Galantamine-curcumin hybrids as dual-site binding acetylcholinesterase inhibitors

Article abstract of DOI:10.3390/molecules25153341

Galantamine (GAL) and curcumin (CU) are alkaloids used to improve symptomatically neurodegenerative conditions like Alzheimer's disease (AD). GAL acts mainly as an inhibitor of the enzyme acetylcholinesterase (AChE). CU binds to amyloid-beta (Aβ) oligomers and inhibits the formation of Aβ plaques. Here, we combine GAL core with CU fragments and design a combinatorial library of GAL-CU hybrids as dual-site binding AChE inhibitors. The designed hybrids are screened for optimal ADME properties and BBB permeability and docked on AChE. The 14 best performing compounds are synthesized and tested in vitro for neurotoxicity and anti-AChE activity. Five of them are less toxic than GAL and CU and show activities between 41 and 186 times higher than GAL.

Full text of DOI:10.3390/molecules25153341

Article
Galantamine-Curcumin Hybrids as Dual-Site
Binding Acetylcholinesterase Inhibitors
1,2,
2
1
1
Georgi Stavrakov *, Irena Philipova , Atanas Lukarski , Mariyana Atanasova ,
1
1
1,3
1
Dimitrina Zheleva , Zvetanka D. Zhivkova , Stefan Ivanov , Teodora Atanasova ,
1
1,
Spiro Konstantinov and Irini Doytchinova *
1
Faculty of Pharmacy, Medical University of Sofia, Sofia 1000, Bulgaria;
alukarski@pharmfac.mu-sofia.bg (A.L.); matanasova@pharmfac.mu-sofia.bg (M.A.);
dzheleva@pharmfac.mu-sofia.bg (D.Z.); zzhivkova@pharmfac.mu-sofia.bg (Z.D.Z.);
ivanovs@umd.edu (S.I.); tatanasova@pharmfac.mu-sofia.bg (T.A.);
skonstantinov@pharmfac.mu-sofia.bg (S.K.)
2
Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, Sofia 1113,
Bulgaria; irena@orgchm.bas.bg
Institute for Bioscience and Biotechnology Research, University of Maryland, Rockville, MD 20850, USA
3
* Correspondence: stavrakov@pharmfac.mu-sofia.bg (G.S.); idoytchinova@pharmfac.mu-sofia.bg (I.D.)
Academic Editor: Jaume Bastida
Received: 21 June 2020; Accepted: 16 July 2020; Published: 23 July 2020
Abstract: Galantamine (GAL) and curcumin (CU) are alkaloids used to improve symptomatically
neurodegenerative conditions like Alzheimer’s disease (AD). GAL acts mainly as an inhibitor of the
enzyme acetylcholinesterase (AChE). CU binds to amyloid-beta (Aβ) oligomers and inhibits the
formation of Aβ plaques. Here, we combine GAL core with CU fragments and design a
combinatorial library of GAL-CU hybrids as dual-site binding AChE inhibitors. The designed
hybrids are screened for optimal ADME properties and BBB permeability and docked on AChE.
The 14 best performing compounds are synthesized and tested in vitro for neurotoxicity and anti-
AChE activity. Five of them are less toxic than GAL and CU and show activities between 41 and 186
times higher than GAL.
Keywords: virtual screening; molecular docking; neurotoxicity; ADME; BBB permeability
1. Introduction
Most of the neurodegenerative diseases are associated with altered levels of the transmitter
acetylcholine (ACh) in selected areas of the brain [1]. The enzyme acetylcholinesterase (AChE)
catalyzes the hydrolysis of ACh in the cholinergic synapses. By inhibiting the AChE, the levels of
ACh increase and the neurodegeneration improves symptomatically. Moreover, AChE interacts with
the amyloid-β (Aβ) peptide throughout the peripheral anionic site (PAS) and forms highly neurotoxic
AChE-Aβ complexes that promote the assembly of Aβ into amyloid fibrils [2]. The amyloid fibrils
grow progressively and irreversibly to amyloid plaques—one of the main characteristics of
Alzheimer’s disease (AD) [3]. Hypothetically, compounds binding simultaneously to the catalytic
active site (CAS) and the PAS of AChE, are able both to prevent the ACh hydrolysis and the onset of
amyloidogenesis. This hypothesis prompted the search for dual-site binding AChE inhibitors as
multitarget agents for AD treatment [4–18]. Unfortunately, since 2003, no new drugs have been
approved for the treatment of AD [19].
Galantamine (GAL) (Figure 1) is an alkaloid isolated from the bulbs and flowers of Leucojum
aestivum L. and Galanthus sp. It is an inhibitor of AChE [20] and is among the few drugs approved for
Molecules 2020, 25, 3341; doi:10.3390/molecules25153341
www.mdpi.com/journal/molecules

Molecules 2020, 25, 3341
2 of 18
the treatment of AD. In addition, GAL has been identified as an allosteric nicotinic acetylcholine
2+
receptor (nAChR) modulator [21]. The stimulation of nAChRs increases the intracellular Ca levels
and facilitates the release of noradrenaline. Both effects improve the cognitive function of the brain
[21]. GAL treatment of rat microglia significantly increases the phagocytosis of amyloid β (Aβ)
peptide and facilitates the clearance of Aβ in the brain of rodents with AD [22]. This multitarget action
of GAL makes it a valuable drug for AD treatment and stimulates the search for new GAL derivatives
with higher affinity to AChE [23–31].
Curcumin (CU) (Figure 1) is a natural polyphenolic compound isolated from the rhizomes of
Curcuma longa L. It has a symmetrical molecular structure and its IUPAC name is (1E,6E)-1,7-Bis(4-
hydroxy-3-methoxyphenyl)hepta-1,6-diene-3,5-dione. There are two more curcuminoids—
demethoxycurcumin and bisdemethoxycurcumin. All three compounds are biologically active. The
main features of the curcuminoid family are the following structural motifs: a benzene ring
conjugated with α-β unsaturated carbonyl moiety, hydroxyl groups on the fourth position in the
benzene ring, and methoxy group present in two of the three members at third position. Usually, CU
exists as two tautomeric forms as one of the carbonyl groups transforms into an enolic group. CU
binds to Aβ oligomers and fibrils and inhibits the β-sheet formation [32]. It was found that the
aromatic methoxy and/or hydroxy groups interact with the V12 and 16KLVFFA21 residues of the Aβ
[33,34], while the aryl rings make π-π stacking with the aromatic residues [32]. The enone group and
the unsaturated carbon spacer are also essential for the anti-Aβ aggregation activity [33–35].
In the present study, we designed a set of GAL-CU hybrids as dual-site binding AChE inhibitors
with potentially anti-amyloid aggregation activity. The hybrids underwent two-step virtual
screening. First, they were screened for optimal ADME properties and blood-brain barrier (BBB)
permeability and then, a docking-based virtual screening on human AChE was performed. The
highest scored hybrids were synthesized and tested for cytotoxicity and AChE activity. Five of them
were more active and less toxic than GAL and CU themselves.
2. Results
2.1. Design of GAL-CU Hybrids
The GAL-CU hybrid structures were composed of a core, a linker, and an aromatic group. The
GAL molecule was used as a core. The linkers started from GAL amine’s methyl group and ended
with aromatic moieties mimicking the CU structure (Figure 1). Eight aromatic substituents (assigned
by the letters a–h) were designed to mimic the CU phenyl rings (Figure 2). The first substituent a
consisted of a single phenyl ring. The substituents b–d contained a phenyl ring with methoxy groups
at m- or p-position, or both. The substituent e was similar to d, but the two methoxy groups were
replaced by a more constrained 1,3-dioxole ring. The remaining substituents f–h contained phenyl
rings with bioisosteric methyl groups at m- or p-position, or both.

