2-Aminoimidazole-based antagonists of the 5-HT6 receptor – A new concept in aminergic GPCR ligand design

Article abstract of DOI:10.1016/j.ejmech.2019.06.001

A new strategy in the design of aminergic GPCR ligands is proposed – the use of aromatic, heterocyclic basic moieties in place of the evergreen piperazine or alicyclic and aliphatic amines. This hypothesis has been tested using a benchmark series of 5-HT₆R antagonists obtained by coupling variously substituted 2-aminoimidazole moieties to the well established 1-benzenesulfonyl-1H-indoles, which served as the ligands cores. The crystallographic studies revealed that upon protonation, the 2-aminoimidazole fragment triggers a resonance driven conformational change leading to a form of higher affinity. This molecular switch may be responsible for the observed differences in 5-HT₆R activity of the studied chemotypes with different amine-like fragments. Considering the multiple functionalization sites of the embedded guanidine fragment, diverse libraries were constructed, and the relationships between the structure and activity, metabolic stability, and solubility were established. Compounds from the N-(1H-imidazol-2-yl)acylamide chemotype (10a–z) exhibited high affinity for 5-HT₆R and very high selectivity over 5-HT_1A, 5-HT_2A, 5-HT₇ and D₂ receptors (negligible binding), which was attributed to their very weak basicity. The lead compound in the series 4-methyl-5-[1-(naphthalene-1-sulfonyl)-1H-indol-3-yl]-1H-imidazol-2-amine (9i) was shown to reverse the cognitive impairment caused by the administration of scopolamine in rats indicating procognitive potential.

Full text of DOI:10.1016/j.ejmech.2019.06.001

Accepted Manuscript
2
-Aminoimidazole-based antagonists of the 5-HT receptor – A new concept in
6
aminergic GPCR ligand design
Adam S. Hogendorf, Agata Hogendorf, Rafał Kurczab, Justyna Kalinowska-Tłuścik,
Piotr Popik, Agnieszka Nikiforuk, Martyna Krawczyk, Grzegorz Satała, Tomasz
Lenda, Joanna Knutelska, Ryszard Bugno, Jakub Staroń, Wojciech Pietruś, Mikołaj
Matłoka, Krzysztof Dubiel, Rafał Moszczyński-Pętkowski, Jerzy Pieczykolan, Maciej
Wieczorek, Bogusław Pilarski, Paweł Zajdel, Andrzej J. Bojarski
PII:
S0223-5234(19)30518-5
DOI:
https://doi.org/10.1016/j.ejmech.2019.06.001
EJMECH 11404
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 20 November 2018
Revised Date: 17 May 2019
Accepted Date: 1 June 2019
Please cite this article as: A.S. Hogendorf, A. Hogendorf, Rafał. Kurczab, J. Kalinowska-Tłuścik, P.
Popik, A. Nikiforuk, M. Krawczyk, G. Satała, T. Lenda, J. Knutelska, R. Bugno, J. Staroń, W. Pietruś,
Mikoł. Matłoka, K. Dubiel, Rafał. Moszczyński-Pętkowski, J. Pieczykolan, M. Wieczorek, Bogusł. Pilarski,
Paweł. Zajdel, A.J. Bojarski, 2-Aminoimidazole-based antagonists of the 5-HT receptor – A new
6
concept in aminergic GPCR ligand design, European Journal of Medicinal Chemistry (2019), doi: https://
doi.org/10.1016/j.ejmech.2019.06.001.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

ACCEPTED MANUSCRIPT
2
-Aminoimidazole-based antagonists of the
5
-HT receptor – a new concept in aminergic
6
GPCR ligand design
a,*
a
a
b
Adam S. Hogendorf, Agata Hogendorf, Rafał Kurczab, Justyna Kalinowska-Tłuścik, Piotr
a
a
a
a
a
Popik, Agnieszka Nikiforuk, Martyna Krawczyk, Grzegorz Satała, Tomasz Lenda, Joanna
c
a
a
a,b
d
Knutelska, Ryszard Bugno, Jakub Staroń, Wojciech Pietruś, Mikołaj Matłoka, Krzysztof
d
d
d
d
Dubiel, Rafał Moszczyński-Pętkowski, Jerzy Pieczykolan, Maciej Wieczorek, Bogusław
e
c
a
Pilarski, Paweł Zajdel, Andrzej J. Bojarski
a
Institute of Pharmacology, Polish Academy of Sciences, 12 Smętna Street, 30-343 Kraków,
Poland
b
Faculty of Chemistry, Jagiellonian University, 2 Gronostajowa Street, 30-387 Kraków, Poland
c
Faculty of Pharmacy, Jagiellonian University Medical College, 9 Medyczna Street, 30-688
Kraków, Poland
d
Research & Development Centre, Celon Pharma S.A., 41A Mokra Street, Kiełpin, 05-092
Łomianki, Poland
e
CERKO, 96/98 Zwycięstwa Avenue, 81-525 Gdynia, Poland
Corresponding Author:
E-mail: ahogen@if-pan.krakow.pl
Postal address: 12 Smętna Street, 31-343 Kraków
Phone: +48 792287874
Fax: +48 12 637 45 00

