Design, synthesis, biological evaluation and molecular modelling of substituted pyrrolo[2,1-: A] isoquinolinone derivatives: Discovery of potent inhibitors of AChE and BChE

Article abstract of DOI:10.1039/d1nj00345c

We report here the design, synthesis and biological evaluation of a new series of substituted pyrrolo[2,1-a]isoquinolin-3-one derivatives, some of which have strong inhibitory activity against both AChE and BChE enzymes. The design of these new inhibitors was carried out taking rivastigmine as the starting structure. Thus, on the basis of an exhausting molecular modeling study using combined techniques (docking, dynamic molecular simulations and QTAIM calculations), we obtained new ligands possessing stronger inhibitory effects than rivastigmine, the reference compound. QTAIM analysis gave us detailed information about the molecular interactions stabilizing the different ligand-enzyme complexes. These calculations showed the importance of the interaction with the CAS esteratic site for the inhibitory effect of these compounds. Nevertheless, they also indicated that the combination of interactions with CAS and strong interactions with the PAS site might be beneficial for the inhibitory effect.

Full text of DOI:10.1039/d1nj00345c

NJC
View Article Online
View Journal | View Issue
PAPER
Design, synthesis, biological evaluation and
molecular modelling of substituted pyrrolo[2,1-a]
isoquinolinone derivatives: discovery of potent
inhibitors of AChE and BChE†
Cite this: New J. Chem., 2021,
45, 8321
a
b
c
Oscar Parravicini,
Francisco Garibotto,
Diego Cortes
Emilio Angelina,
Roque Spinelli,
a
c
d
de
´
Alvaro S. Siano,
and Ricardo D. Enriz
Laura Vila,
Nuria Cabedo,
e
a
*
We report here the design, synthesis and biological evaluation of a new series of substituted pyrrolo[2,1-a]
isoquinolin-3-one derivatives, some of which have strong inhibitory activity against both AChE and BChE
enzymes. The design of these new inhibitors was carried out taking rivastigmine as the starting structure.
Thus, on the basis of an exhausting molecular modeling study using combined techniques (docking,
dynamic molecular simulations and QTAIM calculations), we obtained new ligands possessing stronger
inhibitory effects than rivastigmine, the reference compound. QTAIM analysis gave us detailed information
about the molecular interactions stabilizing the different ligand–enzyme complexes. These calculations
showed the importance of the interaction with the CAS esteratic site for the inhibitory effect of these
compounds. Nevertheless, they also indicated that the combination of interactions with CAS and strong
interactions with the PAS site might be beneficial for the inhibitory effect.
Received 21st January 2021,
Accepted 25th March 2021
DOI: 10.1039/d1nj00345c
rsc.li/njc
few patients having received a formal diagnosis of AD. Further-
more, despite all research efforts, therapeutic options for AD are
1. Introduction
Currently, Alzheimer’s disease (AD) is without doubt a very very limited. At this time, the only treatments available with
serious problem for our society, but it is even more complex for clinical evidence to AD patients are the cholinesterase inhibitors
the near future. About fifty million people are currently living (ChEI) donepezil, rivastigmine and galantamine for mild to
with Alzheimer’s disease (AD).1 According to the latest Center moderate AD, and memantine (an NMDA receptor antagonist)
for Disease Control and Prevention (CDC) statistics, while which has been approved for moderate to severe AD. All four
other diseases have decreased like, for example, cardiovascular compounds have been rigorously tested in clinical trials and
disorders, those involving AD complications have increased. have been proved to improve patients’ reported outcomes, that
In addition, not just the patients are involved in this particular is, cognition, memory, communication and the ability to per-
and difficult disease. The burden experienced by caregivers is form daily activities. However, it is important to remark that the
also substantial, affecting their physical and mental wellbeing.2,3 above drugs provide symptomatic treatment for the disease,
Despite its societal impact, AD is underdiagnosed, with relatively without strong evidence for disease-modifying properties.4 In
some countries, tacrine is also available for AD treatment.
a
However, it is a substance with a questionable safety profile
and, in fact, it is not involved in general AD therapy.
´
´
Facultad de Quımica, Bioquımica y Farmacia, Universidad Nacional de San Luis,
´
´
Instituto Multidisciplinario de Investigaciones Biologicas (IMIBIO-SL), Ejercito de
los Andes 950, 5700 San Luis, Argentina. E-mail: denriz@unsl.edu.ar
The principal role of acetylcholinesterase (AChE) is the
termination of impulse transmission at cholinergic synapses
by the rapid hydrolysis of acetylcholine (ACh);5 and one of the
earliest and major neurobiological findings in AD is a deficit in
cholinergic neurotransmission in the basal forebrain. Thus, the
reduction of cortical and CSF cholinergic markers, such as
AChE, choline acetyl transferase and ACh itself, is correlated
with both the extent of the neuropathology and the severity of
the cognitive impairment in Alzheimer’s disease and other
degenerative diseases.6,7 Inhibitors of AChE are the principal
b Laboratorio de Estructura Molecular y Propiedades, IQUIBA-NEA, Universidad
Nacional del Nordeste, CONICET, FaCENA, Av. Libertad 5470, Corrientes 3400,
Argentina
c
´
´
´
Laboratorio de Peptidos Bioactivos, Departamento de Quımica Organica, Facultad
´
´
de Bioquımica y Ciencias Biologicas, Universidad Nacional del Litoral, Santa Fe,
Argentina
d
e
´
Instituto de Investigacion Sanitaria INCLIVA, 46010 Valencia, Spain
´
Departamento de Farmacologıa, Facultad de Farmacia, Universidad de Valencia,
46100 Burjassot, Valencia, Spain
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d1nj00345c
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8321

View Article Online
Paper
NJC
On the other hand, we have recently reported new AChE
inhibitors, and in such articles we have conducted molecular
modelling studies that have allowed us to understand in
some detail the molecular interactions that are involved in
the stabilization of different inhibitor–enzyme complexes
for this molecular target.9–13 In particular, for N-benzyl-2-
phenylethanamines and carbamate-type ligands we have
described the importance of interactions with residues W84,
Y121, G118, S122, E199, S200, W279, F330, Y334 and H440 for
the stabilization of cholinesterase complexes. Moreover, we
carried out a comparative study of AChE and BChE, based on
the accessible area within their active sites, which showed that
the active site of BChE is larger and more flexible than AChE.
Taking advantage of having this information, we asked ourselves
if we would be able to design a new structure with inhibitory
effects on AChE, taking as a starting structure those previously
reported in ref. 8. Thus, we report here the synthesis and
biological evaluation of a new series of substituted pyrrolo[2,1-
a]isoquinolin-3-one derivatives, some of them with strong inhi-
bitory activity against both AChE and BChE enzymes.
Fig. 1 Structural features of substituted pyrrolo[2,1-a]isoquinolin-3-one
derivatives and the FDA-approved cholinesterase inhibitor rivastigmine.
therapeutical option for AD; in fact, donepezil, rivastigmine
and galantamine are the three inhibitors that are FDA-approved
for first-line treatment for AD. Although they share the same
mode of action, they differ in terms of their pharmacologic
characteristics and route of administration, which can affect
their safety and tolerability profile. Rivastigmine, available in
both oral and transdermal patch formulations, is a slowly
reversible dual inhibitor of acetyl and butyryl cholinesterase,
selective for the G1 isoform of acetylcholinesterase, without
hepatic metabolism by the CYP-450 system. Considering the
very limited number of drugs currently available for therapeutic
use in the treatment of AD, it is evident that there is a marked
need to find new structures with similar properties to rivastig-
mine that can be used therapeutically. Our study is focused
mainly on such an objective.