Molecules 2020, 25, 3341
3 of 18
Figure 1. Common structure of the designed GAL-CU hybrid molecules.
Figure 2. Aromatic substituents (Ar) used in the designed GAL-CU hybrid molecules.
The linkers were designed to resemble the CU α-β unsaturated fragment. In our previous studies
on GAL derivatives [36–38], it was found that a linker containing five to seven carbon atoms is optimal
for dual-site binding to AChE. Seven types of linkers were suggested for the present GAL-CU hybrids
and the designed compounds were divided into five groups according to the linker type (Figure 3). In
the first two groups of linkers, the place of a double bond according to a carbonyl group was varied, as
well as the linkers’ lengths. The third linker resembles the tautomeric form of CU, with one double bond
on each side of the carbonyl group. The last two groups of linkers incorporate a benzene ring in the
middle because of the possible π-π interaction with Tyr341 and Tyr337 in hAChE [20]. Both m- and p-
isomers were considered. In total, 72 GAL-CU hybrids were designed in the present study.

Molecules 2020, 25, 3341
4 of 18
Figure 3. Linkers used in the designed GAL-CU hybrid molecules.
2.2. ADME Filters
The 72 newly designed GAL-CU hybrids were tested for BBB permeability, gastrointestinal (GI)
absorption and non-specific interactions with numerous biological targets (PAINS filter), as described
in Section 4. Only 44 compounds passed all criteria: 12 from group 1, 15 from group 2, 8 from group
3, 3 from group 4, and 6 from group 5.
2.3. Virtual Screening by Molecular Docking
The 44 molecules that passed the ADME filters were docked to human rAChE, as described in the
Materials and methods section. Only docking poses with root mean square difference (RMSD) < 1,5 Å
of the referent GAL structure were considered. The GoldScores of the newly designed compounds
ranged from 114.54 to 88.28, and all were higher than the GoldScore of GAL (78.25) (Table 1). Among
the top 33% of the best-scored compounds (GoldScore range 114.54–104.00) were compounds with
aromatic substituents a, b, c, f, g and h and linkers from group 4 (series 6) and group 5 (series 8).
2.4. GAL-CU Hybrids Selected for Synthesis
The GAL-CU hybrids selected for synthesis are given in bold in Table 1. Among the top 33% of
the best-scored compounds were selected from series 6 and 8. Compounds 6a, 6b, 8b, 8c, 8f, 8g, and
8h from the top range were selected for synthesis. In our previous studies, good correlations between
the GoldScores and the experimental IC50 values to AChE of GAL derivatives were observed [36–39].
Here, in order to examine this relationship, several compounds from the middle (103.99–94.00) and
low GoldScore range (93.99–84.00) were selected for synthesis as well. Compounds 4b, 4e, 4f and 4h
were selected from the middle range, while compounds 4a, 4c and 4g were selected from the low
range of GoldScore. In total, 14 GAL-CU hybrids were selected for synthesis.
2.5. Synthesis of Selected GAL-CU Hybrids
2.5.1. Synthesis of Hybrids 4a–c,e–h
The synthetic strategy used in the present study relied on the construction of linkers binding
different aromatic moieties to Norgalantamine (NOR) 10 (Figure 4). Ketones of the type 11 were
foreseen as key intermediates, due to possible Aldol condensation with chosen benzaldehydes 12a–
c,e–h. Further on, we envisaged the introduction of a leaving group on the aliphatic end of the linker,
because of the final nucleophilic substitution with NOR 10 towards the desired target compounds
4a–c,e–h (Figure 4).

Molecules 2020, 25, 3341
5 of 18
Figure 4. Retrosynthetic analysis of compounds 4a–c,e–h.
Notably, 1,5-Pentandiol 13 was considered as an available and inexpensive starting material
(Scheme 1). It allows independent functionalization of its two hydroxyl groups. Monoprotection as
tert-butyldimethylsilyl ether and subsequent mild oxidation proceeded with good yields to give the
unstable aldehyde 14 [40]. The latter was immediately subjected to Grignard attack to give a racemic
secondary alcohol, which was oxidized to ketone 11a [41]. Aldol condensation of the ketone with
benzaldehyde was accomplished using catalytic amounts of lithium hydroxide monohydrate as a
base in absolute ethanol. The product was isolated in low yield. Thus, the hydrocarbon skeleton of
the desired linker was generated. The deprotection of the silyl ether proceeded smoothly in slightly
acidic media, and the free alcohol was transformed into iodide 15, which unfortunately decomposed
during the reaction. Obviously, the iodide was not the proper choice for a leaving group because of
its high reactivity.
Scheme 1. Attempted synthesis of iodo-linkers.
Alternatively, a monobromination of 1,5-pentandiol 13 proceeded with excellent selectivity
because the product is water-insoluble and leaves the aqueous phase where the reaction takes place
(Scheme 2) [42]. Mild oxidation afforded 5-bromopentanal 16 [43], which was attacked by CH3MgCl
and subsequently oxidized to give 6-bromohexan-2-one 11b. A key feature for the success of this
shorter route was the application of Aldol condensation under mild reaction conditions. The desired
bromo-linkers 17a–c,e–h were synthesized after days stirring of the bromo-ketone 11b with chosen
benzaldehyde 12a–c,e–h and catalytic amounts of L-proline, and triethylamine in methanol media
[44]. The products of condensation were isolated in moderate yields. Nucleophilic substitution of the
bromo-linkers with NOR 10 in the presence of K2CO3 as a base afforded the desired target compounds
4a–c,e–h in 48% to 64% yield (Scheme 2).

Molecules 2020, 25, 3341
6 of 18
Scheme 2. Synthesis of compounds 4a–c,e–h via bromo-linkers.
2.5.2. Synthesis of Hybrids 6a and 6b
The synthetic strategy towards compounds 6a and 6b again included two key steps: Aldol
condensation for the linker construction and nucleophilic substitution with NOR 10 at a later stage
(Figure 5). The difference compared to compounds type 4 was the change of places of aldehyde and
ketone, due to the change of positions of carbonyl group and double bond in the conjugated enone
system. Thus, we planned the synthesis of benzaldehydes of type 18, subsequent condensation with
chosen acetophenones 19a,b and a final attachment to the alkaloid core (Figure 5).
OH
OH
O
O
R
R
O
O
+
+
X
O
N
NH
O
O
6a,b
10
18a
, X = OTBS
18b, X = Br
19a,b
Figure 5. Retrosynthetic analysis of compounds 6a,b.
The synthesis of 6a started with the monoprotection of 1,3-phenylenedimethanol as tert-
butyldimethylsilyl ether and its oxidation with manganese(IV) oxide to give benzaldehyde 18a [45].
Aldol condensation of the latter with acetophenone in the presence of 20 mol% LiOH. H2O in ethanol
afforded ketone 21 [46], which was deprotected quantitatively and the resulting alcohol was
converted into iodide 22 applying Appel reaction. The iodo-linker was attached to NOR 10 to give
the desired product 6a in 76% yield, after 6 steps and 20% overall yield (Scheme 3).

