N-(7-(1 H-Imidazol-1-yl)-2,3-dioxo-6-(trifluoromethyl)-3,4-dihydroquinoxalin-1(2 H)-yl)benzamide, a New Kainate Receptor Selective Antagonist and Analgesic: Synthesis, X-ray Crystallography, Structure-Affinity Relationships, and in Vitro and in Vivo Pharmacology

Article abstract of DOI:10.1021/acschemneuro.9b00479

Selective pharmacological tool compounds are invaluable for understanding the functions of the various ionotropic glutamate receptor subtypes. For the kainate receptors, these compounds are few. Here we have synthesized nine novel quinoxaline-2,3-diones with substitutions in the 7-position to investigate the structure-activity relationship at kainate and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors. Compound 11 exhibited the highest binding affinity across GluK1-3 while having selectivity toward kainate vs AMPA receptors. Compound 11 potently inhibited glutamate evoked currents at homomeric GluK1 and GluK3 receptors in HEK293 cells with K_b values of 65 and 39 nM, respectively. The binding mode of 11 in the ligand binding domain of GluK1 was investigated by X-ray crystallography, revealing that 11 stabilizes the receptor in an open conformation, consistent with its demonstrated antagonism. Furthermore, 11 was tested for analgesic effects in the mouse tail flick test where it significantly increased tail flick latency at doses where 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo[f]-quinoxaline-7-sulfonamide (NBQX) was ineffective.

Full text of DOI:10.1021/acschemneuro.9b00479

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
N-(7-(1H-imidazol-1-yl)-2,3-dioxo-6-(trifluoromethyl)-3,4-
dihydroquinoxalin-1(2H)-yl)benzamide - a new kainate receptor
selective antagonist and analgesic: Synthesis, X-ray crystallography,
structure-affinity relationships, in vitro and in vivo pharmacology
Stine Møllerud, Rie B. Hansen, Jakob S. Pallesen, Piero Temperini, Diletta
Pasini, Jan Bornholdt, Birgitte Nielsen, Esmira Mamedova, Paulina Chalupnik,
Ana V Paternain, Juan Lerma, Marta Diaz del Castillo, Jesper T. Andreasen, Karla
Frydenvang, Jette Sandholm Kastrup, Tommy N. Johansen, and Darryl S. Pickering
ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.9b00479 • Publication Date (Web): 17 Oct 2019
Downloaded from pubs.acs.org on October 23, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 2 of 38
1
2
3
4
5
6
7
8
9
N-(7-(1H-imidazol-1-yl)-2,3-dioxo-6-(trifluoromethyl)-3,4-
dihydroquinoxalin-1(2H)-yl)benzamide - a new kainate receptor
selective antagonist and analgesic: Synthesis, X-ray crystallography,
structure-affinity relationships, in vitro and in vivo pharmacology
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Stine Møllerud,a Rie B. Hansen,a Jakob Pallesen,a Piero Temperini,a Diletta Pasini,a Jan Bornholt,†, a
Birgitte Nielsen,a Esmira Mamedova,a Paulina Chalupnik,b Ana V. Paternain,c Juan Lerma,c Marta
Diaz del Castillo,a Jesper T. Andreasen,a Karla Frydenvang,a Jette S. Kastrup,*,a Tommy N.
Johansen*,a and Darryl S. Pickering*,a
aDepartment of Drug Design and Pharmacology, Faculty of Health and Medical Sciences,
University of Copenhagen, DK-2100 Copenhagen, Denmark
bDepartment of Technology and Biotechnology of Drugs, Collegium Medicum, Jagiellonian
University, PL 30-688 Kraków, Poland
cInstitute of Neuroscience, CSIC-UMH, ES-03550 San Juan de Alicante, Spain
Abstract
Selective pharmacological tool compounds are invaluable for understanding the functions of the
various ionotropic glutamate receptor subtypes. For the kainate receptors these compounds are few.
Here we have synthesized 9 novel quinoxaline-2,3-diones with substitutions in the 7-position to
investigate the structure-activity relationship at kainate and AMPA receptors. Compound 11
exhibited the highest binding affinity across GluK1-3 while having selectivity towards kainate vs.
AMPA receptors. 11 potently inhibited glutamate evoked currents at homomeric GluK1 and GluK3
receptors in HEK293 cells with Kb values of 65 and 39 nM, respectively. The binding mode of 11 in
the ligand binding domain of GluK1 was investigated by X-ray crystallography, revealing that 11
stabilizes the receptor in an open conformation, consistent with its demonstrated antagonism.
Furthermore, 11 was tested for analgesic effects in the mouse tail flick test where it significantly
increased tail flick latency at doses where NBQX was ineffective.
keywords: glutamate receptor, X-ray diffraction, electrophysiology, pain, analgesia, tail-flick
ACS Paragon Plus Environment

Page 3 of 38
ACS Chemical Neuroscience
1
2
3
4
5
Introduction
6
7
8
9
The majority of fast excitatory neurotransmission in the mammalian central nervous system is
mediated by glutamate acting through ionotropic glutamate receptors (iGluRs). This receptor family
comprises the NMDA, AMPA, kainate and delta receptors.1 The kainate receptors, consisting of the
subunits GluK1-5, are located both at the pre- and postsynaptic membrane. Thus, in addition to a
direct excitatory effect, kainate receptors are believed to modulate the activity of neuronal networks
by facilitating or depressing neurotransmitter release.2, 3 While AMPA and NMDA receptors have
proven to be difficult therapeutic targets, the modulatory role of kainate receptors could potentially
be a more suitable target as fine-tuning the balance of excitatory and inhibitory signaling could be
more desirable than completely blocking the essential physiological functions of AMPA and NMDA
receptors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The physiological function of kainate receptors are less well understood compared to the AMPA and
NMDA receptors. This is partially due to a relative lack of selective pharmacological tool compounds
that can discriminate between AMPA and kainate receptors as well as between kainate receptor
subtypes.4 Extensive effort has been put into the design of kainate receptor selective antagonists,
however only antagonists that selectively inhibit GluK1 receptors have been discovered so far.5 Thus,
there is an unmet need for subunit selective ligands for GluK2-5 to aid in the research on the
individual functions of these receptor subunits in health and disease.
The quinoxaline-2,3-diones, such as 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo[f]-quinoxaline-7-
sulfonamide (NBQX; Figure 1), were among the first antagonists applied in non-NMDA receptor
research. However, these compounds are not optimal for elucidating the functions of kainate receptors
as they are also inhibitors of AMPA receptors.4 Only few compounds with preferential binding at
GluK3 or GluK2 have been reported. Some interesting compounds are: a N1-substituted quinoxaline-
2,3-dione, LU97175 (1), which was reported to preferentially bind kainate receptor subunit GluK3
over GluK1 (>10 fold)6 as well as the 6-substituted quinoxalinedione (2), which showed a slight
preference for GluK3 over GluK1 (4-fold).7 Recently we reported on a series of quinoxaline-2,3-
diones substituted at the N1, 6- and 7-positions wherein compound 3 (N-(7-fluoro-2,3-dioxo-6-
(trifluoromethyl)-3,4-dihydroquinoxalin-1(2H)-yl)-2-hydroxybenzamide) exhibited similar binding
affinity as 1 with an 8-fold preferential binding at GluK3 over GluK1 receptors.8
In continuous efforts to identify useful ligands for functional characterization of kainate receptors,
we have synthesized a series of quinoxaline-2,3-dione analogues with varying substituents in the 7-
position and have pharmacologically characterized them with radioligand binding. Of this series
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 4 of 38
1
2
3
4
5
6
7
8
9
compound 11 showed the highest affinity at GluK3 and was further characterized by whole cell patch
clamp in HEK293 cells expressing recombinant rat GluK1 or GluK3 receptors and in vivo for
analgesic effects in the mouse tail flick test. Finally, to help understand the molecular interactions of
this series of quinoxaline-2,3-dione we report a crystal structure of 11 in complex with the GluK1
ligand-binding domain (LBD).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Chemical structures of selected quinoxaline-2,3-diones with antagonistic effect at
AMPA/kainate receptors.
Results and Discussion
Medicinal chemistry
The first series of 7-substituted target compounds were based on a nucleophilic aromatic
substitution
(SNAr)
strategy
(Scheme
1)
starting
from
1,5-difluoro-2-nitro-4-
(trifluoromethyl)benzene (4). Compound 4 was easily transformed into the key intermediate 5
through a SNAr reaction using benzhydrazide at room temperature followed by a second SNAr
reaction of the 5-fluoro substituent of 5 using selected N- and S-nucleophiles to afford
benzohydrazides 6 – 9. Three different transformations of the benzohydrazides 6 – 9 into the
corresponding N1-benzamido-substituted quinoxaline-2,3-dione were investigated: 6 was converted
to 10 in a one-pot procedure through acylation, reduction using iron in acetic acid and ring closure,
analogous to a previously published strategy.8 As an alternative to this strategy, benzohydrazides 7
and 9 were – in a two-step procedure – transformed into 11 and 13, respectively, first through a Pd/C-
catalyzed reduction of the nitro groups by polymethylhydrosiloxane (PMHS) based on a protocol
developed by Rahaim and Maleczka,9 followed by acylation and ring closure. As a variation of the
second strategy, benzohydrazide 8 was converted to target compound 12 first reducing the nitro group
through a Pd/C-catalyzed hydrogenation followed by acylation and ring closure. For the compounds
investigated, all three methods seemed to be applicable and no preferred method was identified.
ACS Paragon Plus Environment

