Synthesis and biological activities of aryl-ether-, biaryl-, and fluorene-aspartic acid and diaminopropionic acid analogs as potent inhibitors of the high-affinity glutamate transporter EAAT-2

Article abstract of DOI:10.1016/j.bmcl.2005.08.003

Excitatory amino acid transporters (EAATs) play a pivotal role in maintaining glutamate homeostasis in the mammalian central nervous system, with the EAAT-2 subtype thought to be responsible for the bulk of the glutamate uptake in forebrain regions. A complete elucidation of the functional role of EAAT-2 has been hampered by the lack of potent and selective pharmacological tools. In this study, we describe the synthesis and biological activities of novel aryl-ether, biaryl-, and fluorene-aspartic acid and diaminopropionic acid analogs as potent inhibitors of EAAT-2. Compound (16) represents one of the most potent (IC₅₀ = 85 ± 5 nM) and selective inhibitors of EAAT-2 identified to date.

Full text of DOI:10.1016/j.bmcl.2005.08.003

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4985–4988
Synthesis and biological activities of aryl-ether-, biaryl-, and
fluorene-aspartic acid and diaminopropionic acid analogs as potent
inhibitors of the high-affinity glutamate transporter EAAT-2
Alexander Greenfield,a,* Cristina Grosanu,a John Dunlop,b Beal McIlvain,b
Tikva Carrick,b Brian Jow,b Qiang Lu,b Dianne Kowal,b John Williamsc and John Buteraa
aChemical and Screening Sciences, Wyeth Research, CN 8000, Princeton, NJ 08543, USA
bDiscovery Neuroscience, Wyeth Research, CN 8000, Princeton, NJ 08543, USA
cNeurocrine Biosciences, 12790 El Camino Real, San Diego, CA 92130, USA
Received 7 July 2005; revised 2 August 2005; accepted 2 August 2005
Available online 13 September 2005
Abstract—Excitatory amino acid transporters (EAATs) play a pivotal role in maintaining glutamate homeostasis in the mammalian
central nervous system, with the EAAT-2 subtype thought to be responsible for the bulk of the glutamate uptake in forebrain
regions. A complete elucidation of the functional role of EAAT-2 has been hampered by the lack of potent and selective pharma-
cological tools. In this study, we describe the synthesis and biological activities of novel aryl-ether, biaryl-, and fluorene-aspartic acid
and diaminopropionic acid analogs as potent inhibitors of EAAT-2. Compound (16) represents one of the most potent
(IC50 = 85 5 nM) and selective inhibitors of EAAT-2 identified to date.
Ó 2005 Elsevier Ltd. All rights reserved.
It is widely believed that the dysfunction of glutamate
transmission participates in the etiology of a number of
neurodegenerative and neuropsychiatric disorders and
diseases. In the mammalian central nervous system, the
excitatory amino acid transporter (EAAT) family of pro-
teins is responsible for the high-affinity sodium-depen-
dent uptake of glutamate into both astroglial cells and
neurons. Normal EAAT function is required both for
the efficient termination of glutamatergic neurotransmis-
sion and clearance of glutamate from the synaptic cleft,
thereby preventing excitotoxicity. Among the five recent-
ly identified subtypes of excitatory aminoacid transport-
ers (EAAT-1–5), three of them (EAAT-1/GLAST,
EAAT-2/GLT-1, and EAAT-3/EAAC1) are involved in
synaptic glutamate homeostasis. Further classification
distinguishes neuronal transporter EAAT-3 from glial
transporters EAAT-1 and EAAT-2, with the latter being
the major contributor to glutamate uptake1 from the
synapse.
