Design and Synthesis of Selurampanel, a Novel Orally Active and Competitive AMPA Receptor Antagonist

Article abstract of DOI:10.1002/cmdc.201600467

A series of potent quinazolinedione sulfonamide antagonists of the α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptor were designed and synthesized. The structure–activity relationships (SAR) and in vivo activity of the series were investigated. In particular, compound 1 S (selurampanel; N-[7-isopropyl-6-(2-methylpyrazol-3-yl)-2,4-dioxo-1H-quinazolin-3-yl]methanesulfonamide) has shown excellent oral potency against maximal electroshock seizure (MES)-induced generalized tonic–clonic seizures in rodents as well as significant activity in patients suffering from various forms of epilepsy. The X-ray crystal structure of selurampanel bound to the AMPA receptor hGluA was also obtained.

Full text of DOI:10.1002/cmdc.201600467

DOI: 10.1002/cmdc.201600467
Communications
Design and Synthesis of Selurampanel, a Novel Orally
Active and Competitive AMPA Receptor Antagonist
David Orain,[a] Engin Tasdelen,[a] Samuel Haessig,[a] Manuel Koller,[a] Anne Picard,[a]
Celine Dubois,[a] Kurt Lingenhoehl,[b] Sandrine Desrayaud,[c] Phillip Floersheim,[a]
David Carcache,[a] Stephan Urwyler,[b] Joerg Kallen,[d] and Henri Mattes*[a]
A series of potent quinazolinedione sulfonamide antagonists
of the a-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid
(AMPA) receptor were designed and synthesized. The struc-
ture–activity relationships (SAR) and in vivo activity of the
series were investigated. In particular, compound 1S (seluram-
panel; N-[7-isopropyl-6-(2-methylpyrazol-3-yl)-2,4-dioxo-1H-qui-
nazolin-3-yl]methanesulfonamide) has shown excellent oral po-
tency against maximal electroshock seizure (MES)-induced gen-
eralized tonic–clonic seizures in rodents as well as significant
activity in patients suffering from various forms of epilepsy.
The X-ray crystal structure of selurampanel bound to the
AMPA receptor hGluA was also obtained.
methyl-d-aspartate (NMDA), kainic acid, and a-amino-3-hy-
droxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors.
For every class of iGluR several subunits are known: NR1,
NR2A–D, NR3A,B for NMDA receptors, GluK1–5 for kainate re-
ceptors, and GluA1–4 for AMPA receptors. iGluRs mediate
rapid excitatory neurotransmission and are linked to a variety
of neurodegenerative and psychiatric diseases such as ischemic
brain damage, amyotrophic lateral sclerosis, schizophrenia,
and, most significantly, epilepsy.[5] Overstimulation of the gluta-
matergic system plays a role in triggering seizures associated
with epilepsy.[6]
AMPA receptor antagonists are currently in clinical develop-
ment as AEDs, and there is evidence from various sources that
this mechanism of action is likely to be of therapeutic use in
epilepsy. First, AMPA receptor antagonists prevent excessive
neuronal activation leading to epileptic seizures.[7,8,9] Second,
they potently and efficiently block seizures in animal models of
seizures and epilepsy. Third, changes in the distribution, ex-
pression, and editing of AMPA receptors have been reported in
human epileptic hippocampal tissue.[10–16] Fourth, in Rasmus-
sen’s encephalitis, an autoimmune disease, one of the autoan-
tigens causing the disease is directed against the AMPA GluA3
receptor.[17–19] And finally, there is direct clinical evidence that
the non-competitive AMPA receptor antagonist perampanel,
which has been approved for clinical use, is effective in pa-
tients with partial-onset seizures.[9]
Despite the fact that several antiepileptic drugs (AEDs) are on
the market, there is still a high medical need for improved
treatments of epilepsy, as about 40% of patients are inade-
quately controlled[1] or suffer from adverse events.[2,3] Current
treatment strategies focus on three main biological targets:
the g-aminobutyric acid (GABA) system, voltage-gated sodium
channels, and calcium channels.[4]
Based on this information, AEDs with new mechanisms of
action are of interest. Because glutamate is the principal excita-
tory neurotransmitter of the central nervous system where it
acts at both ionotropic and metabotropic receptors, a novel
approach to treat epilepsy could be by modulation of the glu-
tamatergic system and in particular the AMPA receptor. Metab-
otropic glutamate receptors (mGluRs) are G-protein-coupled
receptors and, to date, eight subtypes, subdivided into three
groups, are described. Ionotropic glutamate receptors (iGluRs)
are classified according to their agonist selectivity toward N-
The quinazolinedione-sulfonamide structural class presented
herein was designed as hybrid between two well-known
classes of AMPA antagonists, namely kynurenic acids and qui-
noxalinediones (Figure 1).[20,23–25] Herein we report the prepara-
tion and biological activity of a set of novel quinazolinedione-
sulfonamides as potent and orally bioavailable AMPA antago-
nists.
