Discovery of N -[(2 s)-5-(6-fluoro-3-pyridinyl)-2,3-dihydro-1 H -inden-2-yl]-2-propanesulfonamide, a novel clinical AMPA receptor positive modulator

Article abstract of DOI:10.1021/jm1005429

A series of AMPA receptor positive allosteric modulators has been optimized from poorly penetrant leads to identify molecules with excellent preclinical pharmacokinetics and CNS penetration. These discoveries led to 17i, a potent, efficacious CNS penetrant molecule with an excellent pharmacokinetic profile across preclinical species, which is well tolerated and is also orally bioavailable in humans.

Full text of DOI:10.1021/jm1005429

J. Med. Chem. 2010, 53, 5801–5812 5801
DOI: 10.1021/jm1005429
Discovery of N-[(2S)-5-(6-Fluoro-3-pyridinyl)-2,3-dihydro-1H-inden-2-yl]-2-propanesulfonamide, a
Novel Clinical AMPA Receptor Positive Modulator†
Simon E. Ward,*,‡ Mark Harries,‡ Laura Aldegheri,§ Daniele Andreotti,§ Stuart Ballantine, Benjamin D. Bax,
Andrew J. Harris,‡ Andy J. Harker,‡ Jesper Lund,§ Rosemary Melarange,‡ Anna Mingardi,§ Claudette Mookherjee,‡
Julie Mosley, Marta Neve,§ Beatrice Oliosi,§ Roberto Profeta,§ Kathrine J. Smith, Paul W. Smith,‡ Simone Spada,§
Kevin M. Thewlis,‡ and Shahnaz P. Yusaf
‡Neurosciences Centre of Excellence for Drug Discovery, GlaxoSmithKline, New Frontiers Science Park, Third Avenue, Harlow,
Essex CM19 5AW, United Kingdom, §Centro Ricerche, GlaxoSmithKline, Via A. Fleming 4, 37100 Verona, Italy, and
Molecular Discovery Research, GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, United Kingdom
Received May 5, 2010
A series of AMPA receptor positive allosteric modulators has been optimized from poorly penetrant
leads to identify molecules with excellent preclinical pharmacokinetics and CNS penetration. These
discoveries led to 17i, a potent, efficacious CNS penetrant molecule with an excellent pharmacokinetic
profile across preclinical species, which is well tolerated and is also orally bioavailable in humans.
Introduction
loop (M2) which constitutes part of the ion channel pore,11
and an intracellular domain comprising motifs for binding
and phosphorylation of transmembrane AMPA regulatory
proteins (TARPs) which play a pivotal role in regulating the
trafficking of AMPARs during synaptic plasticity.12 These
AMPAR auxiliary subunits are also fundamental to synaptic
transmission within the CNS modulating open channel prob-
ability and channel kinetics,13 single channel conductance,14
and sensitivity to polyamine block.15
The understanding of the biophysical and structural char-
acteristics of the AMPAR has advanced considerably over the
past decade, in particular through investigation of the AM-
PAR ligand-binding core, which is formed by two polypeptide
segments S1, located at the M1 region, and S2, located bet-
ween M3 and M4 regions. Studies have shown that the S1S2
domain possesses pharmacological properties that are similar
to those of native membrane bound receptors.16 The crystal
structure of the rat GluA2 S1S2 ligand binding domain has
allowed development of agonists and antagonists which have
benefitedfrom high resolutionstructuralinformation.17 More
recent studies have identified binding sites for positive allos-
teric modulators of AMPAR and investigated mechanisms by
which such molecules exert their pharmacological effects
acting to slow both agonist release and rate of receptor de-
activation.18 Most recently, this structural investigation has
been taken to even greater ends with the successful crystal-
lizationand structuralelucidationofthe ratGluA2receptorat
3.6 A in complex with a competitive antagonist.19
A growing body of evidence indicates that dysfunction of
glutamatergic neurotransmission underlies the pathophysiol-
ogy of many neurological diseases including schizophrenia,
Alzheimer’s disease, Parkinson’s disease, and mood disorders.1,2
Preclinical studies have identified modulators of the R-amino-3-
hydroxyl-5-methyl-4-isoxazole-propionic acid receptor (AMP-
ARa), which slow the rate of receptor deactivation and/or
desensitization, leading to enhanced synaptic activity and effi-
cacy in cognition models.3-5
AMPAR belong to the ligand-gated, ionotropic glutama-
tergic receptor family and are widely distributed in the mam-
malian central nervous system and mediate the vast majority
of fast excitatory neurotransmission in the CNS.6,7 The re-
ceptors are tetrameric assemblies of subunits GluA1-4 (orig-
inally named GluR1-4 or GluRA-D), in which two subunits
join to form a dimer which then combine to form a tetrameric
complex.8,9 Each subunit comprises a large extracellular
domain incorporating the glutamate binding residues in the
ligand-binding domain (LBD),10 three hydrophilic regions
(M1, M3 and M4) which span the membrane and a re-entrant
†The atomic coordinates for the crystal structure of 17i described here
have been deposited at the Protein Data Bank under entry code 2xhd.
*To whom correspondence should be addressed. Phone: þ44 (0) 1279
622894. Fax: þ44 1279 622 790. E-mail: simon.e.ward@gsk.com.
a Abbreviations: AMPA, R-amino-3-hydroxyl-5-methyl-4-isoxazole-
propionate; AMPAR, AMPA receptor; AUC, area under the curve;
BTB, brain tissue binding; CNS, central nervous system; DBU, 1,8-
diazabicyclo[5.4.0]undec-7-ene; DTT, dithiothreitol; IV, intravenous;
FLIPR, fluorescent imaging plate reader; GPCR, G-protein coupled
receptor; hERG, human ether-a-go-go related gene; IPTG, isopropyl-β-
D-thiogalactopyranoside; LBD, ligand-binding domain; LTP, long-term
potentiation, MDCK, Madin-Darby canine; MDR, multidrug-resis-
tant; NADPH, nicotinamide adenosine dinucleotide phosphate;
NMDA, N-methyl-D-aspartic acid; NOAEL, no observed adverse event
level; NOR, novel object recognition; P-gp, P-glycoprotein; PPB, plasma
protein binding; PSA, polar surface area; PXR, pregnane X receptor;
SD, standard deviation; TARP, transmembrane AMPA regulatory
protein.
Despite these advances in basic understanding, designing
molecules to potentiate ion flux through the AMPAR remains
challenging, not least due to the layers of complexity created
by the existence of a number of splice variants20 and sites for
post-translational modification which influence receptor
kinetics21 and binding.22 All in all, this serves to render the
specific composition of ion channels in vivo extremely
complicated and one of the key challenges faced by a drug
discovery program is to have confidence that data generated
r
2010 American Chemical Society
Published on Web 07/08/2010
pubs.acs.org/jmc

5802 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
Figure 1. Clinically evaluated AMPAR positive modulators.
Scheme 1. Synthesis of Phenyl Indane Analoguesa
i
a Reagent and reaction conditions: (a) DBU, CH2Cl2, PrSO2Cl; (b) periodic acid, I2, AcOH, H2SO4; (c) 3-R1-phenyl boronic acid, Cs2CO3,
i
Pd(OAc)2, PPh3, dioxane; (d) 10% Pd/C, EtOH, H2; (e) for 13a PrCOCl, for 13b EtSO2Cl, for 13c 3-chloropropanesulfonyl chloride Et3N, DMF;
(f) LiAlH4, THF, 0 °C; (g) 4-chlorobutanoyl chloride, Et3N, CH2Cl2.
against recombinant systems will translate reliably through to
the native system.
AMPAR positive modulators, in particular combining high
potency with a pharmacokinetic profile which would deliver
high, unbound compound concentrations.25
Many molecules have been described as AMPAR positive
modulators across a range of chemotypes23,24 which, despite
differences in their proposed mechanism, promote long-term
potentiation (LTP)23,24 by enhancing glutamatergic synaptic
activity. LTP is thought to be the underlying mechanism for
cognition23,24 and as such, preclinical cognition models have
been the primary focus for AMPAR modulators although the
aim is to progress molecules in the clinic to treat a range of
diseases such as schizophrenia, depression, and Parkinson’s
disease.23,24 In brief, three major, well-established classes that
have been investigated clinically exist alongside a number
of less extensively exemplified structures (Figure 1). The
first main group derives from the original nootropic agent 1
(Aniracetam) and were subsequently developed into a range
of benzamide derivatives by Cortex including 2 (CX-516)
and 3 (CX-691). The second major class, benzothiadiazines,
exemplified by the diuretic agent 4 (cyclothiazide) was
developed into 5 (S-18986) by Servier. The third broad class
comprises the phenethylsulfonamides discovered by Lilly,
which include clinically investigated 6 (LY450108) and 7
(LY451395). In broad terms the molecules have generally
displayed higher affinities for the AMPAR over time, with
the Lilly phenethylsulfonamides being considerably more
potent than the original benzamide class.
Results
Preparation of Phenyl Indane Analogues. Our initial explo-
ratory activities identified 11a as the lead for this program.
The indane analogues were prepared from amino indane 8 as
described in Scheme 1 via sequential functionalization to
first form the sulfonamide 9 and then the regioselective iodi-
nated derivative 10. This reaction was unsuccessful on the
parent amino indane due to a lack of regiochemical control.
Key intermediate 10 was then reacted under standard Suzuki
coupling conditions to afford a range of analogues 11a-h.
