Quantitative structure-activity relationship and complex network approach to monoamine oxidase A and B inhibitors

Article abstract of DOI:10.1021/jm800656v

The work provides a new model for the prediction of the MAO-A and -B inhibitor activity by the use of combined complex networks and QSAR methodologies. On the basis of the obtained model, we prepared and assayed 33 coumarin derivatives, and the theoretical prediction was compared with the experimental activity data. The model correctly predicted 27 compounds, and most of the active derivatives showed IC₅₀ values in the μM-nM range against both the MAO-A and MAO-B isoforms. Compound 14 shows the same MAO-A inhibitory activity (IC₅₀ = 7.2 nM), as clorgyline used as a reference inhibitor and has the highest MAO-A specificity (1000-fold higher compared to MAO-B). On the other hand, compounds 24 (IC₅₀ = 1.2 nM) and 28 (IC₅₀ = 1.5 nM) show higher activity than selegiline (IC ₅₀ = 19.6 nM) and high MAO-B selectivity with 100-fold and 1600-fold inhibition levels, with respect to the MAO-A isoform.

Full text of DOI:10.1021/jm800656v

6740
J. Med. Chem. 2008, 51, 6740–6751
Quantitative Structure-Activity Relationship and Complex Network Approach to Monoamine
Oxidase A and B Inhibitors
Lourdes Santana,*,† Humberto Gonza´lez-D´ıaz,† El´ıas Quezada,† Eugenio Uriarte,† Matilde Ya´n˜ez,‡ Dolores Vin˜a,‡ and
Francisco Orallo‡
Department of Organic Chemistry, Department of Pharmacology, Faculty of Pharmacy, UniVersity of Santiago de Compostela 15782, Spain
ReceiVed June 1, 2008
The work provides a new model for the prediction of the MAO-A and -B inhibitor activity by the use of
combined complex networks and QSAR methodologies. On the basis of the obtained model, we prepared
and assayed 33 coumarin derivatives, and the theoretical prediction was compared with the experimental
activity data. The model correctly predicted 27 compounds, and most of the active derivatives showed IC50
values in the µM-nM range against both the MAO-A and MAO-B isoforms. Compound 14 shows the
same MAO-A inhibitory activity (IC50 ) 7.2 nM), as clorgyline used as a reference inhibitor and has the
highest MAO-A specificity (1000-fold higher compared to MAO-B). On the other hand, compounds 24
(IC50 ) 1.2 nM) and 28 (IC50 ) 1.5 nM) show higher activity than selegiline (IC50 ) 19.6 nM) and high
MAO-B selectivity with 100-fold and 1600-fold inhibition levels, with respect to the MAO-A isoform.
Introduction
selective and reversible drugs and facilitated the computer-
assisted development of more selective inhibitors. Molecular
docking calculations,17,18 as well as comparative molecular field
analysis (CoMFA)19,20 and 3D quantitative structure-activity
relationships (3D-QSAR),21,22 are among the computational
methods that have been used to predict MAOIs. Moreover, the
estimation of docking parameters for lead drug candidates may
involve time-consuming operations in which the interactions
of large libraries of chemicals with different isoenzymes must
be calculated. In the case of CoMFA, a critical step is the
positioning and subsequent alignment of the molecules, a
requirement that limits its applications to homogeneous series
of compounds. Finally, 3D-QSAR methods also involve the
time-consuming optimization of molecular geometry. Moreover,
to the best of our knowledge, almost all models reported to
predict MAOIs apply only to one series of compounds or ignore
aspects such as the isoenzyme selectivity or the enzymatic source
used in the experimental protocols (animal species and organs
or tissues). As a result, for large and complex databases, the
use of a combined QSAR and complex network approach can
be considered as the most appropriate methodology in the search
for new selective MAO inhibitors.
Mono amine oxidases (MAOs)a are flavoenzymes bound to
the outer mitochondrial membrane and are responsible for the
oxidative deamination of neurotransmitters and dietary amines.1,2
Two isoforms, namely MAO-A and MAO-B, have been
identified on the basis of their amino acid sequences, three-
dimensional structure, substrate preference, and inhibitor
selectivity.3,4 MAO-A has a higher affinity for serotonin and
noradrenaline, whereas MAO-B preferentially deaminates phe-
nylethylamine and benzylamine. These properties determine the
clinical importance of MAO inhibitors. Selective MAO-A
inhibitors such as clorgyline (irreversible) and moclobemide
(reversible) are used in the treatment of neurological disorders
such as depression,5,6 whereas the selective and irreversible
MAO-B inhibitors such as selegiline and rasagiline are useful
in the treatment of Parkinson’s7,8 and Alzheimer’s diseases.9,10
All of these aspects have led to an intensive search for novel
MAO inhibitors (MAOIs), and this effort has increased con-
siderably in recent years. However, earlier MAOIs introduced
into clinical practice were abandoned due to adverse effects,
such as hepatotoxicity, orthostatic hypotension, and the so-called
“cheese effect”, which was characterized by hypertensive
crisis.11,12
Almost all QSAR techniques are based on the use of
molecular descriptors, which are numerical series’ that codify
useful chemical information and enable correlations to be made
between statistical and biological properties.23-26 The principal
deficiency in the use of some molecular indices concerns their
lack of physical meaning. In this respect, the introduction of
novel molecular indices must obey physicochemical laws in
order to ensure a theoretically rigorous interpretation of the
results.24
In recent years, a broad consensus has been reached concern-
ing the necessity for a search for novel MAOIs and the study
of their interaction with MAOs.13-15 The major breakthrough
has been brought about by the crystallization of hMAO-B with
different inhibitors.16 This fact explains the subsequent elucida-
tion and determination of the 3D structure of the active site of
hMAO-A3 and opened new possibilities for the design of more
In the particular case of MAOIs, electron delocalization and
its consequences, such as polarizability and hydrophobicity, may
be a determinant physical factor. This issue can be explained
by considering that two models have been proposed for MAO
catalysis, and these involve an electron transfer or a polar
nucleophilic mechanism.13,16,27 Our research group has just
introduced a novel series of stochastic indices in the so-called
Markov chains invariants for simulation and design (MARCH-
INSIDE) approach. The method is based on the use of Markov
* To whom correspondence should be addressed. Phone: +34-981-
563100. Fax: +34-981 594912. E-mail: qolsant@usc.es.
†
Department of Organic Chemistry, Faculty of Pharmacy, University
of Santiago de Compostela.
‡
Department of Pharmacology, Faculty of Pharmacy, University of
Santiago de Compostela.
a Abbreviations: QSAR, quantitative structure-activity relationship; CN,
complex networks; MAO, monoamine oxidase; MARCH-INSIDE, Markov
chains invariants for simulation and design; MM, Markov models; PCA,
principal components analysis; LDA, linear discriminant analysis; LOO,
leave-one-out.
10.1021/jm800656v CCC: $40.75
2008 American Chemical Society
Published on Web 10/04/2008