An ever-growing interest in the area led to the discovery
of several classes of restricted glutamate analogs as
inhibitors of EAATs (pyrolidine dicarboxylates,2
aminocyclobutane dicarboxylates,3 and carboxycyclo-
propyl glycines4). Many of these relatively compact mol-
ecules act as competitive (transportable) substrates
inducing transport currents and heteroexchange. A close
structural similarity of these earlier series with glutamate
presents a plausible explanation for the poor selectivity
across EAAT subtypes and substantial affinity to other
glutamate receptors (mGluRs and iGluRs). Finally,
their use as pharmacological tools is also affected by
modest micromolar potency and poor physico-chemical
properties. While more recently designed non-transport-
able ligands, such as dl-threo-b-hydroxyaspartate and its
benzylated derivative (TBOA),5 and the novel heptane
dicarboxylate-3-amino-tricyclo[2.2.1.02.6]heptane-1,3-di-
carboxylic acid6 offered improved potency, they still fall
short of being ideal tool molecules. The latest generation
of TBOA-based analogs delivers nanomolar potent
EAAT-2 inhibitors (measured in transfected MDCK
cells), with the most selective agent PMB-TBOA ((2S,
3(S)-3-[3-(4-methoxybenzoylamino)benzyloxy]aspartate)7
exhibiting a 39-fold selectivity over EAAT-3. Manifest-
ing a notably better profile, this series still leaves room
Keywords: Glutamate transporters; EAAT-2 inhibitors.
*
Corresponding author. Tel.: +1 732 274 4044; fax: +1 732 274
4505; e-mail: greenfa@wyeth.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.08.003

4986
A. Greenfield et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4985–4988
O
H
O
O
H
N
N
H
Ar
OH
(iii)
N
Bn
(ii)
(i)
Ar
OH
NHBoc
Ar
O
Ar NH2
O
NH2
O
O
NHBoc
1 - 16
O
O
(vi) , (vii)
(iv), (v)
COOMe
NHBoc
COOH
NH2
Ar COOH
Ar
N
H
Ar
N
H
17 - 28
Scheme 1. Synthesis of aspartamides and diaminopropionic acid analogs. Reagents and conditions: (i) N-Boc-L-aspartic acid benzyl ester, 1-[3-
(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride, CH2Cl2, (ii) H2/Pd on carbon, 50 psi, (iii) trifluoroacetic acid (TFA), CH2Cl2, (iv) 1. N-
hydroxysuccinimide, N,N0-diisopropylcarbodiimide, CH2Cl2. 2. diisopropylethylamine, N-Boc-3-amino-L-alanine, (v) TMS–diazomethane, (vi)
NaOH, MeOH, THF, (vii) trifluoroacetic acid (TFA), CH2Cl2.
The general synthetic routes, leading to aspartate and
2,3-diaminopropionate analogs, are shown in Scheme
1. As outlined in Scheme 1, aspartic acid analogs 1–16
were synthesized through standard carbodiimide-medi-
ated coupling between an amine and an appropriately
protected aspartic acid. It is worth mentioning that step
(ii) in Scheme 1 proved to be challenging in the synthesis
of aspartamides. For example, basic hydrolysis or even
exposure of the coupled protected material to elevated
temperatures led to the contamination of the target
materials with the product of intramolecular cyclization
(1) and its subsequent non-selective opening (2) (Fig. 1).
Hydrogenolysis with vigorous stirring at elevated pres-
sure (50 psi) or Cu2+-mediated debenzylation8 (step
(ii) in Scheme 1, and steps (iv) and (viii) in Scheme 2)
allowed us to overcome these difficulties.
O
O
NH2
HO
Ar
N
H
Ar
N
O
NH2
O
1
2
Figure 1. Typical side products of aspartamide synthesis.
for improvement in selectivity and physico-chemical
properties.
In this study, we present the synthesis and structure–ac-
tivity studies of a structurally distinct series possessing
high potency and selectivity in HEK cell lines together
with a potentially promising overall biopharmacological
profile. Focusing on creating sufficient structural dissim-
ilarity with glutamate and improvement of pharmaco-
logical characteristics, we constructed variable
lipophilic structural fragments with the terminal car-
boxy- and amino-groups of aspartic and 2,3-diamino-
propionic acid correspondingly. In comparison with
TBOA series,7 these inhibitors possess only one chiral
center and a free carboxylic acid group, potentially sim-
plifying the syntheses and improving physico-chemical
properties.
