Reactive derivatives for affinity labeling in the ifenprodil site of NMDA receptors

Article abstract of DOI:10.1016/j.bmcl.2008.04.019

To prepare thiol-reactive ifenprodil derivatives designed as potential probes for cysteine-substituted NR2B containing NMDA receptors, electrophilic centers were introduced in different areas of the ifenprodil structure. Intermediates and final compounds were evaluated by binding studies and by electrophysiology to determine the structural requirements for their selectivity. The reactive compounds were further tested for their stability and for their reactivity in model reactions; some were found suitable as structural probes to investigate the binding site and the docking mode of ifenprodil in the NR2B subunit.

Full text of DOI:10.1016/j.bmcl.2008.04.019

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2765–2770
Reactive derivatives for affinity labeling in the ifenprodil
site of NMDA receptors
Karine Alarcon,a,* Adeline Martz,a Laetitia Mony,b Jacques Neyton,b Pierre Paoletti,b
Maurice Goeldnera and Bernard Foucauda
aLaboratoire de Chimie Bioorganique, CNRS UMR 7175, Faculte´ de Pharmacie, Universite´ Louis Pasteur de Strasbourg,
74 Route du Rhin, F-67401 Illkirch, France
bLaboratoire de Neurobiologie, CNRS UMR 8544, Ecole Normale Supe´rieure, 46 rue d’Ulm, 75005 Paris, France
Received 19 March 2008; revised 7 April 2008; accepted 8 April 2008
Available online 11 April 2008
Abstract—To prepare thiol-reactive ifenprodil derivatives designed as potential probes for cysteine-substituted NR2B containing
NMDA receptors, electrophilic centers were introduced in different areas of the ifenprodil structure. Intermediates and final com-
pounds were evaluated by binding studies and by electrophysiology to determine the structural requirements for their selectivity.
The reactive compounds were further tested for their stability and for their reactivity in model reactions; some were found suitable
as structural probes to investigate the binding site and the docking mode of ifenprodil in the NR2B subunit.
ꢀ 2008 Elsevier Ltd. All rights reserved.
NMDA receptors (NMDARs) are a subtype of ionotro-
pic glutamate receptors widely distributed in the verte-
brate central nervous system. NMDARs play major
roles in both physiological and pathological states of
the CNS, including ischemic stroke, seizures, head trau-
ma and pain. NMDARs occur as hetero-oligomers com-
posed of NR1, NR2 (of which there are four: NR2A–
NR2D), and more rarely NR3 subunits.1,2 Because
non-selective NMDAR antagonists have impeding ad-
verse side effects, attention has focused toward drugs
capable of modulating selectively certain subtypes of
NMDARs. The NR2B-selective type of non-competitive
antagonists has a strong potential in this regard, show-
ing both neuroprotective and analgesic properties to-
gether with little side effects.3–5 It has been recently
shown that the N-terminal domain (NTD) of the
NR2B subunit forms the binding site of ifenprodil (1)6
and Ro256981,7 two prototypic NR2B-selective antago-
nists, while an extensive pharmaco-chemical research
had led to the description of common structural features
for this family of compounds.8–10 Our knowledge of
their binding site is based on homology 3D modeling
and mutational analysis: the NR2B NTD has been mod-
eled after a periplasmic protein from Escherichia coli
(LIVBP),6,11,12 as well as after the glutamate binding
domain of the mGluR1 receptor7: it is proposed to fold
as two lobes separated by a large central cleft.6 Mutating
residues located in the central cleft of the NR2B NTD
abolishes the high sensitivity to ifenprodil and deriva-
tives.6,12–14 Other biochemical evidences6,15–17 also sup-
port the location of the ifenprodil binding site in the
NTD of the NR2B subunit and in its cleft; however,
the effects of the mutation of residues not located in this
domain suggest alternative hypotheses.11
OH
A
N
B
HO
1
In order to propose a docking mode of ifenprodil in the
NTD of NR2B, we pursued a strategy combining cys-
teine-scanning mutagenesis and affinity labeling.18 In
this strategy, the formation of a covalent bond between
a cysteine-reactive ligand derivative and a cysteine-
substituted receptor enables to identify direct interac-
tions between the ligand and a precise region of the
receptor. In the recent years, we have applied this
strategy to explore the glycine binding site (NR1 sub-
unit) of the NMDA receptor.19,20 Our results were ob-
tained using full-length, membrane-inserted NR1/NR2
receptors; they provided our structural homology model
Keywords: Ifenprodil; NR2B; NMDA.