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6741
models (MM)28 to codify useful chemical structure information
in terms of molecular electron delocalization, polarizability,
refractivity, and n-octanol/water partition coefficient.29-31
In previous work, we developed a MARCH-INSIDE meth-
odology to seek a QSAR for MAO inhibitors.32 This QSAR
model was then used to predict the in vitro inhibitory MAO-A
activity of a series of new coumarin derivatives. Most of the
compounds were confirmed to be active, and one of these
compounds (18) proved to have a very high activity (submi-
cromolar range).
On the basis of the above information, in the work described
here, we developed a new methodology in which we combined
complex networks (CN) analysis with a QSAR methodology
to carry out a unified analysis on a very large database. The
work provides a method for the prediction of the MAO-A and/
or -B inhibitory activity. The data set compiled and analyzed is
possibly one of the largest and most up-to-date and includes ex
vivo pharmacological activity information for heterogeneous
series of compounds measured in cellular lines extracted from
different organs (liver, brain, placenta, and platelets) of human,
mouse, rat, baboon, as well as human in vitro data. Experimental
conditions related to the isoenzyme (MAO-A or -B), the animal
species and the organ were encoded with dummy variables. A
very good MARCH-INSIDE-QSAR model was obtained, and
the subsequent combined QSAR-CN analysis may become of
major importance for the prediction of the MAO-A/B activity
of new compounds.
This method is based on the calculation of the different
physicochemical molecular properties as an average of atomic
properties (ap). For instance, it is also possible to derive average
estimations of molecular or group electronegativities (kꢀ), molar
refractivities (kMR), polarizabilities (kR), water/n-octanol parti-
tion coefficients (logkP), and atomic van der Waals volumes
(kVvdw).
It is possible to consider isolated atoms (k ) 0) in the
estimation of the molecular properties. In this case, the prob-
abilities 0p(apj) are determined without considering the formation
of chemical bonds (simple additive scheme). However, it is
possible to consider the gradual effects of the neighboring atoms
at different distances in the molecular backbone. To achieve
this goal, the method uses a MM, which determines the absolute
k
probabilities p(apj) with which the atoms placed at different
distances k affect the contribution of the atom j to the molecular
property in question.
Finally, it is interesting to note that one can sum only the
atoms included in a specific group (G) rather than all atoms. In
this way we can approach specific classes of average properties
such as average electronegativity for sp3 carbon atoms (Csp3
)
or average polarizability for heteroatoms (Het). The number of
chemical groups used was four and includes the following:
heteroatoms (Het), heteroatom-bonded hydrogens (H-Het),
saturated carbon atoms (CSp3), unsaturated carbon atoms
(CSp&Sp2). The number of topological distances k between atoms
considered for calculations was six, including 0, 1, 2, 3, 4, and
5. In total, we calculated a number of molecular descriptors
per molecule 5 × (4 + 1) × 6 ) 150, the 1 in the formula
accounts for the molecule as a whole. This means that we
represented the molecular structure of each molecule by a vector
with 150 components (global and local molecular descriptors).
Bearing in mind our experience in the field of coumarin
compounds, the model was subsequently used to predict the
MAO-A/B inhibitory activity of a designed library of coumarin
derivatives, which were prepared and experimentally assayed
as in vitro inhibitors using recombinant human MAO-A and
MAO-B.
2. The biological data and activity assay conditions were also
considered. The dummy or coding variables assign two possible
values (e.g., 1 and 0) when the case presents a certain condition.
We used dummy variables either as inputs or as outputs of the
model. The variable MAO (A/B)i-score is the output of the
model, i.e., the variable to be predicted by the QSAR. This
variable is 1 when a compound was experimentally confirmed
as an inhibitor of the corresponding enzyme for a given organ
and species and -1 otherwise. We also used dummy variables
as inputs of the model to indicate precisely the conditions of
the experiment (animal species or organ of origin). The input
dummy variables used were MAO enzyme (A ) 1, B ) 0),
liver, brain, placenta, platelets, mouse, human, baboon, and
bovine in this order from left to right in the data source file
(see Supporting Information file). In closing, a case in the data
is represented by 9 conditions + 150 descriptors ) 159 input
values and 1 output value.
To reduce the dimensionality of the problem, we conducted
a PCA analysis. We calculated up to 10 principal components
or factors (Fs), which were later used as inputs for the QSAR
analysis. PCA and classification scores allowed us to select at
random the training and validation series’ of compounds. This
group contains chemicals with IC50 values e25 µM or, in some
cases, compounds that show an inhibition percentage of g50%
at inhibitor concentrations e25 µM. This series is composed at
random by the most representative families of MAO-A/B
inhibitors taken from the literature (Supporting Information);
these include propargyl derivatives, benzamides, phenylethy-
lamines, indoles, coumarins, thioxanthenes, oxadiazolidones, and
diazoheterocyclic compounds. The remaining compounds were
a heterogeneous series of inactive compounds including mem-
bers of the aforementioned families, with IC50 > 25 µM, along
Results and Discussion
QSAR-CN Approach to MAO-A and/or -B Inhibitors. Our
main focus in the QSAR study performed was to consider the
differences in the inhibitory properties with MAOs from
different animal species’ and organs.33,34 Hence, the develop-
ment of a discriminant function35 for the classification of organic
compounds as MAO-A and/or MAO-B inhibitors (2 isoen-
zymes) in four animal species (human, bovine, mouse, and
baboon) and using enzymes from four different origins (liver,
brain, placenta, and platelets) is the key step in the present
approach. We a selected a data set of 3408 cases. A case is
defined as the activity class (inhibitor/nonactive) of one
compound as an inhibitor of MAO-A and/or -B. As mentioned
above, a compound may generate a number of statistical cases
per drug (nscpd) equal to nscpd ) (number of enzymes) ×
(number of animal species) × (number of organ of origins) )
2 × (4) × (4) ) 32. Consequently, a compound with inhibitory
activity reported for only one of the two enzymes, one of the
animal species, and one of the organ targets generates only one
statistical case (as found in almost all QSAR studies). In contrast,
a compound with activity previously reported either for one,
more than one, or none of the enzymes, species, and organs
generates, respectively, zero or a number of statistical cases
higher than 2 and lower than 32.
The structure/activity assay conditions (enzyme, animal
species, enzyme organ of origin) and biological activity results
for all cases were then encoded with continuous or dummy
variables.
1. The chemical structures of compounds were encoded with
continuous variables (global or local molecular descriptors) that
were calculated using the MARCH-INSIDE approach.31

6742 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21
Santana et al.
Figure 1. PCA plot for training and validation cases. Note that almost all cases lie within the principal cloud of points; note also the main axes
that define the four quadrants.
Table 1. Training and Validation Results for the MAO Inhibitors
QSAR
with many other structural patterns extracted from international
databases.36,37 It is common for compounds with even higher
%
nonactive
Train
MAO inhibitors
IC50 values to be considered as active or moderately active, but
we considered 25 µM to be a reasonable limit. The selection of
higher break point values to cluster chemicals by their MAO-
A/B IC50 may generate a series with a clearly disproportionate
size and, therefore, a vastly reduced number of active com-
pounds. The selection here of discriminant techniques instead
of regression techniques was determined by the lack of
homogeneity in the conditions under which these values were
measured. As reported in different sources, numerous IC50 values
lie within a range rather than a single value. In other cases, the
activity is not scored in terms of IC50 values but is quoted as
inhibitory percentages at a given concentration.
The PCA scores for all cases were depicted in a plot of the
first factor F1 versus F2. These factors are the most important,
and together they explain a cumulative variance of F1 (explained
variance) + F2 (explained variance) ) 43.17 + 18.52 ) 61.69%
of the total variance. For training and validation, we selected
at random compounds that lie within the four quadrants of the
cloud of points. The overall distribution of compounds in this
plot is illustrated in Figure 1. Once the training series had been
designed, forward stepwise linear discriminant analysis (LDA)
was carried out in order to derive the QSAR for the MAO-A/
MAO-B inhibitory activity score MAO(A,B)i:
90.8
98.7
94.3
nonactive
MAO inhibitors
total
1300
15
131
1119
Validation
92.6
98.6
95.3
nonactive
MAO inhibitors
total
441
5
35
362
Both
91.3
98.7
94.5
nonactive
MAO inhibitors
total
1741
20
166
1481
series. Details of this information are provided in Table 1. In
addition to these results, the linear relationship between leave-
one-out (LOO) residuals and raw residuals are shown in Figure
2, and these also illustrate the high stability of the model to
data variation.
The results strongly validate the model as a predictor of
appropriate interactions of a selected compound as an MAO-A
and/or MAO-B inhibitor. The model is able to predict for a
new compound one probability value for each of the possible
outputs including assays in cellular lines extracted from different
organs (liver, brain, placenta, and platelets) of human, mouse,
rat, and baboon species. The results for a selected group of
compounds are given in Table 2. For detailed information on
all compounds used, see Tables 1SI, 2SI, and 3SI in the online
Supporting Information.
MAO(A, B)i - score ) 0.14 · F1 + 0.56 · F2 + 0.40 · F3 - 0.02
Rc ) 0.79 U ) 0.38 p < 0.001
(1)
The statistical significance of this model was determined by
examining the canonical regression coefficient (Rc), the U-
statistic (which is also known as Wilk’s λ statistic), and the
p-level (p). We also inspected the percentage of good clas-
sification in training and validation experiments. Overall, the
model correctly classified 3222 out of 3408 inputs (94.5%).
This includes the correct prediction of 1481 out of 1501
results for experimental assays of MAO-A and/or MAO-B
inhibitors (98.7%) and 1741 out of 1907 nonactive compounds.
Briefly, training classification and external validation rates were
above 90% for all positive/negative groups in training/validation
The values of the PCA scores for each compound weighted
with the LDA-QSAR model coefficients were used to calculate
a very large matrix of compound-compound pairwise Euclidean
distances. After selection of a cutoff value, this matrix was
transformed into a Boolean (0, 1) matrix. The matrix can be
represented as a very large graph or network that expresses the
similarity/dissimilarity relationships of very heterogeneous

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6743
Figure 2. Model robustness to LOO variation of the data.
Table 2. Prediction of the Interaction with MAO-A and/or MAO-B Isoenzymes for a Selected Group of Compoundsa
f
f
f
compds
ref Nb
OCc
PCc
Pd
t/cve
F1
F2
F3
MAO
organ
liver
placenta
brain
species
isoquinolin42
isoquinolin42
oxazolinon25
oxazolinon25
quinaz40
quinaz40
tetrazole27
tetrazole27
51
51
57
57
82
82
37
37
25
25
2j
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.53
0.54
0.58
0.59
0.95
0.95
0.89
0.89
t
t
t
cv
cv
cv
t
9.17
9.26
3.12
-2.58
-2.48
-2.03
-2.02
4.70
4.71
2.84
2.84
0.13
0.00
2.04
2.04
1.65
1.66
1.07
1.08
B
A
A
B
A
B
A
B
human
human
bovine
bovine
human
human
mouse
mouse
2j
3.11
brain
37
37
17g
17g
-3.34
-3.35
-1.08
-1.08
platelets
platelets
brain
cv
brain
Nonactive Compounds
amphetamin30
amphetamin30
cypropamine9
cypropamine9
indol4
indol4
indol16
indol16
proparg38
proparg38
4
4
17
17
3a
4a
6
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
0.22
0.22
0.13
0.13
0.33
0.34
0.09
0.09
0.10
0.11
cv
t
cv
t
t
t
t
t
t
t
6.47
6.47
6.57
6.57
8.71
8.70
4.17
4.17
3.54
3.54
-1.41
-1.40
-3.37
-3.36
1.37
-4.06
-4.05
-2.97
-2.96
-7.27
-7.24
-5.54
-5.53
4.05
A
B
A
B
A
B
A
B
A
B
brain
brain
liver
mouse
mouse
human
human
human
human
bovine
bovine
mouse
mouse
43
43
46
46
48
48
64
64
liver
placenta
platelets
brain
brain
liver
6
1.40
4a
4a
15
15
-1.52
-1.52
-8.03
-8.02
4.06
liver
a See Tables 1SI, 2SI, and 3SI of Supporting Information for detailed information of all compounds. b Number of each compound in the corresponding
reference (Supporting Information). c Observed classification (OC) and predicted classification (PC); 1 for active, -1 for nonactive. d Subsequent probability
of a positive activity predicted for each compound. e Train and cross-validation indicator, t for compounds in training and cv for compounds in the external
validation series. f Factor scores (F1 to F3) for each compound.
classes of MAO-A and/or -B inhibitors in different mammalian
species. This complex network is represented in Figure 3.
The topological parameters of the present network are very
similar to the more typical free scale topology parameters. A
number of interesting parameters to characterize the topology
of this network and to compare it to a free scale topology are
shown in Table 3. It can be seen that despite differences in some
topological parameters, the network developed here perfectly
fits to a power law distribution with a regression coefficient
higher than 0.9. This kind of topology is characteristic of a
network with a group of drugs called hubs (very connected
nodes) that is representative of the vast universe of drugs
considered. As a consequence, our network will very probably
associate every new compound to one of these hubs. This
combined QSAR-CN analysis may therefore become of major
importance for the prediction of the binding of new compounds
that could help to explain the nature of interactions. The
clustering of the MAO-B inhibitors N-[4-[(3-fluorophenyl)-
methoxy]phenyl]-5-oxo-3-pyrrolidinyl acetamide (pyrrolidone40)
and N-[4-(3-fluoro-benzyloxy)-phenyl]-2-oxo-3-pyrrolidinylac-
etamide (pyrrolidone70)38,39 using CentiBin software is repre-
sented in Figure 4. These compounds are similar only to 1 and
3 other drugs (see node degree centrality in the right-hand
column of the figure), respectively, but pyrrolidone70 is similar
to 2-fluoro-N-[4-[(3-fluorophenyl)methoxy]phenyl]-5-oxo-3-pyr-
rolidinylacetamide (pyrrolidone43) and N-[-5-oxo-4-[(2,4,6-
trifluorohenyl)methoxy]henyl]-3-yrrolidinylacetamide (pyrroli-
done52), which are typical pyrrolidone scaffold MAO-B
inhibitors.38 Both pyrrolidone43 and pyrrolidone52 are similar
to more than 10 compounds (considering only the compounds
depicted in this subnetwork). These latter two compounds are
therefore candidates to be considered as drug-hubs that are
characteristics of this class.
Rational Design and in Silico Evaluation of Novel
Series of MAO-A and/or MAO-B Inhibitors. The model
developed in this study was used for the design of novel MAO-