Certain of these were reacted further to broaden the scope of
the chemotypes investigated. In particular, nitro derivative
11g was hydrogenated to afford aniline derivative 12, which
could be acylated or sulfonylated to afford a wider range of
amides and sulfonamides 13a-c. Nitrile derivative 11c was
also used as the starting point for analogues, with a methy-
lene spacer between the phenyl ring and nitrogen atom by
subjecting 11e to sequential reduction with lithium alumi-
num hydride to amine 14 and acylation to lactam 15.
Pharmacological Characterization of Analogues. The abil-
ity of the test compounds to potentiate a glutamate receptor-
mediated response was determined primarily by using fluor-
escent calcium-indicator dyes and then additionally for some
From evaluation of these chemotypes, we sought to initi-
ate a program to create our own series of CNS-penetrant

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5803
Table 1. Biological Activity and Physiochemical Data for 3-Substituted Phenyl Indane Analoguesa
a FLIPR generated pEC50 against hGluA2 flip isoform. All values are (0.2 and n = 3 except for 11h (n = 2). Asym max is the fitted maximum
response, relative to 100% defined as the maximal response of cyclothiazide standard. b clogP Daylight Chemical Information Systems Inc., Aliso Viejo,
CA, http://www.daylight.com. c Rat plasma protein binding (rat PPB) values were determined using a 96-well plate equilibrium dialysis method at a
concentration of 1 μg/mL.
Table 2. Rat Pharmacokinetic (PK) Profiles of Key 3-Phenyl Substi-
tuted Indanes
example compounds by measuring glutamate-evoked cur-
rents recorded from human GluA2 flip (hGluA2i), Q/R
AUC0-t
ng h/mL
Brain:Blood
AUC0-t ratio
PAPP(þinh)
efflux
ratio
unedited HEK293 cells. For the primary screen against
hGluA2i homomeric AMPARs, a concentration-response
curve was derived for potentiation of glutamate-induced rises
in intracellular calcium. Further data also confirmed, for
those molecules profiled, a broad spectrum positive modula-
tory activity across AMPARs formed from different subunits
(GluA1i, GluA3i, or GluA4i homomers) and also equivalent
potency at rat and human GluA2i homomeric AMPARs.
As detailed in Table 1, lead 11a displayed potent potentia-
tion of the AMPAR-mediated response, with encouraging
predicted physicochemical properties: low clogP and mod-
erate molecular weight, contributing to acceptable rat
plasma protein binding of 95.9%. Furthermore, this mole-
cule had a clean selectivity profile against key targets and did
not inhibit P450 enzymes at low concentrations (data not
shown). 11a was also progressed to a rat oral pharmacoki-
netic study and, very encouragingly, gave excellent systemic
exposure and time-course profile. However, as detailed in
Table 2 and most likely as a consequence of the high polar
surface area, this molecule suffered restricted access to the
brain due to interaction with P-glycoprotein (P-gp).26 This
was supported by experimental data showing a moderate to
high efflux ratio in the MDCK-MDR1 in vitro P-gp assay.
It was clear from our preliminary investigation that mod-
ification of the secondary sulfonamide group attached to the
indane was not tolerated; the presence of both the secondary
nitrogen and sulfonyl group were crucial for activity. As
such, analogue investigation focused on the substituent on
the phenyl ring, which was restricted to the meta position, as
compd
nm/s a
3
11a
11h
15
1130
161
36
0.1
0.4
0.4
608
698
5.8
3.2
a PAPP (þinh): apical (A) to basolateral (B) transport rates of
molecules (0.5 μM) across Madin-Darby canine multidrug-resistant
(MDCKII-MDR1) cells heterogeneously expressing human P-gp mea-
sured in presence of a potent P-gp inhibitor, GF120918, as an indicator
of passive membrane permeability. Efflux ratio: the ratio of transport
from A-B/B-A in the absence of P-gp inhibitor (inh).
their regioisomeric analogues demonstrated lower potencies.
Removal of this substituent led to a loss of activity, which could
not be restored by installing other polar, hydrogen-bond accept-
ing groups such as reversed sulfonamide 11b, nitrile 11c, or
ketone 11d. Recognizing that the hydrogen bond acceptor may
need to be separated from the phenyl group by a spacer atom,
we prepared the homologated ketone 11e and the methylene-
linked lactam 15, both of which were only moderately active
compounds. Similarly, replacement of the sulfonamide group in
11a with an amide 11f led to a loss of activity, which was ex-
acerbated by increasing the size of the alkyl group to isopropyl
13a. The only substituents which maintained good potency were
the sulfonamides, for which larger alkyls 13b and tertiary sul-
fonamides 11h and 13c gave high potency, demonstrating the
tolerance for increased size and loss of the H-bond donor
capability. Furthermore, this meant that all highly active com-
pounds, regardless of clogP, had a high polar surface area,
although only ethyl sulfonamide 13b was as high as lead 11a.

5804 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
With this in mind, we profiled further exemplars from this set
in rat pharmacokinetic studies. Unfortunately, as outlined in
Table 2, although brain-blood AUC ratios for the tertiary
amide 15 and the tertiary sulfonamide analogue 11h were
improved, and acceptable protein binding maintained, systemic
exposures were poor.
Preparation and Characterization of Pyridyl Indane Ana-
logues. Given the challenges we had encountered in balan-
cing potency with CNS penetration, and despite the lack of
success in truncating the sulfonamide substituent in lead 11a,
we decided to explore the indane substituent with a range
of alternative aromatic rings which would all be of signifi-
cantly lower polar surface area. By doing so, we hoped to
decrease interaction with P-gp, while maintaining good sys-
temic free concentrations thereby maximizing the free drug
concentrations in the brain, even if potentially at the expense of
high potency. A number of systems were investigated, and by
far the most effective were the pyridines, in particular, the
3-pyridyl and 2-pyridyl indane analogues. The initial com-
pound sets were prepared as described in Scheme 2. Use of
earlier key intermediate 10 allowed for rapid preparation of
these derivatives, with the route chosen according to the avai-
lability of the pyridyl boronic acid coupling partners. Inter-
mediate 10 was therefore either reacted directly under Suzuki
cross-coupling conditions with substituted pyridyl boronic
acids to afford the final pyridyl derivatives 17a-g and 18a-
d, or via a two-step process first involving preparation of the
indane boronate ester species 16 with subsequent cross-cou-
pling to afford the target molecules. While these are short
synthetic sequences, the precise conditions of the Suzuki coup-
ling required considerably more optimization and were gen-
erally lower yielding than for phenyl analogues above.
As displayed in Table 3, the initial analogue in this set,
unsubstituted 3-pyridyl indane 17a, gave lower activity in the
primary screen, in line with previously established SAR,
coupled with inhibition of P450 1A2 isoform and moderate
intrinsic clearance in rat microsomes. However, this analo-
gue had significantly improved physicochemical properties:
lower polar surface area, molecular weight, and calculated
lipophilicity. As such, we wanted to further investigate the
potential of these pyridyl derivatives, but in order to main-
tain a low polar surface area to maximize our chances of
improving CNS penetration, we were severely restricted in
the range of substituents available for use. This is highlighted
by inclusion of the cyano pyridyl derivative 17e in the data
set, for which the PSA is restored to an unacceptably high
value of 83 A2.
Scheme 2. Preparation of Pyridyl Indane Analoguesa
a Reagent and reaction conditions: (a) (1,10-bis(diphenylphosphino)-
ferrocene)palladium(II) chloride, CH2Cl2, KOAc, bis(pinacolato)-
diboron, DMSO; (b) Cs2CO3, R-substituted bromopyridine, Pd(OAc)2,
PPh3, dioxane; (c)Cs2CO3, R-substituted pyridyl boronicacid, Pd(OAc)2,
PPh3, dioxane.
Consequently, the range of substituents explored centered
on alkyl and halo derivatives, which gave a range of properties
Table 3. Biological and in Vitro DMPK Profiling of Novel Pyridyl Indane Analoguesa
P450 (μM) c
CLi d (mL/min/g)
compd
R
f
pEC50 Asym max % PSA (A2) MWt clogPg pKa
1A2
2C9 2C19 2D6 3A4
rat
human
b
3-Pyridyls
17a
17b
17c
17d
17e
17f
17g
17h
17i
H
rac
rac
rac
4.8
4.8
101
96
59
59
59
59
83
59
59
59
59
59
316
334
334
344
341
330
330
334
334
330
2.3
2.6
2.6
3.0
2.0
2.5
2.8
2.6
2.6
2.8
5.1
e
0.6 >10 >10
>10 >10
9.0 >10 >10
2.1
<0.5
2.1
1.7
2-F
6-F
>10
>10
>10
2.5
1.1
5.3
4.5
106
59
e
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
2,6-diMe rac
6.7 >10
>10
4.9
0.9
0.7
5-CN
4-Me
6-Me
6-F
rac
rac
rac
R
4.2
4.7
84
64
e
<0.5
4.5
2.0
5.8 >10
5.7 >10
e
>10 >9
>10 >10
1.4
0.7
5.5
107
<4.5
5.6
5.7
6-F
6-Me
S
107
104
e >10
5.8 >10
>10 >10
>10 >10
>10 >10
>10 >10
0.7
3.5
<0.7
<0.6
17j
S
2-Pyridyls
2.8
3.0
18a
18b
18c
18d
18e
5-F
rac
rac
rac
rac
S
5.2
5.5
4.7
5.1
5.7
114
90
59
59
59
59
59
334
330
330
351
334
2.7 >10
4.9 >10
4.9 >10
>10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
>10 >10
0.6
7.3
0.5
0.8
5-Me
3-Me
5-Cl
5-F
89
89
2.8
3.4
14
<0.8
1.6
1.3
<0.5
<0.5
103
2.8
2.7 >10
a FLIPR generated pEC50 against hGluA2 flip isoform. All values are (0.2 and n g 3. Asym max is the fitted maximum response, relative to 100%
defined as the maximal response of cyclothiazide standard. b pKa measured spectroscopically. c Inhibition of P450 isoforms. d Rate of turnover against
rat and human liver microsomes. e No pKa could be determined. f stereochemistry at chiral center, rac = racemic mixture. g clogP Daylight Chemical
Information Systems Inc., Aliso Viejo, CA, http://www.daylight.com.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5805
Scheme 3. Preparation of Single Enantiomer Pyridyl Indane Analoguesa
a Reagent and reaction conditions: (a) (i) HCl, water, (ii) Br2, (iii) HBr; (b) 4-methylmorpholine (R)-CSA or (S)-CSA, MeOH; (c) (i) 1 M NaOH,
CH2Cl2, (ii) iPrSO2Cl, DBU, CH2Cl2; (d) pyridyl boronic acid or substituted bromo-pyridine, Pd(Ph3)4, H2O/1,4-dioxan, Cs2CO3; (e) Bis(pinacolato)
diboron, KOAc, PdCl2(dppf), DMSO.
