Improved synthetic route for the GluN2-specific NMDA receptor glycine site agonist AICP

Article abstract of DOI:10.1016/j.tetlet.2020.151653

(R)-2-Amino-3-(4-(2-ethylphenyl)-1H-indole-2-carboxamido)propanoic acid (AICP) is a N-methyl-D-aspartate (NMDA) receptor glycine site agonist with unprecedented high potency in the low nanomolar range, and a GluN2 subunit-dependent pharmacological profile in terms of potency and agonist efficacy (Jessen et al., 2017 [1]). Here, we report a scalable, practical and cost-efficient synthetic route for AICP, which is an improvement compared to the previously reported route. This improved synthetic route includes a versatile diphenylmethylester (DPM) protection for the amino acid moiety, which can be widely used in the synthesis of other amino acid ligands. Further functional evaluation of AICP at the different ionotropic glutamate receptor (iGluR) classes demonstrates that high affinity binding of AICP to the orthosteric binding site is selective for NMDA receptors over AMPA and kainate receptors. Furthermore, high affinity binding of AICP is not observed at GluN3A, GluN3B, and GluD2 subunits, which also bind glycine and D-serine. Thus, the new approach described here enables scalable synthesis of AICP for the use as a pharmacological tool compound to study the involvement of neuronal NMDA receptor subtypes in normal brain function and disease.

Full text of DOI:10.1016/j.tetlet.2020.151653

Journal Pre-proofs
Improved synthetic route for the GluN2-specific NMDA receptor glycine site
agonist AICP
Fabao Zhao, Nirvan Rouzbeh, Kasper B. Hansen, Rasmus P. Clausen
PII:
S0040-4039(20)30070-8
DOI:
https://doi.org/10.1016/j.tetlet.2020.151653
TETL 151653
Reference:
To appear in:
Tetrahedron Letters
Received Date:
Revised Date:
Accepted Date:
20 November 2019
7 January 2020
17 January 2020
Please cite this article as: Zhao, F., Rouzbeh, N., Hansen, K.B., Clausen, R.P., Improved synthetic route for the
GluN2-specific NMDA receptor glycine site agonist AICP, Tetrahedron Letters (2020), doi: https://doi.org/10.1016/
j.tetlet.2020.151653
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Ltd.

Improved synthetic route for the GluN2-specific NMDA receptor glycine site agonist AICP
1
2
2
1
Fabao Zhao , Nirvan Rouzbeh , Kasper B. Hansen , Rasmus P. Clausen
1
Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences,
University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark
Rasmus Prætorius Clausen, e-mail: rac@sund.ku.dk
2
Department of Biomedical and Pharmaceutical Sciences, Center for Biomolecular
Structure and Dynamics, Center for Structural and Functional Neuroscience, University of
Montana, 32 Campus Dr, Missoula, MT 59812.
Graphical abstract
Br
Br
O
N
H
OH
O
O
Cost-efficient & practical
NO2
N
OH
One-step mild deprotection
H
NH
NH2
DPM as a vaild amino acid
protecting group
O
O
H2N
OH
Boc
AICP
H2N
O
O
HN
HN
Boc
Highlights
•
•
•
•
•
Novel cost-efficient and practical route of AICP synthesis
Preparation of ethyl 4-bromo-1H-indole-2-carboxylate via Reissert indole synthesis
Diphenylmethylester as a TFA deprotectable group for amino acid protection
Selective high affinity binding of AICP to NMDA receptor
Binding to AMPA, kainate receptor and GluN3A, GluN3B, GluD2 subunits is not