Molecules 2020, 25, 3341
7 of 18
Scheme 3. Synthesis of 6a via an iodo-linker.
Aiming for optimization of the synthesis of 6b, we decided to skip the silyl protection by
introducing the leaving group into the very first step (Scheme 4). Monobromination of 1,3-
phenylenedimethanol 20 proceeded selectively with good yield, and the product was oxidized to
give benzaldehyde 18b. Aldol condensation of the latter with acetanisole proceeded with catalytic
ammounts of LiOH.H2O in ethanol to give 23 in moderate yield. Substitution of the bromo-linker
with NOR resulted in the desired product 6b in 59% yield, after 4 steps and 13% overall yield (Scheme
4). Although the synthetic pathway to 6b was 2 steps shorter in comparison with 6a, the overall yield
was lower. When applying Aldol condensations in the presence of benzyl bromides, a concurrent
nucleophilic substitution poisons the base.
Scheme 4. Synthesis of 6b via a bromo-linker.

Molecules 2020, 25, 3341
8 of 18
2.5.3. Synthesis of Hybrids 8b,c,f,g,h
For the synthesis of the target structures 8b,c,f,g,h, we implemented the strategy developed for
compounds of the type 4. We planned the synthesis of acetophenone 24, its condensation with chosen
benzaldehydes 12b,c,f,g,h, and finally the attachment of the constructed linkers to the galantamine
core (Figure 6).
Figure 6. Retrosynthetic analysis of compounds 8b,c,f,g,h.
Benzaldehyde 18a was converted into 3-(hydroxymethyl)acetophenone 25 via reaction with
CH3MgCl, oxidation of the resulting alcohol to 24 and deprotection. Aldol condensation proceeded
in good yields with the chosen benzaldehydes 12b,c,f,g,h [47]. The resulting alcohols 26b,c,f,g,h were
converted into the corresponding iodides 27b,c,f,g,h. Reactions of the latter with NOR 10 gave the
desired target compounds 8b,c,f,g,h (Scheme 5).
Scheme 5. Synthesis of 8b,c,f,g,h.
1
13
The detailed synthetic procedures, analytical data and copies of H and C NMR spectra for the
target compounds are given in a Supplementary File.

Molecules 2020, 25, 3341
9 of 18
2.6. Neurotoxicity of the GAL-CU Hybrids
The neurotoxicity of the GAL-CU hybrids was tested on Neuro-2A cells, as described in
Materials and methods. The IC50 values were in the range from 7.91 to 52.53 µM (Table 1). The IC50s
for GAL and CU were >50 and 26.30 µM, respectively. Only five of the hybrids were less toxic than
CU. One of them, compound 8b, is among the best-scored compounds; two of them, compounds 4e
and 4f, are from the middle GoldScore range; and two of them, compounds 4a and 4b, belong to the
low GoldScore range.
2.7. Anti-AChE Activity
The anti-AChE activity of the GAL-CU hybrids, less toxic than GAL and CU, was measured in
vitro by Ellman’s method, as described in Materials and Methods. As the docking studies were
performed on rhAChE, while for the in vitro measurements was used AChE from electric eel
(eeAChE), it was reasonably to question their relevance. The UniProt alignment of rhAChE (UniProt:
P22303) and eeAChE (UniProt: O42275) have shown that all 17 residues forming the binding gorges
are identical [36]. Additionally, our previous experience has shown that the docking scores predicted
for rhAChE correlate well with the experimental binding affinity measured on eeAChE [36–38].
The IC50 values of the tested compounds are given in Table 1. For comparison, the IC50 values of
GAL and CU as single molecules are given. GAL is a medium AChE inhibitor with IC50 = 3.52 µM,
while CU is less active with IC50 = 68 µM. However, GAL-CU hybrids are more active than both of
them. Compound 4b is the most active with IC50 = 0.02 µM, being 186 times more active than GAL.
The compounds 4a, 4e and 4f have IC50 values in the range of 0.033 to 0.046 µM and are between 110
and 75 times more active. Unexpectedly, the best-scored compound 8b is the least active among the
5 tested compounds, having IC50 of 0.086 µM and only 41 times higher activity than GAL.
The correlation coefficient r between GoldScores and pIC50 (–logIC50) values of the novel hybrids,
GAL and CU was 0.803. Omitting the outlier 8b yielded r = 0.938.
Table 1. GoldScore, neurotoxicity and affinity to AChE of the GAL-CU hybrids. The IC50 values of the
hybrids, less toxic than CU, are given in bold. The binding affinity to AChE was measured only for
the less toxic compounds.
ID
8c
6a
8g
6b
8f
8h
8b
4h
4e
4f
4b
4a
GoldScore
114.54
114.50
111.92
111.15
109.22
108.95
104.01
96.98
IC50 µM Neuro-2A
25.55
IC50 µM eeAChE
Times More Active Than GAL
-
-
-
-
-
-
-
-
-
-
12.14
17.80
7.91
24.32
25.00
28.87
21.93
42.91
34.35
52.53
30.65
23.68
-
-
41
-
98
110
186
75
-
0.086
-
96.07
95.05
94.31
93.74
0.036
0.033
0.020
0.046
-
4g
90.71
4c
GAL
CU
88.28
24.11
-
-
1
74.56 a
88.93
>50 a
3.520
26.30
67.69 b
a [37]; b [48].
2.8. Physicochemical Properties and PK Parameters
The molecular weights (Mw) of the tested compounds were in the range from 459.58 to 523.62
g/mol (Table 2). The pKa values varied in a short range from 6.58 to 7.78 and were close to the pKa of
GAL 7.92. The GAL-CU hybrids were more lipophilic than GAL—their logP values were between
4.48 and 4.99, logD7.4s – between 3.92 and 4.83, while the corresponding values of GAL were 1.75 and

Molecules 2020, 25, 3341
10 of 18
1.12. According to fB, compounds 4a, 4b, 4e and 4f were moderate bases, 70% ionized at physiological
pH like GAL, but 8b was a very weak base—only 13% is ionized. Although the hybrids were more
lipophilic than GAL, they had polar surface areas (PSA) wider than that of GAL, but lower if
normalized to unit mass. There was only one hydrogen-bond donors (HBD) in the molecules (H from
ammonium cation), while the number of hydrogen-bond acceptors (HBA) depended on the number
of O-atoms, but for all active hybrids, was up to 10.
ss
The predicted steady state volume of distribution (VD ) values of the hybrids 4a, 4b, 4e and 4f
ss
were almost three times bigger than the VD of 8b. All compounds had comparable lipophilicity, but
ss
ss
different ionized fractions at pH 7.4. The fraction ionized as a base affects positively VD [49–51]. VD
reflects the drug ability to cross membranes and to bind in tissues [52]. The moderate bases 4a, 4b, 4e
ss
ss
and 4f have VD between 4.71 L/kg and 5.88 L/kg which are close to the mean value of VD 5.90 L/kg
ss
for bases in Obach’s database [53,54]. The very weak base 8b has VD of 1.85 L/kg, which is close to
the mean VD of 1.94 L/kg for neutral molecules [54,55].
ss
The tested compounds bind extensively to plasma proteins (PP). The predicted free fraction of
drug in plasma (fu) were between 0.029 and 0.094, i.e., the extent of PP binding is in the range 91–97%.
Lipophilicity has been identified as a major determinant for the PP binding of basic drugs [56]. GAL
has lower logP and logD7.4, which results in a lower PP binding and fu = 0.83 (Table 2).
The predicted total clearance (CL) values were in the range between 0.305 and 0.466 L/h/kg.
Apparently, the GAL-CU hybrids are low-clearance compounds with CLs below 37% of hepatic blood
ss
flow (QH = 1.26 L/h/kg) [57]. The half-lives (t1/2) were calculated from the corresponding VD and CL
values. For the compounds 4a, 4b, 4e and 4f, they varied between 7.01 h and 10.51 h, while the half-
life of 8b was shorter, close to that of GAL. The moderate half-lives allow convenient multiple-dose
regimens.
Table 2. Physicochemical properties and PK parameters of the active GAL-CU hybrids.
Parameter
Mw
pKa
logP
logD7.4
PSA
HBD
4a
459.58
7.78
4.53 ± 0.8
3.97
4b
489.60
7.77
4.48 ± 0.81
3.92
4e
503.59
7.77
4.64 ± 0.85
4.09
4f
473.60
7.77
4.99 ± 0.80
4.43
8b
523.62
6.58
4.9 ± 0.85
4.83
GAL
287.35
7.92
1.75
1.12
41.93
1
59.00
1
68.23
1
77.46
1
59.00
1
68.23
1
HBA
fBH+
5
0.71
4.71
0.082
0.466
7.01
6
0.70
4.95
0.085
0.412
8.32
7
0.70
5.88
0.094
0.388
10.51
5
0.70
5.31
0.078
0.445
8.28
6
0.13
1.85
0.029
0.305
4.19
4
0.77
2.30 a
0.83 a
0.336 a
4.74
VDss L/kg
fu
CL, L/h/kg
t1/2, h
a [58].
2.9. Molecular Dynamics Simulation of the Complex 4b-AChE
The interactions between the most active GAL-CU hybrid 4b with the enzyme AChE were
investigated further by molecular dynamics (MD), simulating the complex 4b-AChE for 100 ns in
explicit water molecules as a solvent. The compound remained inserted within the binding site and
the complex was stable during the simulation. The analysis of the MD trajectory in the vicinity of the
phenyl –OCH3 revealed that the O-atom makes an additional hydrogen bond with Tyr72 (Figure 7)
which is absent in the complex 4a-AChE. This additional hydrogen bond might be crucial for the
better binding to the enzyme.