in terms of pyridine conjugate acid pKa, EC50, and stability
to microsomal turnover. For the 3-pyridyl indanes, a sub-
stituent para to the indane group on the pyridine ring gave an
increase in potency as exemplified by the fluorine in 17c and
the methyl in 17g. The P450 inhibition profile across the set
17b-g appeared to show an improvement over initial un-
substituted derivative 17a, even for analogue 17f, with no
substituent on the carbon adjacent to the pyridine nitrogen.
The intrinsic clearance for this set was more variable: fluor-
ine on the carbon adjacent to the pyridine nitrogen increased
microsomal turnover in human. The methyl substituted
pyridines 17d, 17f, and 17g showed higher microsomal turn-
over in rat. As expected, all methyl substituted pyridines are
significantly more basic than their halogenated counterparts,
for which no pKa of the conjugate acid could be determined.
However, we were unable to determine whether the increased
rat microsomal turnover was due solely to the increased
basicity of the pyridine or instead was simply a consequence
of the methyl group being a site of potential metabolism.
A similar pattern was replicated for the 2-pyridyl indane
analogues 18a-d. The best AMPAR potentiation potencies
were observed for substituents para to the indane ring for
fluoro derivative 18a and methyl derivative 18b, and all com-
pounds demonstrated an excellent profile against the P450
isoform panel. Similarly to the previous group, more basic
pyridines 18b and 18c were turned over to a far greater extent
by rat microsomes than halo analogues 18a and 18d.
Before committing to further activities in this area, we ran
a series of screening CNS penetration studies in the rat,
which validated our hypothesis of lowering PSA to drive im-
proved PK profiles. Key exemplars 17b and 17g were dosed
to rat and found to have brain-blood ratios of 1.5:1 and
1.4:1 respectively. This data set gave us the confidence to
continue exploration in this area and required us to optimize
our synthetic sequences to allow exploration in the single
enantiomer series.
Preparation and Characterization of Single Enantiomeric
Indanes. Because of the difficulties in separating the individual
enantiomers of the molecules above and the desire to prepare
sufficient quantities of compound for in vivo biological testing,
we sought to identify an improved synthetic route. Scheme 3
outlines the revised synthetic procedure, which drew on a pub-
lished classical resolution as the most pragmatic and efficient
way to access single enantiomer materials.27 Direct bromina-
tion of amino indane 8 generated the hydrobromide salt of
5-bromoindane derivative 19 in excellent yield. This material
could then be efficiently resolved into its constituent enantio-
mers 20 and 22 using the required enantiomer of camphor
sulfonic acid. The seemingly straightforward sulfonylation
required considerable optimization to allow isolation of 21
or 23 in high yield, which was then reacted, as before, under
Suzuki cross-coupling conditions to give target compounds 17i
or an enantiomeric analogue such as 17h. Transformation of
intermediate 21 into the corresponding boronic ester 24 al-
lowed preparation of target compounds 17j and 18e after the
Suzuki cross coupling.
As shown in Table 3, racemic 17c comprised R enantiomer
17h and S enantiomer 17i, the latter being considerably more
potent, and as such, exploration in the single enantiomer
series focused exclusively on the S stereochemistry series.
The three enantiomers characterized in detail, 17i, 17j, and
18e, showed good potentiation of AMPA currents, a lack of
P450 isoform inhibition, and intrinsic clearance in line with
the racemic analogues: low turnover for fluoro analogues 17i
and 18e in rat and human microsomes with moderate turn-
over in the rat system for methyl pyridine 17j. All three of
these compounds showed excellent wider selectivity against a
wide range of other ion channels (including hERG, NMDA,
and kainate channels), enzymes, and GPCRs. Furthermore,
as shown in Figure 2, analogues showed clear potentiation of
AMPAR-mediated currents in whole-cell patch-clamp elec-
trophysiology evaluation of the recombinant hGluA2i cell
line.
Molecules 17i and 18e were also investigated in behavioral
models of cognition and found to improve performance in
rats. Specifically, 17i and 18e (administered by oral gavage as
suspension in methyl cellulose solution) were tested for their
ability to improve a 24 h delay-induced deficit in novel object
recognition (NOR),28 a test of recognition memory in male
Lister Hooded rats (n = 9-12/group). Following single oral

5806 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
Figure 2. Electrophysiological activity of lead analogues. (A) Representative whole-cell current traces recorded from HEK293 cell expressing
hGluA2i homomeric AMPARs. 17i (1-100 μM, upper concentration limited by solubility) produced a concentration-dependent and reversible
increase in charge transfer. Positive modulator, cyclothiazide (CTZ) 30 μM, also increased charge transfer. (B) Concentration-response curve
for the potentiation of hGluA2i-mediated inward currents by 17i. Each data point is the mean response from four different cells and is
normalized to the effect of 30 μM cyclothiazide. pEC50 = 5.19. (C,D) Same data plots for 18e.
administration, both molecules improved performance in
this task with a minimally effective dose (MED) of 0.3 mg/kg
(equivalent to concentrations of 81 ng/mL blood and 122 ng/g
brain for 17i, and 37 ng/mL blood and 70 ng/g brain for 18e).
In the passive avoidance task,29 scopolamine administra-
tion (0.8 mg/kg, intraperitoneal) at 6 h post-training ren-
Detailed Evaluation of Preclinical Profile of 17i. In contrast
to the more basic pyridine analogues, the equilibrium solu-
bility of 17i was uniformly low (0.1 mg/mL or less) across the
pH range 2-10 and in the physiologically relevant fluids
tested (water, simulated gastric fluid, simulated intestinal
fluid [fed or fasted]). Despite this, the bioavailability in pre-
clinical species was good, which may be explained by the very
high permeability of the compound. Furthermore, 17i ex-
hibited excellent stability both as a powder and in solution
under all conditions tested, displaying no evidence for labi-
lity of the R-fluorine on the pyridine ring. The pharmacoki-
netics and oral bioavailability of 17i were investigated in rat,
dog, and monkey following single intravenous (as an infu-
sion over 1 h in saline containing 2% (v/v) N-methylpyrrolid-
one and 10% (w/v) encapsin) and oral (suspended in 1% (w/v)
methyl cellulose in water) administration. As detailed in Table 5,
good bioavailability and low blood clearance were seen across
all species.
In addition to the favorable DMPK (or developability)
properties outlined above and in Tables, 3, 4, and 5, no time-
dependent inhibition toward CYP3A4 and no formation of
glutathione conjugates was observed for 17i. This, together
with weak rat and human PXR activation, gave an overall
low potential for drug-drug interactions through inhibition
or induction of CYP450. Additionally, the in vitro plasma
protein binding of 17i was determined in five species by
equilibrium dialysis at a plasma concentration of 1 μg/mL.
Table 6 demonstrates that plasma protein binding was
moderate and similar across all species.
dered experimentally naıve male Wistar rats (n = 6-12/
¨
group) amnesic when tested 24 h later. 17i and 18e
(administered as for NOR) 3 h prior to training dose-depen-
dently attenuated the scopolamine-induced amnesic deficit
with significant effects at 3 and 10 mg/kg for 17i and 1 and
3 mg/kg for 18e (no concentration data available). The com-
pounds administered in the absence of scopolamine had no
effect on recall of the passive avoidance response nor any
effect on locomotor activity as assessed in an open-field arena.
Comparison of the DMPK properties for 17i, 17j, and 18e,
detailed in Table 4, confirms all three molecules had good
permeability, were not P-gp substrates, showed good CNS
penetration together with good free fraction, and satisfactory
rat pharmacokinetics. Blood clearance was low for 17i and low
to moderate for 18e. The rank order of in vivo clearance for the
three molecules in rat was consistent with that found in vitro,
with 17j showing the greatest clearance. Volume of distribution
was moderate and appeared to be similar for all 3 molecules;
however as a result of the moderate/high clearance, 17j gave the
shortest half-life (<1 h) and the lowest oral bioavailability. In
addition, in vitro microsomal clearance across species appeared
to be higher for 17j, suggesting lower oral exposure across
species and potentially in humans.