ACCEPTED MANUSCRIPT
Abstract
A new strategy in the design of aminergic GPCR ligands is proposed – the use of
heterocyclic basic moieties in place of the evergreen piperazine or alicyclic and aliphatic
amines. This hypothesis has been tested using a benchmark series of 5-HT R antagonists
6
obtained by coupling variously substituted 2-aminoimidazole moieties to the well
established 1-benzenesulfonyl-1H-indoles, which served as the ligands cores. The
crystallographic studies revealed that upon protonation, the 2-aminoimidazole fragment
triggers a resonance driven conformational change leading to a form of higher affinity.
This molecular switch may be responsible for the observed differences in 5-HT R
6
activity of the studied chemotypes with different amine-like fragments. Considering the
multiple functionalization sites of the embedded guanidine fragment, diverse libraries
were constructed, and the relationships between the structure and activity, metabolic
stability,
N-(1H-imidazol-2-yl)acylamide chemotype (10a–z) exhibited high affinity for 5-HT
and very high selectivity over 5-HT1A, 5-HT2A, 5-HT and D receptors (negligible
and
solubility
were
established.
Compounds
from
the
6
R
7
2
binding), which was attributed to their very weak basicity. The lead compound in the
series 4-methyl-5-[1-(naphthalene-1-sulfonyl)-1H-indol-3-yl]-1H-imidazol-2-amine (9i)
was shown to reverse the cognitive impairment caused by the administration of
scopolamine in rats indicating procognitive potential.
Keywords

ACCEPTED MANUSCRIPT
2
-aminoimidazole; serotonin; 5-HT antagonist; non-basic; GPCR; Alzheimer’s disease.
6
Introduction
Until now, the chemical space of ligands of aminergic G protein-coupled receptors
GPCR) has been expanding mainly by the employment of novel bioisosteric building
(
blocks, resulting in a large diversity of core scaffolds, aromatic systems, linkers,
hydrogen bond donors and acceptors. This is in sharp contrast to the very narrow pool of
amine-like groups that have been used to replace the aminoalkyl chains of endogenous
neurotransmitters and classical ligands. The commonly used bioisosteres include
piperazine,[1] piperidine,[2] tetrahydropyridine,[3] morpholine, pyrrolidine[4] and
azetidine. More sophisticated saturated heterocyclic and cycloalkyl structures include
spirocyclic amines[5,6] and bicyclic amines.[7] Interestingly, hardly any attempts have
been made to employ aromatic basic groups in the design of aminergic GPCR ligands. In
our previous paper, we characterized a series of 5-(1H-indol-3-yl)-1-alkylimidazoles
7
which were potent and selective 5-HT receptor agonists.[8] This discovery supported
the idea of introducing aromatic basic moieties into serotonin receptor ligands.
2-Aminoimidazole (2-AI) remains an underexplored basic scaffold in the GPCR ligand
field, and it has been found to be a common molecular framework of numerous marine
alkaloids and synthetic antibacterial (anti-biofilm) agents. The serotonergic activity of
marine alkaloids containing the 2-AI motif was revealed in 1984, well before the
discovery and cloning of several subtypes of serotonin receptors (Figure 1).[9,10]

ACCEPTED MANUSCRIPT
Affinity of some 2-AI alkaloids for other GPCR’s was also reported in the following
years.[11–14]
Figure 1. Marine alkaloids exhibiting aminergic GPCR activity. The compounds
keramadine[9] and hymenidine,[10] at 15 µM, reversed the contractile effect of 1 µM
serotonin on an isolated rabbit aorta, while the contraction caused by KCl or
noradrenaline was not affected. The compounds archerine[11] and taurodispacamide
A[12] at high concentrations blocked the effects of histamine in a screening assay
conducted on the guinea pig ileum. The compound hymenin[13] acted as a competitive
antagonist of α-adrenoreceptors in vascular smooth muscles in a screening assay.
Dibromophakellin showed agonistic activity against α2B adrenoreceptor with an EC of
50
4
.2 µM.[14]

ACCEPTED MANUSCRIPT
Contrary to their well-recognized anti-biofilm properties, synthetic
-aminoimidazole derivatives have not been shown to interact with aminergic GPCRs, or
2
with other monoamine binding CNS-related targets. The widespread occurrence of 2-AI
in natural bioactive compounds and the previous use of this fragment in medicinal
chemistry suggest that 2-AI might serve as a guanidine mimic that can be chemically
modified to have the desired basicity (Figure 2).[15]
Figure 2. Comparison of the calculated pK of aminoimidazoles, 2-aminothiazole and
a
other cyclic amidines vs. the classical 5-HT
were made for structures containing the standardized 5-HT
-benzenesulfonyl-1H-indole. A comparison of the basicity of various amidines can be
found in a review by Sullivan et. al.[15]
6
R ligand amine groups. The calculations
6
R ligand core:
1
5
-HT is a G coupled serotonin receptor that is involved in regulating cholinergic
6 s
transmission.[16,17] This receptor has long been a hope for Alzheimer's disease patients

ACCEPTED MANUSCRIPT
since its antagonists were shown to significantly relieve the disease symptoms in animal
models, possibly by restoring the physiological acetylcholine levels in the brain.[18]
Numerous structurally diverse antagonists were reported, a majority of which contributed
to the arylsulfonyl chemotypes.[19–22] The most advanced of several clinical candidates,
Idalopirdine (LuAE58054) did not show convincing activity in phase III clinical trial.[23]
The development of Intepirdine (SB-742457) and Cerlapirdine (WAY-262,531) have
been halted. An 1-arylsulfonyl-1H-indole derivative SUVN-502 was shown safe and
well-tolerated during phase I clinical trial.[1,24] Among the clinical leads, AVN-211
holds a remarkable place as an antagonist developed for treating schizophrenia, and it
was confirmed to relieve positive symptoms while producing procognitive
effects.[25,26]
The primary objective of this work was to validate the usefulness of 2-AI as a basic
fragment in aminergic GPCR ligands. Once this was achieved by employing 2-AI as a
bioisostere of the classical piperazine/alkylamino fragment in a series of potent and
6
selective 5-HT R antagonists, the next goal was to tune the ADME properties of the
series to find suitable lead compounds. Differently functionalized 2-aminoimidazoles
were used to obtain compounds based on a 2-aminoimidazole scaffold while pursuing
the optimal solubility, activity and metabolic stability.
Results
Chemistry

