Synthesis and biological assessment of novel 2-thiazolylhydrazones and computational analysis of their recognition by monoamine oxidase B

Article abstract of DOI:10.1016/j.ejmech.2011.12.027

Monoamine oxidase B (MAO-B) is a promising target for the treatment of neurodegenerative disorders. We report the synthesis and the biological evaluation of halogenated derivatives of 1-aryliden-2-(4-phenylthiazol-2-yl) hydrazines. The fluorinated series shows interesting activity and great selectivity toward the human recombinant MAO-B isoform expressed in baculovirus infected BTI insect cells. The multiple crystal structures alignment of the enzyme highlighted pronounced induced fit (IF) adaptations with respect to bound ligands. Therefore, IF docking (IFD) experiments and molecular dynamic (MD) simulations were carried out to reveal the putative binding mode and to explain the experimentally observed differences in the activity of 1-(aryliden-2-(4-(4- chlorophenyl)thiazol-2-yl)hydrazines. The importance of water molecules within the binding site was also investigated. These are known to play an important role in the binding site cavity and to mediate protein-ligand interactions. Detailed analyses of the trajectories provide insights on the chemical features required for the activity of this scaffold. In particular it was highlighted the importance of fluorine atom interacting with the water close to the cofactor and the influence of steric bulkiness of substituents in the arylidene moiety. Free energy perturbation (FEP) analysis confirmed experimental data. The information we deduced will help to develop novel high-affinity MAO-B inhibitors.

Full text of DOI:10.1016/j.ejmech.2011.12.027

European Journal of Medicinal Chemistry 48 (2012) 284e295
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and biological assessment of novel 2-thiazolylhydrazones and
computational analysis of their recognition by monoamine oxidase B
,
c
d
Simona Distinto a, Matilde Yáñez b, Stefano Alcaro a , M. Cristina Cardia , Marco Gaspari ,
M. Luisa Sanna c, Rita Meleddu c, Francesco Ortuso a, Johannes Kirchmair e, Patrick Markt e,
Adriana Bolasco f, Gerhard Wolber g, Daniela Secci f, Elias Maccioni c
*
a Dipartimento di Scienze della Salute, Università degli Studi “Magna Græcia” di Catanzaro, Campus “Salvatore Venuta”, Viale Europa, 88100 Catanzaro (CZ), Italy
b Departamento de Farmacología and Instituto de Farmacia Industrial, Universidad de Santiago de Compostela, Campus Universitario Sur, E-15782 Santiago de Compostela, Spain
c Dipartimento Farmaco Chimico Tecnologico, University of Cagliari, Via Ospedale 72, 09124 Cagliari, Italy
d Dipartimento di Medicina Sperimentale e Clinica, Università degli Studi “Magna Græcia” di Catanzaro, Campus “Salvatore Venuta”, Viale Europa, 88100 Catanzaro (CZ), Italy
e Department of Pharmaceutical Chemistry, University of Innsbruck, Innrain 52c, A-6020 Innsbruck, Austria
f Dipartimento di Studi di Chimica e Tecnologia delle Sostanze Biologicamente Attive, University “La Sapienza” of Rome, P.le A. Moro 5, 00185 Rome, Italy
g Department of Pharmaceutical Chemistry, University “Freie” of Berlin, Königin-Luise-Str. 2, 14195 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Monoamine oxidase B (MAO-B) is a promising target for the treatment of neurodegenerative disorders.
We report the synthesis and the biological evaluation of halogenated derivatives of 1-aryliden-2-(4-
phenylthiazol-2-yl)hydrazines. The fluorinated series shows interesting activity and great selectivity
toward the human recombinant MAO-B isoform expressed in baculovirus infected BTI insect cells. The
multiple crystal structures alignment of the enzyme highlighted pronounced induced fit (IF) adaptations
with respect to bound ligands. Therefore, IF docking (IFD) experiments and molecular dynamic (MD)
simulations were carried out to reveal the putative binding mode and to explain the experimentally
observed differences in the activity of 1-(aryliden-2-(4-(4-chlorophenyl)thiazol-2-yl)hydrazines.
The importance of water molecules within the binding site was also investigated. These are known to
play an important role in the binding site cavity and to mediate proteineligand interactions. Detailed
analyses of the trajectories provide insights on the chemical features required for the activity of this
scaffold. In particular it was highlighted the importance of fluorine atom interacting with the water close
to the cofactor and the influence of steric bulkiness of substituents in the arylidene moiety. Free energy
perturbation (FEP) analysis confirmed experimental data. The information we deduced will help to
develop novel high-affinity MAO-B inhibitors.
Received 2 March 2011
Received in revised form
25 November 2011
Accepted 18 December 2011
Available online 24 December 2011
Keywords:
FEP
GRID
Induced fit docking
MAO-B inhibitors
Molecular dynamics
Ó 2011 Elsevier Masson SAS. All rights reserved.
1. Introduction
Alzheimer’s disease, Huntington’s chorea, and amyotrophic lateral
sclerosis [3,4].
It is common knowledge that both isoforms of monoamine
oxidase, A and B, (MAO, EC 1.4.3.4), play a key role in the metabo-
lism of neurotransmitters and are important targets for the treat-
ment of psychiatric and neurological diseases. In particular, the
relevance of MAO-B in the pathogenesis of Parkinson’s disease (PD)
and the therapeutic potential of MAO-B selective inhibitors in this
pathology has been pointed out [1,2]. Moreover, the interest toward
the B isoform of MAO has grown since the detection of increased
MAO-B levels in a number of neurodegenerative disorders such as
Thus, MAO-B inhibitors are considered promising agents for
the treatment of several neurodegenerative disorders. Selegiline
[R-(ꢀ)-deprenyl], an irreversible MAO-B inhibitor has been used in
combination with l-Dopa for PD treatment for some time. Evidence
on the effect of slowing down the progression of PD suggests these
inhibitors could be useful therapeutic agents in the early stages of
the disease. Recently, rasagiline, (N-propargyl-1-R-aminoindan,
Azilect), a novel selective and irreversible MAO-B inhibitor, has
been approved for PD therapy [5] while safinamide, a selective and
reversible MAO-B inhibitor, is currently undergoing clinical phase
III for the treatment of early stages of PD [6].
On this basis, we focused our interest on the design of potential
MAO-B inhibitors. We have recently reported the synthesis and the
* Corresponding author. Tel.: þ39 0961 3694197; fax: þ39 0961 515509.
E-mail address: alcaro@unicz.it (S. Alcaro).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.12.027