Molecules 2020, 25, 3341
11 of 18
Figure 7. Hydrogen bonds (in yellow dashes) in the complex 4b-AChE. Most of the time, the GAL’s
ammonium group makes a hydrogen bond with Tyr337. At the given snapshot (45 ns), the N-atom
makes an intramolecular hydrogen bond with the linker carbonyl oxygen. A second hydrogen bond
is formed between the methoxyphenyl O-atom and Tyr72. The latter bond determines the higher
affinity of 4b to AChE.
3. Discussion
A combinatorial library of 72 GAL-CU hybrid compounds was designed in the present study as
dual-site binding AChE inhibitors. They were constructed from GAL binding core, eight aromatic
substituents mimicking the CU molecule and nine linkers, gathered in five groups. Initially, the
compounds were screened for GI absorption, BBB permeability and target specificity. Forty-four
compounds passed these tests and moved to molecular docking on AChE. The GoldScores of the
tested compounds showed that all of them were potential good binders, better than GAL and CU
themselves. The compounds were checked for synthetic feasibility and 14 of them were selected for
further synthesis. The synthesized hybrids were tested initially for neurotoxicity on Neuro-2A cells,
and those of them that were less toxic than CU underwent a test for affinity to AChE. The anti-AChE
tests showed that all five hybrids were between 41 and 186 times more active than GAL.
The best-scored docking poses of the tested for AChE affinity compounds 4a, 4b, 4e, 4f and 8b
and their interactions with the enzyme are visualised at Figure 8. The GAL binding core is well
embedded in the bottom of AChE gorge, forming a dense network of hydrogen bonds: between
GAL’s ammonium group and Tyr337, between OCH and Ser203, and between OH and Glu202. The
3
structural water molecule here is involved in the network as well. A cation-π interaction is formed
between the positively charged ammonium atom and Tyr337 and π-π contacts between the aromatic
ring and His447 (Figure 8A and B). An extra T-shaped π-π contact is formed between the aromatic
rings of 4a, 4b and 4f and Tyr124 (Figure 8A).

Molecules 2020, 25, 3341
12 of 18
A
B
C
Figure 8. Best-scored docking poses between the compounds 4a,b,f (A), 4e (B), and 8f (C) and AChE.
Hydrogen bonds are shown in yellow dashes, cation-π interactions are shown in blue lines and π-π
(including p-π) interactions, are given in red lines. The structural water molecule within the binding
site is presented in ball-and-stick. Interactions are visualized by YASARA (http://www.yasara.org).
2
The linker also takes part in the interaction with the enzyme. The sp C atoms make p-π contacts
with Tyr72 and/or Trp286. Additionally, the oxygen atom of the carbonyl group forms a hydrogen
bond with Tyr124 (4e and 8b, Figure 8B and C). Surprisingly, no π-π stacking was observed between
the linker’s benzene and any of the aromatic residues in the binding site. The terminal aromatic
moieties fit well into the PAS. There are two different orientations, but all of them make π-π stacking
with Trp286 and/or Tyr72.
The newly synthesized GAL-CU hybrids are weak bases, highly lipophilic, moderate or weakly
ionized at physiological pH. They bind to plasma proteins in >90%, distribute extensively, cross the
BBB, and are cleared slowly with half-lives up to 12 h.
A special attention deserves compound 4b as the most active and the least toxic compound in
the set. It is very close to 4a and 4f, having only one additional methoxy group bound in the terminal
phenyl ring. The molecular docking did not detect any significant difference between 4a, 4b and 4f
in their interactions with the enzyme AChE. However, the two-and-a-half-fold difference in the
activities compared to 4a made us look at the complex 4b-AChE by MD. The MD simulation showed
that the phenyl -OCH
3
group of 4b forms an additional hydrogen bond with Tyr72 (Figure 7). At
some of the initial frames, His287 also is coming quite close to -OCH
3
but not enough to form a long-
lasting hydrogen bond. The additional hydrogen bond with Tyr72 which is absent in the complex 4a-
AChE explains the better binding of 4b to AChE.
In summary, the designed, synthesized and tested in the present study hybrid molecules
between GAL and CU, showed higher affinities to AChE and less neurotoxicities than both
galantamine and curcumin themselves. The novel structures have optimal ADME properties and are
able to cross the BBB. They could be considered as perspective multi-target drug candidates. The
most active compound 4b will be directed to more detailed in vitro and in vivo studies.
4. Materials and Methods
4.1. Compounds
Galantamine hydrobromide was purchased from Galen-N Ltd., Sofia, Bulgaria. 1,5-Pentandiol,
98%; 1,3-benzenedimethanol, 98%; acetophenone, 98%; benzaldehyde, 98%; 4-methoxybenzaldehyde,
98%; piperonyl alcohol, 98%; 4-methylbenzaldehyde, 97%; 3-methylbenzaldehyde, 97%; 3,4-
dimethylbenzaldehyde; iodine, resublimiert p.a.; triphenylphosphine, 99%; L-(-)-proline, 99+%; tert-

