the sulfur atom linking
the two phenyl rings was replaced by an oxygen, sulfinyl, sulfonyl, amino or carbon group. Three of these exhibited good to high
in vitro affinity (0.5 nM < Ki < 11 nM) and selectivity for the SERT over the other monoamine transporters.
Ó 2005 Elsevier Ltd. All rights reserved.
The serotonin transporter (SERT) plays a pivotal role in
HO
X
N
N
CH3
NH2
the regulation of serotoninergic neurotransmission
by the reuptake of serotonin from the synaptic cleft into
the presynaptic nerve.1,2 Dysfunction of serotoninergic
neurotransmission has been shown to be involved in sev-
eral neurodegenerative3,4 and psychiatric disorders.5–7
In vivo SERT imaging in humans by positron emission
tomography (PET) or by single photon emission com-
puted tomography (SPECT) would assist in the early
diagnosis and follow-up of treatment in these diseases.
CH3
1
2'
R
S
2
1'
B
A
R4'
5'
403U76: R = S , X = Cl
IDAM: R = S , X = I
ODAM: R = O , X = I
ADAM: R = I
DASB: R = CN
MADAM: R = CH3
Figure 1. Potent SPECT or PET diphenylsulfide derivatives for SERT
imaging.
(KiSERT = 0.012 nM) and selectivity for SERT compared
to the DAT (dopamine transporter) and NET.9 Howev-
er, its in vivo properties are far from ideal requiring fur-
ther structural modifications to enhance its in vivo
behavior such as reduced peripheral metabolism, in-
creased brain uptake, and higher target to non-target ra-
tio. The substitution of the sulfur atom of IDAM by an
oxygen (ODAM, Fig. 1)10 results in decreased in vitro
affinity for the SERT (KiSERT = 0.171 nM), increased
brain uptake, slower kinetics, and peripheral metabo-
lism.10,11 These differences suggest that a minor change
in the chemical structure can result in significant changes
The diphenylsulfide, 403U76 (Fig. 1), has been reported
as a SERT and NET (norepinephrine transporter) inhib-
itor which provided the framework for the recent devel-
opment of selective SERT imaging agents.8 Its iodinated
analog, 5-iodo-2-[2-(dimethylaminomethyl)phenylthio]-
benzylalcohol (IDAM), displays high in vitro affinity
Keywords: SERT; Diphenylsulfide derivatives; MADAM derivatives;
Monoamine transporters.
*
Corresponding author. Tel.: +33247367242; fax: +33247367224;
e-mail: emond@univ-tours.fr
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.066

1298
J. Vercouillie et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1297–1300
to binding affinity and in vivo biodistribution
parameters.
O
N
O2N
O
N
O2
N
R1
a or b
R2
R
+
Moreover, it has recently been found that an amino
function at position 20 of ring B in addition to the
N,N-dimethylaminomethyl group in position 1 of the
phenyl ring A yields compounds such as ADAM,12
DASB,13 or MADAM14 (Fig. 1) with high in vitro and
in vivo affinity and selectivity for the SERT. Preliminary
in vivo results in monkeys with [11C]MADAM15 sup-
ported its use for in vivo SERT exploration by PET.
Moreover, [123I]ADAM16 and [11C]DASB17 have been
fully validated and, in view of their biological properties,
proposed as potent radiopharmaceuticals to explore the
SERT in humans by PET or SPECT.