Page 5 of 38
ACS Chemical Neuroscience
1
2
3
4
5
Scheme 1. Synthesis of Target Compounds 10 – 13.a
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) benzohydrazide, DMSO rt, 30 min (b) appropriate nucleophile,
DMSO, heat (c) ethyl chlorooxoacetate, pyridine, and then iron powder, AcOH, reflux (10);
Pd(OAc)2, KF, PHMS THF/H2O, rt and then ethyl chlorooxoacetate, pyridine, DMF (11 and 13);
Pd/C, H2 (60 psi), EtOH and then ethyl chlorooxoacetate, pyridine, DMF (12).
Scheme 2. Synthesis of Target Compounds 14 – 25.a
a
Reagents and conditions: (a) KNO3, H2SO4, 0 °C (b) benzohydrazide, DMSO, 100 °C (c) ethyl
chlorooxoacetate, TEA, DCM and then iron powder, AcOH, reflux; (d) appropriate boronic acid, 2-
aminobiphenyl-based second generation Buchwald precatalyst, K3PO4 (18-20); 3-
acetamidobenzeneboronic acid, Pd(dtbpf)Cl2, K2CO3 (21) (e) Pd(OAc)2, KF, PHMS THF/H2O, rt and
then ethyl chlorooxoacetate, pyridine, DMF (22 and 23); ethyl chlorooxoacetate, pyridine, and then
iron powder, AcOH, reflux (24); Pd/C, H2, DMF and then ethyl chlorooxoacetate, pyridine, DMF
(25).
The synthetic strategy for the second series of 7-substituted target compounds is shown in Scheme 2.
First, 2-bromo-4-fluoro-trifluoromethylbenzene (14) was selectively nitrated as reported earlier10 at
0 °C to give 15 in 37% yield. The fluoro atom of 15 was easily substituted by benzohydrazide at room
temperature to give the key intermediate 16 which was converted to the 7-bromosubstituted
quinoxaline-2,3-dione 17 via acylation, reduction using iron in acetic acid and ring closure. Attempts
were made to use arylbromide 17 in Suzuki cross coupling reactions. Combinations of different
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 6 of 38
1
2
3
4
5
6
7
palladium sources and phosphine ligands, such as Pd2(dba)3, Pd(OAc)2, Pd(dtbpf)Cl2,
.
Pd(dppf)Cl2 DCM, the 2-aminobiphenyl-based second generation Buchwald precatalyst,11, 12 t-Bu3P,
XPhos and SPhos, were investigated using different solvents and bases as well as conventional or
8
9
microwave heating. In most cases, only unconverted starting material 17 could be seen even at high
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
.
temperatures. Only when using the second generation Buchwald precatalyst11 or Pd(dppf)Cl2 DCM,13
cross coupling products were observed, but it was not possible to optimize the conditions and to avoid
the formation of side products. Instead, Suzuki cross coupling reactions using aryl bromide 16 and
the second generation Buchwald precatalyst11 or Pd(dtbpf)Cl2, were straightforward and afforded 18
– 21 almost without the formation of side products. Compound 18 and 19 were converted to the
corresponding quinoxalinediones 22 and 23, respectively, using the Pd/C-catalyzed PMHS reduction
followed by acylation and ring closure. The phenyl-substituted quinoxaline-2,3-dione 24 was
prepared from 20 using the one-pot procedure through acylation, reduction using iron in acetic acid
and ring closure, and finally, compound 25 was prepared from 21 through a Pd/C-catalyzed
hydrogenation followed by acylation and ring closure.
In vitro pharmacology
The synthesized quinoxaline-2,3-diones were characterized in receptor binding studies at
recombinant iGluRs (Table 1). For compound, 11, which had the highest affinity at GluK1-3, the
binding affinities were additionally determined at the soluble LBDs of GluK1, GluK3 and GluA2
yielding Ki values (mean ± SEM) of 43 ± 7 nM, 6.8 ± 2.5 nM and 670 ± 190 nM, respectively.
Additionally, to functionally characterize compound 11, antagonist affinities (Kb) were determined
by whole cell patch clamping. Compound 11’s ability to inhibit glutamate evoked currents in cultured
HEK293 cells expressing either homomeric rat GluK1 or GluK3 receptors was tested at a glutamate
concentration (200 µM and 10 mM, respectively) close to the receptor EC50 values (212 µM and 14.4
mM, respectively) (Figure S1). 11 potently inhibited glutamate responses with similar apparent Kb
values of 65 nM (95% confidence interval (CI): 52-82 nM) and 39 nM (95% CI: 26-60 nM) for
GluK1(Q)1b and GluK3a, respectively (Figure 2). This is in accordance with no statistically significant
difference between the binding affinity at GluK1 and GluK3 (P > 0.05). Compared to compound 1,
where Kb values of 115 nM (GluK1) and 50 nM (GluK3)8 were seen, 11 is comparably potent in
inhibiting glutamate responses. This is in accordance with modest increases in the measured binding
affinity of 11 (Ki: GluK1 = 166 nM, GluK3 = 78 nM) compared to 1 (Ki: GluK1 = 697 nM, GluK3 =
ACS Paragon Plus Environment

Page 7 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
187 nM;8).These electrophysiological studies thus confirmed an antagonistic effect at homomeric
GluK1 and GluK3 receptors.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1: Binding Affinities of Quinoxaline-2,3-dione Analogues (Ki (µM), mean ± SEM)
Native
Compound
GluK1
GluK2
GluK3
GluK5
GluA2
NMDAR
1a
0.697 ± 0.123 0.488 ± 0.058 0.187 ± 0.021 23.9 ± 4.0 1.52 ± 0.25
nd
5.5 ± 0.4
9.6 ± 1.2
nd
10
11
12
13
17
22
23
24
25
4.19 ± 0.60
0.166 ± 0.011 0.523 ± 0.137 0.078 ± 0.011
1.97 ± 0.23 1.68 ± 0.14 0.192 ± 0.038
0.606 ± 0.100 0.651 ± 0.149 0.181 ± 0.042
4.58 ± 0.38 3.50 ± 0.60
18.2 ± 1.2
5.67 ± 0.25
10.1 ± 3.2
 100
5.20 ± 0.30
10-50
33.0 ± 3.6
> 100
22 ± 3
4.53 ± 0.76
8.79 ± 1.47
6.12 ± 0.67
6.23 ± 0.51
8.16 ± 1.04
1.58 ± 0.17
> 10 [89%]
1.5 ± 0.4
4.2 ± 0.7
3.7 ± 0.5
nd
1.61 ± 0.21
3.80 ± 1.00
1.04 ± 0.20 0.460 ± 0.184
2.45 ± 0.27 1.53 ± 0.40
57.8 ± 12.5
43.0 ± 10.4
 100
0.958 ± 0.076 1.69 ± 0.25 0.737 ± 0.022
2.91 ± 0.32 2.07 ± 0.29 1.50 ± 0.30
1.29 ± 0.14 0.941 ± 0.092 0.360 ± 0.093
 10
a ref 8, numbers in brackets is percentages of remaining specific radioligand binding at 10 µM
quinoxalinedione derivative, nd: not determined.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 8 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Antagonism of recombinant homomeric kainate receptors by compound 11. (A) Inhibition
curve of 11 at GluK1(Q)1b receptors against 200 µM glutamate (n = 5-7 cells). Kb= 65 nM (95% CI: 52-82
nM). (B) Inhibition curve of 11 at GluK3a receptors against 10 mM glutamate (n = 3-11 cells). Kb= 39 nM
(95% CI: 26-60 nM). Shown are mean ± SEM of the pooled, normalized data. Curves were constrained at 0%
and 100% for bottom and top, respectively. Example traces for each curve are given to the right.
Structural analysis
The X-ray crystal structure of the GluK1-LBD in complex with 11 was determined at 2.3 Å resolution
with clearly-defined electron density of 11 present in the binding site (Table 2, Figure 3A). The 2-
one and 3-one of the quinoxaline-2,3-dione moiety of 11 make hydrogen bonds with Arg523 of
GluK1-LBD (Figure 3A). The 3-one forms an additional contact to the backbone nitrogen atom
whereas the 2-one has a hydrogen bond to a water molecule. The unsubstituted nitrogen of the
ACS Paragon Plus Environment