[a] Dr. D. Orain, E. Tasdelen, S. Haessig, Dr. M. Koller, A. Picard, C. Dubois,
Dr. P. Floersheim, Dr. D. Carcache, Dr. H. Mattes
Global Discovery Chemistry, Novartis Institute for Biomedical Research,
4002 Basel (Switzerland)
The synthesis of quinazolinedione-sulfonamides 1 was ac-
complished by following the general procedure highlighted for
the 6-isopropyl derivatives (Scheme 1). Starting from commer-
cially available 4-isopropyl-1-methyl-2-nitrobenzene, Vilsmaier
E-mail: henri.mattes@novartis.com
[b] Dr. K. Lingenhoehl, Dr. S. Urwyler
Previously Neuroscience Disease Area, Novartis Institute for Biomedical Re-
search, 4002 Basel (Switzerland)
[c] Dr. S. Desrayaud
Metabolism and Pharmacokinetics, Novartis Institute for Biomedical Re-
search, 4002 Basel (Switzerland)
[d] Dr. J. Kallen
Center for Proteomic Chemistry, Novartis Institute for Biomedical Research,
4002 Basel (Switzerland)
The ORCID identification number(s) for the author(s) of this article can
be found under http://dx.doi.org/10.1002/cmdc.201600467.
Figure 1. AMPA antagonists.
ChemMedChem 2016, 11, 1 – 6
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
fluoromethylbenzoic acid methyl was separated.[23] Detailed ex-
perimental procedures can be found in Allgeier et al.[23]
The affinity of new compounds (Tables 1 and 2) at the AMPA
receptor was assessed by measuring the inhibition of the bind-
ing of [3H]CNQX at the AMPA receptor according to previously
described methods.[26] The in vivo activity of the compounds
was assessed in the audiogenic seizure test in DBA/2 (Dilute
Brown Agouti) mice which is a model of reflex epilepsy.[27–29] As
this model does not discriminate between different drug
mechanisms and because seizures cannot be provoked in
humans by auditory stimulation, best compounds were also
assessed in the maximal electroshock seizure (MES) test in ro-
dents, which is highly predictive for the activity of drugs
against generalized tonic–clonic seizures in humans.[30,31]
It was found that various substituents at position 7 imparted
moderate affinity and oral activity to the corresponding ana-
logues, as shown in Table 1. In a first optimization cycle, 6-imi-
Table 1. Binding affinity, in vivo activity of selected quinazolinediones 1.
Scheme 1. Representative synthesis of quinazolinedione-sulfonamide. Re-
agents and conditions: a) reflux (1508C oil bath), 20 h, 95%; b) KMnO4, KOAc,
tBuOH/H2O, 258C 1.5 h, quant. (crude); c) 1. SOCl2, reflux, 2. MeOH, 86%;
d) H2, MeOH, 10% Pd/C, 97%; e) I2, Ag2SO4, EtOH, 60%; f) [bistriphenylphos-
phine]dichloropalladium dioxane (0.1 equiv), RÀSnBu3 (1.2 equiv), 858C,
18 h; g) 1. 4-chlorophenylchloroformate (1 equiv), Et3N (2 equiv), CH2Cl2, RT,
4 h, then evaporation, 2. methanesulfonyl hydrazide (1.2 equiv), THF, RT,
30 min, then 1m NaOH (1 equiv), RT, 30 min.