ACCEPTED MANUSCRIPT
5
-Aryl-2-aminoimidazoles can be synthesized using several different protocols,
although only two were found practical for the compounds that we designed and were
tolerant of broad scope of reagents applied in the present study. Both synthetic methods
rely on the condensation of bromoketones with reagents carrying the guanidine synthon.
The
first
method
involves
the
formation
salts
of
5
2
2
-hydroxy-1H,2H,3H-4λ -imidazo[1,2-a]pyrimidin-4-ylium
from
-aminopyrimidine derivatives and haloketones.[27–29] Strict control of the reaction
conditions leads to either an alcohol or an aromatized product (Scheme 1). The
nucleophilic addition of hydrazine and the subsequent cleavage of the pyrimidinium ring,
which results in the formation of pyrazole lead to substituted or unsubstituted
2
-aminoimidazole derivatives.
Scheme 1. Two approaches for synthesizing 5-aryl-2-aminoimidazoles. Both pathways
use bromoketones as starting materials.
The cleavage of an aromatized salt yields a simple 1,4,5-substituted 2-aminoimidazole
product (13a), whereas the cleavage of a non-aromatic salt leads to a rearrangement that

ACCEPTED MANUSCRIPT
resembles the Dimroth mechanism, giving rise to 2,4,5-substituted 2-aminoimidazole
(
13b-d, Scheme 2).[27] The acylated and unsubstituted 2-AI’s could be concisely
prepared from acylguanidines, via condensation with bromoketones and subsequent
deprotection with acid. Acylguanidines can be prepared from guanidine free base and
esters.[30]
Scheme 2. Approaches to alkyl-substituted 2-aminoimidazole synthesis. The order of
aromatization and hydrazine addition is crucial to the structure of the product. A
Dimroth-type rearrangement may be responsible for the alkyl group migration.[27]
The initial experiments involved the condensation of unprotected indol-3-yl based
bromoketones with amidine derivatives that did not yield the desired
products.Condensation of unprotected 2-bromo-1-(1H-indol-3-yl) propan-1-one with
acetylguanidine did not yield the desired 2-aminoimidazole derivative, instead the
+
formation of a product M /Z=412 was observed which could potentially be explained by
the dimerization of the substrate and condensation with acetylguanidine.
Thus, the indoles were first protected with the appropriate arylsulfonyl groups

ACCEPTED MANUSCRIPT
giving rise to compounds 1a–h; the 3 position was next functionalized via Friedel-Crafts
acylation that employed acid anhydrides and aluminum chloride, yielding 2a–l (Scheme
3). The final compounds that were based on 4- and 6-imidazole-substituted
1-arylsulfonylindole scaffolds (9l, 9m and 10x–z) were synthesized starting from the
appropriate indolenitriles (Scheme 3, procedure 3). The nitriles were treated with
alkyllithiums (Grignard reagents could potentially be used instead but were found less
suitable in reagent screening) followed by hydrolysis in dilute hydrochloric acid.[31] The
resulting 4- and 6-acylindoles 3a–c were protected with an arylsulfonyl group via phase
transfer catalyzed on water reaction with arylsulfonyl chlorides yielding 4a–c. A mild
2
bromination using CuBr and ethyl acetate-chloroform mixture was found to be a very
convenient way to produce the desired bromoketones 5a–o in high yields, and it has a
very simple work-up.[32] It was found that only the acetylindole derivatives (2a–c and
4
a) were prone to overbromination, whereas the α-substituted ketones could be reacted
with an excess of CuBr2, which resulted in 100% conversion.
-[1-(Benzenesulfonyl)-2-methyl-1H-indol-3-yl]-2-bromopropan-1-one 5p, which could
not be synthesized by direct bromination was obtained via Friedel-Crafts acylation of
a
1
1-benzenesulfonyl-2-methyl-1H-indole (1i) with bromopropionyl bromide (Scheme 3).
The acylguanidines 6a–c were obtained from the appropriate ethyl esters and freshly
prepared guanidine free base. The alkyl substituted 2-aminoimidazoles were synthesized
using modified microwave-assisted method.[28] The starting
a

ACCEPTED MANUSCRIPT
N-alkylpyrimidin-2-amines 7a–c were prepared from 2-chloropyrimidine and the
appropriate amines (for low nucleophilicity amines, such as 2,2-difluoroethylamine long
reaction times were required).[33]
The condensation of bromoketones with N-alkylpyrimidin-2-amines was conducted
in pressure tubes immersed in an oil bath and led to 8a, 8c, 8d and 12a–c. Screening
experiments revealed that acetonitrile was the solvent of choice, and that the highest
temperature at which the product did not aromatize was 80°C. Compound 8b was
synthesized in a similar manner, starting from 2-aminopyridine.
2-Acetamidoimidazoles 10a–z were conveniently prepared via the condensation of
bromoketones with a large excess of acylguanidines 6a–c in warm DMF.[34] The
products were often easily precipitated by the addition of water, which solubilized the
unreacted acylguanidine while the byproducts were simply triturated out with ethyl
acetate or acetone. Compounds 10a, 10b, 10d, 10f–h, 10n, 10o and 10y were converted
into hydrochloride salt. The unprotected 2-aminoimidazoles 9a–m were prepared from
their acetylated counterparts (10a–c, 10e–g, 10j, 10m, 10p, 10s, 10u and 10y), by
hydrolysis in boiling dilute hydrochloric acid. The resulting hydrochlorides 9a–m could
not be converted back to the free bases, as a chemical reaction occurs at pH>9. This
finding was in contrast to the more chemically stable 2-acetamidoimidazoles 10a–z,
which remain intact in highly basic solutions.
1
H,2H,3H-Imidazo[1,2-a][1,3]diazole
11a
and
11c
and

