Discovery of novel benzofuran-based compounds with neuroprotective and immunomodulatory properties for Alzheimer's disease treatment

Article abstract of DOI:10.1016/j.ejmech.2019.05.080

To address the multifactorial nature of Alzheimer's Disease (AD), a multi-target-directed ligand approach was herein developed. As a follow-up of our previous studies, a small library of newly designed 2-arylbenzofuran derivatives was evaluated towards ch

Full text of DOI:10.1016/j.ejmech.2019.05.080

Accepted Manuscript
Discovery of novel benzofuran-based compounds with neuroprotective and
immunomodulatory properties for Alzheimer's disease treatment
Serena Montanari, Ali Mokhtar Mahmoud, Letizia Pruccoli, Alessandro Rabbito,
Marina Naldi, Sabrina Petralla, Ignacio Moraleda, Manuela Bartolini, Barbara Monti,
Isabel Iriepa, Federica Belluti, Silvia Gobbi, Vincenzo Di Marzo, Alessandra Bisi,
Andrea Tarozzi, Alessia Ligresti, Angela Rampa
PII:
S0223-5234(19)30498-2
DOI:
https://doi.org/10.1016/j.ejmech.2019.05.080
EJMECH 11390
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 11 April 2019
Revised Date: 21 May 2019
Accepted Date: 29 May 2019
Please cite this article as: S. Montanari, A.M. Mahmoud, L. Pruccoli, A. Rabbito, M. Naldi, S. Petralla,
I. Moraleda, M. Bartolini, B. Monti, I. Iriepa, F. Belluti, S. Gobbi, V. Di Marzo, A. Bisi, A. Tarozzi,
A. Ligresti, A. Rampa, Discovery of novel benzofuran-based compounds with neuroprotective and
immunomodulatory properties for Alzheimer's disease treatment, European Journal of Medicinal
Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.05.080.
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
Graphical abstract

ACCEPTED MANUSCRIPT
DISCOVERY OF NOVEL BENZOFURAN-BASED COMPOUNDS WITH
NEUROPROTECTIVE AND IMMUNOMODULATORY PROPERTIES FOR
ALZHEIMER’S DISEASE TREATMENT
1
2
3
2
1
Serena Montanari, Ali Mokhtar Mahmoud, Letizia Pruccoli, Alessandro Rabbito, Marina Naldi,
1
4
1
1
4,5
Sabrina Petralla, Ignacio Moraleda, Manuela Bartolini, Barbara Monti, Isabel Iriepa,
1
1
2
1
3
Federica Belluti, Silvia Gobbi, Vincenzo Di Marzo, Alessandra Bisi, Andrea Tarozzi, Alessia
2
1
Ligresti, Angela Rampa *
1
Department of Pharmacy and Biotechnology, Alma Mater Studiorum-University of Bologna, Via
2
Belmeloro 6, 40126 Bologna, Italy; Endocannabinoid Research Group, Institute of Biomolecular
Chemistry, National Research Council, Via Campi Flegrei 34, 80078 Pozzuoli, NA, Italy;
3
Department for Life Quality Studies, Alma Mater Studiorum-University of Bologna, Corso
4
d’Augusto 237, 47921 Rimini, Italy; Departamento de Química Orgánica y Química Inorgánica,
5
Universidad de Alcalá, 28805-Alcalá de Henares, Madrid, Spain; Instituto de Investigación
Química Andrés M. del Río (IQAR), Universidad de Alcalá, 28805-Alcalá de Henares, Madrid,
Spain
Corresponding Author
*Phone: +39-051-2099710. Fax: +39-051-2099734. E-mail: angela.rampa@unibo.it.
1

ACCEPTED MANUSCRIPT
Abbreviations: Alzheimer’s disease (AD); acetylcholinesterase (AChE); butyrylcholinesterase
(BuChE); multi-target-directed ligands (MTDLs); endocannabinoid system (ECS); inducible nitric
oxide synthase (iNOS); triggering receptor expressed on myeloid cells 2 (TREM2);
lipopolysaccharide (LPS); Glyceraldehyde 6-phosphate dehydrogenase (GAPDH)
2

ACCEPTED MANUSCRIPT
Abstract
To address the multifactorial nature of Alzheimer’s Disease (AD), a multi-target-directed ligand
approach was herein developed. As a follow-up of our previous studies, a small library of newly
designed 2-arylbenzofuran derivatives was evaluated for its biological activities towards
cholinesterases and cannabinoid receptors. The two most promising compounds, 8 and 10, were
then assessed for their neuroprotective activity and for their ability to modulate the microglial
phenotype. Compound 8 emerged as able to fight AD from several directions: it restored the
cholinergic system by inhibiting butyrylcholinesterase, showed neuroprotective activity against
Aβ1-42 oligomers, was a potent and selective CB ligand and had immunomodulatory effects,
2
switching microglia from the pro-inflammatory M1 to the neuroprotective M2 phenotype.
Derivative 10 was a potent CB inverse agonist with promising immunomodulatory properties and
2
could be considered as a tool for investigating the role of CB receptors and for developing potential
2
immunomodulating drugs addressing the endocannabinoid system.
Keywords
MTDL; benzofuran scaffold; Aβ peptide; neuroprotection; CB receptors; anti-inflammatory M2
phenotype.
3

