B. Biscussi et al.
Journal of Photochemistry & Photobiology, A: Chemistry 418 (2021) 113375
with a slow thermal reversion. These results probe that these azo-
compounds are efficient photoswitchable molecules with potential
application in therapies based on synaptic communications. Compound
(6ꢀ 31 G (d, p) basis set) incorporated in the Gaussian 09 program
[26–28]. Also the protein from the X-ray crystal structure of Torpedo
californica AChE (PDB code 2ACE [29]) was prepared (i. e. hydrogen
addition, charges calculation and atom types assignment). The program
Autodock version 4.2.5.1 was chosen for the docking research using the
implemented empirical free energy function [30]. The simulations were
prepared, run and analyzed using the graphical interface program
AutoDock Tools. A box including the active site, as well as the peripheral
anionic site of the enzyme was defined as the simulation space. The
auxiliary program Autogrid 4 was used for calculating the atomic
interaction energy on a 0.375 Å grid between probes corresponding to
each inhibitor atom type and the different protein atom types. All di-
hedrals in the compounds were able to rotate freely.
19 (Z) proved to be more effective AChEI (IC50 = 0.43 μM) than its E
isomer, becoming the most active of the series. A molecular modelling
study allowed us to propose that these inhibitors are located at the PAS
of the enzyme. That means that, in addition to inhibit the enzymatic
activity, these compounds can interfere in the β-amyloid pro-
aggregating AChE-induced activity. It is to note that the molecular dy-
namic simulations allowed us to understand why the inhibitor 19 is
more effective in its Z geometry. These studies suggest the importance in
the rational design of the interaction optimization at the enzyme binding
sites and demonstrate that azoderivatives reported here can be consid-
ered as promising candidate compounds for the development of new
multifunctional drugs for the treatment of AD.
The Lamarckian genetic algorithm protocol was chosen for the
docking performance and 256 independent simulations with a 150
members population size were run for each compound. Other parame-
ters were set as default. After docking, the poses obtained were grouped
according to all atom root-mean-square deviation (rmsd). The tolerance
for each group is a RMSD of 2 Å position from the lowest-energy
conformation. Then, a ranking was done with the clusters energy
related to the lowest-energy conformer within each cluster.
4. Experimental section
4.1. Quantitative AChE inhibition assay
AChE from electric eel (500 U) was purchased from Sigma (Buenos
Aires, Argentina). The enzymatic inhibition was determined in vitro
using the Ellman′s spectrophotometric method with minor modifica-
tions [16,19]. Buffer phosphate A (8 mM K2HPO4, 2.3 mM NaH2PO4)
was used to obtain 5 U/mL stock solution of the enzyme. Further dilu-
tion was carried out with buffer phosphate B (8 mM K2HPO4, 2.3 mM
NaH2PO4, 0.15 M NaCl, 0.05 % Tween 20, pH 7.6) to produce 0.3 U/mL
final enzyme solution. Substrate solution (0.6 mM ATCI) containing the
Elman′s reactive (0.5 mM DTNB) was prepared in 0.1 M Na2HPO4 so-
lution (pH 7.5). Samples were dissolved in buffer phosphate B with 1.25
% CHCl3 and 13.75 % of MeOH as a cosolvent mixture. Enzyme solution
4.5. Molecular dynamics simulations
MD simulations were performed starting from the Crystal structure
of Torpedo californica acetylcholinesterase complexed with Xe, solved at
2.3 Å resolution (PDB entry: 3M3D [31]). The xenon atom was removed.
The system was immersed in an octahedral box of TIP3P water mole-
cules [32], the limits of which were at a minimum distance of 10 Å from
the protein. Due to periodic boundary conditions, Ewald sums were used
to treat long-range electrostatic interactions [33] for all systems. To
maintain the covalent bonds between heavy atoms and hydrogen atoms,
the SHAKE algorithm was used to maintain the geometric constraint of
the bond during molecular dynamics simulations [34]. The integration
of Newton′s equations was performed to predict the atomic positions in a
future time step of each atom of the protein, ligands, and water mole-
cules, by the force fields parm99, gaff and TIP3P, respectively, all of
them implemented in AMBER [35,36]. To regulate the temperature and
pressure, the Berendsen thermostat and barostat were used throughout
the simulation. The first step consisted of a short and gradual geometric
optimization of all the atoms in the system in order to avoid that when
starting the simulation, two or more atoms were too far from equilib-
rium position, which can cause undesired collisions. Subsequently, brief
simulations of molecular dynamics were carried out using a 0.1 fs time
step to increase gradually the kinetic energy of the atoms. To do this, the
systems were slowly heated from 100 to 300 K, under constant volume
conditions. The last step of the balancing step, to allow the systems to
reach the proper density, consisted of a brief simulation at a constant
temperature of 300 K using a 0.1 fs time step, under constant pressure of
1 bar. The atomic positions and velocities at the end of this equilibration
step were taken as the starting point for the production of 100 ns MD
simulations at 300 K temperature using a 2 fs time step.
(300
μ
L) and sample solution (300
μ
L) were mixed in a test tube and
incubated at r.t. (60 min). The reaction was started by adding the sub-
strate solution (600
μL). The absorbance was recorded at 405 nm for 120
s at 25 ◦C on a JASCO V-630BIO (Tokyo, Japan) Spectrophotometer
equipped with an EHCS-760 Peltier. Enzymatic activity was calculated
by comparing the reaction rates between the sample and the blank. All
reactions were performed in triplicate. The sample concentration
reflecting 50 % inhibition (IC50) was calculated by nonlinear regression
of the response curve versus log (concentration), using GraphPadPrism
5. Tacrine was used as the reference inhibitor.
4.2. Quantitative AChE inhibition assay (isomer Z)
Inhibitory activity of the Z isomer was determined in the absence of
light. Sample solution in MeOH (30 μL) was irradiated with UV light at
360 nm for 10 min. Then, 270 μL of buffer B and 300 μL of AChE solution
were added and incubated 60 min avoiding the light. The whole process
was performed in the dark to prevent Z → E isomerization. The test was
continued as indicated in the previous point.
4.3. UV–vis spectroscopy
Switching experiments were done with an 8 W mercury arc lamp
with filter of 360 nm from Pleuger Antwerp Brussels. The Z → E isom-
erization was accelerated by illumination with white light bulb of 60 W
to evaluate photoreversion without decomposition. UV ꢀ vis spectros-
copy data were measured with a Varian Cary 60 UV/Vis Spectropho-
tometer (Agilent, USA). Thermal Z → E isomerization was slow enough
allowing the evaluation of all pss separately.
Author statement
Brunella Biscussi was responsible for the data curation, formal
analysis, investigation, methodology, visualization, writing original
draft, review and editing; Maria Alejandra Sequeira for data curation,
formal analysis, investigation, methodology, visualization, writing
original draft; Victoria Richmond for the conceptualization, data cura-
tion, formal analysis, investigation, software, writing original draft; Pau
˜
4.4. Molecular docking determinations
Arroyo Manez for the conceptualization, data curation, funding acqui-
sition, investigation, software, supervision and writing original draft;
Ana Paula Murray for the conceptualization, data curation, formal
analysis, funding acquisition, investigation, methodology, project
administration, resources, supervision, writing original draft, review
In order to prepare the ligands and the protein for the docking study,
the geometry of compounds 17 (E), 19 (Z) and 19 (E) were optimized
with semiempirical calculations (AM1) and the Hartree-Fock method
6
Arroyo Ma?ez, Pau