Compd
R6
R7
IC50 [mm][a]
AS p.o.
ED50 [mgkgÀ1
[b]
]
1A
1B
1C
1D
H
H
H
H
tBu
Cl
CF3
NO2
4.9Æ0.1
3.6Æ0.2
0.7Æ0.1
1.2Æ0.1
9.7
15
24
reaction yielded enamine intermediate 2,[21] which was oxida-
tively cleaved to carboxylic acid 3 using potassium permanga-
nate. Esterification of the carboxylic acid followed by reduction
of the nitro group yielded compound 5. Regioselective iodina-
tion with iodine and silver sulfate afforded 6.[22] Subsequent
palladium-mediated carbon–carbon couplings of this inter-
mediate with various heteroaryl stannane reagents provided
the required 4-isopropyl-5-heteroaryl-disubstituted 2-amino-
benzoates 7. Final ring closure of 7 was effected by treatment
with 4-chlorophenylchloroformate to generate the activated
carbamate derivative, which was then allowed to undergo re-
action with methanesulfonyl hydrazide to furnish the expected
products 1 after base-catalyzed cyclization. The 6-chloro, 6-
methyl, and 6-trifluoromethyl analogues were synthesized in
a similar manner by starting from the corresponding commer-
cially available 2-aminobenzoates.
0% @ 30 mgkgÀ1
1E
CF3
0.046Æ0.01
0% @ 30 mgkgÀ1
[a] Rat AMPA [3H]CNQX binding. [b] Audiogenic seizure in DBA/2 mice.
Values are the geometric mean of at least three independent experi-
ments.
dazolyl and 6-triazolyl derivatives were synthesized.[23,24] These
substituents were reported to substantially increase affinity of
related scaffolds, for example, quinoxalinedione, for the AMPA
receptor.[32] In line with published data, the new analogues em-
bedding these 6-substituents showed high affinity for the
AMPA receptor. However, despite good improvement in bind-
ing affinity, the resulting products showed no in vivo activity,
following oral administration of the drug, as shown in Table 1.
Pharmacokinetic profiling of 1E[25] indicated a very low bio-
availability of this compound (10%), which may explain its lack
of oral activity.
Compound 1F was prepared similarly from the appropriate
2-aminobenzoate derivative. It was synthesized by starting
from commercially available methyl 2-nitro-4-trifluoromethyl-5-
fluorobenzoate by nucleophilic aromatic substitution with
methanesulfonylacetic acid benzyl ester followed by reduction
of the nitro group with concomitant decarboxylation. Finally,
the 2-aminobenzoate derivative required for the preparation of
compound 1G was synthesized by starting from commercially
available methyl 2-nitro-4-trifluoromethyl-5-fluorobenzoate by
nucleophilic aromatic substitution with methylamine. Follow-
ing reduction of the nitro group, the bis-aniline was treated
with acetyl chloride to yield a mixture of mono- and bis-acety-
lated anilines, from which 5-(acetylmethylamino)-2-amino-4-tri-
The next challenge was thus to prepare quinazolinedione
derivatives which would show oral activity in animal models of
epilepsy. The availability of a number of in-house X-ray crystal
structures in conjunction with molecular modelling efforts
were important assets to assist the selection of new substitu-
ents which would potentially interact with the receptor, keep-
ing in mind that the ultimate goal was to influence the phar-
macokinetic properties of the molecules in order to gain
in vivo activity.
&
ChemMedChem 2016, 11, 1 – 6
www.chemmedchem.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
X-ray analysis of hGluA2 complexes with a number of close
analogues of 1E[24,25,40] allowed us to identify important
ligand–receptor interactions. The 3D structure highlights the
exquisite complementarity of the quinazolinedione scaffold for
the AMPA binding site. The sulfonamide nitrogen atom accepts
a hydrogen bond from NH2-Arg485 and is thus possibly nega-
tively charged in the complex. The quinazolinedione ring
system makes favorable p–p stacking interactions with Tyr450
and establishes a network of hydrogen bonds with CO-Pro478,
NH-Thr480, and NH1-Arg485. An interaction is observed be-
tween the 3-nitrogen atom of the imidazolyl substituent and
Thr686. The importance of this interaction correlates with im-
proved binding affinity. Moreover, the five-membered hetero-
cycle establishes favorable hydrophobic contacts with Met708.