ACCEPTED MANUSCRIPT
5H,6H,7H,8H-imidazo[1,2-a]pyrimidine 11b were synthesized via cyclization from
2-acetamidoimidazoles and 1,2-dibromoethane or 1,3-dibromopropane, respectively
(
Scheme 3, procedure 11). Both possible isomers were detected by LC-MS and the peak
integral ratio was 1:5 however, only the major peak was isolated, and its structure was
determined via 2D NMR. The deprotection of 11a with hydrochloric acid yielded 11c.
2-Aminothiazoles 14a–e were synthesized from thiourea and bromoketones (Scheme 3,
procedure 12).

ACCEPTED MANUSCRIPT
Scheme 3. Synthetic figure for the final 2-aminoimidazole and 2-aminothiazole
compounds. The final compounds, 8a–c, 9a–m, 10a–z, 11a–c, 13a–d and 14a–e, were

ACCEPTED MANUSCRIPT
synthesized from common substrates, i.e., bromoketones via condensation with
embedded amidines. Reagents and conditions: Procedure 1: 1. NaH, THF, 2. arylsulfonyl
chloride; procedure 2: AlCl , DCM, acid anhydride; procedure 3: 1. Alkyllithium, THF,
3
-
78°C 2. conc. HCl, reflux; procedure 4: NaOH, TBAB, arylsulfonyl chloride, H O,
2
toluene; procedure 5: AcOEt and CHCl (1:1), CuBr , reflux; procedure 6: EtOH, r.t.;
3
2
procedure 7: EtOH, Et N, MW; procedure 8: 2-aminopyridine or 2-aminopyrimidine,
3
MeCN, 120°C; procedure 9: conc. HCl, reflux; procedure 10: acylguanidine, DMF;
procedure 11: Cs
2
CO
2: MeCN, DMAP, 80°C; procedure 13: MeCN, N
00W; procedure 14: thiourea, MeCN, 40°C.
3
, DMF, BrCH
2
CH
2
Br or BrCH
2
CH
2 2
CH Br, 60°C; procedure
1
2
2
H
4
xH
2
O, MW: 10 minutes, 100°C,
Structure-affinity and structure-ADME relationships
The pilot compounds 9a (R7 = H, Table 1) and 9b (R7 = Me), which were structurally
the simplest within the series showed high 5-HT R affinity (K = 10 nM and 14 nM,
6
i
respectively) and moderate selectivity over 5-HT , 5-HT and D receptors, 9b being
1
A
2A
2
more selective over 5-HT R and D R. Extending the alkyl chain at the 5th position of
1
A
2
the imidazole to an ethyl decreased the potency even further (9c, R7 = Et, K = 29 nM).
i
The incorporation of a methoxy group at the 5th position of the indole nucleus resulted
in a significant enhancement in binding affinity; compounds 9d (K = 2 nM) and 9e (K =
i
i
3
nM) were the most potent derivatives within the series.[35] The change from
benzenesulfonyl group to 3-fluorobenzenesulfonyl (9d and 9e vs. 9g and 9h),

ACCEPTED MANUSCRIPT
1
-naphthylsulfonyl (9i, 9j) and 2-naphthylsulfonyl (9k) resulted in lowered affinity. The
shift of the 2-aminoimidazole moiety to the 4 position of the indole diminished the
affinity roughly threefold (9a vs 9l). A major enhancement of 5-HT R binding was
6
observed with the shift of the 2-aminoimidazole to the 6 position of the indole in
compound 9m (K = 4 nM), but this was accompanied by a sharp drop in solubility
i
(
which was in the 1–10 mg/mL range for 9m).
Table 1. Structure and binding data of compounds 9a–m. The amine-like group within
the chemotype is 1H-imidazol-2-amine (2-aminoimidazole).
K [nM]
i
IM
*
ID
R1
R5
R7
5
-HT1A
5-HT2A
891±114
100±13
107±9
5-HT6
10±2
14±3
29±4
2±1
5-HT7
2455±377
9120±842
3467±403
1148±178
7079±947
N.A.
D2
9
9
a
Ph
Ph
H
H
Me
Et
3
3
3
3
3
3
3
126±17
N.A.
N.D.
N.A.
N.A.
N.A.
N.D.
87±7
b
H
794±81
1585±186
1202±173
N.D.
9
c
Ph
H
9
d
Ph
OMe
OMe
OMe
OMe
H
794±155
676±82
661±97
N.D.
9
e
Ph
Me
Et
3±1
9
f
Ph
19±2
9±2
1622±99
7079±802
9
g
3-F-Ph
H
N.D.