2. Results
2.1 Synthesis
The synthesis of compounds 1–11 has been previously reported.8
Four new carbamates 12–15 were synthesized by reaction
between the corresponding hydroxy-pyrroloisoquinolone and
the appropriate phenyl or ethyl isocyanate. 8-Phenylcarbamate-
1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]isoquinolin-3-one (12) was
obtained from 8-hydroxy-pyrroloisoquinolinone (3), and 8-
chloro-9-phenylcarbamate-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]
isoquinolin-3-one (13) was prepared from 8-chloro-9-hydroxy-
pyrroloisoquinolone (7). Whereas, both 8,9-bis(ethylcarbamate)- (14)
Our research group has previously reported the synthesis of
8-substituted and 8,9-disubstituted pyrrolo[2,1-a]isoquinolin-3-
one derivatives (compounds 1–11, Fig. 1) whose general struc-
tures have a slight structural resemblance with rivastigmine.8
Scheme 1 Synthesis of 8-substituted and 8,9-disubstituted pyrrolo[2,1-a]isoquinolin-3-one derivatives. Reagents and conditions: (a) ethyl succinyl
chloride, NaOH 5%, CH2Cl2, rt, overnight; (b) POCl3, CH3CN, N2, reflux, 4 h; (c) NaBH4; MeOH, rt, 2 h; (d) conc. HCl–EtOH (1 : 1) for 2, reflux, 3 h, or BBr3,
for 6 and 10, CH2Cl2, rt, 2 h; (e) ethyl or phenyl isocyanate, acetone, reflux, 3 h.
8322
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
and 8,9-bis(phenylcarbamate)-1,2,3,5,6,10b-hexahydropyrrolo and 4 and 8 were not clear enough. In fact, at this stage of our
[2,1-a]isoquinolin-3-ones (15) were prepared from (ꢀ)-trolline study, we were not clear on the mechanism of action of the
(11) (Scheme 1).
active compounds 7 and 8. Thus, in order to better understand
the behaviour of these new inhibitors at a molecular level, in
our next step we conducted a molecular modelling study in
which we paid particular attention to the molecular interac-
tions that would stabilize the molecular complexes.
2.2 Searching for new inhibitors of AChE
Previously, the research group of the University of Valencia
reported the synthesis of compounds 1–11 (Fig. 1).8 Some of
these compounds seemed interesting to us from a structural
point of view since they resemble rivastigmine in some way
(particularly compound 8). Therefore, our first step in this
study was to evaluate the inhibitory activity against AChE and
BChE of these eleven compounds. These results are shown in
Fig. S1–S3 (ESI†) and Table 1. As can be seen, none of the
mono-substituted derivatives (compounds 1–4) showed any
inhibitory activity. Regarding the rest of the tested compounds,
compounds 5, 6, 9–11 did not show any significant inhibitory
effect. However, compounds 7 and 8 showed a strong inhibitory
activity against both enzymes, AChE and BChE. This first result
was quite encouraging considering that compounds 7 and 8
displayed a strong inhibitory activity. It should be noted that
both compounds displayed a stronger activity against AChE
than the reference compound (rivastigmine).
2.3 Molecular modelling
To understand more deeply our experimental results, we car-
ried out a molecular modelling study. We conducted this study
in four steps using different modelling techniques. In the first
step, we conducted a docking study; in the second one, we
performed simulations using molecular dynamics calculations.
With these data we performed a per residue analysis and in the
last step quantum mechanics calculations were made to eval-
uate in detail the molecular interactions that stabilize the
different ligand–receptor complexes. The Torpedo californica
AChE (TcAChE) gorge is a narrow, 20 Å long pocket composed
of two binding sites for the cationic substrate acetylcholine,
termed the catalytic anionic site (CAS) and the peripheral
anionic site (PAS), lying at the base and entrance of the gorge,
respectively.
The lack of activity of compounds 2, 6 and 10 was not at all
surprising since in both cases the presence of the benzyloxy or
methoxy group in the replacement of OH has given, in general,
less active compounds when these molecules and other struc-
turally related ones have been tested against other molecular
targets.14–16 However, for this particular series of compounds,
the presence of a hydroxyl group might not be sufficient for the
ligands to develop inhibitory effects. In this regard, a co-
substitution of a Cl atom together with another polar moiety
could be crucial since 7 and 8 displayed the lowest IC50 values.
In the case of compounds 1, 5 and 9, a different behaviour is
expected since these compounds are linear and such great
molecular flexibility can justify their different activities. How-
ever, the dissimilar activities observed for compounds 3 and 7
Between the two sites, the gorge narrows due to a constric-
tion formed by two aromatic residues, F330 and Y121 (bottle-
neck region). Both CAS and PAS are lined mostly by aromatic
residues that interact with the enzyme substrate ACh. When
ACh or an analogue substrate like acetylthiocholine (ATCh)
binds to the PAS, it forms interactions with residues Y70, Y121,
W279, F330, F331 and Y334. A characteristic cation–p inter-
action between a quaternary amine and W279 as well as an H-
bond between the acetyl moiety and Y121 are formed (PDB
codes 2C4H, and 2C58). On the other hand, when ACh/ATCh
binds to the CAS, a reaction occurs and the acetyl group
covalently bound to catalytic serine interacts with residues
from the ‘‘esteratic subsite’’ formed by:
(a) the catalytic triad E327, H440 and S200, (b) the backbone
of residues A201, G118 and G119 that together form the
‘‘oxyanion hole’’ and (c) the aromatic residues W233, F288,
F290 that hold the methyl group in place (PDB codes 2ACE, and
2C58). At the same time, the positively charged choline moiety
interacts with CAS residues W84, Y130, E199, F330, F331, and
Y442 that form the ‘‘anionic subsite’’. As in the PAS, the choline
quaternary amine forms a cation–p interaction with tryptophan
W84. Our docking study indicates that compounds 7 and 8 are
bound at the same site to that previously reported for rivastig-
mine (Fig. 2).12 However, some differences might be appre-
Table 1 Enzyme inhibition of substituted pyrrolo[2,1-a]isoquinolin-3-one
derivatives against acetyl- and butyryl-cholinesterase. All values were
expressed as IC50 ꢀ SD, with a confidence of 95%. The IC50 values were
determined by regression analyses of three replicate determinations
IC50 (mM)
Compound
AChE
BChE
SIb
1
2
3
4
5
6
450
450
450
450
450
450
450
450
450
450
450
450
—
—
—
—
—
—
ciated between compounds
7 and 8. While the spatial
arrangement observed for compound 8 is very similar to that
previously reported for rivastigmine,12 compound 7 is located a
little bit shifted to the bottom of the gorge (Fig. 2). In contrast,
compounds 1, 5 and 9 led to a high number of possible
conformations, which might be related to their lack of activity.
In turn, molecular dynamics calculations allowed us to
perform a per residue analysis for these compounds. As we
expected, the histogram obtained for compound 8 is closely
7
8
9
10
11
Riva
19.25 ꢀ 2.88
11.03 ꢀ 0.34
450
450
450
4.26 ꢀ 0.78
1.74 ꢀ 0.31
450
450
450
4.52
6.34
—
—
—
34.18 ꢀ 6.00
0.08 ꢀ 0.00
427.25
a
b
Rivastigmine (Riv) was used as a positive control. SI (selectivity
index) = IC50 (AChE)/IC50 (BChE).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8323

View Article Online
Paper
NJC
Fig. 2 Spatial view of AChE–4 (cyan), AChE–7 (magenta) and AChE–8
(green) complexes overimposed for comparison.
related to that of rivastigmine (Fig. 3A and B). The main
interactions of both compounds are those with D72, W84,
G118, Y121, S200, W279, F290, F330, Y334 and H440. It should
be noted that the same interactions are found for compound 7.