From the efficacy and DMPK data, both 17i and 18e
were progressed to safety and tolerability studies and 17i
was selected as the molecule to advance as a clinical drug
candidate.
The metabolism of 17i was investigated in vitro in liver
microsomes from rat, dog, monkey, and human. Five metabo-
lites were characterized at low abundance relative to unchanged
17i, with four being the result of aliphatic hydroxylation.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5807
Table 4. Preclinical Species DMPK Profiling of Pyridyl Indane Analogues a
17i
17j
18e
blood clearance (mL/min/kg)
T1/2 (h)
Vss (L/kg)
5.4
5.0
55
24
0.6
2.4
116
1
1.0
2.2
436
0.5
72
2.4
416
oral Cmax (ng/mL)
oral Tmax (h)
F(%)
3
61
25
1.3
brain-blood AUC0-t ratio
microsomal CLi (mL/min/g liver) m/r/d/ma/mo/h
r/h PPB (%)
2.1
0.5/0.7/<0.6/0.6/1.1/<0.7
1.4
<0.5/1.6/<0.5/0.5/1.5/<0.5
0.6/3.5/0.7/1.1/2.3/<0.6
86.7/90.8
97.4
688
92.2/-
96.7
633
90.3/-
97.9
557
r BTB (%)
PAPP(þinh) (nm/s)b
P-gp efflux ratio
1.1
0.9
1.1
a PK and CNS penetration data generated in rat, from 1 mg/kg IV or 3 mg/kg po serial and composite profiles. Species abbreviations used: m, r, d, ma,
mo, h, refer to mouse, rat, dog, marmoset, monkey, and human, respectively. Plasma protein binding (PPB) and brain tissue binding (BTB) values were
determined using a 96-well plate equilibrium dialysis method at a concentration of 1 μg/mL. b Permeability assay defined for Table 2.
Table 5. Pharmacokinetic Parameters for 17i Obtained in Male Rat, Dog and Monkey Following Intravenous and Oral Administration a
parameter
Sprague-Dawley rat
beagle dog
cynomolgus monkey
doses
IV 1 mg/kg
po 3 mg/kg
5.4 [4.6-6.3]
6
IV 0.125 mg/kg
po 0.5 mg/kg
3.4 [3.0-4.0]
11
IV 0.125 mg/kg
po 0.5 mg/kg
1.4 [1.3-1.5]
3
blood clearance (mL/min/kg)
% liver blood flowb (%)
Vss (L/kg)
2.4 [1.8-2.7]
5.0 [4.5-5.5]
61 [49-70]
416 [328-496]
3 [2-8]
4.8 [4.1-6.1]
16 [13-20]
61 [52-70]
116 [85-132]
2.0 [1.3-3.0]
17 [15-21]
1.8 [1.7-1.9]
15 [14-16]
51 [51-52]
115 [101-132]
6 [5-8]
IV half-life (t1/2) (h)
oral bioavailability (%)
oral Cmax (ng/mL)
oral Tmax (h)
oral half-life (t1/2) (h)
6.6 [6.3-7.0]
24 [22-25]
a All parameters were calculated from blood concentration-time data. All data are reported as mean [and range], for Tmax the median [and range] are
given for n = 3 measurements. b Calculated using the following liver blood flows (mL/min/kg): rat 85, dog 31, cynomolgus monkey 44.
Table 6. In Vitro Plasma Protein Binding of 17i in Preclinical Species
and Human at a Plasma Concentration of 1 μg/mLa
mouse
rat
dog
monkey
human
85.0 ( 1.1
86.7 ( 1.4
79.5 ( 1.0
84.8 ( 0.7
90.8 ( 0.5
a Results expressed as mean ( SD of 12 determinations.
Metabolites M1, M2, and M4 (see Figure 3) were produced
by oxidation of the aliphatic part of the indane ring system
and probably by hydroxylation of one or both of the benzylic
carbons which would result in the formation of diastereoi-
somers. It was not possible to determine whether M1, M2,
and M4 were hydroxylated on the three carbons available or
whether two were diastereoisomeric (other diastereoisomers
may have been present but chromatographically unresolved
from M1, M2, or M4).
In terms of relative metabolite peak areas, formation of M1
was an important metabolic route for each species. Formation of
M4 was also important for rat and dog but minor in human and
monkey. Formation of M2 and M3 was a relatively minor route
for each species. Aryl hydroxylation (M5) was a major meta-
bolic route in the rat but was not significant in dog, monkey, or
human. All the metabolites present in human liver microsomes
were also observed in rat, dog, and monkey microsomes.
Development candidate 17i exhibited no evidence of geno-
toxic potential in the in vitro predevelopment genetic toxicity
screens. There were no significant cardiovascular risks identi-
fied from a study in cynomolgus monkeys with single oral
doses up to 2.5 mg/kg, or in a hERG assay for potential effects
on QT interval. Furthermore, evaluation in preclinical safety
screening up to 6 weeks in duration established no observed
adverse effect levels (NOAEL) with acceptable safety margins
to progress into clinical evaluation.
Figure 3. Summary of major metabolic pathways of 17i in liver
microsomes from rat, dog, human, and monkey.
The first time in human study conducted in healthy volun-
teers explored the pharmacokinetics of single doses of 17i in
the 0.25-6 mg dose range (N = 64) and of a multiple dose of
0.1 mg (N=15) administered daily for 28 d. According to
both single and repeat dose administrations, 17i appeared to
be rapidly absorbed, with Cmax being attained between 0.5
and 3 h postdose. The apparent half-life was high on average
(107-168 h) with estimated individual values up to ∼300 h
(average t1/2 was consistent in the 0.1-6 mg dose range).

5808 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
cation unless otherwise stated. All reactions were conducted at
ambient temperature unless otherwise stated. Flash chromatog-
raphy was carried out using prepacked Isolute Flash or Biotage
silica-gel columns as the stationary phase and analytical grade
solvents as the eluent. Catch and release purification was carried
out using SCX (strong cation exchanger) cartridges, consisting
of bonded-phase silica with sulfonic acid functional groups.
Mass directed preparative HPLC was carried out using a 19 mm
ꢀ100 mm or 30 mm ꢀ100 mm, 5 μm, reversed-phase Waters
Atlantis column as the stationary phase and a gradient from
water þ0.1% formic acid to acetonitrile þ0.1% formic acid as
the eluent. The eluent was monitored by a Waters 996 photo-
diode array and a Micromass ZQ mass spectrometer. All yields
reported are of purified, isolated material. NMR spectra were
obtained at 298K at the frequency stated using either a Bruker
DPX400 or an Oxford Instruments 250 MHz machine and run
as a dilute solution of CDCl3 unless otherwise stated. All NMR
spectra were reference to tetramethylsilane (TMS δH 0, δC 0). All
coupling constants are reported in hertz (Hz), and multiplicities
are labeled s (singlet), bs, (broad singlet), d (doublet), t (triplet),
q (quartet), dd (doublet of doublets), dt (doublet of triplets) and
m (multiplet).
Purity was determined by liquid chromatography/mass spec-
trometry (LC/MS) using an Agilent 1100 HPLC system with a
4.6 mm ꢀ50 mm, 3 μm, reversed phase Waters Atlantis column
as the stationary phase. A gradient elution from 97% water
þ0.05% formic acid/3% acetonitrile þ0.05% formic acid to
97% acetonitrile þ0.05% formic acid over 3 min plus a further
minute continuing this mixture at a flow rate of 1.5 mL/min was
used as the eluent. Retention time is reported as min (with per-
centage intensity for DA/ELSD for the relevant peak). Spectro-
scopic monitoring was performed using an Agilent 1100 diode
array (DA) detector or a Sedex evaporative light scattering
detector (ELSD). Total ion current traces were obtained for
electrospray positive and negative ionization (ESþ/ES-) and
atmospheric pressure chemical positive and negative ionization
(APþ/AP-). All final products were of >95% purity by HPLC
unless otherwise stated. For all key molecules, high resolution
mass spectrometry data were acquired as accurate mass cen-
troided data using a Micromass Q-Tof 2 hybrid quadrupole
time-of-flight mass spectrometer, equipped with a Z-spray
interface.
Figure 4. Crystal structure of 17i in complex with hGluA2 S1S2
ligand binding domain. The compound (yellow carbons) binds at
the dimer interface; a glutamate (red, solid) is bound at the core of
each ligand binding domain.
Observed apparent oral clearance was very small, with an
apparent volume of distribution in excess of total body
water. A dose proportional increase in both Cmax and AUC-
(0-inf) was observed following escalating single doses from
0.25 to 1.75 mg (higher doses were administered to parallel
rather than interlocking cohorts). The accumulation ratio
derived from the subjects receiving daily doses of 17i for 28 d
was 5.6-fold, thus confirming the PK linearity assumption.
The administration of both single and multiple doses of 17i
was well tolerated: neither safety issues were raised nor
withdrawals due to adverse events occurred.
Crystal Structure Determination of 17i in Complex with
hGluA2 S1S2 Ligand Binding Domain. To determine the
binding mode of 17i, we solved the crystal structure of the
hGluA2 S1S2 ligand binding domain in complex with 17i at
1.8 A resolution. The structure showed the compound bound
to the 2-fold axis of the dimer (Figure 4) in a similar manner to
other AMPAR positive modulators.18,19,30 The compound
binds in a similar manner to that described for 6,19 with the
isopropylsulphonamide group of 17i occupying a deep pocket
at the dimer interface (Supporting Information Figure).