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
285
biological activity of a series of 2-thiazolylhydrazones [7]. Encour-
aged by these results we have synthesized and studied a new series
of 1-arylidene-2-(4-arylthiazol-2-yl)hydrazines to investigate the
influence of the introduction of 4-chloro- or 4-fluoro-phenyl
moiety in the position four of the thiazole ring. The importance of
fluorine in medicinal chemistry is well recognized and discussed
extensively in literature [8,9]. The introduction of fluorine causes
modulation of electronic, lipophilic, and steric parameters, which
determine pharmacodynamic and pharmacokinetic properties and
may also influence biological activity. In our study, fluorine seems
to play a pivotal role for both activity and selectivity toward the B
isoform.
In recent years, thanks to the availability of several crystal
structures, the mechanism and the basis of selective inhibition of
MAO-B has become clearer. A requirement that seems to be
important for selectivity is the ability of the inhibitors to fit within
both substrate and entrance cavity; since the elongated bipartite
cavity of MAO-B is bigger (700 Å3) than the single round substrate
cavity of MAO-A (550 Å3) [10,11]. Rational drug design relies on in
silico methods. Their application is growing rapidly in both
academia and industry, as such methods allow for increasing lead
retrieval rates and understanding the molecular mechanism of
action of known and new active compounds [12].
All reactions were performed using isopropyl alcohol as solvent
and were monitored by TLC analysis. The obtained reaction prod-
ucts (1e6)a and (1e6)b precipitated upon cooling, and were
filtered and purified by crystallization from ethanol or ethanol/
isopropanol. All the synthesized products have been characterized
by analytical and spectroscopic methods (see Experimental
section).
The hMAO activity was evaluated with microsomal MAO iso-
forms prepared from insect cells (BTI-TN-5B1-4) infected with
recombinant baculovirus containing cDNA inserts for hMAO-A or
hMAO-B and following the general procedure (see Experimental
section) [19]. The tested drugs (new compounds and reference
inhibitors) inhibited the control enzymatic MAO activities and the
inhibition was concentration-dependent. The corresponding IC50
values are shown in Table 1.
3. Computational methods
Several computational methods were applied on three levels to
gain insight about: i) structure of the enzyme; ii) molecular
recognition; iii) free energy evaluation of complexes (Table 2).
3.1. Structure alignment
Several computational studies on this target have been pub-
lished and help by elucidating the activity of different series of
compounds [13e16]. Is it known that small modifications on small
molecules may lead to different activity or target selectivity [17,18].
In order to better understand the differences in activity of this
series of synthesized compounds, where fluorine was substituted
by isosteric chlorine, IFD experiments of the two series have been
carried out and followed by MD simulations for the most active
4-fluorophenyl derivative, 5a, and the homologous 4-chlorophenyl
compound 5b. Moreover, to confirm experimental data we have
performed FEP analysis.
The starting geometries for the MD simulations were acquired
by means of docking the compounds into the MAO-B active site.
Several X-ray structures co-crystallized with different MAO-B
inhibitors are available from the Protein Data Bank (PDB) [20].
With the purpose to identify differences in the binding mode and to
gain knowledge on the conserved waters within the binding
pocket, we aligned twenty complexes with resolution better than
2.5 Å and without mutations (PDB codes: 2byb, 2bk3, 2c64, 2c65,
2c66, 2c67, 2c70, 1oj9, 1oja, 1ojb, 1ojc, 1s2q, 1s2y, 1s3b, 1s3e, 2v5z,
2v60, 2v61, 2vrm, 2vrl). The 3D-structures and the sequences of
crystallized complexes were aligned with MOE-Align [21] (Fig. 1).
The alignment revealed that, while most of the residues in the
substrate cavity are almost perfectly superimposed and are char-
acterized by low b-factor values, the residues of the access cavity
are quite flexible (e.g. AA 99-105, Tyr326, Ile199, Phe118, Trp119).
2. Chemistry and biochemistry
The synthetic procedure for the preparation of the novel 2-
thiazolylhydrazone derivatives (1e6)a and (1e6)b was performed
as reported in our previous communication [7]. The appropriate
aryl aldehydes react directly with thiosemicarbazide, and the
obtained thiosemicarbazones react either with 2-bromo-40-
chloroacetophenone or 2-chloro-40-fluoroacetophenone to give the
4-substituted thiazole ring derivatives (Scheme 1).
3.2. Structural water
Nine water molecules within a 4.5 Å distance range radius from
the crystallized ligands, conserved in the investigated PDB
O
X
R
R
S
Y
O
S
H2N
iPrOH
N
H
NH2
iPrOH
N
N
H
NH2
H
R
R= H. 4-Cl, 3,4-OCH3, 4-OCH3, 2-CH3 , 4-CH3
X=Cl, Y= F, 1a-6a
X=Br, Y=Cl, 1b-6b
N
NH
N
S
Y
Scheme 1. Synthetic pathway to compounds 1ae6a, 1be6b.

286
S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
Table 1
IC50 values for the inhibitory effects of test drugs (new compounds and reference inhibitors) on the enzymatic activity of human recombinant MAO isoforms expressed in
baculovirus infected BTI insect cells.
Compounds
R
MAO-A (IC50
)
MAO-B (IC50)(
m
M) Compounds
R
MAO-A (IC50
)
MAO-B (IC50))(mM)
1a
***
0.79 ꢁ 0.04
1b
***
***
2a
3a
***
***
1.32 ꢁ 0.05
2.39 ꢁ 0.10
2b
3b
***
***
***
***
4a
5a
***
***
9.24 ꢁ 0.36
0.19 ꢁ 0.01
4b
5b
***
***
***
***
6a
**
44.74 ꢁ 1.68
6b
**
***
Drugs
Structure
MAO-A (IC50)(
m
M) MAO-B (IC50)(
m
M) Drugs
Iproniazide
Structure
MAO-A (IC50)(
m
M) MAO-B (IC50)(mM)
Clorgyline
0.0046 ꢁ 0.0003a
67.25 ꢁ 1.02a
61.35 ꢁ 1.13
6.56 ꢁ 0.76
7.54 ꢁ 0.36
Deprenyl
0.019 ꢁ 0.00086
Moclobemide
361.38 ꢁ 19.37
*
All IC50 values shown in this table are the mean ꢁ S.E.M. from five experiments.
* Inactive at 1 mM (highest concentration tested).
** Inactive at 100
*** 100
M inhibits the corresponding MAO activity by approximately 40e45%. At higher concentration the compounds precipitate.
Level of statistical significance: P < 0.01 versus the corresponding IC50 values obtained against MAO-B, as determined by ANOVA/Dunnett’s.
mM (highest concentration tested). At higher concentration the compounds precipitate.
m
a
complexes, have been identified. These are likely to be important
for mediating interactions of inhibitors and were therefore taken
into account in the docking experiments. In fact, several studies
have highlighted the importance of water bound to residues in the
binding pocket in numerous proteins [22,23]. Previous docking
studies have already pointed out the impact of water molecules in
proximity of the MAO-B flavin adenine dinucleotide (FAD) cofactor
[24e26] and recent crystallization studies have shown their role in
recognition and stabilization of the interaction between the ligand
and its binding site [11,27]. To estimate the significance of these
Table 2
Computational approaches applied.
Approach
Program
Aim
i) Structure
analysis
- Structure alignment
PDB models
MOE alignment
Investigation of flexible areas.
- Structural water
study
GRID analysis with OH2 probe
Electron density map analysis with Coot
Analysis of energetic favorable areas for specific OH2 (water) probe.
Model analysis -i.e. checking that the atomic model agrees with the
experimentally derived electron density.
B-factors analysis from PDB files
B-factor is an indicator of thermal atom motion, therefore, it allows
to evaluate the strength of evidence for their presence.
It enables the analysis of energetic favorable areas.
- Molecular Interaction
Fields (MIF)
GRID analysis with N1, DRY, F probes
ii) Molecular
recognition
- Docking
Induced fit Docking protocol implemented
into Schrödinger Suite
Exploration of binding modes and the associated conformational changes
within receptor active sites.
- Dynamic simulation
Desmond
It provides insights concerning the internal motions of the complexes by
considering it a dynamic model in explicit aqueous solvent.
FEP calculations for the mutation of a ligand functional group.
iii) Energy
evaluation
- Mutational free energy
perturbation (FEP)
Desmond