Page 9 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
quinoxaline-2,3-dione forms a hydrogen bond with the backbone oxygen atom of Pro516 and the
amide oxygen atom to a water molecule. Additionally,  stacking between the quinoxaline-2,3-
dione moiety and Tyr489 (not shown) and potential  stacking of the benzamide substituent at the
N1-position in the quinoxaline-2,3-dione with protein backbone of Gly688-Ser689 are observed.
Finally, there is a sulfate ion present in the binding site as also previously observed in GluK1-LBD
structures in complex with quinoxaline-2,3-diones8, 14 as well as a putative chloride ion. The
unsubstituted nitrogen atom of the imidazole substituent of 11 is probably protonated (complex
crystallized at pH 4.5), making a salt bridge with the sulfate ion in the binding site possible. The
chloride ion is located under the carbonyl carbon atom and N1 of the quinoxaline-2,3-dione of 11.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The complex crystalized with one molecule in the asymmetric unit and forms a dimer with a
symmetry related molecule (Figure 3B). The dimer interface involves both D1 and D2 lobe residues
with a large buried surface area of 1278 Å2 comparable to that observed in a structure with another
quinoxaline-2,3-dione compound (1339 Å2)8. Domain opening of the complex relative to a glutamate
bound GluK1-LBD structure (PDB ID 2F36, molA) is 32.5°, which is in accordance with an
antagonistic effect.15
Figure 3. A) Interactions between 11 and GluK1-LBD (PDB ID 6SBT). The protein backbone is shown as
cartoon and residues interacting with 11 are shown as beige sticks and 11 in stick representation with carbon
atoms colored green. The sulfate ion located in the binding site is shown as sticks and water molecules
interacting with 11 and a putative chloride ion are displayed as red and green spheres, respectively. Potential
hydrogen bonds are shown as black dashed lines. A 2Fo-Fc omit map carved around the ligand and contoured
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 10 of 38
1
2
3
4
5
6
7
at 1 sigma is shown in grey. B) Overall conformation of the GluK1-LBD dimer (molA grey, symmetry molA
beige) shown as cartoon with surface representation. 11 (green) and sulfate are shown in stick representation.
Domain opening is indicated by an arrow.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Crystal Data, Data Collection and Structure Refinement of 11 in Complex with GluK1-LBD.
Crystal data
Space group
H32
Unit cell: a,b,c (Å)
α, β, γ (°)
Wavelength (Å)
89.27, 89.27, 156.32
90.0, 90.0, 120.0
1.07227
Molecules in asymmetric unit of the crystal
Data collection
Resolution (Å)a
1
44.64-2.30 (2.42-2.30)
13.3 (12.8)
100 (100)
11.1 (49.6)
6.2 (1.5)
Average redundancy
Completeness (%)
Rmerge (%)b
I/σ(I)
CC1/2
0.99 (0.95)
27.1
Wilson B (Å2)
Refinement
Rworkc/Rfreed(%)
20.5/24.9
249/1
2/2/2/61
0.002/0.53
0.0/98.0
0.0/0.0/0.0
Amino acid residues/11
Sulfate/chloride/glycerol/water
R.m.s. deviation bond length (Å)/angles (°)
Ramachandran outliers/favoured (%)
Rotamer outliers (%)/Cβ outliers (%)/Clash score
Average B-values (Å2) for:
Amino acid residues/11
Sulfate/chloride/glycerol/water
49.6/42.5
77.5/68.0/100.1/35.1
a Values in parenthesis correspond to outermost resolution shell. b Rmerge is calculated as follows: Ii(hkl) is the
intensity of an individual measurement of the reflection with Müller indicies hkl, and I(hkl) is the intensity
from multiple observations. Rmerge = ∑hkl∑i│Ii(hkl)-I(hkl)│/∑hkl∑i│Ii(hkl)│. c Rwork = ∑hkl |Fobs - Fcalc| / ∑hkl |Fobs|
where Fobs and Fcalc are the observed and calculated structure factor amplitudes, respectively, for reflection
hkl.d Rfree is equivalent to Rwork, but calculated with 5% of reflections omitted from the refinement process.
Structure-affinity relationships
In order to develop potential pharmacological tools for the characterization of kainate receptors, we
have studied a series of N1-substituted quinoxaline-2,3-diones (all analogues of compound 1,6, 8; see
Figure 1) with different substituents in the 7-position. Five of these compounds (10 – 12, 22 and 23;
Scheme 1 and 2) are 5-membered heterocyclic analogues of compound 1, and three of these
compounds (11, 12 and 23) are equipotent or more potent as compared with starting compound 1;
compound 11 being the most potent compound within the series both at GluK1 and at GluK3 (Table
ACS Paragon Plus Environment

Page 11 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
1), whereas compound 22 and, notably, compound 10 are less potent. Based on the X-ray structure
of compound 11 co-crystallized with the GluK1-LBD, it is clear that these compounds all with 5-
membered heterocyclic substituents most likely adopt conformations in which the heterocyclic
substituents are close to orthogonal to the plane of the quinoxaline-2,3-dione similar to what is seen
for compound 11. These 5-membered heterocyclic substituents are similar in size and thus can easily
be accommodated at GluK1 receptors (Figure 3A). The lack of affinity differences of these 5-
membered heterocyclic substituents across GluK1-3 indicates that the GluK2 and GluK3 binding
pockets can also accommodate such bulky substituents. Interesting loss of affinities (6-fold at GluK1,
9-fold at GluK2 and 19-fold at GluK3) were observed for compound 10 when compared with
compound 1. This loss of affinity may possibly be explained by electronic differences; the
pyrrolidinyl of compound 10 being electron donating whereas the pyrrolyl of compound 1 being
electron withdrawing. Therefore, an electron donating substituent at position 7, such as pyrrolidinyl,
seems to reduce the ligand binding affinity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The unsubstituted nitrogen atom of the imidazole substituent in the co-crystal structure of compound
11 was observed to form interaction with a sulfate ion located in the binding site. Most likely the
sulfate ion will not be present physiologically and water-mediated contacts might be formed instead.
Therefore, the basicity at this heteroatom – compared to the remaining heterocycles included in the
present series – could contribute positively to the binding of compound 11 to the receptors.
Two compounds within the series (24 and 25) have larger 6-membered aromatic rings as substituents
in the 7-position. Affinity data (Table 1) show that both these substituents are accepted by GluK1 - 3
receptors; compound 25 with an additional amido substituent being the more potent of the two.
Compound 13, also containing an amido substituent, but through a flexible side chain, is also well
accepted by GluK1 - 3 receptors. Even though the GluK1-LBD co-crystal structure of compound 11
reveals the presence of residues capable of donating and accepting hydrogen bonds, the binding
affinities of compound 13 and 25 seem to indicate that neither of the two amido substituents are
involved in important receptor interactions through hydrogen bonding.
The series of compounds were also investigated at recombinant GluK5, GluA2 and at native
NMDA receptors, the latter using [3H]CGP 39653 as radioligand. At GluK5 receptors and at native
NMDA receptors affinities between 1.5 and >100 µM were observed. Most interestingly, at GluK5
recombinant receptors the order of potency observed for compounds 1, 10 and 11 follow the same
trend as was seen at GluK1 and GluK3 receptors; compound 11 being 2.4 – 4.6-fold more potent
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 12 of 38
1
2
3
4
5
6
7
8
9
than compound 1 and compound 10 being 4 – 19-fold less potent than compound 1. At GluA2
receptors, however, all compounds except compound 25 showed affinities within a narrow range
(4.5 – 18 µM), illustrating that the structure-affinity relationships observed at GluK1 - 3 receptors
could not be transferred to GluA2 receptors. Thus, with a Ki of 5.67 µM at GluA2 receptors
compound 11 showed the largest discrimination between GluK3 and GluA2 (73-fold) compared to
a corresponding 8-fold discrimination as seen for compound 1. Compound 25 were the only new
compound within the series with a GluA2 receptor affinity comparable to that of compound 1.
Interesting at GluA2 receptors, the 3-amido substituent of compound 25 apparently contributes
positively to the binding of compound 25 as compared to compound 24; compound 25 being 6-fold
more potent than compound 24 at GluA2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In vivo pharmacology
To assess the acute analgesic effect of 11 (0, 0.3, 1.2, 4.8 and 9.8 nmol i.t.) NMRI mice were tested
for tail withdrawal response by the tail flick test. Animals were tested at baseline and 15, 30 and 60
min after drug administration. NBQX (8.0 nmol i.t.) and morphine (3.5 nmol i.t. or 5 mg/kg s.c.)
treatment were included for comparison.
Tail flick test. Prior to injection a baseline response was determined for each treatment group.
Intrathecal administration of 11 resulted in a dose-dependent increase in the tail flick latency in
response to a noxious thermal stimulus applied to the tail (Figure 4A). The ANCOVA indicated a
significant main effect of treatment (F4,31 = 4.12; P < 0.001) and time (F2,63 = 21.53; P < 0.001)
but no treatment by time interaction (F8,63 = 1.21; P = 0.31). Pairwise comparisons within each time
point showed that 11 significantly increased the latency for the two highest doses of 11 (4.8 and 9.8
nmol) at 15 min post drug administration and for the highest dose at 30 min compared to vehicle (15
min, 10.5 ± 0.5 s and 11.7 ± 1.6 s vs. 7.0 ± 1.2 s, P < 0.05 and P < 0.001; 30 min, 8.4 ± 0.8 vs. 4.6 ±
0.5 s, P < 0.01; n = 5-11). One mouse that was administered the highest dose of 11 had a minor paresis
of the right hind paw at 15 and 30 min post drug administration; at 60 min, the paresis had resolved.
This could be an adverse effect of 11, but may also be due to the i.t. drug delivery.
For NBQX (8.0 nmol i.t.) there was a near-significant main effect of treatment (F1,7 = 2.51; P =
0.072), but not for time (F2,16 = 2.04; P = 0.16) or treatment by time interaction (F2,16 = 2.45; P =
0.12). Still the pairwise comparisons within each time point did show a significant increase the tail
flick latency 30 min post drug administration when compared to vehicle treatment at the same time
point (3.3 ± 0.4 s vs. 6.5 ± 0.2 s, P < 0.01, n = 5, Figure 4B). One mouse that was administered NBQX
ACS Paragon Plus Environment