N-(2,3-Dihydro-1H-inden-2-yl)-2-propanesulfonamide (9).
2-Aminoindan hydrochloride (5.16 g, 30 mmol) was suspended
in dry dichloromethane (100 mL) and cooled with stirring under
argon to 0 °C. To the suspension was added 1,8-diazabicyclo-
[5.4.0]undec-7-ene (14 g, 92 mmol) followed by the dropwise
addition of isopropylsulfonyl chloride (8.56 g, 60 mmol). The
cooling bath was removed and the mixture stirred at room
temperature for 1 h. The reaction mixture was diluted with
dichloromethane (50 mL) and washed with 1 M hydrochloric
acid (2 ꢀ 50 mL). The organic layer was separated, dried over
sodium sulfate, and evaporated under reduced pressure to give a
yellow oil (11.8 g). The crude product was purified by chroma-
tography on a 50 g Isolute Flash silica-gel column eluting from
20 to 50% ethyl acetate in 40-60 °C petroleum ether to give the
Conclusions
In summary, a series of amino indane sulfonamide deriva-
tives was rapidly optimized toward a series of AMPAR positive
modulators with excellent drug-like physicochemical properties
and good developability characteristics. This study delivered a
number of molecules with encouraging rat pharmacokinetic
properties, and in particular 17i, 17j, and 18e, which were
extensively profiled. The best overall profile was identified in
17i, and although we were initially concerned about the lability
of the ortho-fluoro pyridine, this group proved to be extremely
stable and show no sign of chemical degradation, bioactivation
risk, or metabolic susceptibility. 17i has been proven to be safe
and well tolerated in initial phase I healthy volunteer investiga-
tion and has shown the surprisingly long half-life, in excess of
that predicted from preclinical species.
1
title compound as a pale-yellow solid (6.88 g, 96%). H NMR
(400 MHz, CDCl3) δ 1.39 (6H, d, J = 7 Hz), 2.91 (2H, m), 3.18
(1H, m), 3.31 (2H, m), 4.31 (2H, m), 7.21 (4H, m).
N-(5-Iodo-2,3-dihydro-1H-inden-2-yl)-2-propanesulfonamide
(10). N-(2,3-Dihydro-1H-inden-2-yl)-2-propanesulfonamide
(1.75 g, 7.32 mmol) was dissolved in glacial acetic acid
(30 mL) and then treated with concentrated sulfuric acid (0.8
mL) followed by water (2.8 mL) with stirring. This mixture was
then treated with periodic acid (0.38 g, 1.67 mmol) then iodine
(800 mg, 3.15 mmol), and the whole mixture was stirred at 60 °C
for 4 h. The reaction mixture was allowed to cool and then
partitioned between ethyl acetate (150 mL) and 10% aqueous
sodium metabisulfite (100 mL). The organic layer was sepa-
rated and dried over sodium sulfate and evaporated under
Experimental Section
Chemistry. General Remarks. Starting materials were obtai-
ned from commercial suppliers and used without further purifi-

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5809
vacuum to give the title compound as a yellow oil (2.67 g, 100%).
MS m/z 364 (M - 1). 1H NMR (400 MHz, CDCl3) δ 1.39
(6H, m), 3.39 (2H, m), 3.18 (1H, m), 3.28 (2H, m), 4.38
(1H, m), 4.63 (1H, m), 6.97 (1H, d, J = 8 Hz), 7.51 (1H, m),
7.56 (1H, m).
General Procedure for Suzuki Coupling Reactions to Prepare
11a-h. A mixture of molecule 10 (1.08 mmol) and substituted
phenyl boronic acid (1.62 mmol, Combi-Blocks) in dioxane
(50 mL) was treated with cesium carbonate (1.62 mmol) and
water (15 mL) and degassed with argon for 2 min. Palladium(II)
acetate (0.05 mmol) and triphenylphosphine (0.22 mmol) were
added, and the mixture was stirred at 80 °C under argon for 16 h.
The reaction mixture was cooled and then partitioned between
ethyl acetate (50 mL) and water (50 mL). The organic layer was
separated and dried over Na2SO4 and evaporated in vacuo and
purified using flash chromatography.
N-(5-{4-[(Methylsulfonyl)amino]phenyl}-2,3-dihydro-1H-inden-
2-yl)-2-propanesulfonamide (11a). Yield 30%. MS (ES-) m/z
407 (M - 1). 1H NMR (400 MHz, CDCl3) δ 1.39 (6H, d, J =
7 Hz), 2.97 (2H, m), 3.05 (3H, s), 3.21 (1H, m), 3.37 (2H, m), 4.34
(1H, m), 6.52 (1H, bs), 7.18 (1H, m), 7.24 (1H, m,), 7.29 (1H, d,
J = 7 Hz), 7.40 (4H, m). HRMS (ESI) m/z 431.1064 ([M þ Na]þ
calcd for C19H24NaN2O4S2þ 431.1075).
m), 3.30 (2H, m), 3.38 (2H, m), 4.35 (2H, m), 4.51 (2H, s), 7.22
(1H, m), 7.28 (1H, d, J = 8 Hz), 7.40 (4H, m), 7.47 (1H, m).
HRMS (ESþ) m/z 413.1883 ([M þ H]þ calcd for C23H29N2O3Sþ
413.1899).
(S)-5-Bromo-2-aminoindan (Camphorsulfonate Salt) (20).
Compound 20 was prepared using a similar method to that
described in the literature,27 i.e., by resolution of the free base
form of racemic 5-bromo-2-aminoindane using (1R)-(-)-10-
camphorsulphonic acid to obtain (S)-5-bromo-2-aminoindan
(1R)-(-)-10-camphorsulfonate salt. The absolute configura-
tion of (S)-5-bromo-2-aminoindan (1R)-(-)-10-camphorsul-
fonate salt was confirmed by X-ray crystallography. Further-
more, the enantiomeric purity of (S)-5-bromo-2-aminoindan
(1R)-(-)-10-camphorsulfonate salt was checked by HPLC
using the following conditions: Column, chiralpak AD-H
5 μm, 250 mm ꢀ 4.6 mm. Mobile phase: A, n-hexane; B, etha-
nol þ0.1% isopropyl amine (gradient: isocratic 8% B, flow
rate: 0.8 mL/min; UV WL range: 200-400 nm; analysis time:
17 min). Enantiomer 1 was recovered as 0.84% a/a from the
racemate (retention time = 11.9 min). Enantiomer 2 was
recovered as 99.16% a/a from the racemate (retention time=
12.8 min).
N-[(2S)-5-Bromo-2,3-dihydro-1H-inden-2-yl]-2-propanesulfo-
namide (21). Compound 20 was treated with NaOH (1 M
solution in water to reach pH = 10) in isopropyl acetate as
solvent. The free base form of 20 was converted to 21 as for
racemic mixture 9.
N-[5-(3-Cyanophenyl)-2,3-dihydro-1H-inden-2-yl]-2-propane-
sulfonamide (11c). Yield 73%. MS (ES-) m/z 339 (M - 1). 1H
NMR (400 MHz, CDCl3): 1.41 (6H, d, J = 7 Hz), 2.98 (2H, m),
3.21 (1H, m), 3.38 (2H, m), 4.37 (2H, m), 7.32 (1H, d, J = 8 Hz),
7.39(2H, m),7.53(1H, m),7.62(1H, m), 7.78(1H, m), 7.83(1H, m).
N-(5-{3-[Methyl(methylsulfonyl)amino]phenyl}-2,3-dihydro-
1H-inden-2-yl)-2-propanesulfonamide (11h). Yield 63%. MS
N-[(2S)-5-(6-Fluoro-3-pyridinyl)-2,3-dihydro-1H-inden-2-yl]-
2-propanesulfonamide (17i). Compound 21 was reacted with (6-
fluoro-3-pyridinyl)boronic acid in a similar process used for the
preparation of 17c. The enantiomeric purity of 17i was checked
by HPLC using the following conditions: column, Chiralpak
AS-H 5 μm, 250 mm ꢀ 4.6 mm. Mobile phase: A, n-hexane; B,
ethanol. Gradient, isocratic 30% B; flow rate, 0.8 mL/min; UV
WL range, 200-400 nm; analysis time, 20 min. Enantiomer 1
was recovered as 2.08% a/a from the racemate (retention time =
16.3 min). Enantiomer 2 was recovered as 97.92% a/a from the
racemate (retention time=17.7 min). Enantiomer 2 was con-
firmed to be N-[(2S)-5-(6-fluoro-3-pyridinyl)-2,3-dihydro-1H-
inden-2-yl]-2-propanesulfonamide by X-ray crystallography.
Relevant IR bands: 3154 cm-1 Str. NH, 1600 cm-1 Str. CdN,
CdC, 1314 cm-1 Str. SO2, 1141-1125 cm-1 Str. SO2, Str. C-F.
1H NMR as 17c. 13C NMR (150 MHz, DMSO-d6): δ 162.3,
145.1, 141.8, 140.9, 134.5, 134.4, 125.3, 125.0, 122.8, 109.5, 54.1,
51.7, 39.9, 40.0, 16.4. HRMS (ESþ) m/z 335.1220 ([M þ H]þ
calcd for C17H20N2O2FSþ 335.1230).