Diaminopropionic acid analogs were prepared via
in situ formation of hydroxysuccinimide ester with the
appropriate aryl carboxylic acid, followed by coupling
with diaminopropionic acid.9 To simplify isolation and
purification of the intermediate, the crude material was
subjected to esterification with trimethylsilyl diazome-
thane. Final hydrolysis of a methyl ester with diluted
(1 N) sodium hydroxide and removal of Boc-group with
TFA afforded analogs 17–28.
O
O
H
N
H
N
F
F
NH2
NH2
F
F
Br
O
Br
O
F
F
NO
2 (ii)
F
Br
O
F
F
Br
O
Bn
(iv)
(iii)
(i)
OH
OH
O
O
O
NH2
O
O
O
NHBoc
O2N
(16)
(13)
(v)
O
H
N
H
N
F
F
F
F
F
F
Bn
(viii)
NO
2 (vi)
F
F
(vii)
O
NH2
NHBoc
O
O
O
O
O
H
N
HO
HO
COOMe
Br
OH
(ix)
(x)
R
(xi)
(xii)
H2N
R
(xiii)
R
Br
O
NH2
O2N
O2N
O2N
O2N
R
Scheme 2. Syntheses of the analogs with highly functionalized cores. Reagents and conditions: (i) 2-bromo-4,5-difluorophenol, NaH, DMSO, (ii)
SnCl2, EtOH, 80 °C, (iii) N-Boc-L-aspartic acid-a-benzyl ester, 1-[3-dimethylamino)propyl]-3-ethylcarbodiimide, CH2Cl2, (iv) 1. trifluoroacetic acid
(TFA), CH2Cl2, 2. CuSO4, EtOH, 40 °C, (v) Pd(OAc)2/Na2CO3/DMA, 120 °C, (vi) H2/PtO2, MeOH, THF, (vii) N-Boc-L-aspartic acid-a-benzyl
ester, 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride, CH2Cl2, (viii) 1. H2/Pd on carbon, 50 psi, 2. trifluoroacetic acid (TFA),
CH2Cl2, (ix) LiBH4, E2O, (x) Pd(OAc)2, Tol3P, arylboronic acid, dioxane, (xi) polyphosphoric acid, 110 °C, (xii) H2, PtO2, MeH, THF, (xiii) N-Boc-
L-aspartic acid-a-benzyl ester, 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride, CH2Cl2, 2. H2/Pd on carbon, 50 psi, 3. trifluoroacetic
acid (TFA), CH2Cl2.

A. Greenfield et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4985–4988
4987
Table 1. The inhibitory properties of aspartamides and diaminoprop-
ionic acid analogs
Table 1 (continued)
Inhibition (IC50, lM)a
EAAT-3
0.6
EAAT-1
1
EAAT-2
O
H
N
Ar
OH
CH3
19
20
21
0.4
0.3
0.2
O
NH2
CF3
Inhibition (IC50, lM)a
EAAT-3
10
EAAT-1
2
EAAT-2
0.1
3
20
2
CH3
F3C
1
2
3
CH3
CH3
1.5
CH3
CH3
F
Cl
Cl
0.6
0.3
0.3
0.2
0.2
0.6
22
23
24
0.85
5
0.7
6
0.1
1
Cl
F
F
F
Cl
F
CF3
CF3
4
0.14
0.13
0.1
CH3
CF3
9
63
67
CH3
5
6
0.6
1
0.4
3
0.2
0.5
0.7
0.08
0.1
0.2
0.3
0.1
0.5
0.3
0.13
0.08
0.7
CF3
F
Cl
25
26
27
28
3
100
22
0.3
3
0.3
1
Cl
CH3
F
9
7
5.6
0.05
0.2
0.65
0.45
2.4
1.8
1
6
F3C
CH3
0.42
2
0.7
F
8
0.1
0.15
0.3
0.2
3.8
1
13
O
Cl
9
a Concentration required to inhibit glutamate uptake in HEK cells by
50%. The error margin is 10%.
* IC50 6 10 nM in MDCK cell line.