Page 13 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
had paralysis of the right hind limb after drug administration and was culled. This could be an adverse
effect of NBQX16, but may also be due to the i.t. drug delivery.
7
8
9
Morphine administration resulted in a significant main effect of treatment (F2,20 = 5.28; P < 0.05)
and time (F2,40 = 7.01; P < 0.01) but no treatment by time interaction (F4,40 = 0.69; P = 0.60).
Pairwise comparisons within each time point showed that i.t. administration of morphine (3.5 nmol)
gave a significant increase in the tail flick latency at 30 min post drug administration when compared
to vehicle treatment at the same time point (7.8 ± 0.8 s vs. 4.7 ± 0.5 s, P < 0.05, n = 7-11, Figure 4C).
Subcutaneous administration of morphine (5 mg/kg) resulted in a significant increase in the tail flick
latency 15 and 30 min post drug administration when compared to i.t. vehicle treatment (15 min, 10.4
± 0.3 s vs. 7.0 ± 1.2 s, P < 0.05; 30 min, 8.4 ± 0.8 vs. 4.7 ± 0.5 s, P < 0.01; n = 6-11). One mouse that
was administered i.t. morphine had a stiff upright tail and was excluded from the study. Two mice
administered subcutaneous morphine were slightly hyperactive at 30 and 60 min post drug
administration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although not statistically significant, it appears that the light isoflurane anaesthesia may affect the
tail flick latency 15 min post drug administration where a minor increase in the latency is observed
for the vehicle treated mice. This effect may possibly mask any minor analgesic effect of the lower
doses of 11 and the effect of NBQX and i.t. morphine at 15 min.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 14 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. The effect of 11 on acute nociceptive behaviour in the mouse tail flick test. All drugs were
administered after baseline (BL) latencies were obtained and the analgesic effect was assessed at 15, 30 and
60 min post drug administration. (A) 11 (0.3-9.8 nmol, 5 µl, i.t), (B) NBQX (8.0 nmol, 5 µl, i.t) and (C)
morphine (3.5 nmol, 5 µl, i.t. or 5 mg/kg, s.c.). Data are presented as mean + SEM, n = 5-11 per group, *P <
0.05, **P < 0.01 and ***P < 0.001 compared to vehicle treatment at same time point (two-way RM ANCOVA
followed by pairwise planned comparisons using Fisher’s LSD).
ACS Paragon Plus Environment

Page 15 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Locomotor activity. ANOVA indicated a significant main effect of treatment (F1,13 = 11.95; P <
0.01, Figure 5) and time (F8,104 = 17.11; P < 0.001) but no treatment by time interaction (F8,104 =
0.97; P = 0.46). Pairwise comparisons within each time point showed that 11 significantly increased
locomotor activity at the following time points: 10 min (P < 0.001), 15 min (P < 0.01), 20 min (P <
0.05) and 25 min (P < 0.05).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20
20
100
100
80
80
**
60
60
VEH
EM1117
VEH
15
15
40
40
20
20
0
0
10
10
VEH 11
VEH EM17
***
**
*
*
5
5
5
5
10 15 20 25 30 35 40 45
10 15 20 25 30 35 40 45
Time (min)
Time (min)
Figure 5. Effect of 11 on locomotor activity. 11 (9.8 nmol) caused an overall increase in locomotor activity
(F1,13 = 11.95; P < 0.01), specifically increasing locomotor activity at 10 min (P < 0.001), 15 min (P < 0.01),
20 min (P < 0.05) and 25 min (P < 0.05), n = 7-8.
Rotarod test. The ANOVA showed no significant main effect of treatment (F1,15 = 3.16; P = 0.10,
Figure 6), time (F1,15 = 1.24; P = 0.28) or treatment by time interaction (F1,15 = 1.24; P = 0.28).
Pairwise comparison revealed no significant effect of 11 at the tested time points.
VEH
VEH
11
EM17
350
350
300
300
250
250
200
200
150
150
100
100
50
50
0
0
15 min
15 min
30 min
30 min
Figure 6. Effect of 11 (9.8 nmol) on motor coordination as measured with the rotarod test. 11 showed no
significant effect on time spent on the rod, n = 9.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 16 of 38
1
2
3
4
5
6
7
8
9
The canonical quinoxalinedione NBQX has been employed for many years in research as a non-
NMDA iGluR competitive antagonist and has been previously demonstrated to show analgesic
properties in mice.16, 17 However, these early generation quinoxalinediones failed to proceed on to
clinical trials due to problems of precipitation in the kidneys leading to nephrotoxicity.18 One aim of
our current study was thus to develop more potent quinoxalinedione analogues in the hopes that lower
doses of compound may be better tolerated in vivo and still produce analgesia. Here we have
preliminarily tested our lead compound (11) in the mouse tail flick test of acute pain where it was
found to indeed show analgesia and to be more efficacious than NBQX. It is of note that in our
radioligand binding experiments NBQX is 43-fold selective for GluA2 vs. GluK3 while 11 has the
opposite selectivity, being 73-fold selective for GluK3 vs. GluA2. 11 is therefore a kainate receptor
selective antagonist compared with NBQX. This supports the notion that the analgesic effects
observed by the early generation quinoxalinediones were likely mediated via kainate receptors rather
than AMPA receptors, consistent with a proposed role of kainate receptors in nociception.19 Since 11
is non-selective amongst GluK1-3 we cannot know from our data which subtype of the kainate
receptors is the predominant mediator of analgesia seen in this study. 11 had no adverse effects on
motor coordination in the rotarod test, as previously observed for NBQX.17 However, 11 was found
to produce an increase in locomotor activity and further experiments are needed to examine this and
to expand on the side effect profile of 11 as well as to determine its pharmacokinetic profile.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusion
Based on the preference in binding at kainate receptors over AMPA receptors observed for compound
1 and our previous work investigating substitutions in the N1 and 6-position we have synthesized
nine compounds with focus on varying the substituent in the 7-position. The quinoxaline-2,3-dione
derivatives were tested in a radioligand binding assay and found to have relatively high binding
affinity at kainate receptors. 11 showed the highest binding affinity across GluK1-3 along with the
largest (73-fold) selectivity between GluA2 and GluK3. 11 potently inhibited glutamate evoked
currents at homomeric GluK1 and GluK3 receptors recombinantly expressed in HEK293 cells with
Kb values in the low nanomolar range. As a representative of the series of synthesized compounds 11
was crystallized with GluK1-LBD and induced a domain opening comparable to other antagonists.
Finally, 11 showed analgesic effects in a mouse test of pain response at doses of 4.8 and 9.8 nmol,
being more potent than NBQX. 11 could serve as a lead candidate for developing new potent and
kainate receptor selective quinoxaline-2,3-diones.
ACS Paragon Plus Environment