N-[(2R)-5-(6-Fluoro-3-pyridinyl)-2,3-dihydro-1H-inden-2-yl]-
2-propanesulfonamide (17h). compound 17h was prepared by an
analogous sequence using (S)-camphor sulfonic acid as resol-
ving agent. The enantiomeric purity of 17h obtained was
checked by HPLC using the same conditions as for the 17i.
Enantiomer 1 was recovered as 99.04% a/a from the racemate
(retention time = 16.62 min). Enantiomer 2 was recovered as
0.96% a/a from the racemate (retention time = 18.29 min).
HRMS (ESþ) m/z 335.1230 ([M þ H]þ calcd for C17H20N2-
O2FSþ 335.1230).
N-[(2S)-5-(6-Methyl-3-pyridinyl)-2,3-dihydro-1H-inden-2-yl]-
2-propanesulfonamide (17j). A mixture of N-[(2S)-5-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)-2,3-dihydro-1H-inden-2-
yl]-2-propanesulfonamide 24 (1.27 g, 3.48 mmol), 2-methyl-5-
bromopyridine (1.1 eq, 658 mg, 3.82 mmol), and cesium
carbonate (1.7 g, 5.22 mmol) in a 3:1 mixture of 1,4-dioxan:
water (13 mL) was degassed with argon for 5 min. Polymer
bound tetrakis(triphenylphosphine)-palladium (5% mol, 1.74 g,
0.17 mmol) was then added, and the whole mixture stirred at 125
°C for 30 min in a microwave reactor. The reaction mixture was
allowed to cool and filtered over a celite pad. The filtrate was
partitioned between ethyl acetate (50 mL) and water (50 mL).
Phases were separated, and the aqueous was back extracted with
ethyl acetate (2 ꢀ 50 mL). Combined organic layers were washed
1
(ES-) m/z 359 (M - 1). H NMR (400 MHz, CDCl3) δ 1.38
(6H, m), 2.99 (2H, m), 3.18 (1H, m), 3.38 (2H, m), 4.35 (2H, m),
7.34 (1H, d J = 8 Hz), 7.45 (2H, m), 7.60 (1H, m), 7.88 (1H, dd,
J = 8 and 1 Hz), 8.19 (1H, m), 8.41 (1H, m).
N-{5-[3-(Aminomethyl)phenyl]-2,3-dihydro-1H-inden-2-yl}-2-
propanesulfonamide (14). A 1.0 M solution of lithium aluminum
hydride in tetrahydrofuran (8 mL) was cooled to 0 °C in an ice/
methanol bath with stirring under argon. The solution was then
treated dropwise with a solution of N-[5-(3-cyanophenyl)-2,3-
dihydro-1H-inden-2-yl]-2-propanesulfonamide (11c) (0.67 g,
1.97 mmol) in dry tetrahydrofuran (20 mL) over 15 min. The
cooling bath was removed and the whole mixture stirred at
reflux for 2 h. The reaction mixture was allowed to cool to room
temperature and quenched with water (0.8 mL), 10% NaOH
(0.8 mL), and water again (1.2 mL). The reaction mix was dried
over sodium sulfate, filtered, and the filtrate evaporated under
reduced pressure to give the 14 as a yellow oil (0.67 g, 100%). MS
(ESþ) m/z 345 (M þ 1). 1H NMR (400 MHz, CDCl3): 1.40 (6H,
d, J = 7 Hz), 1.66 (2H, m), 2.96 (2H, m), 3.19 (1H, m), 3.36 (2H,
m), 3.91 (2H, m), 4.33 (1H, m), 4.67 (1H, m), 7.27 (2H, m), 7.40
(4H, m), 7.49 (1H, s).
N-(5-{3-[(2-Oxo-1-pyrrolidinyl)methyl]phenyl}-2,3-dihydro-1H-
inden-2-yl)-2-propanesulfonamide (15). A solution of N-{5-[3-
(aminomethyl)phenyl]-2,3-dihydro-1H-inden-2-yl}-2-propane-
sulfonamide (14) (69 mg, 0.2 mmol) in dichloromethane (5 mL)
was treated with triethylamine (40 mg, 0.4 mmol), followed by
4-chlorobutanoyl chloride (32 mg, 0.22 mmol), and the whole
mixture stirred at room temperature under argon for 1 h. The
reaction mixture was evaporated under reduced pressure to
remove the dichloromethane and then dimethylformamide
was added (5 mL) followed by potassium carbonate (60 mg,
0.43 mmol) and the mixture stirred at 100 °C for 2 h. The
reaction mixture was cooled and then partitioned between
dichloromethane and 0.5 M hydrochloric acid, and the organic
layer was evaporated under reduced pressure to give a brown oil
which was purified by column chromatography on a 1 g SCX
column eluting from 0 to 50% ethyl acetate in 40-60 °C
petroleum ether to give 15 as a beige solid (16 mg, 19%). MS
(ESþ) m/z 413 (M þ 1). 1H NMR (400 MHz, CDCl3): 1.41 (6H,
d, J = 7 Hz), 2.00 (2H, m), 2.45 (2H, m), 2.96 (2H, m), 3.21 (1H,

5810 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
with brine (2 ꢀ 50 mL), dried, and evaporated under reduced
pressure to give an orange oil, which was purified by Flash silica-
gel column, eluting from 40 to 45% ethyl acetate in cyclohexane
to give the title compound as a white solid (655 mg, 57%). MS
front of a single patch-clamped cell, allowing rapid exchange and
precise application times of solutions (for more information see
http://www.cellectricon.se/). The first channel contained normal
buffer for baseline current measurement. The second channel
contained 3 mM glutamate, which was applied to the cell for
500 ms to record a control (agonist alone) response. The third
channel containednormal buffer, which washed off glutamate for
1-3 min. The fourth channel, which contained either a com-
pound of the invention or a reference compound, was moved in
front of the cell for one minute. The fifth channel contained
glutamate in the presence of the test (or reference) compound,
which was applied to the cell for 500 ms. The sixth channel con-
tained normal buffer, which washed off the glutamate plus test (or
reference) compound for 1-3 min. This procedure was repeated for
increasing concentrations of either a compound of the invention or
a reference compound. The activity of a compound of the invention
is determined by measuring the peak current amplitude or the area
under the curve (500 ms) for the glutamate response in the presence
of the compound of the invention (or reference) and expressing it
as % of potentiation of the response to glutamate alone.
P450 CYPEX Assay. Inhibition (IC50) of human CYP1A2,
2C9, 2C19, 2D6, and 3A4 was determined using Cypex Bacto-
somes expressing the major human P450s. A range of concen-
trations (0.1, 0.2, 0.4, 1, 2, 4, and 10 μM) of test compound were
prepared in methanol and preincubated at 37 °C for 10 min in
50 mM potassium phosphate buffer (pH 7.4) containing recom-
binant human CYP450 microsomal protein (0.1 mg/mL; Cypex
Limited, Dundee, UK) and probe-fluorescent substrate. The
final concentration of solvent was between 3 and 4.5% of the
final volume. Following preincubation, NADPH regenerating
system (7.8 mg glucose 6-phosphate, 1.7 mg NADP, and 6 units
glucose-6-phosphate dehydrogenase/mL of 2% (w/v) NaHCO3;
25 μL) was added to each well to start the reaction. Production
of fluorescent metabolite was then measured over a 10 min time
course using a Spectrafluor plus plate reader. The rate of
metabolite production (AFU/min) was determined at each
concentration of compound and converted to a percentage of
the mean control rate using Magellan (Tecan software). The
inhibition (IC50) of each compound was determined from the
slope of the plot using Grafit v5 (Erithacus software, UK).
Miconazole was added as a positive control to each plate.
CYP450 isoform substrates used were ethoxyresorufin (ER;
1A2; 0.5 μM), 7-methoxy-4-triflouromethylcoumarin-3-acetic
acid (FCA; 2C9; 50 μM), 3-butyryl-7-methoxycoumarin (BMC;
2C19; 10 μM), 4-methylaminomethyl-7-methoxycoumarin
(MMC; 2D6; 10 μM), diethoxyflourescein (DEF; 3A4; 1 μM),
and 7-benzyloxyquinoline (7-BQ; 3A4; 25 μM). The test was
performed in three replicates.
1
(ESþ) m/z 331 (M þ 1). H NMR (400 MHz, CDCl3): 1.41
(6H, m), 2.60 (3H, s), 2.97 (2H, m), 3.20 (1H, m), 3.38 (2H, m),
4.35 (2H, m), 7.21 (1H, d, J = 8 Hz), 7.30 (1H, m), 7.38 (2H, m),
7.73 (1H, dd, J = 8 Hz and 2 Hz), 8.68 (1H, d, J = 2 Hz).
HRMS (ESþ) m/z 331.1471 ([M þ H]þ calcd for C18H23N2O2Sþ
331.1480).