Molecules 2020, 25, 3341
13 of 18
butyldimethylchlorsilan, 98%; manganese(IV) oxide, 88%, electrolytically precipitated, active;
sodium hydride, 60% dispersion in mineral oil; pyridinium chlorochromate, 98%; lithium hydroxide
monohydrate, 98+%; hydrobromic acid, pure, ca. 48 wt% solution in water were purchased from
Acros Organics, Alfa Aeser or Merck, Sofia, Bulgaria.
4.2. ADME Prediction
Two free online servers were used to predict blood brain barrier (BBB) permeability of the
studied compounds—SwissADME (http://www.swissadme.ch) [59] and BBB Predictor
(https://www.cbligand.org/BBB/predictor.php) [60]. SwissADME server predicts permeability based
on BOILED-Egg (brain or intestinal estimated permeation) method, based on the lipophilicity and
polarity of small molecules [61]. The BBB predictor uses support vector machine (SVM) and
LiCABEDS (ligand classifier of adaptively boosting ensemble decision stumps) algorithms on four
types of fingerprints for 1593 compounds with known BBB permeability [62]. Additionally, GI
absorption and PAINS (pan assay interference structures) alerts of the compounds were checked by
SwissADME online platform. The GI absorption prediction is based on the BOILED-Egg method [61].
PAINS alerts help identification of frequent hitters or promiscuous compounds in many biochemical
high-throughput screens based on substructural features [63].
4.3. Virtual Screening by Molecular Docking
The hybrid molecules were constructed with Avogadro software and minimized with MMFF94s
force field, then structures were docked into the X-ray structure of human recombinant
acetylcholinesterase (rhAChE, pdb ID: 4EY6, R = 2.40 Å) [20]. The docking simulations were performed
by GOLD v.5.2.2 (CCDC Ltd., Cambridge, UK) using a protocol previously optimized in terms of
scoring function, rigid/flexible ligand and binding site, radius of the binding site, presence/absence of
a water molecule (HOH846) within the binding site, number of genetic algorithm (GA) runs [41–44].
The docking simulations in the present study were performed at the following settings: scoring
function GoldScore, flexible ligand, flexible binding site, radius of the binding site 10Å, a structural
water molecule within the binding sire (HOH 846), 100 GA runs. The following amino acid residues
were set as flexible during docking calculations: Tyr72, Asp74, Trp86, Tyr124, Ser125, Trp286, Phe297,
Tyr337, Phe338, Tyr341. GAL from X-ray structure was used as a referent molecule. The best-scored
compounds were analyzed for synthetic feasibility and protein-ligand interactions.
4.4. Neurotoxicity Test
Murine neuroblastoma Neuro-2A cells (German collection DSMZ, Braunschweig, Germany)
were cultivated under standard conditions: complete medium (90% DMEM, 10% heat-inactivated
FBS, and non-essential amino acids); 37 °C and 5% CO2 in fully humidified atmosphere. The cell line
was kept in the logarithmic growth phase by splitting 1:4 once a week using trypsin/EDTA. About
30% of the cells grew like neuronal cells. For the experimental evaluation of the cytotoxicity Neuro-
2A, cells were plated in 96-well flat-bottomed cell culture plates at the recommended density of 1 ×
6
2
10 cells/25cm . After 24 h, the cells were treated with various concentrations of the investigational
compounds, and after 72-h incubation, an MTT-dye reduction assay was performed [64]. Briefly, at
the end of incubation, an MTT stock solution (10mg/mL in PBS) was added (0.01 mL/well). Plates
were further incubated at 37 °C for 4 h. Next, the formazan crystals were dissolved by the addition
of 0.110 mL/well 5% formic acid in 2-propanol (v/v). Absorption was measured at 580nm wavelength
on an automated ELISA reader Labexim LMR1. At least six wells per concentration were used, and
data were processed using the GraphPad Prism 5.0 software 2.0 (San Diego, CA, USA).
4.5. Assessment of AChE Inhibitory Activity
AChE activity was assayed as described by Ellman et al. [65], with some modifications [58].
Fifteen µL of Electrophorus electricus AChE (Sigma-Aldrich, Darmstadt, Germany) in buffer phosphate
(pH 7.6) and 15 µL of the tested compounds (1.4 – 4350 µM in methanol) were dissolved in 200 µL in

Molecules 2020, 25, 3341
14 of 18
the same buffer, and plated in 96-well plates. The mixtures were incubated for 30 min at room
temperature before the addition of 30 µL of the substrate solution (15 µL 0.5 M DTNB, 15 µL 0.6 mM
ATCI in buffer, pH 7.6). The absorbance was read on a Microplate Reader Biochrom EZ 800 at 403
nm, after three minutes. Enzyme activity was calculated as a percentage compared to an assay using
a buffer without any inhibitor, according to the following equation:
AChE inhibition (%) = ((Abscontrol − Abssample)/Abscontrol) × 100
(1)
where Abscontrol is the absorbance of the control, containing 15 µL enzyme in buffer, 200 µL buffer and 15
µL solvent (methanol/water for GAL); Abssample is the absorbance of the sample, containing 15 µL enzyme
in buffer, 200 µL buffer and 15 µL solution of the tested compound (in methanol/water for GAL).
4.6. Calculation of Physicochemical Properties and Prediction of Pharmacokinetic (PK) Parameters
The main physicochemical properties pKa, logP, logD7.4, PSA, number of HBD and HBA in the
molecules of the tested compounds were calculated using ACD/logD v. 9.08 (ACD Inc., Toronto,
Canada). As the novel hybrids are weak bases, at physiological pH they exist as cations as well as
neutral molecules. The ionized fraction at pH = 7.4 (fBH+) was calculated according to the equation:
ꢄ
ꢃ
ꢀꢁꢂ
=
.
ꢌ
(2)
(ꢇ.ꢈꢉꢊꢋ
)
ꢄꢅꢄꢆ
The key PK parameters were predicted by quantitative structure-activity relationships (QSPRs)
models derived previously [53–57]. Three PK parameters were modelled: the steady state volume of
ss
distribution (VD ), free fraction of drug in plasma (fu) and unbound clearance (CLu). The CL and half-
life (t1/2) were calculated following the equations:
ꢍꢎ
ꢍꢎꢏ =
(3)
(4)
ꢀ
ꢏ
ꢒꢓ2 × ꢔꢕꢖꢖ
ꢍꢎ
ꢐꢄ/ꢑ
=
The AChEI galantamine (GAL) is given as a reference compound.
4.7. Molecular Dynamics Protocol
The best-scored pose of the most active hybrid in complex with AChE was used as a starting
structure for the MD simulation by Amber 18 [66]. The small molecule was parametrized using
GAFF2.11 force field and AM1-BCC charges, and the complex was solvated in truncated octahedral
box with TIP3P water and 0.15 M NaCl. The system was subjected to energy minimization, heating
to 300 K, density equilibration, preproduction equilibration and production dynamics for 100 ns.
Frames were saved every 0.1 ns.
Supplementary Materials: The supplementary material contains: synthesis and analytical data of compounds
4a–c,e–h; synthesis and analytical data of compounds 6a,b; synthesis and analytical data of compounds
8b,c,f,g,h; copies of 1H and 13C NMR spectra for the target compounds.
Author Contributions: Conceptualization: I.D.; Design of combinatorial library: I.D., G.S., I.P., A.L.; ADME and
BBB screening: A.L., M.A.; Docking: A.L., M.A.; Synthesis: G.S., I.P.; Neurotoxicity test: T.A., A.L., S.K.; AChE
Activity: D.Z.; PK Predictions: Z.D.Z.; MD simulations: M.A, S.I.
Funding: This work was supported by the Bulgarian National Science Fund (Grant DN03/9/2016) and Grant No
BG05M2OP001-1.001-0003, financed by the Science and Education for Smart Growth Operational Program
(2014–2020) and co-financed by the European Union through the European Structural and Investment funds.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Tata, A.M.; Velluto, L.; D’Angelo, C.; Reale, M. Cholinergic system dysfunction and neurodegenerative
diseases: cause or effect? CNS Neurol. Disord. Drug Targets 2014, 13, 1294–1303,
doi:10.2174/1871527313666140917121132.

