Aryl- and heteroaryl-substituted phenylalanines as AMPA receptor ligands

Article abstract of DOI:10.1111/cbdd.13048

A series of racemic unnatural amino acids was rationally designed on the basis of recently published X-ray structures of the GluA2 LBD with bound phenylalanine-based antagonists. Twelve new diaryl- or aryl/heteroaryl-substituted phenylalanine derivatives

Full text of DOI:10.1111/cbdd.13048

DR. EWA SZYMANSKA (Orcid ID : 0000-0003-3025-7002)
Article type : Research Article
Aryl- and heteroaryl-substituted phenylalanines as AMPA receptor ligands
a,
a
b
b
Ewa Szymańska, * Paulina Chałupnik, Tommy N. Johansen, Birgitte Nielsen, Ana Maria
b
b
c
a
Cuñado Moral, Darryl S. Pickering, Anna Więckowska, Katarzyna Kieć-Kononowicz
a
Department of Technology and Biotechnology of Drugs, Jagiellonian University Medical College,
Medyczna 9, PL 30-688 Kraków, Poland
b
Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences,
University of Copenhagen, 2100 Copenhagen Ø, Denmark
c
Department of Medicinal Chemistry, Jagiellonian University Medical College,
Medyczna 9, PL 30-688 Kraków, Poland
Keywords
ionotropic glutamate receptors, AMPA, GluA2, competitive antagonist, molecular modeling
*
Corresponding author: Ewa Szymanska
tel.: +48 012 620 55 80; fax: +48 012 620 55 96, e-mail: ewa.szymanska@uj.edu.pl
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/cbdd.13048
This article is protected by copyright. All rights reserved.

Abstract
A series of racemic unnatural amino acids was rationally designed on the basis of recently
published X-ray structures of the GluA2 LBD with bound phenylalanine-based antagonists.
Twelve new diaryl- or aryl/heteroaryl-substituted phenylalanine derivatives were synthesized
and evaluated in vitro in radioligand binding assays at native rat ionotropic glutamate
receptors. The most interesting compound in this series, (RS)-2-amino-3-(3'-hydroxy-5-(1H-
pyrazol-4-yl)-[1,1'-biphenyl]-3-yl)propanoic acid 7e, showed the binding affinity of 4.6 M
for native AMPA receptors and over four-fold lower affinity for kainic acid receptors.
Furthermore, 7e was evaluated at recombinant homomeric rat GluA2 and GluA3 receptors.
Recently reported X-ray structures 5CBR and 5CBS, representing two distinct antagonist
binding modes, were used as templates for molecular docking of the synthesized series.
Binding data supported with molecular modeling confirmed that aryl/heteroaryl-substituted
phenylalanine analogues effectively bind to AMPA receptors with low micromolar affinity
and high selectivity over native NMDA and kainate receptors. These properties make 7e a
promising lead for the further development of new AMPA receptor ligands.
Introduction
Glutamate (Glu) is the principal excitatory neurotransmitter in the mammalian central
nervous system and exerts its physiological effects through two receptor classes:
metabotropic receptors and ligand gated ion channels (iGluRs). The latter class of receptors is
divided based on agonist binding affinities into three functionally distinct subclasses: N-
methyl-D-aspartic acid (NMDA), -2-amino-3-(3-hydroxy-5-methyl-4-isoxazolyl)propionic
acid (AMPA) and kainic acid (KA) receptors. Similarly to other members of this class,
AMPA receptors are cation-permeable tetrameric (dimer of dimers) ion channels composed
of potential combinations of GluA1, GluA2, GluA3 or GluA4 subunits. They mediate the
This article is protected by copyright. All rights reserved.

most of fast excitatory neurotransmission in the brain and are involved in physiological
processes important for memory and learning, motor control, as well as neural plasticity and
development. On the other hand, the excessive activation of AMPA receptors can be linked to
[
1,2]
neurodegenerative conditions such as Alzheimer’s disease,
amyotrophic lateral sclerosis
[
1,3]
[2,4]
(ALS)
and parkinsonism.
Furthermore, AMPA receptors have also been suggested to
[
3,5]
[6,7]
play an important role in neuropathic pain , depression,
as well as the seizure spread
[
1,2,8,9]
and neuronal damage associated with epilepsy.
In these diseases, blocking AMPA
receptor over-activation is a proposed strategy for regulation of glutamatergic
neurotransmission.
In last two decades, many antagonists were developed that act either competitively at the (S)-
Glu recognition site or noncompetitively at the AMPA ion channel site. Both types of
AMPA/KA receptor antagonists have been demonstrated to possess a broad spectrum of
[
1,8-
anticonvulsant activity in animal models of both focal and generalized epileptic seizures.
1
1]
They also appeared to be effective in preventing neuronal loss in animal models of global
[
1,12]
[3]
ischemia
and were found to produce pain relief after oral surgery and in migraine
[
13]
patients. Recently, first-in-class noncompetitive AMPA receptor inhibitor, perampanel, has
[
14]
been approved for the adjunctive treatment of partial-onset seizures in the US and Europe.
In our search for new competitive AMPA receptor antagonists, we performed a rational
design on the basis of known ligand binding modes within the GluA2 binding site. Known
competitive AMPA/KA receptor antagonists belong to a few structurally distinct classes. One
of them contains expanded versions of α-amino acid agonists - compounds built of an α-
amino acid part linked through a heterocyclic ring system with a distal, negatively charged
[
15,16]
[17]
group.
analysis of interactions at the available X-ray structures of the GluA2 ligand binding core
GluA2 LBD) co-crystallized with such antagonists revealed that the heterocyclic ring of the
Isoxazol-based (S)-ATPO (1, Fig. 1) is a typical example of this category. The
(
This article is protected by copyright. All rights reserved.

ligand acts mainly as a rigid linker between the α-amino acid fragment and the distal acidic
group and stabilizes the open conformation of the receptor (which prevents opening of the ion
channel). This part of molecule could be replaced by other rigid spacers, for instance a
benzene ring, offering possibilities to introduce substituents exploiting various regions of the
binding pocket and potentially increasing the affinity for the receptor. Recently we have
reported a class of unnatural amino acids built on the phenylalanine scaffold as competitive
[
15,18-21]
AMPA/KA receptor antagonists.
Structure-activity relationships have been developed
for two alternate directions of structural modifications, including monocyclic and bicyclic
phenylalanines, and within both series the potent and selective antagonists of AMPA
receptors and/or KA receptors have been identified (Fig. 1A).
The superposition of the X-ray structures of GluA2 binding site bound to antagonists (S)-
ATPO
(1,
PDB
code:
1N0T),
(S)-2-amino-3-(2-(2-carboxyethyl)-5-chloro-4-
nitrophenyl)propionic acid ((S)-2, PDB code: 3TZA), and (S)-2-amino-3-(3,4-dichloro-5-(5-
hydroxypyridin-3-yl)phenyl)propanoic acid ((S)-3b, PDB code: 5CBR) shows that the tert-
butyl group of (S)-ATPO occupies the same cavity of the binding site as the chlorine atoms in
the 5-position of the phenylalanine ring in the structures of (S)-2 and (S)-3b. Molecular
docking experiments suggest that the cavity is deep enough to accommodate not just chlorine,
but also larger substituents with either linear or flat shapes (like for example an aromatic
function, Fig. 1B).
This article is protected by copyright. All rights reserved.