observed
Abstract
(
R)-2-Amino-3-(4-(2-ethylphenyl)-1H-indole-2-carboxamido)propanoic acid (AICP) is
a N-methyl-D-aspartate (NMDA) receptor glycine site agonist with unprecedented high
potency in the low nanomolar range and a GluN2 subunit-dependent pharmacological
1
profile in terms of potency and agonist efficacy . Here, we report a scalable, practical and
cost-efficient synthetic route for AICP, which is an improvement compared to the
previously reported route. This improved synthetic route includes a versatile
diphenylmethylester (DPM) protection for the amino acid moiety, which can be widely
used in the synthesis of other amino acid ligands. Further functional evaluation of AICP
at the different ionotropic glutamate receptor (iGluR) classes demonstrates that high
affinity binding of AICP to the orthosteric binding site is selective for NMDA receptors
over AMPA and kainate receptors. Furthermore, high affinity binding of AICP is not
observed at GluN3A, GluN3B, and GluD2 subunits, which also bind glycine and D-serine.
Thus, the new approach described here enables scalable synthesis of AICP for the use
as a pharmacological tool compound to study the involvement of neuronal NMDA
receptor subtypes in normal brain function and disease.
Key words
Ethyl 4-bromo-1H-indole-2-carboxylate; Amino acid protection; Subunit selectivity;
Two-electrode voltage-clamp electrophysiology.

Introduction
NMDA receptors represent a functional class of ionotropic glutamate receptors
2
(
iGluR) that play vital roles in the central nervous system (CNS) . Aberrant NMDA
receptor signaling has been implicated in neurological disorders, such as schizophrenia,
3-6
depression, epilepsy and Alzheimer’s disease . Seven subunits of NMDA receptors
7
have been identified, named GluN1, GluN2A-D and GluN3A-B . NMDA receptors are
tetrameric assemblies of subunits and most receptors in the CNS are comprised of two
GluN1 subunits, which bind glycine or D-serine, and two GluN2 subunits that bind
7
glutamate . Despite the distinct developmental and regional distribution in the CNS
among the different NMDA receptor subtypes, the endogenous agonist, glycine, D-serine
and glutamate (Chart 1) display almost no subtype selectivity. Considerable efforts have
been made to develop subtype-selective orthosteric agonists targeting NMDA receptors,
8
albeit with limited success so far . Interestingly, unlike the endogenous agonists glycine
and D-serine, the agonist efficacy of the GluN1 glycine site agonist, D-cycloserine, is
9
GluN2-subunit dependent . Electrophysiological recordings of NMDA receptor activation
revealed that the maximal response to D-cycloserine compared to that of glycine (i.e.
relative agonist efficacy) at the different NMDA receptor subtypes is ~90% at GluN1/2A,
~
60% at GluN1/2B, ~190% at GluN1/2C, and ~100% at GluN1/2D receptors.
D-cycloserine is therefore a partial agonist at GluN1/2B receptors, but a superagonist at
8
GluN1/2C receptors . More recently, we demonstrated that AICP has ~90% relative
efficacy at GluN1/2A, ~10% at GluN1/2B receptors, ~350% at GluN1/2C, and ~25% at
GluN1/2D receptors. Furthermore, the potencies of AICP at different NMDA receptor

subtypes are in the low nanomolar range (2-66 nM), which is 1000-fold more potent than
1
D-cycloserine . This pharmacological profile of AICP at NMDA receptor subtypes is
unprecedented among existing NMDA receptor agonists and allosteric modulators, and
we recently used AICP to identify a unique role of neuronal GluN2C-containing NMDA
receptors in the nucleus reticularis of the thalamus, thereby demonstrating
10
GluN2C-selective in vivo targeting of NMDA receptors by AICP .
a) endogenous agonists
b) high potency, selective agonists
O
O
O
HO
OH
HN
NH2
H2N
OH
glycine
NH2
O
D-cycloserine
D-serine
O
N
O
O
O
OH
H
HO
OH
NH
NH2
NH2
glutamate
AICP
Chart 1 (a) Non-selective endogenous NMDA receptor agonists; (b) GluN2-dependent NMDA receptor agonists
Urwyler et al.11 reported a synthetic route for AICP (Scheme 1), which
used
and
4
-(2-ethylphenyl)-1H-indole-2-carboxylic
acid
(
R)-3-amino-2-benzyloxycarbonylaminopropionic acid methyl ester as starting materials,
going through Suzuki coupling, amide coupling and two deprotection steps to afford the
product. In this route, both the starting material, 4-bromo-1H-indole-2-carboxylic acid and
the protected amino acid are highly expensive. Moreover, two deprotecting steps are
required and methyl ester protecting group was hydrolyzed by NaOH, which might cleave
the amide bond. Thus, this route was not suitable for production of large amounts of
AICP for carrying out detailed in vitro and in vivo pharmacological characterization. We
therefore report a new improved synthetic route for AICP we have developed, which
begins from the cost-efficient materials 1-bromo-2-methyl-3-nitrobenzene and