Page 17 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
Methods
9
Chemistry
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthetic details can be found in the supporting information. Commercially available chemicals
were used without any purification. Anhydrous reactions were carried out in oven or flame dried
glassware. Dry solvents (DMF, THF and DCM) were dried using a solvent purification system. Thin
layer chromatography was performed on TLC silica gel 60 F254 and compounds were visualized using
UV light (254 nm). For flash column chromatography Merck silica gel 60 (0.063–0.200 mm) was
used as the stationary phase. LC-MS was performed at an Agilent 1100 HPLC with an XBridge 3.5
μm C-18 column (100 × 4.6 mm) using a linear gradient elution from buffer A (H2O:MeCN:HCOOH,
95:5:0.1) to buffer B (H2O:MeCN:HCOOH, 5:95:0.1) over 10 min. Flow rate: 1.0 mL/min. The
HPLC was coupled to a Hewlett Packard 1100 series mass spectrometer with an electrospray
ionization source. Preparative HPLC was carried out on a Dionex ultimate 3000 system with binary
pump, UV/VIS detector and 10 mL injection loop. The column was a Gemini® 5 μm NX-C18 110
Å, LC Column 250 × 21.2 mm. Appropriate linear gradient programs were set up using mobile phase
A (H2O:HCOOH, 100:0.1) and mobile phase B (H2O:MeCN:HCOOH, 10:90:0.1) with a flow rate of
15 mL/min. 1H-NMR was performed on a 400 MHz Bruker Avance III equipped with a 5 mm broad
band probe at 400.09 MHz. Samples were dissolved in DMSO-d6 or chloroform-d and analyzed at
300 K and referenced using DMSOat 2.50 ppm or chloroform at 7.26 ppm. 13C-NMR was performed
at 100.61 MHz using DMSO at 39.51 ppm or chloroform at 77.20 ppm as reference. Purity of the
tested compounds was determined by analytical HPLC on an UltiMate HPLC system
(ThermoScientific) consisting of a LPG-3400A pump (1.0 mL/min), a WPS-3000SL autosampler,
and a DAD-3000D diode array detector using a Gemini-NX C18 column (4.6 × 250 mm, 3m, 110Å).
Gradient elution (0 to 100% B (MeCN-H2O-TFA 90:10:0.1) in solvent A (H2O-TFA (100:0.1) over
10 min was used. Compounds were dissolved in MeCN. Data were acquired and processed using
Chromeleon Software v. 6.80. All compounds tested had a purity at 254 and at 280 nm of at least
95%, unless otherwise stated.
Synthesis of target compound 11 and intermediate compounds 5 and 7.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 18 of 38
1
2
3
4
5
6
7
8
9
N'-(5-Fluoro-2-nitro-4-(trifluoromethyl)phenyl)benzohydrazide 5. Benzohydrazide (3.95 g,
29.0 mmol) was added to a stirred solution of 4 (6.50 g, 28.6 mmol) in dry DMSO (40 mL). The
reaction mixture was stirred for 30 min at rt and under N2 and then poured into a 0.1% aqueous
NaHCO3 solution (400 mL). Stirring was continued for 30 min. The precipitate was filtered, washed
with H2O and dried in vacuo to give 5 (9.34 g, 95%) as a yellow solid. 1H-NMR (DMSO-d6) δ 10.95
(s, 1H), 10.10 (s, 1H), 8.45 (d, J = 7.3 Hz, 1H), 7.98 (d, J = 7.0 Hz, 2H), 7.64 (t, J = 7.3 Hz, 1H), 7.55
(t, J = 7.2 Hz, 2H), 7.20 (d, J = 13.4 Hz, 1H). 13C-NMR (DMSO-d6) δ 166.56, 150.19, 132.67, 132.48,
129.00, 128.22, 127.84, 127.76, 102.59 (q, J = 27 Hz), 99.97.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N'-(5-(1H-Imidazol-1-yl)-2-nitro-4-(trifluoromethyl)phenyl)benzohydrazide 7. Imidazole
(817 mg, 12.0 mmol) was added to a stirred solution of 5 (1.029 g, 3.0 mmol) in dry DMSO (6 mL)
under N2. After heating at 100 °C for 45 min, the reaction mixture was poured into H2O (60 mL) and
aqueous KH2PO4 (1M, 2 mL), and stirred for 30 min. The beige precipitate formed was filtered,
washed with H2O and dried. Recrystallization from EtOH (40 mL) afforded 7 (970 mg, 83 %) as a
yellow solid. 1H-NMR (DMSO-d6) δ 10.93 (s, 1H), 10.15 (s, 1H), 8.53 (s, 1H), 7.93 (d, J = 7.4 Hz,
2H), 7.84 (s, 1H), 7.61 (t, J = 7.3 Hz, 1H), 7.52 (t, J = 7,4 Hz, 2H), 7.40 (s, 1H), 7.25 (s, 1H), 7.08 (t,
13
J = 1.1 Hz, 1H). C-NMR (DMSO-d6) δ 166.57, 147.83, 141.12, 138.54, 132.67, 132.42, 130.12,
129.52, 129.02, 128.14, 126.98, 123.32 (q, J = 270 Hz), 122.18, 116.16, 113.64 (q, J = 32 Hz).
N-(2,3-Dioxo-7-(1H-imidazol-1-yl)-6-(trifluoromethyl)-3,4-dihydroquinoxalin-1(2H)-yl)-
benzamide 11. Pd(OAc)2 (3 mg, 0.014 mmol,) and 7 (110 mg, 0.32 mmol,) were dissolved in freshly
distilled dry THF (5 mL) under N2. Aqueous KF (0.048 g, 0.84 mmol, in 0.8 mL degassed H2O) was
added, followed by dropwise addition of PMHS (0.1 mL, 1.68 mmol). The reaction was stirred at rt
for 3 h, Et2O (10 mL) was added, and the stirring continued for 5 min. The layers were separated and
the aqueous layer was extracted with Et2O (2 × 20 mL). The combined organic phases were filtered
through Celite flushing with EtOAc. The filtrate was dried with MgSO4, filtered and concentrated in
vacuo. The residue was dissolved in dry DMF (4 mL), ethyl chlorooxoacetate (0.035 mL, 0.32 mmol)
and pyridine (0.067 mL, 0.83 mmol) was added and the reaction mixture was stirred under N2 at rt
for 1h and at 100 °C overnight. The crude product was purified by preperative HPLC. Appropriate
1
fractions were concentrated and freeze-dried to give 11 as a white solid (93 mg, 70 %). H-NMR
(DMSO-d6) δ 12.93 (s, 1H), 11.90 (s, 1H), 8.81 (s, 1H), 8.05 (d, J = 7.6 Hz, 2H), 7.88 (d, J = 8.1 Hz,
ACS Paragon Plus Environment