6744 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21
Santana et al.
Figure 3. Free Kamada-Kawai visualization of the similarity/dissimilarity network for MAO-A and/or MAO-B inhibitors (I) and a free scale
random network (II) for visual comparison purposes.
A/B inhibitors. As described previously, a specific way to new
lead discovery involves several general common steps when
used in terms of a QSAR: (a) construction of a suitable
molecular database of compounds that either show the property
in question or do not, (b) calculation of the molecular descrip-
tors, (c) construction of the model, (d) estimation of the
biological activity using QSAR, (e) synthesis and characteriza-
tion of selected compounds, and (f) assay of the candidate
compounds in order to corroborate the predicted biological
activity.
The coumarin analogues are a family of natural and/or
synthetic compounds with different pharmacological activities,
one of which is MAO inhibitory activity.40,41 In many cases, it
is known that activity and selectivity are determined by the
nature of the substituents at the 7- and the 4/3-positions.42-44
On the other hand, in a previous investigation,32 we developed
QSAR studies for the MAO-A inhibitory activity in a series of
coumarin derivatives and found that the most active compounds
were the 7-acetonyloxy-substituted compounds, which showed
higher activity than their cyclic analogues and were also more
active than the 7-hydroxy precursors.
interesting 7-(ꢁ-ketoether)coumarin derivatives. The compounds
were synthesized according to Scheme 1, and details are given
in the Experimental Section.
Chemistry. The coumarin derivatives 8, 9, 11, 13-15, 21, 22,
24-29, 32, and 33 were efficiently synthesized according to the
synthetic protocol outlined in Scheme 1. The remaining compounds
were prepared as reported in our previous communications.32,45-47
Pechmann condensation of 2-alkyl resorcinol with the cor-
responding ketoester afforded the 3,4-cyclopentene/cyclohexene-
7-hydroxy coumarins 8 and 9. The Williamson reaction of the
7-hydroxycoumarins 1-3, 8, and 9 with 2-chloroketones or 2,3-
dibromopropene gave the corresponding ethers 11, 13-15, 21,
22, 24, and 25. Compounds 22, 23, and 25 were oxidized with
DDQ to give the corresponding 3,4-benzocoumarin derivatives
26-28. Finally, cyclization of coumarin ketoethers 11, 14, and
15 in strongly alkaline solution led to the furanocoumarins 29,
32, and 33.
MAO Inhibition Assay. The potential effects of the test drugs
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
10-acetyl-3,7-dihydroxyphenoxazine as reagent 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.
On the basis of the above information, and with the aim of
exploring the structure-affinity and MAO-A/B selectivity rela-
tionships, in the present work, we designed a series of 33
coumarin derivatives (Table 4) that include the aforementioned
chemical diversity with particular attention paid to the most

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6745
Table 3. Comparative Study of MAO-A and/or -B Inhibitors Network
and Free-Scale Topology
7-hydroxy derivatives have been shown to be inactive, even
with the introduction of bulky substituents in any other positions
of the coumarin (compounds 1-10). The introduction of
acetonyl/bromoallyloxy groups in the 7-position of the coumarin
yielded compounds (11-28) with greater MAO-A and MAO-B
inhibitory activity. The cyclization of this group or the formation
of the furocoumarin analogues (29-33) again resulted in a
reduction of both MAO-A and MAO-B iso-enzyme inhibitory
activities with respect to the acyclic analogues. Finally, the most
relevant findings are that: (a) the introduction of bulky groups
in the 7-acetonyl substituent increase MAO-A selectivity
(13-19), (b) introduction of bulky groups such as cyclohexyl
or phenyl in the 3,4-positions of the 7-acetonyl derivatives
increased both MAO-A and MAO-B inhibitory activities with
concomitant loss of selectivity, whereas replacement of the
acetonyl substituent at position 7 by the bromoallyloxy group
resulted in compounds 24 and 28, which had very high MAO-B
inhibitory activity (IC50 of about 1.2 and 1.5 nM) and the highest
MAO-B selectivity (approximately 100-fold and 1600-fold) with
respect to the MAO-A isoform.
In summary, we have developed good theoretical QSAR-CN
models to predict inhibitory MAO activity. On the basis of this
model, we have found new coumarin derivatives with inhibitory
activity comparable to those of clorgyline and selegiline,
respectively, which are used as reference inhibitors and have a
very high MAO-A and MAO-B selectivity. These findings have
encouraged us to continue our investigations into the design of
more potent and selective analogues by introducing appropriate
substituents into the coumarin scaffold.
networks
MAOi (I) scale free (II)
Classic Topological Properties
number of nodes (drugs): n
1174
3873
1174
2921
number of arcs (similar drug-drug pairs): m
average node degree (similar drugs)
average drug-drug Euclidean distance
average drug-drug topological distance
max m possible: maxm ) n!/(n - 2)!·2!
edge densitya
2.6
2.7
5.0
4.2
688551
0.4
688551
0.6
Free-Scale Typical Properties for Semilog10 Plot of Distribution
Histogram
free-scale regression coefficient: R
average value
standard deviation
0.95
1.24
0.02
0.99
1.55
0.36
Other Important Topological Properties
Zagreb group index 1: M1
Zagreb group index 2: M2
Randic connectivity index: Xr
Platt index: F
99452
938872
583.5
91706
1.99
60328
352307
435.9
54486
1.94
index of relinking: P
a The drug-to-drug average Euclidean distance could not be determined
for the scale free network because the generation of this network with Pajek
omits the Euclidean distance step. Average topological distance could not
be calculated for MAOi-QSAR network with CentiBin due to loss of
connectivity; however, edge density I is slightly higher than edge density
II. Consequently, we can conclude that the average distance for network I
is only slightly lower than the average distance of the free scale network.
The production of H2O2 catalyzed by MAO isoforms can be
detected using the previously mentioned reagent, a nonfluores-
cent, highly sensitive, and stable probe that reacts with H2O2
in the presence of horseradish peroxidase to produce a fluores-
cent product, resorufin. In this study, hMAO activity was
evaluated using the above method following the general
procedure described previously by us.48
The tested drugs (new compounds and reference inhibitors)
inhibited the control enzymatic MAO activities and the inhibi-
tion was concentration dependent. The corresponding IC50 values
and MAO-B selectivity ratios [IC50 (MAO-A)]/[IC50 (MAO-
B)] are shown in Table 5.
The assayed compounds themselves do not react directly with
the 10-acetyl-3,7-dihydroxyphenoxazine, which indicates that
these drugs do not interfere with the measurements.
In our experiments and under our experimental conditions,
hMAO-A displayed a Michaelis constant (Km) of 457.17 (
38.62 µM and a maximum reaction velocity (Vmax) of 185.67
( 12.06 nanomol/min/mg protein, whereas hMAO-B showed
a Km of 220.33 ( 32.80 µM and a Vmax of 24.32 ( 1.97
nanomol/min/mg protein (n ) 5).
Experimental Section
Computational Methods Used for QSAR-CN Analysis. The
calculation of the molecular descriptors was implemented in the
in-house software MARCH-INSIDE.49
To perform the multispecies activity analysis with a CN approach
we carried out the following steps:
1. We calculated the weighted principal component loadings
(multiplying the coefficients of the QSAR equation by the factor
value included in the QSAR equation) for a large number of selected
drugs; in the calculations, we included MAO-A and/or -B inhibitors
assayed for at least one of the animal species and organs of origin
studied.
2. All of the loadings were used as inputs for the STATISTICA
software50 employed to calculate drug-drug multispecies dis-
similarities in the form of weighted Euclidean distances (wEd). The
values can be arranged in the form of a Euclidean distance matrix
of drugs vs STATISTICA drugs. Each drug may appear more than
once depending on the different assays reported for the particular
drug.
3. Microsoft Excel was used to transform the drug pair distances
matrix derived with Statistica into a Boolean matrix. The elements
of this matrix are equal to 1 if two drugs tested under given
conditions have a short wEd, or the same if the QSAR-predicted
results for these drugs in this test are very similar The threshold or
cutoff value used was a distance of 0.005. The line command used
in Excel to transform the distance matrix into a Boolean matrix
was f ) if (A$1)$B2,0, if (B2>0.0051, 0, 1)). This allows the
transformation of distance into Boolean values and makes the main
diagonal elements 0 while avoiding loops in the future network.
The Boolean matrix was saved as a.txt format file.
4. After renaming the.txt file as a.mat file, it was read with the
software CentiBin.51 The software can be used not only to represent
the network but also to highlight all drugs (nodes) connected to a
specific drug and to calculate numerous parameters including node
degree (the number of drugs with similar predicted results).
5. Pajek software52 submenu parameters, which appear in the
bar menu option network information, were used to request the
Conclusions
In this study, we have developed a QSAR-CN analysis for
the design of novel MAO inhibitors using a very large database
of heterogeneous compounds and their activities measured in
cellular lines from different organs and species. The model
correctly classified 3222 out of 3408 inputs (94.5%) and was
then used to predict the activity of 33 compounds in our
designed coumarin library. All of the active compounds were
correctly predicted by the model, and a total of 27 (81.82%)
and 24 (72.73%) compounds were correctly predicted as
MAO-A and MAO-B inhibitors, respectively.
On the other hand, the SAR analysis seems to corroborate
the importance of the structural requirements of the substituents
in the coumarin moiety in order to achieve good activity and
MAO-A/B selectivity, as indicated in the literature.42-44 The