ACCEPTED MANUSCRIPT
1
. Introduction
Alzheimer’s disease (AD) is the most common form of dementia, with the strongest social impact
for incidence and cost of care: its debilitating nature leads to enormous financial and emotional
stress on patients and caregivers. AD is the sixth leading cause of death, currently affecting more
than 44 million people worldwide. Its incidence is expected to increase in the next decades due to
the increase in the average age of the world population [1].
AD is a multifactorial disorder and the mechanism underlying its pathogenesis is not completely
clear yet, although genetic and biochemical clues suggest that dysfunction of the basal acetylcholine
(ACh) forebrain signaling [2], deposition of β-amyloid protein (Aβ) [3], τ protein
hyperphosphorylation [4] and neuroinflammation elicited by activated microglia play a significant
role in the disease. Indeed, recent studies have reported that, during the activation processes,
microglia acquires different pro-inflammatory M1 and anti-inflammatory M2 phenotypes [5]. M1
phenotype, elicited by β-amyloid protein and τ hyperphosphorylation, becomes relevant at the
injury site in the later stage of AD when the neurorepair role of the M2 phenotype is dampened [5].
However, in AD, microglia may also exhibit mixed activation phenotypes [6].
To date, only palliative treatments are available for AD. The limited effectiveness of current
therapies highlights the need for intensifying research efforts devoted to the development of new
agents for preventing or slowing down the disease progression. Management of AD through the
modulation of multiple mechanisms rather than a single dysregulated pathway should have a higher
chance of success [7]. Therefore, exploring novel potential therapeutic targets appears as an urgent
need.
Cannabinoids also have attracted interest in AD because of the beneficial reduction the typical
neurotoxic events of the disease [8]. The endocannabinoid system (ECS) is generally deemed as a
4

ACCEPTED MANUSCRIPT
part of the neuroprotective endogenous mechanisms of the central nervous system (CNS) and has
been proposed as a target for the treatment of neuroinflammation. ECS is a complex system of
receptors, endogenous ligands, and enzymes catalyzing ligand formation and degradation. It helps
regulating fundamental processes in both central and peripheral nervous systems. ECS regulation
and dysregulation might counteract disturbances of such functions [9].
To date, two subtypes of cannabinoid Gi/o-coupled receptors, CB and CB , have been fully
1
2
characterized and cloned. CB is mainly present in neurons and glial cells and regulates important
1
brain functions including cognition and memory processes. Furthermore, recent studies have
revealed that CB receptor density increases in early AD and decreases during later stages. Among
1
cannabinoid receptors, the role of CB is the most controversial in terms of effects on microglial
1
activation [10]. However, recent evidence has shown that CB can promote M1 polarization of
1
monocytes/macrophages during chronic inflammatory processes [11]. Therefore, as for
macrophage-like cells, the activation of CB could be prodromal of the late microglia M1
1
phenotype in the progression of AD.
On the other hand, CB receptors are localized in immune system cells and modulate their migration
2
and the release of cytokines. Within the CNS, CB receptors are mainly located in microglia, the
2
major immune cells of the brain. Recent studies have demonstrated selective CB upregulation in
2
response to neuroinflammation, as seen during neurodegeneration. In particular, in AD patients,
CB are overexpressed in microglia cells surrounding Aβ plaques [12]. Notably, CB can reduce
2
2
neuroinflammation by promoting the M1 to M2 phenotype transformation in microglia [13-14].
To accomplish multifunctional compounds for the treatment of AD, a promising strategy is the
molecular hybridization approach, which allows achieving multitarget-direct ligands (MTDLs) able
to hit more than one biological target associated with the disease [7]. Taking this into account, we
have previously demonstrated that purposely designed benzofuran derivatives can act as MTDLs
towards targets involved in AD [15]. Indeed, the benzofuran scaffold is considered a privileged
5

ACCEPTED MANUSCRIPT
structure and a very important pharmacophore in drug discovery because of its presence in a great
number of natural products and biologically active compounds [16]. Furthermore, the benzofuran
derivatives SKF-64346 and LY320135 (Figure 1) have been shown to inhibit Aβ aggregation [17]
and act as high affinity antagonist/inverse agonist at CB receptor [18], respectively. In previous
1
papers we reported different 2-arylbenzofuran hybrid molecules [15, 19] rationally designed to
include the benzofuran scaffold and a key pharmacophore from some
acetylcholinesterase/butyrylcholinesterase (AChE/BuChE) inhibitors identified in our previous
studies. From this collection, compound 1 (Figure 1) emerged as a promising MTDL lead,
displaying both anticholinesterase activity and anti-aggregation properties towards Aβ, along with
neuroprotective effects towards human neuronal SH-SY5Y cells. Furthermore, 1 showed good
selectivity and moderate affinity for CB receptors [19].
1
Figure 1. Design of the lead compound 1.
6

ACCEPTED MANUSCRIPT
As a follow-up of our previous studies, and aiming at optimizing/improving the biological activities
and further exploring the chemical space of the selected targets, here we report the design, synthesis
and biological evaluation of a new series of 2-arylbenzofuran derivatives, based on the structure of
lead compound 1 (Figure 2).
At first, the N-methyl-N-benzylamine alkoxy side chain was moved from ring A to ring B and the
polymethylene alkoxy spacer was varied to determine its optimal length (2-5). This linker chain was
also removed, leading to a direct linking of mono or di-alkylated amines to ring B (6-18). Finally,
taking into account that click chemistry has found increasing applications in drug discovery, this
approach was exploited to achieve compound 19, by introducing a substituted triazole.
Figure 2. Design strategy for compounds 2-19.
7