Br
Cl
10
11
12
13
14
15
16
17
Examples of the synthesis of the targets with highly elab-
orated lipophilic cores are shown in Scheme 2. Yields
usually ranged from good to excellent (40–90%). The
structural integrity of the analogs was confirmed by tra-
ditional analytical techniques (NMR, MS, CHN, and
IR).
Cl
F3C
F
F
O
It is noteworthy that among the plethora of nitro-group
reduction methods, only SnCl2 in ethanol10 (step (ii) in
Scheme 2) showed excellent regioselectivity preserving
the bromine atom intact. A potentially problematic
Pd-catalyzed formation of dibenzofuran11 (v) and
polyphosphoric acid mediated cyclization of biaryl
methanols into corresponding fluorenes12 (xi), both
requiring strenuous conditions (110–120 °C), took place
without major complications, routinely providing prod-
ucts with satisfactory yield (40–70%).
CH3
N
0.7
2.9
5
O
F
14.5
3.8
1
O
F
Br
O
F
F
2
The EAAT inhibitory properties of compounds, pre-
pared in the course of the study, are shown in Table 1.
18
0.5
0.1
0.06*
The compounds under study were tested in a HEK cell
line expressing each of the human transporter subtypes
CF3
(continued on next page)

4988
A. Greenfield et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4985–4988
Table 2. Pharmacological profile of bromo-ether 16
System
HEK
85
Oocytes
130
Synaptosomes
35
mGluR
iGluR
Inhibition (IC50, nM)
5
1
7
No effect
No effect
EAAT-1–3.13 Compound 18 was also tested for its
EAAT-2 uptake inhibitory effect on MDCK cells in an
effort to benchmark its potency with TBOA series.7 It
showed potent inhibition of EAAT-2 uptake with an
IC50 < 10 nM, which is comparable with the best TBOA
analogs reported to date.
bromo-4,5-difluorophenoxy)phenyl]-L-asparagine)—a
potent, selective, competitive non-substrate inhibitor of
EAAT-2. As one of the most potent and selective
EAAT-2 inhibitors identified to date, compound 16 repre-
sents a unique addition to the arsenal of pharmacological
tools which can be used to elucidate further the role of spe-
cific EAAT subtypes and to improve our understanding
of hyperglutamatergic and neurodegenerative disorders.
Overall, both aspartamide and diaminopropionamide
series routinely elicited high EAAT-2 inhibitory potency
in HEK cells with an IC50 < 100 nM (3, 8, 16, and 18).
Linear arrangement of amino acid fragment with respect
to distal aromatic ring is a prerequisite for potency in
diaminopropionamide series (weakly active 24). The
selectivity proved to be an elusive and complex issue.
As exemplified by compounds 26, 27, and 28, breaking
of aryl–aryl bond with concomitant loss of rigidity did
not deteriorate EAAT-2 inhibition but gave rise to some
selectivity (10- to 20-fold) versus EAAT-1 and EAAT-3.
On the other hand, fluorenone 28, not only lost EAAT-2
potency, but also reversed the sense of inhibitory prefer-
ence. The best results achieved in diaminopropionamide
series are represented by analogs 20 and 25. Possessing
significant EAAT-2 potency, they also showed excellent
separation for EAAT-1 (60- to 300-fold) and a moderate
one for EAAT-3 (ꢀ10-fold). Linear, in respect to amino
acid residue, arrangement in biaryl analogs, while lack-
ing selectivity, revealed a uniformly high blocking poten-
cy against EAAT-2 transporter in both series
(aspartamides- and diaminopropionamides). Planarity
derived from the conversion of biaryls into fluorenes
(8–12, 27) did not affect the properties, leaving them po-
tent non-selective inhibitors (with the exception of 12) of
EAAT-2 transporter. On the other hand, perturbations
in the proximity of aryl–aryl linkage in some cases (1,
20–22) produced the desired selectivity trend. Our at-
tempts to exploit these results led to the synthesis and
evaluation of the ethers 13–16. While potent EAAT-2
blockers dibenzofuran 13 and phenoxazine 14 showed
no selectivity, less rigid ethers 15 and, especially, 16
showed both increased potency and desired inhibitory
preference. In addition to being the most potent com-
pound, bromo-ether 16 (EAAT-2 IC50 ꢀ 85 nM) was
the most selective with 59- and 45-fold selectivity over
EAAT-1 and EAAT-3, respectively. Because of superior
combination of potency and selectivity, the compound
16 was fully characterized pharmacologically. It showed
high potency in rat cortical synaptosomes and EAAT-2
expressing oocytes (Table 2). In addition, the compound
did not show cross-receptor reactivity (failed to activate
both ionotropic and metabotropic glutamate receptors)
and proved to be a competitive non-substrate inhibitor
of EAAT-2 (by failure to activate transporter-like cur-
rent when applied to oocytes expressing EAAT-1–3
transporters14).