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
287
3.3. Docking experiments
In order to overcome the limitation of most traditional docking
techniques, that consider the protein rigid, we decided to apply the
IFD workflow included in the Schrödinger Software Suite [34]. This
is a novel method for modeling the conformational changes
induced by ligand binding, which uses the combination of the
docking program Glide [34], to account for ligand flexibility, and
Prime [34], for side-chain rearrangements and minimization of the
residues within the binding pocket [35].
First, we docked three reference reversible inhibitors available as
co-crystallized structures and then our synthesized compounds. The
compounds co-crystallized are two coumarin derivatives C17 and
C18 available in PDB 2v60 and 2v61 respectively, and the safinamide,
SAG, a benzyloxy-benzylaminopropionamide in 2v5z [11]. The input
structures were built using Maestro and optimized employing
Macromodel applying MMFFs (Merck Molecular Force Field). Water
was considered implicitly; the PRCG (Polak-Ribier Conjugate
Gradient) method was used for minimization, allowing 500
maximum iterations and a 0.05 kcal/(molÅ) convergence threshold.
The structures obtained were docked into PDB structure 2v61 after
preparation of the enzyme employing the Schrödinger protein
preparation wizard [34]. The workflow includes the assignment of
correct atom bond orders, the addition of hydrogen, the optimization
of hydrogen bonds, the determination of the most favorable ligand
protonation state, and the minimization of the protein. Only the
water molecules conserved within the binding site were preserved.
The IFD workflow consists of a grid setup, followed by three docking
steps: Glide rigid docking, Prime induced fit docking, and Glide re-
docking. A grid extended to 26 Å around the co-crystallized ligand
was defined. The ligands were docked using a scaling factor of 0.5 for
the van der Waals radii of both the receptor and the ligands,
generating a maximum of 20 poses. The calculated proteineligand
complexes were then refined using Prime: residues within a 5 Å
distance range from the ligand were energy minimized using the
OPLS 2005 force field. Only the poses showing Prime energies below
or equal to 30 kcal/mol, above the lowest conformation energy, were
re-docked using the extra precision method (XP) and scored. The XP
Score turned out to be the best score for the docked reference
inhibitors and, therefore, it was also used to select the poses for the
synthesized compounds (see Supplementary data (SD) Table S1, S2).
Fig. 1. Alignment of twenty MAO-B PDB complexes. The residues (sticks) and conserved
water molecules (spheres) are colored according to b-factor: blue indicates low, violet
intermediate and red high temperatures respectively. Ligands co-crystallized (sticks) and
FAD cofactor (spacefill) are colored by CPK notation. (For interpretation of the references
to colour in this figure legend, the reader is referred to the web version of this article.)
water molecules GRID maps [28] were computed and analyzed. The
GRID program calculates favorable interaction sites between
a certain target and a selected probe, mimicking a chemical moiety
[29]. The approach has been applied successfully for the investiga-
tion of water molecules in several studies [30,31]. The water (OH2)
probe energies of interaction were calculated with a grid spacing of
0.5 Å (NPLA (Number of PLanes of grid point per Å) ¼ 2). A cut-off
value of ꢀ8 kcal/mol was chosen to detect, with the combination
of MINIM and FILMAP utilities, the most favorable locations for water
molecules. Furthermore, the analysis of the water molecules was
extended to electron density maps [32] using Coot [33] software and
the b-factors reported in PDB structural data (Fig. 2).
Fig. 2. Water analysis. a) LigandScout visualization of the water probe GRID maps calculated from PDB entry 2v61 (dark blue) in which C18 7-[(3-chlorophenyl)methoxy]-4-
(methylaminomethyl)chromen-2-one is co-crystallized with hMAO-B. Minimum points of the GRID maps are illustrated as green spheres; conserved co-crystallized water
molecules in the binding site are indicated as red spheres, ligand as sticks, and the cofactor in space fill representation. The receptor binding pocket is visualized as a wireframe
model colored according to lipophilicity: light blue for hydrophilic and pale yellow for hydrophobic residues. b) Electron density map of PDB complex 2v61 visualized in Coot, with
the 2FoꢀFc map (blue), the positive density of the FoꢀFc map (i.e., parts of the electron density not represented in the model, in green), and the negative density (i.e., parts of the
model that are not backed up by electron density, in red). Protein, ligand, and water are colored by b-factor: From dark blue for low b-factor to cyan and green for higher b-factor.
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

288
S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
Fig. 3. Comparison of the crystallized (in blue) and docked poses (in gray) of ligands. a) C18-7-[(3-chlorophenyl)methoxy]-4-(methylaminomethyl)chromen-2-one, b) SAG-(S)-
(þ)-2-[4-(fluorobenzyloxy-benzylamino)propionamide] and c) C17-7-[(3-chlorophenyl)methoxy]-2-oxo-chromene-4-carbaldehyde docked into the MAO 2v61 PDB model, with the
FAD displayed as CPK colored space fill rendering and the closest residues as wireframe. The electron density maps of the original complex are reported in d) C18-2v61, e) SAG-2v5z
and f) C17-2v60 panels. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3.3.1. Docking of reference compounds
catalytic cavity pointing toward the cofactor (binding mode II).
Coexistence of multiple binding modes has to be taken into
consideration after docking experiments [38]. As the aim of this
work is to gain some insight into the binding mode in order to
optimize activity by keeping the great selectivity for the MAO-B
isoform, we decided to follow up the docking study with a molecular
dynamics simulation on the complex of 5a, the best compounds of
the a series, and 5b, the analog compound where F was replaced
with Cl. The poses with the lowest XP Score of the two different
binding modes of both compounds (SD Table S2) were submitted to
MD simulation by Desmond.
The IFD workflow reproduced the binding mode of three co-
crystallized compounds with an RMSD (Root-Mean-Square Devia-
tion) ranging from 1.3 to 2.7 Å (Fig. 3, SD Table S1).
It is always the object of scientific discussion whether the root-
mean-square distance (RMSD) calculation is valid for docking
performance evaluation and several papers address the drawbacks of
RMSD, suggesting other approaches for the evaluation of predicted
and experimental pose, like IBAC (Interactions-Based Accuracy
Classification) [36] and RSR (Real Space R-factor) [37]. This latter
descriptor takes the electron density maps into account. In our case,
the pose of C18-2v61 (Fig. 3a) was well predicted except for the
methylamino substituent, however, the electron density maps of this
particular area raise doubts about the exclusive location of the
methylamino substituent as proposed by the X-ray structural model.
(Fig. 3d) These observations indicate the docked orientation of this
portion also as plausible. The co-crystallized compound C17, re-
ported in 2v60 PDB complex (Fig. 3c,f), is predicted with the lowest
accuracy considering the RMSD. Also in this case, the portion not
well predicted (i.e., the coumarin moiety) has no directional inter-
action (i.e. hydrogen bonds) with the protein and, by considering the
electron density maps, we judged this docking pose as acceptable
too. Finally, SAG-2v5z was docked correctly and showed RMSD equal
to 1.44 Å (Fig. 3b,e).
3.3.2. Docking of synthesized compounds
The input structures were built using Maestro as described above.
The E-configuration of these compounds is known to be energeti-
cally favorable e as also confirmed by a previous computational
study performed on a similar molecule [7]. Therefore, the E-config-
uration was used for the docking experiments. Visual inspection of
the docking results highlighted many of the compounds double
binding on the best-scored poses. The two configurations resulted
rotated 180ꢂ: with the halogen atom close to the entrance cavity
toward the external part of the enzyme (binding mode I) or in the
Fig. 4. Superimposition of average MD poses of 5a-I (cyan), 5a-II (dark blue), 5b-I
(light blue), 5b-II (marine) onto the crystallographic model 2v61 (gray) showing the
flexibility of some residues of Phe99-Tyr112 loop. (For interpretation of the references
to colour in this figure legend, the reader is referred to the web version of this article.)