*
Corresponding author. Tel.: +0390244161; fax: +0390244306.
e-mail: alarcon@bioorga.u-strasbg.fr
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.04.019

2766
K. Alarcon et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2765–2770
with an experimental basis and established a general
docking mode for antagonists at the glyB site. These
were remarkably consistent with the crystal structure
of the agonist binding domain of the NR1 subunit
(S1–S2 soluble fragment).21
phenol group into a methoxy derivative (to avoid reac-
tion in the ortho position25), a Mannich reaction with
paraformaldehyde and 4-benzylpiperidine under acidic
catalysis yielded condensation products where two ami-
nomethyl groups have been introduced.26 Finally, the
elimination of one benzylpiperidine group on silica gel
produced the propenone derivatives 8 and 9.
This letter describes the stepwise approach toward thiol-
reactive ifenprodil derivatives suitable for structural
investigations of the binding site for NR2B selective
inhibitors of the NMDARs. The affinity and reactivity
of the derivatives should indeed be high enough for
the occurrence of a covalent bond formation with a cys-
teine-mutated receptor while they occupy the binding
site.22
In compound 12, ifenprodil’s asymmetric centers have
been suppressed and a reactive-NCS function was
introduced on aromatic ring A. A high yield synthesis
of the nitro precursor 11 was achieved by coupling p-
hydroxybromophenacyle27 to 4-(4-nitrobenzyl)-piperi-
dine, which was prepared in a three-step procedure
from 4-benzylpiperidine.28 Compound 10 was obtained
from commercially available 4-benzylpiperidine using
the same synthetic pathway. Ligand 12 was synthe-
sized by the hydrogenation of the nitro compound
11 followed by treatment with thiophosgene in a tetra-
hydrofurane/aqueous Na2CO3 solution mixture.
Compound 13, designed to estimate the influence of
asymmetry in ifenprodil, was obtained by the reduc-
tion of precursor 10 with sodium borohydride. Com-
pounds 2–13 were fully characterized by 1H NMR
and HRMS. The purity of the reactive compounds
was checked by HPLC.29
A reliable docking of ifenprodil-like compounds being
our final purpose, we have synthesized molecules consis-
tent with their main structural features: two aryl rings,
one located in a hydrophobic environment (ring A)
and a ring substituted by polar groups (ring B), while
a tertiary amine, at the center of the molecule, is
connected to each ring by a linker.23 A distance of 9–
˚
11 A between the two rings, in an extended conforma-
tion of the molecule, was also suggested.9 Electrophilic
substituents were introduced in carbon positions which
were selected to explore the different areas of the binding
site corresponding to these components of the ligand’s
structure. Intermediate and final compounds were tested
as ligands for the NR2B NTD site, and, for some of
them, as effectors of the recombinant NR1/NR2B
NMDARs. The reactive compounds were also tested
for their stability and their reactivity toward thiols
in model conditions. Their potential as structural
probes for the ifenprodil binding site of NMDARs is
discussed.
The affinities of compounds 2–13 were measured by
equilibrium binding to rat brain membranes in competi-
3
3
tions against H-ifenprodil as in Ref. 30: H-ifenprodil
(2.501 Gbq./mmol from Perkin-Elmer Life Sciences) was
incubated with the membranes for 2 h at 4 ꢁC in 5 mM
Tris–chloride buffer, pH 7.4, in the absence or presence
of 10 lM Ro-084304 for the determination of the non-
specific binding. KIs were calculated from the measured
IC50s31 and from a KD of 11 1.5 nM (n = 5) for ifen-
prodil, that we measured following a previously pub-
lished protocol30 (Table 1).