6746 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21
Santana et al.
Figure 4. View of a subnetwork corresponding to a cluster of MAO-B inhibitors that exemplify the identification of a compound with a pyrrolidine
scaffold with other similar analogues.
calculation of several parameters that describe network topology
including: Zagreb group M1 and M2 indices, Platt index F, Randic
connectivity index Xr, and index of relinking. We used these
parameters to compare the topology of the network constructed by
QSAR for MAO inhibitors with a free scale random network. We
generated this random network with Pajek, containing the same
number of nodes (drugs) as our network.
Chemistry. Melting points were determined using a Reichert
Kofler thermopan or in capillary tubes on a Bu¨chi 510 apparatus
and are uncorrected. IR spectra were recorded on a Perkin-Elmer
1640FT spectrophotometer. 1H and 13C NMR spectra were recorded
on a Bruker AMX spectrometer at 300 and 75.47 MHz, respectively,
using TMS as internal standard (chemical shifts in δ values, J in
Hz). Mass spectra were obtained using a Hewlett-Packard 5988A
spectrometer. Elemental analyses were performed using a Perkin-
Elmer 240B microanalyzer and were within ( 0.4% of calculated
values in all cases. Silica gel (Merck 60, 230-00 mesh) was used
for flash chromatography (FC). Analytical thin layer chromatog-
raphy (TLC) was performed on plates precoated with silica gel
(Merck 60 F254, 0.25 mm).
was recovered by filtration and washed with water to yield the
desired compound, which was purified by crystallization.
8-Methyl-3,4-cyclopentene-7-hydroxycoumarin (8). Yield 82%;
mp 259 °C (EtOH). 1H NMR (DMSO-d6) 2.07 (m, 2H,
CH2CH2CH2), 2.16 (s, 3H, Me-C8), 2.70 (t, J ) 7.3, 2H), 2.98 (t,
J ) 7.5, 2H), 6.84 (d, J ) 8.4, 1H, H-6), 7.24 (d, J ) 8.4, 1H,
H-5), 10.30 (s, 1H, OH). 13C NMR (DMSO-d6) 8.21 (Me-C8),
22.03 (CH2CH2CH2), 29.88, 31.60, 110.57 (C8), 110.63, 111.66
(C6), 121.89, 123.00 (C5), 153.35, 157.02, 158.25, 159.46 (C2).
IR 3362, 2920, 1682, 1577, 1275, 1081, 812. MS m/z 217 ([M +
1]+, 14), 216 (M+, 100), 188 (58), 187 (65), 145 (6), 128 (3). Anal.
(C13H12O3) C, H.
8-Methyl-3,4-cyclohexene-7-hydroxycoumarin (9). Yield 71%;
mp 279-280 °C (EtOH). 1H NMR (DMSO-d6) 1.69 (m, 4H,
CH2(CH2)2CH2), 2.11 (s, 3H, Me-C8), 2.35 (m, 2H), 2.64 (m, 2H),
6.79 (d, J ) 8.6, 1H, H-6), 7.30 (d, J ) 8.6, 1H, H-5), 10.16 (s,
1H, OH). 13C NMR (DMSO-d6) 8.30 (Me-C8), 21.29, 21.61, 23.78
(CH2-C4), 24.95, 110.61 (C8), 111.87 (C6), 112.25, 118.27, 121.78
(C5), 148.13, 151.40, 157.94, 161.45 (C2). IR 3275, 2932, 1676,
1606, 1374, 1266, 1102, 804. MS m/z 231 ([M + 1]+, 15), 230
(M+, 100), 215 (47), 202 (33), 174 (60). Anal. (C14H14O3) C, H.
General Procedure for the Preparation of Oxoether Deriva-
tives 11, 13-15, 21, 22, 24, and 25. To a solution of the substituted
7-hydroxycoumarin derivatives 1-3, 8, or 9 (62 mmol) in dry
acetone (300 mL) were added K2CO3 (4 g) and the corresponding
2-chloroketones or 2,3-dibromopropene (120 mmol). The mixture
General Procedure for the Preparation of 3,4-Cycloalkane-7-
hydroxycoumarins (8, 9). 2-Methylresorcinol (80.554 mmol) was
dissolved in 12 M H2SO4 (120 mL). The appropriate ethyl
2-oxocycloalkanecarboxylate (96.65 mmol) was added, and the
solution was stirred at room temperature for 15 h. The reaction
mixture was then poured into ice/water (100 mL), and the precipitate

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6747
Table 4. Structure of the Coumarin Derivatives 1-33
compd
R3
R4
R5
R7/(CH2)n
OH
OH
OH
OH
OH
OH
OH
OH
R8
1
2
3
4
5
6
7
8
H
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
OMe
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
H
H
H
OMe
H
H
H
H
Me
Me
Me
H
OMe
Me
H
OMe
H
OMe
Me
Me
OMe
H
OMe
OMe
Me
Me
H
NH2
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
9
OH
OH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
H
H
H
H
H
H
H
Me
Me
H
MeCOCH2O-
MeCOCH2O-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)3-
-(CH2)4-
MeCOCH2O-
MeCOCH2O-
MeCOCH2O-
MeCOCH2O-
CH2C(Br)CH2-
CH2C(Br)CH2-
MeCOCH2O-
MeCOCH2O-
CH2C(Br)CH2-
NHAc
H
H
H
H
H
H
H
OMe
H
H
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)3-
-(CH2)4-
Ph
OMe
Me
Me
OMe
Me
Me
Me
OMe
Me
H
Ph
Ph
H
H
H
H
H
H
NH2
H
OMe
H
Me
Me
-(CH2)2-
-(CH2)2-
-(CH2)3-
H
H
H
Scheme 1. Synthetic Strategy for Prepared Compounds
was heated under reflux for 24 h. The mixture was cooled, and the
solid residue was filtered off. The solvent was removed under
reduced pressure, and the crude product was purified by FC and/or
crystallization to give the desired compound.
7-Acetonyloxy-5-methoxycoumarin (11). Yield 77%; mp 160 °C.
1H NMR (CDCl3) 2.30 (s, 3H, Me-CO), 3.91 (s, 3H, MeO), 4.63
(s, 2H, CH2O), 6.19 (d, J ) 9.7, 1H, H-3), 6.29 (d, J ) 2.1, 1H,
H-6), 6.38 (d, J ) 2.1, 1H, H-8), 7.97 (d, J ) 9.7, 1H, H-4). 13C