Figure 1. (A) Structures of selected -amino acid-based competitive antagonists of AMPA and kainate receptors.
B) General structure of target compounds.
(
Thus, the main aim of the present work was to test the influence of a second large substituent
in the 5-position of the phenylalanine ring on AMPA receptor affinity (Fig. 1B). To achieve
this goal, we have designed and synthesized a new series of 3-aryl-5-aryl/heteroaryl-
substituted phenylalanines. All target amino acids were pharmacologically characterized by
radioligand binding at native AMPA, KA and NMDA receptors in rat brain membranes.
Additionally, the most potent compound was also examined at homomeric recombinant rat
GluA2 and GluA3 receptors, expressed in Sf9 insect cell membranes. A molecular modeling
study was performed in attempt to explain the pharmacological data on the synthesized
compounds observed.
Methods and materials
Chemistry
All reagents were purchased from commercial suppliers and were used without further
purification. Melting points (mp) were determined on a MEL-TEMP II (LD Inc., USA)
1
melting point apparatus and are uncorrected. H NMR spectra were recorded at 300 MHz and
1
3
C-NMR spectra were recorded at 75 Hz using a Varian-Mercury-VX 300 MHz PFG
6
spectrometer, with a solvent signal as an internal standard. Dimethyl-d -sulfoxide (DMSO-d )
6
was used as a solvent. Chemical shifts (δ) were expressed in parts per million (ppm). Mass
spectra (LC/MS) were performed on Waters TQ Detector mass spectrometer coupled to a
Waters Acquity Ultra Performance Liquid Chromatography (UPLC). The UPLC purity of all
final compounds was determined (%). Elemental analyses (C, H, N) were performed on an
Elemental Analyser Vario El III (Hanau, Germany). For column chromatography
purification, a CombiFlash Rf+ apparatus (Teledyne Isco Inc., USA) was used and mixtures
This article is protected by copyright. All rights reserved.

of petroleum ether/ethyl acetate or dichloromethane/ethyl acetate were applied as the mobile
phase. The target racemic amino acids were purified by liquid chromatography using RP-18
silica gel (LiChroprep 40-63 μm; Merck) and a mixture of methanol and 15 mM acetic acid
as the mobile phase.
General procedure for preparation of (RS)-2-amino-3-(3'-hydroxy-5-aryl-[1,1'-biphenyl]-3-
yl)propanoic acids (7a-7l)
A mixture of diethyl 2-acetamido-2-substituted malonate 6a-l (0.6 mmol), 36% HCl (3 mL),
acetic acid (1.5 mL) and water (1.5 mL) was refluxed for 24 h. The mixture of acids was
reduced in vacuo; to the solid residue water (10 mL) was added and pH was adjusted to ~5
with 5% NaOH solution. The mixture was evaporated under reduced pressure. The raw
product was purified by a reverse-phase flash column chromatography using R-18 silica gel
and a gradient mobile phase consisting of 15 mM aqueous acetic acid and methanol to give
white or off-white crystals of 7a-7l. The final amino acids were obtained in their zwitterionic
form as racemates.
(
RS)-2-Amino-3-(3'-hydroxy-5-(3-thienyl)-[1,1'-biphenyl]-3-yl)propanoic acid (7a)
Yield: 38%, mp 259-261 °C. Anal. for C19 S: Calc.: C, 67.24; H, 5.05; N, 4.13. Found:
C, 67.52; H, 5.12; N, 4.03. LC/MS: purity 100%, m/z 340.13 [M +]. H NMR (DMSO-d
H
17NO
3
+
1
6
) δ:
.97 (dd, J=14.4, 8.5 Hz, 1H, CH), 3.27 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.56 (dd, J=8.2,
.1 Hz, 1H, CH), 6.74‒6.80 (m, 1H, Ar), 7.12 (d, J=7.8 Hz, 1H, Ar), 7.17 (t, J=1.8 Hz, 1H,
2
4
Ar), 7.23 (t, J=7.8 Hz, 1H, Ar), 7.43 (s, 1H, Ar), 7.56‒7.64 (m, 3H, Ar), 7.73 (s, 1H, Ar),
.90 (dd, J=2.9, 1.8 Hz, 1H, Ar).
RS)-2-Amino-3-(3'-hydroxy-5-(2-thienyl)-[1,1'-biphenyl]-3-yl)propanoic acid (7b)
Yield: 42%, mp 248-249 °C. Anal. for C19 S: Calc.: C, 67.24; H, 5.05; N, 4.13. Found:
7
(
H17NO
3
+
1
6
C, 66.97; H, 5.09; N, 4.08. LC/MS: purity 100%, m/z 340.13 [M +1]. H NMR (DMSO-d ) δ:
This article is protected by copyright. All rights reserved.