The synthetic pathways for these compounds can be
found in Scheme 1. In compounds 4 and 5, the hydrox-
yle group of ring B in ifenprodil has been replaced by a
chloroacetamido group or an isothiocyanate group,
respectively; moreover, one asymmetric center has been
suppressed by converting, on the linker, the secondary
alcohol into a ketone. Its synthesis involves first a Fri-
edel-Crafts24 reaction between acetanilide and 2-bromo-
propionyl chloride. Subsequently, the substitution of the
bromide by 4-benzylpiperidine yielded precursor 2. The
thiol-reactive probes 4 and 5 were finally obtained by the
deprotection of the aromatic amine affording compound
3, followed by functionalization.
The replacement of the p-hydroxyl group in ring B of
ifenprodil by an amino group (compound ( ) 7) de-
creases its affinity by ca. one order of magnitude, while
the conversion of this amine group of precursor 3 into
amide 2, chloroacetamide 4 and isothiocyanate 5 further
shifts the KI value by a factor of 1.8, 8.5, and 12, respec-
tively: this is consistent with the magnitude of the H-
bond donor property of substituents in this position.8,10
The size and rigidity of these substituents are also likely
to interfere: thus, the higher KI value of 5 compared to 4
may be due to the fact that the isothiocyanate group,
although smaller, is less flexible than the chloroacet-
amide group.
Compounds ( ) 6 and ( ) 7 conserved the structure of
ifenprodil; they differed only by the presence of the ami-
no function instead of the phenolic group. Their synthe-
sis involved the reduction of ketone 2, affording 4
diastereoisomers which were separated by column chro-
matography as erythro- and threo-racemic mixtures.
The final deprotections yielded racemic threo-( ) 6
and erythro-( ) 7 derivatives.
Similarly, the effect of substituting the p -position of
ring A is evidenced by the KI values of compounds
10, 11 and 12: the substitution of a hydrogen by a ni-
tro-group and its transformation into an isothiocya-
nate group moderately influence the KI value.
Indeed, the total penalty for introducing this reactive
group in this position is only a factor of 1.7; this po-
sition on ring A thus appears to be less sensitive to
substitution than the corresponding position on ring
B, in agreement with previous findings.8,10
Compounds 8 and 9, whose linker, compared to the ref-
erence ifenprodil molecule, contains a Michael acceptor
together with an additional carbon atom required a dif-
ferent synthetic pathway. After conversion of the B ring

K. Alarcon et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2765–2770
2767
O
O
O
iii
N
i, ii
N
N
N
iv or v
O
O
N
H
O
R
N
H
H N
2
3
2
Cl
4
5
R =
HN
vi, iii
R = NCS
OH
threo ( ) 6
erythro
( ) 7
H N
2
O
O
O
O
N
vii
ix
viii
N
1
1
R
N
1
1
R
R
R
1
1
1
1
1
R = OMe
R = NO
2
8
9
R = OH
R
R
= OMe
R
R
= OMe
1
1
1
R = NO
= NO
= NO
2
2
2
O
Br
, xiii
O
O
x, xi, xii
xiv, v
HO
N
N
2
NCS
R
HN
2
HN
2
R
R
2
2
2
HO
HO
R = H
12
R
R
= H
= NO
2
10
11
R
R
= H
= NO
vi
2
2
2
OH
N
HO
13
Scheme 1. Reagents and conditions: (i) ClCOCH(CH3)Br, AlCl3, CS2, reflux, 17%; (ii) 4-benzylpiperidine, Et3N, CH3CN, reflux, 66%; (iii) HCl 37%,
reflux, 89%; (iv) ClCOCH2Cl, Et3N, CH2Cl2, 0 ꢁC, 69%; (v) CSCl2, NaHCO3, THF/H2O 1:1, rt, 96%; (vi) NaBH4, MeOH, 40 ꢁC, 88%; (vii) CH3I,
K2CO3, acetone, reflux, 100%; (viii) HCHO, 4-benzylpiperidine, HCl, EtOH, reflux, 100%; (ix) silica gel heptane/ethyl acetate 1:1, 40%; (x)
(CF3CO)2O, CH2Cl2, rt, 100%; (xi) HNO3, CF3COOH, À30 ꢁC to rt, 37%; (xii) NaOH, EtOH, rt, 61%; (xiii) DBU, THF, rt, 84%; (xiv) H2, Pd/C,
MeOH/CH2Cl2 1:1, rt, 100%.