Nevertheless, interactions obtained for this compound are
weaker in comparison with that observed for compound 8
(compare Fig. 3A and C). Note that compound 8 displays a
stronger inhibitory activity with respect to compound 7
(Table 1); therefore, these results are in line with our experi-
mental results. In contrast to these results, compound 4 shows
a different behaviour from that observed for 8. At least two
aspects show clear differences between these two compounds.
One is the lack of interaction of compound 4 with S200. This
can be clearly seen in the histogram obtained for this com-
pound (Fig. 3D). It is well known that S200 plays a key role in
the inhibition of AChE.17,18 The other important difference is
that 4 adopts a completely different spatial arrangement than 7
and 8. It should be noted that compound 4 orients the pyrrole
ring towards the bottom of the active site and the chain with
the carbamate substituent towards the surface, in exactly the
opposite form of how compound 8 is located. This can be
appreciated clearly in Fig. 2.
At this stage of our study, we focused our efforts on design-
ing and obtaining new compounds structurally related to
compound 8 that possess inhibitory activity against AChE.
2.4 Design of new inhibitors of AChE and BChE
Fig. 3 Histograms of interaction energies partitioned with respect to the
AChE amino acid sequence when complexed with compound 8 (A),
rivastigmine (B), compound 7 (C) and compound 4 (D). The X-axis denotes
the residue number of AChE and the Y-axis denotes the interaction energy
between the compounds and a specific residue. Negative values and
positive values are favourable or unfavourable to binding, respectively.
From Fig. 2 it is possible to observe that there is some room to
incorporate a bulkier substituent at the carbamate group. Thus,
we design four new ligands (compounds 12–15). In compounds
12 and 13 we replace the ethyl group of 4 and 8 by a phenyl
group, respectively, whereas compounds 14 and 15 are the
diethyl and biphenyl carbamates of compound 11. Next, we
performed MD simulations for these new compounds (com-
pounds 12–15) (Fig. 4) and the results were compared with
8324
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
Table 2 Enzyme inhibition of substituted pyrrolo[2,1-a]isoquinolin-3-one
derivatives against acetyl- and butyryl-cholinesterase. All values were
expressed as IC50 ꢀ SD, with a confidence of 95%. The IC50 values were
determined by regression analyses of three replicate determinations
IC50 (mM)
Compound
AChE
BChE
SIb
12
13
14
450
450
—
0.76
—
—
427.25
11.21 ꢀ 1.51
14.68 ꢀ 2.00
1.84 ꢀ 0.09
450
0.08 ꢀ 0.00
450
15
450
34.18 ꢀ 6.00
Riva
a
b
Rivastigmine (Riv) was used as a positive control. SI (selectivity
index) = IC50 (AChE)/IC50 (BChE).
steric factors since compound 14 would be too bulky to enter
the active site. Similar results were obtained for compound 12
(see Fig. 4B and Fig. S4, ESI†). The same results but even more
noticeably were observed for compound 15 which is a logical
result considering that the steric effects in compound 15 are
stronger than those in compound 14.
From these results, we can expect that 13 displays inhibitory
effects while compounds 12, 14 and 15 do not display or at least
display a weaker activity than 13. Thus, we decided to synthe-
size compounds 12–15 (see details of the synthesis in
Section 2.1).
In the next step, we evaluated the inhibitory activity on both
enzymes and these results are shown in Table 2. It might be
seen that our experimental results are in complete agreement
with the simulations, corroborating the results obtained by
molecular modelling. Note that while compound 13 is one of
the most active of this series, 12, 14 and 15 showed no activity
against AChE. A somewhat surprising result was that obtained
for compound 14 against BChE, since although it did not show
any activity on AChE, it showed a strong inhibitory activity
against BChE (Table 2). To explain these results better, we
performed molecular simulations for compounds 13 and 14
interacting at the active site of BChE. These results are shown
in Fig. 5.
It is interesting to note that the histograms obtained for
compounds 13 and 14 interacting with BChE are very similar
(Fig. 5A and B) and they are closely related with that obtained
for rivastigmine as well (Fig. 5C). These results indicate that in
the case of BChE, compound 14 is able to enter into the bottom
of the binding pocket and, therefore, it can properly interact at
the CAS. Such a situation might be better appreciated in Fig. 6,
which shows the spatial orderings adopted by 14 at the binding
sites of both enzymes (AChE and BChE). These results might
explain, at least in part, the selectivity for BChE observed in
compound 14.
Fig. 4 Histograms of interaction energies partitioned with respect to the
AChE amino acid sequence when complexed with compound 13 (A),
compound 12 (B) and compound 14 (C). The X-axis denotes the residue
number of AChE and the Y-axis denotes the interaction energy between
the compounds and a specific residue. Negative values and positive values
are favourable or unfavourable to binding, respectively.
those obtained for compounds 7 and 8. Whereas the histogram
obtained for compound 13 is closely related to that of com-
pound 8 (compare Fig. 3A and 4A), the histograms obtained for
compounds 12 and 14 (Fig. 4B and C) showed significant
differences. The histogram obtained for compound 14 shows
strong interactions, however the lack of interaction with S200,
which is a key interaction for the inhibitory activity, is very
noticeable. At the same time, those interactions with the
residues from the peripheral anionic site (W279 and Y334)
and from the bottleneck (F330) are clearly increased. This
would be a clear indication that compound 14 could not enter
into the bottom of the pocket but would remain at the surface
of the active site. This different behaviour could be explained by
Finally, considering compound 15, our exploratory docking
results showed that this ligand is too large for both active sites
(AChE and BChE). The inability of compound 15 to fully enter
the binding pocket does not allow the carbamate group to reach
the catalytic triad. This is an expected result since changing two
ethyl groups for two phenyl groups implies a considerable
increase in size.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8325

View Article Online
Paper
NJC
Upon going from A to D in Fig. 7 the ligands are more deeply
buried in the gorge. Compound 14 has two bulky N-ethyl
carbamate groups attached to the tricyclic core that prevent it
from accessing the CAS through the bottleneck. When one of
the bulky carbamate groups in 14 is replaced by chlorine, as in
compound 8, entrance to the CAS through the bottleneck is
allowed. However, compound 8 does not completely enter the
CAS but it remains halfway between both gorge sites.
According to a previous study describing TcAChE substrate
trafficking, the substrate analog ATCh first binds to the PAS
with the acetyl group oriented towards the active site and the
quaternary amine forming a cation–p interaction with the
W279 ring. The observed orientation allows the substrate to
slide straight through the gorge without the necessity of
reorienting before productive interaction at the active site, as
the CH3CO group needs to approach the catalytic serine. In this
sense, binding at the PAS may serve as a ‘‘filter’’ that orients the
ACh molecule before it continues its journey towards the CAS.20
Analogously, in complexes of compounds 14 and 8 the carba-
mate moiety is also pointing down towards the active site while
the electropositive amide from the g-lactam ring is stacked over
the W279 ring (Fig. 7A and B).
On the other hand, the binding of compound 13 to the CAS
also closely resembles that of the substrate: the carbonyl oxygen
atom points toward the oxyanion hole (i.e. the backbone of
S200 and A201 in Fig. 7) and the activated carbon atom is
properly positioned for nucleophilic attack by S200 (PDB code
2ACE). Furthermore, the electron-deficient g-lactam ring of
compound 13 is stacked on top of W84 in the same way as
the substrate quaternary amine is also stacked over the same
ring through a cation–p interaction.