2
4
1
.96 (dd, J=14.1, 8.2 Hz, 1H, CH), 3.24 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.52 (dd, J=7.9,
.4 Hz, 1H, CH), 6.77 (dd, J=7.9, 1.5 Hz, 1H, Ar), 7.07‒7.16 (m, 3H, Ar), 7.24 (t, J=7.6 Hz,
H, Ar), 7.45 (s, 1H, Ar), 7.51‒7.58 (m, 3H, Ar), 7.65 (s, 1H, Ar).
(
RS)-3-(5-(5-Acetyl-2-thienyl)-3'-hydroxy-[1,1'-biphenyl]-3-yl)-2-aminopropanoic acid (7c)
Yield: 49%, mp 189-190 °C. Anal. for C21 S • 1.5 H O: Calc.: C, 61.75; H, 5.43; N,
.43. Found: C, 61.99; H, 5.60; N, 3.25. LC/MS: purity 100%, m/z 382.13 [M +1]. H NMR
DMSO-d ) δ: 2.52 (s, 3H, COCH ), 3.00 (dd, J=14.4, 7.9 Hz, 1H, CH), 3.25 (dd, J=14.1,
.1 Hz, 1H, CH), 3.56 (dd, J=7.6, 4.7 Hz, 1H, CH), 6.78 (dd, J=7.9, 1.47 Hz, 1H, Ar), 7.12
H19NO
4
2
+
1
3
(
6
3
4
(
d, J=8.0 Hz, 1H, Ar), 7.16 (d, J=1.8 Hz, 1H, Ar), 7.24 (t, J=7.6 Hz, 1H, Ar), 7.55 (s, 1H,
Ar), 7.63 (s, 1H, Ar), 7.70 (d, J=4.1 Hz, 1H, Ar), 7.76 (s, 1H, Ar), 7.93 (d, J=4.1 Hz, 1H, Ar).
RS)-5-(5-(2-Amino-2-carboxyethyl)-3'-hydroxy-[1,1'-biphenyl]-3-yl)thiophene-2-carboxylic
(
acid (7d)
5
Yield: 24%, mp 241-242 °C. Anal. for C20H17NO S: Calc.: C, 62.65; H, 4.47; N, 3.65. Found:
+
1
C, 62.38; H, 4.49; N, 3.53. LC/MS: purity 100%, m/z 384.16 [M +1]. H NMR (DMSO-d
6
) δ:
2
4
7
.95 (dd, J=14.1, 8.2 Hz, 1H, CH), 3.24 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.50 (dd, J=8.2,
.7 Hz, 1H, CH), 6.75‒6.80 (m, 1H, Ar), 7.08‒7.16 (m, 3H, Ar), 7.24 (t, J=7.8 Hz, 1H, Ar),
.44 (s, 1H, Ar), 7.51‒7.54 (m, 1H, Ar), 7.55‒7.57 (m, 1H, Ar), 7.65 (t, J=1.5 Hz, 1H, Ar).
(RS)-2-Amino-3-(3'-hydroxy-5-(1H-pyrazol-4-yl)-[1,1'-biphenyl]-3-yl)propanoic acid (7e)
Yield: 44%, mp 265-267 °C. Anal. for C18
H
17
N
3
O
3
• 1.5 H
2
O: Calc.: C, 61.70; H, 5.75; N,
+
1
1.99. Found: C, 61.99; H, 5.70; N, 11.77. LC/MS: purity 100%, m/z 324.18 [M +1].
1
6
H NMR (DMSO-d ) δ: 2.91 (dd, J=14.4, 8.5 Hz, 1H, CH), 3.25 (dd, J=14.1, 3.5 Hz, 1H,
CH), 3.52 (dd, J=8.5, 3.8 Hz, 1H, CH), 6.76 (dd, J=7.9, 1.5 Hz, 1H, Ar), 7.07‒7.15 (m, 2H,
Ar), 7.22 (t, J=7.6 Hz, 1H, Ar), 7.32 (s, 1H, Ar), 7.48 (s, 1H, Ar), 7.64 (s, 1H, Ar), 8.08 (br.s.,
2
H, Ar).
This article is protected by copyright. All rights reserved.

(
RS)-2-Amino-3-(3-hydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7f)
Yield: 62%, mp 247-249 °C. Anal. for C21 : Calc.: C, 75.66; H, 5.74; N, 4.20. Found:
C, 75.55; H, 6.11; N, 4.20. LC/MS: purity 99%, m/z 334.22 [M +1]. H NMR (DMSO-d
H19NO
3
+
1
6
) δ:
3
4
7
.02 (dd, J=14.1, 7.6 Hz, 1H, CH), 3.26 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.54 (dd, J=7.9,
.4 Hz, 1H, CH), 6.74‒6.80 (m, 1H, Ar), 7.11‒7.17 (m, 2H, Ar), 7.20‒7.27 (m, 1H, Ar),
.32‒7.39 (m, 1H, Ar), 7.41‒7.53 (m, 4H, Ar), 7.65 (s, 1H, Ar), 7.69‒7.75 (m, 2H, Ar).
(RS)-2-Amino-3-(2,3''-dihydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7g)
Yield: 37%, mp 250-252 °C. Anal. for C21
H19NO
4
• ½ CH
3
COOH: Calc.: C, 69.64; H, 5.58;
+
N, 3.69. Found: C, 69.71; H, 5.31; N, 3.84. LC/MS: purity 100%, m/z 350.17 [M +1].
1
6
H NMR (DMSO-d ) δ: 2.99 (dd, J=14.1, 7.6 Hz, 1H, CH), 3.23 (dd, J=14.1, 4.1 Hz, 1H,
CH), 3.52 (dd, J=7.9, 4.4 Hz, 1H, CH), 6.75 (dd, J=8.2, 1.8 Hz, 1H, Ar), 6.86 (td, J=7.2,
.2 Hz, 1H, Ar), 6.96 (dd, J=8.2, 1.2 Hz, 1H, Ar), 7.05 (d, J=7.6 Hz, 1H, Ar), 7.09 (t,
1
J=2.1 Hz, 1H, Ar), 7.14 (td, J=7.8, 1.6 Hz, 1H, Ar), 7.21 (t, J=7.7 Hz, 1H, Ar), 7.32 (dd,
J=7.6, 1.8 Hz, 1H, Ar), 7.38 (s, 1H, Ar), 7.43 (s, 1H, Ar), 7.61 (s, 1H, Ar).
(
RS)-2-Amino-3-(3,3''-dihydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7h)
Yield: 54%, mp 228-230 °C. Anal. for C21 • 1.5 H O: Calc.: C, 67.01; H, 5.89; N,
.72. Found: C, 67.02; H, 5.83; N, 3.61. LC/MS: purity 100%, m/z 350.17 [M +1]. H NMR
DMSO-d ) δ: 3.03 (dd, J=14.1, 7.6 Hz, 1H, CH), 3.26 (dd, J=14.4, 3.8 Hz, 1H, CH), 3.58
dd, J=7.9, 4.4 Hz, 1H, CH), 6.76 (dd, J=8.8, 2.3 Hz, 2H, Ar), 7.11 (d, J=7.6 Hz, 2H, Ar),
.15 (t, J=1.5 Hz, 2H, Ar), 7.23 (t, J=7.8 Hz, 2H, Ar), 7.48 (d, J=1.2 Hz, 2H, Ar), 7.58 (d,
J=1.8 Hz, 1H, Ar).
RS)-2-Amino-3-(3,4''-dihydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7i)
Yield: 51%, mp 252-254 °C. Anal. for C21 : Calc.: C, 72.19; H, 5.48; N, 4.01. Found:
C, 72.39; H, 5.66; N, 3.94. LC/MS: purity 100%, m/z 350.17 [M +1]. H NMR (DMSO-d
.01 (dd, J=14.1, 7.6 Hz, 1H, CH), 3.24 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.56 (dd, J=7.6,
H
19NO
4
2
+
1
3
(
(
6
7
(
H19NO
4
+
1
6
) δ:
3
This article is protected by copyright. All rights reserved.