ACCEPTED MANUSCRIPT
9
h
3-F-Ph
1-Naph
1-Naph
2-Naph
Ph
OMe
H
Me
Me
Me
Me
H
3
3
3
3
4
6
N.A.
N.A.
N.A.
N.D.
N.D.
N.A.
1122±215
490±52
9±1
8±2
N.A.
2512±303
1096±91
3715±411
N.A.
N.D.
9
i
1175±144
1318±162
2344±427
3981±511
708±65
9
j
OMe
OMe
H
1660±201
1047±173
2291±359
1175±93
10±3
9±2
9
k
9
l
34±4
4±1
9
m
Ph
H
Me
776±63
*
IM – position at the indole being the 2-aminoimidazole attachement point. N.D. – not
determined N.A. – not active, K > 10000 nM
i
The intermediate 10a (Table 2), which was used to obtain compound 9a (procedures
1
0 and 9 from Scheme 4), exhibited considerable affinity for the 5-HT receptor (K = 79
6 i
nM), despite its very low basicity (calculated pK = 3.11). Compound 10a not only was
a
considered promising (because it had less hydrogen bond donors and higher
lipophilicity), but its lower basicity contributed to a very high selectivity over the related
targets at the cost of low solubility. For most of the 1-(arylsulfonyl)-1H-indole cores,
triplets of analogs namely an unsubstituted 2-aminoimidazole, 2-acetamidoimidazole and
2
-propanamidoimidazole (the order complies with decreasing in vitro potency for most
entries) were synthesized. There were no sharp SAR trends among compounds 10b–10i;
0e had the highest affinity (K = 10 nM), and 10b, 10f and 10g exhibited very high
1
i
metabolic stability (Table 7). In general, compounds 9a–m showed fine selectivity, and
the acylated chemotype (compounds 10a–z) provided ultimate selectivity i.e. did not
bind to targets other than 5-HT R, which could potentially be attributed to their very low
6
basicity.

ACCEPTED MANUSCRIPT
A comparison of the highly basic naphthylsulfonyl-indole-based compounds 9i, 9j
and 9k (9 nM ≤ K ≤ 10 nM), and their low-basicity counterparts: 10p, 10r, 10s, 10t and
i
1
0u (10 nM ≤ K ≤ 15 nM and 23 nM, respectively) clearly showed that the activity of
i
the naphthyl derivatives was not greatly affected by the 5-methoxy substituent on indole
or by the switch from 1-naphthyl to 2-naphthyl. The only compound that did not follow
this pattern was 10w (K = 455 nM), which might be beyond the capacity limits of the
i
receptor binding pocket. The substitution at the 3 position of the phenylsulphonyl moiety
with a fluorine atom was investigated in the series 10n, 10o (vs 9e, 9g), and 9g, 9h (vs
9
d, 9e) and neither produced any significant effect on the activity in the case of
low-basicity compounds nor enhanced the metabolic stability, while 9g and 9h (K = 8
i
nM and 9 nM, respectively) were found approximately three-fold less potent than 9d and
9
e. The depth of the pocket housing the acyl group in the low-basicity compounds was
determined by the synthesis of 10j (K = 3 nM). The chemical space was expanded with
i
1
0y and 10z, the indol-6-yl-based compounds (K = 15 nM and 17 nM, respectively).
i
Table 2. Structure and binding data of the compounds 10a–z. The amine-like group
within the chemotype is N-(1H-imidazol-2-yl)acylamide (2-acylamidoimidazole).

ACCEPTED MANUSCRIPT
pK [M]
i
*
ID
R1
R2 R5
R7
R10
IM
5
-HT1A 5-HT2A 5-HT6
5-HT7
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.D.
N.A.
N.A.
N.D.
D2
1
0a
0b
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
H
H
H
H
Me
Et
Ac
Ac
Ac
Ac
Ac
Ac
3
3
3
3
3
3
3
3
3
3
3
N.A.
N.A.
N.D.
N.A.
N.A.
N.A.
N.D.
N.D.
N.A.
N.D.
N.D.
N.A.
N.D.
N.A.
79±12
25±4
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
1
1
0c
0d
0e
0f
H
H
N.A. 263±31
1
H
OMe
OMe
OMe
OMe
OMe
H
H
N.D.
N.A.
N.A.
N.D.
N.D.
N.D.
N.A.
18±3
10±2
31±5
14±2
16±3
83±5
3±1
1
H
Me
Et
1
H
1
0g
0h
H
Me
Et
CH
CH
3
3
CH
CH
Ac
2
2
CO
CO
1
H
1
0i
0j
0k
Me
H
Me
Me
H
1
OMe
F
2
PhCH CO
1
Me
Ac
N.A. 759±92
7943±1 107±8 8511±743 N.D.
52
1
0l
Ph
Me
F
H
CH
3
CH
2
CO
3
7
1
0m 3-F-Ph
H
H
H
H
H
H
H
H
OMe
OMe
OMe
H
H
Ac
Ac
CH
Ac
CH
Ac
CH
Ac
3
3
3
3
3
3
3
3
N.D.
N.A.
N.A.
N.D.
N.D.
N.D.
N.D.
N.D.
N.A.
N.A.
N.A.
N.D.
N.A.
N.A.
N.D.
N.A.
50±7
14±3
9±2
N.A.
N.A.
N.A.
N.D.
N.A.
N.A.
N.D.
N.A.
N.A.
N.A.
N.A.
N.D.
N.A.
N.A.
N.D.
N.A.
1
0n 3-F-Ph
0o 3-F-Ph
0p 1-Naph
Me
Me
Me
Me
Me
Me
Me
1
CH
CH
CH
3
3
3
2
2
2
CO
CO
CO
1
13±1
15±4
10±2
10±1
23±3
1
0r 1-Naph
0s 1-Naph
H
1
OMe
OMe
OMe
1
0t 1-Naph
1
0u 2-Naph