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
289
Table 3
Ewald method (PME) [42] was used for long range electrostatic
interaction. The analysis of the trajectories were performed with
Desmond simulation analysis event and VMD [43]. Finally, the
average structure of the last ns of the simulation of both binding
modes, named I and II, was energy minimized and used as a start-
ing point to run FEP analysis by Desmond software [44], in order to
estimate the free energy variation as functional of F vs Cl.
Relative binding, solvation and difference free energies in kcal/mol in the FEP
analysis.
Compound
Mutation
D
G-complex
D
G-solvent
DDG
5a-I
5a-II
F / Cl
F / Cl
0.75 ꢁ 0.12
3.40 ꢁ 0.15
0.55 ꢁ 0.08
0.68 ꢁ 0.09
0.21
2.72
3.4. Molecular dynamic simulations and free energy perturbation
analysis
4. Results and discussion
Tested compounds (novel compounds as well as and reference
inhibitors) themselves were unable to react directly with the Amplex
Red reagent, demonstrating no interference with the measurements.
In our experiments, hMAO-A displayed a Michaelis constant (Km) of
The complexes were solvated with a 10 Å box of TIP3P (Trans-
ferable Intermolecular Potential 3-Point) water [39] and counter
ions were added to neutralize the system net charge. The solvated
models were relaxed and energy minimized, and subsequently the
MTK_NPT (Martyna-Tobias-Klein with constant Number of parti-
cles, Pressure and Temperature) ensemble was employed. The
default stages in the relaxation process for the NPT ensemble
include two energy minimizations, and four simulation steps.
During the energy minimizations, two runs of 2000 iteration steps
were processed using the steepest descent method: during the first
run, the protein structure was fixed by a force restraint constant of
50 kcal/(molÅ) and in the second all restraints were released. With
the first simulation, at NVT (constant Number of particles, Volume,
and Temperature) ensemble, the system reached a temperature of
10 K. In the following three simulations in the NPT ensemble, the
system was heated up to 300 K and the pressure was kept constant
at 1 bar using the Berendsen thermostatebarostat. During the
production phase, temperature and pressure were kept constant
using the Nosè-Hoover thermostatebarostat. The energy and
trajectory were recorded every 1.2 ps and 4.8 ps, respectively. For
multiple time step integration, RESPA (REversible reference System
Propagator Algorithm) [40] was applied to integrate the equation of
motion with Fourier-space electrostatics computed every 6 fs, and
all remaining interactions computed every 2 fs. All chemical bond
lengths involving hydrogen atoms were fixed with SHAKE [41].
Short range cut-off was set to 9 Å and the smooth particle mesh
514 ꢁ 46.8
301.4 ꢁ 27.9 nmol/min/mg protein, whereas hMAO-B showed a Km of
104.7 ꢁ 16.3 M and a Vmax of 28.9 ꢁ 6.3 nmol/min/mg protein
(n ¼ 5). All synthesized compounds were inactive toward MAO-A
below 100 M, suggesting the 1-aryliden-2-(4-(4-halophenyl)thia-
mM and a maximum reaction velocity (Vmax) of
m
m
zol-2-yl)hydrazine as a promising candidate scaffold for the design of
selective MAO-B inhibitors. Moreover, compounds (1e6)b, bearing
a chlorine atom in the para position of the phenyl moiety, turned out
to be inactive toward the B isoform. However, compounds (1e6)a,
were all active in the nMemM range (Table 1). The substitution at
the phenyl moiety in position two of thiazole showed to modulate
the activity within the a series. In particular, in cases where no
substitution (1a) or a 2-methyl group was introduced in this ring
(2a), the compounds exhibited activity in the nanomolar range,
while substitution in other phenyl positions with chlorine, methyl or
methoxy group led to an activity decrease. Minor changes in this
scaffold result in wide changes in activity.
Firstly, we analyzed the alignment of twenty available
complexes. The result of the structure comparison is shown in
Fig. 1: the substrate cavity is quite rigid; no remarkable shift in the
alpha carbon (Ca) position was observed, in accordance to previ-
ously reported findings [45]. However, it needs to be pointed out,
that a few residues within the substrate cavity (i.e., Gln206, Cys172,
Fig. 5. Average conformations of MD trajectories. Compounds are represented in sticks, starting geometries in ball and sticks, FAD in space fill models, respectively. Pharmacophoric
features are visualized as green (HB-donor), red (HB-acceptor) arrows, and yellow spheres (hydrophobic interactions). The surface of the binding pocket is represented in wireframe,
colored in accordance to lipophilicity: pale yellow indicates lipophilic and light blue hydrophilic residues. Binding modes and 2D depictions of interactions are reported for 5a-I, 5a-
II, 5b-I and 5b-II respectively in panels a), b), c) and d). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

290
S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
Leu171, and Phe343) were characterized by side chain conforma-
tional flexibility among the different crystal structures (SD
Table S3a). The shift of these chains could influence the interac-
tions between enzyme and inhibitors. Moreover, the higher
b-factors (Fig. 1) and Ca-RMSD values (SD Table S3b) of residues
forming the entrance cavity clearly indicate this portion of the
a gatekeeper between both cavities was published. Rotation of the
Ile199 side chain allows for fusion of the two cavities and could be
observed upon binding of inhibitors across both cavities [46].
In addition, an important role of the Phe99-Tyr112 loop in regu-
lating substrate recognition has been reported [47,48]. Hence,
induced fit docking experiments appear to be a suitable approach
for reproducing experimental observations.
enzyme is rather flexible. In previous studies, the role of Ile199 as
Fig. 6. MD trajectory analysis. a) Distances between F in 5a-II and O-W1431; b) number of hydrogen bonds between F and W1431; c) distance of NH-5a-II from O-Tyr326 and O-
W1294; d) number of hydrogen bonds between 5a-II-Tyr326 and 5a-II W1294.

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
291
Water molecules located within the binding pocket (W1122,
W1207, W1220, W1221, W1294, W1295, W1387, W1388, W1431;
numbering according to 2v61) were preserved. Besides the fact that
these water molecules are conserved in several crystal structures
(Fig. 1) and show multiple interactions with the protein, a corre-
spondence with the most favorable water positions obtained from
GRID was observed: energy minimum points for water probe of the
GRID-molecular interaction fields were close to the position of
waters observed in the crystal structure (Fig. 2a). Furthermore, the
analysis of the electron density maps demonstrated all nine water
molecules are well defined and characterized by a low b-factor (i.e.,
b-factor range is between 9 and 16) (Fig. 2b). In the 2v61 model
water molecule (W1434) was not considered since it was missing in
the majority of crystal structures. Moreover, it was characterized by
a b-factor value of 23, rather high if compared to the surrounding
waters and residues; also, its corresponding electron density map
resulted quite poor at this point (Fig. 2b).
In order to describe realistically proteineligand interactions, it is
necessary to take into account the structural modification in both
the receptor and the ligand during the host/guest recognition.
However, in classic docking approaches, the receptor is usually
treated as being rigid or semi-rigid. The analysis of the alignment,
as discussed above, indicated the entrance cavity of the enzyme as
rather flexible and this aspect may be important in the recognition
of bulky compounds. Using the IF approach, the side chains near the
inhibitor were kept flexible. The XP Score was not able to explain
the huge difference in activity of the two series of compounds (SD
Table S1). Looking at the best-scored poses, it appeared clear that
two main binding modes are possible. We decided to focus our
attention on the best compound 5a and the correspondent chloro-
substituted 5b. Both binding modes were analyzed. In the binding
mode I the phenyl substituent in four position of the thiazole is
orientated toward the entrance cavity, conversely in the II toward
the FAD cofactor. Then we ran a 3 ns MD simulation for each
molecule except for the 5b-II where, due to increasing RMSD, and
we decided to extend the simulation up to 5 ns (Fig. S4, SD).
The resulting poses of IFD were thereby taken as a starting
structure for the MD simulations. All four complexes resulted as
being stable. As shown in Fig. 3, the residues in the catalytic cavity
are rather rigid, while residues in the entrance cavity (Phe103 and
Trp119 in particular) moved slightly in order to better accommo-
date the compounds. The average structure of each complex was
energy minimized and analyzed (Fig. 4).
Furthermore, to obtain more information about a preferential
binding mode and to corroborate the experimental data, we have
performed FEP simulations focusing our attention on the effect of
group mutation: fluorine with chlorine. Table 3 shows the esti-
mation of free energy variation (DDG) for such substituent change.
Confirming the experimental data, the FEP analysis demonstrates
that the substitution is unfavorable. Moreover, only the energetic
difference of the second pose is able to explain the large difference
in terms of activity. From this analysis, we assume that the binding
mode proposed by the second pose is the most favorable.
The ability of the compound 5a to be placed across both cavities
illustrates the reason why these compounds are selective for MAO-
B. The arylidene portion is accommodated within the cavity access
of the enzyme and is stabilized by hydrophobic interactions with
Trp119, Leu164, Leu167, Leu171, Ile316. Pharmacophoric interac-
tions are visualized using LigandScout [49]. The analysis of the
interactions stabilizing the complex highlighted a key role of the
water molecules complex stabilization: during the whole MD
simulation the fluorine atom interacts with W1431 molecule
(Figs. 5b and 6a,b). In the crystal structure this structural water is
involved in HB with FAD and Lys296. After docking and during the
MD it bridges the compound 5a to Gly58 and Lys296. In the last part
of the dynamic simulation (Fig. 6c,d) a second molecule, W1294, in
the crystal structure interested in HB contacts respectively with
Gln206, Thr201 and Ile199, is also implicated in bridging the
compound 5a to the protein (Tyr326).
The visual inspection of the binding mode of the compound 5a
suggested that substitution with bulky groups in positions 3 and 4
of the phenyl ring should prevent a good fit in this pocket causing
the loss of complex interactions mentioned above. This observation
provides a rational explanation for the decreased activity of some of
the compounds of the a series. This is even more evident in the case
of compound 5b, which is kept more distant from the cofactor and,
to accommodate it, the Phe103 has to assume an open conforma-
tion. The simple substitution of the F with Cl does not allow the
Fig. 7. GRID maps and 5a-II binding mode: yellow DRY probe, blue N1 probe, green F probe. (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)