4
2
7
.1 Hz, 1H, CH), 6.76 (dd, J=8.2, 1.8 Hz, 1H, Ar), 6.84 (d, J=8.3 Hz, 2H, Ar), 7.08‒7.16 (m,
H, Ar), 7.22 (t, J=7.7 Hz, 1H, Ar), 7.42 (d, J=8.2 Hz, 2H, Ar), 7.52 (d, J=8.8 Hz, 2H, Ar),
.56 (s, 1H, Ar).
(
RS)-5'-(2-Amino-2-carboxyethyl)-3''-hydroxy-[1,1':3',1''-terphenyl]-3-carboxylic acid (7j)
Yield: 41%, mp 240-241 °C. Anal. for C22 : Calc.: C, 70.02; H, 5.07; N, 3.71. Found:
C, 69.99; H, 5.28; N, 3.83. LC/MS: purity 97%, m/z 378.15 [M +1]. H NMR (DMSO-d
H19NO
5
+
1
6
) δ:
3
4
1
.06 (dd, J=14.1, 7.6 Hz, 1H, CH), 3.29 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.65 (dd, J=7.6,
.1 Hz, 1H, CH), 6.77 (dd, J=7.6, 1.2 Hz, 1H, Ar), 7.11‒7.19 (m, 2H, Ar), 7.23 (t, J=7.6 Hz,
H, Ar), 7.52‒7.60 (m, 3H, Ar), 7.68 (s, 1H, Ar), 7.94 (t, J=7.8 Hz, 2H, Ar), 8.24 (s, 1H, Ar).
(
RS)-2-Amino-3-(2,2''-dihydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7k)
Yield: 42% mp 250-251 °C. Anal. for C21 • H O: Calc.: C, 68.65; H, 5.76; N, 3.81.
Found: C, 68.44; H, 5.52; N, 3.93. LC/MS: purity 100%, m/z 350.17 [M +1]. H NMR
H19NO
4
2
+
1
(DMSO-d
6
) δ: 2.94 (dd, J=14.1, 8.2 Hz, 1H, CH), 3.21 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.45
(dd, J=7.9, 4.4 Hz, 1H, CH), 6.84 (td, J=7.4, 1.0 Hz, 2H, Ar), 6.95 (dd, J=8.2, 1.2 Hz, 2H,
Ar), 7.12 (td, J=7.6, 1.4 Hz, 2H, Ar), 7.28 (dd, J=7.6, 1.8 Hz, 2H, Ar), 7.33 (d, J=1.8 Hz, 2H,
Ar), 7.60 (t, J=1.2 Hz, 1H, Ar).
(
RS)-2-Amino-3-(4,4''-dihydroxy-[1,1':3',1''-terphenyl]-5'-yl)propanoic acid (7l)
Yield: 56% mp 256-257 °C. Anal. for C21 • ⅔ H O: Calc.: C, 69.79; H, 5.67; N, 3.88.
Found: C, 69.84; H, 5.69; N, 3.85. LC/MS: purity 100%, m/z 350.17 [M +1]. H NMR
H19NO
4
2
+
1
(DMSO-d
6
) δ: 2.99 (dd, J=14.1, 8.2 Hz, 1H, CH), 3.22 (dd, J=14.1, 4.1 Hz, 1H, CH), 3.52
(dd, J=7.6, 4.7 Hz, 1H, CH), 6.85 (d, J=8.8 Hz, 4H, Ar), 7.36 (d, J=1.8 Hz, 2H, Ar), 7.51 (d,
J=8.6 Hz, 5H, Ar).
This article is protected by copyright. All rights reserved.

Pharmacological methods
Native rat brain AMPA receptor binding. Rat brain membrane preparations used in native
receptor binding experiments were prepared according to the method described by Ransom
[22]
[23]
[24]
[25]
and Stec. Affinity for AMPA KA and NMDA receptor sites was determined using
3
3
3
as radioligands 5 nM  HAMPA, 5 nM  HKA, and 2 nM [ H](RS)-(E)-2-amino-4-
3
phosphonomethyl-3-heptenoic acid ( HCGP39653) (all from Perkin Elmer, Waltham, MA)
[
26]
with some modifications as previously described.
Recombinant rat AMPA receptor binding. GluA2(R)
o
o
and GluA3 were inserted into
recombinant baculoviruses, receptors expressed by the infection of Sf9 insect cells and
3
infected Sf9 cell membranes utilized for radioligand binding assays using [ H]AMPA
[
27]
radioligand as previously detailed.
HPLC determination of the relative lipophilicity for the selected amino acids
RP-HLPC. HPLC chromatography was performed on Waters liquid chromatography system
with 486 Tunable Absorbance Detector and two 515 gradient pumps. Sample solutions in
methanol (1 mg/mL, 5 µL) were injected on a LiChrospher® 100 RP18, 5 μm, 100 × 4.6 mm
(Merck) column. All measurements were done in duplicates, at λ = 254 nm and at ambient
temperature. The mixtures of acetonitrile (HPLC grade, Sigma Aldrich) and water
(
deionized) were used as mobile phases with acetonitrile content between 70% and 30% (v/v)
pH = 7) in 5% or 10% increments and the flow rate of 1 mL/min. For each compound the
(
retention times (t
isocratic capacity factors, log k, were calculated according to the equation: log k = log[(t
)/t
Log k
concentration in mobile phase (CACN) according to equation: log k = aCACN + log k
Supporting Information, Figure S1, Table S1, page S76).
This article is protected by copyright. All rights reserved.
r 0
) and the dead times (t ) at each mobile phase were determined. The
r
−
t
0
0
]
w
values were calculated by extrapolation of log k values to the zero acetonitrile
[
28,29]
w
(

Prediction of physico-chemical properties
Major protonation form at a specified pH and isoelectric point values for the final amino
[
30]
acids 7a, 7e and 7i were predicted by ChemAxon software.
The octanol-water partition
coefficients for single protonation state of a compound, as well as pH-dependent logD values
[
31]
were estimated, based on a modified version of the method of Viswanadhan, implemented
in the software. The logD values calculated for the neutral species in their isoelectric point
w
were correlated with log k values, obtained by HPLC experiment (Supporting Information,
Table S2, Figure S2, page S77).
To locate the most acidic and basic sites in a molecules of selected amino acids, the pKa
[
30]
values of all proton receiving or donating groups were estimated by ChemAxon software
Supporting Information, Table S3, page S77).
Molecular modeling
The ligand database and protein templates as well as docking calculations were carried out
(
[32,33]
using Schrödinger molecular modeling environment.
The possible protonation and tautomeric states as well as all stereoisomers for the final amino
acids were generated with LigPrep module. The optimization of ligands structure was
performed using a conformational search module as implemented in MacroModel (Monte
Carlo multiple minimization method, OPLS2005 force field in water, minimization with
Truncated Newton Conjugate Gradient option until the root means square of conjugate
-1
-1
gradient was below 0.05 kJmol Å ). The lowest energy conformer was kept for each
ligand.
The crystal structures of the ligand-binding domain of GluA2 subunit in complex with bound
antagonists (S)-2-amino-3-(3,4-dichloro-5-(5-hydroxypyridin-3-yl)phenyl)propanoic acid
(
(S)-3b, PDB code: 5CBR) and (R)-2-amino-3-(3'-hydroxybiphenyl-3-yl)propanoic acid ((R)-
[21]
3
a, PDB code: 5CBS) were used as templates. Prior to docking, the protein structures were
This article is protected by copyright. All rights reserved.