One of the most salient characteristics of these carbamate
derivatives is the overall curved shape provided by the tricyclic
core, which allows a good fit of these compounds within the
CAS. Potent carbamate inhibitor physostigmine also has a
similar curved tricyclic core and its binding mode resembles
that of compound 13 where the curved ring of the molecule is
lying on top of the W84 indole ring.21 Therefore, agreement
between our structural models and previous structural evidence
of substrate analogue binding to both the CAS and the PAS
enable us to be confident in the reliability of our computational
predictions, at least from a qualitative point of view.
Fig. 5 Histograms of interaction energies partitioned with respect to the
BChE amino acid sequence when complexed with compound 13 (A),
compound 14 (B) and rivastigmine (C). The X-axis denotes the residue
number of BChE and the Y-axis denotes the interaction energy between
the compounds and a specific residue. Negative values and positive values
are favourable or unfavourable to binding, respectively.
2.5.1 Charge density analysis of molecular interactions.
Regarding the stability of the structural models, the stacked
bars in Fig. 8 provide a dissected view of the anchoring of
carbamate compounds to the different sites within the
TcAChE gorge.
As indicated by the stacked bars in Fig. 8, compound 14
binds almost completely to the PAS, compounds 7 and 13 are
mostly anchored to the CAS, whereas compound 8 is in between
both sites. An unexpected result revealed by the charge density
data in Fig. 8 is that anchoring to the PAS is actually stronger
than to the CAS, as indicated by the contribution of the
PAS residues to the overall anchoring of compounds 14 and 8
(i.e. yellow stacked bars). Nevertheless, the activity of these
carbamates seems to be conditioned to their ability to reach the
2.5 Molecular interactions analysis from QTAIM calculations
In a recent article, we demonstrated that QTAIM calculations
are a very useful tool to evaluate in an accurate and detailed way
the molecular interactions that stabilize ligand–receptor com-
plexes possessing different degrees of structural complexity. In
fact, one of the systems reported in this paper is AChE.19 On the
basis of those results and with the aim of better understanding
the molecular interactions involved in the different complexes
reported here, we carried out a QTAIM study for the most
representative ones. Thus, we selected compounds 7, 8, 13
and 14 complexed with AChE and compounds 13 and 14
complexed with BChE. Fig. 7 shows the binding mode of the
compounds along the AChE enzyme gorge.
8326
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
Fig. 6 Active sites of Torpedo californica AChE (A) and Homo sapiens BChE (B) when complexed with compound 14, which is represented in ball-and-
stick format and coloured in blue. The gorge of each enzyme is depicted by its molecular surface in semi-transparent grey. The main amino acid residues
from both active sites are also shown. Residues from the CAS including the catalytic triad and oxyanionic subsite are depicted in magenta. Amino acids
from the acyl-binding pocket and the anionic subsite are represented in orange and yellow, respectively. Residues from the PAS and the bottleneck
region are depicted cyan and green, respectively.
Known AChE inhibitors like potent donepezil and galantamine
also target both the PAS and CAS simultaneously (PDB entries
4EY7, 1EVE and 1W6R, 4EY6, respectively).
Charge density molecular graphs in Fig. 9 and 10 show in
more detail the molecular interactions that contribute to the
anchoring of compounds 13 and 8, respectively.
Compound 13 mainly interacts with residues from the CAS
at the gorge base and the carbamoyl group is well positioned
within the esteratic subsite for nucleophilic attack by the
reactive serine residue. Fig. 9 shows that the carbamoyl oxygen
atom is hydrogen bonded to the oxyanion hole as evidenced by
the bond paths connecting this atom to backbone H atoms
from A201, G117 and G119, thus enhancing the electrophilicity
of the carbamoyl carbon atom so that it might then be
eventually attacked by the reactive S200 oxygen atom.
Also, several C–Hꢁ ꢁ ꢁp interactions between the N-phenyl
group and ring atoms from aromatic residues W233, F288
and F290 contribute to binding of the carbamoyl moiety at
the esteratic subsite. Moreover, the terminal g-lactam ring from
the tricyclic core is stacked over W84, forming several C–Hꢁ ꢁ ꢁp
interactions between them that together make an important
contribution to the anchoring of compound 8 (see Fig. 7).
On the other hand, as compound 8 has not fully accessed
the CAS the carbamoyl group is too far from the reactive S200
and it is not properly oriented within the oxyanion hole
(Fig. 10). Misplacement of the carbamoyl group is reflected in
the overall weaker binding to the esteratic subsite compared
with compound 13 (see Fig. 7). However, interactions of the
tricyclic core structure with aromatic residues from the PAS
compensate for the anchoring loss from the CAS and so the
Fig. 7 Binding modes of compounds 14 (A), 8 (B), 7 (C) and 13 (D) within
the TcAChE gorge. Residues from the CAS-esteratic, CAS-anionic and PAS
are depicted in blue, red and yellow, respectively. In order to reach the CAS
at the bottom of the gorge, the ligand must slide through the bottleneck
formed by side-chains of residues Y121 and F330.
CAS regardless of its anchoring strength and, so, compound 14
is inactive since it is not able to enter the CAS. Exclusive
binding to the PAS is only observed when high concentrations
of the substrate analogue are present in the crystallization
solution, suggesting that the PAS is occupied only
transiently,20 which is also reflected in the very high IC50 value
reported for the binding of compound 14 at the TcAChE gorge.
On the other hand, compound 8 seems to benefit from
anchoring at both sites simultaneously. It is likely that com-
pound 8 was captured in a transient intermediate state along
the way to the enzyme active site at the bottom of the gorge.
Unlike compound 14, there is no reason why 8 could not reach
the gorge base since it is even less bulky than compound 13
(Fig. 7). However, because of the stronger interactions with the
PAS residues, anchoring at this intermediate state (i.e. in
between both sites) is stronger than binding at the bottom of
the gorge (compare the stacked bars for 8 and 13 in Fig. 8).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8327

View Article Online
Paper
NJC
Fig. 8 Sum of charge density values (in atomic units) at the intermolecular bond critical points between TcAChE and the selected compounds. The total
height of stacked bars indicates the overall anchoring strength of the compounds within the gorge which can be decomposed into anchoring to the CAS
esteratic subsite (CAS-esteratic), CAS anionic subsite (CAS-anionic) and PAS site.
Fig. 9 Charge density molecular graph of compound 13 (magenta) at the gorge of TcAChE. Residues from the CAS-esteratic, CAS-anionic and PAS are
depicted in blue, red and yellow, respectively. Topological elements of the charge density associated with the intermolecular interactions are depicted
with yellow lines (Bond Paths, BPs) and small red spheres (Bond Critical Points, BCPs).
overall binding of compound 8 ends up being stronger than for Fig. 11 shows in more detail the interactions of compound 14 at
compound 13.
the BChE gorge.
Moreover, although the overall structures of BChE and AChE
As can be seen in Fig. 11, compound 14 is certainly anchored
are similar, their interactions with the same ligands are gen- at the bottom of the gorge between aromatic residues W82 and
erally different due to slight differences in their residue W231 that make up the walls of the CAS at both ends of its
sequences. Bottleneck aromatic residues F330 and Y121 that longer (longitudinal) axis. However, unlike rivastigmine that
separate the PAS from the CAS in AChE are replaced in BChE by bind its carbamate moiety to the esteratic subsite, our model-
the less bulky A328 and Q119, respectively. As a result, there is ling results suggest that compound 14 binds in the opposite
no longer such a bottleneck between the PAS and CAS in BChE, way, namely with the carbamate moiety placed at the CAS-
such that ligands that could not access the CAS in AChE due to anionic subsite. The fact the compound 14 can bind to the CAS
their size might now be able to do so in BChE. This seems to be of BChE, albeit not in the expected orientation, correlates well
the case for bulky compound 14, as indicated by the interaction with the inhibition data, namely that it is able to inhibit BChE
profile in Fig. 5B which resembles that of rivastigmine (Fig. 5C). but with much less potency than rivastigmine.