292
S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
compound to be stabilized by any hydrogen bond, but only by
hydrophobic interactions (Fig. 5d). To further support the II binding
mode of 5a we calculated the GRID maps with the DRY probe, for
detection of favorable hydrophobic areas; N1, to locate favorable
areas for amino groups and Fluorine (F) probe. The analysis and
visualization of these maps confirmed our hypothesis of binding
mode (Fig. 7). Therefore a reason of a series better activity could be
the presence of a less bulky and more electronegative group (F vs
Cl) which allows a better accommodation in the binding pocket and
make possible the stabilization of the complex with H bond.
Additionally we also investigated the ADME profile of the
compound with Qikprop [34] summarized in Table S5 of SD.
The prediction indicates that the compound has good drug-like
characteristics: good oral absorption, does not show any violation
of Lipinski’s [50] and Jorgesen’s [51] rules, good activity for central
nervous system which allowed us to be more confident that such
compound would target the area directly involved in the etiology of
neurodegenerative disorders.
6.1.2. General procedure
For the preparation of compounds 1e6. In a two-necked flask
a mixture of arylidenethiosemicarbazide (0.0051 mol), and 2-
halogenacetophenone (0.0061 mol) are refluxed, under vigorous
stirring in 20 ml of 2-propanol. The reaction is refluxed for a period
ranging between 30 and 1200, until completion by TLC (ethyl acetate/
hexane). The formation of a foaming product, which precipitates, is
observed upon cooling down to room temperature. The solid
precipitate is filtered and crystallized from ethanol or ethanol/2-
propanol. Elemental analyses were within ꢁ0.4% of the theoretical
values. All compounds exhibit well-detectable molecular ions in EI
conditions. HRMS full scan and MS/MS are available as SD.
6.1.3. Analysis
6.1.3.1. 1-Benzylidene-2-[4-(4-fluorophenyl)thiazol-2-yl]hydrazine
(1a). Pale yellow solid, yield 82%; mp 189e191 ꢂC; 1H NMR
(300 MHz, DMSO-d6)
C5H-thiaz.); 7.51e7.59 (m, 3H, Haromatics); 7.79 (d, 2H, Haromatics
d
¼ 7.37 (t, 2H, Haromatics, J ¼ 8.4); 7.44 (s, 1H,
,
J ¼ 7.7); 8.02 (dd, 2H, Haromatics, J ¼ 7.3, 5.4); 8.19 (s, 1H, CH]N);
5. Conclusions
9.25 (s, 1H, NH, D2O-exch.).
We have synthesized and investigated the biological activity of
two closely related series of compounds that show a high gap
inhibition activity. In order to understand the possible binding
mode and to rationalize the activity, we have performed several
computational investigations. Due to the flexibility of the entrance
cavity, we have carried out IFD experiment, followed by MD
simulations and FEP calculations. The modeling methods employed
allowed us to derive structureeactivity relationships that high-
lighted the importance of the presence of a fluorine substituent
interacting with the water close to the cofactor as well as the
importance of substituent position in the arylidene moiety. The key
interactions can then be used to derive conclusions and will be
taken into account for further modifications on this promising
scaffold which showed great selectivity for the MAO-B isoform.
Moreover, relevant pharmacophoric features have been resolved
and will constitute the basis for future work.
6.1.3.2. 1-(4-Chlorobenzylidene)-2-[4-(4-fluorophenyl)thiazol-2-yl]
hydrazine (2a). Pale yellow solid, yield 74%; mp 195e197 ꢂC; 1H
NMR (300 MHz, DMSO-d6)
d
¼ 7.24 (t, 2H, Haromatics, J ¼ 8.8); 7.33 (s,
1H, C5H-thiaz.); 7.49 (d, 2H, Haromatics, J ¼ 8.4); 7.68 (d, 2H, Haromatics
,
J ¼ 8.4); 7.89 (dd, 2H, Haromatics, J ¼ 8.3, 5.8); 8.04 (s, 1H, CH]N);
12.33 (s, 1H, NH, D2O-exch.).
6.1.3.3. 1-(3,4-Dimethoxybenzylidene)-2-[4-(4-fluorophenyl)thiazol-
2-yl]hydrazine (3a). Pale pink powder, yield 95%; mp 194e196 ꢂC;
1H NMR (400 MHz, DMSO-d6)
d
¼ 3.95 (s, 6H, OCH3); 6.70 (s, 1H,
C5H-thiaz.); 6.92 (d, 1H, Haromatics, J ¼ 8.4); 7.7.16e7.22 (m, 3H,
Haromatics) 7.29 (d, 1H, Haromatics, J ¼ 2.0); 7.73 (dd, 2H, Haromatics
J ¼ 8.5, 5.0); 8.16 (s, 1H, CH]N); 9.01 (s, 1H, NH, D2O-exch.).
,
6.1.3.4. 1-(4-Methoxybenzylidene)-2-[4-(4-fluorophenyl)thiazol-2-
yl]hydrazine (4a). Yellow powder, yield 94%; mp 198e200 ꢂC; 1H
NMR (300 MHz, DMSO-d6)
d
¼ 3.79 (s, 3H, OCH3); 7.00 (d, 2H,
6. Experimental section
Haromatics, J ¼ 8.6); 7.23 (t, 2H, Haromatics, J ¼ 8.8); 7.28 (s, 1H, C5H-
thiaz.); 7.60 (d, 2H, Haromatics, J ¼ 8.6); 7.88 (dd, 2H, Haromatics, J ¼ 8.2,
5.9); 8.02 (s, 1H, CH]N); 9.65 (s, 1H, NH, D2O-exch.).
6.1. Chemistry
6.1.1. General methods
6.1.3.5. 1-(2-Methylbenzylidene)-2-[4-(4-fluorophenyl)thiazol-2-yl]
hydrazine (5a). Brilliant yellow powder, yield 100%; mp
Melting points were uncorrected and were determined on
a Reichert Kofler thermopan apparatus. TLC analyses were per-
formed on silica gel 60 F254 plates; spots were visualized by UV
light. All synthesized compounds were purified by crystallization
from an appropriate solvent (ethanol, ethanol/2-propanol).
Elemental analyses were obtained on a Perkin Elmer 240 B
microanalyzer. Electron ionization (EI) mass spectra were obtained
195e198 ꢂC; 1H NMR (400 MHz, DMSO-d6)
d
¼ 2.56 (s, 3H, CH3);
7.33e7.40 (m, 6H, Haromatics þ C5H-thiaz.); 7.82e7.85 (m, 1H, Har-
omatics); 8.02 (dd, 2H, Haromatics, J ¼ 7.3, 5.4); 8.47 (s, 1H, CH]N); 9.17
(s, 1H, NH, D2O-exch.).
6.1.3.6. 1-(4-Methylbenzylidene)-2-[4-(4-fluorophenyl)thiazol-2-yl]
hydrazine (6a). Pale yellow powder, yield 92%; mp 194e195 ꢂC; 1H
by a Fisons QMD 1000 mass spectrometer (70 eV, 200
mA, ion
source temperature 200 ꢂC). The samples were introduced directly
into the ion source. High resolution mass spectra (HRMS) were
recorded using a QSTAR XL hybrid quadrupole time-of-flight mass
spectrometer from AB Sciex (Foster City, CA). Analytes were solu-
bilized in methanol at a concentration of 0.5 mg/mL and subse-
quently diluted in infusion solvent (acetonitrile:formic acid:water
NMR (300 MHz, DMSO-d6)
d
¼ 2.46 (s, 3H, CH3); 7.34e7.42 (m, 5H,
Haromatics þ C5H-thiaz.); 7.67 (d, 2H, Haromatics, J ¼ 6.4); 8.02 (dd, 2H,
Haromatics, J ¼ 7.3, 5.4), 8.14 (s, 1H, CH]N) 9.65 (s, 1H, NH, D2O-
exch.).
6.1.3.7. 1-Benzylidene-2-[4-(4-chlorophenyl)thiazol-2-yl]hydrazine
50/0.1/19.9 v/v) at a concentration of 5
were infused at 5 L/min into the mass spectrometer operating in
positive ion mode.
mg/mL. Analyte solutions
(1b). White powder, yield 89%; mp 222e224 ꢂC; 1H NMR
m
(300 MHz, DMSO-d6)
d
¼ 7.38e7.48 (m, 6H, Haromatics þ C5H-thiaz.);
7.66 (d, 2H, Haromatics, J ¼ 7.9); 7.87 (d, 2H, Haromatics, J ¼ 8.5); 8.05 (s,
1H NMR spectra were recorded on a Varian (300 MHz) or on
a Varian Unity 600 (600 MHz); deuterated DMSO was used as
1H, CH]N); 8.75 (s, 1H, NH, D2O-exch.).
solvents. Chemical shifts are expressed as
using TMS as an internal standard. Coupling constants J are valued
d
units (parts per million)
6.1.3.8. 1-4-Chlorobenzylidene-2-[4-(4-chlorophenyl)thiazol-2-yl]
hydrazine (2b). Brown powder, yield 100%; mp 215e217 ꢂC; 1H
in Hertz (Hz).
NMR (300 MHz, DMSO-d6)
d
¼ 7.53 (s, 1H, C5H-thiaz.); 7.57e7.63