With the aim to restore oral activity, novel molecules were
designed according to the following strategy. To preserve
strong interactions with Thr686, substituents at position 6 con-
taining good hydrogen bond acceptors (high pKb), in a topolo-
gy similar to that found in 1E, were introduced in the quinazo-
linedione scaffold. Preference was given to moieties bearing
non-basic hydrogen bond acceptors in order to avoid the for-
mation of zwitterionic species at physiological pH. In addition,
substitution patterns the could possibly increase lipophilicity
(increase cLogP) without having detrimental effects on solubili-
ty were selected, because a delicate balance between these
two parameters was thought to be required for good oral ab-
sorption and brain penetration. Special attention was given to
the identification of substitution patterns bringing the pKa of
the sulfonamide moiety as close as possible to physiological
pH in order to achieve a percentage, as high as possible, of
neutral species in solution at physiological pH. Finally, the
overall size of the analogues was kept small, with Mr <450 Da.
A set of compounds potentially fulfilling these criteria were
synthesized and allowed the discovery of a number of potent
and orally active compounds which are highlighted in Table 2.
As exemplified in the 6-trifluoromethyl series, introducing
a flexibly linked hydrogen bond acceptor at position 6 yielded
at most a threefold improvement in affinity for the sulfone an-
alogue 1G relative to the naked trifluoromethyl analogue 1C.
It is well accepted that hydrogen bonds are sensitive to the
local chemical environment[33,34] and may offer limited binding
energy gains when located in polar and hydrated surface re-
gions of a protein. Introducing tetrahydrofuran at position 6
(in 1H) yielded a compound equipotent to the sulfone ana-
logue and displaying moderate oral activity. Embedding the
hydrogen bond acceptor into a five- or six-membered hetero-
aromatic ring as in 1I–1K led again to a twofold improvement
in affinity. This result may reflect the favorable hydrophobic
contacts with Met708 and/or the increased hydrophobic
shielding of the aforementioned hydrogen bond. These com-
pounds, however, lack oral activity in the MES model. To fur-
ther increase the hydrophobic character near the hydrogen
bond acceptor, the size or number of alkyl substituents on the
heterocycles was increased as exemplified by 1L and 1M. This
led to fairly potent AMPA antagonists displaying good oral ac-
tivity: 1O. In parallel, to increase the pKa of the sulfonamide,
the 6-trifluoromethyl group was replaced with monofluoro-
Table 2. Binding affinity, in vivo activity of selected quinazolinediones 1.
Compd
R6
R7
IC50 [mm][a]
0.20Æ0.05
MES p.o.
[b]
ED50 [mgkgÀ1
]
1F
CF3
NE
1G
1H
1I
CF3
CF3
CF3
0.36Æ0.05
0.23Æ0.05
0.14Æ0.02
NT
20%[c]
NE
1J
CF3
CF3
0.12Æ0.02
0.12Æ0.02
NE
NE
1K
1L
CF3
CF3
0.96Æ0.1
NT
18
1M
0.086Æ0.03
1N
1O
CH2F
CHF2
0.57Æ0.1
NT
7.4
0.25Æ0.05
1P
CHF2
0.095Æ0.03
20
1Q
1R
iPr
2Æ0.3
NT
Cl
0.32Æ0.05
20%[c]
1S
1T
iPr
iPr
0.19Æ0.05
0.36Æ0.05
7
40%[c]
1U
iPr
0.31Æ0.05
40%[c]
[a] Rat AMPA [3H]CNQX binding. [b] Maximal electroshock seizure test in
mice (1 h). Values are the geometric mean of at least three independent
experiments. NT: not tested. NE: not effective at 50 mgkgÀ1 p.o. [c] Per-
cent inhibition at 50 mgkgÀ1 p.o.