R2 = OH, SH or I
R1 = Br or NH2
R
S
O
NH
NCH3
NEt
NBn
NPh
1
4
5
6
7
8
9
c
d
N
N
H2N
O2N
h
e
R
R
R
S
SO
SO2
O
NH
R
S
SO
SO2
O
1b
1a
On the basis of these results, we hypothesized that a ser-
ies of derivatives of the diphenylsulfide MADAM with
different bridge atoms should provide insight into SERT
binding properties.18 This hypothesis was supported by
a molecular modeling study which predicted that substi-
tution of the sulfur bridge by an amine function or a
methylene bridge should generate SERT compounds
with subnanomolar affinity.19
f
g
2b
3b
4b
5b
2a
3a
4a
5a
6a
7a
8a
9a
NH
6b NCH3
7b NEt
NCH3
NEt
NBn
NPh
NBn
NPh
8b
9b
Scheme 1. Reagents and conditions: (a) DMF, K2CO3, 80 °C, 5 h
(R=S and O); (b) DMF, K2CO3, Cu, reflux, 3 h (R=NH); (c) DMF,
NaH, 0–75 °C, 45 min to overnight and corresponding halide (methyl
iodide, ethyl iodide, and benzyl bromide); (d) iodobenzene, nitrotol-
uene, K2CO3, Cu, reflux, 7 h; (e) B2H6-THF, reflux 5 h and then rt
overnight; (f) AcOH, NaBO3 (1.2 equiv), reflux, 3 h; (g) AcOH, NaBO3
(2.4 equiv), reflux, 3 h; (h) SnCl2/HCl/MeOH, 5 °C to rt overnight.
The pharmacological evaluation of ÔO–Õ and ÔCH2–Õ
bridge analogs of ADAM has recently been reported.20
This study revealed that replacement of the sulfur atom
of ADAM by an oxygen or methylene group differently
affected SERT affinity and selectivity in vitro. All these
findings suggest that the atom linking the two phenyl
rings is critical in determining in vivo properties and
may be involved in the SERT ligand recognition
mechanism.
um carbonate, and copper in nitrotoluene (28% yield).
Reduction of the amide function of compounds 1, 4–9,
and 13 was achieved using diborane–THF complex to
provide the amino derivatives 1a,15 4a–9a, and 13a.
The sulfinyl (2a) and sulfonyl (3a) derivatives were pre-
pared by oxidation of compound 1a in acetic acid using
sodium perborate as oxidant21 (63% and 60% yield,
respectively). Scheme 2 describes the preparation of
derivatives 10b–12b. Compound 10a was synthesized
by nucleophilic addition of 4-lithio-3-nitrotoluene to
2-dimethylaminomethylbenzaldehyde. 4-Lithio-3-nitro-
toluene was prepared by halogen-metal exchange using
n-butyllithium in THF, as previously reported.22 2-
Dimethylaminomethylbenzaldehyde was prepared by
quenching 2-lithio-dimethylaminomethylbenzylamine
with DMF.23
To explore this hypothesis, we have synthesized and
evaluated in vitro eleven derivatives of MADAM with
different linkers between the phenyl groups.
All of the compounds (1b–13b) were synthesized as
outlined in Schemes 1–3.
Scheme 1 describes the preparation of compounds 1b–9b
which follows the same reaction sequence. The first
step gave derivatives 1, 4, and 5 (R = S, O, and NH,
respectively). Compounds 115 and 4 were prepared by
aromatic nucleophilic substitution, using 4-bromo-3-
nitrotoluene, potassium carbonate as base, dimethyl-
formamide (DMF), and 2-(mercapto or hydroxy)-N,
N-dimethylbenzamide. To synthesize 5, an Ullmann
condensation was performed with 2-iodo-N,N-dimeth-
ylbenzamide, K2CO3, copper, DMF, and 4-methyl-2-
nitroaniline (79% yield). The same reaction has been
used to prepare the diphenyl derivative 13 (Scheme 3).