processed with the Protein Preparation Wizard option as a part of Schrodinger package.
Water molecules and other non-ligand small molecules were removed from the protein. The
protonation, orientation and tautomeric states of Asp, Glu, Arg, Lys and His were adjusted to
match a pH of 7.4 and the hydrogen bond network was assigned using the PROPKA option.
The structure of the protein-ligand complex was minimized using the OPLS2005 force field
with the convergence of heavy atoms to RMSD of 0.3 Å.
The ligand binding cavity was confined to a box with the crystal ligand in the center (size of
the box set automatically), according to the receptor grid generation procedure (OPLS2005
force field). The studies of docking to the rigid receptor binding pocket were performed using
Glide program, with Glide extra docking precision (XP) mode. All poses were subjected to
post-docking minimization with strain correction terms option. Per-residue interaction scores
for the residues within 6Å of a grid center were calculated. The final poses were analyzed
according to the XP docking score values obtained.
For the graphic presentation of selected structures with the highest docking scores, PyMOL
[34]
software was used.
Results and discussion
Chemistry
The general procedure for the synthesis of the target amino acids is outlined in Scheme 1.
The treatment of commercially available 3,5-dibromotoluene with N-bromosuccinimide
(NBS) under free radical conditions, followed by the nucleophilic substitution with the
sodium salt of diethyl acetamidomalonate, performed in one step without the isolation of
benzyl bromides, led to the dibromo derivative 4. In the next step, a short optimization of
Suzuki cross-coupling conditions was needed to find the effective way for a selective
substitution of only one bromine atom at the phenylalanine ring of 4. Various palladium
This article is protected by copyright. All rights reserved.

sources, solvents and bases were examined. Eventually, use of palladium acetate,
triphenylphosphine and tripotassium ortophosphate as the base as well as toluene with a small
addition of ethanol (50 : 1) as a solvent turned out to produce the most effective conditions,
under which the yield of mono-substituted product was optimal. The resulting key biphenyl
intermediate (5) was reacted in the next coupling step with a number of aryl boronic acids in
conditions, optimized previously for diethyl 2-acetamido-2-(3-halogenobenzyl)malonate
[
15,35]
derivatives.
palladium(II) dichloride and triethylamine to obtain the tricyclic derivatives of phenylalanine
. In these conditions, the symmetrical compounds 6h, 6k and 6l were prepared directly from
The reaction was performed in the presence of bis(triphenylphosphine)
6
dibromo derivative 4 as a result of Suzuki coupling with 2.3 equivalents of the respective
phenylboronic acid. The final amino acids were prepared by reflux of 6a-l in acidic medium
3
(HCl : CH COOH: water), and purified by pH neutralization followed by reverse-phase
liquid chromatography. All the amino acids were obtained as racemic mixtures in the form of
zwitterions.
Scheme 1. NBS, benzoyl peroxide, CCl
hydroxyphenylboronic acid, Pd(OAc) , PPh
acid, PdCl (PPh , triethylamine, DME/water, 50 °C; (v) HCl/CH
3
4
, reflux; (ii) NaH, diethyl acetamidomalonate, DMF, r.t.; (iii) 3-
, K PO , toluene/ethanol, reflux; (iv) aryl- or heteroarylboronic
COOH, reflux.
2
3
3
4
2
3
)
2
This article is protected by copyright. All rights reserved.

Binding Pharmacology
Most of the target phenylalanine derivatives display a profile of affinity for native iGluRs
comparable with (S)-ATPO and (RS)-3a (Table 1). In the homogenate binding studies, the
compounds expressed negligible affinity for NMDA and KA receptors (K > 100 M, for
i
NMDA receptors data not shown), except for 7e that showed measurable binding to native
kainate receptors (Table 1).
The commonly observed AMPA receptor affinity within the series of amino acids
synthesized (Table 1) is in the micromolar range. In most cases, the introduction of the
additional aromatic function into the 5-position of the phenylalanine ring is accepted by the
receptor but does not induce a significant change in the affinity of compounds, compared to
the parent 3a. The rank order of AMPA receptor potency among the compounds containing
3
1
-hydroxyphenyl substituent in the 3-position is:
H-pyrazol-4-yl > H ~ 3-thienyl ~ 5-carboxy-2-thienyl > 2-thienyl ~ 4-hydroxyphenyl ~ 3-
hydroxyphenyl > 3-carboxyphenyl > 5-acetyl-2-thienyl > phenyl > 2-hydroxyphenyl.
The remaining symmetrical meta-terphenyl-based amino acids (7k and 7l) show reduced or
unmeasurable binding affinity compared to 7h.
The pyrazolyl group was found to be beneficial for AMPA receptor binding. The most potent
AMPA receptor ligand within the series, (RS)-3-(5-(1H-pyrazol-4-yl)-[1,1'-biphenyl]-3-yl)-2-
aminopropanoic acid (7e), displays an affinity determined as 4.6 μM (two-fold increase
compared to the parent 3a). At the same time, introduction of a pyrazole ring into the 5-
position induces micromolar affinity of 7e for native kainate receptors, resulting in the
decrease of AMPA/KA receptors selectivity in comparison with 3a.
This article is protected by copyright. All rights reserved.

3
3
a
Table 1. Binding pharmacology at native rat brain membranes using [ H]AMPA and  HKA as radioligands
1
2
3
3
Compound
R
R
 HAMPA
 HKA
K
i
(M)
K
i
(M)
b
b
(
S)-ATPO
RS)-3a
16
> 100
c
c
(
H
3-OH
3-OH
9.9
> 100
7
a
9.5
> 100
> 100
> 100
> 100









5.02±0.03
7
b
3-OH
3-OH
3-OH
3-OH
3-OH
3-OH
3-OH
3-OH
13
4.89±0.02
7
c
22
4.67±0.05
7
d
11
4.98±0.06
7
e
4.6
21
5.37±0.08
4.67±0.01
7
f
28
> 100
> 100
> 100
> 100
4.57±0.08
7
g
36
4.45±0.02
7
h
14
4.86±0.03
7
i
13
4.90±0.08
This article is protected by copyright. All rights reserved.