D-asparagine, and employs protecting groups that can be removed in one step.
Previous work (Urwyler et al.)
O
+
H2N
O
O
AICP
HN
Cbz
N
OH
H
•
•
Expensive starting material
Two-step deprotection
This work
Br
Br
O
N
OH
NO2
H
AICP
O
O
H2N
O
OH
Boc
H2N
O
HN
HN
Boc
•
•
•
Cost-efficient & practical
One-step mild deprotection
DPM as a vaild amino acid protecting group
Scheme 1 Synthesis route for AICP.
Results and discussion
The key intermediate in the synthesis route is ethyl 4-bromo-1H-indole-2-carboxylate
9
. Two methods (Scheme 2), Hemetsberger indole synthesis12 and Reissert indole
13
synthesis , were explored by us to obtain this product. The first synthesis started from
the Knoevenagel condensation between 2-bromobenzaldehyde and ethyl
-azidoacetate 2 to afford 3, which was followed by reflux in xylene to give 9 via a nitrene
1
2
intermediate 4. Unfortunately, this method failed to provide 3 in our hands due to
unexpected side reactions. The reported yield of the Hemetsberger indole synthesis is
often relatively low and starting material ethyl azidoacetate 2 is an unstable and
expensive reagent, so we looked for another method. Reissert indole synthesis began
from the treatment of 1-bromo-2-methyl-3-nitrobenzene 5 with NaH, which was then
reacted with diethyl oxalate 6 to give 7. The nitro group of 7 was then reduced by iron to

form the aniline intermediate 8, which was cyclized automatically to afford compound 9.
Compared to Hemetsberger indole synthesis, the materials of second route are more
economical and the overall yield was 64.7% over two steps.
Br
O
Br
O
Br
O
a
b
O
H
+
N3
O
O
O
N3
N
Br
1
2
3
4
O
O
Hemetsberger indole synthesis
N
H
Br
Br
O
O
Br
O
O
9
c
d
+
O
O
O
O
64.7%, 2 steps
O
NO2
O
NO2
NH2
5
6
7
8
Reissert indole synthesis
Scheme 2 Two possible paths to synthesize key intermediate 9. Reagents and conditions: (a) NaOEt, EtOH, 0℃, 4h; (b)
xylene, reflux, 1h; (c) NaH, DMF, 0℃ to r.t, 4h; (d) Fe, EtOH, AcOH, reflux, 4h.
As for the amino acid part, (R)-3-amino-2-benzyloxycarbonylaminopropionic acid
methyl ester was used in the previously reported route. This implicates a two-step
deprotection procedure and the use of NaOH for hydrolysis of methyl ester that tend to
also cleave the amide bond. We therefore explored the use of two protecting groups that
can be cleaved in one deprotection step under mild conditions to reduce the amount of
side products. In first attempt, the α-amino motif was protected with a tert-butoxycarbonyl
(Boc) group, whereas the carboxylic acid was protected as a tert-butyl ester, since these
two groups can be easily removed in one step by trifluoroacetic acid (TFA) with high yield.
The synthesis began from Boc-D-asparagine 10, and the carboxylic acid was protected
with a tert-butyl group through nucleophilic substitution with tert-butyl bromide to get 11.
Treatment of compound 11 with [bis(trifluoroacetoxy)iodo]benzene (PIFA) via a Hofmann
rearrangement was carried out to afford the dual protected amino acid 12. However, a
large quantity of tert-butyl bromide (50 equivalents) was needed to get a high yield in this