The nitro group of derivatives 1a–13a was finally re-
duced using tin(II) chloride in concentrated hydrochlo-
ric solution and methanol (Schemes 1–3) to give the
corresponding amino derivatives 1b–13b (25–94%
yield).24
The amino compound 5 was used as starting material in
the preparation of the corresponding alkylated deriva-
tives: N-Me, N-Et, N-benzyl, and N-phenyl (6–9, respec-
tively). To prepare compounds 6–8, the reaction
conditions were: sodium hydride, DMF, and the corre-
sponding halo-derivatives (71–85% yield), whereas to
prepare compound 9, Ullmann reaction conditions were
used which included reaction of 5, iodobenzene, potassi-
In vitro affinities of derivatives 1b–13b were evaluated
by in vitro competition studies using tritiated ligands
of SERT ([3H]paroxetine), DAT ([3H]GBR12935), and
NET ([3H]nisoxetine) as previously detailed.25,26 For
each compound and each transporter, Ki values were
only determined when 100 nM of a target compound
inhibited at least 50% of tritiated ligand binding
(IC50 < 100 nM). Replacement of the MADAM sulfur

J. Vercouillie et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1297–1300
1299
N
N
N
O2N OH
O2N
O2N
R
Br
OHC
a
b or c
+
R
10a
11a CHOMe
12a
CO
R
N
H2N
10a
11a
12a
10b CHOH
11b CHOMe
d
R
12b
CO
Scheme 2. Reagents and conditions: (a) n-BuLi/THF/ꢀ100 °C; (b) DMF, NaH, methyl iodide, 0 °C, 45 min; (c) acetone, Jones reagent, 0 °C to rt,
overnight; (d) SnCl2/HCl/MeOH, 5 °C to rt overnight.
In addition, a MADAM analog was prepared in which
O
N
O2N
O2N
O
N
the two phenyl rings were directly linked (13b, Scheme
3) resulting in no SERT affinity (IC50 > 100 nM). It
could therefore be assumed that the bridge atom plays
a critical role in the binding and is necessary to obtain
ligands with high SERT affinity. These results contrast
with recent molecular modeling predictions where Well-
sow and Kovar19 predicted that the replacement of the
DASB sulfur bridge by an oxygen, an amino (NH or
NR) or a ketone group should lead to SERT derivatives
with subnanomolar affinity. Some of these predictions
were confirmed by our results, such as compounds 4b
and 5b (R=O and NH) which displayed Ki values of
0.53 and 10.28 nM, respectively. However, compounds
6b–9b and 12b (R = NR0 or CO), predicted to be SERT
ligands with high affinity, displayed IC50 > 100 nM.
Br
a
I
+
13
N
N
O2N
H2N
b
c
13a
13b
Scheme 3. Reagents and conditions: (a) DMF, Cu, reflux, 6 h; (b)
B2H6/THF, reflux 5 h, RT overnight; (c) SnCl2/HCl/MeOH, 5 °C to
RT overnight.
bridge with other atoms, or groups, led to compounds
without DAT or NET affinity (IC50 > 100 nM, Table
1). Because a N,N-dimethylaminomethyl group at the
1- and an amino group at 20-positions of the diphenyl-
sulfide structure have been found to be important for
the SERT selectivity,14 it can be suggested that these
substituents are responsible for low DAT and NET
affinities observed for compounds 1b–13b. Compounds
2b, 4b, and 5b exhibited moderate to high affinity to
the SERT (Ki = 4, 0.53, and 10 nM, respectively, Table
1), whereas the remaining compounds showed poor
SERT affinities (IC50 > 100 nM).
In addition, the O-bridged and CH2-bridged derivatives
of ADAM have recently been reported to display Ki
values of 0.37 and 48.6 nM for the SERT, respectively,
compared to 0.013 nM for ADAM.20 These results clearly
demonstrate that the bridge atom is a major component to
be considered in designing new highly potent SERT li-
gands and that direct involvement of this atom in
SERT-ligand recognition could be envisaged. To explore
this hypothesis, we calculated the partial charge on the
linking atom of the newly synthesized derivatives. As all
these compounds interact at the same binding site, and
positive and negative charges were found for both potent
and non-potent compounds (Table 1), it could be as-
sumed that this atom does not have a direct role in SERT
binding site recognition. However, because groups at
positions 1 and 20 have been found to play an important
role in SERT-ligand recognition,14,19 active or inactive
spatial conformations generated by the nature of the link-
ing atom may explain the differences in SERT affinity.