Molecules 2020, 25, 3341
15 of 18
2.
Inestrosa, N.C.; Dinamarca, M.C.; Alvarez, A. Amyloid-cholinesterase interactions. Implications for
Alzheimer’s disease. FEBS J. 2008, 275, 625–632.
3.
4.
Rambaran, R.N.; Serpell, L.C. Amyloid fibrils. Abnormal protein assembly. Prion 2008, 2, 112–117.
Bartolucci, C.; Haller, L.A.; Jordis, U.; Fels, G.; Lamba, D. ProbingTorpedo californicaAcetylcholinesterase
Catalytic Gorge with Two Novel Bis-functional Galanthamine Derivatives. J. Med. Chem. 2010, 53, 745–751,
doi:10.1021/jm901296p.
5.
6.
Kozurkova, M.; Hamulakova, S.; Gazova, Z.; Paulíková, H.; Kristian, P. Neuroactive Multifunctional
Tacrine Congeners with Cholinesterase, Anti-Amyloid Aggregation and Neuroprotective Properties.
Pharmaceuticals 2011, 4, 382–418, doi:10.3390/ph4020382.
Scherzer-Attali, R.; Pellarin, R.; Convertino, M.; Frydman-Marom, A.; Egoz-Matia, N.; Peled, S.; Levy-Sakin,
M.; Shalev, D.E.; Caflisch, A.; Gazit, E.; et al. Complete Phenotypic Recovery of an Alzheimer’s Disease
Model by
a Quinone-Tryptophan Hybrid Aggregation Inhibitor. PLoS ONE 2010, 5, e11101,
doi:10.1371/journal.pone.0011101.
7.
Bolea, I.; Juárez-Jiménez, J.; Ríos, C.D.L.; Chioua, M.; Pouplana, R.; Estarellas, C.; Unzeta, M.; Marco-
Contelles, J.; Samadi, A. Synthesis, Biological Evaluation, and Molecular Modeling of Donepezil andN-[(5-
(Benzyloxy)-1-methyl-1H-indol-2-yl)methyl]-N-methylprop-2-yn-1-amine Hybrids as New Multipotent
Cholinesterase/Monoamine Oxidase Inhibitors for the Treatment of Alzheimer’s Disease. J. Med. Chem. 2011,
54, 8251–8270, doi:10.1021/jm200853t.
8.
9.
Simoni, E.; Daniele, S.; Bottegoni, G.; Pizzirani, D.; Trincavelli, M.L.; Goldoni, L.; Tarozzo, G.; Reggiani, A.;
Martini, C.; Piomelli, D.; et al. Combining Galantamine and Memantine in Multitargeted, New Chemical
Entities Potentially Useful in Alzheimer’s Disease. J. Med. Chem. 2012, 55, 9708–9721, doi:10.1021/jm3009458.
Marco-Contelles, J.; Bautista-Aguilera, Óscar, M.; Esteban, G.; Chioua, M.; Nikolic, K.; Agbaba, D.;
Moraleda, I.; Iriepa, I.; Soriano, E.; Samadi, A.; et al. Multipotent cholinesterase/monoamine oxidase
inhibitors for the treatment of Alzheimer’s disease: Design, synthesis, biochemical evaluation, ADMET,
molecular modeling, and QSAR analysis of novel donepezil-pyridyl hybrids. Drug Des. Dev. Ther. 2014, 8,
1893–1910, doi:10.2147/DDDT.S69258.
10. Wang, L.; Esteban, G.; Ojima, M.; Bautista-Aguilera, O.M.; Inokuchi, T.; Moraleda, I.; Iriepa, I.; Samadi, A.;
Youdim, M.B.; Romero, A.; Soriano, E.; Herrero, R.; Fernández Fernández, A.P.; et al. Donepezil +
propargylamine + 8-hydroxy- quinoline hybrids as new multifunctional metal-chelators, ChE and MAO
inhibitors for the potential treatment of Alzheimer’s disease. Eur. J. Med. Chem. 2014, 80, 543–561.
11. Nepovimova, E.; Uliassi, E.; Korabecny, J.; Altamira, L.E.P.; Samez, S.; Pesaresi, A.; Garcia, G.E.; Bartolini,
M.; Andrisano, V.; Bergamini, C.; et al. Multitarget Drug Design Strategy: Quinone–Tacrine Hybrids
Designed To Block Amyloid-β Aggregation and To Exert Anticholinesterase and Antioxidant Effects. J.
Med. Chem. 2014, 57, 8576–8589, doi:10.1021/jm5010804.
12. Korabecny, J.; Andrs, M.; Nepovimova, E.; Doležal, R.; Babkova, K.; Horova, A.; Maliňák, D.; Mezeiová, E.;
Gorecki, L.; Hepnarova, V.; et al. 7-Methoxytacrine-p-Anisidine Hybrids as Novel Dual Binding Site
Acetylcholinesterase Inhibitors for Alzheimer’s Disease Treatment. Molecules 2015, 20, 22084–22101,
doi:10.3390/molecules201219836.
13. Singh, M.; Kaur, M.; Chadha, N.; Silakari, O. Hybrids: A new paradigm to treat Alzheimer’s disease. Mol.
Divers. 2015, 20, 271–297, doi:10.1007/s11030-015-9628-9.
14. Piazzi, L.; Rampa, A.; Bisi, A.; Gobbi, S.; Belluti, F.; Cavalli, A.; Bartolini, M.; Andrisano, V.; Valenti, P.;
Recanatini, M. 3-(4-{[Benzyl(methyl)amino]methyl}phenyl)-6,7-dimethoxy-2H-2-chromenone (AP2238)
Inhibits Both Acetylcholinesterase and Acetylcholinesterase-Induced β-Amyloid Aggregation: A Dual
Function Lead for Alzheimer’s Disease Therapy. J. Med. Chem. 2003, 46, 2279–2282, doi:10.1021/jm0340602.
15. Hernández-Rodríguez, M.; Correa-Basurto, J.; Martínez-Ramos, F.; Padilla-Martínez, I.I.; Benítez-Cardoza,
C.G.; Mera-Jiménez, E.; Rosales-Hernández, M.C. Design of Multi-Target Compounds as AChE, BACE1,
and Amyloid-β1-42 Oligomerization Inhibitors: In Silico and In Vitro Studies. J. Alzheimer’s Dis. 2014, 41,
1073–1085, doi:10.3233/jad-140471.
16. Meena, P.; Nemaysh, V.; Khatri, M.; Manral, A.; Luthra, P.M.; Tiwari, M. Synthesis, biological evaluation
and molecular docking study of novel piperidine and piperazine derivatives as multi-targeted agents to
treat Alzheimer’s disease. Bioorganic Med. Chem. 2015, 23, 1135–1148, doi:10.1016/j.bmc.2014.12.057.
17. Guzior, N.; Bajda, M.; Skrok, M.; Kurpiewska, K.; Lewiński, K.; Brus, B.; Pišlar, A.; Kos, J.; Gobec, S.;
Malawska, B. Development of multifunctional, heterodimeric isoindoline-1,3-dione derivatives as