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
293
(m, 4H, Haromatics); 7.80 (d, 2H, Haromatics, J ¼ 8.4); 7.99 (d, 2H, Har-
omatics, J ¼ 8.4); 8.44 (s, 1H, CH]N); 12.36 (s, 1H, NH, D2O-exch.).
resorufin was quantified at 37 ꢂC in a multidetection microplate
fluorescence reader (FLX800TM, Bio-TekÒ Instruments, Inc.,
Winooski, VT, USA) based on the fluorescence generated (excitation,
545 nm, emission, 590 nm) over a 15 min period, in which the
fluorescence increased linearly.
Control experiments were carried out simultaneously by
replacing the test drugs (new compounds and reference inhibitors)
with appropriate dilutions of the vehicles. In addition, the potential
ability of the test drugs to modify the fluorescence generated in the
reaction mixture due to non-enzymatic inhibition (e.g., for directly
reacting with AmplexÒ Red reagent) was determined by adding
these drugs to solutions containing only the AmplexÒ Red reagent
in a sodium phosphate buffer.
6.1.3.9. 1-(3,4-Dimethoxybenzylidene)-2-[4-(4-chlorophenyl)thiazol-
2-yl]hydrazine (3b). Pale pink powder, yield 90%; mp 214e216 ꢂC;
1H NMR (300 MHz, DMSO-d6)
d
¼ 3.79 (s, 3H, OCH3); 3.81 (s, 3H,
OCH3); 7.01 (d, 1H, Haromatics, J ¼ 8.3); 7.17 (d, 1H, Haromatics, J ¼ 8.3);
7.26 (s, 1H, Haromatics); 7.37 (s, 1H, C5H-thiaz.); 7.46 (d, 2H, Haromatics
,
J ¼ 8.3); 7.86 (d, 2H, Haromatics, J ¼ 8.3); 7.97 (s, 1H, CH]N); 8.98 (s,
1H, NH, D2O-exch.).
6.1.3.10. 1-(4-Methoxybenzylidene)-2-[4-(4-chlorophenyl)thiazol-2-
yl]hydrazine (4b). Yellow solid, yield 92%; mp 232e235 ꢂC; 1H NMR
(300 MHz, DMSO-d6)
d
¼ 3.79 (s, 3H, OCH3); 7.00 (d, 2H, Haromatics
,
To determine the kinetic parameters of hMAO-A and hMAO-B
(Km and Vmax), the corresponding enzymatic activity of both iso-
forms was evaluated (under the experimental conditions described
above) in presence of a number (a wide range) of p-tyramine
concentrations.
J ¼ 8.5); 7.37 (s, 1H, C5H-thiaz.); 7.46 (d, 2H, Haromatics, J ¼ 8.3); 7.60
(d, 2H, Haromatics, J ¼ 8.5); 7.86 (d, 2H, Haromatics, J ¼ 8.3), 8.02 (s, 1H,
CH]N); 10.29 (s, 1H, NH, D2O-exch.).
6.1.3.11. 1-(2-Methylbenzylidene)-2-[4-(4-chlorophenyl)thiazol-2-yl]
The specific fluorescence emission (used to obtain the final
results) was calculated after subtraction of the background activity,
which was determined from vials containing all components except
the hMAO isoforms, which were replaced by a sodium phosphate
buffer solution.
hydrazine (5b). Red powder, yield 94%; mp 210e212 ꢂC; 1H NMR
(300 MHz, DMSO-d6)
d
¼ 2.56 (s, 3H, CH3); 7.37e7.41 (m, 3H,
Haromatics); 7.51 (s, 1H, C5H-thiaz.); 7.58 (d, 2H, Haromatics, J ¼ 8.4);
7.82 (t, 1H, Haromatics, J ¼ 7.7); 7.99 (d, 2H, Haromatics, J ¼ 8.5), 8.41
(s, 1H, CH]N); 8.86 (s, 1H, NH, D2O-exch.).
6.2.2. Data presentation and statistical analysis
6.1.3.12. 1-(4-Methylbenzylidene)-2-[4-(4-chlorophenyl)thiazol-2-yl]
Unless specified otherwise, the results shown in the text and
tables are expressed as mean ꢁ standard error of the mean (S.E.M.)
from n experiments. Significant differences between two means
(P < 0.01) were determined by one-way analysis of variance
(ANOVA) followed by the Dunnett’s post-hoc test.
hydrazine (6b). Pale yellow powder, yield 96%; mp 224e226 ꢂC; 1H
NMR (300 MHz, DMSO-d6)
d
¼ 2.46 (s, 3H, CH3); 7.37e7.39 (m, 3H,
Haromatics, and C5H-thiaz.); 7.52 (s, 1H, NH, D2O-exch.); 7.59 (d, 2H,
Haromatics, J ¼ 7.7); 7.68 (d, 2H, Haromatics, J ¼ 7.7); 8.00 (d, 2H, Har-
omatics, J ¼ 7.3), 8.14 (s, 1H, CH]N); 9.84 (s, 1H, NH, D2O-exch.).
To study the possible effects of the test drugs (new compounds
or reference inhibitors) on MAO isoform enzymatic activity, we
evaluated the variation of fluorescence per unit of time (fluores-
cence arbitrary U/min) and, indirectly, the rate of hydrogen
peroxide (H2O2) production, and therefore the pmol/min of resor-
ufin produced in the reaction between H2O2 and AmplexÒ Red
reagent. For this purpose, several concentrations of resorufin were
used to prepare a standard curve with X ¼ pmol resorufin and
Y ¼ fluorescence arbitrary U. Note that the value of resorufin
production is similar to the pmol of p-tyramine oxidized to p-
hydroxyphenylacetaldehyde/min, since the stoichiometry of the
reaction (p-tyramine oxidized by MAO isoforms/resorufin
produced) is 1:1.
6.2. Biochemistry studies
6.2.1. Determination of hMAO isoform activity
The potential effects of the tested compounds on hMAO activity
were investigated by measuring their effects on the production of
hydrogen peroxide from p-tyramine (a common substrate for both
hMAO-A and hMAO-B), using the AmplexÒ Red MAO assay kit
(Molecular Probes, Inc., Eugene, Oregon, USA) and microsomal
MAO isoforms prepared from insect cells (BTI-TN-5B1-4) infected
with recombinant baculovirus containing cDNA inserts for hMAO-A
or hMAO-B (SigmaeAldrich Química S.A., Alcobendas, Spain).
The production of H2O2 catalyzed by MAO isoforms can be
detected using 10-acetyl-3,7-dihydroxyphenoxazine (AmplexÒ Red
reagent), a non-fluorescent and highly sensitive probe that reacts
with H2O2 in the presence of horseradish peroxidase to produce the
fluorescent product resorufin. In this study hMAO activity was
evaluated using the above fluorimetric method following the
general procedure described previously by us [19,52].
In these experiments, the inhibitory activity of the tested drugs
(new compounds and reference inhibitors) is expressed as IC50, i.e.,
the concentration of these compounds required for a 50% reduction
of the control MAO isoform enzymatic activity, estimated by least-
squares linear regression, using the OriginÔ 5.0 program (Microcal
Software, Inc., Northampton, MA, USA), with X ¼ log of tested
compound molar concentration and
Y
¼
the corresponding
0.1 mL of sodium phosphate buffer (0.05 M, pH 7.4) containing
the test drugs (new compounds or reference inhibitors) in various
concentrations and adequate amounts of recombinant hMAO-A or
hMAO-B required and adjusted to obtain the same reaction velocity
in our experimental conditions, i.e., to oxidize (in the control group)
percentage of inhibition of control resorufin production obtained
with each concentration. This regression was performed using data
obtained with 4e6 different concentrations of each tested
compound which inhibited the control MAO isoform enzymatic
activity by between 20 and 80%. Kinetic parameters (Km and Vmax
)
165 pmol of p-tyramine/min (hMAO-A: 1.1
activity: 150 nmol of p-tyramine oxidized to p-hydrox-
yphenylacetaldehyde/min/mg protein; hMAO-B: 7.5 g protein;
m
g protein; specific
of hMAO-A and hMAO-B were estimated using the GraphPad Prism
5.0 software (GraphPad Software Inc., San Diego, CA, USA).
m
specific activity: 22 nmol of p-tyramine transformed/min/
mg protein), were incubated for 15 min at 37 ꢂC in a flat-black-
bottom 96-well microtestÔ plate (BD Biosciences, Franklin Lakes,
NJ, USA) placed in a dark fluorimeter chamber. After this incubation
period, the reaction was started by adding (final concentrations)
6.2.3. Drugs and chemicals
The drugs, vehicle and chemicals used in the experiments were
the new compounds, moclobemide (a generous gift from F.
Hoffmann-La Roche Ltd., Basel, Switzerland), dimethyl sulfoxide
(DMSO), R-(ꢀ)-deprenyl hydrochloride, iproniazid phosphate
(purchased from SigmaeAldrich, Spain), resorufin sodium salt,
clorgyline hydrochloride, p-tyramine hydrochloride, sodium
200
1 mM p-tyramine. The production of H2O2 and, consequently, of
m
M AmplexÒ Red reagent, 1 U/mL horseradish peroxidase and