route. As a result, we explored diphenylmethylester (DPM) as an alternative to the
tert-butyl group in the acid protection. In this route, compound 10 was reacted with
diphenylchloromethane to get 13. Finally, compound 14 was prepared through Hofmann
rearrangement by converting the amide into an amino group. Only 1.5 equivalents of
diphenylchloromethane is enough for the acid protection, which is more efficient
compared to tert-butyl protection. Compound 14 was finally used as the amino part for
AICP synthesis. The DPM protecting group can also be conveniently cleaved by TFA in
one step, similar to tert-butyl group.
O
O
c
H2N
O
H2N
O
52.1%
a
O
HN
HN
Boc
11
Boc
12
O
53.4%
H2N
OH
Boc
O
HN
b
10
c
O
O
72.5%
H2N
O
80.6%
H2N
O
O
HN
HN
Boc
Boc
13
14
Scheme 3 Two paths to synthesize dual protected Amino acid. Reagents and conditions: (a) tert-Butyl bromide, K
DMF, 55℃, 48h; (b) diphenylchloromethane, NEt , CH CN, 55℃, overnight; (c) (bis(trifluoroacetoxy)iodo)benzene,
pyridine, EtOAc, CH CN, r.t, 48h.
2 3
CO ,
3
3
3
The key intermediate 4-bromo-1H-indole-2-carboxylate 9 was then treated with
-ethylphenylboronic acid through Suzuki coupling to give 15, which was then hydrolyzed
2
to compound 16. 16 was reacted with 14 using the coupling reagent EDCI and HOBt to
afford 17, which was then deprotected in TFA directly to afford the final product AICP in
one step.

Br
a
b
O
O
O
O
O
9
2.9%
95.2%
N
H
N
H
N
OH
H
9
15
16
O
O
c
O
O
d
NH
OH
NH
O
35.5%, 2steps
NH
NH2
NH
HN
Boc
AICP
17
Scheme 4 Synthesis route of AICP. Reagents and conditions: (a) (2-ethylphenyl)boronic acid, Pd(PPh
EtOH, H O, Toluene, N , 85℃, overnight; (b) i), NaOH, MeOH, H O, 3h; ii), HCl, H O; (c) EDCI, HOBt, NEt
r.t, overnight; (d) TFA, r.t, 3h.
3
)
4
, Na
2 3
CO ,
2
2
2
2
3
, 14, DCM,
To evaluate the pharmacological activity of the synthesized AICP, we determined the
effects of AICP on current responses from different iGluR subtypes. The recombinant
iGluRs were expressed in Xenopus oocytes and current responses in the absence and
presence of 1 µM AICP were measured using two-electrode voltage-clamp
electrophysiology. AICP selectively modulated responses from GluN1/2A-D NMDA
receptor subtypes activated by 10 µM glutamate and 1 µM glycine with a profile that is
1
consistent with previously published results . That is, AICP increased GluN1/2A
responses, consistent with a subsaturating glycine concentration (glycine EC50 is 1 µM)
and the full agonist activity of AICP at this NMDA receptor subtype (Figure 1). Responses
from GluN1/2C were increased by AICP due to superagonist activity at this subtype,
whereas AICP reduced responses from GluN1/2B and GluN1/2D due to partial agonist
activity at these subtypes. These results obtained at recombinant GluN1/2A-D NMDA
receptor subtypes provide confidence that the new synthetic route produces AICP
1
compound with activity consistent with previous studies . AICP did not affect responses
from AMPA receptors (GluA1-4) or kainate receptors (GluK1-2) activated by 30 µM
kainate or 100 µM glutamate, respectively (Figure 1). Thus, high-affinity binding of AICP

is observed at GluN1/2A-D NMDA receptors, but is not observed at the agonist binding
sites of the evaluated AMPA and kainate receptors. In addition, 1 µM AICP did not
modulate responses activated by 100 µM glycine from GluN1FATL/3A and GluN1FATL/3B
receptors, which include two mutations in the glycine binding pocket of GluN1 (F484A +
T518L) to prevent desensitization (Figure 1). Finally, 1 µM AICP did not modulate
Lc
spontaneously active current responses from GluD2 , which possess the Lurcher
mutation14 (Figure 1). In summary, the functional evaluation of AICP demonstrates that
high affinity binding of AICP to the agonist binding site is selective for NMDA receptors
over AMPA and kainate receptors, and effects of AICP were not observed at the
glycine-binding GluN3A, GluN3B, and GluD2 subunits.