References and notes
1. Dunlop, J.; Zaleska, M. M.; Eliasof, S.; Moyer, J. A.
Excitatory amino acid transporters as emerging targets for
central nervous system therapeutics. Emerging Therapeutic
Targets 1999, 3, 543.
2. Bridges, R. J.; Stanley, M. S.; Anderson, M. W.; Cotman,
C. W.; Chamberlin, A. R. J. Med. Chem. 1991, 34, 717.
3. Fletcher, E. J.; Mewett, K. N.; Drew, C. A.; Allan, R. D.;
Johnston, G. A. Neurosci. Lett. 1991, 121, 133.
4. Nakamura, Y.; Kataoka, K.; Ishida, M.; Shinozaki, H.
Neuropharmacology 1993, 32, 833.
5. Lebrun, B.; Sakaitani, M.; Shimamoto, K.; Yasuda-
Kamatani, Y.; Nakajima, T. J. Biol. Chem. 1997, 272,
20336.
6. Dunlop, J.; Eliasof, S.; Stack, G.; McIlvain, H. B.;
Greenfield, A.; Kowal, D.; Petroski, R.; Carrick, T. Br.
J. Pharmacol. 2003, 140, 839.
7. Shimamoto, K.; Sakai, R.; Takaoka, K.; Yumoto, N.;
Nakajima, T.; Amara, S. G.; Shigeri, Y. Mol. Pharmacol.
2004, 65, 1008.
8. Prestidge, R. L.; Harding, D. R. K.; Battersby, J. E.;
Hancock, W. S. J. Org. Chem. 1975, 40, 3287.
9. Bark, S. J.; Schmid, S.; Hahn, K. M. J. Am. Chem. Soc.
2000, 122, 3567.
10. Bellamy, F. D.; Ou, K. Tetrahedron Lett. 1984, 25, 839.
11. Ames, D. E.; Opalko, A. Synthesis 1983, 3, 234.
12. Klein, M.; Boche, G. Synthesis 1999, 7, 1246.
13. Uptake in stable cell lines. Stable HEK cell lines
expressing each of the human glutamate transporter
subtypes EAAT-1–3 were plated at 50,000 cells/well in
96-well culture plates the day, prior to the measurement
of glutamate uptake. Uptake assays were performed in
DulbeccoÕs phosphate-buffered saline (D-PBS) in the
presence of 1 lM glutamate and 0.2 lCi/ml L-[3H]gluta-
mate in a final volume of 100 ll for 20 min at room
temperature. Assays were stopped by aspiration fol-
lowed by two ice-cold D-PBS washes and [3H] accumu-
lation in the wells determined by liquid scintillation
counting. Uptake was linear for incubation times up to
30 min, thus data were analyzed as true rates. Non-
specific uptake was corrected for by performing all
experiments in the absence and presence of sodium.
Sodium-independent uptake accounted for 610% of
total uptake and was subtracted prior to any further
data calculation.
14. Dunlop, J. H.; McIlvain, B.; Carrick, T.; Jow, B.; Lu, Q.;
Kowal, D.; Lin, S.; Greenfield, A.; Grosanu, C.; Fan, K.;
Petroski, R.; Williams, J.; Foster, A.; Butera, J. Mol.
Pharmacol. Fast Forward. Published on line 07/13/2005 as
doi:10.1124/mol.105.012005.
In conclusion, we have designed and characterized a novel
series of EAAT-blockers, exemplified by 16 (N4-[4-(2-