ACCEPTED MANUSCRIPT
1
0w 2-Naph
H
H
OMe
H
Me
Me
CH
3
3
CH
2
2
CO
CO
3
4
N.D.
N.A.
N.D.
N.A. 455±29
N.A. 501±73
N.A.
N.D.
N.D.
N.A.
N.D.
1
1
0x
0y
Ph
Ph
CH
CH
N.A.
15±3
5248±
H
H
H
H
Me
Me
Ac
6
6
3
14
N.D.
7413±1 17±4
77
N.D.
3631±
562
1
0z
Ph
CH
3
CH
2
CO
4
*
IM – position at the indole being the N-(1H-imidazol-2-yl)acetamide attachement point.
N.D. – not determined N.A. – not active, K > 10000 nM
i
i i
Bicyclic 2-acetamidoimidazoles 11a (K = 585 nM Table 3) and 11b (K = 280 nM)
exhibited low affinity and extremely low water solubility. Compound 11a was insoluble
in most organic solvents and crystallized readily from warm DMSO. This scaffold was
investigated in an effort to optimize the ADME and receptor affinity but was
discontinued because no improvement in affinity was observed while the solubility
decreased steeply.
Table 3. Structure and binding data for compounds 11a–e. The amine-like group within
the chemotype is 1H,2H,3H-imidazo[1,2-a][1,3]diazole in 11a and 11c
5
H,6H,7H,8H-imidazo[1,2-a]pyrimidine in 11b.
ID
R7
R10
X
pK [M]
i

ACCEPTED MANUSCRIPT
5
-HT1A
N.A.
N.A.
N.A.
5-HT2A
N.A.
5-HT
6
5-HT
7
D
2
1
1a
Me
Me
Me
Ac
Ac
H
CH
CH CH
CH
2
585±157
280±61
93±18
N.A.
N.A.
N.A.
N.A.
N.A.
11b
2
2
N.A.
*
11c
2
5129±891
1230±243
*
i
Compound 11c was obtained and tested as the hydrochloride. N.A. – not active, K >
10000 nM
Compound 8a (structurally similar to 11a–c) (Table 4) which served as an
intermediate in the synthesis of 9a exhibited considerable binding affinity for 5-HT R (K
6
i
=
50 nM) and high selectivity over the related targets. This rather serendipitous discovery
encouraged
us
to
explore
the
receptor
affinity
of
1,3-substituted
5
1
H-4λ -imidazo[1,2-a]pyrimidin-4-ylium salts. Although they had better solubility than
did 11a–c, the 8a derivatives, 8b (78 nM) and 8c (250 nM), did not exhibit any
improvement in binding affinity, 8a did not have good chemical properties (e.g., highly
susceptible to nucleophiles) and was later found to be metabolically unstable (Table 7)
thus, the chemotype was not extended any further.
Table 4. Structure and binding data for compounds 8a–c. The amine-like group within
the chemotype is imidazo[1,2-a]pyrimidine of 8a and 8c and imidazo[1,2-a]pyridine in
8b.

ACCEPTED MANUSCRIPT
p
Ki [M]
ID
R5
R7
X
5
-HT1A
N.A.
N.D.
N.A.
5-HT2A
N.A.
5-HT6
50±6
5-HT7
N.A.
N.A.
N.A.
D2
N.D.
8
8
a
H
H
N
C
N
b
OMe
OMe
Me
Me
N.A.
78±9
3162±528
N.A.
8
c
N.A.
250±17
N.D. – not determined, N.A. – not active, K > 10000 nM
i
Several N-alkylated derivatives of 2-aminoimidazole were also synthesized,
including the fluorinated compounds 13a (K = 157 nM) and 13b (K = 21 nM, Table 5).
i
i
The isomers 13a and b displayed highly different activities and compound 13b with the
exocyclic nitrogen substituted was more potent. Compound 13c was found to be
significantly less potent than its non-basic counterpart 10j.
Table 5. Structure and binding data for compounds 13a–d. The amine-like group within
the chemotype is a N-alkyl-1H-imidazol-2-amine in 13a and 1-substituted
1
H-imidazol-2-amine in 13b–d.

ACCEPTED MANUSCRIPT
i
pK [M]
ID
R8
R10
5
-HT1A
N.A.
N.A.
N.A.
N.A.
5-HT2A
3236±725
1230±187
1905±423
N.D.
5-HT
6
5-HT
N.A.
N.A.
N.A.
7
D
2
1
3a
3b
CH
2
CHF
2
H
157±41
21±4
3802±586
3981±819
2138±414
1
H
CH
2
CHF
2
1
3c
H
H
EtPh
83±6
1
3d
Bn
21±3
2455±319 1175±101
N.D. – not determined
-Aminothiazoles 14a–e (Table 6) were found to be active, as the switch to this
2
amine-like moiety neither decreased the 5-HT
6
receptor binding nor yielded any hit
= 44 nM.
compounds. The most potent thiazole 14c had a K
i
Table 6. Structure and binding data for compounds 14a–e. The amine-like group in this
chemotype is 1,3-thiazol-2-amine (2-aminothiazole).
ID
R1
R5
pK [M]
i