N-[(2S)-5-(5-Fluoro-2-pyridinyl)-2,3-dihydro-1H-inden-2-yl]-
2-propanesulfonamide (18e). A mixture of N-[(2S)-5-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)-2,3-dihydro-1H-inden-2-
yl]-2-propanesulfonamide 24 (48.5 g, 0.133 mol), 2-bromo-5-
fluoropyridine (26.9 g, 0.15 mol), and cesium carbonate (1 M
water solution, 270 mL, 0.266 mol) in 1,4-dioxan (490 mL) was
degassed with argon for 5 min. Polymer bound tetrakis(tri-
phenylphosphine)-palladium (1% mol, 2.66 g, 1.33 mmol) was
then added and the whole mixture stirred at 83 °C for 3 h. The
reaction mixture was allowed to cool and filtered over a celite
pad. The filtrate was partitioned between ethyl acetate (500 mL)
and water (500 mL). Phases were separated, and the aqueous
was back-extracted with ethyl acetate (2 ꢀ 500 mL). Combined
organic layers were washed with brine (2 ꢀ 500 mL), dried, and
evaporated under reduced pressure to give an orange oil, which
was purified by Flash silica-gel column, eluting from 10 to 20%
ethyl acetate in cyclohexane to give the title compound as a white
solid (35 g, 79%). MS (APCI) m/z 335 (M þ 1). 1H NMR (400
MHz, CDCl3): 1.40 (6H, m), 2.96 (2H, m), 3.20 (1H, m), 3.38
(2H, m), 4.34 (2H, m), 7.31 (1H, m), 7.47 (1H, m), 7.71 (2H, m),
7.81 (1H, m), 8.52 (1H, m). HRMS (ESþ) m/z 335.1217 ([M þ
H]þ calcd for C17H20N2O2FSþ 335.1230).
Calcium Influx Fluorescence Assay. Stable HEK293-hGluR2-
iQ/R unedited cells were grown in a DMEM/F12, 10% FBS in a
5% CO2 incubator in 175 cm2 T-flasks. Two-three d before
experiment, cells were detached and seeded into 384-well black
PDL-coated plates with clear bottom up to confluence. On the
day of the experiment, cell plates were washed 3 times with assay
buffer (in mM): 20 HEPES, 145 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2,
5.5 glucose, pH 7.3, and loaded with 2 μM Fluo4 AM for 60 min
at room temperature. Cell plates were washed again to remove
unloaded Fluo4 and placed into a FLIPR. A dual addition pro-
tocol was used to add compound solution 5 min before 100 μM
glutamate stimulation. The glutamate-induced increased fluor-
escence was analyzed by using ActivityBase software and com-
pared with positive control cyclothiazide.
Whole Cell Voltage-Clamp ElectrophysiologyAssay. This assay
involved the electrophysiological characterization of AMPA
receptor positive modulators using HEK293 cells stably expres-
sing human GluA2 flip (unedited) subunits which form a func-
tional homotetrameric AMPA receptor. The extracellular re-
cording solution contained (in mM): 140 NaCl, 2 KCl, 1 MgCl2,
2 CaCl2, 12 N-[2-hydroxyethyl]-piperazine-N-[2-ethanesulfonic
acid (HEPES), 10 D-glucose, pH 7.35. The intracellular solution
contained (in mM): 150 CsCl, 10 HEPES, 10 ethylene glycol-bis-
(γ-aminoethylether)-N,N,N0,N0,-tetra-acetic acid (EGTA), pH
7.3. For perforated patch recordings intracellular solution con-
taining amphotericin B (240 μg/mL) was used to backfill the pipet
while intracellular solution alone was used to fill just the tip (the
patch clamp pipettes have a resistance of between 2 and 5 MΩ).
Amphotericin B creates small pores in the cell membrane beneath
the electrode which allow small ions to pass across the membrane
(and therefore allow electrical control of the cell) without the
dialysis of second messenger molecules out of the cell, which
could result in metabolic rundown of the cell leading to incon-
sistent receptor activation.31 The membrane potential of the cell
was held at -60 mV and perforated patch clamp electrophysiol-
ogy performed using EPC9 or 10 patch clamp system. The cell
was positioned in front of the first of 16 linearly arranged
channels. The system moves one channel and then the next in
Intrinsic Clearance (CLi) Assay. Intrinsic clearance (CLi)
values were determined in rat and human liver microsomes.
Test compounds (0.5 μM) were incubated at 37 °C for 30 min in
50 mM potassium phosphate buffer (pH 7.4) containing 0.5 mg
microsomal protein/mL. The reaction was started by addition of
cofactor (NADPH; 8 mg/mL). The final concentration of sol-
vent was 1% of the final volume. At 0, 3, 6, 9, 15, and 30 min, an
aliquot (50 μL) was taken, quenched with acetonitrile contain-
ing an appropriate internal standard, and analyzed by HPLC-
MS/MS. The intrinsic clearance (CLi) was determined from the
first-order elimination constant by nonlinear regression using
Grafit v5 (Erithacus software, UK), corrected for the volume of
the incubation and assuming 52.5 mg microsomal protein/g liver
for all species. Values for CLi were expressed as mL/min/g liver.
The lower limit of quantification of clearance was determined
to be when <15% of the compound had been metabolized by
30 min, and this corresponded to a CLi value of 0.5 mL/min/g
liver. The upper limit was 50 mL/min/g liver.
Determination of the X-ray Structure of Human AMPA Hu-
man Glutamate Receptor GluA2 S1S2 Ligand Binding Domain
with 17i. Construct Generation, Expression, and Purification. To
construct the pET15b-humanGluA2flipS1-GlyThr-S2 plasmid

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15 5811
the DNA fragments, S1 and S2 were amplified in separate
polymerase chain reactions. The pCDNA3.2-humanGluA2flip
plasmid was used as a template for amplification of the hGluA2-
flip S1 and S2 fragments. The primers for the S1 reaction were,
50hGluA2flipS1: 50-CGATCCATATGGGCTCTGGAAATG-
ACACCTCTGGG-30 and 30hGluA2flipS1: 50-TTCGATGGG-
GGTGCCCTTCTTGATCATGATAGATATC-30; the pri-
mers for amplification of S2 were 50hGluA2flipS2: 50-ATCAA
GAAGGGCACCCCCATCGAAAGTGCTGAGGATC-30 and
30hGluA2flipS2: 50-GGATCCTCGAGGGCTCCACATTCA-
CCTTTATC-30; the underlined sequences are complementary
to the template. Each 100 mL of PCR used 100 ng of pCDNA-
3.1-humanGluA2flip template, 25 pmol of each primer, 40 mM
of each dNTP, 5U of Pfu DNA polymerase, and 25 cycles of
amplification. The major products were of the predicted size and
were gel purified and eluted into 30 mL of water. Two ml of each
product and primers 50hGluA2flipS1 and 30hGluA2flipS2 were
used in a second round of PCR using the complementary over
lapping regions to amplify the full length S1-GlyThr-S2 frag-
ment. The S1-GlyThr-S2 fragment and the vector, pET15b,
were digested with NdeI and XhoI. Subsequently, all digest pro-
ducts were gel purified; S1-GlyThr-S2 was ligated into the
pET15b vector to generate pET15b-humanGluA2flipS1-GlyThr-
S2. This was verified by sequencing.
Expression and Purification of S1S2 Ligand Binding Domain.
pET15b-humanGluA2flipS1-GlyThr-S2 was transformed into
Escherichia coli BL21(DE3) for expression. Overnight cultures
grown in LB containing 100 μg/mL carbenicillin were diluted
into fresh media, grown at 37 °C to an OD600 of ∼0.8 and ind-
uced with 1 mM IPTG. After 4 h induction, cells were harvested
by centrifugation and inclusion bodies prepared. Cell paste,
resuspended to 25% (w/v) in Tris buffer (50 mM Tris, pH 8.0,
0.2 M NaCl, 1 mM EDTA), was incubated with 20 μg/mL
lysozyme and the cells broken on a Manton Gaulin homogeni-
zer. After addition of Triton X-100 to 2%(w/v), the pellet was
harvested by centrifugation and washed twice by resuspending
in Tris buffer containing 1 M NaCl and 5 mM DTT, each time
recovering the pellet by centrifugation. The final pellet was re-
suspended in a minimum volume of Tris buffer plus 5 mM DTT.
Then 500 mg aliquots of resuspended inclusion body were solu-
bilized in guanidine and refolded and purified exactly according
to the method of Chen et al.32
sidues (Arg231, Thr232, Pro233, Val246) that differ in sequence
between the rat and human S1S2 ligand binding domain are not
close to the positive modulator binding site and do not affect
binding of 17i. The atomic coordinates for the crystal structure
described here have been deposited at the Protein Data Bank
under entry code 2xhd.
Acknowledgment. We thank Dr. Bill Leavens for generat-
ing the HRMS data.
Supporting Information Available: Full experimental proce-
dures and characterization for all compounds not included
above. Plasma protein binding of 17i in mouse, rat, dog, cyno-
molgus monkey, and human and brain tissue binding in the rat.
Methods for membrane permeability, P-glycoprotein transport,
and interaction and analysis of blood, plasma, or brain samples
from in vivo and protein binding studies. High resolution mass
spectrometry methods and comparison of the crystal struc-
ture of 6 and 17i. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Sanacora, G.; Zarate, C. A.; Krystal, J. H.; Manji, H. K. Targeting
the glutamatergic system to develop novel, improved therapeutics
for mood disorders. Nature Rev. Drug Discovery 2008, 7, 426–437.
(2) Zarate, C. A., Jr.; Manji, H. K. The role of AMPA receptor
modulation in the treatment of neuropsychiatric diseases. Exp.
Neurol. 2008, 211, 7–10.
(3) Black, M. D. Therapeutic potential of positive AMPA modulators
and their relationship to AMPA receptor subunits. A review of
preclinical data. Psychopharmacology (Berlin) 2005, 179, 154–163.