ACCEPTED MANUSCRIPT
In summary, a small library of eighteen new derivatives, whose structures are collected in Tables 1
and 2, was synthesized. All compounds were assayed for their biological activities towards selected
targets involved in AD, namely human AChE and BuChE, and CB receptors. After a preliminary
screening, the two most promising compounds (8 and 10) were assessed for their neuroprotective
activity and for their ability to modulate the microglial phenotype switch from the pro-inflammatory
M1 to the anti-inflammatory M2 phenotype.
2
. Results
2.1. Chemistry
According to Scheme 1, 2-(4-methoxyphenyl)benzofuran underwent Friedel Craft acylation, using
acyl chlorides 28-31 in order to obtain compounds 32-35, which were treated with NaI affording the
corresponding iodo analogues (36-39), more reactive species for the following nucleophilic
substitution, that was accomplished by refluxing compounds 36-39 with N-methyl-N-benzylamine,
and TEA in toluene to give the desired compounds 2-5.
The key acyl intermediates 28-31 were prepared via alkylation of the hydroxyl group of the 4-
hydroxyethyl benzoate in basic conditions with 1-bromo-ω-chloroalkanes in order to obtain 20-23,
which via saponification with potassium hydroxide led to compounds 24-27. These carboxylic acids
were then converted into the corresponding acyl chlorides (28-31) by treatment with thionyl
chloride.
a
Scheme 1. Synthesis of compounds 2-5.
8

ACCEPTED MANUSCRIPT
a
Reagents and conditions: i) Br(CH ) Cl, K CO , acetone, reflux; ii) KOH, EtOH, reflux; iii)
2
n
2
3
SOCl , reflux; iv) SnCl , DCM, 0° C, r.t.; v) NaI, MeCOEt, reflux; vi) N-methyl-N-benzylamine,
2
4
TEA, toluene, reflux.
When the Friedel Craft acylation was carried out with 4-nitrobenzoylchloride, compound 40 was
obtained (Scheme 2). The nitro group was reduced to the corresponding amine 41 by refluxing with
SnCl ; this intermediate was then reacted with the selected benzyl bromide in the presence of KI
2
and K CO and in a microwave oven, to obtain the mono- and di-alkylated compounds (6-9, 11-16).
2
3
Compound 8 was methylated in a biphasic mixture of DCM/NaOH 50% using CH I in presence of
3
tetrabutylammonium hydrogensulfate obtaining 10; with the same procedure, using the appropriate
bromoalkane, alkylated compounds 17-18 were obtained from 41. Compound 19 was prepared via a
9

ACCEPTED MANUSCRIPT
click reaction, CuI-catalyzed Huisgen 1,3-dipolar cycloaddition, combining 41 with the alkyne 42
in the presence of tert-butyl nitrite and azidotrimethylsilane (TMSN ) in CH CN.
3
3
a
Scheme 2. Synthesis of compounds 6-19.
O
NO2
O
O
Cl
i
O
+
O N
O
O
2
40
F
ii
R2
O
N
O
N
O
NH2
O
R3
iv
iii
O
O
O
O
O
from 8
10
6-9, 11-16
41
vi
v
N
N
N
H
N
F
O
O
O
F
O
(
)n
42
O
O
O
O
19
17-18 n =2,4
a
Reagents and conditions: i) SnCl , DCM, 0 °C; ii) SnCl , EtOH, reflux; iii) selected benzyl
4
2
bromide, KI, K CO , CH CN, 110 °C, 150W, MW; iv) CH I, tetrabutylammonium hydrogensulfate,
2
3
3
3
DCM/NaOH 50%, r.t.; v) tert-butyl nitrite, TMSN , CH CN, r.t.; vi) 1-bromopentane or 1-
3
3
bromoheptane, tetrabutylammonium hydrogensulfate, DCM/NaOH 50%, r.t.
2
2
.2. Biological evaluation
.2.1. Cholinesterase inhibition
1
0

ACCEPTED MANUSCRIPT
The evaluation of the inhibitory activity toward human recombinant AChE (hAChE) and BuChE
from human serum (hBuChE) was performed according to Ellman's method [20], and the inhibitory
potencies, expressed as IC values, are listed in Tables 1 and 2, together with that of
50
anticholinesterase drug galanthamine.
Table 1. Inhibitory activities exerted by compounds 2-5 and the reference compound galanthamine
towards human AChE and BuChE expressed as IC values.
5
0
O
N
n
O
O
OCH3
b
%
inhib. hAChE
% inhib. hBuChE
IC50 hBuChE (µM)
Compd
n
a
a
[I] = 25 µM
[I] = 25 µM
c
c
1
IC50 = 40.7 ± 3.5
38.1 ± 2.2
2
1
2
3
4
-
n.a.
n.a.
65.0 ± 1.0
35.0 ± 2.7
35.2 ± 0.7
> 90
14.0 ± 0.9
135 ± 6
3
4
n.a.
219 ± 94
1.09 ± 0.07
5
11.6 ± 3.6
d
Galantamine
>90%
54.2 ± 1.8
20.7 ± 1.5
a
%
inhibition of human recombinant AChE or BuChE from human serum at 25 µM. % inhibition
are expressed as mean ± standard error of the mean (SEM) of at least two experiments, each
b
performed in duplicate. IC inhibitory values for inhibition of hBuChE are expressed as mean ±
5
0
1
1