ACCEPTED MANUSCRIPT
5
-HT1A
N.D.
N.D.
N.D.
N.A.
N.A.
5-HT2A
2455±612
N.D.
5-HT
6
5-HT
N.A.
N.D.
N.D.
N.A.
N.A.
7
D
2
1
4a
4b
Ph
OMe
OMe
OMe
H
105±19
65±5
N.A.
N.D.
1
2-Naph
1-Naph
1-Naph
Ph
1
4c
4d
4e
N.D.
44±3
N.D.
1
N.A.
69±8
N.A.
1
H
N.D.
100±13
5129±1158
N.D. – not determined
The metabolic stability, which was a key factor used to establish the lead compounds
was assessed in an in vitro experiment using rat liver microsomes. Compounds 10b, 10f,
10g were found to be very stable (Table 7). Most of the tested 2-acylamidoimidazoles
exhibited pronounced stability, when compared to compounds 8a, 9i and 9m. Only
compounds 8a, 10a and 9m were found to have lower stability than the benchmark
compound Idalopirdine (LuAE58054).
Table 7. Metabolic stability: clearance (rat liver microsomes) and half-lives of selected
compounds.
ID
Clearance [µL/min/mg]
t1/2 [min]
8
a
230.3
60.3
104
6
9
i
23
13
<3
9
m
1
0a
>460

ACCEPTED MANUSCRIPT
1
1
0b
0d
7.3
190
32
43.0
10.8
17.1
43.4
53.5
35.4
39.0
1
0f
129
81
1
0g
1
1
0h
0n
32
26
1
0o
39
1
3b
36
Idalopirdine
Intepirdine
Donepezil
99.3
6.6
N.C.
N.C.
79
17.6
N.C. – not calculated
Physicochemical descriptors (i.e., ClogP, pK and TPSA) were calculated using
a
ChemAxon Instant JChem and the water solubility range was determined experimentally
for compounds 9b, 9d, 9i, 10b, 10y and 13b (Table 8). The selected compounds were
screened for several anti-targets including adrenergic, dopamine, histamine, muscarinic
and serotonin receptors and the level of binding to the hERG potassium channel was
measured (Table 8). The low binding of all of the tested compounds to hERG suggests a
low risk of cardiac side effects mediated by this channel. Compounds 9b and 9d showed
marginal binding, and 10b,and 9i did not exhibit any functional activity at 5-HT2BR. The
functional assays clearly demonstrated the ability of compounds 10b, 10y and 9i to block
6 b
the 5-HT R at low concentration. The K constants were calculated based on the

ACCEPTED MANUSCRIPT
observed inhibition of cAMP production (induced by 5-CT) at different concentrations of
the tested compound. The values obtained were consistent with the results of the binding
experiments.
Table 8. Calculated descriptors, solubility ranges and anti-target profiles of the selected
compounds.
*
*
9
b
9d
2.15
103.0
368.41
8.60
high
N.D.
33.1
N.D.
N.D.
N.D.
6.7
9i
3.96
93.77
402.47
8.64
medium
4.8
10b
10y
13b
3.58
89.01
446.47
7.47
medium
13
CLogP
TPSA
M.W.
2.44
93.77
352.41
8.41
high
N.D.
12.2
N.D.
N.D.
N.D.
3.8
2.50
96.85
394.45
2.90
low
2.50
96.85
394.45
3.10
pKa
Solubility
very low
3.1
5
-HT R K [nM]
1.3
6
b
-6
**
alpha1A (% inh. at 10 M)
10.7
70.1
-1.3
-1.5
-4.2
4.0
-6
**
(
% inh. at 10 M)
-5.9
-1.9
1.0
-5
H (h)
1
(%agonist at 10 M)
N.D.
N.D.
-13.4
N.D.
22.4
N.D.
N.D.
N.D.
N.D.
-10.7
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
-4.0
-5
(
% antagonist at 10 M)
25.7
-0.3
-
6
**
**
M (h) (% inh. at 10 M)
1
-6
M (h) (% inh. at 10 M)
7.6
5.6
N.D.
52.7
21.3
50.3
N.D.
28.0
N.D.
N.D.
5
-6
**
(
% inh. at 10 M)
71.6
N.D.
N.D.
82.0
N.D.
N.D.
-5
5
-HT2C (h)
(% agonist at 10 M)
-5
(
% antagonist at 10 M)

ACCEPTED MANUSCRIPT
-
6
**
hERG (h) (% inh. 10 M)
N.D.
N.D.
4.5
N.D.
N.D.
1.7
25.7
-2.8
2.4
-1.9
N.D.
N.D.
N.D.
N.D.
23.0
N.D.
N.D.
-6
**
Function 5-HT R (h) (% inh. 10 M)
2
B
-6
**
5
*
-HT2B (h) (% inh. at 10 M)
N.D.
Compound was tested as hydrochloride.
*
*
Experiments were performed at Eurofins Cerep. Water solubility ranges: high >10
mg/mL; medium 1–10 mg/mL; low 0.1–1 mg/mL; very low <0.1mg/mL.
Physicochemical descriptors: partition coefficient (ClogP), topological surface area
(
TPSA), molecular weight (M.W.) and acidity dissociation constant (pK ) were
a
calculated using ChemAxon software.
Lead compound selection
The non-basic chemotypes (i.e., 10, 11 and 14) were omitted from further
development due to insufficient water solubility. The highest water solubility was
observed for compounds from chemotypes 8 and 9. Compound 9b showed mediocre
blood brain barrier permeability in mice (unpublished data). Compound 9i was thus
chosen as the lead compound based on the calculated physicochemical descriptors (i.e.,
higher lipophilicity than 9b), in vitro activity, solubility data and metabolic stability
assay results.
The cytotoxicity of compound 9i was assessed in the HepG2 cell line by measuring
both cell membrane damage and cell viability. The compound was mildly cytotoxic (i.e.,
it had an IC = 15.7 µM in the PrestoBlue assay) and damaged cell membranes at a
50