6748 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21
Santana et al.
Table 5. In Vitro and in Silico Evaluation of MAO Inhibitory Activities of Compounds 1-33 and Reference Inhibitorsa
MAO-A
MAO-B
compd
MAO-A IC50
MAO-B IC50
**
**
71.32 ( 3.17 µM
80.41 ( 5.16 µM
**
**
**
**
**
50.20 ( 1.49 µM
44.10 ( 2.80 µM
**
856 ( 19.50 µM
10.60 ( 0.96 µM
3.87 ( 0.15 µM
***
MAO-A/B
obsdf
predictedf
Pe
obsdf
predictedf
Pe
1
2
3
4
5
6
7
8
**
-
-
-
-
-
-
-
-
-
-
-
-
+
+
+
+
+
+
+
-
+
+
+
+
+
+
+
+
+
-
+
-
-
-
+
-
+
-
-
-
-
-
-
-
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
-
0.38
0.22
0.40
0.48
0.37
0.33
0.26
0.29
0.27
0.26
0.83
0.85
0.87
0.87
0.87
0.72
0.72
0.83
0.71
0.68
0.71
0.69
0.66
0.94
0.94
0.99
0.99
0.99
0.91
0.92
0.87
0.68
0.71
0.82
0.33
0.79
0.42
-
-
-
-
-
-
-
-
-
-
-
-
-
+
+
-
-
+
-
+
+
+
+
+
+
-
+
+
-
+
-
-
-
-
+
+
-
-
-
-
-
-
-
-
-
-
-
+
+
+
+
+
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
-
+
+
-
0.38
0.22
0.41
0.49
0.08
0.33
0.26
0.29
0.27
0.26
0.83
0.69
0.87
0.88
0.87
0.42
0.72
0.83
0.80
0.68
0.71
0.69
0.66
0.94
0.94
0.99
0.99
0.99
0.92
0.61
0.87
0.37
0.68
0.77
0.99
0.82
0.92
80.30 ( 5.82 µM
<0.80d
>1.40d
>1.24d
-
**
**
**
48.20 ( 3.61 µM
63.53 ( 3.72 µM
>**
60.11 ( 3.47 µM
**
<0.48d
<0.63d
-
9
<0.60d
>1.99d
>2.27d
<0.53d
0.0031
0.00067
0.10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
**
5.31 ( 0.12 µM
2.66 ( 0.18 µMb
7.16 ( 0.52 nMb
0.40 ( 0.03 µMb
20.35 ( 1.12 µM
15.45 ( 0.56 µM
40.50 ( 2.2 nMb
43.00 ( 2.93 µM
121 ( 7.43 nMb
235 ( 18.52 nMc
3.78 ( 0.34 nMb
396 ( 16.63 nMb
130 ( 11.77 nMb
695 ( 28.54 nMc
2.56 ( 0.14 µMb
11.93 ( 0.61 µM
2.38 ( 0.06 µMb
255 ( 10.50 µM
12.20 ( 0.46 µM
57.92 ( 2.24 µM
32.80 ( 1.61 µMb
**
<0.41d
<0.31d
0.0073
<0.43d
0.0093
1.42
0.22
0.033
110
2.94
0.091
1.27
1,597
0.98
***
5.57 ( 0.32 µM
**
12.98 ( 0.91 µΜ
165 ( 7.04 nΜ
17.3 ( 1.48 µM
12.12 ( 1.03 µM
1.18 ( 0.15 nM
236 ( 5.03 nM
28.13 ( 1.14 µM
9.36 ( 0.26 µM
1.49 ( 0.12 nM
260 ( 12.72 µM
16.53 ( 0.74 µM
**
73.20 ( 4.23 µM
**
61.35 ( 1.13 µM
19.60 ( 0.86 nM
7.54 ( 0.36 µM
*
0.74
<0.58d
0.45
-
clorgyline
selegiline
iproniazide
moclobemide
4.46 ( 0.32 nMb
67.25 ( 1.02 µMb
6.56 ( 0.76 µM
361 ( 19.37 µM
0.00007
3431
0.87
<0.36d
a Each IC50 value is the mean ( SEM from five experiments (n ) 5). * Inactive at 1 mM (highest concentration tested). ** Inactive at 100 µM (highest
concentration tested). At higher concentrations, the compounds precipitate. *** Inactive at 50 µM (highest concentration tested). At higher concentrations
the compounds precipitate. b Level of statistical significance:bP < 0.01 or bP < 0.05 versus the corresponding IC50 values obtained against MAO-B as
determined by ANOVA/Dunnett’s. c Level of statistical significance:bP < 0.01 or bP < 0.05 versus the corresponding IC50 values obtained against MAO-B
as determined by ANOVA/Dunnett’s. d Values obtained under the assumption that the corresponding IC50 against MAO-A/B is the highest concentration
tested. e P Subsequent probability predicted by the model. f Observed and predicted groups for the selected compounds: (+) if the IC50 < 25 µM for the
observed group, and if P > 0.5 for predicted group.
NMR (CDCl3) 26.62 (Me-CO), 56.10 (MeO), 72.84 (CH2O), 93.31
(C8), 95.27 (C6), 104.70, 111.67 (C3), 138.61 (C4), 156.61, 157.24,
161.27, 161.47 (C2), 203.30 (Me-CO). IR 3010, 1728, 1614, 1365,
1122, 1065, 816; MS m/z 249 ([M + 1]+, 14), 248 (M+, 100), 220
(3), 205 (74), 191 (3), 177 (47). Anal. (C13H12O5) C, H.
8-Methoxy-4-methyl-7-(2′-oxocyclopentyloxy)coumarin (13).
Yield 96%; mp 174-176 °C. 1H NMR (CDCl3) 1.88-2.55 (m,
6H (CH2)3), 2.39 (d, J ) 1.1, 3H, Me), 3.97 (s, 3H, MeO), 4.77 (t,
J ) 8.6, 1H, H-1′), 6.17 (d, J ) 1.1, 1H, H-3), 6.99 (d, J ) 8.9,
1H, H-6), 7.27 (d, J ) 8.9, 1H, H-5). 13C NMR (CDCl3) 17.25
(C4′), 18.78 (Me-C4), 29.58 (C5′), 35.23 (C3′), 61.57 (MeO), 80.81
(C1′), 112.39 (C3), 112.87 (C6), 115.86, 119.24 (C5), 137.47,
147.97, 152.50, 153.76 (C4), 160.52 (C2), 213.10 (C2′). IR 2994,
2932, 1755, 1718, 1604, 1300, 1114, 850. MS m/z 289 ([M + 1]+,
15), 288 (M+, 100), 231 (19), 205 (65), 178 (56). Anal. (C16H16O5)
C, H.
4,8-Dimethyl-7-(2′-oxocyclohexyloxy)coumarin (15). Yield 96%;
mp 187-189 °C. H NMR (CDCl3) 1.80-2.70 (m, 8H (CH2)4),
1
2.37 (s, 6H, Me-C4, Me-C8), 4.73 (dd, J ) 5.2, 9.2, 1H, H-1′),
6.14 (d, J ) 1.1, 1H, H-3), 6.62 (d, J ) 8.8, 1H, H-6), 7.33 (d, J
) 8.8, 1H, H-5). 13C NMR (CDCl3) 8.9 (Me-C8), 19.0 (Me-C4),
23.1, 28.1, 34.9, 40.7, 81.5 (C1′), 109.3 (C3), 112.6 (C6), 114.6,
115.4, 122.6 (C5), 152.9, 153.2, 158.7, 161.8 (C2), 207.9 (C2′).
IR 2945, 1727, 1706, 1598, 1283, 1119, 804. MS m/z 287 ([M +
1]+, 7), 286 (M+, 37), 190 (75), 162 (100), 115 (16). Anal.
(C17H18O4) C, H.
7-Acetonyloxy-3,4-cyclopentene-8-methylcoumarin (21). Yield
1
76%; mp 144 °C. H NMR (CDCl3) 2.18 (m, 2H, CH2CH2CH2),
2.34 (s, 3H, MeCO), 2.38 (s, 3H, Me-C8), 2.87 (t, J ) 7.5, 2H),
3.01 (t, J ) 5.9, 2H), 4.63 (s, 2H, CH2O), 6.64 (d, J ) 8.6, 1H,
H-6), 7.21 (d, J ) 8.6, 1H, H-5). 13C NMR (CDCl3) 9.01 (Me-
C8), 22.86 (CH2CH2CH2), 27.10 (MeCO), 30.76, 32.40, 73.74
(CH2O), 107.73 (C6), 113.69, 114.87, 122.98 (C5), 125.29, 153.72,
156.73, 157.98, 160.78 (C2), 205.29 (MeCO). IR 2956, 1733, 1706,
1615, 1376, 1286, 1120, 805. MS m/z 273 ([M + 1]+, 18), 272
(M+, 100), 229 (46), 215 (43), 187 (33), 128 (44). Anal. (C16H16O4)
C, H.
7-Acetonyloxy-3,4-cyclohexene-8-methylcoumarin (22). Yield
79%; mp 161-162 °C. 1H NMR (CDCl3) 1.81 (m, 4H,
CH2(CH2)2CH2), 2.33 (s, 3H, MeCO), 2.37 (s, 3H, Me-C8), 2.54
(m, 2H), 2.72 (m, 2H), 4.62 (s, 2H, CH2O), 6.64 (d, J ) 8.8, 1H,
H-6), 7.34 (d, J ) 8.8, 1H, H-5). 13C NMR (CDCl3) 8.39 (Me-
4,8-Dimethyl-7-(2′-oxocyclopentyloxy)coumarin (14). Yield 90%;
1
mp 187 °C. H NMR (CDCl3) 2.09 (m, 4H, CH(CH2)2), 2.31 (s,
3H, Me-C8), 2.39 (d, J ) 1.1, 3H, Me-C4), 2.47 (m, 2H, H-3′),
4.70 (t, J ) 7.8, 1H, H-1′), 6.14 (d, J ) 1.1, 1H, H-3), 6.94 (d, J
) 8.8, 1H, H-6), 7.39 (d, J ) 8.8, 1H, H-5). 13C NMR (CDCl3)
8.40 (Me-C8), 17.26 (Me-C4), 18.70 (C4′), 29.67 (C5′), 35.24 (C3′),
79.97 (C1′), 109.52 (C3), 112.12 (C6), 114.35 (C8), 115.15, 122.28
(C5), 152.64, 152.74, 158.69, 161.52 (C2), 213.41 (C2′). IR 2922,
1722, 1605, 1288, 1119; MS m/z 273 ([M + 1]+, 16), 272 (M+, 92),
201 (32), 190 (90), 162 (100), 115 (26). Anal. (C16H16O4) C, H.