A
AICP
glutamate
AICP
glutamate
glycine
glycine
GluN1/2B
GluN1/2C
AICP
AICP
-solution
glycine
Na
+
GluN1FATL/3B
GluD2Lc
B
250
2
1
1
00
50
00
5
0
0
C
L
A
/2B /2C /2D
/3A
/3B uA uA uA uA uK uK
1
2
3
4
1
2
/
2
L
L
2
1
1
1
1
l
l
l
l
l
l
D
AT
AT
F
N
N
N
N
F
G
G
G
G
G
G
lu
lu
lu
lu
lu
1
1
G
G
G
G
G
N
N
lu
lu
G
G
Figure 1 Selectivity of AICP for modulation of GluN1/2 NMDA receptors. A) Representative two-electrode voltage-clamp
recordings of current responses from recombinant glutamate receptors expressed in Xenopus oocytes. Current
responses were recorded in the absence and presence of 1 µM AICP. For GluD2Lc recordings, the oocytes were initially
+
+
perfused with a Na -free NMDG-solution and then perfused with Na - solution to determine the amplitude of the
+
Lc
spontaneously active Na -permeable GluD2 receptor response (see Methods). B) Summary of modulation by 1 µM
AICP of current responses from recombinant GluN1/2 and GluN1FATL/3 NMDA receptors, GluA1-4 AMPA receptors,
GluK1-2 kainate receptors, and spontaneously active GluD2Lc receptors. Responses from GluN1/2 receptors were
activated by 10 µM glutamate in the continuous presence of 1 µM glycine. Responses from GluN1FATL/3 receptors were
activated by 100 µM glycine. Responses from GluA1-4 were activated by 30 µM kainate and responses from GluK1-2
receptors were activated by 100 µM glutamate (see Methods for details). Scale bars are 200 nA (vertical) and 30 sec
(horizontal). Data are normalized to the control response in the absence of AICP and are mean ± SEM of 6-11 oocytes.
Conclusion

In summary,
a
new methodology for the synthesis of AICP from
1-bromo-2-methyl-3-nitrobenzene and D-asparagine has been disclosed. Compared to
the previous synthetic route, cost-efficient starting materials and a one-step deprotection
were used to pave the way for large scale synthesis of AICP and other analogues. A
practical synthetic path of 4-bromo-1H-indole-2-carboxylic acid has been developed, and
dual protection of the amino acid (i.e. both the carboxyl acid and α-amino group) was
performed. The DPM protection group can be widely used for other amino acid protection.
The synthesized AICP selectively modulated NMDA receptors over AMPA and kainate
receptors in a functional assay.
Acknowledgement
The authors gratefully acknowledge financial support from the National Institutes of
Health (NIH) (P20GM103546 and R01NS097536). We thank Gina Bullard for excellent
technical assistance. The authors thank China Scholarship Council for financial support.
KBH is a PI on a research grant from Janssen to University of Montana.
Appendix A. Supplementary data
Supplementary
data
to
this
article
can
be
found
online
at
XXXXXXXXXXXXXXXXXXXX
References
1.
Jessen, M.; Frederiksen, K.; Yi, F.; Clausen, R. P.; Hansen, K. B.; Brauner-Osborne, H.;