ChemMedChem 2016, 11, 1 – 6
www.chemmedchem.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
methyl, difluoromethyl, chloro, and various alkyl groups as ex-
emplified in 1N–1S. These variations led to the most potent
orally active AMPA antagonists 1O and 1S of this series. In-
creasing the heterocycle substituent size further and/or incor-
porating various weak hydrogen bond acceptor functionalities
as exemplified by 1T or 1U did not further improve overall ac-
tivity.
Table 3. In vivo activity and properties of selected quinazolinediones 1.
[a]
Compd
MES p.o. ED50 [mgkgÀ1
]
logD pKa Solubility [gLÀ1
]
1 h
4 h
8 h
pH 1/pH 7.4
1M
1O
1S
18
7.4
7
28
NE
NE
25.5
2.6 6.2
1.4 6.4
1.3 6.7
0.1/1.6
0.5/5.5
0.016/0.11
20%[b]
11
selurampanel
An X-ray structure (Figure 2) was obtained for 1S co-crystal-
lized with a construct of the human receptor hGluA2 (Swis-
sProt P42262), similar to that reported by Koller et al.,[24] but
with the mutation Y702.[38] The central scaffold of 1S forms
a favorable p–p stacking interaction with Tyr450, as well as
a hydrogen bond network with residues Pro478, Thr480, and
Arg485. An additional hydrogen bond with Thr686 was also
identified. The methyl group on the pyrazole of 1S points
toward a water-filled cavity (containing for example, Glu705),
and the isopropyl group makes van der Waals interactions with
hydrophobic parts of E402, Y405, P478, and M708.
[a] Maximal electroshock seizure test in mice; values are the geometric
mean of at least three independent experiments; NE: not effective at
50 mgkgÀ1 p.o. [b] Percent inhibition at 50 mgkgÀ1 p.o.
tives, 1M and the 7-CHF2 derivative 1O suppressed general-
ized tonic–clonic seizures as induced by MES using 1 h pre-
treatment, their activity dropped considerably following 4 h
pre-treatment and disappeared following 8 h pre-treatment
(Table 3). The 7-iPr derivative 1S, on the other hand, potently
suppressed generalized tonic–clonic seizures, at peak oral ED50
values of 6.1 mgkgÀ1 and 8.4 mgkgÀ1 in mice and rats, respec-
tively. As shown in Table 3, this effect lasted for at least 8 h in
mice. Seizure suppression in these tests is indicative of anti-
convulsant efficacy against generalized as well as partial seiz-
ures.
Compounds 1m, 1O, and 1S potently and dose-dependent-
ly antagonized MES-induced generalized tonic–clonic seizures
in mice with ED50 values around 7 mgkgÀ1 after 1 h pre-treat-
ment. When pretreatment time was extended, clear differences
appeared between these compounds. While the 7-CF3 deriva-
The oral absolute bioavailability and disposition pharmacoki-
netics of 1S, as well as its in vivo brain penetration, corroborat-
ed these findings. When tested in mice (30 mgkgÀ1 p.o.), 1S
exhibited good pharmacokinetic properties with reasonable
AUC (208116 pmolmLÀ1 h) although with a low brain/plasma
ratio of <0.1. In a study with a 3 mgkgÀ1 i.v. dose, a mouse
plasma half-life of 3.3 h was determined, with a moderate
volume of distribution (Vdss =1.3 LkgÀ1) and a low clearance of
5.4 mLminÀ1 kg. Compound 1S has a good selectivity profile
with no apparent binding to a panel of over 60 receptors and
enzymes. The compound shows a low risk of hERG interaction
with an IC50 value of >30 mm in the hERG cell-based assay. It
also showed no substantial inhibition of all tested cytochromes
P450 up to 10 mm. It should finally be pointed out that not all
possibilities to interfere with the glutamatergic system seem
well suited for chronic treatment purposes in patients. As most
classes of NMDA antagonists have been associated with severe
psychotomimetic effects,[35] cardiovascular side effects,[36] im-
pairment of learning and memory,[37] and neurotoxicity,[38] se-
lectivity of 1S for the AMPA receptor versus the other iono-
tropic receptors was assessed (Table 4). Compound 1S inhibit-
ed the binding of radioligands to the glutamate
([3H]CGP39653) or glycine ([3H]MDL105519) sites of the NMDA
receptor only at much higher concentrations. Similarly,
Figure 2. X-ray crystal structure at 1.65 ꢁ resolution of the ligand binding
domain of an hGluA construct (carbon atoms in yellow, nitrogen atoms in
blue, oxygen atoms in red, and sulfur atoms in brown) bound to 1S (carbon
atoms in cyan). Selected water molecules and interactions (distances in ꢁ)
are shown in white. Details of this X-ray structure determination will be pub-
lished elsewhere.