ACCEPTED MANUSCRIPT
similar concentration level (i.e., 18.4 µM for Toxi Light assay). Moreover, 9i was found
to be non-mutagenic in a mini AMES test up to 10 µM and inhibited CYP3A4 and
CYP2D6 at a concentration of 1.7 µM and 6.4 µM, respectively. The compound was
non-toxic in female rats (LD > 2000 mg/kg) with no apparent adverse effects observed
5
0
at the highest dose.
In vivo pharmacology
The procognitive properties of compound 9i were determined; it had the potential to
reverse the scopolamine-induced novel object recognition (NOR) impairment, which
serves as a model of Alzheimer’s disease cholinergic system failure. There were no
significant differences in the time spent exploring two identical objects in the
familiarization phase in any group (data not shown).
One hour after the inter trial interval (ITI), vehicle-treated, but not scopolamine-treated
rats spent significantly more time exploring a novel object than they spent exploring a
familiar one. Thus, the administration of scopolamine (1.25 mg/kg) abolished the ability
to discriminate novel and familiar objects in a recognition (T2) trial. This deficit was
reduced by administering compound 9i (at 1 but not 0.3 mg/kg) and donepezil (at 1 but
not 0.3 mg/kg).
It has been concluded that the lack of efficacy of the current Alzheimer’s disease
treatments may be overcome with the use of combined therapy. One of the latest

ACCEPTED MANUSCRIPT
approaches involves the administration of a 5-HT R antagonist together with an
6
acetylcholinesterase inhibitor.[36] Thus, the efficacy of 9i as a part of an 5-HT R
6
antagonist-acetylcholinesterase inhibitor cocktail was tested. Co-administration of
inactive doses of compound 9i (0.3 mg/kg) with donepezil (0.3 mg/kg), facilitated
cognitive performance. Because the analyses of exploration time during the recognition
trial and the discrimination index (DI) yielded the same results, only the DI data are
presented (Figure 3). The DI data were analyzed by a one-way ANOVA followed by a
Newman–Keul’s post hoc test.
Figure 3. Effects of 9i, donepezil and their combined administration on
scopolamine-induced cognitive impairment in rats. Animals were treated with the
experimental compounds and scopolamine at 120 and 30 min before the learning trial,
and were tested 1 hour later in the recognition trial. The data are expressed as the mean ±
standard error of the mean of the discrimination index (DI). N = 9–10 animals per group.

ACCEPTED MANUSCRIPT
Symbols: ***p < 0.001, a significant reduction in the DI compared with the
vehicle-treated group; # p < 0.05, ### p < 0.001, a significant increase in the DI
compared with the scopolamine-treated group. Note that one rat from the scopolamine +
9i (1 mg/kg) group was excluded to fulfill the D'Agostino & Pearson and Shapiro-Wilk
normality tests criteria (Prism 7 software).
Determination of the basicity of compounds 9i and 10b
In an acidic solution, 2-aminoimidazole could be expected to undergo a two stage
protonation:
+
+
2
-AI +H = 2-AIH
pK1
+
+
2+
2
-AIH +H = [2-AIH ]
pK2
2
Based on the presented equilibria (protonation/deprotonation) two dissociations
constants should be expected: pK for NH group and pK for N(1) in imidazole ring
system. The experimental data reports only one base centre with pK = 8.46 for
-aminoimidazole[37] and pK = 7.18 for 2-aminobenzimidazole suggesting the
1
2
2
a
2
a
equivalence of N(1) and (N3) nitrogen atoms in imidazoline ring system (Supporting
Information figure S1). Due to a very high stability of the 2-aminoimidazolium cation,
the second protonation is not observed in aquous solutions and requires the action of
superacids.[38] The difference between pK of imidazole and the suggested imidazoline
a
moiety (∆pK =1.46) for 2-aminoimidazole vs imidazole and (∆pK =1.73) for
a
a

ACCEPTED MANUSCRIPT
2-aminobenzimidazole vs benzimidazole, let us postulate imidazoline tautomeric forms
for both 2-aminoimidazole and benzimidazole derivatives.
Results from the titration curve of 9i (Figure 4) confirm that its acid-base properties are
similar to imidazole and 2-aminobenzimidazole (Supporting Information figure S1).
Contrary to 9i, compound 10b seemed to be resistant to protonation which could be
explained as a result of acetylation of the amine group at (2-NH ) position resulting in
2
formation of an intramolecular hydrogen bond leading to very low solubility. The
relative order of the pK values was as follows:
a
pK
a 10b< pKa benzimidazole< p
Ka 9i < pKa imidazole < pKa 2-aminobenzimidazole < pKa 2-aminoimidazole
1
2
1
0
9
8
7
6
5
4
3
2
1
1
(9i)
(10b)
0
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6
Figure 4. Titration curve pH = f(VKOH) of 9i (0.0025 M) +HCl (0.004 M), V = 4 mL and
0
1
0b (0.0044 M) + HCl (0.008 M), V = 4.5 mL as titrand D in tertiary solvent system
0
H O/DMSO/Polysorbate 80 with 0.098 M KOH as titrant T.
2
We have also considered the influence of tautomeric equilibria on the pK values of the
a