Kilburn, P.; Damholt, A., Identification of AICP as a GluN2C-Selective N-Methyl-D-Aspartate
Receptor Superagonist at the GluN1 Glycine Site. Mol. Pharmacol. 2017, 92 (2), 151-161.
2.
Hansen, K. B.; Yi, F.; Perszyk, R. E.; Menniti, F. S.; Traynelis, S. F., NMDA Receptors in
the Central Nervous System. In NMDA Receptors: Methods and Protocols, Burnashev, N.;
Szepetowski, P., Eds. Springer New York: New York, NY, 2017; pp 1-80.
3.
Balu, D. T., Chapter Twelve - The NMDA Receptor and Schizophrenia: From
Pathophysiology to Treatment. In Advances in Pharmacology, Schwarcz, R., Ed. Academic
Press: 2016; Vol. 76, pp 351-382.
4.
Deutschenbaur, L.; Beck, J.; Kiyhankhadiv, A.; Mühlhauser, M.; Borgwardt, S.; Walter,
M.; Hasler, G.; Sollberger, D.; Lang, U. E., Role of calcium, glutamate and NMDA in major
depression and therapeutic application. Progress in Neuro-Psychopharmacology and
Biological Psychiatry 2016, 64, 325-333.
5.
Symonds, J. D.; Zuberi, S. M.; Johnson, M. R., Advances in epilepsy gene discovery and
implications for epilepsy diagnosis and treatment. Current Opinion in Neurology 2017, 30 (2),
1
93-199.
Wang, R.; Reddy, P. H., Role of Glutamate and NMDA Receptors in Alzheimer's Disease.
6.
J Alzheimers Dis 2017, 57 (4), 1041-1048.
Traynelis, S. F.; Wollmuth, L. P.; McBain, C. J.; Menniti, F. S.; Vance, K. M.; Ogden, K.
7.
K.; Hansen, K. B.; Yuan, H.; Myers, S. J.; Dingledine, R., Glutamate receptor ion channels:
structure, regulation, and function. Pharmacol. Rev. 2010, 62 (3), 405-96.
8.
Maolanon, A. R.; Risgaard, R.; Wang, S. Y.; Snoep, Y.; Papangelis, A.; Yi, F.; Holley, D.;
Barslund, A. F.; Svenstrup, N.; Hansen, K. B.; Clausen, R. P., Subtype-Specific Agonists for

NMDA Receptor Glycine Binding Sites. ACS Chem Neurosci 2017, 8 (8), 1681-1687.
Schade, S.; Paulus, W., D-Cycloserine in Neuropsychiatric Diseases: A Systematic
Review. Int. J. Neuropsychopharmacol. 2016, 19 (4).
0. Liu, J.; Shelkar, G. P.; Zhao, F.; Clausen, R. P.; Dravid, S. M., Modulation of burst firing
9.
1
of neurons in nucleus reticularis of the thalamus by GluN2C-containing NMDA receptors.
Molecular Pharmacology 2019, mol.119.116780.
11. Urwyler, S.; Floersheim, P.; Roy, B. L.; Koller, M., Drug design, in vitro pharmacology,
and structure-activity relationships of 3-acylamino-2-aminopropionic acid derivatives, a novel
class of partial agonists at the glycine site on the N-methyl-D-aspartate (NMDA) receptor
complex. J. Med. Chem. 2009, 52 (16), 5093-107.
1
2. Hemetsberger, H.; Knittel, D., Synthese und Thermolyse von α-Azidoacrylestern.
Monatshefte für Chemie / Chemical Monthly 1972, 103 (1), 194-204.
3. Reissert, A., Einwirkung von oxalester und natriumäthylat auf nitrotoluole. synthese
nitrirter phenylbrenztraubensäuren. Berichte der deutschen chemischen Gesellschaft 1897,
1
3
1
1
0 (1), 1030-1053.
4. Yuzaki, M.; Aricescu, A. R., A GluD Coming-Of-Age Story. Trends Neurosci. 2017, 40 (3),
38-150.
Declaration of interests
☒
The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this

paper.
☐The authors declare the following financial interests/personal relationships which
may be considered as potential competing interests:
Graphical abstract
Br
Br
O
N
H
OH
O
O
Cost-efficient & practical
NO2
N
OH
One-step mild deprotection
H
NH
NH2
DPM as a vaild amino acid
protecting group
O
O
H2N
OH
Boc
AICP
H2N
O
O
HN
HN
Boc

Highlights
•
•
•
•
•
Novel cost-efficient and practical route of AICP synthesis
Preparation of ethyl 4-bromo-1H-indole-2-carboxylate via Reissert indole synthesis
Diphenylmethylester as a TFA deprotectable group for amino acid protection
Selective high affinity binding of AICP to NMDA receptor
Binding to AMPA, kainate receptor and GluN3A, GluN3B, GluD2 subunits is not observed