8328
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
with W279. In this arrangement, Y121 partially covers the CAS
entrance as can be seen in Fig. 11. In BChE, both residues W279
and Y121 are replaced by A277 and Q119 respectively, so that
they are no longer able to interact with each other and conse-
quently Q119 is displaced away from the CAS entrance. Because
of these substitutions (as well as the substitution of F330 by
A328), larger ligands can access the bottom of the gorge in
HsBChE, as compared with TcAChE.
Besides the role of the specialized structure formed by
residues W279 and Y121 as a ligand size-based filter, these
residues at the PAS of TcAChE are also involved in ligand pre-
orientation by accommodating the hydrolysable tail (i.e. the
carbamate moiety) in such a way that it can slide straight
through the CAS-esteratic subsite, as already discussed pre-
viously. Thus, the lack of a pre-orientational filter at the PAS of
BChE might be allowing the inverted binding mode of com-
pound 14 at the CAS of the enzyme.
Fig. 10 Charge density molecular graph of compound 8 (magenta) at the
gorge of TcAChE. Residues from the CAS-esteratic, CAS-anionic and PAS
are depicted in blue, red and yellow, respectively. Topological elements of
the charge density associated with the intermolecular interactions are
depicted with yellow lines (Bond Paths, BPs) and small red spheres (Bond
Critical Points, BCPs).
Fig. 12 shows the interactions of compound 13 as it binds to
the HsBChE gorge. The binding mode of compound 13 to
TcAChE is also shown in grey for comparison. As can be seen
in Fig. 12, compound 13 is anchored at the CAS of both TcAChE
and HsBChE; however, it is able to reach the bottom of the
gorge base only in the first case. This difference is likely due to
the substitution of residue V400 in TcAChE (in grey in Fig. 12)
To understand the origin of the binding mode differences of by the bulkier F398 in HsBChE, which prevents the N-phenyl
the compounds to BChE and AChE, Fig. 11 and 12 shows key carbamate moiety of compound 13 from fully accessing the
active site residues from TcAChE that are mutated in HsBChE. CAS-esteratic subsite. As a consequence, the hydrolysable car-
In particular, Fig. 11 shows residues W279 and Y121 from the bamate moiety of compound 13 is not properly placed for
TcAChE structure (in grey) that have been previously super- nucleophilic attack by the reactive serine in HsBChE, as it does
posed with the HsBChE structure. These residues form a in TcAChE. These modelling results are also in line with the
specialized structure at the PAS of TcAChE in which Y121 is inhibition data, which show that compound 13 is a slightly
held firmly on top of the CAS by T-shaped C–Hꢁ ꢁ ꢁp interactions more potent inhibitor of TcAChE than HsBChE.
Fig. 11 Charge density molecular graph of compound 14 (magenta) at the gorge of HsBChE. Residues from the CAS-esteratic, CAS-anionic and PAS are
depicted in blue, red and yellow, respectively. Residues in grey are not from BChE structure but from a superimposed TcAChE structure. Topological
elements of the charge density associated with the intermolecular interactions are depicted with yellow lines (Bond Paths, BPs) and small red spheres
(Bond Critical Points, BCPs).
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8329

View Article Online
Paper
NJC
Fig. 12 Charge density molecular graph of compound 13 (magenta) at the gorge of HsBChE. Residues from the CAS-esteratic, CAS-anionic and PAS are
depicted in blue, red and yellow, respectively. The binding mode of compound 13 to TcAChE as well as residue V400 from TcAChE have been
superimposed for comparison (in grey). Topological elements of the charge density associated with the intermolecular interactions are depicted with
yellow lines (Bond Paths, BPs) and small red spheres (Bond Critical Points, BCPs).
Although the results obtained from molecular modelling were purified by silica gel 60 (40–63 mm, Merck) column
simulations represent additional support for the possible effect chromatography. The quoted yields are of the purified material.
on AChE and BChE, further studies, as for example a kinetic
3.2 Synthesis
analysis, could give a better and more complete understanding
about the inhibitory effect displayed by these compounds.
3.2.1 Synthesis of 8,9-disubstituted pyrrolo[2,1-a]isoquino-
linone derivatives. The general synthetic path for compounds
1–11 has been previously reported. 8-Substituted (12) and 8,9-
disubstituted (13–15) pyrrolo[2,1-a]isoquinolinone carbamate
derivatives were prepared according to ref. 8.
3. Methodology
3.1 General experimental procedures and instrumentation
3.2.2 General procedure for the synthesis of carbamate
derivatives (8, 12–15)
All solvents and reagents were purchased from Scharlab S.L.
(Barcelona, Spain) and Sigma-Aldrich (St Louis, MO) commer-
cial sources. Solvents were dried and freshly distilled under a
nitrogen atmosphere if necessary. 1H and 13C NMR spectra
were recorded using CDCl3 as the reference solvent with two
drops of CD3OD if necessary in a Bruker AC-300 or AC-500
spectrometer (Bruker Instruments, Kennewick, WA). Chemical
shifts (d) are reported in ppm relative to an internal deuterated
solvent reference. Multiplicities of the 13C NMR resonances
were assigned by DEPT experiments. The assignments of all
compounds were made by COSY, DEPT, HSQC and HMBC. All
reactions were monitored by analytical thin-layer chromatogra-
phy (TLC), carried out on silica gel 60 F254 plates (Merck
Group, Darmstadt, Germany), and visualized by UV irradiation
at 254 nm. ESIMS analyses were performed using UHPLC
apparatus (Shimadzu, LCMS-8040) coupled to a tandem mass
8-Hydroxy-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]isoquinolin-3-
one (3). 8-Chloro-9-hydroxy-1,2,3,5,6,10b-hexahydropyrrolo
[2,1-a]isoquinolin-3-one (7) or 8,9-dihydroxy-1,2,3,5,6,10b-
hexahydropyrrolo[2,1-a]isoquinolin-3-one (11) (1 equiv.) was
dissolved in dry acetone (10 mL). Subsequently, the appropriate
ethyl or phenyl isocyanate (1.5 equiv. or 3 equiv. for 12, 13 or 14,
15 respectively) was added. The reaction mixture was heated
under reflux for 3 h and then concentrated to dryness. The
residue was redissolved in 10 mL of CH2Cl2 and washed with
H2O (3 ꢂ 10 mL). The organic layer was dried with anhydrous
Na2SO4, filtered and evaporated under reduced pressure. The
crude product was purified by column chromatography on
silica gel (CH2Cl2/MeOH 97 : 3).