294
S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
phosphate and horseradish peroxidase (supplied in the AmplexÒ
Red MAO assay kit from Molecular Probes).
[15] G.H. Braun, D.M.M. Jorge, H.P. Ramos, R.M. Alves, V.B. da Silva, S. Giuliatti,
S.V. Sampaio, C.A. Taft, C. Silva, Molecular dynamics, flexible docking, virtual
screening, ADMET predictions, and molecular interaction field studies to
design novel potential MAO-B inhibitors, J. Biomol. Struct. Dyn. 25 (2008)
347e355.
[16] K. Boppana, P.K. Dubey, S. Jagarlapudi, S. Vadivelan, G. Rambabu, Knowledge
based identification of MAO-B selective inhibitors using pharmacophore and
structure based virtual screening models, Eur. J. Med. Chem. 44 (2009)
3584e3590.
[17] H. Kubinyi, Similarity and dissimilarity: a medicinal chemist’s view, perspect,
Drug Discov. 9-11 (1998) 225e252.
[18] C.G. Wermuth, Selective optimization of side activities: another way for drug
discovery, J. Med. Chem. 47 (2004) 1303e1314.
[19] M. Yáñez, N. Fraiz, E. Cano, F. Orallo, Inhibitory effects of cis- and trans-
resveratrol on noradrenaline and 5-hydroxytryptamine uptake and on
monoamine oxidase activity, Biochem. Biophys. Res. Commun. 344 (2006)
688e695.
[20] H. Berman, K. Henrick, H. Nakamura, Announcing the worldwide protein data
bank, Nat. Struct. Biol. 10 (2003) 980.
Appropriate dilutions of the above drugs were prepared every
day immediately before use in deionized water from the following
concentrated stock solutions kept at ꢀ20 ꢂC: the new compounds
(0.1 M) in DMSO; R-(ꢀ)-deprenyl, moclobemide, iproniazid, resor-
ufin, clorgyline, p-tyramine and horseradish peroxidase (0.1 M) in
deionized water.
Due to the photosensitivity of some chemicals (e.g., AmplexÒ
Red reagent), all experiments were performed in the dark. In all
assays, neither deionized water (Milli-QÒ, Millipore Ibérica S.A.,
Madrid, Spain) nor appropriate dilutions of the vehicle used
(DMSO) had significant pharmacological effects.
[21] Molecular Operating Environment (MOE), Chemical Computing Group, Inc,
Montreal, 2008, 10.
Acknowledgments
[22] C. Barillari, J. Taylor, R. Viner, J.W. Essex, Classification of water molecules in
protein binding sites, J. Am. Chem. Soc. 129 (2007) 2577e2587.
[23] N. Huang, B.K. Shoichet, Exploiting ordered waters in molecular docking,
J. Med. Chem. 51 (2008) 4862e4865.
[24] M. Catto, O. Nicolotti, F. Leonetti, A. Carotti, A.D. Favia, R. Soto-Otero,
E. Mendez-Alvarez, A. Carotti, Structural insights into monoamine oxidase
inhibitory potency and selectivity of 7-substituted coumarins from ligand-
and target-based approaches, J. Med. Chem. 49 (2006) 4912e4925.
[25] L. Novaroli, A. Daina, E. Favre, J. Bravo, A. Carotti, F. Leonetti, M. Catto,
P.A. Carrupt, M. Reist, Impact of species-dependent differences on screening,
design, and development of MAO-B inhibitors, J. Med. Chem. 49 (2006)
6264e6272.
The work was supported by Fondazione Banco di Sardegna-
Sassari-Italy, Ministerio de Sanidad
y Consumo (Spain; FISS
PI061537) and Consellería de Innovación e Industria de la Xunta de
Galicia (Spain; INCITE07PXI203039ES, INCITE08E1R203054ES and
08CSA019203PR). The authors wish to thank Mr Roberto Maxia for
the technical support.
Appendix. Supplementary data
[26] F. Chimenti, E. Maccioni, D. Secci, A. Bolasco, P. Chimenti, A. Granese, O. Befani,
P. Turini, S. Alcaro, F. Ortuso, R. Cirilli, F. La Torre, M.C. Cardia, S. Distinto,
Synthesis, molecular modeling studies, and selective inhibitory activity
against monoamine oxidase of 1-thiocarbamoyl-3,5-diaryl-4,5-dihydro-(1H)-
pyrazole derivatives, J. Med. Chem. 48 (2005) 7113e7122.
Supplementary data related to this article can be found online at
doi:10.1016/j.ejmech.2011.12.027.
[27] C. Binda, F. Hubalek, M. Li, Y. Herzig, J. Sterling, D.E. Edmondson, A. Mattevi,
Crystal structures of monoamine oxidase B in complex with four inhibitors of
the N-propargylaminoindan class, J. Med. Chem. 47 (2004) 1767e1774.
[28] GRID, in, Molecular Discovery Ltd., Pinner, Middlesex, UK.
References
[1] M.B.H. Youdim, D. Edmondson, K.F. Tipton, The therapeutic potential of
monoamine oxidase inhibitors, Nat. Rev. Neurosci. 7 (2006) 295e309.
[2] L.W. Elmer, J.M. Bertoni, The increasing role of monoamine oxidase type B
inhibitors in Parkinson’s disease therapy, Expert Opin. Pharmacother. 9 (2008)
2759e2772.
[3] J. Saura, J.M. Luque, A.M. Cesura, M. Da Prada, V. Chan-Palay, G. Huber,
J. Loffler, J.G. Richards, Increased monoamine oxidase B activity in plaque-
associated astrocytes of Alzheimer brains revealed by quantitative enzyme
radioautography, Neuroscience 62 (1994) 15e30.
[29] P.J. Goodford,
A computational procedure for determining energetically
favorable binding sites on biologically important macromolecules, J. Med.
Chem. 28 (1985) 849e857.
[30] R.C. Wade, Solvation of the active site of cytochrome P450-cam, J. Comput.
Aided Mol. Des. 4 (1990) 199e204.
[31] R.H. Henchman, J.A. McCammon, Structural and dynamic properties of water
around acetylcholinesterase, Protein Sci. 11 (2002) 2080e2090.
[32] G.J. Kleywegt, M.R. Harris, J.-y. Zou, T.C. Taylor, A. Wahlby, T.A. Jones, The
Uppsala electron-density server, Acta Crystallogr. Sect. D: Biol. Crystallogr. 60
(2004) 2240e2249.
[4] M.J. Kumar, D.G. Nicholls, J.K. Andersen, Oxidative a-ketoglutarate dehydro-
genase inhibition via Subtle Elevations in monoamine oxidase B levels results
in loss of spare respiratory capacity: implication for Parkinson’s disease, J. Biol.
Chem. 278 (2003) 46432e46439.
[33] P. Emsley, K. Cowtan, Coot: model-building tools for molecular graphics, Acta
Crystallogr. Sect. D: Biol. Crystallogr. 60 (2004) 2126e2132.
[34] Schrödinger Suite, in, Schrödinger, LLC, New York, NY, USA.
[35] W. Sherman, T. Day, M.P. Jacobson, R.A. Friesner, R. Farid, Novel procedure for
modeling ligand/receptor induced fit effects, J. Med. Chem. 49 (2006) 534e553.
[36] R.T. Kroemer, A. Vulpetti, J.J. McDonald, D.C. Rohrer, J.-Y. Trosset,
F. Giordanetto, S. Cotesta, C. McMartin, M. Kihlen, P.F.W. Stouten, Assessment
of docking poses: interactions-based accuracy classification (IBAC) versus
crystal structure deviations, J. Chem. Inf. Comput. Sci. 44 (2004) 871e881.
[37] D. Yusuf, A.M. Davis, G.J. Kleywegt, S. Schmitt, An alternative method for the
evaluation of docking performance: RSR vs RMSD, J. Chem. Inf. Model. 48
(2008) 1411e1422.
[5] J.J. Chen, D.M. Swope, K. Dashtipour, Comprehensive review of rasagiline,
a second-generation monoamine oxidase inhibitor, for the treatment of Par-
kinson’s Disease, Clin. Ther. 29 (2007) 1825e1849.
[6] R.G. Fariello, Safinamide, Neurotherapeutics 4 (2007) 110e116.
[7] F. Chimenti, E. Maccioni, D. Secci, A. Bolasco, P. Chimenti, A. Granese, O. Befani,
P. Turini, S. Alcaro, F. Ortuso, M.C. Cardia, S. Distinto, Selective inhibitory
activity against MAO and molecular modeling studies of 2-thiazolylhydrazone
derivatives, J. Med. Chem. 50 (2007) 707e712.
[8] W.K. Hagmann, The many roles for fluorine in medicinal chemistry, J. Med.
Chem. 51 (2008) 4359e4369.
[38] P. Kallblad, R.L. Mancera, N.P. Todorov, Assessment of multiple binding modes
in ligandeprotein docking, J. Med. Chem. 47 (2004) 3334e3337.
[39] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Klein,
Comparison of simple potential functions for simulating liquid water, J. Chem.
Phys. 79 (1983) 926e935.
[40] D.A. Gibson, E.A. Carter, Time-reversible multiple time scale ab initio molec-
ular dynamics, J. Phys. Chem. 97 (1993) 13429e13434.
[41] J.P. Ryckaert, G. Ciccotti, H.J.C. Berendsen, Numerical integration of the
[9] K.L. Kirk, Fluorine in medicinal chemistry: recent therapeutic applications of
fluorinated small molecules, J. Fluorine Chem. 127 (2006) 1013e1029.
[10] D.E. Edmondson, C. Binda, J. Wang, A.K. Upadhyay, A. Mattevi, Molecular and
mechanistic properties of the membrane-bound mitochondrial monoamine
oxidases, Biochem. (Mosc.) 48 (2009) 4220e4230.
[11] C. Binda, J. Wang, L. Pisani, C. Caccia, A. Carotti, P. Salvati, D.E. Edmondson,
A. Mattevi, Structures of human monoamine oxidase
B complexes with
selective noncovalent inhibitors: safinamide and coumarin analogs, J. Med.
Chem. 50 (2007) 5848e5852.
cartesian equations of motion of
a system with constraints: molecular
dynamics of n-alkanes, J. Comput. Phys. 23 (1977) 327e341.
[42] T. Darden, D. York, L. Pedersen, Particle mesh Ewald: an N.log(N) method for
Ewald sums in large systems, J. Chem. Phys. 98 (1993) 10089e10092.
[43] W. Humphrey, A. Dalke, K. Schulten, VMD: visual molecular dynamics, J. Mol.
Graph. 14 (1996) 33e38, 27e38.
[44] Desmond Molecular Dynamics System, D.E Shaw Research, New York, NY,
2008.
[45] D.E. Edmondson, C. Binda, A. Mattevi, Structural insights into the mechanism
of amine oxidation by monoamine oxidases A and B, Arch. Biochem. Biophys.
464 (2007) 269e276.
[12] J. Kirchmair, S. Distinto, D. Schuster, G. Spitzer, T. Langer, G. Wolber,
Enhancing drug discovery through in silico screening: strategies to increase
true positives retrieval rates, Curr. Med. Chem. 15 (2008) 2040e2053.
[13] W.J. Geldenhuys, A.S. Darvesh, M.O. Funk, C.J. Van der Schyf, R.T. Carroll,
Identification of novel monoamine oxidase B inhibitors by structure-based
virtual screening, Bioorg. Med. Chem. Lett. 20 (2010) 5295e5298.
[14] F. Chimenti, D. Secci, A. Bolasco, P. Chimenti, A. Granese, S. Carradori,
E. Maccioni, M.C. Cardia, M. Yanez, F. Orallo, S. Alcaro, F. Ortuso, R. Cirilli,
R. Ferretti, S. Distinto, J. Kirchmair, T. Langer, Synthesis, semipreparative HPLC
separation, biological evaluation, and 3D-QSAR of hydrazothiazole derivatives
as human monoamine oxidase B inhibitors, Bioorg. Med. Chem. 18 (2010)
5063e5070.
[46] C. Binda, M. Li, F. Hubálek, N. Restelli, D.E. Edmondson, A. Mattevi, Insights
into the mode of inhibition of human mitochondrial monoamine oxidase B