Molecules 2020, 25, 3341
16 of 18
cholinesterase and β-amyloid aggregation inhibitors with neuroprotective properties. Eur. J. Med. Chem.
2015, 92, 738–749, doi:10.1016/j.ejmech.2015.01.027.
18. Basile, L. Virtual Screening in the Search of New and Potent Anti-Alzheimer Agents; Humana Press, New
York, NY, USA, 2017; Volume 132, pp. 107–137.
19. Hung, S.-Y.; Fu, W.-M. Drug candidates in clinical trials for Alzheimer’s disease. J. Biomed. Sci. 2017, 24, 47,
doi:10.1186/s12929-017-0355-7.
20. Cheung, J.; Rudolph, M.J.; Burshteyn, F.; Cassidy, M.S.; Gary, E.N.; Love, J.; Franklin, M.C.; Height, J.J.
Structures of Human Acetylcholinesterase in Complex with Pharmacologically Important Ligands. J. Med.
Chem. 2012, 55, 10282–10286, doi:10.1021/jm300871x.
21. Dajas-Bailador, F.; Heimala, K.; Wonnacott, S. The Allosteric Potentiation of Nicotinic Acetylcholine
Receptors by Galantamine Is Transduced into Cellular Responses in Neurons: Ca2+ Signals and
Neurotransmitter Release. Mol. Pharmacol. 2003, 64, 1217–1226, doi:10.1124/mol.64.5.1217.
22. Takata, K.; Kitamura, Y.; Saeki, M.; Terada, M.; Kagitani, S.; Kitamura, R.; Fujikawa, Y.; Maelicke, A.;
Tomimoto, H.; Taniguchi, T.; et al. Galantamine-induced Amyloid-β Clearance Mediated via Stimulation
of Microglial Nicotinic Acetylcholine Receptors. J. Boil. Chem. 2010, 285, 40180–40191,
doi:10.1074/jbc.m110.142356.
23. Han, S.; Sweeney, J.; Bachman, E.; Schweiger, E.; Forloni, G.; Coyle, J.; Davis, B.; Joullie, M. Chemical and
pharmacological characterization of galanthamine, an acetylcholinesterase inhibitor, and its derivatives. A
potential application in Alzheimer’s disease? Eur. J. Med. Chem. 1992, 27, 673–687, doi:10.1016/0223-
5234(92)90087-h.
24. Poschalko, A.; Welzig, S.; Treu, M.; Nerdinger, S.; Mereiter, K.; Jordis, U. Synthesis of (±)-6Hbenzofuro
[3a,3,2,ef][3] benzazepine, an unnatural analog of (-)-galanthamine. Tetrahedron 2002, 58, 1513–1518.
25. Jordis, U.; Treu, M.; Mereiter, K. 12H-[2]-Benzothiepino[6 ,5a,5-bc]benzofuran: Synthesis of a Sulfur-Analog
of Galanthamine. HETEROCYCLES 2001, 55, 1727, doi:10.3987/com-01-9278.
26. Bores, G.M.; Kosley, R.W. Galanthamine derivatives for the treatment of Alzheimer’s disease. Drugs Future
1996, 21, 621–635.
27. Mary, A.; Renko, D.Z.; Guillou, C.; Thal, C. Potent acetylcholinesterase inhibitors—design, synthesis, and
structure-activity relationship of bisinteracting ligands in the galanthamine series. Bioorg. Med. Chem. 1998,
6, 1835–1850.
28. Guillou, C.; Mary, A.; Renko, D.Z.; Gras, E.; Thal, C. Potent acetylcholinesterase inhibitors: Design,
synthesis and structure-activity relationships of alkylene linked bis-galanthamine and galanthamine-
galanthaminium salts. Bioorganic Med. Chem. Lett. 2000, 10, 637–639, doi:10.1016/s0960-894x(00)00059-7.
29. Herlem, D.; Martin, M.T.; Thal, C.; Guillou, C. Synthesis and structure-activity relationships of open D-
Ring galanthamine analogues. Bioorganic Med. Chem. Lett. 2003, 13, 2389–2391.
30. Czollner, L.; Frohlich, J.; Jordis, U.; Kalz, B.; Frantsits, W.J.; Welzig, S. Novel Derivatives of 4a,5,9,10,11,12-
hexahydrobenzofuro[3a,3,2][2]-benzazepine, Method for the Production Thereof and use Thereof in the
Production of Medicaments. Int. Patent WO 2005/030333, 7 April 2005.
31. Jia, P.; Sheng, R.; Zhang, J.; Fang, L.; He, Q.; Yang, B.; Hu, Y. Design, synthesis and evaluation of
galanthamine derivatives as acetylcholinesterase inhibitors. Eur. J. Med. Chem. 2009, 44, 772–784,
doi:10.1016/j.ejmech.2008.04.018.
32. Zhao, L.N.; Chiu, S.W.; Benoit, J.; Chew, L.Y.; Mu, Y. The effect of curcumin on the stability of Aβ dimers.
J. Phys. Chem. B 2012, 116, 7428–7435.
33. Masuda, Y.; Fukuchi, M.; Yatagawa, T.; Tada, M.; Takeda, K.; Irie, K.; Akagi, K.; Monobe, Y.; Imazawa, T.;
Takegoshi, K. Solid-state NMR analysis of interaction sites of curcumin and 42-residue amyloid beta-
protein fibrils. Bioorg. Med. Chem. 2011, 19, 5967–5974.
34. Orlando, R.; Gonzales, A.M.; Royer, R.E.; Deck, L.M.; Jagt, D.L.V. A Chemical Analog of Curcumin as an
Improved Inhibitor of Amyloid Abeta Oligomerization. PLoS ONE 2012, 7, e31869,
doi:10.1371/journal.pone.0031869.
35. Balasubramanian†, K. Molecular Orbital Basis for Yellow Curry Spice Curcumin’s Prevention of
Alzheimer’s Disease. J. Agric. Food Chem. 2006, 54, 3512–3520, doi:10.1021/jf0603533.
36. Atanasova, M.; Stavrakov, G.; Philipova, I.; Zheleva-Dimitrova, D.; Yordanov, N.; Doytchinova, I.
Galantamine derivatives with indole moiety: Docking, design, synthesis and acetylcholinesterase
inhibitory activity. Bioorganic Med. Chem. 2015, 23, 5382–5389, doi:10.1016/j.bmc.2015.07.058.