8-Phenylcarbamate-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]iso-
spectrometry (MS/MS) triple quadrupole instrument equipped quinoline-3-one (12). Colourless oil (13 mg, 27% yield). 1H NMR
with an electrospray ionization (ESI) ion source (Shimadzu, (300 MHz, CDCl3): d = 7.44 (2H, d, J = 8.0 Hz, H-20, H-60),
Kyoto, Japan). High-resolution electron ionization mass spec- 7.34 (1H, t, J = 7.4 Hz, H-40), 7.16–7.08 (2H, m, H-30, H-50), 6.97
trometry (HREIMS) was determined by a TripleTOF 5600 LC/ (1H, d, J = 8.4 Hz, H-10), 6.75 (1H, dd, J = 2.6, 8.4 Hz, H-9),
MS/MS system (AB SCIEX, Framingham, MA, U.S.A.). Residues 6.63 (1H, d, J = 2.6 Hz, H-7), 6.20 (1H, brs, NH), 4.78–4.69
8330
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
(1H, m, H-10b), 4.32–4.21 (1H, m, Ha-5), 3.11–2.47 (6H, m, Hb- 0.25 U mLꢃ1 of EeAChE (E.C. 3.1.1.7, type VI-S from the
5, H-6, Ha-1, H-2), 1.92–1.80 (1H, m, Hb-1); 13C NMR (75 MHz, electric eel, Sigma-Aldrich), 0.24 mM of acetylthiocholine
CDCl3): d = 173.1 (CO-3), 154.8 (NHCO), 149.1 (C-8), 137.2 (C-10), (ATCh), 0.2 mM of DTNB and the corresponding compound
135.1 (C-6a), 129.4 (C-10a), 129.2 (CH-40), 126.0 (CH-10), 125.9 (50–3.125 mM). The reaction started after the addition of
(CH-30, CH-50), 119.0 (CH-20, CH-60), 115.3 (CH-7), 114.3 (CH-9), the substrate to a 15 min pre-incubated solution of AChE with
56.5 (CH-10b), 37.0 (CH2-5), 31.8 (CH2-2), 28.6 (CH2-6), 27.7 the compound. Absorbance at 405 nm was measured after
(CH2-1); HREIMS: m/z 322.1374 [M]+ (calcd for C19H18N2O3, 5 min using a 96-well microplate reader (Thermo Fisher
322.1310).
FC Multiskan). The AChE inhibition percentage (I%) was cal-
culated with eqn (1). Each compound was tested in triplicate.
The AChE residual activity percentage (equal to 100 ꢃ I%)
values were used to calculate the required concentration to
inhibit 50% of enzyme activity (IC50) with a four-parameter
logistic curve.
8-Chloro-9-phenylcarbamate-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]
isoquinoline-3-one (13). Colourless oil (7.5 mg, 25% yield). 1H NMR
(500 MHz, CDCl3): d = 7.46 (2H, d, J = 8.0 Hz, H-20, H-60), 7.35 (1H,
t, J = 7.4 Hz, H-40), 7.26 (1H, s, H-7), 7.17–7.12 (2H, m, H-30, H-50),
7.03 (1H, s, H-10), 4.75 (1H, t, J = 7.8 Hz, H-10b), 4.30 (1H, ddd,
J = 2.6, 6.2, 12.9 Hz, Ha-5), 3.08–3.01 (1H, m, Hb-5), 2.95–2.87
(1H, m, Ha-6), 2.80–2.75 (1H, m, Hb-6), 2.67–2.62 (1H, m, Ha-1),
2.61–2.55 (1H, m, Ha-2), 2.51–2.45 (1H, m, Hb-2), 1.90–1.86 (1H, m,
Hb-1); 13C NMR (125 MHz, CDCl3): d = 173.1 (CO-3), 150.5 (NHCO),
145.3 (C-9), 137.3 (C-10a), 137.0 (C-10), 132.8 (C-6a), 130.6 (CH-7),
129.2 (CH-40), 125.7 (CH-30, CH-50), 124.2 (C-8), 120.4 (CH-10),
118.7 (CH-20, CH-60), 56.3 (CH-10b), 36.7 (CH2-5), 31.6 (CH2-2),
27.9 (CH2-6), 27.2 (CH2-1); HREIMS: m/z 357.1011 [M]+ (calcd for
C19H17ClN2O3, 357.1000).
% I = (AI ꢃ Ab) ꢂ 100/(A0 ꢃ Ab)
(1)
A0: control (no inhibitor); Ab: blank (no enzyme, no inhibitor);
AI: sample.
The same procedure was applied for the BChE inhibitory
assay, with modifications in the enzyme and substrate used,
which were BChE from human serum and S-butyrylthiocholine
iodide (BTCI; 7 mM), respectively.
In both cases, rivastigmine was used as a positive control.
3.4 Molecular modelling
8,9-Bis(ethylcarbamate)-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]
isoquinoline-3-one (14). Colourless oil (12 mg, 36% yield).
1H NMR (300 MHz, CDCl3 + 2 drops CD3OD): d = 6.97 (1H, s,
H-10), 6.95 (1H, s, H-7), 5.21–5.17 (2H, m, NH), 4.71 (1H, t, J =
6.7 Hz, H-10b), 4.25 (1H, ddd, J = 12.7, 6.0, 2.8 Hz, Ha-5), 3.34–
3.24 (4H, m, CH2NH), 3.03 (1H, td, J = 1.0, 4.2 Hz, Hb-5), 2.95–
2.80 (1H, m, Ha-6), 2.78–2.67 (1H, m, Hb-6), 2.65–2.40 (3H, m,
Ha-1, H-2), 1.95–1.80 (1H, m, Hb-1), 1.19 (3H, t, J = 7.0 Hz,
CH3CH2NH); 13C NMR (75 MHz, CDCl3 + 2 drops CD3OD): d =
173.7 (CO-3), 157.8 (NHCO), 141.7 (C-9), 141.5 (C-8), 135.1
(C-10a), 131.5 (C-6a), 123.6 (CH-10), 119.7 (CH-7), 56.5 (CH-
10b), 36.8 (CH2-5), 35.9 (CH2NH), 31.5 (CH2-2), 27.8 (CH2-6),
27.0 (CH2-1), 14.7 (CH3CH2NH); HREIMS: m/z 362.1714 [M]+
(calcd for C18H23N3O5, 362.1710).
Calculations were performed after synthesizing previously
reported compounds 1–11. Conversely, all simulations regard-
ing new derivatives 12–15 were conducted before their
synthesis.
3.4.1 Molecular docking. The following X-ray structures
available in the Protein Data Bank (http://www.rcsb.org) were
used for molecular modelling studies: Torpedo californica acetyl-
cholinesterase (TcAChE; PDB code: 1DX6)24 and Homo sapiens
butyrylcholinesterase (HsBChE; PDB code: 6EUL).25 Water and
ligand molecules were removed from both PDB structures
before calculations. Compounds 1–15 and receptor structures
were converted from PDB to PDBQT format using AutoDock-
Tools 1.5.4.26 Gasteiger charges were added for all the com-
pounds whereas non-polar hydrogen atoms were merged.
The docking simulations were performed using AutoDock 4.2
software.26 In all simulations, the receptor structure was set as
rigid while all the torsions of the ligand were allowed to rotate
during docking. 3D affinity maps were defined as a cubic box
centred at the active site of each molecular target. Grid dimen-
sions (XYZ) were 60 ꢂ 60 ꢂ 60 points separated by 0.375 Å. The
maximum number of energy evaluations was set to 2.5 ꢂ 106,
and the maximum number of generations was 2.7 ꢂ 104
whereas 200 poses were collected. Other parameters were set
to default values. The collected conformations were clustered
into different families, taking into account the root-mean
square deviation (RMSD) of the backbone, and were then
ranked according to the binding free energy. The structure
from the most populated cluster with the lower relative free
energy was chosen for further calculations.
8,9-Bis(phenylcarbamate)-1,2,3,5,6,10b-hexahydropyrrolo[2,1-a]
isoquinoline-3-one (15). Colourless oil (15 mg, 41% yield).