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6749
C8), 21.43, 21.67, 23.89, 25.21, 26.70 (MeCO), 73.40 (CH2O),
107.18 (C6), 114.24 (C8), 114.79, 120.97, 121.08 (C5), 147.13,
151.22, 157.09, 162.05 (C2), 205.13 (MeCO). IR 2942, 1738, 1698,
1605, 1416, 1292, 1025, 803. MS m/z 287 ([M + 1]+, 38), 286
(M+, 100), 271 (12), 258 (5), 243 (60), 230 (21). Anal. (C17H18O4)
C, H.
MS m/z 347 ([M + 2]+, 2), 345 (M+, 12), 344 (13), 265 (60), 225
(100), 171 (17). Anal. (C17H13BrO3) C, H.
General Procedure for the Preparation of the Furocoumarin
derivatives 29, 32, 33. To a solution of the corresponding ketoether
11, 14, or 15 (3.39 mmol) in ethyl alcohol (200 mL) was added
0.1 M NaOH (200 mL). The mixture was heated under reflux for
12 h, acidified with HCl, and concentrated to half-volume and left
overnight. The resulting precipitate was filtered off and purified
by FC to give the desired compound.
5-Methoxy-4′-methylfuro[3,2-g]coumarin (29). Yield 70%; mp
140 °C, 1H NMR (CDCl3) 2.40 (d, J ) 1.3, 3H, Me-C4′), 3.99 (s,
3H, MeO), 6.34 (d, J ) 9.8, 1H, H-3), 7.21 (s, 1H, H-8), 7.37 (d,
J ) 1.3, 1H, H-5′), 8.06 (d, J ) 9.8, 1H, H-4), 13C NMR (CDCl3)
9.26 (Me-C4′), 64.94 (MeO), 96.58 (C8), 109.20, 113.63 (C3),
114.68, 119.13, 138.57 (C4), 142.52 (C5′), 150.99, 152.35, 158.36,
161.04 (C2), IR 3110, 2945, 1732, 1717, 1634, 1607, 1113, 820,
MS m/z 231 ([M + 1]+, 14), 230 (M+, 100), 215 (78), 187 (49),
159 (25), 131 (8). Anal. (C13H10O4) C, H.
7-(ꢀ-Bromoallyloxy)-3,4-cyclopentene-8-methylcoumarin (24).
Yield 78%; mp 140 °C. 1H NMR (CDCl3) 2.19 (m, 2H,
CH2CH2CH2), 2.38 (s, 3H, Me-C8), 2.90 (t, J ) 7.5, 2H, CH2-C4),
3.03 (t, J ) 7.6, 2H, CH2-C3), 4.72 (s, 2H, CH2O), 5.71 (d, J )
2.1, 1H, C)CH), 6.01 (d, J ) 2.1, 1H, C)CH), 6.77 (d, J ) 8.6,
1H, H-6), 7.23 (d, J ) 8.6, 1H, H-5). 13C NMR (CDCl3) 8.61 (Me-
C8), 22.56 (CH2CH2CH2), 30.40 (CH2-C4), 32.07 (CH2-C3),
72.02 (CH2O), 108.09 (C6), 113.26 (C8), 114.81, 118.00 (CH2dC),
122.47 (C5), 124.94, 126.52, 153.40, 156.40, 157.63, 160.58 (C2).
IR 2919, 1716, 1611, 1373, 1282, 1109, 803. MS m/z (%): 336
([M + 2]+, 12), 335 (M+, 12), 255 (76), 215 (29), 187 (100), 128
(15). Anal. (C16H15BrO3) C, H.
4′,5′-Cyclopentene-4,8-dimethylfuro[3,2-g]coumarin (32). Yield
30%; mp 153 °C, 1H NMR (CDCl3) 2.48 (s, 3H, Me-C8), 2.57 (s,
3H, Me-C4), 2.59 (m, 2H, CH2CH2CH2), 2.78 (m, 2H, CH2-C5′),
2.90 (m, 2H, CH2-C4′), 6.24 (s, 1H, H-3), 7.41 (s, 1H, H-5). 13C
NMR (CDCl3) 8.9 (Me-C8), 19.7 (Me-C4), 23.1 (CH2CH2CH2),
25.6 (CH2-C4′), 27.6 (CH2-C5′), 110.6 (C8), 111.3 (C3), 113.2 (C5),
116.2, 121.7, 123.3, 148.9, 153.8, 161.4, 162.1, 164.9 (C2). IR 2922,
2850, 1706, 1558, 1480, 1396, 1125, 1092. MS m/z 254 (M+, 80),
226 (M+ - CO, 100), 225 (98), 199 (18), 183 (15), 149 (56). Anal.
(C16H14O3) C, H.
4′,5′-Cyclohexene-4,8-dimethylfuro[3,2-g]coumarin (33). Yield
56%; mp 189-190 °C. 1H NMR (CDCl3) 1.85-2.00 (m, 4H,
CH2(CH2)2CH2), 2.50 (d, J ) 1.1, 3H, Me-C4), 2.58 (s, 3H, Me-
C8), 2.65 (m, 2H, CH2-C5′), 2.78 (CH2-C4′), 6.24 (d, J ) 1.1, 1H,
H-3), 7.41 (s, 1H, H-5). 13C NMR (CDCl3) 8.9 (Me-C8), 19.7 (Me-
C4), 20.8 (CH2-C5′), 22.8, 23.1, 23.9 (CH2-C4′), 109.0 (C8), 110.7
(C3), 113.0 (C5), 113.3, 116.2, 126.3, 150.1, 152.0, 153.7, 156.3,
162.0 (C2). IR 3070, 2923, 1716, 1639, 1403, 1183, 1100, 873.
MS m/z 269 ([M + 1]+, 11), 268 (M+, 100), 240 (29), 212 (19),
167 (13), 127 (10). Anal. (C17H16O3) C, H.
Biological Assay. Enzymatic MAO-A and MAO-B activity of
compounds was determined by a fluorimetric method following a
previously described protocol.41 Briefly, 0.1 mL of sodium
phosphate buffer (0.05 M, pH 7.4) containing various concentrations
of the test drugs (new compounds or reference inhibitors) and
appropriate amounts of recombinant hMAO-A or hMAO-B (Sigma-
Aldrich Quimica S.A., Alcobendas, Spain) and adjusted to obtain
in our experimental conditions the same reaction velocity in the
presence of both isoforms (i.e., to oxidize (in the control group)
the same concentration of substrate: 165 pmol of p-tyramine/min
(hMAO-A: 1.1 µg protein; specific activity: 150 nmol of p-tyramine
oxidized to p-hydroxyphenylacetaldehyde/min/mg protein; hMAO-
B: 7.5 µg protein; specific activity: 22 nmol of p-tyramine
transformed/min/mg protein) were incubated for 15 min at 37 °C
in a flat-black-bottom 96-well microtest plate (BD Biosciences,
Franklin Lakes, NJ) placed in the dark fluorimeter chamber. After
this incubation period, the reaction was started by adding (final
concentrations) 200 µM of 10-acetyl-3,7-dihydroxyphenoxazine
reagent (Amplex Red assay kit, Molecular Probes, Inc., Eugene,
OR), 1 U/mL horseradish peroxidase, and 1 mM p-tyramine. The
production of H2O2 and, consequently, of resorufin, was quantified
at 37 °C in a multidetection microplate fluorescence reader
(FLX800, Bio-Tek Instruments, Inc., Winooski, VT) based on the
fluorescence generated (excitation, 545 nm, emission, 590 nm) over
a 15 min period, during which the fluorescence increased linearly.
7-(ꢀ-Bromoallyloxy)-3,4-cyclohexene-8-methylcoumarin (25).
Yield 76%; mp 135-136 °C. 1H NMR (CDCl3) 1.82 (m, 4H,
CH2(CH2)2CH2), 2.35 (s, 3H, Me-C8), 2.56 (m, 2H, CH2-C4), 2.72
(m, 2H, CH2-C3), 4.71 (s, 2H, CH2O), 5.76 (d, J ) 1.6, 1H,
CdCH), 6.01 (d, J ) 1.6, 1H, CdCH), 6.75 (d, J ) 8.8, 1H, H-6),
7.35 (d, J ) 8.8, 1H, H-5). 13C NMR (CDCl3) 8.78 (Me-C8), 21.86
(CH2-CH2C3), 22.11 (CH2-CH2C4), 24.31 (CH2-C3), 25.65
(CH2-C4), 72.36 (CH2O), 108.32 (C6), 114.80 (C8), 115.04,
118.36 (CH2dC), 121.26, 121.35 (C5), 127.03, 147.64, 151.57,
157.47, 162.60 (C2). IR 3071, 2935, 1708, 1605, 1114, 755. MS
m/z 350 ([M + 2]+, 4), 349 (M+, 22), 269 (89), 229 (100), 201
(58), 187 (40). Anal. (C17H17BrO3) C, H.
General Procedure for the Preparation of 3,4-Benzocoumarins
26-28. To a solution of the acyclic ether 22, 23,45 or 25 (0.30
mmol) in toluene (15 mL) was added DDQ (0.60 mmol). The
solution was heated under reflux for 5 h. The mixture was cooled,
the precipitate filtered off, and the solvent evaporated under reduced
pressure. The resulting residue was purified by FC to give the
desired compound.
3,4-Benzo-7-acetonyloxy-8-methylcoumarin (26). Yield 64%; mp
172-174 °C. 1H NMR (CDCl3) 2.35 (s, 6H, Me-C8, MeCO), 4.60
(s, 2H, CH2O), 6.63 (d, J ) 8.8, 1H, H-6), 7.46 (m, 1H, CH-CHC3),
7.73 (m, 2H, H-5, CH-CHC4), 7.91 (d, J ) 8.0, 1H, CH C4), 8.28
(d, J ) 7.9, 1H, CH-C3). 13C NMR (CDCl3) 8.45 (Me-C8), 26.67
(MeCO), 73.19 (CH2O), 107.43 (C6), 111.95, 114.95, 119.72,
120.44 (CH-C4), 121.12 (CH-CHC3), 127.72 (C5), 130.19 (CH-
C3), 134.60 (CH-CHC4), 134.94, 150.15, 156.95, 161.13 (C2),
204.98 (MeCO). IR 2925, 1716, 1607, 1468, 1284, 1125, 766. MS
m/z 283 ([M + 1]+, 18), 282 (M+, 100), 239 (60), 225 (43), 181
(51), 152 (47). Anal. (C17H14O4) C, H.
3,4-Benzo-7-acetonyloxy-8-methoxycoumarin (27). Yield 56%;
1
mp 172-174 °C. H NMR (CDCl3) 2.33 (s, 3H, MeCO), 4.05 (s,
3H, MeO), 4.72 (s, 2H, CH2O), 6.80 (d, J ) 9.0, 1H, H-6), 7.53 (t,
J ) 8.1, 1H, CH-CHC3), 7.70 (d, J ) 9.0, 1H, H-5), 7.78 (t, J )
7.7, 1H, CH-CHC4), 7.99 (d, J ) 8.0, 1H, CH-C4), 8.36 (d, J )
7.9, 1H, CH-C3). 13C NMR (CDCl3) 27.01 (MeCO), 62.15 (MeO),
74.62 (CH2O), 110.94 (C6), 114.12, 118.00 (CH-C4), 120.50,
121.88 (C5), 128.76 (CH-CHC3), 131.04 (CH-C3), 135.29, 135.36
(CH-CHC4), 137.94, 146.26, 152.78, 161.07 (C2), 205.12 (MeCO).
IR 2935, 1730, 1608, 1475, 1304, 1121, 812. MS m/z 299 ([M +
1]+, 18), 298 (M+, 100), 255 (27), 241 (67), 197 (25), 170 (19).
Anal. (C17H14O5) C, H.
3,4-Benzo-7-(b-bromoallyloxy)-8-methylcoumarin (28). Yield
82%; mp 128 °C. 1H NMR (CDCl3) 2.38 (s, 3H, Me-C8), 4.72 (s,
2H, CH2O), 5.72 (d, J ) 1.9, 1H, CdCH), 6.03 (d, J ) 1.9, 1H,
CdCH), 6.80 (d, J ) 8.8, 1H, H-6), 7.49 (m, 1H, CH-CHC4), 7.78
(m, 1H + 1H, CH-CHC3 + H-5), 7.98 (d, J ) 8.0, 1H, CH-C3),
8.34 (d, J ) 8.0, 1H, CH-C4). 13C NMR (CDCl3) 8.95 (Me-C8),
72.33 (CH2O), 108.68 (C6), 112.36, 115.67, 118.39 (CdCH2),
120.25, 120.84 (CH-C4), 121.63 (CH-CHC3), 127.01, 128.18 (C5),
130.75 (CH-CHC4), 135.12 (CH-C3), 135.57, 150.66, 157.49,
161.78 (C2). IR 2921, 1726, 1608, 1470, 1283, 1115, 891, 766.
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 possible
capacity of the above test drugs to modify the fluorescence
generated in the reaction mixture due to nonenzymatic inhibition
was determined by adding these drugs to solutions containing only
the Amplex Red reagent in a sodium phosphate buffer.