ACCEPTED MANUSCRIPT
c
standard error of the mean (SEM) of at least two experiments, each performed in duplicate. Data
d
from ref [15], involving the same experimental conditions. IC (hAChE) = (2.77 ± 0.13) µM. n.a.:
5
0
=
% inhibition not significant (<10%) at the tested concentration.
With respect to the lead compound 1, all the new derivatives lost their hAChE inhibitory activity,
while some of them showed an increased inhibition of hBuChE, thus displaying a different
selectivity profile. In particular, in the first series (2-5, Table 1), the shift of the N-methyl-N-
benzylamine alkoxy side chain from ring A to ring B, while making the compounds inactive on
hAChE, led to an increased potency toward hBuChE, being compound 5 a good and selective
BuChE inhibitor (35 folds more active than 1 and 19 folds more active than galanthamine).
Moreover, a correlation between potency and side chain length was observed, being the two- and
five-methylene linkers (compounds 2 and 5) considerably more favorable than three or four
(compounds 3 and 4).
Table 2. Inhibitory activities exerted by of compounds 6-19 and the reference compound
galanthamine on human AChE and BuChE expressed as IC values.
5
0
Compd
R
R1
% inhib
hAChE
% inhib
hBuChE
IC BuChE
50
b
(µM)
a
a
[
I] = 25 µM
[I] = 25 µM
1
2

ACCEPTED MANUSCRIPT
6
7
8
H
H
n.a.
n.a.
n.a.
n.d.
n.d.
30.9 ± 1.8
n.a.
60.8 ± 1.1
75.0 ± 1.4
13.4 ± 0.5
F
9
1
9.4 ± 2.38
n.a.
6.03 ± 0.74
F
F
F
1
0
1
CH
3
n.a.
n.a.
n.d.
n.d.
1
1
H
n.a.
2
n.a.
n.a.
67.0 ± 0.1
11.7 ± 0.4
1
1
3
4
H
H
15.3 ± 1.2
n.a.
n.d.
n.d.
1
6.5 ± 3.2
n.a.
CN
1
5
H
n.a.
n.d.
1
1
6
7
H
H
n.a.
n.a.
22.4 ± 1.5
72.2 ± 0.2
n.d.
3.90 ± 0.90
1
1
8
9
H
n.a.
n.a.
30.2 ± 2.9
n.a.
n.d.
n.d.
1
3

ACCEPTED MANUSCRIPT
Galantamine
-
>90%
54.2 ± 1.8
20.7 ± 1.5
a
%
inhibition of human recombinant AChE or BuChE from human serum at 25 µM. % inhibition
are expressed as mean ± standard error of the mean (SEM) of at least two experiments, each
b
performed in duplicate. IC inhibitory values for inhibition of hBuChE. IC values are expressed
5
0
50
as mean ± standard error of the mean (SEM) of at least two experiments, each performed in
duplicate. n.a. = % inhibition not significant (<10%) at the tested concentration. n.d. = not
determined.
In the second series (6-19, Table 2), only compounds 8, 9, 12 and 17 showed inhibitory activity on
hBuChE in the micromolar range, being compound 17 the most potent within this series.
Noteworthy, these four compounds exhibited higher potency with respect to the lead compound 1,
which showed balanced, albeit low, potency on both cholinesterases. Due to their weak potency,
clear structure-activity relationships (SARs) cannot be drawn for this subset of derivatives.
2
.2.2. CB and CB receptors affinity
1 2
The binding affinities of compounds 2-19 at human recombinant CB and CB receptors were
1
2
3
determined by a competition assay using a high affinity ligand, i.e. [ H]-CP-55,940. Results are
reported in Table 3, together with the Ki (inhibition constant) value for compound 1 and for
SR144528 as reference CB receptor ligand.
2
In the first series (compounds 2-5, Table 3), moving the N-methyl-N-benzylamine alkoxy side chain
from ring A to ring B caused a loss of selectivity for CB with respect to the appreciable result
1
observed for lead 1. In particular, the affinity for the CB receptor decreased with the increase of the
1
linker length (compounds 2-5); furthermore, while 1 was CB -selective, the new hybrids bearing the
1
shorter linker chain (2- or 3-methylene units linker, derivatives 2 and 3, respectively) showed a
slightly higher affinity for CB receptors, while compounds 4 and 5, bearing a longer spacer chain,
2
1
4