Table 4. Binding of 1S and reference compounds to iGluRs.
Compd
AMPA C50 [mm]
NMDA glu IC50 [mm]
SR[b]
80
1.8
>540
NMDA gly IC50 [mm]
SR[b]
Kainate IC50 [mm]
SR[b]
DNQX
kynurenic acid
1S
0.5
101
0.186
40
184
>100
9.5
20
27
19
0.2
145
2.3
>1000
>100
4
>10
>540
[a] Radioligand binding experiments were carried out using rat whole-brain membranes; GraphPad Prism software was used for the analysis of radioligand
binding data. [b] Selectivity ratio: (IC50 receptor X)/(IC50 AMPA); values are the geometric mean of at least three independent experiments.
&
ChemMedChem 2016, 11, 1 – 6
www.chemmedchem.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
[3H]kainate binding to the kainic acid receptor was only mar-
ginally displaced by 1S.
[13] G. Seifert, K. Huttmann, J. Schramm, C. Steinhꢃuser, J. Neurosci. 2004,
24, 1996–2003.
[14] G. Seifert, W. Schroder, S. Hinterkeuser, T. Schumacher, J. Schramm, C.
Steinhꢃuser, Epilepsia 2002, 43, 162–167.
[15] W. Vollmar, J. Gloger, E. Berger, G. Kortenbruck, R. Kçhling, E.-J. Speck-
mann, U. Musshoff, Neurobiol. Dis. 2004, 15, 371–379.
[16] K. Zilles, M. S. Qꢂ, R. Kçhling, E. J. Speckmann, Neuroscience 1999, 94,
1051–1061.
[17] R. E. Twyman, L. C. Gahring, J. Spiess, S. W. Rogers, Neuron 1995, 14,
755–762.
In summary, the present work provides a new class of orally
active AMPA antagonists. We demonstrated that 6- and 7-posi-
tion-substituted quinazolinedione-sulfonamide analogues dis-
play potent AMPA antagonistic activity in vitro. N-[7-Isopropyl-
6-(2-methyl-2H-pyrazol-3-yl)-2,4-dioxo-1,4-dihydro-2H-quinazo-
lin-3-yl]methanesulfonamide (1S) described herein is a highly
selective and competitive AMPA receptor antagonist. Moreover,
1S has a good mouse pharmacokinetic profile, low human mi-
crosomal clearance, and has been shown to antagonize, when
given orally, MES-induced generalized tonic–clonic seizures in
[18] X. P. He, M. Patel, K. D. Whitney, S. Janumpalli, A. Tenner, J. O. McNa-
mara, Neuron 1998, 20, 153–163.
[19] L. C. Gahring, N. G. Carlson, S. W. Rogers, Autoimmunity 1998, 28, 243–
248.
[20] T. J. Wilding, J. E. Huettner, Mol. Pharmacol. 1996, 49, 540–546.
[21] A. D. Batcho, W. Leimgruber, Org. Synth. 1985, 63, 214.
[22] A. J. Culshaw, S. Bevan, M. Christiansen, P. Copp, A. Davis, C. Davis, A.
Dyson, E. K. Dziadulewicz, L. Edwards, H. Eggelte, A. Fox, C. Gentry, A.
Groarke, A. Hallett, T. W. Hart, G. A. Hughes, S. Knights, P. Kotsonis, W.
Lee, I. Lyothier, A. McBryde, P. McIntyre, G. Paloumbis, M. Panesar, S.
Patel, M. P. Seiler, M. Yaqoob, K. Zimmerman, J. Med. Chem. 2006, 49,
471–474.
[23] H. Allgeier, Y. Auberson, D. Carcache, P. Floersheim, C. Guibourdenche,
W. Froestl, J. Kallen, M. Koller, H. Mattes, J. Nozulak, D. Orain, J. Renaud
(Novartis AG), Int. PCT Pub. No. WO2006108591 A1, 2006.