6750 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21
Santana et al.
(16) Binda, C.; Hubalek, F.; Li, M.; Herzig, Y.; Sterling, J.; Edmondson,
D. E.; Mattevi, A. Crystal Structures of Monoamine Oxidase B in
Complex with Four Inhibitors of the N-Propargylaminoindane Class.
J. Med. Chem. 2004, 47, 1767–1774.
(17) Chimenti, F.; Maccioni, E.; Secci, D.; Bolasco, A.; Chimenti, P.;
Granese, A.; Befani, O.; Turini, P.; Alcaro, S.; Ortuso, F.; Cirilli, R.;
La Torre, F.; Cardia, M. C.; Distinto, S. 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. 2005, 48, 7113–7122.
(18) Harkcom, W. T.; Bevan, D. R. Molecular docking of inhibitors into
monoamine oxidase B. Biochem. Biophys. Res. Commun. 2007, 360,
401–406.
(19) Gallardo-Godoy, A.; Fierro, A.; McLean, T. H.; Castillo, M.; Cassels,
B. K.; Reyes-Parada, M.; Nichols, D. E. Sulfur-Substituted R-Alkyl
Phenethylamines as Selective and Reversible MAO-A Inhibitors:
Biological Activities, CoMFA Analysis, and Active Site Modeling.
J. Med. Chem. 2005, 48, 2407–2419.
To determine the kinetic parameters of hMAO-A and hMAO-B
(Km and Vmax), the corresponding enzymatic activity of both
isoforms was evaluated (under the experimental conditions de-
scribed above) in the presence of a number of p-tyramine
concentrations.
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 MAO isoforms, which were replaced by a sodium phosphate
buffer solution.
Acknowledgment. We are grateful to the Xunta de Galicia
(PGIDIT05PXIB20304PR, PGIDIT05BTF20302PR, and
INCITE07PXI203039ES) and Ministerio de Sanidad y Con-
sumo (FIS PI061537 and PI061457) for financial support.
(20) Mevedev, A. E.; Veselovsky, A. V.; Shvedov, V. I.; Tikhonova, T. A.;
Moskvitinia, T. A.; Fedotova, O. A.; Axenova, L. N.; Kamyshanskaya,
A. Z.; Kirkel, A. Z.; Ivanov, A. Inhibition of Monoamino Oxidase by
Pirlindole Analogues: 3D-QSAR and CoMFA Analysis. J. Chem. Inf.
Comput. Sci. 1998, 38, 1137–1144.
(21) Chimenti, F.; Bolasco, A.; Manna, F.; Secci, D.; Chimenti, P.; Granese,
A.; Befani, O.; Turini, P.; Cirilli, R.; La Torre, F.; Alcaro, S.; Ortuso,
F.; Langer, T. Synthesis, biological evaluation and 3D-QSAR of 1,3,5-
trisubstituted-4,5-dihydro-(1H)-pyrazole derivatives as potent and
highly selective monoamine oxidase A inhibitors. Curr. Med. Chem.
2006, 13, 1411–1428.
Supporting Information Available: Table 1SI including the
complete list of compounds used in training sets, an indication of
the compound number in the corresponding bibliographic reference,
the observed and predicted classification, and subsequent prob-
abilities of a positive result predicted for each compound. Table
2SI with the factor scores (F1 to F10) for each compound. Table
3SI with the conditions in terms of enzyme, species and organs of
all experimental results used as input for the QSAR model.
Bibliographic references for database compounds. Elemental analy-
sis data for synthesized compounds. Information from pharmaco-
logical studies. This material is available free of charge via the
Internet at http://pubs.acs.org.
(22) Gnerre, C.; Catto, M.; Francesco, L.; Weber, P.; Carrupt, P.-A.;
Altomare, C.; Carotti, A.; Testa, B. Inhibition of Monoamine Oxidase
by Functionalized Coumarin derivatives: Biological Activities, QSAR,
and 3D-QSARs. J. Med. Chem. 2000, 43, 4747–4758.
(23) Nunez, M. B.; Maguna, F. P.; Okulik, N. B.; Castro, E. A. QSAR
modeling of the MAO inhibitory activity of xanthone derivatives.
Bioorg. Med. Chem. Lett. 2004, 14, 5611–5617.
(24) Todeschini, R.; Consonni, V. Handbook of Molecular Descriptors;
Mannhold, R., Kubinyi, H., Timmerman, H., Eds.; Wiley-VCH:
Weinheim, 2000.
(25) Kier, L. B.; Hall, L. H. The Kappa indices for modeling molecular
shape and flexibility in Topological Indices and Related Descriptors
in QSAR and QSPR; Devillers, J., Balaban, A. T., Eds.; Gordon and
Breach Science Publishers: Amsterdam, 1999, pp 445-489.
(26) Estrada, E.; Uriarte, E.; Montero, A.; Teijeira, M.; Santana, L.; De
Clercq, E. A Novel Approach for the Virtual Screening and Rational
Design of Anticancer Compounds. J. Med. Chem. 2000, 43, 1975–
1985.
(27) Erdem, S. S.; Karahan, O.; Yildiz, I.; Yelekci, K. A computational
study on the amine-oxidation mechanism of monoamine oxidase:
insight into the polar nucleophilic mechanism. Chem. Org. Biomed.
Chem. 2006, 4, 646–658.
(28) Freund, J. A., Poschel, T. Eds. Stochastic Processes in Physics,
Chemistry, and Biology. In Lecture Notes in Physics; Springer-Verlag:
Berlin, Germany, 2000.
(29) Gonza´lez-D´ıaz, H.; Agu¨ero, G.; Cabrera, M. A.; Molina, R.; Santana,
L.; Uriarte, E.; Delogu, G.; Castan˜edo, N. Unified Markov thermo-
dynamics based on stochastic forms to classify drugs considering
molecular structure, partition system, and biologic species. Distribution
of the antimicrobial G1 on rat tissues. Bioorg. Med. Chem. Lett. 2005,
15, 551–557.
(30) Gonza´lez-D´ıaz, H.; Cruz-Monteagudo, M.; Molina, R.; Tenorio, E.;
Uriarte, E. Predicting multiple drugs side effects with a general drug-
target interaction thermodynamic Markov model. Bioorg. Med. Chem.
2005, 13, 1119–1129.
(31) Gonza´lez-D´ıaz, H.; Torres-Go´mez, L. A.; Guevara, Y.; Almeida, M. S.;
Molina, R.; Castan˜edo, N.; Santana, L.; Uriarte, E. Markovian
chemicals “in silico” design (MARCH-INSIDE), a promising approach
for computer-aided molecular design III: 2.5D indices for the discovery
of antibacterials. J. Mol. Model. 2005, 11, 116–123.
References
(1) Dostert, P.; Strolin Benedetti, M.; Jafre, M. Structural modifications
in oxazolidinone series leading to type A or B selective monoamine
oxidase inhibitors. In Monoamine Oxidase: Basic and Clinical
Frontiers; Kamijo, K., Usdin, E., Nagausu, T., Eds.; Excerpta Medica:
Amsterdam, 1982; pp 197-208.
(2) Singer, T. P. Monoamine oxidases. In Chemistry and Biochemistry of
FlaVoenzymes (II); Muller, F., Ed.; CRC Press: London, 1991; pp
437-470.
(3) De Colibus, L.; Li, M.; Binda, C.; Lustig, A.; Edmondson, D. E.;
Mattevi, A. Three-dimensional structure of human monoamine oxidase
A (MAO A): relation to the structures of rat MAO A and human MAO
B. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 12684–12689.
(4) Binda, C.; Li, M.; Huba´lek, F.; Restelli, N.; Edmondson, D. E.; Mattevi,
A. Insights into the mode of inhibition of human mitochondrial
monoamine oxidase B from high-resolution crystal structures. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 9750–9755.
(5) Yamada, M.; Yasuhara, H. Clinical pharmacology of MAO inhibitors:
safety and future. Neurotoxicology 2004, 25, 11–20.
(6) Rudorfer, M. V.; Potter, V. Z. Antidepressants. A comparative review
of the clinical pharmacology and therapeutic use of the “newer” versus
the “older” drugs. Drugs 1989, 37, 713–738.
(7) Palhagen, S.; Heinonen, E.; Hagglund, J.; Kaugesaar, T.; Maki-Ikola,
O.; Palm, R. Selegiline slows the progression of the symptoms of
Parkinson disease. Neurology 2006, 66, 1200–1206.
(8) Guay, D. R. Rasagiline (TVP-1012): a new selective monoamine
oxidase inhibitor for Parkinson’s disease. Am. J. Geriatr. Pharmaco-
ther. 2006, 4, 330–346.
(9) Riederer, P.; Danielczyk, W.; Grunblatt, E. Monoamine oxidase-B
inhibition in Alzheimer′s desease. Neurotoxicology 2004, 25, 271–
277.
(10) Youdim, M. B. H.; Fridkin, M.; Zheng, H. Novel bifunctional drugs
targeting monoamine oxidase inhibition and iron chelation as an
approach to neuroprotection in Parkinson’s disease and other neuro-
degenerative diseases. J. Neural Transm. 2004, 111, 1455–1471.
(11) Cesura, A. M.; Pletscher, A. The new generation of monoamine oxidase
inhibitors. Prog. Drug Res. 1992, 38, 171–297.
(12) Youdim, M. B. The advent of selective monoamine oxidase A inhibitor
antidepressants devoid of the cheese reaction. Acta Psychiatr. Scand.,
Suppl. 1995, 386, 5–7.
(13) Edmondson, D. E.; Mattevi, A.; Binda, C.; Li, M.; Hubalek, F.
Structure and mechanism of monoamine oxidase. Curr. Med. Chem.
2004, 11, 1983–1993.
(14) Tipton, K. F.; Boyce, S.; O’Sullivan, J.; Davey, G. P.; Healy, J.
Monoamine oxidases: Certainties and uncertainties. Curr. Med. Chem.
2004, 11, 1965–1982.
(15) Edmondson, D. E.; Binda, C.; Mattevi, A. Structural insights into the
mechanism of amine oxidation by monoamine oxidases A and B. Arch.
Biochem. Biophys. 2007, 464, 269–276.
(32) Santana, L.; Uriarte, E.; Gonza´lez-D´ıaz, H.; Zagotto, G.; Soto-Otero,
R.; Me´ndez-Alvarez, E. A QSAR Model for in Silico Screening of
MAO-A Inhibitors. Prediction, Synthesis, and Biological Assay of
Novel Coumarins. J. Med. Chem. 2006, 49, 1149–1156.
(33) Fierro, A.; Osorio-Olivares, M.; Cassels, K.; Edmondson, D. E.;
Sepulveda-Boza, S.; Reyes-Parada, M. Human and rat monoamine
oxidase-A are differentially inhibited by (S)-4-alkylthioamphetamine
derivatives: insights from molecular modeling studies. Bioorg. Med.
Chem. 2007, 15, 5198–5206.
(34) Novaroli, L.; Daina, A.; Favre, E.; Bravo, J.; Carotti, A.; Leonetti, F.;
Catto, M.; Carrupt, P. A.; Reist, M. Impact of Species-Dependent
Differences on Screening, Design, and Development of MAO-B
Inhibitors. J. Med. Chem. 2006, 49, 6264–6272.