1H NMR (300 MHz, CDCl3): d = 7.41–7.03 (10H, m, 2xH-20,
2xH-30, 2xH-40, 2xH-50, 2xH-60), 7.12–7.06 (1H, s, H-10),
7.03 (1H, s, H-7), 4.79–4.66 (1H, m, H-10b), 4.27–4.15 (1H, m,
Ha-5), 3.07–2.82 (2H, m, Hb-5, Ha-6), 2.81–2.69 (1H, m, Hb-6),
2.68–2.41 (3H, m, Ha-1, H-2), 1.98–1.76 (1H, m, Hb-1); 13C NMR
(75 MHz, CDCl3): d = 173.3 (CO-3), 150.8 (NHCO), 141.2, 140.9
(C-8, C-9), 136.4 (C-10), 136.0 (C-6a), 132.3 (C-10a), 129.2 (CH-40),
124.5 (CH-10), 124.0 (CH-30, CH-50), 120.1 (CH-7), 118.7
(CH-20, CH-60), 56.4 (CH-10b), 36.8 (CH2-5), 31.6 (CH2-2), 28.1
(CH2-6), 27.1 (CH2-1); HREIMS: m/z 475.1977 [M + H2O]+ (calcd
for C26H23N3O5, 475.1976).
3.3 AChE and BChE inhibitory activity
End-point inhibitory activity was tested according to a pre-
3.4.2 Molecular dynamics (MD). MD simulations of all
viously reported method.22 For each sample, an Ellman’s complexes and later analysis of their trajectories were per-
reaction23 was performed in a final volume of 200 mL with formed with the Amber16 package.27 Antechamber software
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8331

View Article Online
Paper
NJC
in the AmberTools package28 was used to generate the para- of the Multiwfn software.39 This type of calculation has been
meters for MD simulations with ff99SB29 and GAFF30 force used in recent works since it ensures a reasonable compromise
fields. The best ranked conformation of each complex was then between the wave function quality required to obtain reliable
soaked in truncated octahedral periodic boxes of explicit water values of the derivatives of r(r) and the computer power
boxes using the TIP3P model.31 Na+ or Clꢃ ions were placed available.12,19,40–45
using the Leap module to neutralize the negative and positive
charges of the AChE and BChE complexes, respectively. The
energy of each system was minimized with the sander module
using a steepest-descent algorithm for 1000 steps. Subse-
4. Conclusions
quently, the complexes were equilibrated during 500 ps at We report here a new series of inhibitors of AChE and BChE.
constant volume. The SHAKE algorithm32 was applied allowing Some of the substituted pyrrolo[2,1-a]isoquinolinone deriva-
for an integration time step of 2 fs. The systems were heated up tives reported here exhibit a significant inhibitory effect against
to 300 K, a Langevin thermostat33 was applied and the collision both the AChE and BChE enzymes. Among the series reported
frequency was 1.0 psꢃ1. After minimization and heating, three here, compounds 7, 8, 13 and 14 were those that displayed
MD production simulations of 20 ns were carried out at a 298 K more strong inhibitory activities; most of them possessed
target temperature. Thus, in total 60 ns were simulated for each inhibitory effects even stronger than rivastigmine (the com-
complex. Coordinates were saved for analysis every 10 ps. All pound used as the reference). It should be noted that com-
productions were performed under NVT conditions and the pound 14 was the only one showing a selective inhibitory
particle mesh Ewald method (PME)34 was applied using a grid activity against BChE. The molecular modelling study showed
spacing of 1.2 Å, a spline interpolation order of 4 and a real that the larger size of the active site of BChE with respect to
space direct sum cutoff of 10 Å. Post MD analysis was per- AChE could explain the selective activity of compound 14.
formed with the program CPPTRAJ.35
QTAIM analysis gave us detailed information about the mole-
3.4.3 MM-PBSA free energy calculation and per-residue cular interactions stabilizing the different ligand–enzyme com-
analysis. MMPBSA calculations were conducted to calculate plexes. Such information could be useful for the study and
the relative binding energies of the complexes.36 Snapshots design of new inhibitors of both enzymes.
from the corresponding last 1000 ps of the MD trajectories were
considered. The explicit water molecules and counter ions were
removed from the snapshots.
Author contributions
Afterwards, a per-residue decomposition of the relative free
energy was performed using the mm_pbsa program in the O. P. and R. D. E. contributed to the conceptualization of this
AMBER package. This analysis led us to determine which work. O. P., L. V., N. C. and D. C. carried out the synthesis and
amino acids of the AChE and BChE catalytic sites are involved characterization of the compounds. O. P. and F. G. performed
in the interactions with each ligand.
the molecular modelling studies. O. P., E. A. and R. D. E.
3.4.4 Quantum mechanics calculations and topological conducted the QTAIM calculations. R. S. and A. S. S. carried
analysis of the electron density distribution. MD trajectories out the bioassays studies and their analysis. O. P., E. A., N. C.,
were first clustered based on the root mean square deviation D. C. and R. D. E. wrote and corrected the article.
(RMSD) of the complex backbone atoms employing the
CPPTRAJ tool included in the Amber package. Then, the
representative structure from the most populated clusters was
selected for the QTAIM analysis.
Conflicts of interest
Reduced 3D model systems representing the different L–R There are not conflicts to declare.
binding pockets of the selected zones were constructed from
the representative structures obtained by MD simulations by
keeping only receptor residues that interact directly with the
ligands. All the amino acids that are within a radius of 5 Å
Acknowledgements
distance from each ligand atom were considered. Those resi- The manuscript was written through contributions of all
dues that showed significant interactions with the ligand in the authors. The continued funding of the UNSL is strongly appre-
free energy decomposition approach were also included.
ciated. This work has also been partially funded by grants from
The charge density of the reduced model systems was then ANPCyT (PICT 2015-1769). This study was also supported by
computed by DFT methodology with the PBE hybrid functional grants from the Spanish Ministry of Economy and Competi-
and the 6-31G(d) as basis set, as implemented in the Gaussian tiveness (SAF2017-89714-R) and the Carlos III Health Institute
16 package.37 The topological analysis of charge density was (PI18/01450) and the European Regional Development Fund.
performed on the different reduced models to evaluate the NC is an investigator from the ‘‘Miguel Servet’’ programme
interactions between each ligand and the molecular target in (CP15/00150), from Carlos III Institute of Health, co-funded by
the context of the quantum theory of atoms in molecules the European Social Fund. RDE, FG, EA, RS and AS are
(QTAIM).38 These calculations were carried out with the help members of CIC-CONICET-ARGENTINA.
8332
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
22 I. Sanchis, R. Spinelli, N. Aschemacher, M. V. Humpola and
A. Siano, Amino Acids, 2020, 52, 387–396.
23 G. L. Ellman, K. D. Courtney, V. Andres and R. M.
Featherstone, Biochem. Pharmacol., 1961, 7, 88–95.
24 H. M. Greenblatt, G. Kryger, T. Lewis, I. Silman and
J. L. Sussman, FEBS Lett., 1999, 463, 321–326.
25 S. N. Dighe, E. De la Mora, S. Chan, S. Kantham, G. McColl,
J. A. Miles, S. K. Veliyath, B. Y. Sreenivas, Z. D. Nassar,
I. Silman, J. L. Sussman, M. Weik, R. P. McGeary,
M. O. Parat, X. Brazzolotto and B. P. Ross, Commun. Chem.,
2019, 2, 1–14.
26 G. M. Morris, R. Huey, W. Lindstrom, M. F. Sanner,
R. K. Belew, D. S. Goodsell and A. J. Olson, J. Comput. Chem.,
2009, 30, 2785–2791.
27 D. A. Case, R. M. Betz, W. Botello-Smith, D. S. Cerutti,
I. T. E. Cheatham, T. A. Darden, R. E. Duke, T. J. Giese,
H. Gohlke, A. W. Goetz, N. Homeyer, S. Izadi, P. Janowski,
J. Kaus, A. Kovalenko, T. S. Lee, S. LeGrand, P. Li, C. Lin,
T. Luchko, R. Luo, B. Madej, D. M. York, P. A. K. M. York
and P. A. Kollman, Univ. California, San Fr., 2016, 10779586.