In our cysteine affinity labeling approach, one should
apply to cysteine-substituted receptors only reactive li-
gands of well defined configuration, for a useful inter-
pretation of their effect in terms of binding-site
structure and ligand docking. Asymmetric centers gener-
ate mixtures of enantiomeric pairs. We therefore at-
tempted to suppress the two asymmetric centers in the
spacer arms while keeping as much as possible of their
properties in length and polarity. With regard to the
asymmetric carbon atom closest to ring B (a carbon),
compounds ( ) 6 and ( ) 7 represent two diastereoiso-
meric pairs of enantiomers. Forming a ketone on this
a carbon (instead of an alcohol function) shifts the KI
value from 97 to 58 nM (compounds ( ) 7 and 3). How-
ever, when there is no asymmetric center on the b car-
bon, the same change (from compound 13 to
compound 10) induces an opposite shift in KI value,
from 62 to 141 nM, while the suppression of the asym-
metric center on the b carbon by the replacement of
the linker’s methyl group of ifenprodil 1 with a hydrogen
uents on ring B (the p-amino- or p-hydroxy-substituted
compounds could not be isolated).
The stability of the reactive compounds 4, 5, 8, 9 and
different times, after dilution 1/10 in frog’s Ringer 12,
were analyzed by HPLC: aliquots were injected, at buf-
fer (pH 7.4) of a 10 mM stock solution in DMSO, and
the time-dependent decrease of the peak area of each
compound was followed. Their reactivity toward thiols
was evaluated, using the same technique, by determining
their half-life in the presence of excess of N-acetyl cys-
teine methyl ester (NACME). As the reaction rate varies
linearly with the concentration of the reagents (pseudo-
first order reaction), these were adjusted for an accurate
measurement of the decay of the probes. Conditions and
results can be found in Table 2. All five compounds re-
act with thiols within minutes at the concentrations
used; their stability in Ringer’s buffer is appropriate
for their use in receptor binding and activity assays.
The compared reactivity of these compounds is consis-
tent with their chemical structure: isothiocyanates react
faster than chloroacetamides (5 and 12 vs 4) and, among
Michael acceptors (both highly reactive), the electron
attracting p-nitro substituent confers to 9 a 110-fold
higher reactivity over 8, in which electron-donor meth-
oxy group is in the same position.