[24] M. Koller, K. Lingenhoehl, M. Schmutz, I.-T. Vranesic, J. Kallen, Y. P. Auber-
son, D. A. Carcache, H. Mattes, S. Ofner, D. Orain, S. Urwyler, Bioorg.
Med. Chem. Lett. 2011, 21, 3358–3361.
mice for at least 8 h with an ED50 value of ~25 mgkgÀ1
.
The favorable pharmacological profile and adequate phar-
macokinetic properties led to the selection of 1S (selurampa-
nel) as a clinical candidate. In a proof-of-concept (POC) study,
selurampanel significantly decreased the sensitivity of patients
with photosensitive epilepsy to flashing lights.[39] Results of fur-
ther POC studies in patients with therapy-refractory partial
seizures, in patients with migraine pain, and in patients with
chronic subjective tinnitus have been or will be reported short-
ly.[40,41] Finally, a phase 2 multicenter, randomized, double-blind,
placebo-controlled, crossover study in patients with partial-
onset seizures was run, and results will be reported in due
course.
[25] D. Orain, S. Ofner, M. Koller, D. A. Carcache, W. Froestl, H. Allgeier, V. Ra-
setti, J. Nozulak, H. Mattes, P. Floersheim, J. Kallen, K. Lingenhoel, S. Ur-
wyler, S. Desrayaud, Bioorg. Med. Chem. Lett. 2012, 22, 996–999.
[26] T. Honorꢄ, J. Drejer, E. Ø. Nielsen, M. Nielsen, Biochem. Pharmacology
1989, 38, 3207–3212.
[27] A. G. Chapman, S. E. Smith, B. S. Meldrum, Epilepsy Res. 1991, 9, 92–96.
[28] P. Desos, J. M. Lepagnol, P. Morain, P. Lestage, A. A. Cordi, J. Med. Chem.
1996, 39, 197–206.
[29] M. Shimizu-Sasamata, S. Kawasaki-Yatsugi, M. Okada, S. Sakamoto, S.
Yatsugi, J. Togami, K. Hatanaka, J. Ohmori, K. Koshiya, S. Usuda, K.
Murase, J. Pharmacol. Exp. Ther. 1996, 276, 84–92.
[30] M. Schmutz, C. Portet, A. Jeker, K. Klebs, A. Vassout, H. Allgeier, R. Heck-
endorn, G. E. Fagg, H. R. Olpe, H. van Riezen, Naunyn-Schmiedeberg’s
Arch. Pharmacol. 1990, 342, 61–66.
Acknowledgements
We acknowledge Anke Blechschmidt and Rene Hemmig for their
excellent technical assistance. The crystallographic experiments
were performed on the X06SA beamline at the Swiss Light
Source, Paul Scherrer Institute, Villigen, Switzerland.
[31] A. W. Wamil, M. Schmutz, C. Portet, K. F. Feldmann, M. J. McLean, Eur. J.
Pharmacol. 1994, 271, 301–308.
[32] J. Ohmori, S. Sakamoto, H. Kubota, M. Shimizu-Sasamata, M. Okada, S.
Kawasaki, K. Hidaka, J. Togami, T. Furuya, K. Murase, J. Med. Chem. 1994,
37, 467–475.
Keywords: AMPA
quinazolinediones · selurampanel
· antagonists · epilepsy · glutamate ·
[1] P. Kwan, M. J. Brodie, N. Engl. J. Med. 2000, 342, 314–319.
[2] R. S. Fisher, B. G. Vickrey, P. Gibson, B. Hermannd, P. Penoviche, A. Scher-
erf, S. Walkerg, Epilepsy Res. 2000, 41, 39–51.
[33] W. C. Wimley, K. Hristova, A. S. Ladokhin, L. Silvestro, P. H. Axelsen, S. H.
White, J. Mol. Biol. 1998, 277, 1091–1110.
[3] R. S. Fisher, B. G. Vickrey, P. Gibson, B. Hermannd, P. Penoviche, A. Scher-
erf, S. Walkerg, Epilepsy Res. 2000, 41, 53–61.