Page 19 of 38
ACS Chemical Neuroscience
1
2
3
4
2H), 7.75 (s, 1H), 7.67 (t, J = 7.2 Hz, 1H), 7.60 – 7.53 (m, 3H). 13C-NMR (DMSO-d6) δ 165.52,
5
6
7
162.35, 153.74, 132.85, 131.49, 130.98, 128.63, 128.17, 125.19, 122.98.
8
9
Crystallization
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Rat GluK1-LBD was expressed and purified as previously described.20 In short, the LBD construct
was expressed in the E. coli cell line Origami 2. The construct contains an N-terminal His fusion tag
(leaving a glycine after tag removal) and the protein was purified by nickel affinity chromatography
on a HisTrap column followed by ion exchange and size exclusion chromatography.
11 was dissolved in protein buffer (10 mM HEPES pH 7.0, 20 mM NaCl) and pH was increased to
approximately 9 with concentrated NaOH to increase solubility before mixing with GluK1-LBD.
Solid 11 was added to saturate the solution and protein-ligand solution was equilibrated for 3 h at 6°C
prior to setup.
GluK1-LBD in complex with 11 was crystallized by the hanging drop vapor diffusion method at 6°C.
Drops consisted of 1 µL reservoir solution and 1 µL complex solution (4.0 mg/mL GluK1-LBD and
saturated 11 with excess of solid 11). Volume of reservoir solution was 0.5 mL and crystals grew
within one week from set up. The crystal used for data collection was achieved using reservoir
solution containing 20% PEG4000, 0.3 M lithium sulfate, 0.1 M phosphate-citrate, pH 4.5.
X-ray data collection and processing
Crystals were soaked in cryo-buffer (reservoir solution containing 20% glycerol) and flash cooled in
liquid nitrogen. X-ray diffraction data were collected at the ID29 beamline (ESRF, Grenoble, France)
at a wavelength of 1.07227 Å to 2.3 Å resolution. Data processing was performed with XDS21 and
SCALA within the CCP4i suite of programs.22 Molecular replacement using PHASER (implemented
in CCP4i;23) was performed using individual D1 and D2 lobes of PDB entry 4QF9 (molA) as search
models. Automated model building was initially performed using AutoBuild in PHENIX.24 Next, 11
coordinates were created in Maestro [Maestro version 10.7, Schrödinger, LLC, New York, NY,
2016], geometry optimized [MacroModel 11.3, Schrödinger, LLC New York, NY, 2016]] and fitted
into the electron density in the ligand binding site. A restraint file for 11 was generated with eLBOW
keeping geometry obtained in Maestro.25 The structure was refined in PHENIX using isotropic B
factors, TLS and riding hydrogen atoms. The structure was manually inspected and corrected in
COOT26 between refinement steps.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 20 of 38
1
2
3
4
5
Structure analysis
6
7
8
9
Domain opening was calculated relative to the structure of the fully closed structure of GluK1-LBD
in complex with glutamate (PDB entry 2F36, molA) using the DynDom server.27 Buried surface areas
were calculated using the PISA server28 and figures were prepared using PyMOL [The PyMOL
Molecular Graphics System, version 1.8.6.1, Schrödinger, LCC].
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Radioligand binding assays
Native NMDA receptors. Binding at native NMDA receptors was carried out as previously
detailed29 in rat brain synaptosomal membranes which had been washed four times in ice-cold 50
mM Tris-HCl buffer, pH 7.4, containing 2.5 mM CaCl2 on the day of assaying and using 2 nM
[3H]CGP 39653. The final pellet was resuspended in ice-cold buffer to an approximate concentration
of 0.4–0.5 mg of protein/mL. The binding was carried out in 96-wells plates in a total volume of 250
µL and equilibrated at 0 °C for 60 min before rapid filtration through GF/C filters (Perkin-Elmer Life
Sciences), using a 96-well Packard FilterMate cell harvester, and three washes with 250 μL of ice-
cold binding buffer. To the dried filters was added 30 L/well of Microscint scintillation fluid
(Perkin-Elmer Life Sciences), and the amount of filter bound radioactivity was quantified in a Packard
TopCount microplate scintillator counter. Nonspecific binding was determined using 1.0 mM Glu.
Full length receptors. Ligand binding affinities at recombinant rat homomeric GluA2, GluK1-3 and
GluK5 were determined as previously detailed.30-32 In short, ligands were diluted in assay buffer
(GluA2: 50 mM Tris-HCl, 100 mM KSCN, 2.5 mM CaCl2 pH 7.2 at 4 °C; GluK1-3,5: 50 mM Tris-
HCl pH 7.1 at 4 °C), mixed with sf9 insect cell membranes expressing the respective receptors and
radioligand (GluA2: [3H]-(RS)-AMPA (57.5 Ci/mmol; PerkinElmer, Waltham, MA), GluK1: [3H]-
(S)-NF608 (16.3 Ci/mmol)30, GluK2,3,5: [isopropenyl-3H]-kainic acid (43.6-47.2 Ci/mmol;
PerkinElmer) in a total volume of 0.25 mL followed by 2h equilibration at 4 °C. GluK1-3 assays were
filtered through GF/B-type glass fiber filters in microtitre plate format (UniFilter-96, PerkinElmer)
and washed twice with ice-cold assay buffer on a FilterMate manifold (PerkinElmer). The filters were
dried at 70 °C for 1 h and 50 µL/well Microscint 20 (PerkinElmer) was added. Radioactivity was
detected with a TopCounter (PerkinElmer). GluA2 and GluK5 assays were filtered on GF/C type
filters using Millipore 12-well filtration manifolds (Merck Life Science, Denmark) and washed twice
with ice-cold assay buffer. Filters were placed in pony vials (PerkinElmer), 3 mL Ultima Gold
scintillation fluid (PerkinElmer) added and DPM determined using a TriCarb 2900 scintillation
counter (PerkinElmer).
ACS Paragon Plus Environment

Page 21 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Soluble ligand binding domain constructs. Binding to GluA2-LBD (50 ng protein/tube), GluK1-LBD
(50 ng/tube) and GluK3-LBD (250 ng/tube) was conducted essentially as previously described.33, 34
LBD assay buffers are the same as for full length receptor binding but additionally contain 10% (v/v)
glycerol. Experiments were performed at 4 °C and bound and free radioligand were separated by
filtration onto 0.22 µM mixed cellulose ester filters (Tisch Scientific, North Bend, OH) on a 12-well
Millipore filtration manifold, washing twice with 1 mL ice-cold assay buffer. Filters were dried, then
solubilized in 2 mL Filter Count scintillation fluid (PerkinElmer) and radioactivity determined by
liquid scintillation counting using a TriCarb 2900 scintillation counter.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Binding data analysis. All competition curves (n  3 per ligand) were determined in triplicate at 12–
16 ligand concentrations. Data were analyzed using GraphPad Prism 6 (GraphPad Software, San
Diego, CA) to determine ligand affinity (Ki) and Hill coefficient (nH) using the one-site Ki and four-
parameter logistics equations, respectively.
Electrophysiology
Cell culturing and transfection. HEK293 cells were maintained in Dulbecco’s modified eagle
medium supplemented with fetal bovine serum and penicillin/streptomycin. Cells were seeded on
poly-D-lysine-coated coverslips and transiently transfected with 3 µg rat GluK1(Q)1b or 5 µg rat
GluK3a cDNA (in pCis vector) using 2 µL lipofectamine (Invitrogen, CA) per transfection. Cells
were co-transfected with 1 µg GFP for visualization of transfected cells and used 24-72 h post
transfection.
Recording procedure. Membrane currents were recorded using the patch-clamp technique under
whole-cell configuration. Patch pipettes had resistances of 3–5 MΩ. Currents were filtered at 10 kHz
(two-pole Butterworth filter, −12 db/octave) and transferred to a personal computer for analysis.
Extracellular solution contained (in mM): 140 NaCl, 2.5 KCl, 1.8 CaCl2, 10 HEPES, 15 glucose, 1
MgCl2, pH 7.3-7.4. Pipette solution (in mM): 130 CsMeSO4, 4 NaCl, 10 HEPES, 10 TEA, 1 EGTA,
5 Qx314 bromide (Sigma Chemical, MO), 2 ATP, 0.5 GTP, pH 7.3, 290 mOsm. (S)-glutamate was
diluted from stock solutions on the day of experiment. A 25 mM solution of 11 in DMSO was
prepared and stored at -20 °C. On the day of experiment 11 was diluted and added to (S)-glutamate
solutions (maximal final DMSO concentration: 0.04%). All experiments were performed at a holding
potential of -60 mV and at room temperature. The cells were rapidly perfused using a linear array of
7 glass tubes and rapid change (< 2 ms) between solutions was controlled by a motor. Kb values were
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 22 of 38
1
2
3
4
5
6
7
8
9
determined using glutamate concentrations (GluK1: 200 µM, GluK3: 10 mM) near the respective
EC50 values (212 µM, 14.4 mM, respectively. Supplementary Information, Figure S1). 11 was pre-
applied for 20-30 seconds before the same concentration of 11 was co-applied with glutamate. Data
were collected and pooled from 27 cells for GluK1(Q)1b and 29 cells for GluK3a.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electrophysiology data analysis. Data were displayed using pClamp software (Axon Instruments,
Foster City, CA, USA), and responses were normalized to a saturating (GluK1: 10 mM, GluK3: 100
mM) glutamate response (= 100% response). Data were analyzed using GraphPad Prism 6 (GraphPad
Software, San Diego, CA) to determine IC50 and Hill slope (4-parameter logistic equation). IC50
values of the pooled, normalized curves were converted to Kb values using the modified Cheng-
Prusoff equation.
In vivo pharmacology
Animals. A total of 90 six week old male NMRI mice (Envigo, Venray, The Netherlands) were
included in the study. The animals were housed in groups in individually ventilated GM500+ cages
(524 cm2) at a 12-hour light-dark cycle (lights on at 07:00 AM) in a temperature (20-24 ˚C) and
humidity controlled environment. Upon arrival, animals were left to acclimatize for a minimum of
one week prior to initiation of experiments. They were provided with Tapvei 2HV wood-chip bedding
material (Harjumaa, Estonia) and allowed free access to water and standard diet (Altomin 1314;
Brogaarden, Lynge, Denmark) except during behavioural testing, which was conducted between 9
am and 4 pm. For environmental enrichment, mice were provided with paper ropes, an S-brick
(Tapvei, Harjumaa, Estonia), a red translucent shelter and corn hidden in the bedding. The experiment
was approved by the Danish Animal Experiments Inspectorate, Ministry of Environment and Food
(License no. 2012-15-2934-00165, C3) and conducted under the guidelines of the International
Association for the Study of Pain.35
Drug treatment. 11 (synthesised in house at the Department of Drug Design and Pharmacology,
University of Copenhagen, Denmark), NBQX (Tocris Bioscience, Bristol, United Kingdom) and
morphine hydrochloride (10 mg/ml, Skanderborg Produktionsapotek, Denmark) were
dissolved/diluted in 4% DMSO/10 mM NaH2PO4 (pH = 8.0) and administered intrathecally (i.t., 26-
gauge needle) at the lumbar enlargement of mice as a single bolus injection of 5 µl. Morphine
administered subcutaneously (s.c.) was diluted in sterile saline. The highest dose of 11 (9.8 nmol) and
ACS Paragon Plus Environment