Molecules 2020, 25, 3341
17 of 18
37. Stavrakov, G.; Philipova, I.; Zheleva-Dimitrova, D.; Atanasova, M.; Konstantinov, S.; Doytchinova, I.
Docking-based Design of Galantamine Derivatives with Dual-site Binding to Acetylcholinesterase. Mol.
Informatics 2016, 35, 278–285, doi:10.1002/minf.201600041.
38. Stavrakov, G.; Philipova, I.; Zheleva-Dimitrova, D.; Valkova, I.; Salamanova, E.; Konstantinov, S.;
Doytchinova, I. Docking-based design and synthesis of galantamine-NDASH-camphane hybrids as
inhibitors of acetylcholinesterase. Chem. Boil. Drug Des. 2017, 90, 709–718, doi:10.1111/cbdd.12991.
39. Atanasova, M.; Yordanov, N.; Dimitrov, I.; Berkov, S.; Doytchinova, I. Molecular Docking Study on
Galantamine Derivatives as Cholinesterase Inhibitors. Mol. Informatics 2015, 34, 394–403,
doi:10.1002/minf.201400145.
40. McDougal, P.G.; Rico, J.G.; Oh, Y.I.; Condon, B.D. A convenient procedure for the monosilylation of
symmetric 1,n-diols. J. Org. Chem. 1986, 51, 3388–3390, doi:10.1021/jo00367a033.
41. Diez, D.; Nunez, M.G.; Beneitez, A.; Moro, R.F.; Marcos, I.S.; Basabe, P.; Broughton, H.B.; Urones, J.G.
Organocatalytic cynthesis of an alkyltetrahydropyran. Synlett 2009, 3, 390–394.
42. Chong, J.M.; Heuft, M.A.; Rabbat. P. Solvent effects on the monobromination of alpha,omega-diols: A
convenient preparation of omega-bromoalkanols. J. Org. Chem. 2000, 65, 5837–5838.
43. Kelkar, S.V.; Joshi, G.S.; Reddy, G.B.; Kulkarni, G.H. A Novel Approach to the Synthesis of 7(Z), 11(Z)-
Nonacosadiene, Pheromone of Drosophila Melano Gaster and 9(Z)-Tricosene, Pheromone of Musca
Domestica. Synth. Commun. 1989, 19, 1369–1379, doi:10.1080/00397918908054546.
44. Wang, J.-F.; Lei, M.; Li, Q.; Ge, Z.-M.; Wang, X.; Li, R. A novel and efficient direct aldol condensation from
ketones and aromatic aldehydes catalyzed by proline–TEA through a new pathway. Tetrahedron 2009, 65,
4826–4833, doi:10.1016/j.tet.2009.04.052.
45. Chen, L.; Tilley, J.; Trilles, R.V.; Yun, W.; Fry, D.; Cook, C.; Rowan, K.; Schwinge, V.; Campbell, R. N-acyl-
L-phenylalanine derivatives as potent VLA-4 antagonists that mimic a cyclic peptide conformation. Bioorg.
Med. Chem. Lett. 2002, 12, 137–140, doi:10.1016/s0960-894x(01)00711-9.
46. Roman, B.; De Ryck, T.; Verhasselt, S.; Bracke, M.; Stevens, C.V. Further studies on anti-invasive
chemotypes: An excursion from chalcones to curcuminoids. Bioorganic Med. Chem. Lett. 2015, 25, 1021–1025,
doi:10.1016/j.bmcl.2015.01.027.
47. Kumar, S.; Thorat, K.G.; Ravikanth, M. Synthesis and Properties of Covalently Linked AzaBODIPY–
BODIPY Dyads and AzaBODIPY-(BODIPY)2 Triads. J. Org. Chem. 2017, 82, 6568–6577,
doi:10.1021/acs.joc.7b00542.
48. Ahmed, T.; Gilani, A.H. Inhibitory effect of curcuminoids on acetylcholinesterase activity and attenuation
of scopolamine-induced amnesia may explain medicinal use of turmeric in Alzheimer’s disease. Pharmacol.
Biochem. Behav. 2009, 91, 554–559, doi:10.1016/j.pbb.2008.09.010.
49. Ghafourian, T.; Barzegar-Jalali, M.; Hakimiha, N.; Cronin, M.T.D.; Barzegar-Jalali, M. Quantitative
structure-pharmacokinetic relationship modelling: Apparent volume of distribution. J. Pharm. Pharmacol.
2004, 56, 339–350, doi:10.1211/0022357022890.
50. Ghafourian, T.; Barzegar-Jalali, M.; Dastmalchi, S.; Khavarikhorasani, T.; Hakimiha, N.; Nokhodchi, A.
QSPR models for the prediction of apparent volume of distribution. Int. J. Pharm. 2006, 319, 82–97,
doi:10.1016/j.ijpharm.2006.03.043.
51. Lombardo, F.; Obach, R.S.; DiCapua, F.M.; Bakken, G.A.; Lu, J.; Potter, D.M.; Gao, F.; Miller, M.D.; Zhang,
Y. Hibrid mixture discriminant analysis—random forest computational model for the prediction of volume
of distribution of drugs in human. J. Med. Chem. 2006, 49, 2262–2267.
52. Smith, D.A.; Allerton, C.; Kalgutkar, A.; van der Waterbeemd, H.; Walker, D.K. Pharmacokinetics and
Metabolism in Drug Design, 3rd ed.; Wiley-VCH, Weinheim, Germany, 2012; p. 70.
53. Zhivkova, Z.; Mandova, T.; Doytchinova, I. Quantitative Structure—Pharmacokinetics Relationships
Analysis of Basic Drugs: Volume of Distribution. J. Pharm. Pharm. Sci. 2015, 18, 515–527, doi:10.18433/j3xc7s.
54. Obach, R.S.; Lombardo, F.; Waters, N. Trend Analysis of a Database of Intravenous Pharmacokinetic
Parameters in Humans for 670 Drug Compounds. Drug Metab. Dispos. 2008, 36, 1385–1405,
doi:10.1124/dmd.108.020479.
55. Zhivkova, Z. Quantitative structure—Pharmacokinetics relationship for the steady state volume of
distribution of basic and neutral drugs. J. Pharm. Pharm. Sci. 2018, 7, 94–105.
56. Zhivkova, Z. Quantitative structure—Pharmacokinetics relationships for plasma protein binding of basic
drugs. J. Pharm. Pharm. Sci. 2017, 20, 349–359.

Molecules 2020, 25, 3341
18 of 18
57. Zhivkova, Z. Quantitative structure—Pharmacokinetics relationships for clearance of basic drugs with
consideration of the major elimination pathay. J. Pharm. Pharm. Sci. 2017, 20, 135–137.
58. Ortiz, J.E.; Berkov, S.; Pigni, N.B.; Theoduloz, C.; Roitman, G.; Tapia, A.; Bastida, J.; Feresin, G.E. Wild
Argentinian Amaryllidaceae, a New Renewable Source of the Acetylcholinesterase Inhibitor Galanthamine
and Other Alkaloids. Molecules 2012, 17, 13473–13482, doi:10.3390/molecules171113473.
59. Daina, A.; Michielin, O.; Zoete, V. SwissADME: A free web tool to evaluate pharmacokinetics, drug-
likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. 2017, 7, 42717,
doi:10.1038/srep42717.
60. Liu, H.; Wang, L.; Lv, M.; Pei, R.; Li, P.; Pei, Z.; Wang, Y.; Su, W.; Xie, X.-Q. AlzPlatform: An Alzheimer’s
Disease Domain-Specific Chemogenomics Knowledgebase for Polypharmacology and Target
Identification Research. J. Chem. Inf. Model. 2014, 54, 1050–1060, doi:10.1021/ci500004h.
61. Daina, A.; Zoete, V. A BOILED-Egg To Predict Gastrointestinal Absorption and Brain Penetration of Small
Molecules. ChemMedChem 2016, 11, 1117–1121, doi:10.1002/cmdc.201600182.
62. Ma, C.; Wang, L.; Xie, X.-Q. Ligand Classifier of Adaptively Boosting Ensemble Decision Stumps
(LiCABEDS) and Its Application on Modeling Ligand Functionality for 5HT-Subtype GPCR Families. J.
Chem. Inf. Model. 2011, 51, 521–531, doi:10.1021/ci100399j.
63. Whitty, A. Growing PAINS in academic drug discovery. Futur. Med. Chem. 2011, 3, 797–801,
doi:10.4155/fmc.11.44.
64. Momekov, G.; Ferdinandov, D.; Bakalova, A.; Zaharieva, M.; Konstantinov, S.; Karaivanova, M. In vitro
toxicological evaluation of a dinuclear platinum(II) complex with acetate ligands. Arch. Toxicol. 2006, 80,
555–560, doi:10.1007/s00204-006-0078-0.
65. Ellman, G.L.; Courtney, K.; Andres, V.; Featherstone, R.M. A new and rapid colorimetric determination of
acetylcholinesterase activity. Biochem. Pharmacol. 1961, 7, 88–95, doi:10.1016/0006-2952(61)90145-9.
66. Case, D.A.; Ben-Shalom, I.Y.; Brozell, S.R.; Cerutti, D.S.; Cheatham III, T.E.; Cruzeiro, V.W.D.; Darden, T.A.;
Duke, R.E.; Ghoreishi, D.; Gilson, M.K.; Gohlke, et al. AMBER 2018, University of California, San Francisco,
CA, USA, 2018.
Sample Availability: Samples of the compounds 4a–c,e–h, 6a,b, 8b,c,f,g,h are available from the authors.
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).