ACCEPTED MANUSCRIPT
were CB -selective. Thus, a proper chain length seems to be a relevant feature for the modulation of
1
both activity and selectivity. In this respect, compound 3 showed the highest potency and selectivity
(4-fold) towards CB2.
Removal of the spacer chain, with the direct linking of alkylated amines to ring B, led to improved
affinity for the CB receptor, which seemed to be favorably modulated by the aryl substituent.
2
Indeed, the introduction of a fluorine atom in para position induced a 10-fold increase in the
potency (8 vs 6, Ki = 20 nM and 170 nM, respectively). A similar behavior was observed with the
introduction of a CN group (14), while a methoxy function (15) led to a 50-fold reduction of
activity. The bioisosteric substitution of the benzene with a pyridine (13) reduced potency by an
order of magnitude, while the cyclopropyl analogue 16 was equipotent to 6. With the substitution of
the benzene ring with a heptamethylene alkyl chain, as in compound 18, a 2.5-fold decrease in
potency was observed. Conversely, a shorter pentamethylene chain (17) led to a 2.5-fold increase in
activity. Di-alkylated amines bound to B ring showed a loss of potency (12) or a decrease in activity
with respect to mono-alkylated derivatives (9 vs 8, and, to a lesser extent, 7 vs 6). Noteworthy, the
introduction of a small methyl group on compound 8, allowed achieving the most potent compound
of the series (10, Ki = 2.7 nM, 600 folds more active than SR144528).
In the second series, appreciable CB affinity in the nanomolar range was observed for derivatives
1
8
, 10, 14, 16-18. The introduction of a F (8) or a CN (14) substituent on the aromatic ring seemed to
be favorable for receptor binding. The introduction of a cyclopropyl (16) and, mainly, a short five-
methylene unit chain (17, Ki = 80 nM) led to a significant increase in activity. Once again, the most
potent compound in this series was the methyl di-alkylated 10 (Ki = 20 nM), which proved to be a
potent ligand for both receptors, with a selectivity index (SI) towards CB of 7.4. In this respect, the
2
removal of a methyl group yielding the mono-alkylated 8, led to increased selectivity up to 39 folds
(
CB over CB ). Nevertheless, the most selective compound of the series was 14, again a mono
2 1
alkylated derivative, bearing a CN group on the benzene ring. Compound 14 showed a SI of two
1
5

ACCEPTED MANUSCRIPT
orders of magnitude (SI = 130) towards CB . The length of the side chain also affected selectivity:
2
the presence of a 5-unit chain led to a comparable potency at both receptors (17, Ki = 80 and 70
nM), while a 7-methylene chain allowed to better discriminate between the two receptor subtypes,
i.e. derivative 18 showed a 4-fold higher affinity for CB2.
Finally, compound 19, bearing a substituted triazole, was inactive at both CB and CB receptors.
1
2
3
Table 3. Effect of compounds 2-19 and of reference compound SR144528 on [ H]-CP55940
binding to the human recombinant CB and CB receptors.
1
2
a
IC on CB
Ki on CB1
µM
IC on CB
Ki on CB2
µM
SI
5
0
1
50
2
Compd
µM
µM
m ± SD
m ± SD
m ± SD
m ± SD
b
b
b
b
1
2
3
4
5
6
7
8
9
0.79
0.32
>10
>10
1.52 ± 0.28
2.06 ± 0.20
3.52 ± 0.52
7.20 ± 1.47
>10
0.43 ± 0.08
0.50 ± 0.05
0.86 ± 0.13
1.76 ± 0.36
>10
2.63 ± 0.33
0.44 ± 0.09
>10
0.69 ± 0.09
0.12 ± 0.02
>10
0.6
4.2
>10
>10
0.66 ± 0.04
3.64 ± 0.24
0.07 ± 0.001
1.65 ± 0.15
0.17 ± 0.01
0.92 ± 0.06
0.02 ± 0.00
0.43 ± 0.04
> 59
> 11
39
>10
>10
3.21 ± 1.37
>10
0.78 ± 0.33
>10
> 23
7.4
1
0
0.08 ± 0.03
0.02 ± 0.01
0.01 ± 0.001 0.0027 ± 0.0000
1
6

ACCEPTED MANUSCRIPT
11
12
13
14
15
16
17
18
19
>10
>10
>10
1.15 ± 0.23
0.30 ± 0.06
>10
> 33
>10
>10
>10
>10
5.09 ± 1.45
0.09 ± 0.00
3.86 ± 0.46
0.41 ± 0.06
0.26 ± 0.001
1.54 ± 0.29
>10
1.34 ± 0.38
0.02 ± 0.00
1.02 ± 0.12
0.11 ± 0.01
0.07 ± 0.00
0.41 ± 0.08
>10
> 7.5
130
> 10
4.3
9.12 ± 0.17
>10
2.60 ± 0.05
>10
1.67 ± 0.59
0.30 ± 0.07
5.84 ± 2.32
>10
0.47 ± 0.17
0.08 ± 0.02
1.66 ± 0.66
>10
1.1
4
SR1445 541.9 ± 103.03
116.8 ± 22.20
7.16 ± 1.83
1.82 ± 0.46
64.2
2
8
a
b
SI: selectivity index for CB , calculated as Ki(CB )/Ki(CB ); Data from ref [19]
2
1
2
2
.2.3. CB receptor functional activity
2
The most potent CB ligand 10 was further evaluated to assess its capability of activating CB
2
2
receptor in a functional assay. A cAMP Hunter™ assay using an enzyme fragment
complementation chemiluminescent detection kit was performed to assess whether compound 10
could modulate intracellular cAMP levels in NKH-477-stimulated CHO cells overexpressing the
human recombinant CB receptor. In the selected assay conditions, 10 induced an increase of cAMP
2
levels following stimulation with NKH-477 (a water-soluble analogue of forskolin), suggesting an
inverse agonist behavior (Figure 3).
1
7