Page 23 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
s.c. morphine was assessed in a separate experiment, however, as no significant differences were
found in baseline tail flick latencies the data was pooled.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Tail flick test. The reflex response to a noxious thermal stimulus was measured in normal uninjured
NMRI male mice using the tail flick test. Mice were placed in cylindrical Plexiglas test boxes on a
framed glass pane (Ugo Basile Inc., Comerio, Italy) and habituated for approximately 45 minutes. A
radiant heat source (Plantar Test, Ugo Basile Inc., Comerio, Italy) was focused on the ventral surface
of the tail ⅓-¼ down from the base of the tail. The apparatus was calibrated to give a baseline tail
flick latency of approximately 4–6 s (20 s cut-off) and the tail flick latency was measured to the
nearest 0.1 s. Before drug testing mice were acclimatised to the test chamber on two separate days
and a baseline measurement performed to familiarize the mice with the testing procedure. A baseline
measurement (two measurements separated by 5 min) was made just prior to the drug test. Mice were
lightly anaesthetized with isoflurane (Baxter A/S, Allerød, Denmark) administered either drug or
vehicle and the tail flick latency measured 15, 30 and 60 min post drug administration. The
experimenter was blinded to drug treatment.
Tail flick data analysis. Data were analyzed with a two-way repeated measure analysis of covariance
(RM-ANCOVA) with baseline as covariate, treatment as independent factor and time as repeated
measure followed by pairwise planned comparisons using Fisher’s LSD. One-way ANOVA followed
by Tukey’s multiple comparisons test was used to compare baseline tail flick latencies prior pooling
of the data. For all statistical analyses, a probability value of P < 0.05 was considered significant.
Spontaneous activity - locomotor activity (LA) test. Mice (n = 7-8) were individually placed for 30
minutes in transparent cages (L: 37 cm x W: 21 cm x H: 15 cm) with a thin layer of bedding.
Locomotor distance was automatically recorded by a camera mounted above the arenas and stored on
a computer equipped with Ethovision XT (Noldus, Holland).
Motor coordination - rotarod test. Motor function was evaluated using an accelerating rotarod
(MedAssociates, Inc., VT). The rotarod (3.2 cm diameter) speed was increased from 4 to 40 rpm over
a 300 s period with the minimum time possible to spend on the rod designated as 0 s and the maximum
cut-off time set at 310 s. Mice (n = 9) were tested 15 and 30 minutes after treatment. When an animal
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 24 of 38
1
2
3
4
5
6
7
fell off the rotating drum, a photobeam was automatically broken to record the amount of time spent
on the rotating rod.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Locomotor and coordination data analysis. Data were analyzed with a two-way repeated measure
analysis of variance (RM-ANOVA) with treatment as independent factor and time as repeated
measure followed by pairwise planned comparisons using Fisher’s LSD. To ensure normality and
variance homogeneity, which is required for ANOVA/ANCOVA, data were log-transformed before
statistical analysis, but displayed graphically on the original scale.
ASSOCIATED CONTENT
Supporting Information.
The Supporting Information available free of charge on the ACS Publication website at DOI: XXX,
includes:
Synthesis and characterization of compounds; Figure S1, Glutamate concentration-response
curves of transfected HEK293 cells (PDF)
Molecular formula strings (CSV)
Accession Codes
The structure of compound 11 in GluK1-LBD is deposited in the protein data bank (PDB ID
6SBT). Authors will release the atomic coordinates and experimental data upon article publication.
Author Information
ACS Paragon Plus Environment