7
j
3-OH
17
> 100
4.78±0.06
7
k
2-OH
4-OH
> 100
> 100
> 100
7
l
19
4.72±0.05
a
i
Data are given as mean of at least three separate experiments conducted in triplicate. Mean pK +/- SEM (M)
b
[17]
c
[21]
are indicated in brackets. IC50 values (from Møller et al. ). IC50 values (from Szymanska et al.
)
In order to determine the influence of the pyrazolyl moiety in the 5-position of the
phenylalanine ring on the subtype-selectivity among the individual AMPA receptors, the
most potent amino acid 7e has been subsequently characterized at the cloned rat GluA2 and
GluA3 receptors. The compound presents micromolar binding affinity to both homomeric
receptors with only two-fold preference for GluA2 in relation to GluA3 (K
and K = 38.4 ± 6.4 μM, respectively).
Physicochemical properties of selected amino acids
i
= 14.7 ± 3.4 μM
i
To evaluate the lipophilicity of the synthesized amino acids, a well-established technique of
reversed-phase high performance liquid chromatography (RP-HPLC) was applied as an
[
36]
alternative to the classical shake-flask method. As this technique allows to determine the
relative lipophilicity of the investigated compounds, phenylalanine was used as a reference.
The measurements were performed for amino acids 7a, 7e and 7i.
Additionally, the obtained log k
w
values were compared to logD values calculated by
[30]
ChemAxon software
for the neutral species at their isoelectric points. High linear
correlation between experimental and predicted lipophilicity was observed (correlation
2
coefficient R = 0.989, Supporting Information, FigureS2, page S77). As can be expected, all
This article is protected by copyright. All rights reserved.

the synthesized amino acids show lipophilicity higher than phenylalanine, and among them
the 3-thienyl derivative 7a is the most lipophilic (Table 2).
Table 2. Isoelectric points and lipophilicity parameters for the selected amino acids
Compound
pI
logD (pI)
log k
w
Phenylalanine
5.96
5.98
5.82
5.58
-1.18
1.59
0.11
1.50
-0.444
0.525
-0.061
0.396
7
a
7
e
7
i
Molecular modeling
To analyze the structure-activity relationship of the new target compounds, docking studies
were undertaken. The recently reported X-ray crystal structures of phenylalanines (S)-3b and
[
21]
(
R)-3a bound to the GluA2 LBD (PDB code: 5CBR and 5CBS, respectively) were used as
templates for docking of both (S)- and (R)-isomers of 7a-7l. To examine the ligand-protein
non-bonded and polar interactions energy for each final docking pose, the overall terms of the
docking score function were analyzed along with the energy contributions calculated for the
individual residues.
Docking to 5CBR template
The results of docking experiments clearly indicate the pyrazolyl compound (S)-7e as the
highest ranked ligand when using 5CBR X-ray structure as a template. The docking pose of
(
S)-7e (Fig. 2) is in agreement with the canonical binding mode, observed in the published X-
ray structures for most of the α-amino acid-based competitive antagonists bound to the
[
37]
GluA2 LBD, including (S)-3b. The pyrazole ring of (S)-7e is located in the GluA2 LBD in
the similar position as the 5-chlorine atom of (S)-3b, inside the cavity formed by the residues
of D1 and D2 (Fig. 3). This space seems to be large enough to accommodate bigger
substituents with a flat shape, for example an aromatic or a heterocycle ring. Such a situation
This article is protected by copyright. All rights reserved.

has been described for one of the AMPA receptors agonists, BnTetAMPA (PDB code:
[
38]
2
P2A), where a benzyl moiety of the ligand is deeply buried in the cavity of GluA2 LBD
mentioned (Fig. 3). However, none of competitive antagonists co-crystallized with the
GluA2 LBD (or the full-length GluA2 receptor) so far reaches this pocket to such an extent.
Furthermore, the partially polar character of residues in this part of the receptor suggests a
capacity of the hydrogen bond formation with a ligand possessing the suitable feature.
In the docking pose of 7e, the pyrazole ring is only partially buried in the receptor pocket
occupied by the benzyl function of BnTetAMPA (Fig. 3). However, in contrast to
BnTetAMPA agonist, where the benzyl moiety has some flexibility, the benzene ring of 7f-7l
is linked to the phenylalanine core in a rigid way that prevents adopting a more favored
position by pyrazole ring. No hydrogen bonding is observed for the pyrazolyl moiety of 7e,
instead, the heterocycle ring interacts with the surrounding residues with van der Waals and
favorable electrostatic forces.
Compounds containing a thiophene ring in the 5-position of the phenylalanine ring (except
for the acetylthiophene analogue 7c) were successfully docked to the 5CBR template in
docking poses similar to that of (S)-7e, but with lower docking scores, whereas the meta-
terphenyl compounds (7f-7l) are unable to adopt this binding mode. Due to the larger size and
the rigid connection with the phenylalanine core, the benzene ring in the position 5 of 7f-7l
would cause a steric clash with the surrounding residues, especially Tyr405.
This article is protected by copyright. All rights reserved.

Figure 2. The highest ranked docking pose of 7e at the GluA2 LBD (5CBR template). The template antagonist
S)-3b (yellow) and the docking pose of 7e (purple) are shown with sticks; the selected residues of GluA2 LBD
(
are shown with green thin sticks. Residue numbering according to the rat GluA2 sequence P19491 (flop isoform
without signal peptide).
Figure 3. Superposition of the GluA2 LBD X-ray template 5CBR with the docked pose of 7e and the GluA2
LBD X-ray structure in complex with BnTetAMPA agonist (PDB code: 2P2A), aligned with D1 domain. The
docked compound 7e (purple) and BnTetAMPA ligand (yellow) are shown with sticks. Selected residues of
5
CBR protein (green) and 2P2A (orange) are shown with thin sticks. The shape of the binding site (in mesh)
was calculated for BnTetAMPA agonist at the GluA2 LBD (2P2A structure).
This article is protected by copyright. All rights reserved.