`
atom, in 13, produces a sixfold decrease in affinity, from
a KD value of 11.2 nM to a KI value of 62 nM. In a race-
mic mixture, the presence of the less active compound
decreases the apparent affinity of its isomer only mod-
estly.32 Similarly, the KD for ifenprodil and the KI values
in Table 1 (compounds 2, 3, 4, 5, ( ) 6, ( ) 7, and 13) are
composite values. The maximum effect of the suppres-
sion of the two asymmetric centers of ifenprodil is a de-
crease in affinity by a factor of 13 (from ifenprodil to
compound 10), which is likely to confer to the achiral
derivatives sufficient affinity to undergo functionaliza-
tion for affinity labeling experiments. However, com-
pounds 8 and 9 have no asymmetric center, but their
high KI values can also result from unfavourable substit-
At submicromolar concentrations, ifenprodil selec-
tively inhibits currents carried by NR1/NR2B recep-
tors.33 This high-affinity inhibition is non-competitive
(toward the agonist glutamate) and does not depend
on the transmembrane voltage (as expected for a bind-
ing site located outside the ionic channel). The biolog-
ical activity of the reactive compounds in Table 2 was

2768
K. Alarcon et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2765–2770
Table 1. Binding data for ifenprodil and compounds 2–13
ifen
Name or number
Compound
OH
KIa
Fold shift KIn/KD
N
Ifenprodil ( ) 1b
11 1.5c
1
HO
O
O
N
2
107 11
58 5.3
493 56
679 34
300 35
97 10
10
5.3
45
62
27
8.8
48
23
13
18
22
16
N
H
O
N
3
H2N
O
N
Cl
O
4
N
H
O
N
5
S
C
N
OH
OH
N
N
( ) 6b
( ) 7b
8
H2N
H2N
O
O
527 33
248 44
141 19
200 14
245 33
N
O
N
9
O2N
O
N
10
HO
HO
HO
HO
O
N
N
N
11
NO2
O
12
NCS
OH
13
62
8
a Values are given in nM SEM (n = 3).
b Ifenprodil, compounds 6 and 7 were tested as racemic erythro- or threo-derivatives (see text).
c KD.
evaluated by performing two-electrode voltage-clamp
measurements on Xenopus oocytes expressing wild-
type NR1/NR2B receptors (ref 6). NMDA currents
were induced by saturating concentrations of L-gluta-
mate and glycine (100 lM each), and the effects of
the different compounds were assessed by measuring
the change in current size induced by the application
of the compound during a NMDA response. Com-

K. Alarcon et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2765–2770
Table 2. Stability in buffer and reactivity versus thiols of reactive compounds
2769
Compound
t1/2 for stability
Reactivity
[NACME] (mM)
500
0.5
10
0.1
100
[Compound] (mM)
t1/2 (min)
Relative reactivity
4
>1 day
20 h
5 h
0.1
0.005
0.1
114
3.24
7
5
704 · 103
814
92 · 103
2.5
8
9
12
5 h
>1 daya
0.01
1
62
23
a An apparent decrease of concentration of compound 12 results from its slow precipitation in buffer, without the formation of degradation products.
Table 3. Activity of thiol-reactive ifenprodil derivatives at wild-type
NR1/NR2B receptors
voltage dependent) and 4 (IC50 = 230 lM [n = 3]);
compound 8 was found inactive, as with the NR2A sub-
unit (no effect up to 100 lM). Altogether, these results
show that the synthesized compounds have a common
and subunit-specific target on NMDARs, the NR2B
NTD.
ifenprodil
IC50compound/IC50
Compound
IC50 (lM)
1 (Ifenprodil)
4
0.11
1
14 7 (n = 5)
70 10 (n = 3)
16 6 (n = 4)
130
640
145
8
12
In summary, we have synthesized ifenprodil-derived
compounds to design NR2B selective ligands with a
reactive group in different parts of the molecule. These
ligands were assayed for their ability to be recognized
by the ifenprodil binding site and to produce the ex-
pected NR2B-NTD mediated allosteric antagonism of
NMDAR mediated currents. Characterized as stable
in buffer but reactive with thiol groups, these com-
pounds, like ifenprodil, present a pharmacological pro-
file consistent with their specific binding to the
modulatory NTD of the NR2B subunit. Although these
compounds displayed a decreased potency compared to
ifenprodil, they appear to be valuable tools for a further
characterization of the ifenprodil binding site, its topo-
logical analysis, and the docking of cognate ligands.
pound 5, which displayed the highest KI value,
showed only very weak inhibition of NR1/NR2B
receptors (8% current inhibition with 10 lM of 5).