ACCEPTED MANUSCRIPT
300
200
100
0
1
0
NKH
-10
-9
-8
-7
-6
-5
-4
log[M]
Figure 3. Concentration-response curve showing the effect of increasing concentrations of
compound 10 on NKH-477-induced cAMP levels in stable CHO cells expressing the human CB2
receptor. Data, reported as mean ± SEM of three independent experiments, were normalized to the
maximal and minimal response observed considering the NKH-477 stimulus alone as 100% of the
response.
To confirm this hypothesis, 10 was tested in the presence of a previously determined EC CB -
8
0
2
agonist challenge (JWH133, 4µM). Accordingly, we found that 10 dose-dependently displaced the
challenge of JHW133 and increased cAMP levels following the NKH-477 stimulus (Figure 4).
400
300
200
100
0
1
0
NKH
EC80
-10
-9
-8
-7
-6
-5
-4
log[M]
1
8

ACCEPTED MANUSCRIPT
Figure 4. Concentration-response curve showing the effect of increasing concentrations of
compound 10 on NKH-477-induced cAMP levels and in presence of an EC80 CB -ligand challenge
2
(4µM of JWH-133), in stable CHO cells expressing the human CB2.
2.2.4. Neuroprotective activity
The cytotoxicity of compounds 8 and 10 was evaluated in human SH-SY5Y neuroblastoma cells
using the MTT assay. After 24 h of treatment with different concentrations of compounds 8 or 10
(1.25 – 40 µM) residual cell viability was assessed. Data showed that concentrations up to 5 µM of
both derivatives did not affect neuronal cell viability (data not shown). Therefore, we selected a
range of not cytotoxic concentrations (0.625 – 5 µM) to investigate the neuroprotective activity of
compounds 8 and 10 against Aβ1-42 oligomer (OAβ1-42)-induced neurotoxicity in SH-SY5Y cells.
The neuroprotective activity was evaluated after 4 h of treatment with OAβ1-42 (10 µM) in the
presence of compounds 8 or 10 using the MTT formazan exocytosis assay. Only the treatment with
compound 8 significantly counteracted, in a dose-dependent manner, the neurotoxicity induced by
OAβ1-42 (Figure 5).
50
Compound 8
Compound 10
40
3
0
**
20
***
10
0
2
M
M
M
M
4
-
µ
µ
µ
µ
β 1
5
5
5
5
.
A
.62
2
.
2
O
1
0
OAβ1-42
1
9

ACCEPTED MANUSCRIPT
Figure 5. Compound 8, unlike compound 10, counteracts the neurotoxicity induced by OAβ1-42 in
SH-SY5Y cells. The neurotoxicity was evaluated using the MTT formazan exocytosis assay as
described in the materials and methods section. Data are expressed as percentages of neurotoxicity
and reported as mean ± SEM of at least three independent experiments (**p < 0.01 and ***p <
0.001 versus cells treated with OAβ1-42 at two-way ANOVA with Bonferroni post hoc test).
2.2.5. Activation of Microglia
The possible immunomodulatory effects of compounds 8, 10 and of the lead 1 were evaluated by
analyzing the expression of inducible nitric oxide synthase (iNOS), which is widely considered a
marker of M1 neurotoxic microglia, and of the triggering receptor expressed on myeloid cells 2
(
TREM2), a marker of M2 neuroprotective microglia, in immortalized murine N9 cells, after
exposure to lipopolysaccharide (LPS; 100 ng/mL), a widely–used pro-inflammatory stimulus
Figure 6 and 7) [21-22].
(
After treatment with compounds 8 and 10, reduction of the M1 marker iNOS was observed with a
maximum effect at 5 µM (inhibition: 83% for compound 8; 60% for compound 10), with no
concomitant decrease in TREM2 expression (Figure 6). This suggests an immunomodulatory effect
for both compounds.
2
0

ACCEPTED MANUSCRIPT
Figure 6. Western blot analysis of iNOS and TREM2 expression as microglial polarization markers
(A, B) and their relative densitometries (C, D, E, F) to evaluate the immunomodulatory effects of
compounds 8 and 10 in N9 cell line. GAPDH was used as loading control. Densitometric results are
expressed as percentage of LPS only and are the mean ± SE of three different experiments.
*p<0.05, **p<0.01 compared to LPS condition; Bonferroni’s test after ANOVA.
Regarding the lead compound 1, a toxic effect was observed at the lowest assayed concentration (5
µM). Indeed, as reported in Figure 7, glyceraldehyde 6-phosphate dehydrogenase (GAPDH, loading
control) expression appeared considerably reduced, suggesting the occurrence of cell death.
2
1

ACCEPTED MANUSCRIPT
Figure 7. Western blot analysis of iNOS and TREM2 expression as microglial polarization markers
(A) and their relative densitometries (B, C) to evaluate the immunomodulatory effects of compound
1
in N9 cell line.
2.2.6. Molecular docking studies
Finally, being compound 8 also endowed with a significant hBuChE inhibitory activity, a molecular
docking simulation was performed using Autodock Vina software [23], in order to explore the
binding mode of this derivative toward hBuChE and provide more insights into the possible
interactions with the enzyme. We applied docking studies based on the available crystal structure of
the hBuChE complexed with tacrine (PDB ID: 4BDS) with a grid box of sufficient size to cover
whole protein for docking.
The docking simulation of 8 discloses that this compound can be efficiently accommodated inside
the active site gorge with a molecular docking score of -10.7 kcal/mol (Figure 8a). Three-
2
2