Docking to 5CBS template
In contrast to the binding mode, adopted by most of the α-amino acid-based antagonists
inside the binding cleft of GluA2, we have recently discovered and described an alternative
and unique AMPA receptor antagonist binding mode, represented by the X-ray structure of
[
21]
GluA2 LBD with bound (R)-3a (PDB code: 5CBS). In this complex, the α-amino acid part
of (R)-3a interacts with the GluA2 LBD in a similar way as (S)-3b, while the 3-
hydroxyphenyl moiety is positioned in a cavity which is not occupied by other reported
antagonists (Fig. 4). It is important to note that (S)-3b in 5CBR structure induces only partial
LBD domain opening (13.5°), whereas the domain opening in 5CBS (18.8-20.4°) is
comparable with the apo structure of GluA2 LBD, which is more suitable for molecular
[
21]
docking of large ligands. Thus, in an attempt to explain AMPA receptor affinity observed
for meta-terphenyl compounds (7f-7l) we decided to perform a new docking simulation using
5
CBS template.
The highest docking scores were obtained for (R)-enantiomers of terphenyl-based
compounds, namely the 3,4”-dihydroxy-, 3,3”-dihydroxy- but also for 3,2”-dihydroxy-,
3
-hydroxy- and even 2,2”-dihydroxy-terphenyl derivatives (7i, 7h, 7g, 7f and 7k,
respectively, Fig. 4). Slightly lower docking scores are observed for pyrazolyl- and thienyl-
substituted compounds. All the ligands adopt the binding mode presented by (R)-3a in 5CBS
X-ray structure with the 3-hydroxy- or 4-hydroxy group of the distal benzene ring involved in
a hydrogen bond with Asp728. The third aromatic substituent in the 5-position (benzene,
thiophene, pyrazole) is directed outside the binding cleft and easily accommodated between
Arg485 of D1 and Ser654 of D2, forming the -cation interaction with the nitrogen atoms of
Arg485 side chain (interaction not shown on Fig. 4). The polar groups present in this part of
the molecule (hydroxyl groups, carboxyl groups, nitrogen atom within the ring of pyrazole)
act most frequently as hydrogen bond donors or acceptors for Thr482.
This article is protected by copyright. All rights reserved.

In terms of the energy of interaction, however, the main contribution to the docking score
function comes from the phenylalanine core together with the 3-substituent, while the
substituent in the position 5, placed in the specific location in the D1-D2 hinge region, exerts
a rather low input to the overall binding energy. This calculation seems to be in agreement
with the pharmacological results for most of the compounds showing binding affinity for
AMPA receptors comparable with 3a (except for 7k, which displayed a high docking score
but negligible affinity, Table 1). Thus, it appears to be possible that at least some of the
compounds described can adopt a (R)-3a-like binding mode, while for the most potent 7e the
(
S)-3b-like binding mode seems to be more favorable.
Figure 4. Docking results to the GluA2 LBD X-ray template 5CBS. The X-ray ligand (R)-3a (purple) and the
high scored docking poses are shown with sticks. Selected residues of 5CBS protein are shown with green thin
sticks.
Conclusion
In conclusion, our work allowed us to design and perform the synthesis of a new series of
aryl and heteroaryl-substituted phenylalanines acting as AMPA receptor ligands. The
introduction of two aromatic substituents into the 3- and 5-positions of the phenylalanine ring
This article is protected by copyright. All rights reserved.

resulted in compounds with medium micromolar AMPA receptor affinity and high selectivity
versus NMDA and kainate receptors. Within the group investigated, the compound
substituted with a pyrazolyl moiety in the 5-position of the phenylalanine ring (7e) was found
to be the most potent AMPA receptor ligand with four-fold preference for native AMPA over
KA receptors and two-fold preference in affinity for the cloned GluA2 receptors compared to
GluA3. Molecular docking calculations with use of previously described X-ray structures of
GluA2 LBD bound to phenylalanine-based AMPA receptor antagonists allowed to compare
two alternative binding modes of the compounds synthesized. The results suggest that the
pyrazolyl-substituted compound 7e is well accommodated in the GluA2 binding cleft,
adopting the canonical binding mode observed in X-ray structures for most of the known α-
amino acid-based antagonists. From this point of view, in order to optimize the ligand
scaffold, it seems advantageous to introduce a flexible linker bridge between the substituents
in the 3- and 5-positions of the phenylalanine part. On the other hand, for larger terphenyl
amino acids, the alternative (R)-3a-like binding mode seems to be more preferable. The data
coming from this work will support future design and further development of studies on
selective phenylalanine-based AMPA receptor ligands.
Conflict of interest
The authors declare no financial conflict of interest regarding the publication of this
manuscript.
Acknowledgements
The financial support of this research by the National Science Centre Poland (DEC-
2
014/15/B/NZ7/00908) is gratefully acknowledged.
This article is protected by copyright. All rights reserved.