Compounds 4, 8, and 12 were much better at inhibit-
ing NR1/NR2B receptors, with IC50s of 14, 70, and
16 lM, respectively (Table 3). For comparison, in
the same test, the IC50 for ifenprodil was found to
be of 0.11 lM, a value very similar to what has been
previously reported.6 When tested by electrophysiol-
ogy, the compounds appear less potent compared to
binding studies, as their activity is decreased by at
least two orders of magnitude compared to ifenprodil.
Moreover, compounds 4 and 12 show similar IC50s,
whereas their KI values are twofold different. However,
a general trend is followed: the derivatives which have
the lowest KI value also have the lowest IC50. Surpris-
ingly, compound 9 was found to be inactive, since it
had no effect on NMDA responses at 10 lM. The
example of compound 9 suggests the existence of
NR2B-NTD compounds functionally ‘silent’, that is,
compounds unable to modulate the receptor activity
in spite of their ifenprodil-competitive binding.
Acknowledgments
We wish to thank the Roche Company for the kind gift
of Ro-084304 as well as the Centre National de la
´
Recherche Scientifique, the Universite Louis Pasteur,
the Institut National de la Sante et de la Recherche
´
´
Medicale, the Agence Nationale de la Recherche and
the Ministere de la Recherche et de la Technologie for
`
To ascertain that the inhibitory effects seen with the
compounds were specific to the NR2B subunit and
engaging the NR2B NTD, we repeated some of the
above experiments on oocytes expressing either wild-
type NR1/NR2A receptors or NR1/NR2B receptors
truncated for the entire NR2B-NTD (NR1/N2B DNTD;
see Ref. 14). As expected for compounds targeting selec-
tively the NR2B NTD, compounds 4, 8, and 12 have lit-
tle effects on NR1/NR2A receptors with EC50s
estimated to be >100 lM (25% [n = 3] and 11%
[n = 3] inhibition for compounds 4 and 8 at 100 lM).
The IC50 for compound 12 was shifted from 16
(NR2B, n = 4) to 150 lM (NR2A, n = 5) and became
voltage dependent, as expected if the observed inhibition
on NR1/NR2A receptors is mostly due to channel block
by the compound.6 When the NTD segment was trun-
cated from the NR2B subunit, similar rightward shifts
were observed forcompounds 12 (IC50 = 570 lM, [n = 4],
their material and financial support.
References and notes
1. Cull-Candy, S. G.; Brickley, S.; Farrant, M. Curr. Opin.
Neurobiol. 2001, 11, 32.
2. Paoletti, P.; Neyton, J. Curr. Opin. Pharmacol. 2007,
7, 39.
3. Chizh, B. A.; Headley, P. M.; Tzschentke, T. M. Trends
Pharmacol. Sci. 2001, 22, 636.
4. Parsons, C. G. Eur. J. Pharmacol. 2001, 429, 71.
5. Kemp, J. A.; McKernan, R. M. Nat. Neurosc. 2002, 5,
1039.
6. Perin-Dureau, F.; Rachline, J.; Neyton, J.; Paoletti, P.
J. Neurosci. 2002, 22, 5955.
7. Malherbe, P.; Mutel, V.; Broger, C.; Perin-Dureau, F.;
Kemp, J. A.; Neyton, J.; Paoletti, P.; Kew, J. N. C.
J. Pharmacol. Exp. Ther. 2003, 307, 897.

2770
K. Alarcon et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2765–2770
8. Chenard, B. L.; Menniti, F. S. Curr. Pharm. Des. 1999, 5,
381.
23. Tamiz, A. P.; Whittemore, E. R.; Zhou, Z. L.; Huang, J.
C.; Drewe, J. A.; Chen, J. C.; Cai, S. X.; Weber, E.;
Woodward, R. M.; Keana, J. F. W. J. Med. Chem. 1998,
41, 3499.
24. Gershon, H.; Scala, A.; Parmegiani, R. J. Am. Chem. Soc.
1965, 8, 877.