S. Distinto et al. / European Journal of Medicinal Chemistry 48 (2012) 284e295
295
from high-resolution crystal structures, Proc. Natl. Acad. Sci. U. S. A. 100
(2003) 9750e9755.
[50] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and
computational approaches to estimate solubility and permeability in drug
discovery and development settings, Adv. Drug Delivery Rev. 46 (2001) 3e26.
[51] W.L. Jorgensen, E.M. Duffy, Prediction of drug solubility from structure, Adv.
Drug Delivery Rev. 54 (2002) 355e366.
[52] F. Chimenti, E. Maccioni, D. Secci, A. Bolasco, P. Chimenti, A. Granese,
S. Carradori, S. Alcaro, F. Ortuso, M. Yáñez, F. Orallo, R. Cirilli, R. Ferretti, F. La
Torre, Synthesis, stereochemical identification, and selective inhibitory
activity against human monoamine oxidase-B of 2-methylcyclohexylidene-
(4-arylthiazol-2-yl)hydrazones, J. Med. Chem. 51 (2008) 4874e4880.
[47] C. Binda, F. Hubálek, M. Li, D.E. Edmondson, A. Mattevi, Crystal structure of
human monoamine oxidase B, a drug target enzyme monotopically inserted
into the mitochondrial outer membrane, FEBS Lett. 564 (2004) 225e228.
[48] D.E. Edmondson, A. Mattevi, C. Binda, M. Li, F. Hubalek, Structure and
mechanism of monoamine oxidase, Curr. Med. Chem. 11 (2004) 1983e1993.
[49] G. Wolber, T. Langer, LigandScout: 3-D pharmacophores derived from
protein-bound ligands and their use as virtual screening filters, J. Chem. Inf.
Model. 45 (2005) 160e169.