References
[
[
[
1]
2]
3]
P.K.-Y. Chang, D. Verbich, R.A. McKinney, Eur J Neurosci 2012, 35, 1908.
D. Lemoine, R. Jiang, A. Taly, T. Chataigneau, A. Specht, T. Grutter, Chem Rev 2012, 112, 6285.
S.F. Traynelis, L.P. Wollmuth, C.J. McBain, F.S. Menniti, K.M. Vance, K.K. Ogden, K.B. Hansen, H.
Yuan, S.J. Myers, R. Dingledine, Pharmacol Rev 2010, 62, 405.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
4]
5]
6]
7]
8]
9]
10]
11]
12]
13]
14]
15]
16]
17]
K.A. Johnson, P.J. Conn, N.C. M., CNS Neurol Disord Drug Targets 2009, 8, 475.
D. Bleakman, A. Alt, E.S. Nisenbaum, Semin Cell Dev Biol 2006, 17, 592.
D. Bleakman, A. Alt, J.M. Witkin, CNS Neurol Disord Drug Targets 2007, 6, 117.
A. Alt, J.M. Witkin, D. Bleakman, Curr Pharm Des 2005, 11, 1511.
E. Russo, R. Gitto, R. Citraro, A. Chimirri, G. De Sarro, Expert Opin Investig Drugs 2012, 21, 1371.
M.A. Rogawski, Epilepsy Curr 2011, 11, 56.
M. Mula, Expert Opin Emerg Drugs 2013, 18, 87.
R. Citraro, R. Aiello, V. Franco, G. De Sarro, E. Russo, Expert Opin Ther Targets 2014, 18, 319.
D. Catarzi, V. Colotta, F. Varano, Med Res Rev 2007, 27, 239.
K. Chan, A. MaassenVanDenBrink, Drugs 2014, 74, 1165.
K. Traynor, Am J Health Syst Pharm 2012, 69, 2024.
E. Szymanska, D.S. Pickering, B. Nielsen, T.N. Johansen, Bioorg Med Chem 2009, 17, 6390.
P. Stawski, H. Janovjak, D. Trauner, Bioorg Med Chem 2010, 18, 7759.
E.H. Møller, J. Egebjerg, L. Brehm, T.B. Stensbøl, T.N. Johansen, U. Madsen, P. Krogsgaard-Larsen,
Chirality 1999, 11, 752.
18] E. Szymanska, K. Frydenvang, A. Contreras-Sanz, D.S. Pickering, E. Frola, Z. Serafimoska, B.
Nielsen, J.S. Kastrup, T.N. Johansen, J Med Chem 2011, 54, 7289.
19] R. Venskutonyte, K. Frydenvang, E.A. Valades, E. Szymanska, T.N. Johansen, J.S. Kastrup, D.S.
Pickering, ChemMedChem 2012, 7, 1793.
20] N. Sköld, B. Nielsen, J. Olsen, L. Han, L. Olsen, U. Madsen, J.L. Kristensen, D.S. Pickering, T.N.
Johansen, Bioorg Med Chem 2014, 22, 5368.
21] E. Szymanska, K. Frydenvang, D.S. Pickering, C. Krintel, B. Nielsen, A. Kooshki, L.G. Zachariassen,
L. Olsen, J.S. Kastrup, T.N. Johansen, J Med Chem 2016, 59, 448.
[
[
[
[
[22]
[23]
[24]
[25]
R.W. Ransom, N.L. Stec, J Neurochem 1988, 51, 830.
T. Honore, M. Nielsen, Neurosci Lett 1985, 54, 27.
D.J. Braitman, J.T. Coyle, Neuropharmacology 1987, 26, 1247.
M.A. Sills, G. Fagg, M. Pozza, C. Angst, D.E. Brundish, S.D. Hurt, E.J. Wilusz, M. Williams, Eur J
Pharmacol 1991, 192, 19.
26] M.B. Hermit, J.R. Greenwood, B. Nielsen, L. Bunch, C.G. Jørgensen, H.T. Vestergaard, T.B. Stensbøl,
C. Sanchez, P. Krogsgaard-Larsen, U. Madsen, H. Bräuner-Osborne, Eur J Pharmacol 2004, 486, 241.
[
[
1
[
27]
14.
28]
U. Madsen, D.S. Pickering, B. Nielsen, H. Bräuner-Osborne, Neuropharmacology 2005, 49 Suppl. 1,
H. Van de Waterbeemd, M. Kansy, B. Wagner, H. Fischer, Lipophilicity measurement by reversed-
phase high performance liquid chromatography (RP-HPLC), in Lipophilicity in Drug Action and Toxicology,
Pliska V, Testa B, van de Waterbeemd H (eds). VCH: New York, 1996, 73.
[29]
D. Clédat, J.-C. Bollinger, H. Hoja, J. Debord, B. Pénicaut, Quantitative Structure- Activity
Relationships 1998, 17, 1.
[30]
[31]
[32]
MarvinView.NET (v17.3.2700.1567), 201n (2017), ChemAxon (http://www.chemaxon.com).
V.N. Viswanadhan, A.K. Ghose, G.R. Revankar, R.K. Robins, J Chem Inf Comput Sci 1989, 29, 163.
Schrödinger Release 2016-1: Maestro, version 10.5, LigPrep, version 3.7, Epik, version 3.5,
MacroModel, version 11.1, Impact version 7.0, Prime version 4.3, Schrödinger, LLC, New York, NY, 2016.
[
2
[
2
33]
016.
34]
012.
Small-Molecule Drug Discovery Suite 2016-1: Glide, version 7.0, Schrödinger, LLC, New York, NY,
The PyMOL Molecular Graphics System, version 1.5.0.4. Schrödinger, LLC, New York, NY, 2003-
[35]
[36]
[37]
[38]
B.H. Kaae, P. Krogsgaard-Larsen, T.N. Johansen, J Org Chem 2004, 69, 1401.
D. Casoni, A. Kot-Wasik, J. Namieśnik, C. Sârbu, J Chromatogr A 2009, 1216, 2456.
J. Pohlsgaard, K. Frydenvang, U. Madsen, J.S. Kastrup, Neuropharmacology 2011, 60, 135.
S.B. Vogensen, K. Frydenvang, J.R. Greenwood, G. Postorino, B. Nielsen, D.S. Pickering, B. Ebert, U.
Bolcho, J. Egebjerg, M. Gajhede, J.S. Kastrup, T.N. Johansen, R.P. Clausen, P. Krogsgaard-Larsen, J Med
Chem 2007, 50, 2408.
This article is protected by copyright. All rights reserved.

Additional Supporting Information may be found in the online version of this article:
Appendix S1. Procedures, yields and analytical data of compounds 4, 5 and 6a-l
1
13
Appendix S2. H-NMR and C-NMR Spectra of compounds 4, 5 and 6a-l
1
Appendix S3. H-NMR Spectra of compounds 7a-l
Appendix S4. HPLC determination of the relative lipophilicity for the selected amino acids
Appendix S5. Prediction of physico-chemical properties
List of Figures, Schemes and Tables
Figure 1. (A) Structures of selected -amino acid-based competitive antagonists of AMPA
and kainate receptors. (B) General structure of target compounds.
Figure 2. The highest ranked docking pose of 7e at the GluA2 LBD (5CBR template). The
template antagonist (S)-3b (yellow) and the docking pose of 7e (purple) are shown with
sticks; the selected residues of GluA2 LBD are shown with green thin sticks. Residue
numbering according to the rat GluA2 sequence P19491 (flop isoform without signal
peptide).
Figure 3. Superposition of the GluA2 LBD X-ray template 5CBR with the docked pose of 7e
and the GluA2 LBD X-ray structure in complex with BnTetAMPA agonist (PDB code:
2
P2A), aligned with D1 domain. The docked compound 7e (purple) and BnTetAMPA ligand
(yellow) are shown with sticks. Selected residues of 5CBR protein (green) and 2P2A (orange)
are shown with thin sticks. The shape of the binding site (in mesh) was calculated for
BnTetAMPA agonist at the GluA2 LBD (2P2A structure).
This article is protected by copyright. All rights reserved.

Figure 4. Docking results to the GluA2 LBD X-ray template 5CBS. The X-ray ligand (R)-3a
purple) and the high scored docking poses are shown with sticks. Selected residues of 5CBS
protein are shown with green thin sticks.
(
Scheme 1. NBS, benzoyl peroxide, CCl
r.t.; (iii) 3-hydroxyphenylboronic acid, Pd(OAc)
aryl- or heteroarylboronic acid, PdCl (PPh
HCl/CH COOH, reflux.
4
, reflux; (ii) NaH, diethyl acetamidomalonate, DMF,
2
2
, PPh
3 3 4
, K PO , toluene/ethanol, reflux; (iv)
2
3
)
, triethylamine, DME/water, 50 °C; (v)
3
3
3
Table 1. Binding pharmacology at native rat brain membranes using [ H]AMPA and  HKA
a
as radioligands
a
Footnote: Data are given as mean of at least three separate experiments conducted in
b
triplicate. Mean pKi +/- SEM (M) are indicated in brackets. IC50 values (from Møller et
[
17]
c
[21]
al. ). IC50 values (from Szymanska et al.
)
Table 2. Isoelectric points and lipophilicity parameters for the selected amino acids
This article is protected by copyright. All rights reserved.