QSAR and CN Approach to MAO A and B Inhibitors
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 21 6751
(35) Van Waterbeemd, H. Discriminant analysis for activity prediction. In
Chemometric Methods in Molecular Design. Method and Principles
in Medicinal Chemistry, Vol. 2; Wiley-VCH: Weinhiem, 1995; pp
265-282.
(44) Carotti, A.; Altomare, C.; Catto, M.; Gnerre, C.; Summo, L.; De Marco,
A.; Rose, S.; Jenner, P.; Testa, B. Lipophilicity Plays a Major Role in
Modulating the Inhibition of Monoamine Oxidase B by 7-Substituted
Coumarins. Chem. BiodiVersity 2006, 3, 134–149.
(45) Dalla Via, L.; Uriarte, E.; Quezada, E.; Dolmella, A. M. G.; Ferlin,
M. G.; Gia, O. Novel Pyrone Side Tetracyclic Psoralen Derivatives:
Synthesis and Photobiological Activity. J. Med. Chem. 2003, 46, 3800–
3810.
(46) Gia, O.; Marciani-Magno, S.; Gonza´lez-D´ıaz, H.; Quezada, E.; Santana,
L.; Uriarte, E.; Dalla Via, L. Design, Synthesis and Photobiological
Properties of 3,4-Cyclopentene Psoralens. Bioorg. Med. Chem. 2005,
13, 809–817.
(47) Dalla Via, L.; Uriarte, E.; Santana, L.; Marciani-Magno, S.; Gia, O.
Methyl derivatives of tetracyclic psoralen analogues: antiproliferative
activity and interaction with DNA. ARKIVOC 2004, 5, 131–146.
(48) Ya´n˜ez, M.; Fraiz, N.; Cano, E.; Orallo, F. Inhibitory effects of cis-
and trans-resveratrol on noradrenaline and 5-hydroxytryptamine uptake
and on monoamine oxidase activity. Biochem. Biophys. Res. Commun.
2006, 344, 688–695.
(49) Gonza´lez-D´ıaz, H.; Molina-Ruiz, R.; Hernandez, I. MARCH-INSIDE
(MARkoV CHains INVariants for SImulation & DEsign), version 3.0.
Windows supported version released under request to the main author.
Contact E-mail: humberto.gonzalez@usc.es.
(50) Statistica Software, version 8.0; StatSoft Iberica Portugal Lda.: Lisboa,
Portugal, 2008.
(51) Junker, B. H.; Koschu¨tzki, D.; Schreiber, F. Exploration of biological
network centralities with CentiBiN. BMC Bioinform. 2006, 7, 219.
(52) Batagelj, V.; Mrvar, A. Pajek: Program for Large Network Analysis.
Home page: http://vlado.fmf.uni-lj.si/pub/networks/pajek/.
(36) Kleeman, A.; Engel, J.; Kutscher, B.; Reichert, D. Pharmaceutical
Substances: Syntheses, Patents and Applications, 4th Ed.; George
Thieme Verlag: Stuttgart, 2001.
(37) Negwer, M. Organic Chemical Drugs and their Synonyms: An
International SurVey; Akademie-Verlag: Berlin, 1987.
(38) Iding, H.; Jolidon, S.; Krummenacher, D.; Rodriguez Sarmiento, R. M.;
Thomas, A. W.; Wirz, B.; Wostl, W.; Wyler, R. (F. Hoffmann-La
Roche A.-G., Switzerland). Preparation of arylpyrrolidones as monoam-
ine oxidase-B (MAO-B) inhibitors. Patent WO 2004026825 A1, 2004.
(39) Jolidon, S.; Rodriguez-Sarmiento, R. M.; Thomas, A. W.; Wostl, W.;
Wyler, R. (F. Hoffmann-La Roche A.-G., Switzerland). Preparation
of 4-pyrrolidinophenyl benzyl ether derivatives as monoamine oxidase
B inhibitors. Patent WO 2004026826 A1, 2004.
(40) Borges, F.; Roleira, F.; Mihazes, N.; Santana, L.; Uriarte, E. Simple
coumarins and analogues in medicinal chemistry: occurrence, synthesis
and biological activity. Curr. Med. Chem, 2005, 12, 887–916.
(41) Chimenti, F.; Secci, D.; Bolasco, A.; Chimenti, P.; Granese, A.; Befani,
O.; Turini, P.; Alcaro, S.; Ortuso, F. Inhibition of monoamine oxidases
by coumarin-3-acyl derivatives: biological activity and computational
study. Bioorg. Med. Chem. Lett. 2004, 14, 3697–3703.
(42) Catto, M.; Nicolotti, O.; Leonetti, F.; Carotti, A.; Soto-Otero, R.;
Mendez-Alvarez, E.; Carotti, A. Structural Insights into Monoamine
Oxidase Inhibitory Potency and Selectivity of 7-Substituted Coumarins
from Ligand- and Target-Based Approaches. J. Med. Chem. 2006,
49, 4912–4925.
(43) Binda, C.; Wang, J.; Pisani, L.; Caccia, C.; Carotti, A.; Salvati, P.;
Edmondson, D. E.; Mattevi, A. Structures of Human Monoamine
Oxidase B Complexes with Selective Noncovalent Inhibitors: Safi-
namide and Coumarin Analogs. J. Med. Chem. 2007, 50, 5848–5852.
JM800656V