(4) Granger, R.; Staubli, U.; Davis, M.; Perez, Y.; Nilsson, L.; Rogers,
G. A.; Lynch, G. A drug that facilitates glutamatergic transmission
reduces exploratory activity and improves performance in a learn-
ing-dependent task. Synapse 1993, 15, 326–329.
(5) Lynch, G.; Gall, C. M. Ampakines and the threefold path to
cognitive enhancement. Trends Neurosci. 2006, 29, 554–562.
(6) Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F. The
glutamate receptor ion channels. Pharmacol. Rev. 1999, 51, 7–61.
(7) Ozawa, S.; Kamiya, H.; Tsuzuki, K. Glutamate receptors in the
mammalian central nervous system. Prog. Neurobiol. 1998, 54,
581–618.
(8) Mansour, M.; Nagarajan, N.; Nehring, R. B.; Clements, J. D.;
Rosenmund, C. Heteromeric AMPA receptors assemble with a
preferred subunit stoichiometry and spatial arrangement. Neuron
2001, 32, 841–853.
(9) Tichelaar, W.; Safferling, M.; Keinanen, K.; Stark, H.; Madden,
D. R. The Three-dimensional Structure of an Ionotropic Gluta-
mate Receptor Reveals a Dimer-of-Dimers Assembly. J. Mol. Biol.
2004, 344, 435–442.
(10) Armstrong, N.; Gouaux, E. Mechanisms for activation and antag-
onism of an AMPA-sensitive glutamate receptor: crystal structures
of the GluR2 ligand binding core. Neuron 2000, 28, 165–181.
(11) Gouaux, E. Structure and function of AMPA receptors. J. Physiol.
2004, 554, 249–253.
(12) Collingridge, G. L.; Isaac, J. T.; Wang, Y. T. Receptor trafficking
and synaptic plasticity. Nature Rev. Neurosci. 2004, 5, 952–962.
(13) Milstein, A. D.; Nicoll, R. A. Regulation of AMPA receptor gating
and pharmacology by TARP auxiliary subunits. Trends Pharma-
col. Sci. 2008, 29, 333–339.
(14) Tomita, S.; Adesnik, H.; Sekiguchi, M.; Zhang, W.; Wada, K.;
Howe, J. R.; Nicoll, R. A.; Bredt, D. S. Stargazin modulates
AMPA receptor gating and trafficking by distinct domains. Nature
2005, 435, 1052–1058.
LC/MS analysis confirmed purification of protein with the
expected mass of 29237 Da. Purified protein was dialyzed ex-
haustively against 10 mM Hepes pH 7.0, 20 mM NaCl, 1 mM
EDTA, concentrated to 18-20 mg/mL, flash frozen in liquid
nitrogen, and stored at -80 °C.
Crystallization. Protein was incubated with solid compound at
4 °C overnight. The sample was then spun at 4 °C/13.2 rpm/15 min
in a microfuge and screened with a grid of (18, 20, 22, 24% peg 8K,
0.1 M sodium acetate at pH 5.0, 0.2 M ammonium sulfate) in a 24-
well VDX plate using the hanging drop method (1 þ 1 μL drops).
Crystal quality was improved by hair-seeding. Crystals were
harvested by transferring into well solution plus 15% glycerol for
approximately 10 s, followed by flash freezing in liquid nitrogen.
Data Collection and Structure Solution. X-ray diffraction data
were collected at the ESRF on BL23. Data were processed with
DENZO and scaled with SCALEPACK,33 resulting in an Rmerge
of 7.3%. The structure was solved by molecular replacement
using the rat GluA2 S1S2J-N754S with CTZ (PDB code 1LBC)
using AMORE.34 Prior to molecular replacement, the cyclothia-
zide and water molecules were removed and the residues that dif-
fered between the rat and human protein were changed to alan-
ine. Model building was carried out using the program O,35 and
refinement was carried out with CNX using standard positional
refinement and B-factor refinement protocols.36 The final Rwork
was 19.8% and Rfree was 23.6% for data between the resolution
of 20 and 1.8 A. The final model contains 569 water molecules, 2
glutamate molecules, 5 sulfate molecules, and 2 17i molecules
(final statistics in Supporting Information Table 1). The four re-
(15) Soto, D.; Coombs, I. D.; Kelly, L.; Farrant, M.; Cull-Candy, S. G.
Stargazin attenuates intracellular polyamine block of calcium-
permeable AMPA receptors. Nature Neurosci. 2007, 10, 1260–
1267.
(16) Kuusinen, A.; Arvola, M.; Keinanen, K. Molecular dissection of
the agonist binding site of an AMPA receptor. EMBO J. 1995, 14,
6327–6332.
(17) Armstrong, N.; Sun, Y.; Chen, G. Q.; Gouaux, E. Structure of a
glutamate-receptor ligand-binding core in complex with kainate.
Nature 1998, 395, 913–917.
(18) Jin, R.; Clark, S.; Weeks, A. M.; Dudman, J. T.; Gouaux, E.;
Partin, K. M. Mechanism of positive allosteric modulators acting
on AMPA receptors. J. Neurosci. 2005, 25, 9027–9036.

5812 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 15
Ward et al.
(19) Sobolevsky, A. I.; Rosconi, M. P.; Gouaux, E. X-ray structure,
symmetry and mechanism of an AMPA-subtype glutamate recep-
tor. Nature 2009, 462, 745–756.
(29) Foley, A. G.; Murphy, K. J.; Hirst, W. D.; Gallagher, H. C.;
Hagan, J. J.; Upton, N.; Walsh, F. S.; Regan, C. M. The 5-HT(6)
receptor antagonist SB-271046 reverses scopolamine-disrupted
consolidation of a passive avoidance task and ameliorates spatial
task deficits in aged rats. Neuropsychopharmacology 2004, 29,
93–100.
(30) Hald, H.; Ahring, P. K.; Timmermann, D. B.; Liljefors, T.;
Gajhede, M.; Kastrup, J. S. Distinct structural features of
cyclothiazide are responsible for effects on peak current ampli-
tude and desensitization kinetics at iGluR2. J. Mol. Biol. 2009,
391, 906–917.
(31) Virginio, C.; Giacometti, A.; Aldegheri, L.; Rimland, J. M.;
Terstappen, G. C. Pharmacological properties of rat alpha 7
nicotinic receptors expressed in native and recombinant cell sys-
tems. Eur. J. Pharmacol. 2002, 445, 153–161.
(32) Chen, P. E.; Wyllie, D. J. Pharmacological insights obtained from
structure-function studies of ionotropic glutamate receptors. Br.
J. Pharmacol. 2006, 147, 839–853.
(33) Otwinowski, Z. Data collection and processing. In Proceedings of
CCP4 Study Weekend; Sawyer, L., Ed.; SERC Daresbury Laboratory:
Warrington, UK: 1993; pp 56-62.
(34) Bailey, S. The CCP4 suite: programs for protein crystallography.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 1994, 50, 760–763.
(35) Jones, T. A.; Zou, J. Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and
the location of errors in these models. Acta Crystallogr., Sect. A:
Found. Crystallogr. 1991, 47, 110–119.
(36) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges, M.;
Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: A new software suite for macro-
molecular structure determination. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 1998, 54, 905–921.
(20) Sommer, B.;Keinanen, K.;Verdoorn, T. A.;Wisden, W.;Burnashev,
N.; Herb, A.; Kohler, M.; Takagi, T.; Sakmann, B.; Seeburg, P. H.
Flip and flop: a cell-specific functional switch in glutamate-
operated channels of the CNS. Science 1990, 249, 1580–1585.
(21) Leever, J. D.; Clark, S.; Weeks, A. M.; Partin, K. M. Identifica-
tion of a site in GluR1 and GluR2 that is important for modulation
of deactivation and desensitization. Mol. Pharmacol. 2003, 64, 5–10.
(22) Harpsoe, K.; Liljefors, T.; Balle, T. Prediction of the binding mode
of biarylpropylsulfonamide allosteric AMPA receptor modulators
based on docking, GRID molecular interaction fields and 3D-
QSAR analysis. J. Mol. Graphics Modell. 2008, 26, 874–883.
(23) Morrow, J. A.; Maclean, J. K.; Jamieson, C. Recent advances in
positive allosteric modulators of the AMPA receptor. Curr. Opin.
Drug Discovery Dev. 2006, 9, 571–579.
(24) Francotte, P.; de Tullio, P.; Fraikin, P.; Counerotte, S.; Goffin, E.;
Pirotte, B. In search of novel AMPA potentiators. Recent Pat. CNS
Drug Discovery 2006, 1, 239–246.
(25) Ward, S. E.; Bax, B. D.; Harries, M. Challenges for and current
status of research into positive modulators of AMPA receptors.
Br. J. Pharmacol. 2010, 160, 181–190.
(26) Subramanian, G.; Kitchen, D. B. Computational models to predict
blood-brain barrier permeation and CNS activity. J. Comput-
Aided Mol. Des. 2003, 17, 643–664.
(27) Prashad, M.; Hu, B.; Har, D.; Repic, O.; Blacklock, T. J.;
Acemoglu, M. Efficient and practical syntheses of (R)-(5-amino-
2,3-dihydro-1H-inden-2-yl)carbamic acid methyl ester. Adv. Synth.
Catal. 2001, 343, 461–472.
(28) Ennaceur, A.; Delacour, J. A new one-trial test for neurobiological
studies of memory in rats. 1: Behavioral data. Behav. Brain Res.
1988, 31, 47–59.