ACCEPTED MANUSCRIPT
dimensional (3D) and two-dimensional (2D) representations of the most energetically profitable
pose of compound 8 docked into the active site of BuChE were given in Figures 8b and 8c.
The complex of protein with the best-docked pose of inhibitor showed that the benzofuran moiety is
located at the entrance of the active-site gorge with both side chains oriented deep inside the gorge.
The phenyl ring aromaticity clearly plays a dominant role in stabilizing inhibitor binding. The p-F-
benzyl moiety is oriented toward the bottom of the active site and it binds in the CAS region of the
enzyme establishing strong ᴫ-ᴫ stacking interactions with Trp82 and His438 of the catalytic-triad.
The methoxy-phenyl ring forms ᴫ-ᴫ T-shaped interactions with Phe329 in the acyl-binding pocket
of the enzyme. The amino-phenyl ring forms ᴫ-ᴫ stacking interaction with Tyr332 of the PAS.
Apart from these ᴫ-ᴫ interactions, a hydrogen bond was established between the oxygen atom of the
methoxy group and the hydroxyl group of Ser198 of the catalytic triad.
Notably, the F atom in the phenyl ring interacted with amino acids to form halogen interactions. F
atom held Gly115 and Glu197 at one binding site resulting in the presence of F•••C=O and F•••O
interactions (Figures 8b and 8c). These residues were near and in the sub-site known as “oxyanion
hole” formed by Gly115, Gly116 and Gly117 and having an essential role in catalysis. These non‒
covalent interactions were fairly strong, with the bond distances being around 3.60 Å.
In conclusion, docking studies revealed the capability of compound 8 to bind to CAS and PAS of
hBuChE and induce its inhibitory effect.
2
3

ACCEPTED MANUSCRIPT
Figure 8. Docking poses of compound 8 inside gorge cavity of hBuChE. a) Compound 8 is
coloured in pink. Different subsites of the active site were differently coloured: catalytically anionic
site (CAS) in green, oxyanion hole (OH) in red, choline binding site in violet (CBS), acyl binding
pocket (ABP) in yellow, and peripheral site (PAS) in blue. b) 3D representation of the amino acids
in the binding site interacting with the ligand 8. c) 2D representation of the amino acids in the
binding site interacting with the ligand 8.
3. Discussion
In this paper the potential multitarget profile of two series of benzofuran derivatives, focused on the
structure of the previously reported lead compound 1, was assessed. From the results, some
considerations can be drawn.
2
4

ACCEPTED MANUSCRIPT
In the first series (compounds 2-5), moving the side chain from A to B ring allowed achieving
selectivity towards hBuChE, while affinity for the two cannabinoid receptors was modulated by the
length of the linker. In particular, compound 5, with a 5-methylene chain, showed a balanced effect
on the two targets, with a good inhibitory activity on hBuChE and appreciable affinity and
selectivity for CB receptors. This finding is of particular interest since BuChE has recently gained
1
a renewed interest as a therapeutic target in AD, since changes in its activity along disease
progression were highlighted. Indeed, the brain levels of BuChE are un-altered or progressively
increased during AD progression, while AChE activity declines [24]. On the other hand, CB1
antagonists caused improvement in memory in several preclinical models of cognitive impairment
[
25-26].
In the second series (6-18), the linker chain was removed, leading to a direct binding of mono or di-
alkylated amines to B ring. Again, from a multitarget point of view, derivatives 8, 9 and 17 emerged
as good hBuChE inhibitors and CB receptors ligands. In particular, compound 9 was endowed with
a peculiar selectivity for CB receptors. Notably, derivatives 8 (p-F-benzyl alkylated) and 9 (bis-p-
2
F-benzyl alkylated) contain fluorine atoms. Organofluorine compounds have recently been claimed
as one of the most promising class of drug candidates in medicinal chemistry [27-29]. Indeed,
fluorine and fluoro-functional groups can be of crucial importance in affecting potency and
pharmacokinetic properties of bioactive compounds, as the unique nature of fluorine strongly
impacts on the chemical properties of the products [27]. On the other hand, 17 contains a n-pentyl
alkyl chain, able to mimic the side chain of classical natural and synthetic cannabinoid ligands.
Aiming at evaluating the possible involvement of CB receptor in neuroinflammation, compound 8,
2
showing affinity for both receptors, although with a preference for CB receptor (SI = 39), and 10,
2
endowed with the highest CB binding affinity (Ki = 2.7 nM) were selected for further studies.
2
2
5