28 J. Wang, W. Wang, P. A. Kollman and D. A. Case, J. Mol.
Graphics Modell., 2006, 25, 247–260.
29 W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. Merz,
D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell and
P. A. Kollman, J. Am. Chem. Soc., 1995, 117, 5179–5197.
30 J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman and
D. A. Case, J. Comput. Chem., 2004, 25, 1157–1174.
31 W. L. Jorgensen, J. Chandrasekhar, J. D. Madura,
R. W. Impey and M. L. Klein, J. Chem. Phys., 1983, 79,
926–935.
References
1 M. Prince, A. Comas-Herrera, M. Knapp, M. Guerchet and
M. Karagiannidou, World Alzheimer Report 2016 Improving
healthcare for people living with dementia coverage, QualIty
and costs now and In the future.
2 H. Brodaty and M. Donkin, Dialogues Clin. Neurosci., 2009,
11, 217–228.
3 M. Ma, D. Dorstyn, L. Ward and S. Prentice, Aging Ment.
Heal., 2018, 22, 1395–1405.
4 L. Patel and G. T. Grossberg, Drugs Aging, 2011, 28, 539–546.
5 P. Taylor, in The Pharmacological Basis of Therapeutics, ed.
A. Gilman, A. Nies, T. Rall and P. Taylor, MacMillan, New
York, United States, 5th edn, 1990, pp. 131–150.
6 D. M. Bowen, C. B. Smith, P. White and A. N. Davison, Brain,
1976, 99, 459–496.
7 P. Davies, Brain Res., 1979, 171, 319–327.
8 L. Moreno, J. Parraga, A. Galan, N. Cabedo, J. Primo and
D. Cortes, Bioorg. Med. Chem., 2012, 20, 6589–6597.
9 F. Carmona-Viglianco, D. Zaragoza-Puchol, O. Parravicini,
A. Garro, R. D. Enriz, G. E. Feresin, M. Kurina-Sanz and
A. A. Orden, New J. Chem., 2020, 44, 9466–9476.
´
´
´
10 A. Hudcova, A. Kroutil, R. Kubınova, A. D. Garro,
¨
L. J. Gutierrez, D. Enriz, M. Oravec and J. Csollei, Molecules,
2020, 25, 1751.
11 J. E. Ortiz, A. Garro, N. B. Pigni, M. B. Agu¨ero, G. Roitman,
A. Slanis, R. D. Enriz, G. E. Feresin, J. Bastida and A. Tapia,
Phytomedicine, 2018, 39, 66–74.
12 T. Padrtova, P. Marvanova, K. Odehnalova, R. Kubinova,
O. Parravicini, A. Garro, R. Enriz, O. Humpa, M. Oravec and
P. Mokry, Molecules, 2017, 22, 2048.
13 J. E. Ortiz, N. B. Pigni, S. A. Andujar, G. Roitman,
F. D. Suvire, R. D. Enriz, A. Tapia, J. Bastida and
G. E. Feresin, J. Nat. Prod., 2016, 79, 1241–1248.
32 S. Miyamoto and P. A. Kollman, J. Comput. Chem., 1992, 13,
952–962.
33 J. A. Izaguirre, D. P. Catarello, J. M. Wozniak and R. D. Skeel,
J. Chem. Phys., 2001, 114, 2090.
´
14 J. Parraga, S. A. Andujar, S. Rojas, L. J. Gutierrez, N. El
34 U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee
and L. G. Pedersen, J. Chem. Phys., 1995, 103(19), 8577–8593.
35 D. R. Roe and T. E. Cheatham, J. Chem. Theory Comput.,
2013, 9, 3084–3095.
Aouad, M. J. Sanz, R. D. Enriz, N. Cabedo and D. Cortes, Eur.
J. Med. Chem., 2016, 122, 27–42.
15 O. Parravicini, M. L. Bogado, S. Rojas, E. L. Angelina, S. A.
´
Andujar, L. J. Gutierrez, N. Cabedo, M. J. Sanz, M. P. Lopez-
36 S. Genheden and U. Ryde, Expert Opin. Drug Discovery, 2015,
10, 449–461.
Gresa, D. Cortes and R. D. Enriz, J. Mol. Model., 2017, 23, 273.
16 J. Parraga, N. Cabedo, S. Andujar, L. Piqueras, L. Moreno,
A. Galan, E. Angelina, R. D. Enriz, M. D. Ivorra, M. J. Sanz
and D. Cortes, Eur. J. Med. Chem., 2013, 68, 150–166.
17 P. Bar-On, C. B. Millard, M. Harel, H. Dvir, A. Enz, J. L. Sussman
and I. Silman, Biochemistry, 2002, 41, 3555–3564.
18 M. B. Colovic, D. Z. Krstic, T. D. Lazarevic-Pasti,
A. M. Bondzic and V. M. Vasic, Curr. Neuropharmacol.,
2013, 11(3), 315–335.
37 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato,
A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts,
B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov,
J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini,
F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson,
D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega,
G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell,
J. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark,
J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov,
T. A. Keith, R. Kobayashi, J. Normand, K. Raghavachari,
A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
J. M. Millam, M. Klene, C. Adamo, R. Cammi,
19 S. Rojas, O. Parravicini, M. Vettorazzi, R. Tosso, A. Garro,
´
L. Gutierrez, S. Andu´jar and R. Enriz, Eur. J. Med. Chem.,
2020, 208, 112792.
20 J.-P. Colletier, D. Fournier, H. M. Greenblatt, J. Stojan,
J. L. Sussman, G. Zaccai, I. Silman and M. Weik, EMBO J.,
2006, 25, 2746–2756.
21 D. Barak, A. Ordentlich, D. Stein, Q. S. Yu, N. H. Greig and
A. Shafferman, Biochem. J., 2009, 417, 213–222.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 8321–8334 | 8333

View Article Online
Paper
NJC
J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, 42 L. J. Gutierrez, E. E. Barrera Guisasola, N. Peruchena and
J. B. Foresman and D. J. Fox, Gaussian, Gaussian, Inc.,
Wallingford CT, 2016.
38 R. F. W. Bader, Acc. Chem. Res., 1985, 18, 9–15.
39 T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
R. D. Enriz, Mol. Simul., 2016, 42, 196–207.
43 E. L. Angelina, S. A. Andujar, R. D. Tosso, R. D. Enriz and
N. M. Peruchena, J. Phys. Org. Chem., 2014, 27, 128–134.
44 J. E. Ortiz, A. Garro, N. B. Pigni, M. B. Agu¨ero, G. Roitman,
A. Slanis, R. D. Enriz, G. E. Feresin, J. Bastida and A. Tapia,
Phytomedicine, 2018, 39, 66–74.
´
´
´
40 L. J. Gutierrez, O. Parravicini, E. Sanchez, R. Rodrıguez,
J. Cobo and R. D. Enriz, J. Biomol. Struct. Dyn., 2018, 1–18.
´
41 E. E. Barrera Guisasola, L. J. Gutierrez, R. E. Salcedo, 45 S. A. Andujar, R. D. Tosso, F. D. Suvire, E. Angelina,
F. M. Garibotto, S. A. Andujar, R. D. Enriz and
N. Peruchena, N. Cabedo, D. Cortes and R. D. Enriz,
´
A. M. Rodrıguez, Comput. Theor. Chem., 2016, 1080, 56–65.
J. Chem. Inf. Model., 2012, 52, 99–112.
8334
|
New J. Chem., 2021, 45, 8321–8334
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021