Page 25 of 38
ACS Chemical Neuroscience
1
2
3
4
Corresponding Authors
5
*D.S.P. on pharmacology: Email: picker@sund.ku.dk Phone: +45 35334342.
*T.N.J. on chemistry: E-mail: tnj@sund.ku.dk. Phone: +45 35336412.
*J.S.K. on crystallography: E-mail: jsk@sund.ku.dk . Phone: +45 35336486.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ORCID
Karla Frydenvang: 0000-0002-5823-3478
Darryl S. Pickering: 0000-0001-8687-6094
Jette Sandholm Kastrup: 0000-0003-2654-1510
Tommy N. Johansen: 0000-0001-5918-0833
Stine Møllerud: 0000-0001-6828-8539
Present Address
† J.B.: LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark.
Author Contributions
Design and synthesis: J.P., J.B., E.M., P.C., and T.N.J. Pharmacology: S.M., P.T., D.P., B.N.,
A.V.P., J.L., and D.S.P. In vivo studies: R.B.H., M.D.C., and J.T.A. Crystallography: S.M., K.F.,
and J.S.K. All authors interpreted data and contributed to data analysis. All authors have given
approval to the final version of the manuscript.
Notes
The authors declare no competing financial interest.
Acknowledgements
Heidi Peterson is thanked for help with expression and purification of the ligand-binding domain of
GluK1. ESRF, Grenoble, France is thanked for providing beam time.
This research was generously supported by: the Lundbeck Foundation (S.M., D.S.P., K.F., and
J.S.K.), GluTarget (J.B., T.N.J.), Danscatt (S.M., K.F., J.S.K.), European Union Horizon 2020
Research and Innovation Program under the Marie Skłodowska-Curie grant agreement No. 642720
(M.D.C).
Abbreviations used
AMPA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid); GluK1-LBD, ligand-binding
domain of GluK1; iGluRs, ionotropic glutamate receptors; i.t., intrathecal; LBD, ligand-binding
domain; LU 97175, 1-benzamido-7-pyrrol-1-yl-6-trifluoromethylquinoxaline-2,3-(1H,4H)-dione;
NBQX, 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo[f]-quinoxaline-7-sulfonamide; [3H]NF608, (S)-
2-amino-3-(6-[3H]-2,4-dioxo-3,4-dihydrothieno[3,2-d]pyrimidin-1(2H)-yl)-propanoic acid; NMDA,
N-methyl-D-aspartate; s.c., subcutaneous.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 26 of 38
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 27 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
References
7
8
9
1.
Traynelis, S. F.; Wollmuth, L. P.; McBain, C. J.; Menniti, F. S.; Vance, K. M.; Ogden,
K. K.; Hansen, K. B.; Yuan, H.; Myers, S. J.; Dingledine, R. Glutamate receptor ion channels:
structure, regulation, and function. Pharmacol Rev 2010, 62, 405-96.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.
Evans, A. J.; Gurung, S.; Henley, J. M.; Nakamura, Y.; Wilkinson, K. A. Exciting
Times: New Advances Towards Understanding the Regulation and Roles of Kainate Receptors.
Neurochem Res 2019, 44, 572-584.
3.
292-311.
4.
and therapeutic potential. Neuropharmacology 2009, 56, 90-113.
5. Larsen, A. M.; Bunch, L. Medicinal chemistry of competitive kainate receptor
antagonists. ACS Chem Neurosci 2011, 2, 60-74.
Lerma, J.; Marques, J. M. Kainate receptors in health and disease. Neuron 2013, 80,
Jane, D. E.; Lodge, D.; Collingridge, G. L. Kainate receptors: pharmacology, function
6.
Loscher, W.; Lehmann, H.; Behl, B.; Seemann, D.; Teschendorf, H. J.; Hofmann, H.
P.; Lubisch, W.; Hoger, T.; Lemaire, H. G.; Gross, G. A new pyrrolyl-quinoxalinedione series of
non-NMDA glutamate receptor antagonists: pharmacological characterization and comparison with
NBQX and valproate in the kindling model of epilepsy. Eur J Neurosci 1999, 11, 250-62.
7.
Demmer, C. S.; Rombach, D.; Liu, N.; Nielsen, B.; Pickering, D. S.; Bunch, L.
Revisiting the Quinoxalinedione Scaffold in the Construction of New Ligands for the Ionotropic
Glutamate Receptors. ACS Chem Neurosci 2017, 8, 2477-2495.
8.
Pallesen, J. S.; S, M.; Frydenvang, K.; Pickering, D. S.; Bornholdt, J.; Nielsen, B.;
Pasini, D.; Han, L.; Marconi, L.; Kastrup, J. S.; Johansen, T. N. N1-Substituted Quinoxaline-2,3-
diones as Kainate Receptor Antagonists: X-ray Crystallography, Structure-Affinity Relationships
and in vitro Pharmacology. ACS Chem Neurosci 2019.
9.
aromatic and aliphatic nitro groups. Org Lett 2005, 7, 5087-90.
10. Cooper, D. G. F., I. T.; Garzya, V.; Jin, J.; Louchart, Y.; Walker, G.; Wyman, P. A.
Benzimidazoles which have activity at m1 receptor and their uses in medicine. 2007.
11. Kinzel, T.; Zhang, Y.; Buchwald, S. L. A new palladium precatalyst allows for the
fast Suzuki-Miyaura coupling reactions of unstable polyfluorophenyl and 2-heteroaryl boronic
acids. J Am Chem Soc 2010, 132, 14073-5.
12.
W. Suzuki-Miyaura cross-coupling reaction on copper-trans-A(2)B corroles with excellent
functional group tolerance. Tetrahedron 2011, 67, 4243-4252.
13.
Uesugi, M. Wrenchnolol derivative optimized for gene activation in cells. J Am Chem Soc 2009,
131, 4774-82.
14.
Bowie, D.; Frydenvang, K.; Kastrup, J. S.; Bunch, L. Binding mode of an alpha-amino acid-linked
quinoxaline-2,3-dione analogue at glutamate receptor subtype GluK1. ACS Chem Neurosci 2015, 6,
845-54.
15.
structures of kainate receptors. Neuropharmacology 2017, 112, 16-28.
16. Lutfy, K.; Cai, S. X.; Woodward, R. M.; Weber, E. Antinociceptive effects of NMDA
Rahaim, R. J., Jr.; Maleczka, R. E., Jr. Pd-catalyzed silicon hydride reductions of
Konig, M.; Reith, L. M.; Monkowius, U.; Knor, G.; Bretterbauer, K.; Schoefberger,
Jung, D.; Shimogawa, H.; Kwon, Y.; Mao, Q.; Sato, S.; Kamisuki, S.; Kigoshi, H.;
Demmer, C. S.; Moller, C.; Brown, P. M.; Han, L.; Pickering, D. S.; Nielsen, B.;
Møllerud, S.; Frydenvang, K.; Pickering, D. S.; Kastrup, J. S. Lessons from crystal
and non-NMDA receptor antagonists in the tail flick test in mice. Pain 1997, 70, 31-40.
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 28 of 38
1
2
3
4
5
6
7
17.
Blackburn-Munro, G.; Bomholt, S. F.; Erichsen, H. K. Behavioural effects of the
novel AMPA/GluR5 selective receptor antagonist NS1209 after systemic administration in animal
models of experimental pain. Neuropharmacology 2004, 47, 351-62.
18.
Turski, L.; Schneider, H. H.; Neuhaus, R.; McDonald, F.; Jones, G. H.; Lofberg, B.;
8
9
Schweinfurth, H.; Huth, A.; Kruger, M.; Ottow, E. Phosphonate quinoxalinedione AMPA
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
antagonists. Restor Neurol Neuros 2000, 17, 45-59.
19.
Ruscheweyh, R.; Sandkuhler, J. Role of kainate receptors in nociception. Brain Res
Brain Res Rev 2002, 40, 215-22.
20.
Naur, P.; Vestergaard, B.; Skov, L. K.; Egebjerg, J.; Gajhede, M.; Kastrup, J. S.
Crystal structure of the kainate receptor GluR5 ligand-binding core in complex with (S)-glutamate.
FEBS Lett 2005, 579, 1154-60.
21.
22.
Kabsch, W. Xds. Acta Crystallogr D Biol Crystallogr 2010, 66, 125-32.
Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.; Emsley, P.; Evans, P. R.;
Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.; McCoy, A.; McNicholas, S. J.; Murshudov, G. N.;
Pannu, N. S.; Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S. Overview of the
CCP4 suite and current developments. Acta Crystallogr D Biol Crystallogr 2011, 67, 235-42.
23.
Read, R. J. Phaser crystallographic software. J Appl Crystallogr 2007, 40, 658-674.
24. Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis, I. W.; Echols, N.;
McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M. D.; Storoni, L. C.;
Headd, J. J.; Hung, L. W.; Kapral, G. J.; Grosse-Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.;
Oeffner, R.; Read, R. J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H.
PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta
Crystallogr D Biol Crystallogr 2010, 66, 213-21.
25.
Moriarty, N. W.; Grosse-Kunstleve, R. W.; Adams, P. D. electronic Ligand Builder
and Optimization Workbench (eLBOW): a tool for ligand coordinate and restraint generation. Acta
Crystallogr D Biol Crystallogr 2009, 65, 1074-80.
26.
Coot. Acta Crystallogr D Biol Crystallogr 2010, 66, 486-501.
27. Hayward, S.; Lee, R. A. Improvements in the analysis of domain motions in proteins
from conformational change: DynDom version 1.50. J Mol Graph Model 2002, 21, 181-3.
28. Krissinel, E.; Henrick, K. Inference of macromolecular assemblies from crystalline
state. J Mol Biol 2007, 372, 774-97.
29. Assaf, Z.; Larsen, A. P.; Venskutonyte, R.; Han, L.; Abrahamsen, B.; Nielsen, B.;
Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and development of
Gajhede, M.; Kastrup, J. S.; Jensen, A. A.; Pickering, D. S.; Frydenvang, K.; Gefflaut, T.; Bunch, L.
Chemoenzymatic synthesis of new 2,4-syn-functionalized (S)-glutamate analogues and structure-
activity relationship studies at ionotropic glutamate receptors and excitatory amino acid
transporters. J Med Chem 2013, 56, 1614-28.
30.
Alcaide, A.; Marconi, L.; Marek, A.; Haym, I.; Nielsen, B.; Møllerud, S.; Jensen, M.;
Conti, P.; Pickering, D. S.; Bunch, L. Synthesis and pharmacological characterization of the
selective GluK1 radioligand (S)-2-amino-3-(6-[3H]-2,4-dioxo-3,4-dihydrothieno[3,2-d]pyrimidin-
1(2H)-yl)propanoic acid ([3H]-NF608). Med. Chem. Commun. 2016, 7, 9.
31.
affinity GluK5 kainate receptor. Eur J Pharmacol 2016, 788, 315-320.
32. Sagot, E.; Pickering, D. S.; Pu, X.; Umberti, M.; Stensbol, T. B.; Nielsen, B.;
Møllerud, S.; Kastrup, J. S.; Pickering, D. S. A pharmacological profile of the high-
Chapelet, M.; Bolte, J.; Gefflaut, T.; Bunch, L. Chemo-enzymatic synthesis of a series of 2,4-syn-
functionalized (S)-glutamate analogues: new insight into the structure-activity relation of ionotropic
glutamate receptor subtypes 5, 6, and 7. J Med Chem 2008, 51, 4093-103.
ACS Paragon Plus Environment

Page 29 of 38
ACS Chemical Neuroscience
1
2
3
4
5
6
33.
Krintel, C.; Frydenvang, K.; Ceravalls de Rabassa, A.; Kaern, A. M.; Gajhede, M.;
Pickering, D. S.; Kastrup, J. S. L-Asp is a useful tool in the purification of the ionotropic glutamate
receptor A2 ligand-binding domain. FEBS J 2014, 281, 2422-30.
7
8
34.
Venskutonyte, R.; Larsen, A. P.; Frydenvang, K.; Gajhede, M.; Sagot, E.; Assaf, Z.;
9
Gefflaut, T.; Pickering, D. S.; Bunch, L.; Kastrup, J. S. Molecular recognition of two 2,4-syn-
functionalized (S)-glutamate analogues by the kainate receptor GluK3 ligand binding domain.
ChemMedChem 2014, 9, 2254-9.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
35.
Zimmermann, M. Ethical guidelines for investigations of experimental pain in
conscious animals. Pain 1983, 16, 109-10.
ACS Paragon Plus Environment