Table 1. Affinities of target compound for the SERT, NET, and DAT
Compound
R
Ki/IC50 (nM)a
Partial
charge
on Rb
SERT
NET DAT
1b
2b
S
SO
1.65 0.10 >100 >100
4.15 0.57 >100 >100
0.19
1.52
2.89
3b
4b
SO2
O
>100
>100 >100
0.53 0.04 >100 >100 ꢀ0.17
10.28 3.60 >100 >100 ꢀ0.26
5b
6b
NH
NMe
NEt
NBn
NPh
CHOH
CHOMe
CO
>100
>100
>100
>100
>100
>100
>100
>100
>100 >100 ꢀ0.26
>100 >100 ꢀ0.24
>100 >100 ꢀ0.23
>100 >100 ꢀ0.20
>100 >100 ꢀ0.09
Finally, comparison of SERT affinity of compound 1b
with that of compounds 2b and 3b could provide
insights into in vivo pharmacological properties. As
recently reported, diphenylsulfide has been shown to be
oxidized by cytochrome P450 monooxygenase and/or
by flavin-containing monooxygenase into the corre-
sponding sulfoxide and thereafter into its sulfone.27–29
MADAM could thus be oxidized in vivo into 2b which
also binds to the SERT and further oxidized into 3b which
does not bind to the SERT. This metabolic pathway for
other diphenylsulfides could lead to potent and/or non-
potent SERT metabolites, and thus must be considered
in the development of any SERT imaging agent.
7b
8b
9b
10b
11b
12b
13b
>100 >100
0.32
>100 >100 ꢀ0.03
>100 >100
—
—
a Ki values have been determined when IC50 < 100 nM. Ki values are
means of four experiments realized in duplicate. Radioligands
(DuPont NEN) [3H]paroxetine (SERT), [3H]nisoxetine (NET), and
[3H]GBR12935 (DAT).
b After ‘‘Ampac/Mopac’’ (vers. Insight II 98.0) calculations, in MSI
(San Diego, USA) package.

1300
J. Vercouillie et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1297–1300
11. Acton, P. D.; Kung, M. P.; Mu, M.; Plossl, K.; Hou, C.;
Siciliano, M.; Oya, S.; Kung, H. F. Eur. J. Nucl. Med.
1999, 26, 854.
12. Choi, S. R.; Hou, C.; Oya, S.; Mu, M.; Kung, M. P.;
Siciliano, M.; Acton, P. D.; Kung, H. F. Synapse 2000, 38,
403.
Based on these results, further development of the phen-
oxy derivative 4b for radiolabeling with carbon-11 and
in vivo pharmacological evaluation by PET is currently
underway.
13. Wilson, A. A.; Ginovart, N.; Schmidt, M.; Meyer, J. H.;
Threlkeld, P. G.; Houle, S. J. Med. Chem. 2000, 43,
3103.
14. Emond, P.; Vercouillie, J.; Innis, R.; Chalon, S.; Mavel, S.;
Frangin, Y.; Halldin, C.; Besnard, J. C.; Guilloteau, D.
J. Med. Chem. 2002, 45, 1253.
Acknowledgments
´
This work was supported by the Region Centre (France)
and CEA (LRC 21V). Radioligand binding assays were
performed by the National Institute of Mental Health
Psychoactive Drug Screening Program, supported by
NO1MH80005, at Case Western Reserve UniversityÕs
Biochemistry Department under the direction of B.L.
Roth (supported by KO2MH01366). We thank SAVIT
(Tours, France) for chemical analyses.
15. Tarkianen, J.; Vercouillie, J.; Guilloteau, D.; Gulyas, B.;
´
Sovago, J.; Cselenyi, Z.; Emond, P.; Chalon, S.; Sandell,
J.; Hiltunen, J.; Farde, L.; Halldin, C. J. Labelled Compd.