[4] R. L. Macdonald in Epilepsy: A Comprehensive Textbook, 2nd ed. (Eds.: J.
Engel, T. A. Pedley), Lippincott-Raven Publishers, Philadelphia, New York,
1998, 1383–1391.
[34] A. S. Ladokhin, S. H. White, J. Mol. Biol. 1999, 285, 1363–1369.
[35] W. Koek, J. H. Woods, G. D. Winger, J. Pharmacol. Exp. Ther. 1988, 245,
969–974.
[36] K. W. Muir, D. G. Grosset, E. Gamzu, K. Lees, Br. J. Clin. Pharmacol. 1994,
38, 33–38.
[5] R. Dingledine, K. Borges, D. Bowie, S. F. Traynelis, Pharmacol. Rev. 1999,
51, 7–61.
[37] G. Collingridge, Nature 1987, 330, 604–605.
[38] J. W. Olney, Prog. Brain Res. 1990, 86, 37–51.
[6] H. F. Bradford, Prog. Neurobiol. 1995, 47, 477–511.
[7] K. Ohno, R. Tsutsumi, N. Matsumoto, H. Yamashita, Y. Amada, J.-I. Shishi-
kura, H. Inami, S.-I. Yatsugi, M. Okada, S. Sakamoto, T. Yamaguchi, J.
Pharmacol. Exp. Ther. 2003, 306, 66–72.
[8] T. Namba, K. Morimoto, K. Sato, N. Yamada, S. Kuroda, Brain Res. 1994,
638, 36–44.
[9] T. Hanada, J. Receptor, J. Recept. Ligand Channel Res. 2014, 7, 39–50.
[10] I. Blꢂmcke, H. Beck, B. Scheffler, P. R. Hof, J. H. Morrison, H. K. Wolf, J.
Schramm, C. E. Elger, O. D. Wiestler, Acta Neuropathol. 1996, 92, 576–
587.
[39] D. Kasteleijn-Nolst Trenitꢄ, B. Steinhoff, C. Brandt, T. Mayer, F. Rosenow,
G. Imbert, K. Kucher, Antiepileptic Drug and Device Trials XI Conference,
Aventura, FL (USA), April 27–29, 2011.
[40] B. Gomez-Mancilla, R. Brand, T. P. Jꢂrgens, H. Gçbel, C. Sommer, A.
Straube, S. Evers, M. Sommer, V. Campos, H. O. Kalkman, S. Hariry, N.
Pezous, D. Johns, H.-C. Diener, Cephalalgia 2014, 34, 103–113.
[41] E. Faught, Expert Opin. Invest. Drugs 2014, 23, 107–113.
[11] N. C. de Lanerolle, T. Eid, G. von Campe, I. Kovacs, D. D. Spencer, M.
Brines, Eur J Neurosci. 1998, 10, 1687–1703.
[12] T. Eid, I. Kovacs, D. D. Spencer, N. C. de Lanerolle, Eur. J. Neurosci. 2002,
15, 517–527.
Received: September 14, 2016
Revised: October 24, 2016
Published online on && &&, 0000
ChemMedChem 2016, 11, 1 – 6
www.chemmedchem.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

COMMUNICATIONS
D. Orain, E. Tasdelen, S. Haessig, M. Koller,
A. Picard, C. Dubois, K. Lingenhoehl,
S. Desrayaud, P. Floersheim, D. Carcache,
S. Urwyler, J. Kallen, H. Mattes*
Strong anticonvulsant: We report the
design and synthesis of the AMPA re-
ceptor antagonist selurampanel. The X-
ray crystal structure of selurampanel
bound to the AMPA receptor hGluA,
which is also described, highlights the
interactions of the compound with
hGluA. This compound has shown ex-
cellent oral potency against MES-in-
duced generalized tonic–clonic seizures
in rodents as well as significant activity
in patients suffering from various forms
of epilepsy.
&& – &&
Design and Synthesis of
Selurampanel, a Novel Orally Active
and Competitive AMPA Receptor
Antagonist
&
ChemMedChem 2016, 11, 1 – 6
www.chemmedchem.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!