25. Cromwell, N. H. J. Am. Chem. Soc. 1946, 68, 2634.
9. Kornberg, B. E.; Nikam, S. S.; Wright, J. L.; Kesten, S.
R.; Meltzer, L. T.; Coughenour, L.; Barr, B.; Serpa, K. A.;
McCormick, J. Bioorg. Med. Chem. Lett. 2004, 5, 1213.
10. Nikam, S. S.; Meltzer, L. Curr. Pharm. Des. 2002, 8, 845.
11. Masuko, T.; Kashiwagi, K.; Kuno, T.; Nguyen, N. D.;
Pahk, A. J.; Fukuchi, J.; Igarashi, K.; Williams, K. Mol.
Pharmacol. 1999, 55, 957.
12. Paoletti, P.; Perin-Dureau, F.; Fayyazuddin, A.; LeGoff,
A.; Callebaut, I.; Neyton, J. Neuron 2000, 28, 911.
13. Rachline, J.; Perin-Dureau, F.; Le Goff, A.; Neyton, J.;
Paoletti, P. J. Neurosci. 2005, 25, 308.
14. Herin, G. A.; Aizenman, E. Eur. J. Pharmacol. 2004, 500,
101.
15. Hawkins, L. M.; Chazot, P. L.; Stephenson, F. A. Br. J.
Pharmacol. 1999, 1p, 126.
16. Gallagher, M. J.; Huang, H.; Pritchett, D. B.; Lynch, D.
R. J. Biol. Chem. 1996, 271, 9603.
17. Wong, E.; Ng, F. M.; Yu, C. Y.; Lim, P.; Lim, L. H.;
Traynelis, S. F.; Low, C. M. Protein Sci. 2005, 14, 2275.
18. Foucaud, B.; Perret, P.; Grutter, T.; Goeldner, M. Trends
Pharmacol. Sci. 2001, 22, 170.
26. Girreser, U.; Heber, D.; Schutt, M. Synthesis 1998, 715.
¨
27. Givens, R. S.; Park, C.-H. Tetrahedron Lett. 1996, 37,
6259.
28. Imamura, S.; Nishikawa, Y.; Ichikawa, T.; Hattori, T.;
Matsushita, Y.; Hashiguchi, S.; Kanzaki, N.; Iizawa, Y.;
Baba, M.; Sugihara, Y. Bioorg. Med. Chem. 2005, 13, 397.
29. HPLC retention time on a Analytical C18 Acclaim (4.6 ·
300 mm) column using a 30 min linear gradient from 0%
to 50% acetonitrile in 0.1% TFA water solution at 1 ml/
min, for compounds 4 and 12, and 20 min similar linear
gradient for compounds 8 and 9: 22.7, 25.5, 20.6 and 19.9
min, respectively. For compound 5, the linear gradient was
as for compounds 4 and 12, except for the acetonitrile
concentration, from 20% to 70%; retention time: 24.7 min.
30. Mutel, V.; Buchy, D.; Klingelschmidt, A.; Messer, J.;
Bleuel, Z.; Kemp, J. A.; Richards, J. G. J. Neurochem.
1998, 70, 2147.
19. Foucaud, B.; Laube, B.; Schemm, R.; Goeldner, M.; Betz,
H. J. Biol. Chem. 2003, 278, 24011.
20. Kreimeyer, A.; Laube, B.; Sturgess, M.; Goeldner, M.;
Foucaud, B. J. Med. Chem. 1999, 42, 4394.
31. Cheng, Y.-C.; Prusoff, W. H. Biochem. Pharmacol. 1973,
22, 3099.
´
32. Avenet, P.; Leonardon, J.; Besnard, F.; Graham, D.;
21. Furukawa, H.; Gouaux, E. EMBO J. 2003, 22, 2873.
22. Foucaud, B.; Alarcon, K.; Sakr, E.; Goeldner, M. Curr.
Chem. Biol. 2007, 1, 271.
Frost, J.; Deporteere, H.; Langer, S. Z.; Scatton, B. Eur. J.
Pharmacol. 1996, 296, 209.
33. Williams, K. Mol. Pharmacol. 1993, 44, 851.