ACCEPTED MANUSCRIPT
First of all, the most potent CB ligand (10) was evaluated to assess its capability of functionally
2
activating the CB receptor. Under the functional assay conditions, this compound induced an
2
increase of cAMP levels, suggesting an inverse agonist behavior.
The compounds were then evaluated to establish their neuroprotective activity against OAβ1-42
-
induced neurotoxicity in SH-SY5Y cells. Inflammation, indeed, represents a consistent feature in
AD brain, and mounting evidence suggests that a neuroinflammatory response may be implicated in
disease development. Insoluble deposits of Aβ are strong candidates for initiating the inflammatory
response, whereas activated microglia and astrocytes could trigger a self-propagating
neuroinflammatory process resulting in brain damage. Treatment with compound 8 significantly
counteracted, in a dose-dependent manner, the OAβ1-42-induced neurotoxicity while 10 was
inactive. Recent studies have reported that both CB and CB receptors are expressed in SH-SY5Y
1
2
cells, but have also highlighted the lack of a functional role for CB receptor agonists in response to
Aβ exposure in these cells [30]. Therefore, the neuroprotective activity exerted by 8 may be due to
its putative mode of action, i.e., inverse agonist at CB receptor, a property for which this
2
compound proved to be 39 folds more selective than the inactive derivative 10. However, it cannot
be excluded at this stage that the neuroprotective activity of 8 could also result from a direct, non-
CB receptor-mediated, action on Aβ peptide. In this regard, we have previously demonstrated that
the benzoyl group of compound 1 is crucial [15] for the interaction with the hydrophobic residues of
Aβ peptide such as Ile31, Ile32 and Met35, that are critical for both neurotoxicity and aggregation
,
processes [31-32]. Indeed, the neuroprotective effects of benzofuran derivative 8, but not 10,
carrying an additional secondary amino group, may be related to its ability to form further donor
hydrogen bonds with the same amino acid residues of Aβ peptide.
Microglial cells represent the main immune cells of the brain and are widely considered a promising
target to counteract neuroinflammatory diseases, including AD. In physiological conditions, they
constantly survey the parenchyma, while in neuropathological conditions they become activated,
2
6

ACCEPTED MANUSCRIPT
and play different roles depending on the duration and the gravity of the insult. Indeed, at an early
stage of the disease, in order to protect neurons and fight the pathology, microglia acquire a M2
phenotype; later, when the disease progresses and expands, a M1 phenotype develops, and these
cells become neurotoxic, producing proinflammatory cytokines, and leading to neuroinflammation.
For these reasons, immunomodulation, i.e. the shift of microglial cells from the neurotoxic M1 to
neuroprotective M2 phenotype, is now widely considered a promising therapeutic approach to
tackle neurodegenerative diseases, including AD [33-34].
Accumulating evidence indicates that cannabinoids can modulate the function of activated
microglia and the release of cytotoxic factors [35]. It has been also shown that both CB receptors
are expressed in microglia and astrocytes. Unlike the CB receptor, which is constitutively
1
expressed in microglia, the expression of CB receptor is inducible and is modulated in response to
2
the activation stage of microglia. In particular, CB₂ receptor levels undergo time-dependent changes
during microglia activation, with the initial attempt to counteract inflammation [36-37].
Although much of the literature dealing with CB immunomodulatory effects focuses on agonists
2
[
38], the in vitro studies involving microglia continue to produce inconsistent results on the role of
CB and CB receptors in the action of cannabinoids. CB agonists have raised great interest as
1
2
2
candidates for neuroprotection in neurodegenerative disorders. By contrast, CB antagonists/inverse
2
agonists have been shown to be promising agents for immunomodulation therapies. Experimental
studies have documented beneficial effects of the administration of CB antagonists/inverse
2
agonists in animal models of various inflammatory disorders, arthritic bone damage and
experimental autoimmune encephalomyelitis [39]. Indeed, recent studies showed that CB and CB2
1
receptor antagonists can directly suppress iNOS induction and ROS generation in activated
microglia, leading to a feasible neuroprotective outcome [40]. Moreover, the beneficial effect of
SMM-189, a potent and selective CB inverse agonist, in a mild traumatic brain injury model,
2
suggested a possible anti-inflammatory activity, due to its ability of reducing the M1 and increasing
2
7

ACCEPTED MANUSCRIPT
the M2 phenotype of human and murine microglia in vitro [41]. Again, studies on the triaryl bis-
sulphone, Sch.414319, also led to speculate that inverse agonists to the CB receptor may serve as
2
immune modulators [42]. Here we observed that compounds 8 and 10 were able to strongly reduce
iNOS expression (an M1 marker) induced by LPS treatment, without changing TREM2 expression
(an M2 marker). This indicates an immunomodulatory effect of both compounds even at low
concentrations (5µM). Notably, in the same experimental conditions, the lead compound 1, a
selective CB ligand, was cytotoxic and induced cell death even at the lowest tested concentration,
1
endorsing a potential key role of CB receptors in immunomodulation.
2
In conclusion, compound 8 can face AD from several directions: it restores the cholinergic system
by inhibiting BuChE, exerts a neuroprotective activity against Aβ1-42 oligomers and is a potent CB
receptor ligand, with 39-fold selectivity for CB . Compound 8 is able to exert an interesting
2
multifunctional action by combining anti-inflammatory and neuroprotective effects. In addition,
since pro-inflammatory M1 microglia predominate in the advanced stage of AD and aggravate
disease progression, compound 8 by interfering with the activation of M1, could be effective in
halting lesion development.
Derivative 10 proved to be a potent CB inverse agonist with promising immunomodulatory
2
properties. Being the development of CB receptor antagonists/inverse agonists still very limited, a
2
clear estimation of the therapeutic impact of such a class of ligands must await initial clinical testing
and data in humans. Indeed, the mechanisms whereby CB modulation affects inflammation and
2
immune cell function are complex and detailed pathways remain to be elucidated. CB receptor
2
antagonists are, therefore, useful tools for studying signaling pathways associated with CB binding.
2
In this respect, 10 could be considered as a new probe for a better understanding of the role of CB2
receptors and to develop potential immunomodulating drugs acting via the endocannabinoid system.
4.
EXPERIMENTAL
2
8