Radiopharm. 2001, 1, S193.
16. Acton, P. D.; Choi, S. R.; Hou, C.; Plossl, K.; Kung, H. F.
J. Nucl. Med. 2001, 42, 1556.
17. Wilson, A. A.; Ginovart, N.; Hussey, D.; Meyer, J.;
Houle, S. Nucl. Med. Biol. 2002, 29, 509.
Supplementary data
´
18. Emond, P.; Vercouillie, J.; Ragusa, T.; Mavel, S.; Dolle,
F.; Innis, R. B.; Chalon, S.; Guilloteau, D. J. Labelled
Compd. Radiopharm. 2003, 46, S160.
19. Wellsow, J.; Kovar, K. A.; Machulla, H. J. J. Pharm.
Pharm. Sci. 2002, 5, 245.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.2005.
11.066.
20. Kung, H. F.; Newman, S.; Choi, S. R.; Oya, S.; Hou, C.;
Zhuang, Z. P.; Acton, P. D.; Plossl, K.; Winkler, J.; Kung,
M. P. J. Med. Chem. 2004, 47, 5258.
References and notes
21. McKillop, A.; Tarbin, J. A. Tetrahedron 1987, 43,
1753.
22. Parham, W. E.; Picciliri, R. M. J. Org. Chem. 1977, 42,
257.
23. Viswanathan, C. T.; Wilkie, C. A. J. Organomet. Chem.
1973, 54, 1.
1. Masson, J.; Sagne, C.; Hamon, M.; El Mestikawy, S.
Pharmacol. Rev. 1999, 51, 439.
2. Pineyro, G.; Blier, P. Pharmacol. Rev. 1999, 51, 533.
3. Chinaglia, G.; Landwehrmeyer, B.; Probst, A.; Palacios,
J. M. Neuroscience 1993, 54, 691.
24. All new compounds gave analytical and spectroscopic
results consistent with the assigned structure.
25. Glennon, R. A.; Lee, M.; Rangisetty, J. B.; Dukat,
M.; Roth, B. L.; Savage, J. E.; McBride, A.; Rauser,
L.; Hufeisen, S.; Lee, D. K. J. Med. Chem. 2000, 43,
1011.
4. Tejani-Butt, S. M.; Yang, J.; Pawlyk, A. C. NeuroReport
1995, 6, 1207.
5. Owens, M. J.; Nemeroff, C. B. Clin. Chem. 1994, 40, 288.
6. Staley, J. K.; Malison, R. T.; Innis, R. B. Biol. Psychiat.
1998, 44, 534.
7. Stein, D. J. Lancet 2002, 360, 397.
26. Rothman, R. B.; Baumann, M. H.; Savage, J. E.; Rauser,
L.; McBride, A.; Hufeisen, S. J.; Roth, B. L. Circulation
2000, 102, 2836.
8. Ferris, R. M.; Brieaddy, L.; Mehta, N.; Hollingsworth, E.;
Rigdon, G.; Wang, C.; Soroko, F.; Wastila, W.; Cooper,
B. J. Pharm. Pharmacol. 1995, 47, 775.
27. Nnane, I. P.; Damani, L. A. J. Pharm. Biomed. Anal. 2002,
27, 315.
9. Oya, S.; Kung, M. P.; Acton, P. D.; Mu, M.; Hou, C.;
Kung, H. F. J. Med. Chem. 1999, 42, 333.
28. Nnane, I. P.; Damani, L. A. Life Sci. 2003, 73, 359.
29. Nnane, I. P.; Damani, L. A. Biomed. Chromatogr. 2004,
19, 87.
10. Acton, P. D.; Mu, M.; Plossl, K.; Hou, C.; Siciliano, M.;
Zhuang, Z. P.; Oya, S.; Choi, S. R.; Kung, H. F. Eur. J.
Nucl. Med. 1999, 26, 26.